de Oliveira et al BMC Microbiology 2013 13107httpwwwbiomedcentralcom1471-218013107

RESEARCH ARTICLE Open Access

Intermolecular interactions of the malatesynthase of Paracoccidioides sppKarine Martins de Oliveira1 Benedito Rodrigues da Silva Neto1 Juliana Alves Parente1 Roosevelt Alves da Silva2Guilherme Oliveira Quintino2 Aline Raquel Voltan3 Maria Joseacute Soares Mendes-Giannini3Ceacutelia Maria de Almeida Soares1 and Maristela Pereira1

Abstract

Background The fungus Paracoccidioides spp is the agent of paracoccidioidomycosis (PCM) a pulmonary mycosisacquired by the inhalation of fungal propagules Paracoccidioides malate synthase (PbMLS) is important in theinfectious process of Paracoccidioides spp because the transcript is up-regulated during the transition frommycelium to yeast and in yeast cells during phagocytosis by murine macrophages In addition PbMLS acts as anadhesin in Paracoccidioides spp The evidence for the multifunctionality of PbMLS indicates that it could interactwith other proteins from the fungus and host The objective of this study was to identify and analyze proteins thatpossibly bind to PbMLS (PbMLS-interacting proteins) because protein interactions are intrinsic to cell processes andit might be possible to infer the function of a protein through the identification of its ligands

Results The search for interactions was performed using an in vivo assay with a two-hybrid library constructed inS cerevisiae the transcripts were sequenced and identified In addition an in vitro assay using pull-down GSTmethodology with different protein extracts (yeast mycelium yeast-secreted proteins and macrophage) was performedand the resulting interactions were identified by mass spectrometry (MS) Some of the protein interactions wereconfirmed by Far-Western blotting using specific antibodies and the interaction of PbMLS with macrophages wasvalidated by indirect immunofluorescence and confocal microscopy In silico analysis using molecular modelingdynamics and docking identified the amino acids that were involved in the interactions between PbMLS andPbMLS-interacting proteins Finally the interactions were visualized graphically using Osprey software

Conclusion These observations indicate that PbMLS interacts with proteins that are in different functional categoriessuch as cellular transport protein biosynthesis modification and degradation of proteins and signal transduction Thesedata suggest that PbMLS could play different roles in the fungal cell

Keywords Paracoccidioides spp Malate synthase Protein-protein interactions

BackgroundIn vivo the Paracoccidioides spp transition from myceliumto yeast cells is governed by an increase in temperature thatoccurs upon contact of the mycelia or conidia with thehost The fungus a complex of several phylogenetic speciescauses paracoccidioidomycosis (PCM) a human systemicmycosis The infection begins with the inhalation of fungalpropagules which reach the epithelium of the alveoliwhere the mycelium differentiates to the yeast pathogenic

Correspondence maristelaufggmailcom1Laboratoacuterio de Biologia Molecular Instituto de Ciecircncias BioloacutegicasUniversidade Federal de Goiaacutes Goiacircnia GO BrazilFull list of author information is available at the end of the article

copy 2013 de Oliveira et al licensee BioMed CenCreative Commons Attribution License (httpdistribution and reproduction in any medium

form [1] Although most clinical forms of the disease areasymptomatic severe and progressive infections involvingpulmonary and extra-pulmonary tissues occur [2] A highpercentage (80) of cases of the disease is reported inBrazil where PCM is the leading cause of death among thesystemic mycoses PCM is the eighth-leading cause of mor-tality among infectious and parasitic diseases which estab-lishes it as a serious public health problem [3-5]Paracoccidioides malate synthase (PbMLS) appears to be

important to the infectious process of Paracoccidioidesspp because the transcript is up-regulated during the tran-sition from mycelium to yeast during the infectious phase[6] and in yeast cells during phagocytosis by murine

tral Ltd This is an Open Access article distributed under the terms of thecreativecommonsorglicensesby20) which permits unrestricted use provided the original work is properly cited

de Oliveira et al BMC Microbiology 2013 13107 Page 2 of 14httpwwwbiomedcentralcom1471-218013107

macrophages [7] PbMLS participates in the glyoxylatepathway which enables the fungus to assimilate two-carbon compounds and in the allantoin degradation path-way of the purine metabolism which allows the fungus touse nitrogen compounds [8] In addition to being a crucialenzyme in the metabolism of Paracoccidioides spp PbMLSis located in peroxisomes and in the cell wall of the fun-gus It is capable of binding to extracellular matrix compo-nents such as fibronectin and collagen types I and IV andis also secreted by the fungus Furthermore it has beendemonstrated that this enzyme plays a role as an adhesinhaving the ability to mediate host cell adhesion and intern-alization of Paracoccidioides spp in a significant role in theestablishment of infection [9] Therefore there is evidenceof PbMLS functionality which drives the investigation ofthese functions through studies of protein interactionsThe availability of all of the sequences of the Para-

coccidioides spp genome and the appearance of varioustechniques for the screening of protein-protein interac-tions makes it possible to discover the functions of fungalproteins of interest from the identification of their ligands[10] Therefore this study was performed to identifyParacoccidioides spp proteins that might interact withPbMLS through techniques such as the yeast two-hybridsystem (which is the most suitable method for identifyingbinary interactions) and affinity purifications coupled withmass spectrometry (MS) analyses (pull-down) to discovermulti-protein assemblies that enable us to infer otherfunctions of this enzyme and corroborate evidence of theirmultiple locations in the fungal cell The interactions werealso evaluated by in silico analysis

ResultsTracking of protein interactions in vitro by pull-downassaysThe pull-down technique detects the physical interactionsbetween proteins most directly as a result it is a usefultool in the confirmation of protein-protein interactionspredicted by other techniques [11] Here pull-down assayswere performed to search for interactions between PbMLSand other proteins of Paracoccidioides Pb01 from differentextracts because the fungus expresses different proteinsdepending on the phase [12] which could lead to differentPbMLS-interacting proteinsThe recombinant proteins GST and PbMLS fused to

GST (PbMLS-GST) were expressed purified by using an af-finity resin and visualized by SDS-PAGE (Additional file 1Figure S1A lanes 1 and 2 respectively) The predicted massfor the hybrid protein PbMLS-GST was 864 kDa (609 kDafor PbMLS and 255 kDa for GST) The proteins designatedas 1 2 3 and 4 were subjected to proteolysis and identifica-tion by MS The proteomic identification data are compiledin Additional file 2 Table S1 The results indicated that

proteins 1 and 2 correspond to PbMLS (both arePAAG_04542) but protein 2 is most likely a result of itsproteolysis or incomplete translation Protein 3 was identi-fied as membrane protein F of E coli The co-purification ofproteins from E coli has been described [13] Protein 4 cor-responds to GSTAfter purification the GST bound to resin was incubated

with protein extracts from Paracoccidioides Pb01 mycelium(Additional file 1 Figure S1B) yeast (Additional file 1Figure S1C) yeast-secreted (Additional file 1 Figure S1D)and macrophage (Additional file 1 Figure S1E) to excludenonspecific bindings that occur only in the presence ofGST The presence of only GST in lane 1 (Additional file 1Figures S1B S1C S1D and S1E) indicated the absence ofnon-specific bindings to GST Next the supernatant wasremoved and incubated with PbMLS-GST bound to resinThe protein complexes formed during incubation were pre-cipitated and resolved by SDS-PAGE (lane 2 ndash Additionalfile 1 Figures S1B S1C S1D and S1E)Proteins that interacted with PbMLS which are listed

from 5 to 66 (Additional file 1 Figure S1B S1C S1D andS1E) were removed from the gel and identified by MS(Additional file 2 Table S1) Proteins that interact withPbMLS and that were detected by different pull-down as-says were listed (Additional file 3 Table S2) The searchagainst the NCBI non-redundant database using the MSMS data was performed using MASCOT software v 24[14] Functional characterization was performed usingUniProt databases [15] and MIPS [16]A total of 45 PbMLS-interacting proteins were identified

(Additional file 3 Table S2) Of these 18 proteins werefrom macrophage and 27 were from ParacoccidioidesPb01 15 were from mycelium 18 were from yeast and 11were yeast-secreted Some proteins were found in morethan one extract (4 proteins in mycelium yeast and yeast-secreted 11 proteins in mycelium and yeast 1 protein inmycelium and yeast-secreted) No protein was foundin both yeast and yeast-secreted extracts Of the 27Paracoccidioides Pb01 proteins 13 were exclusively ex-tract (found only in mycelium yeast or yeast-secreted) Of18 macrophage proteins 13 were exclusive to macro-phage with 5 related to cytoskeleton A total of 3 proteins(heat shock protein 60 kDa heat shock protein 70 kDaand fructose 1 6 bisphosphate aldolase) were also identi-fied in the pull-down assays with Paracoccidioides Pb01mycelium andor yeast cells

Tracking of protein interactions in vivo by a two-hybridassayTo detect new interactions between PbMLS and otherParacoccidioides Pb01 proteins two-hybrid assays wereperformed The Y187 strain of S cerevisiae that harborsthe bait (PbMLS) fused to the binding domain (BD) ofthe GAL4 transcription factor and the strain AH109 that

de Oliveira et al BMC Microbiology 2013 13107 Page 3 of 14httpwwwbiomedcentralcom1471-218013107

harbors the prey (cDNA library of ParacoccidioidesPb01) fused to the activation domain (AD) of GAL4were placed in the same system to promote diploidsThe diploid yeast-expressing proteins that interacted

were finally selected in medium that contained a chromo-genic substrate (X-α-GAL) to observe the transcriptionalactivation of the reporter gene mel1 a GAL4-regulatedgene coding for the α-galactosidase enzyme A total of 24clones showed the activation of the reporter gene mel1 byturning blue (data not shown) which confirmed that therewas interaction between PbMLS and the gene productslisted in the Additional file 4 Table S3To identify gene products that interacted with PbMLS

the cDNAs of the clones were sequenced after PCR ampli-fication ESTs (Expressed Sequence Tags) were processedusing the bioinformatics tool Blast2GO The functionalclassification was based on the homology of each ESTagainst the GenBank database using the BLAST algorithm[17] with a significant homology cutoff of le 1e-5 and func-tional annotation by MIPS [16] Additionally sequenceswere grouped into functional categories through the PED-ANT 3 database [18] The analysis indicated the presenceof several functional categories of genes and cell functionsrelated to cellular transport protein fate protein synthesisnucleotide metabolism signal transduction cell cycle and

Figure 1 Map of interactions between MLS and other proteins generainteractions obtained by a two-hybrid assay Protein interactions obtained(B) yeast (C) and yeast secretions (D) The blue lines indicate protein interaprotein interactions with MLS that are already described in The GRID interaThe colored dots show the functional classifications of the proteins

DNA processing and hypothetical protein (Additional file 4Table S3)

Construction of protein interaction mapsA comprehensive genetic interaction dataset has been de-scribed for the model yeast S cerevisiae [19] Becausegenes that act in the same pathway display similar patternsof genetic interactions with other genes [19-22] we inves-tigated whether Paracoccidioides Pb01 protein sequencesthat interacted with PbMLS and were tracked by the pull-down and two-hybrid assays (Additional file 3 Table S2and Additional file 4 Table S3 respectively) were found inthe structural genome database of S cerevisiae [23] Thosesequences and others from The GRID protein interactiondatabase [24] of S cerevisiae were used to construct pro-tein interaction maps generated by the Osprey NetworkVisualization System [25] (Figure 1) Protein sequencesfrom macrophage were not used because some of themwere not found in the S cerevisiae database The bluelines indicate protein interactions with MLS fromParacoccidioides Pb01 experimental data The green linesindicate protein interactions with MLS already describedin The GRID interaction database [24] of S cerevisiae Apink line corresponds to both The colored dots show thefunctional classification of proteins

ted by the Osprey Network Visualization System [25] (A) Proteinby pull-down assays with protein extracts of Paracoccidioides myceliumctions with MLS from the experimental data The green lines indicatection database [24] of S cerevisiae The pink line corresponds to both

de Oliveira et al BMC Microbiology 2013 13107 Page 4 of 14httpwwwbiomedcentralcom1471-218013107

Protein interactions obtained by a two-hybrid assay areshown in Figure 1A Protein interactions obtained by pull-down assays with protein extracts of ParacoccidioidesPb01 mycelium yeast and yeast-secretions are shown inFigure 1B C and D respectively Ubiquitin (YLL039C)was the only protein that interacted with MLS that wasfound in both Paracoccidioides and S cerevisiae The otherproteins were identified in Paracoccidioides Pb01 or Scerevisiae but not in both Although some proteins identi-fied in Paracoccidioides Pb01 have homologous proteinsin S cerevisiae (Additional file 5 Table S4) these proteinscould not yet be identified as interacting with PbMLSMost of the Paracoccidioides Pb01 proteins that interactedwith PbMLS were related to the metabolism category

Confirmation of the interactions by Far-Western blotassaysFar-Western blot assays were conducted to confirm the in-teractions between PbMLS and other proteins from the

Figure 2 Confirmation of the interactions by Far-Western blot assaysMembranes were reacted with Paracoccidioides protein extracts of myceliumsubsequently incubated with anti-rabbit IgG anti-enolase anti-triosephosphwith anti-rabbit IgG conjugated to alkaline phosphatase Negative control wanti-triosephosphate isomerase and anti-actin respectively without preincuwas obtained by incubating the PbMLS with the polyclonal anti-PbMLS antsecretions and macrophages (lanes 1 2 3 and 4 respectively) were subjectwere incubated with PbMLS and subsequently primary antibody anti-PbMobtained by incubating each protein extract with anti-PbMLS antibody witindicate the proteins (Additional file 2 Table S1) that interact with PbMLS t

fungus identified by pull-down assays PbMLS wassubjected to SDS-PAGE and was electro blotted Themembranes were reacted with protein extracts of Para-coccidioides Pb01 mycelium yeast and macrophage(Figure 2A lanes 1 2 and 3 respectively) and were subse-quently incubated with rabbit IgG anti-enolase anti-triosephosphate isomerase and anti-actin respectively Thereactions were revealed with anti-rabbit IgG conjugated toalkaline phosphatase Positive signals to the three extractsindicated the presence of an interaction between PbMLSand enolase triosephosphate isomerase and actin Negativecontrol was obtained by incubating PbMLS with the anti-bodies anti-enolase anti-triosephosphate isomerase andanti-actin respectively without preincubation with the pro-tein extracts (Figure 2A lanes 4 5 and 6 respectively) Posi-tive control was obtained by incubating the PbMLS withthe polyclonal anti-PbMLS antibody (Figure 2A lane 7)Another Far-Western blot assay was performed

using membranes that contained protein extracts of

(A) PbMLS was subjected to SDS-PAGE and electro blotted(lane 1) yeast (lane 2) and macrophage (lane 3) and were

ate isomerase and anti-actin respectively The reactions were revealedas obtained by incubating PbMLS with the antibodies anti-enolasebation with the protein extracts (lanes 4 5 and 6) The positive controlibody (lane 7) (B) Protein extracts of Paracoccidioides mycelium yeasted to SDS-PAGE and blotted onto nylon membrane The membranesLS and secondary antibody anti-rabbit IgG Negative control washout preincubation with PbMLS (lanes 5 6 7 and 8) The numbershat are confirmed by this technique

Figure 3 Binding of PbMLS to the macrophage surfaceImmunofluorescence microscopy that shows the binding of PbMLSto J774 A1 mouse macrophage cells (A) Negative control wasperformed with the unrelated protein BSA (B) Arrows indicatePbMLS (green) binding to the macrophage cell surfaces blueindicates the macrophage nucleus

Figure 4 Interaction between Paracoccidioides yeast cells andpneumocytes by confocal laser scanning microscopy Infectedcell monolayers were fixed and permeabilized Primary anti-PbMLSand secondary antibodies Alexa Fluor 594 goat anti-rabbit IgG (red)were used The specimens were analyzed by laser confocalmicroscopy using DIC (A) and fluorescence (B)

de Oliveira et al BMC Microbiology 2013 13107 Page 5 of 14httpwwwbiomedcentralcom1471-218013107

Paracoccidioides Pb01 mycelium yeast yeast secretionsand macrophage (Figure 2B lanes 1 2 3 and 4 respec-tively) The membranes were incubated with PbMLS andsubsequently were incubated with antibody anti-PbMLSand secondary antibody anti-rabbit IgG Several proteinsidentified in the pull-down assays interacted with PbMLSat this point which suggested the veracity of the interac-tions Negative control was obtained by incubating eachprotein extract with the anti-PbMLS antibody withoutpreincubation with PbMLS (Figure 2B lanes 5 6 7 and 8)The numbers identify the proteins that interacted withPbMLS as shown in Additional file 2 Table S1

PbMLS binds to the surface of macrophagesBecause the results from Far-Western blot assays revealedseveral macrophage proteins interacting with PbMLS weperformed immunofluorescence microscopy to visualizewhether PbMLS could adhere to the surface of the macro-phage cells No binding was observed using BSA as a con-trol (Figure 3A) The arrow indicates PbMLS binding to amacrophage surface (Figure 3B)

PbMLS participates in the adherence of Paracoccidioidesto pneumocyte cellsBecause the fungus initially reaches the lungs the partici-pation of PbMLS in the adherence of ParacoccidioidesPb18 to pneumocyte cells was investigated by using con-focal laser scanning microscopy A549 cells were pretreatedwith anti-PbMLS and infected with Paracoccidioides Pb18isolate After washings with frozen PBS-T the monolayerswere incubated with Alexa Fluor that was 594-conjugatedfor labeling the antibody The arrows indicate PbMLSinteracting with the A549 surface (Figures 4A and B)

Homology modelsIn silico analysis was performed to investigate how the in-teractions identified by pull-down and two-hybrid assayscould occur Some PbMLS-interacting proteins from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle were selected for analysis Because PbMLS partici-pates in the glyoxylate cycle interaction between proteinsfrom different metabolic pathways would be expected Be-cause no crystal structure of PbMLS-interacting proteinsdescribed here was reported a three-dimensional homologymodel for each protein was constructed based on the struc-ture template listed in Additional file 6 Table S5 All of the3D-structure templates used to build models of the proteinshave a resolution of lt 20 Aring and an identity of gt 49 with acoverage of gt 91Homology models of the PbMLS-interacting proteins

have very little conformational change when comparedto their templates (Additional file 6 Table S5) The lar-gest deviations were observed for enolase and fructose

de Oliveira et al BMC Microbiology 2013 13107 Page 6 of 14httpwwwbiomedcentralcom1471-218013107

16 bisphosphate aldolase with 265 Aring and 144 Aring ofroot mean square derivation (RMSD) when superposedon the template when considering the non-hydrogenatoms For enolase there is a significant conformationalchange only in the C-terminal regions and betweenPRO143 and ASN155 (data not shown)Alpha-helix-like secondary-structure patterns were ob-

served in a greater proportion in the homology modelsPbMLS-interacting proteins For almost all of the struc-tures the alpha-helix-like pattern corresponded to morethan 40 of the whole structure while the beta-sheet-like pattern accounted for less than 20 except for theprotein ubiquitin whose quantity of beta-sheet-like pat-tern was greater (Additional file 6 Table S5)Ramachandran plots of homology models were assessed

stereo-chemically through the RAMPAGE web server [26](data not shown) For all of the proteins the Φ and Ψ dis-tributions of the Ramachandran plots were always above94 in the favored regions and less than 35 in theallowed regions The quality factors of the structures wereestimated by the ERRAT web server and are summarizedin Additional file 6 Table S5

Molecular dynamicsAll of the proteins were subjected to at least 20 ns simula-tion using GROMACS software [27] For the proteinsgamma actin 2-methylcitrate synthase triosephosphateisomerase and ubiquitin that time was insufficient toachieve RMSD stability of non-hydrogen atoms with re-spect to the structure homology models In those casesmore simulation time was provided until this conditionwas achieved The times required are listed for each pro-tein For almost all of the proteins the deviations fromtheir homology models were low (approximately 30 Aring)Specifically ubiquitin and 2-methylcitrate synthase hadthe highest RMSDs The increase was 765 Aring and 634 Aringafter 60 ns and 40 ns respectively When only the residuesfrom the interfaces of the complexes were consideredthe RMSDs increased 90 Aring and 587 Aring respectively(Additional file 6 Table S5)The alpha-helix-like pattern was slightly reduced in all

of the proteins that were binding to PbMLS but the beta-sheet-like structures almost did not change Although theRMSDs were high for ubiquitin and 2-methylcitrate syn-thase the alpha-helix-like patterns decreased to only106 and 69 respectively

Molecular docking and molecular dynamics of theprotein-protein complexesMolecular docking between PbMLS and PbMLS-in-teracting proteins was investigated by the GRAMM-Xweb server using the structures stabilized by DM Onlythe best model-structures provided by the server were se-lected These complexes were then subjected to a rapid

DM so that their structures could accommodate and avoidhigh energy at the interface between them thus identifyingresidues in this region Significant conformational changesoccurred in ubiquitin and 2-methylcitrate synthase whenthey were complexed with PbMLS (data not shown) Theresidues contacting at the interface of the complexes areshown in Additional file 7 Table S6 and these amino acidsare highlighted in Figure 5 Some amino acid residues arecommon to different proteins For example ASP379 andGLN380 are residues of PbMLS that interact with enolaseand ubiquitin ASN386 is at the interface for gamma actinand ubiquitin LEU388 is common to triosephosphateisomerase and glyceraldehyde-3-phosphate dehydrogenaseand ASP401 is common to 2-methylcitrate synthase andmalate dehydrogenaseThe protein-protein complexes relaxed by DM were pro-

vided to the Fiberdock web server which determined theglobal energy for each complex (Additional file 7 TableS6) The results showed that fructose 1 6 bisphosphate al-dolase and ubiquitin were well stabilized when complexedwith PbMLS The ASP265 residue of PbMLS is present inthe interaction of both proteins

DiscussionOur previous studies showed that PbMLS is required inthe metabolism of Paracoccidioides Pb01 acting in theglyoxylate cycle and in the allantoin degradation pathwayPbMLS condenses acetyl-CoA from both 2C sources(glyoxylate cycle) and nitrogen sources (from proline andpurine metabolism) to produce malate which is a centralmolecule of the tricarboxylic acid cycle or glyoxylate cycle[8] In addition PbMLS is located in the cytoplasm and onthe fungal cell surface and is secreted behaving like an an-chorless adhesin [9] The strong evidence for PbMLSmultifunctionality increased our interest in researchingthe possibility of new roles for PbMLS through studies ofprotein-protein interactions which aimed to identifyPbMLS-interacting proteinsWe searched for PbMLS-interacting proteins using

Far-Western blot pull-down and two-hybrid techniquesThe two-hybrid and pull-down are used as complemen-tary techniques because the results depend on variants ofthe methods The two-hybrid system is highly sensitive todetecting low-abundance proteins unlike the pull-downsystem which detects high-abundance molecules Add-itionally the two-hybrid system allows identifying strongand weak interactions while the pull-down is not a sensi-tive method for identifying some of the weak interactionsbecause of the wash steps [28] Because the principles ofthe techniques are different we have the capability ofidentifying different proteinsPull-down assays were performed using Paracoccidioides

Pb01 mycelium yeast and yeast-secreted protein extractsbecause protein differences [12] and metabolic differences

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 3 of 14httpwwwbiomedcentralcom1471-218013107

harbors the prey (cDNA library of ParacoccidioidesPb01) fused to the activation domain (AD) of GAL4were placed in the same system to promote diploidsThe diploid yeast-expressing proteins that interacted

were finally selected in medium that contained a chromo-genic substrate (X-α-GAL) to observe the transcriptionalactivation of the reporter gene mel1 a GAL4-regulatedgene coding for the α-galactosidase enzyme A total of 24clones showed the activation of the reporter gene mel1 byturning blue (data not shown) which confirmed that therewas interaction between PbMLS and the gene productslisted in the Additional file 4 Table S3To identify gene products that interacted with PbMLS

the cDNAs of the clones were sequenced after PCR ampli-fication ESTs (Expressed Sequence Tags) were processedusing the bioinformatics tool Blast2GO The functionalclassification was based on the homology of each ESTagainst the GenBank database using the BLAST algorithm[17] with a significant homology cutoff of le 1e-5 and func-tional annotation by MIPS [16] Additionally sequenceswere grouped into functional categories through the PED-ANT 3 database [18] The analysis indicated the presenceof several functional categories of genes and cell functionsrelated to cellular transport protein fate protein synthesisnucleotide metabolism signal transduction cell cycle and

Figure 1 Map of interactions between MLS and other proteins generainteractions obtained by a two-hybrid assay Protein interactions obtained(B) yeast (C) and yeast secretions (D) The blue lines indicate protein interaprotein interactions with MLS that are already described in The GRID interaThe colored dots show the functional classifications of the proteins

DNA processing and hypothetical protein (Additional file 4Table S3)

Construction of protein interaction mapsA comprehensive genetic interaction dataset has been de-scribed for the model yeast S cerevisiae [19] Becausegenes that act in the same pathway display similar patternsof genetic interactions with other genes [19-22] we inves-tigated whether Paracoccidioides Pb01 protein sequencesthat interacted with PbMLS and were tracked by the pull-down and two-hybrid assays (Additional file 3 Table S2and Additional file 4 Table S3 respectively) were found inthe structural genome database of S cerevisiae [23] Thosesequences and others from The GRID protein interactiondatabase [24] of S cerevisiae were used to construct pro-tein interaction maps generated by the Osprey NetworkVisualization System [25] (Figure 1) Protein sequencesfrom macrophage were not used because some of themwere not found in the S cerevisiae database The bluelines indicate protein interactions with MLS fromParacoccidioides Pb01 experimental data The green linesindicate protein interactions with MLS already describedin The GRID interaction database [24] of S cerevisiae Apink line corresponds to both The colored dots show thefunctional classification of proteins

ted by the Osprey Network Visualization System [25] (A) Proteinby pull-down assays with protein extracts of Paracoccidioides myceliumctions with MLS from the experimental data The green lines indicatection database [24] of S cerevisiae The pink line corresponds to both

de Oliveira et al BMC Microbiology 2013 13107 Page 4 of 14httpwwwbiomedcentralcom1471-218013107

Protein interactions obtained by a two-hybrid assay areshown in Figure 1A Protein interactions obtained by pull-down assays with protein extracts of ParacoccidioidesPb01 mycelium yeast and yeast-secretions are shown inFigure 1B C and D respectively Ubiquitin (YLL039C)was the only protein that interacted with MLS that wasfound in both Paracoccidioides and S cerevisiae The otherproteins were identified in Paracoccidioides Pb01 or Scerevisiae but not in both Although some proteins identi-fied in Paracoccidioides Pb01 have homologous proteinsin S cerevisiae (Additional file 5 Table S4) these proteinscould not yet be identified as interacting with PbMLSMost of the Paracoccidioides Pb01 proteins that interactedwith PbMLS were related to the metabolism category

Confirmation of the interactions by Far-Western blotassaysFar-Western blot assays were conducted to confirm the in-teractions between PbMLS and other proteins from the

Figure 2 Confirmation of the interactions by Far-Western blot assaysMembranes were reacted with Paracoccidioides protein extracts of myceliumsubsequently incubated with anti-rabbit IgG anti-enolase anti-triosephosphwith anti-rabbit IgG conjugated to alkaline phosphatase Negative control wanti-triosephosphate isomerase and anti-actin respectively without preincuwas obtained by incubating the PbMLS with the polyclonal anti-PbMLS antsecretions and macrophages (lanes 1 2 3 and 4 respectively) were subjectwere incubated with PbMLS and subsequently primary antibody anti-PbMobtained by incubating each protein extract with anti-PbMLS antibody witindicate the proteins (Additional file 2 Table S1) that interact with PbMLS t

fungus identified by pull-down assays PbMLS wassubjected to SDS-PAGE and was electro blotted Themembranes were reacted with protein extracts of Para-coccidioides Pb01 mycelium yeast and macrophage(Figure 2A lanes 1 2 and 3 respectively) and were subse-quently incubated with rabbit IgG anti-enolase anti-triosephosphate isomerase and anti-actin respectively Thereactions were revealed with anti-rabbit IgG conjugated toalkaline phosphatase Positive signals to the three extractsindicated the presence of an interaction between PbMLSand enolase triosephosphate isomerase and actin Negativecontrol was obtained by incubating PbMLS with the anti-bodies anti-enolase anti-triosephosphate isomerase andanti-actin respectively without preincubation with the pro-tein extracts (Figure 2A lanes 4 5 and 6 respectively) Posi-tive control was obtained by incubating the PbMLS withthe polyclonal anti-PbMLS antibody (Figure 2A lane 7)Another Far-Western blot assay was performed

using membranes that contained protein extracts of

(A) PbMLS was subjected to SDS-PAGE and electro blotted(lane 1) yeast (lane 2) and macrophage (lane 3) and were

ate isomerase and anti-actin respectively The reactions were revealedas obtained by incubating PbMLS with the antibodies anti-enolasebation with the protein extracts (lanes 4 5 and 6) The positive controlibody (lane 7) (B) Protein extracts of Paracoccidioides mycelium yeasted to SDS-PAGE and blotted onto nylon membrane The membranesLS and secondary antibody anti-rabbit IgG Negative control washout preincubation with PbMLS (lanes 5 6 7 and 8) The numbershat are confirmed by this technique

Figure 3 Binding of PbMLS to the macrophage surfaceImmunofluorescence microscopy that shows the binding of PbMLSto J774 A1 mouse macrophage cells (A) Negative control wasperformed with the unrelated protein BSA (B) Arrows indicatePbMLS (green) binding to the macrophage cell surfaces blueindicates the macrophage nucleus

Figure 4 Interaction between Paracoccidioides yeast cells andpneumocytes by confocal laser scanning microscopy Infectedcell monolayers were fixed and permeabilized Primary anti-PbMLSand secondary antibodies Alexa Fluor 594 goat anti-rabbit IgG (red)were used The specimens were analyzed by laser confocalmicroscopy using DIC (A) and fluorescence (B)

de Oliveira et al BMC Microbiology 2013 13107 Page 5 of 14httpwwwbiomedcentralcom1471-218013107

Paracoccidioides Pb01 mycelium yeast yeast secretionsand macrophage (Figure 2B lanes 1 2 3 and 4 respec-tively) The membranes were incubated with PbMLS andsubsequently were incubated with antibody anti-PbMLSand secondary antibody anti-rabbit IgG Several proteinsidentified in the pull-down assays interacted with PbMLSat this point which suggested the veracity of the interac-tions Negative control was obtained by incubating eachprotein extract with the anti-PbMLS antibody withoutpreincubation with PbMLS (Figure 2B lanes 5 6 7 and 8)The numbers identify the proteins that interacted withPbMLS as shown in Additional file 2 Table S1

PbMLS binds to the surface of macrophagesBecause the results from Far-Western blot assays revealedseveral macrophage proteins interacting with PbMLS weperformed immunofluorescence microscopy to visualizewhether PbMLS could adhere to the surface of the macro-phage cells No binding was observed using BSA as a con-trol (Figure 3A) The arrow indicates PbMLS binding to amacrophage surface (Figure 3B)

PbMLS participates in the adherence of Paracoccidioidesto pneumocyte cellsBecause the fungus initially reaches the lungs the partici-pation of PbMLS in the adherence of ParacoccidioidesPb18 to pneumocyte cells was investigated by using con-focal laser scanning microscopy A549 cells were pretreatedwith anti-PbMLS and infected with Paracoccidioides Pb18isolate After washings with frozen PBS-T the monolayerswere incubated with Alexa Fluor that was 594-conjugatedfor labeling the antibody The arrows indicate PbMLSinteracting with the A549 surface (Figures 4A and B)

Homology modelsIn silico analysis was performed to investigate how the in-teractions identified by pull-down and two-hybrid assayscould occur Some PbMLS-interacting proteins from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle were selected for analysis Because PbMLS partici-pates in the glyoxylate cycle interaction between proteinsfrom different metabolic pathways would be expected Be-cause no crystal structure of PbMLS-interacting proteinsdescribed here was reported a three-dimensional homologymodel for each protein was constructed based on the struc-ture template listed in Additional file 6 Table S5 All of the3D-structure templates used to build models of the proteinshave a resolution of lt 20 Aring and an identity of gt 49 with acoverage of gt 91Homology models of the PbMLS-interacting proteins

have very little conformational change when comparedto their templates (Additional file 6 Table S5) The lar-gest deviations were observed for enolase and fructose

de Oliveira et al BMC Microbiology 2013 13107 Page 6 of 14httpwwwbiomedcentralcom1471-218013107

16 bisphosphate aldolase with 265 Aring and 144 Aring ofroot mean square derivation (RMSD) when superposedon the template when considering the non-hydrogenatoms For enolase there is a significant conformationalchange only in the C-terminal regions and betweenPRO143 and ASN155 (data not shown)Alpha-helix-like secondary-structure patterns were ob-

served in a greater proportion in the homology modelsPbMLS-interacting proteins For almost all of the struc-tures the alpha-helix-like pattern corresponded to morethan 40 of the whole structure while the beta-sheet-like pattern accounted for less than 20 except for theprotein ubiquitin whose quantity of beta-sheet-like pat-tern was greater (Additional file 6 Table S5)Ramachandran plots of homology models were assessed

stereo-chemically through the RAMPAGE web server [26](data not shown) For all of the proteins the Φ and Ψ dis-tributions of the Ramachandran plots were always above94 in the favored regions and less than 35 in theallowed regions The quality factors of the structures wereestimated by the ERRAT web server and are summarizedin Additional file 6 Table S5

Molecular dynamicsAll of the proteins were subjected to at least 20 ns simula-tion using GROMACS software [27] For the proteinsgamma actin 2-methylcitrate synthase triosephosphateisomerase and ubiquitin that time was insufficient toachieve RMSD stability of non-hydrogen atoms with re-spect to the structure homology models In those casesmore simulation time was provided until this conditionwas achieved The times required are listed for each pro-tein For almost all of the proteins the deviations fromtheir homology models were low (approximately 30 Aring)Specifically ubiquitin and 2-methylcitrate synthase hadthe highest RMSDs The increase was 765 Aring and 634 Aringafter 60 ns and 40 ns respectively When only the residuesfrom the interfaces of the complexes were consideredthe RMSDs increased 90 Aring and 587 Aring respectively(Additional file 6 Table S5)The alpha-helix-like pattern was slightly reduced in all

of the proteins that were binding to PbMLS but the beta-sheet-like structures almost did not change Although theRMSDs were high for ubiquitin and 2-methylcitrate syn-thase the alpha-helix-like patterns decreased to only106 and 69 respectively

Molecular docking and molecular dynamics of theprotein-protein complexesMolecular docking between PbMLS and PbMLS-in-teracting proteins was investigated by the GRAMM-Xweb server using the structures stabilized by DM Onlythe best model-structures provided by the server were se-lected These complexes were then subjected to a rapid

DM so that their structures could accommodate and avoidhigh energy at the interface between them thus identifyingresidues in this region Significant conformational changesoccurred in ubiquitin and 2-methylcitrate synthase whenthey were complexed with PbMLS (data not shown) Theresidues contacting at the interface of the complexes areshown in Additional file 7 Table S6 and these amino acidsare highlighted in Figure 5 Some amino acid residues arecommon to different proteins For example ASP379 andGLN380 are residues of PbMLS that interact with enolaseand ubiquitin ASN386 is at the interface for gamma actinand ubiquitin LEU388 is common to triosephosphateisomerase and glyceraldehyde-3-phosphate dehydrogenaseand ASP401 is common to 2-methylcitrate synthase andmalate dehydrogenaseThe protein-protein complexes relaxed by DM were pro-

vided to the Fiberdock web server which determined theglobal energy for each complex (Additional file 7 TableS6) The results showed that fructose 1 6 bisphosphate al-dolase and ubiquitin were well stabilized when complexedwith PbMLS The ASP265 residue of PbMLS is present inthe interaction of both proteins

DiscussionOur previous studies showed that PbMLS is required inthe metabolism of Paracoccidioides Pb01 acting in theglyoxylate cycle and in the allantoin degradation pathwayPbMLS condenses acetyl-CoA from both 2C sources(glyoxylate cycle) and nitrogen sources (from proline andpurine metabolism) to produce malate which is a centralmolecule of the tricarboxylic acid cycle or glyoxylate cycle[8] In addition PbMLS is located in the cytoplasm and onthe fungal cell surface and is secreted behaving like an an-chorless adhesin [9] The strong evidence for PbMLSmultifunctionality increased our interest in researchingthe possibility of new roles for PbMLS through studies ofprotein-protein interactions which aimed to identifyPbMLS-interacting proteinsWe searched for PbMLS-interacting proteins using

Far-Western blot pull-down and two-hybrid techniquesThe two-hybrid and pull-down are used as complemen-tary techniques because the results depend on variants ofthe methods The two-hybrid system is highly sensitive todetecting low-abundance proteins unlike the pull-downsystem which detects high-abundance molecules Add-itionally the two-hybrid system allows identifying strongand weak interactions while the pull-down is not a sensi-tive method for identifying some of the weak interactionsbecause of the wash steps [28] Because the principles ofthe techniques are different we have the capability ofidentifying different proteinsPull-down assays were performed using Paracoccidioides

Pb01 mycelium yeast and yeast-secreted protein extractsbecause protein differences [12] and metabolic differences

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 4 of 14httpwwwbiomedcentralcom1471-218013107

Protein interactions obtained by a two-hybrid assay areshown in Figure 1A Protein interactions obtained by pull-down assays with protein extracts of ParacoccidioidesPb01 mycelium yeast and yeast-secretions are shown inFigure 1B C and D respectively Ubiquitin (YLL039C)was the only protein that interacted with MLS that wasfound in both Paracoccidioides and S cerevisiae The otherproteins were identified in Paracoccidioides Pb01 or Scerevisiae but not in both Although some proteins identi-fied in Paracoccidioides Pb01 have homologous proteinsin S cerevisiae (Additional file 5 Table S4) these proteinscould not yet be identified as interacting with PbMLSMost of the Paracoccidioides Pb01 proteins that interactedwith PbMLS were related to the metabolism category

Confirmation of the interactions by Far-Western blotassaysFar-Western blot assays were conducted to confirm the in-teractions between PbMLS and other proteins from the

Figure 2 Confirmation of the interactions by Far-Western blot assaysMembranes were reacted with Paracoccidioides protein extracts of myceliumsubsequently incubated with anti-rabbit IgG anti-enolase anti-triosephosphwith anti-rabbit IgG conjugated to alkaline phosphatase Negative control wanti-triosephosphate isomerase and anti-actin respectively without preincuwas obtained by incubating the PbMLS with the polyclonal anti-PbMLS antsecretions and macrophages (lanes 1 2 3 and 4 respectively) were subjectwere incubated with PbMLS and subsequently primary antibody anti-PbMobtained by incubating each protein extract with anti-PbMLS antibody witindicate the proteins (Additional file 2 Table S1) that interact with PbMLS t

fungus identified by pull-down assays PbMLS wassubjected to SDS-PAGE and was electro blotted Themembranes were reacted with protein extracts of Para-coccidioides Pb01 mycelium yeast and macrophage(Figure 2A lanes 1 2 and 3 respectively) and were subse-quently incubated with rabbit IgG anti-enolase anti-triosephosphate isomerase and anti-actin respectively Thereactions were revealed with anti-rabbit IgG conjugated toalkaline phosphatase Positive signals to the three extractsindicated the presence of an interaction between PbMLSand enolase triosephosphate isomerase and actin Negativecontrol was obtained by incubating PbMLS with the anti-bodies anti-enolase anti-triosephosphate isomerase andanti-actin respectively without preincubation with the pro-tein extracts (Figure 2A lanes 4 5 and 6 respectively) Posi-tive control was obtained by incubating the PbMLS withthe polyclonal anti-PbMLS antibody (Figure 2A lane 7)Another Far-Western blot assay was performed

using membranes that contained protein extracts of

(A) PbMLS was subjected to SDS-PAGE and electro blotted(lane 1) yeast (lane 2) and macrophage (lane 3) and were

ate isomerase and anti-actin respectively The reactions were revealedas obtained by incubating PbMLS with the antibodies anti-enolasebation with the protein extracts (lanes 4 5 and 6) The positive controlibody (lane 7) (B) Protein extracts of Paracoccidioides mycelium yeasted to SDS-PAGE and blotted onto nylon membrane The membranesLS and secondary antibody anti-rabbit IgG Negative control washout preincubation with PbMLS (lanes 5 6 7 and 8) The numbershat are confirmed by this technique

Figure 3 Binding of PbMLS to the macrophage surfaceImmunofluorescence microscopy that shows the binding of PbMLSto J774 A1 mouse macrophage cells (A) Negative control wasperformed with the unrelated protein BSA (B) Arrows indicatePbMLS (green) binding to the macrophage cell surfaces blueindicates the macrophage nucleus

Figure 4 Interaction between Paracoccidioides yeast cells andpneumocytes by confocal laser scanning microscopy Infectedcell monolayers were fixed and permeabilized Primary anti-PbMLSand secondary antibodies Alexa Fluor 594 goat anti-rabbit IgG (red)were used The specimens were analyzed by laser confocalmicroscopy using DIC (A) and fluorescence (B)

de Oliveira et al BMC Microbiology 2013 13107 Page 5 of 14httpwwwbiomedcentralcom1471-218013107

Paracoccidioides Pb01 mycelium yeast yeast secretionsand macrophage (Figure 2B lanes 1 2 3 and 4 respec-tively) The membranes were incubated with PbMLS andsubsequently were incubated with antibody anti-PbMLSand secondary antibody anti-rabbit IgG Several proteinsidentified in the pull-down assays interacted with PbMLSat this point which suggested the veracity of the interac-tions Negative control was obtained by incubating eachprotein extract with the anti-PbMLS antibody withoutpreincubation with PbMLS (Figure 2B lanes 5 6 7 and 8)The numbers identify the proteins that interacted withPbMLS as shown in Additional file 2 Table S1

PbMLS binds to the surface of macrophagesBecause the results from Far-Western blot assays revealedseveral macrophage proteins interacting with PbMLS weperformed immunofluorescence microscopy to visualizewhether PbMLS could adhere to the surface of the macro-phage cells No binding was observed using BSA as a con-trol (Figure 3A) The arrow indicates PbMLS binding to amacrophage surface (Figure 3B)

PbMLS participates in the adherence of Paracoccidioidesto pneumocyte cellsBecause the fungus initially reaches the lungs the partici-pation of PbMLS in the adherence of ParacoccidioidesPb18 to pneumocyte cells was investigated by using con-focal laser scanning microscopy A549 cells were pretreatedwith anti-PbMLS and infected with Paracoccidioides Pb18isolate After washings with frozen PBS-T the monolayerswere incubated with Alexa Fluor that was 594-conjugatedfor labeling the antibody The arrows indicate PbMLSinteracting with the A549 surface (Figures 4A and B)

Homology modelsIn silico analysis was performed to investigate how the in-teractions identified by pull-down and two-hybrid assayscould occur Some PbMLS-interacting proteins from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle were selected for analysis Because PbMLS partici-pates in the glyoxylate cycle interaction between proteinsfrom different metabolic pathways would be expected Be-cause no crystal structure of PbMLS-interacting proteinsdescribed here was reported a three-dimensional homologymodel for each protein was constructed based on the struc-ture template listed in Additional file 6 Table S5 All of the3D-structure templates used to build models of the proteinshave a resolution of lt 20 Aring and an identity of gt 49 with acoverage of gt 91Homology models of the PbMLS-interacting proteins

have very little conformational change when comparedto their templates (Additional file 6 Table S5) The lar-gest deviations were observed for enolase and fructose

de Oliveira et al BMC Microbiology 2013 13107 Page 6 of 14httpwwwbiomedcentralcom1471-218013107

16 bisphosphate aldolase with 265 Aring and 144 Aring ofroot mean square derivation (RMSD) when superposedon the template when considering the non-hydrogenatoms For enolase there is a significant conformationalchange only in the C-terminal regions and betweenPRO143 and ASN155 (data not shown)Alpha-helix-like secondary-structure patterns were ob-

served in a greater proportion in the homology modelsPbMLS-interacting proteins For almost all of the struc-tures the alpha-helix-like pattern corresponded to morethan 40 of the whole structure while the beta-sheet-like pattern accounted for less than 20 except for theprotein ubiquitin whose quantity of beta-sheet-like pat-tern was greater (Additional file 6 Table S5)Ramachandran plots of homology models were assessed

stereo-chemically through the RAMPAGE web server [26](data not shown) For all of the proteins the Φ and Ψ dis-tributions of the Ramachandran plots were always above94 in the favored regions and less than 35 in theallowed regions The quality factors of the structures wereestimated by the ERRAT web server and are summarizedin Additional file 6 Table S5

Molecular dynamicsAll of the proteins were subjected to at least 20 ns simula-tion using GROMACS software [27] For the proteinsgamma actin 2-methylcitrate synthase triosephosphateisomerase and ubiquitin that time was insufficient toachieve RMSD stability of non-hydrogen atoms with re-spect to the structure homology models In those casesmore simulation time was provided until this conditionwas achieved The times required are listed for each pro-tein For almost all of the proteins the deviations fromtheir homology models were low (approximately 30 Aring)Specifically ubiquitin and 2-methylcitrate synthase hadthe highest RMSDs The increase was 765 Aring and 634 Aringafter 60 ns and 40 ns respectively When only the residuesfrom the interfaces of the complexes were consideredthe RMSDs increased 90 Aring and 587 Aring respectively(Additional file 6 Table S5)The alpha-helix-like pattern was slightly reduced in all

of the proteins that were binding to PbMLS but the beta-sheet-like structures almost did not change Although theRMSDs were high for ubiquitin and 2-methylcitrate syn-thase the alpha-helix-like patterns decreased to only106 and 69 respectively

Molecular docking and molecular dynamics of theprotein-protein complexesMolecular docking between PbMLS and PbMLS-in-teracting proteins was investigated by the GRAMM-Xweb server using the structures stabilized by DM Onlythe best model-structures provided by the server were se-lected These complexes were then subjected to a rapid

DM so that their structures could accommodate and avoidhigh energy at the interface between them thus identifyingresidues in this region Significant conformational changesoccurred in ubiquitin and 2-methylcitrate synthase whenthey were complexed with PbMLS (data not shown) Theresidues contacting at the interface of the complexes areshown in Additional file 7 Table S6 and these amino acidsare highlighted in Figure 5 Some amino acid residues arecommon to different proteins For example ASP379 andGLN380 are residues of PbMLS that interact with enolaseand ubiquitin ASN386 is at the interface for gamma actinand ubiquitin LEU388 is common to triosephosphateisomerase and glyceraldehyde-3-phosphate dehydrogenaseand ASP401 is common to 2-methylcitrate synthase andmalate dehydrogenaseThe protein-protein complexes relaxed by DM were pro-

vided to the Fiberdock web server which determined theglobal energy for each complex (Additional file 7 TableS6) The results showed that fructose 1 6 bisphosphate al-dolase and ubiquitin were well stabilized when complexedwith PbMLS The ASP265 residue of PbMLS is present inthe interaction of both proteins

DiscussionOur previous studies showed that PbMLS is required inthe metabolism of Paracoccidioides Pb01 acting in theglyoxylate cycle and in the allantoin degradation pathwayPbMLS condenses acetyl-CoA from both 2C sources(glyoxylate cycle) and nitrogen sources (from proline andpurine metabolism) to produce malate which is a centralmolecule of the tricarboxylic acid cycle or glyoxylate cycle[8] In addition PbMLS is located in the cytoplasm and onthe fungal cell surface and is secreted behaving like an an-chorless adhesin [9] The strong evidence for PbMLSmultifunctionality increased our interest in researchingthe possibility of new roles for PbMLS through studies ofprotein-protein interactions which aimed to identifyPbMLS-interacting proteinsWe searched for PbMLS-interacting proteins using

Far-Western blot pull-down and two-hybrid techniquesThe two-hybrid and pull-down are used as complemen-tary techniques because the results depend on variants ofthe methods The two-hybrid system is highly sensitive todetecting low-abundance proteins unlike the pull-downsystem which detects high-abundance molecules Add-itionally the two-hybrid system allows identifying strongand weak interactions while the pull-down is not a sensi-tive method for identifying some of the weak interactionsbecause of the wash steps [28] Because the principles ofthe techniques are different we have the capability ofidentifying different proteinsPull-down assays were performed using Paracoccidioides

Pb01 mycelium yeast and yeast-secreted protein extractsbecause protein differences [12] and metabolic differences

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

Figure 3 Binding of PbMLS to the macrophage surfaceImmunofluorescence microscopy that shows the binding of PbMLSto J774 A1 mouse macrophage cells (A) Negative control wasperformed with the unrelated protein BSA (B) Arrows indicatePbMLS (green) binding to the macrophage cell surfaces blueindicates the macrophage nucleus

Figure 4 Interaction between Paracoccidioides yeast cells andpneumocytes by confocal laser scanning microscopy Infectedcell monolayers were fixed and permeabilized Primary anti-PbMLSand secondary antibodies Alexa Fluor 594 goat anti-rabbit IgG (red)were used The specimens were analyzed by laser confocalmicroscopy using DIC (A) and fluorescence (B)

de Oliveira et al BMC Microbiology 2013 13107 Page 5 of 14httpwwwbiomedcentralcom1471-218013107

Paracoccidioides Pb01 mycelium yeast yeast secretionsand macrophage (Figure 2B lanes 1 2 3 and 4 respec-tively) The membranes were incubated with PbMLS andsubsequently were incubated with antibody anti-PbMLSand secondary antibody anti-rabbit IgG Several proteinsidentified in the pull-down assays interacted with PbMLSat this point which suggested the veracity of the interac-tions Negative control was obtained by incubating eachprotein extract with the anti-PbMLS antibody withoutpreincubation with PbMLS (Figure 2B lanes 5 6 7 and 8)The numbers identify the proteins that interacted withPbMLS as shown in Additional file 2 Table S1

PbMLS binds to the surface of macrophagesBecause the results from Far-Western blot assays revealedseveral macrophage proteins interacting with PbMLS weperformed immunofluorescence microscopy to visualizewhether PbMLS could adhere to the surface of the macro-phage cells No binding was observed using BSA as a con-trol (Figure 3A) The arrow indicates PbMLS binding to amacrophage surface (Figure 3B)

PbMLS participates in the adherence of Paracoccidioidesto pneumocyte cellsBecause the fungus initially reaches the lungs the partici-pation of PbMLS in the adherence of ParacoccidioidesPb18 to pneumocyte cells was investigated by using con-focal laser scanning microscopy A549 cells were pretreatedwith anti-PbMLS and infected with Paracoccidioides Pb18isolate After washings with frozen PBS-T the monolayerswere incubated with Alexa Fluor that was 594-conjugatedfor labeling the antibody The arrows indicate PbMLSinteracting with the A549 surface (Figures 4A and B)

Homology modelsIn silico analysis was performed to investigate how the in-teractions identified by pull-down and two-hybrid assayscould occur Some PbMLS-interacting proteins from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle were selected for analysis Because PbMLS partici-pates in the glyoxylate cycle interaction between proteinsfrom different metabolic pathways would be expected Be-cause no crystal structure of PbMLS-interacting proteinsdescribed here was reported a three-dimensional homologymodel for each protein was constructed based on the struc-ture template listed in Additional file 6 Table S5 All of the3D-structure templates used to build models of the proteinshave a resolution of lt 20 Aring and an identity of gt 49 with acoverage of gt 91Homology models of the PbMLS-interacting proteins

have very little conformational change when comparedto their templates (Additional file 6 Table S5) The lar-gest deviations were observed for enolase and fructose

de Oliveira et al BMC Microbiology 2013 13107 Page 6 of 14httpwwwbiomedcentralcom1471-218013107

16 bisphosphate aldolase with 265 Aring and 144 Aring ofroot mean square derivation (RMSD) when superposedon the template when considering the non-hydrogenatoms For enolase there is a significant conformationalchange only in the C-terminal regions and betweenPRO143 and ASN155 (data not shown)Alpha-helix-like secondary-structure patterns were ob-

served in a greater proportion in the homology modelsPbMLS-interacting proteins For almost all of the struc-tures the alpha-helix-like pattern corresponded to morethan 40 of the whole structure while the beta-sheet-like pattern accounted for less than 20 except for theprotein ubiquitin whose quantity of beta-sheet-like pat-tern was greater (Additional file 6 Table S5)Ramachandran plots of homology models were assessed

stereo-chemically through the RAMPAGE web server [26](data not shown) For all of the proteins the Φ and Ψ dis-tributions of the Ramachandran plots were always above94 in the favored regions and less than 35 in theallowed regions The quality factors of the structures wereestimated by the ERRAT web server and are summarizedin Additional file 6 Table S5

Molecular dynamicsAll of the proteins were subjected to at least 20 ns simula-tion using GROMACS software [27] For the proteinsgamma actin 2-methylcitrate synthase triosephosphateisomerase and ubiquitin that time was insufficient toachieve RMSD stability of non-hydrogen atoms with re-spect to the structure homology models In those casesmore simulation time was provided until this conditionwas achieved The times required are listed for each pro-tein For almost all of the proteins the deviations fromtheir homology models were low (approximately 30 Aring)Specifically ubiquitin and 2-methylcitrate synthase hadthe highest RMSDs The increase was 765 Aring and 634 Aringafter 60 ns and 40 ns respectively When only the residuesfrom the interfaces of the complexes were consideredthe RMSDs increased 90 Aring and 587 Aring respectively(Additional file 6 Table S5)The alpha-helix-like pattern was slightly reduced in all

of the proteins that were binding to PbMLS but the beta-sheet-like structures almost did not change Although theRMSDs were high for ubiquitin and 2-methylcitrate syn-thase the alpha-helix-like patterns decreased to only106 and 69 respectively

Molecular docking and molecular dynamics of theprotein-protein complexesMolecular docking between PbMLS and PbMLS-in-teracting proteins was investigated by the GRAMM-Xweb server using the structures stabilized by DM Onlythe best model-structures provided by the server were se-lected These complexes were then subjected to a rapid

DM so that their structures could accommodate and avoidhigh energy at the interface between them thus identifyingresidues in this region Significant conformational changesoccurred in ubiquitin and 2-methylcitrate synthase whenthey were complexed with PbMLS (data not shown) Theresidues contacting at the interface of the complexes areshown in Additional file 7 Table S6 and these amino acidsare highlighted in Figure 5 Some amino acid residues arecommon to different proteins For example ASP379 andGLN380 are residues of PbMLS that interact with enolaseand ubiquitin ASN386 is at the interface for gamma actinand ubiquitin LEU388 is common to triosephosphateisomerase and glyceraldehyde-3-phosphate dehydrogenaseand ASP401 is common to 2-methylcitrate synthase andmalate dehydrogenaseThe protein-protein complexes relaxed by DM were pro-

vided to the Fiberdock web server which determined theglobal energy for each complex (Additional file 7 TableS6) The results showed that fructose 1 6 bisphosphate al-dolase and ubiquitin were well stabilized when complexedwith PbMLS The ASP265 residue of PbMLS is present inthe interaction of both proteins

DiscussionOur previous studies showed that PbMLS is required inthe metabolism of Paracoccidioides Pb01 acting in theglyoxylate cycle and in the allantoin degradation pathwayPbMLS condenses acetyl-CoA from both 2C sources(glyoxylate cycle) and nitrogen sources (from proline andpurine metabolism) to produce malate which is a centralmolecule of the tricarboxylic acid cycle or glyoxylate cycle[8] In addition PbMLS is located in the cytoplasm and onthe fungal cell surface and is secreted behaving like an an-chorless adhesin [9] The strong evidence for PbMLSmultifunctionality increased our interest in researchingthe possibility of new roles for PbMLS through studies ofprotein-protein interactions which aimed to identifyPbMLS-interacting proteinsWe searched for PbMLS-interacting proteins using

Far-Western blot pull-down and two-hybrid techniquesThe two-hybrid and pull-down are used as complemen-tary techniques because the results depend on variants ofthe methods The two-hybrid system is highly sensitive todetecting low-abundance proteins unlike the pull-downsystem which detects high-abundance molecules Add-itionally the two-hybrid system allows identifying strongand weak interactions while the pull-down is not a sensi-tive method for identifying some of the weak interactionsbecause of the wash steps [28] Because the principles ofthe techniques are different we have the capability ofidentifying different proteinsPull-down assays were performed using Paracoccidioides

Pb01 mycelium yeast and yeast-secreted protein extractsbecause protein differences [12] and metabolic differences

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 6 of 14httpwwwbiomedcentralcom1471-218013107

16 bisphosphate aldolase with 265 Aring and 144 Aring ofroot mean square derivation (RMSD) when superposedon the template when considering the non-hydrogenatoms For enolase there is a significant conformationalchange only in the C-terminal regions and betweenPRO143 and ASN155 (data not shown)Alpha-helix-like secondary-structure patterns were ob-

served in a greater proportion in the homology modelsPbMLS-interacting proteins For almost all of the struc-tures the alpha-helix-like pattern corresponded to morethan 40 of the whole structure while the beta-sheet-like pattern accounted for less than 20 except for theprotein ubiquitin whose quantity of beta-sheet-like pat-tern was greater (Additional file 6 Table S5)Ramachandran plots of homology models were assessed

stereo-chemically through the RAMPAGE web server [26](data not shown) For all of the proteins the Φ and Ψ dis-tributions of the Ramachandran plots were always above94 in the favored regions and less than 35 in theallowed regions The quality factors of the structures wereestimated by the ERRAT web server and are summarizedin Additional file 6 Table S5

Molecular dynamicsAll of the proteins were subjected to at least 20 ns simula-tion using GROMACS software [27] For the proteinsgamma actin 2-methylcitrate synthase triosephosphateisomerase and ubiquitin that time was insufficient toachieve RMSD stability of non-hydrogen atoms with re-spect to the structure homology models In those casesmore simulation time was provided until this conditionwas achieved The times required are listed for each pro-tein For almost all of the proteins the deviations fromtheir homology models were low (approximately 30 Aring)Specifically ubiquitin and 2-methylcitrate synthase hadthe highest RMSDs The increase was 765 Aring and 634 Aringafter 60 ns and 40 ns respectively When only the residuesfrom the interfaces of the complexes were consideredthe RMSDs increased 90 Aring and 587 Aring respectively(Additional file 6 Table S5)The alpha-helix-like pattern was slightly reduced in all

of the proteins that were binding to PbMLS but the beta-sheet-like structures almost did not change Although theRMSDs were high for ubiquitin and 2-methylcitrate syn-thase the alpha-helix-like patterns decreased to only106 and 69 respectively

Molecular docking and molecular dynamics of theprotein-protein complexesMolecular docking between PbMLS and PbMLS-in-teracting proteins was investigated by the GRAMM-Xweb server using the structures stabilized by DM Onlythe best model-structures provided by the server were se-lected These complexes were then subjected to a rapid

DM so that their structures could accommodate and avoidhigh energy at the interface between them thus identifyingresidues in this region Significant conformational changesoccurred in ubiquitin and 2-methylcitrate synthase whenthey were complexed with PbMLS (data not shown) Theresidues contacting at the interface of the complexes areshown in Additional file 7 Table S6 and these amino acidsare highlighted in Figure 5 Some amino acid residues arecommon to different proteins For example ASP379 andGLN380 are residues of PbMLS that interact with enolaseand ubiquitin ASN386 is at the interface for gamma actinand ubiquitin LEU388 is common to triosephosphateisomerase and glyceraldehyde-3-phosphate dehydrogenaseand ASP401 is common to 2-methylcitrate synthase andmalate dehydrogenaseThe protein-protein complexes relaxed by DM were pro-

vided to the Fiberdock web server which determined theglobal energy for each complex (Additional file 7 TableS6) The results showed that fructose 1 6 bisphosphate al-dolase and ubiquitin were well stabilized when complexedwith PbMLS The ASP265 residue of PbMLS is present inthe interaction of both proteins

DiscussionOur previous studies showed that PbMLS is required inthe metabolism of Paracoccidioides Pb01 acting in theglyoxylate cycle and in the allantoin degradation pathwayPbMLS condenses acetyl-CoA from both 2C sources(glyoxylate cycle) and nitrogen sources (from proline andpurine metabolism) to produce malate which is a centralmolecule of the tricarboxylic acid cycle or glyoxylate cycle[8] In addition PbMLS is located in the cytoplasm and onthe fungal cell surface and is secreted behaving like an an-chorless adhesin [9] The strong evidence for PbMLSmultifunctionality increased our interest in researchingthe possibility of new roles for PbMLS through studies ofprotein-protein interactions which aimed to identifyPbMLS-interacting proteinsWe searched for PbMLS-interacting proteins using

Far-Western blot pull-down and two-hybrid techniquesThe two-hybrid and pull-down are used as complemen-tary techniques because the results depend on variants ofthe methods The two-hybrid system is highly sensitive todetecting low-abundance proteins unlike the pull-downsystem which detects high-abundance molecules Add-itionally the two-hybrid system allows identifying strongand weak interactions while the pull-down is not a sensi-tive method for identifying some of the weak interactionsbecause of the wash steps [28] Because the principles ofthe techniques are different we have the capability ofidentifying different proteinsPull-down assays were performed using Paracoccidioides

Pb01 mycelium yeast and yeast-secreted protein extractsbecause protein differences [12] and metabolic differences

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

Figure 5 Complexes between PbMLS-interacting proteins (red) and PbMLS (green) after protein-protein docking simulations by usingGramm-X and GROMACS software (A) Enolase (B) Fructose 1 6 bisphosphate aldolase (C) Gamma actin (D) Glyceraldehyde-3-phosphateisomerase (E) Malate dehydrogenase (F) 2-Methylcitrate dehydratase (G) Triosephosphate isomerase and (H) Ubiquitin The amino acid residuesthat are involved in complex formation are highlighted

de Oliveira et al BMC Microbiology 2013 13107 Page 7 of 14httpwwwbiomedcentralcom1471-218013107

including changes in the PbMLS transcript expressionlevel [29] were observed between both phases whichcould lead to different PbMLS-interacting proteins In factconsidering mycelium and yeast 4 proteins were exclusiveto mycelium and 7 were exclusive to yeast In addition 5proteins were exclusive to yeast-secreted extract and 15were exclusive to macrophage A total of 13 of those pro-teins were also identified by Far-Western blot These find-ings suggest that PbMLS appears to play a different role inParacoccidioides Pb01 because it interacts with proteinsfrom diverse functional categories

Several significant interactions were found PbMLSinteracted with fatty acid synthase subunit beta whichcatalyzes the synthesis of long-chain saturated fattyacids PbMLS interacted with 2-methylcitrate synthaseand 2-methylcitrate dehydratase which are enzymes ofthe cycle of 2-methylcitrate This cycle is related to themetabolism of propionyl-coenzyme A (and odd-chainfatty acids) unlike the glyoxylate cycle which is relatedto the metabolism of even-chain fatty acids The inter-action of PbMLS with these enzymes suggests its in-volvement in fatty acid metabolism regulation

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 8 of 14httpwwwbiomedcentralcom1471-218013107

The peroxisomal enzyme malate dehydrogenase whichparticipates in the glyoxylate cycle [30] interacts withPbMLS In addition to having the signal peptide AKL thattargets peroxisomes [8] PbMLS was localized in that or-ganelle [9]PbMLS interacts with serine threonine kinase It is

known that protein kinases catalyze the transfer of thegamma phosphate of nucleotide triphosphates (ATP) toone or more amino acids of the protein side chain whichresults in a conformational change that affects the func-tion of the protein resulting in a functional alteration ofthe target protein by altering enzymatic activity cellularlocalization or association with other proteins [31] Thusthe interaction with a protein kinase suggests that PbMLScould be regulated by phosphorylation PbMLS has a var-iety of sites which indicates possible post-translationalmodifications including protein kinase phosphorylationsites [8] We have already described the regulation byphosphorylation of PbICL the other enzyme unique tothe glyoxylate cycle [32]The secretion of PbMLS [9] suggests that it interacts

with fungus proteins themselves and host surface proteinsExtracellular vesicles from Paracoccidioides spp presentproteins with many functions [33] Of 11 PbMLS-interacting proteins 5 were also found in the extracellularvesicle Extracellular proteins are known to play importantroles such as the uptake of nutrients cell-cell communi-cation and detoxification of the environment [34] Morespecifically proteins secreted by pathogenic microorgan-isms appear to play important roles in virulence [35]Corroborating our results many proteins identified in thisstudy such as 2-methylcitrate synthase malate dehydro-genase nucleoside diphosphate kinase pyruvate kinasehsp70-like protein and Cobalamin-independent methio-nine synthase had previously been described as secretedproteins in Paracoccidioides Pb01 secretome from myce-lium and yeast cells [36]The adhesion of pathogens to host cells is considered to

be an essential step in the establishment of infection [37]Several clinically important fungi such as Candidaalbicans Aspergillus fumigatus Histoplasma capsulatumand Cryptococcus neoformans are known to bind to pro-teins of the extracellular matrix (ECM) [38] The adhesinsof fungi are important in the migration invasion differen-tiation and proliferation of microbes Paracoccidioidesyeast cells also have the ability to adhere and invadehost cells [3940] Some adhesins such as PbDfg5p [41]triosephosphate isomerase (PbTPI) [42] glyceraldehyde-3-phosphate dehydrogenase (PbGAPDH) [39] and enolase(PbEno) [43] and PbMLS [9] have been described inParacoccidioides Pb01 Here the interaction betweenPbMLS and enolase and triosephosphate isomerase wasconfirmed by Far-Western blot assay The interaction ofPbMLS with those proteins suggests that the joint action

of those adhesins could promote adhesion to and invasionof host cells acting as potent virulence factorsPbMLS appears to act in the interaction between

Paracoccidioides Pb01 and macrophage because it inter-acts with several macrophage-specific proteins of which 5proteins are related to cytoskeleton which suggests the in-volvement of that structure in the fungus adhesionprocess The PbMLS binding to actin was confirmed byFar-Western blot The cytoskeletons of the macrophagescontrol the movement of the cell membrane which re-flects the movement of the cell as a whole and are also in-volved in processes such as phagocytosis [44] Ourprevious work used Far-Western blotting and flow cytom-etry to show that PbMLS binds to A549 cells Here theparticipation of PbMLS in Paracoccidioides Pb01 adhesionto and invasion of A549 cells was confirmed using con-focal laser scanning microscopySome PbMLS-interacting proteins were selected for in

silico interaction analysis Proteins were chosen from meta-bolic pathways such as the glycolytic pathway the tricarb-oxylic acid cycle the methyl citrate cycle and the glyoxylatecycle because PbMLS participates in the glyoxylate cycleand the interaction between proteins from different meta-bolic pathways would be expected Global energy values foreach complex studied showed that there is good comple-mentarity between PbMLS and most PbMLS-interactingproteins For example the complexes that involve PbMLSand the proteins glyceraldehyde-3-phosphate isomerasemalate dehydrogenase 2-methylcitrate dehydratase andtriosephosphate isomerase have global energies that are lessthan minus55 kcalmol The global energy values found herewere very good For example in a recent study of the inter-actions between D-phosphoglycerate dehydrogenase andphosphoserine aminotransferase from the enteric humanparasite Entamoeba histolytica [45] the best global energieswere approximately minus75 kcalmol Here the best valueswere found for fructose 16 bisphosphate aldolase and ubi-quitin (less than minus100 kcalmol)S cerevisiae MLS-interacting proteins have already

been described Here in silico analysis using the Scerevisiae database showed that PbMLS interacts withother new proteins The only protein that they share isubiquitin This fact and the fact that the interaction be-tween ubiquitin and PbMLS is very stable suggest thatthis interaction is very important Ubiquitin is respon-sible for the conjugation of proteins marking them forselective degradation via the ubiquitin-proteasome sys-tem 26S a process that is essential in the response tocellular stress These proteins however act throughubiquitination changing the function the location andor the traffic protein or are targeted for destruction bythe 26S proteasome [46]In conclusion the molecular interactions that involve

proteins located in subcellular compartments facilitate

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 9 of 14httpwwwbiomedcentralcom1471-218013107

the understanding of mechanisms that are associatedwith each interaction However proteins are not alwaysat the same location in the cell and do not have uniqueroles [47] Here several new PbMLS-interacting proteinsfrom various functional categories were identified whichsuggests that their function is diversified beyond theglyoxylate cycle

ConclusionsThe results of this study indicated that PbMLS interactswith proteins of different functional categories such ascellular transport protein biosynthesis modification anddegradation and signal transduction These data suggestthat PbMLS is found in many locations and plays differ-ent roles in the fungal cell

MethodsParacoccidioides isolate and growth conditionsThe fungus Paracoccidioides isolate Pb01 (ATCC MYA-826) was grown as previously described [39] The yeastand mycelium phase were grown at 36 and 22 degC re-spectively in FavandashNettorsquos medium (1 wv peptone05 wv yeast extract 03 wv proteose peptone05 wv beef extract 05 wv NaCl 4 wv glucose1 wv agar pH 72)

Preparation of protein extracts from Paracoccidioides sppTotal protein extracts from Paracoccidioides spp myce-lium and yeast cells were prepared as previously described[48] Mycelium and yeast cells were frozen and groundwith a mortar and pestle in buffer (20 mM TrisndashHClpH 88 2 mM CaCl2) with protease inhibitors (50 μgmLN-α-ρ-tosyl-L-lysine chloromethylketone 1 mM 4-chloromercuribenzoic acid 20 mM leupeptin 20 mMphenylmethylsulfonyl fluoride and 5 mM iodoacetamide)The mixture was centrifuged at 10000 times g at 4degC for20 min and the supernatant was collected and storedat minus20 degCYeast-secreted proteins of Paracoccidioides spp were

prepared Culture supernatant of yeast cells was obtainedafter 24 h incubation in liquid Fava Nettorsquos medium Thecells were separated by centrifugation at 5000 times g for15 min and the supernatant was filtered in 045 and022 μm filters (MilliPore) Each 50 mL of culture super-natant was concentrated to 500 μL in 25 mM TrisndashHClpH 70 and a protease inhibitor was added The proteinconcentration of all of the samples was determinedaccording to Bradford [49]

Preparation of protein extracts from macrophageJ774 A1 mouse macrophage cells purchased from a CellBank in Rio de Janeiro Brazil [50] were cultured inRPMI 1640 supplemented with fetal bovine serum non-essential amino acids and interferon gamma (1 UmL)

To obtain the protein extract cells were detachedwith 09 saline solution containing trypsin and werecentrifuged at 5000 times g for 10 min Then milliQ waterwas added to lyse the cells and the solution wascentrifuged again Buffer (20 mM TrisndashHCl pH 88 2 mMCaCl2) and protease inhibitors were added to the pelletProtein concentration was determined according to Brad-ford [49]

Heterologous expression and purification of recombinantPbMLSPbMLS recombinant protein was obtained as describedby Zambuzzi-Carvalho et al [8] and Neto et al [9]PbMLS cDNA was cloned into the expression vectorpGEX-4-T3 (GE HealthcareW Chalfont St Giles UK)E coli (BL21 Startrade (DE3) pLys Invitrogen Grand IslandNY) was transformed with pGEX-PbMLS constructionby thermal shock and was grown in LB mediumsupplemented with ampicillin (100 μgmL) at 20degC untilreaching the optical density of 06 at 600 nm Synthesisof the recombinant protein was then initiated by addingisopropyl-β-D-thiogalactopyranoside (IPTG) (Sigma-Al-drich St Louis MO) to a final concentration of 01 mMto the growing culture After induction the cells wereincubated for 16 h at 15degC with shaking at 200 rpmCells were harvested by centrifugation at 10000 times g for10 min The supernatant was discarded and the cellswere resuspended in 1times phosphate-buffered saline (PBS)(014 M NaCl 27 mM KCl 10 mM Na2HPO4 18 mMKH2PO4 pH 74)E coli cells were incubated for 60 min with lysozyme

(100 μgmL) and were lysed by extensive sonication (25 cy-cles of 1 min) The sample was centrifuged at 8000 times g for15 min to obtain the supernatant which contained thesoluble protein fraction The recombinant protein waspurified by affinity chromatography under no denaturingconditions The soluble fraction was placed in a Glutathi-one Sepharosetimes 4B resin column (GE HealthcareW) Theresin was washed five times in 1x PBS and the recombin-ant protein was cleaved by the addition of thrombin prote-ase (50 UmL) The purity and size of the recombinantprotein were evaluated by running the molecule on 12SDS-PAGE followed by Coomassie blue staining E colicells transformed with pGEX-4 T-3 without an insert forthe expression and purification of the protein glutathioneS transferase (GST) were used as the experimentalcontrol

Antibody productionThe purified PbMLS was used to produce anti-PbMLSpolyclonal antibodies in New Zealand rabbits Theimmunization protocol constituted an initial injection of300 μg of purified recombinant protein in completeFreundrsquos adjuvant and two subsequent injections of the

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 10 of 14httpwwwbiomedcentralcom1471-218013107

same amount of the antigen in incomplete Freundrsquos adju-vant Each immunization was followed by a 14-day inter-val After the fourth immunization the serum containingthe anti-PbMLS polyclonal antibody was collected andstored at minus20degC

Pull-down assaysA total of 5 mg of each protein extract of ParacoccidioidesPb01 mycelium yeast yeast secretions and macrophagewas incubated with 20 μL of resin bound to GST for 2 h at4degC under gentle agitation (control) The resin wascentrifuged at 200 times g for 5 min and the supernatant wasplaced into a tube that contained 100 μL of the resinbonded to PbMLS This mixture was incubated for 3 h at4degC with stirring After this period the resin wascentrifuged at 200 times g for 5 min and the supernatant wasdiscarded Both resins were washed four times with 1xPBS buffer and subjected to SDS-PAGE on 15 polyacryl-amide gel followed by staining with Coomassie Blue (GEHealthcareW)Separated by SDS-PAGE the proteins that interacted

with PbMLS in the pull-down assay were excised from thegel and identified by MS Pieces of the gels were soaked in50 μL of acetonitrile The solvent was removed under avacuum and was incubated in 100 mM NH4HCO3 buffercontaining 10 mM 14-dithiothreitol for 1 h at 56degC undergentle agitation The above buffer was removed and re-placed by 55 mM iodoacetamide in 100 mM NH4HCO3

for 45 min at room temperature in the dark The gelpieces were then subjected to alternating 5 min washingcycles with NH4HCO3 and acetonitrile dried down swol-len in 50 μL of 50 mM NH4CO3 containing 125 ngmLsequencing-grades modified porcine trypsin (PromegaMadison WI) and incubated at 37degC overnight Theresulting tryptic peptides were extracted by adding 20 μLof 5 vv acetic acid and removing the solution Thisprocedure was repeated once The extracts were pooleddried under a vacuum and then solubilized in 01 vvtrifluoroacetic acid for MS analysis The proteins of thetryptic digestion samples were analyzed using a MALDI-Synapt MStrade mass spectrometer (Waters-MicromassManchester UK) The peptide mass list obtained for eachspectrum was searched using the MASCOT algorithm [14]Proteins were identified by Peptide Mass Fingerprint (PMF)andor MSMS even considering 1 tryptic cleavage lostscore gt 60 50ndash100 ppm mass error between theoreticaland experimental masses and oxidized methionine as vari-able modification resulting from in-gel digestion

Two-hybrid assaysA cDNA library was obtained using RNA extracted fromParacoccidioides Pb01 yeast cells as described previously[51] The cDNAs were synthesized and cloned into theprey vector pGADT7 to perform yeast two-hybrid

screens using the Matchmaker Two-Hybrid System 3(Clontech Laboratories Polo Alto CA) To screen protein-protein interactions in vivo with the MLS the cDNA en-coding PbMLS was sub-cloned into the bait vectorpGBKT7 The generation of transformants was obtained byintroducing the bait vector into the Saccharomycescerevisiae yeast strain Y187 (MATα trp1-901) and the preyvector into the S cerevisiae strain AH109 (MATα leu2-3)The experimental protocol was performed according to

the Matchmaker GAL4 Two-Hybrid System 3 manual andthe Yeast Protocol Handbook (Clontech) Following cellmating the S cerevisiae diploids that contained thetwo vectors were selected from plates that contained SDndashLeundashTrp minimal media To exclude false-positive clonesthe colonies were replicated using high-stringency platesthat contained SDndashAdendashHisndashLeundashTrp minimal mediaThe screening of positive clones was accomplished bydetecting the bluewhite color of the substrate 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside (X-α-GAL) Aden-ine and histidine were the reporter genes that expressedtogether with lacZ (α-galactosidase reporter gene) A PCRcolony assay was performed on the clones using AD-LD 5primeand AD-LD 3prime supplied oligonucleotides for the pGADT7-Rec bait plasmid The PCR products of the identifiedtransformants were subjected to DNA sequencing using aMegaBACE 1000 sequencer (GE HealthcareW) for auto-mated sequence analysis Sequence homologies to the genesof interest were performed by searching the GenBank data-base using the BLAST algorithm [17]

Construction of protein interaction mapsThe Osprey Network Visualization System [25] wasused to design a complex interaction network to enableviewing and manipulation [52] This program uses TheGRID protein interaction databases [24] and the Sac-charomyces Genome Database - SGD [53] In this wayinteraction maps were obtained from pull-down andtwo-hybrid Paracoccidioides Pb01 protein data Thenames of the proteins correspond to S cerevisiae andthis correspondence was obtained through analysis ofthe structural genome databases of ParacoccidioidesPb01 [54] and S cerevisiae [23]

Far-Western blot assaysFar-Western blot assays were conducted as previously de-scribed [9] PbMLS was submitted to SDS-PAGE and blot-ted onto nylon membrane After blocking for 4 h with15 (wv) BSA in 10 mM PBS-milk and washing threetimes (for 10 min each) in 10 mM triton in PBS (PBS-T)the membranes were incubated with ParacoccidioidesPb01 mycelium protein extract (100 μgmL) yeastcells (100 μgmL) and macrophage protein extract(100 μgmL) diluted in PBS-T with 2 BSA for 90 minand then washed three times (for 10 min each) in PBS-T

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 11 of 14httpwwwbiomedcentralcom1471-218013107

The membranes were incubated for 18 h with rabbit IgGanti-enolase anti-triosephosphate isomerase and anti-actin respectively in PBS-T with 2 BSA (11000 dilu-tion) The blots were washed with PBS-T and incubatedwith the secondary antibodies anti-rabbit IgG (11000 dilu-tion) The blots were washed with PBS-T and subjected toreaction with alkaline phosphatase The reaction wasdeveloped with 5-bromo-4-chloro-3-indolylphosphate nitro-bluetetrazolium (BCIPndashNBT) The negative controlwas obtained by incubating PbMLS with anti-enolaseanti-triosephosphate isomerase and anti-actin antibodieswithout preincubation with the protein extracts The posi-tive control was obtained by incubating the PbMLS withthe anti-PbMLS antibody following the reaction as previ-ously described Another Far-Western blot experimentwas performed using the same protocol but protein ex-tracts of Paracoccidioides Pb01 (mycelium yeast andyeast-secreted) and macrophages were subjected toSDS-PAGE and were blotted onto nylon membrane Themembranes were incubated with PbMLS (100 μgmL) andsubsequently with the primary antibody anti-PbMLS(14000 dilution) and the secondary antibody anti-rabbitimmunoglobulin (11000 dilution) The negative controlwas obtained by incubating each protein extract with anti-PbMLS antibody without preincubation with PbMLS

Immunofluorescence assaysAn immunofluorescence experiment was performed aspreviously described [55] J774 A1 mouse macrophagecells (106 cellsmL) were cultured over cover slips in 6-well plates and were subjected to a recombinant PbMLSbinding assay Mammalian cells were cultured in RPMIsupplemented with interferon gamma (1 UmL) Themedium was removed and the cells were washed 3 timeswith PBS fixed for 30 min with cold methanol and air-dried Either recombinant PbMLS (350 μgmL) or 1 BSA(wv negative control) in PBS was added and incubatedwith fixed macrophage cells at room temperature for 1 hAfter the cells were washed 3 times with PBS anti-PbMLSantibody (11000 dilution) was added The system was in-cubated for 1 h at 37 degC and washed 3 times with PBSThe cells were incubated with anti-rabbit IgG coupledto fluoresce in isothiocyanate (FITC 1100 dilution) for1 h The cells were incubated with 50 μM 4prime 6-diamidino-2-phenylindole (DAPI) for nuclear staining

Confocal laser scanning microscopyA confocal laser scanning microscopy experiment wasperformed as described by Batista et al [56] and Lenziet al [57] A549 cell cultivation and adhesion of theParacoccidioides strain Pb18 were performed The total ad-hesion (infection and invasion) assays were accomplishedin 24 well-plates that contained cover slips at the bottomIn all of the tests a cellular suspension with 106 cellsmL

was standardized After the tripsinization of the cell sus-pension 02 mL was removed from the bottle and dilutedin 18 mL of HAM F12 medium Cells were counted witha hemocytometer after several dilutions until the appropri-ate concentration was defined Later 05 mL of the ad-justed cell concentration was placed in each well of theplates and incubated at 36degC for 24 hThe monolayers were fixed and washed in PBS and

permeabilized in 05 Triton X-100 for 30 min After thepermeabilization step the primary antibody anti-PbMLS(150 in PBS + 3 skimmed milk + 1 BSA) was added for1 h unbound antibody was removed by washing in PBSand then Alexa Fluor 594-conjugated antibody goat anti-rabbit IgG (1400) (150 in PBS + 3 skimmed milk + 1BSA) was added for 1 h followed by three additionalwashings with frozen PBS-T before mounting in 90 gly-cerol in PBS adjusted to pH 85 and containing an anti-fading agent (p-phenylenediamine 1 gL) (Sigma-Aldrich)The specimens were analyzed by laser confocal micros-copy using differential interference contrast microscopy(DIC) and fluorescence (LSM 510-META Zeiss)

3D Structures of PbMLS-interacting proteinsThe 3D structures of proteins binding to PbMLS (PbMLS-interacting proteins) were initially predicted by the hom-ology modeling method using the modeler algorithm onthe ModWeb server [58] The quality of the structurespredicted was measured at NIH-MBI laboratory servers[59] with the ERRAT web server [60] A Ramachandranplot of each protein was checkedconferred on the RAM-PAGE web server [2661] and Verify 3D was used toevaluate the amino acid environments [62] The percent-ages of helical and sheet content were estimated using the2Struc DSSP server [63] and Helix System [64] for linearrepresentation of the secondary structuresMolecular Dynamics (MD) simulations of these struc-

tures were performed using GROMACS software [2765] toimprove the relaxation and orientation of their side chainsand to reproduce the structural stability of the receptor inits native environment [66] The Particles Mesh Ewaldmethod [67] was used to improve treatment approachesthat involve electrostatic interactions with periodic bound-ary conditions which were considered in all directions fromthe box Initially the system was neutralized by addingcounter ions and then it was immediately subjected tominimization using steepest descent energy The simula-tions were completed when the tolerance of 1000 kJmolwas no longer exceeded The first step in the equilibrationof the system was energy relaxation of the solvent for100 ps (pico seconds) only after this step was the systemsubjected to MD With a constant temperature of 300 K1 atm pressure a time-step of 2 fs (femto seconds) andwithout any restriction of the protein conformations the

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 12 of 14httpwwwbiomedcentralcom1471-218013107

simulations were performed for 20 ns (nano seconds) to60 ns depending on the proteinAll of the information concerning the trajectory of

these times was collected every 5 ps The equilibrationof the trajectory was checked by monitoring the equili-bration of the quantities such as the RMSD of non-hydrogen atoms with respect to the initial structureAnalysis of the total energy potential energy and kineticenergy were all obtained using GROMACS softwareRMSD values between final and template structures alsohelped to identify the common segments which corre-sponds to the structurally conserved regionThe average structure of the entire trajectory was also

determined using the g_rms algorithm [68] The first10 ns of the trajectory were not used to determine theaverage structures All of the water molecules were re-moved from the selected structures to proceed with thedocking simulations in the next step

Molecular dockingBy using the structures of PbMLS-interacting proteinsdetermined by MD as described above a global searchof protein-protein interactions was performed usingGRAMM-X software [69] The Protein-Protein DockingWeb Server v120 was used to perform rigid dockingSimulations were performed with no pre-conceived biastoward specific residue interactions and the best model-structure of each complex (PbMLS + PbMLS-interactingproteins) was selected

Refinement of MDMD simulations of the complexes were performed to im-prove the orientation of their side chains and to minimizethe high-magnitude repulsive interactions between atomsShort simulations were performed for the complexes de-fined by the GRAMM-X software again using GROMACSsoftware with the same force field and solvent model pre-viously used to define the 3D-structures of each proteinThe system was defined by a cubic box with periodicboundary conditions and a 9 Aring cut-off for non-bond in-teractions was used for electrostatic interactions treatedby the Particle Mesh Ewald method Overlapping watermolecules were deleted and the systems were neutralizedby adding counter ionsInitially the system was subjected to minimization using

steepest descent energy The simulations were completedwhen the tolerance of 1000 kJmol was no longerexceeded After minimization the system was subjected toa 100 ps simulation in the NVT ensemble and then wasimmediately subjected to a 100 ps simulation in the NPTensemble For both stages T = 300 K and the thermostatrelaxation constant = 01 ps additionally a Berendsenthermostat 1 atm pressure a time-step of 2 fs and pos-ition restraint of the complex were used After that step

the system was subjected to an MD run in the NPT en-semble The simulations were performed for 1 ns with aconstant temperature of 300 K 1 atm pressure a time-step of 2 fs and without any restriction on the complexconformations The structure of the complex used to de-fine the interface region between the proteins was thatobtained at the end of the simulations Fiberdock software[70] was used to estimate the global-energy that was in-volved in this interface

Additional files

Additional file 1 Figure S1 Pull-down assays for the determination ofin vitro interactions between PbMLS and other proteins ofParacoccidioides (A) Purification of GST protein (lane 1) and recombinantPbMLS (lane 2) by affinity resin The proteins detected after thepurification of PbMLS were removed from the gel and identified by MS(Additional file 2 Table S1) GST protein was incubated with proteinextracts of Paracoccidioides mycelium (B) yeast (C) secretions (D) andmacrophages (E) during which we aimed to remove nonspecific bindingproteins (lane 1) After incubation the supernatant was incubated withPbMLS-GST (purified) The protein complex resulting from this interactionwas resolved by SDS-PAGE (lane 2) The proteins numbered wereremoved from the gel and identified by MS (Additional file 2 Table S1)

Additional file 2 Table S1 PbMLS -interacting proteins by using pull-down assays identified by MS

Additional file 3 Table S2 PbMLS-interacting proteins identified bypull-down assays

Additional file 4 Table S3 Gene products interacting with PbMLS byusing two-hybrid assay identified by sequencing

Additional file 5 Table S4 PbMLS-interacting proteins alreadydescribed in the database interactions The GRID indicated in Figure 1

Additional file 6 Table S5 3D Models informations of PbMLS andPbMLS-interacting proteins

Additional file 7 Table S6 Key residues and scores of the protein-protein interaction interface

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsKMO performed pull-down assays Far-Western blot assays andimmunofluorescence microscopy BRSN performed two-hybrid assays andprepared samples for confocal microscopy assays KMO and BRSN preparedthe interaction maps RAS and GOQ performed Molecular Docking andMolecular Dynamics ARV and MJSMG performed confocal microscopyassays KMO BRSN RAS MJSMG JAP CMAS and MP contributed to thediscussion of the data and preparation of the manuscript MP conceiveddesigned and coordinated the study All authors contributed to thediscussion of results All the authors have read and approved the finalmanuscript

AcknowledgementsThis study at the Universidade Federal de Goiaacutes was supported by Ministeacuterioda Ciecircncia e TecnologiaConselho Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (MCTICNPq) Fundo Nacional de Desenvolvimento Cientiacutefico eTecnoloacutegico (FNDCT) Fundaccedilatildeo de Amparo agrave Pesquisa do Estado de Goiaacutes(FAPEG) Coordenaccedilatildeo de Aperfeiccediloamento de Pessoal de Niacutevel Superior(CAPES) Financiadora de Estudos e Projetos (FINEP) and INCT_IF (InstitutoNacional de Ciecircncia e Tecnologia para Inovaccedilatildeo Farmacecircutica) AdditionallyKMO BRSN and GOQ were supported by a fellowship from CNPq Theauthors would like to thank Henrique Leonel Lenzi (In memoriam) andMarcelo Pelajo Machado from Laboratory of Pathology Instituto OswaldoCruz Fiocruz Rio de Janeiro Brazil for help with confocal microscopy

de Oliveira et al BMC Microbiology 2013 13107 Page 13 of 14httpwwwbiomedcentralcom1471-218013107

Author details1Laboratoacuterio de Biologia Molecular Instituto de Ciecircncias BioloacutegicasUniversidade Federal de Goiaacutes Goiacircnia GO Brazil 2Nuacutecleo Colaborativo deBioSistemas Campus Jatobaacute Universidade Federal de Goiaacutes Goiacircnia GOBrazil 3Laboratoacuterio de Micologia Cliacutenica Universidade Estadual PaulistaAraraquara SP Brazil

Received 15 February 2013 Accepted 10 May 2013Published 14 May 2013

References1 Brummer E Castaneda E Restrepo A Paracoccidioidomycosis an update

Clin Microbiol Rev 1993 689ndash1172 Bernard G Kavakama J Mendes-Giannini MJM Kono A Duarte AJ Shikanai-

Yasuda MA Contribution to the natural history of paracocidioidomycosisidentification of primary pulmonary infection in the severe acute form ofthe disease - a case report Clin Infect Dis 2005 401ndash4

3 San-Blas G Nintildeo-Vega G Iturriaga T Paracoccidioides brasiliensis andparacoccidioidomycosis molecular approaches to morphogenesisdiagnosis epidemiology taxonomy and genetics Med Mycol 200240225ndash242

4 Coutinho ZF Silva D Lazeacutera M Petri V Oliveira RM Sasbroza PC Wanke BParacoccidioidomycosis mortality in Brazil Caderno Sauacutede Publica 2002181441ndash1454

5 Prado M Silva MB Laurenti R Travassos LR Taborda CP Mortality due tosystemic mycoses as a primary cause of death or in association withAIDS in Brazil a review from 1996 to 2006 Mem Inst Oswaldo Cruz 2009104513ndash521

6 Bastos KP Bailatildeo AM Borges CL Faria FP Felipe MSS Silva MG Martins WSFiuacuteza RB Pereira M Soares CMA The transcriptome analysis of earlymorphogenesis in Paracoccidioides brasiliensis mycelium reveals noveland induced genes potentially associated to the dimorphic processBMC Microbiol 2007 107ndash29

7 Derengowski LS Tavares AH Silva S Procoacutepio LS Felipe MS Silva-Pereira IUpregulation of glyoxylate cycle genes upon Paracoccidioides brasiliensisinternalization by murine macrophages and in vitro nutritional stresscondition Med Mycol 2008 46125ndash134

8 Zambuzzi-Carvalho PF Cruz AHS Santos-Silva LK Goes AM Soares CMAPereira M The malate synthase of Paracoccidioides brasiliensis Pb01 isrequired in the glyoxylate cycle and in the allantoin degradationpathway Med Mycol 2009 11ndash11

9 Neto BRS Silva JF Mendes-Giannini MJS Lenzi HL Soares CMA Pereira MThe malate synthase of Paracoccidioides brasiliensis is a linked surfaceprotein that behaves as an anchorless adhesion BMC Microbiol 20099272ndash284

10 Auerbach D Thaminy S Hottiger MO Stagljar I The post-genomic era ofinteractive proteomics facts and perspectives Proteomics 20022611ndash623

11 Vikis HG Guan KL Glutathione-S-transferase-fusion based assays forstudying protein-protein interactions Methods Mol Biol 2004 261175ndash186

12 Rezende TC Borges CL Magalhatildees AD de Sousa MV Ricart CA Bailatildeo AMSoares CM A quantitative view of the morphological phases ofParacoccidioides brasiliensis using proteomics J Proteomics 2011 75572ndash587

13 Ellis RJ van der Vies SM Molecular chaperones Annu Rev Biochem 199160321ndash347

14 MASCOT algorithm httpwwwmatrixsciencecom15 UniProt databases http wwwuniprotorg16 MIPS httpmipshelmholtz-muenchendegenreprojyeast17 BLAST algorithm httpwwwncbinlmnihgov18 PEDANT 3 database httppedanthelmholtz-muenchendeindexjsp19 Costanzo M Baryshnikova A Bellay J Kim Y Spear ED Sevier CS Ding H

Koh JL Toufighi K Mostafavi S Prinz J St Onge RP VanderSluis BMakhnevych T Vizeacoumar FJ Alizadeh S Bahr S Brost RL Chen Y CokolM Deshpande R Li Z Lin ZY Liang W Marback M Paw J San Luis BJShuteriqi E Tong AH van Dyk N et al The genetic landscape of a cellScience 2010 327425ndash431

20 Tong A Boone C Synthetic genetic array analysis in Saccharomycescerevisiae Meth Mol Biol 2006 313171ndash192

21 Tong AH Lesage G Bader GD Ding H Xu H Xin X Young J Berriz GF BrostRL Chang M Chen Y Cheng X Chua G Friesen H Goldberg DS Haynes JHumphries C He G Hussein S Ke L Krogan N Li Z Levinson JN Lu H

Meacutenard P Munyana C Parsons AB Ryan O Tonikian R Roberts T et alGlobal mapping of the yeast genetic interaction network Science 2004303808ndash813

22 Collins SR Miller KM Maas NL Roguev A Fillingham J Chu CS SchuldinerM Gebbia M Recht J Shales M Ding H Xu H Han J Ingvarsdottir K ChengB Andrews B Boone C Berger SL Hieter P Zhang Z Brown GW Ingles CJEmili A Allis CD Toczyski DP Weissman JS Greenblatt JF Krogan NJFunctional dissection of protein complexes involved in yeastchromosome biology using a genetic interaction map Nature 2007446806ndash810

23 Structural genome databases of Saccharomyces cerevisiae httpwwwbroadinstituteorgannotationgenomesaccharomyces_cerevisiae

24 The GRID protein interaction databases httpthebiogridorg25 Osprey network visualization system - version 120 httpbiodatamshri

oncaospreyservletIndex26 RAMPAGE web server httpmordredbioccamacuk~rapperrampage

php27 GROMACS software httpwwwgromacsorg28 Cho S Park SG Lee DH Park BC Protein-protein interaction networks

from interactions to networks J Biochem Mol Biol 2004 3745ndash5229 Felipe MS Andrade RV Arraes FB Nicola AM Maranhatildeo AQ Torres FA Silva-Pereira

I Poccedilas-Fonseca MJ Campos EG Moraes LM Andrade PA Tavares AH Silva SSKyaw CM Souza DP Pereira M Jesuiacuteno RS Andrade EV Parente JA Oliveira GSBarbosa MS Martins NF Fachin AL Cardoso RS Passos GA Almeida NF Walter MESoares CM Carvalho MJ Briacutegido MM Transcriptional profiles of the humanpathogenic fungus Paracoccidioides brasiliensis in mycelium and yeast cellsJ Biol Chem 2005 28024706ndash24714

30 Gietl C Malate dehydrogenase isoenzymes cellular locations and role inthe flow of metabolites between the cytoplasm and cell organellesBiochim Biophys Acta 1992 1100217ndash234

31 Hanks SK Quinn AM Hunter T The protein kinase family conservedfeatures and deduced phylogeny of the catalytic domains Science 199824142ndash52

32 Silva AH Brock M Zambuzzi-Carvalho PF Santos-Silva LK Troian RF GoacuteesAM Soares CMA Pereira M Phosphorylation is the major mechanismregulating isocitrate lyase activity in Paracoccidioides brasiliensis yeastcells FEBS Journal 2011 2782318ndash2332

33 Vallejo MC Nakayasu ES Matsuo AS Sobreira TJP Longo LVG Ganiko LAlmeida IC Puccia R Vesicle and vesicle-free extracellular proteome ofParacoccidioides brasiliensis Comparative analysis with other pathogenicfungi J Proteome Res 2012 111676ndash1685

34 Bonin-Debs AL Boche I Gille H Brinkmann U Development of secretedproteins as biotherapeutic agents Expert Opin Biol Ther 2004 4551ndash558

35 Tjalsma H Antelmann H Jongbloed Proteomics of protein secretion byBacillus subtilis separating the ldquosecretsrdquo of the secretome Microbiol andMol Biol Rev 2004 68207ndash233

36 Weber SS Parente AFA Borges CL Parente JA Bailatildeo AM Soares CMAAnalysis of the secretomes of Paracoccidioides mycelia and yeast cellsPLoS ONE 2012 7e52470

37 Marchais V Kempf M Licznar P Lefranccedilois C Bouchara JP Robert R Cottin JDNA array analysis of Candida albicans gene expression in response toadherence to polystyrene FEMS Microbiol 2005 24525ndash32

38 Gonzaacutelez A Gomez BL Diez S Hernandez O Restrepo A Hamilton AJ CanoLE Purification and partial characterization of a Paracoccidioidesbrasiliensis protein with capacity to bind to extracellular matrix proteinsInfect Immun 2004 732486ndash2495

39 Barbosa MS Bao SN Andreotti PF De Faria FP Felipe MSS Feitosa LSMendes-Giannini MJS Soares CMA Glyceraldehyde-3-phosphatedehydrogenase of Paracoccidioides brasiliensis is a cell surface proteininvolved in fungal adhesion to extracellular matrix proteins andinteraction with cells Infect Immun 2006 74382ndash389

40 Mendes-Giannini MJS Hanna SA da Silva JL Andretti PF Vicentini LRBernard G Lenzi HL Soares CP Invasion of epithelial mammalian cells byParacoccidioides brasiliensis leads to cytoskeletal rearrangement andapoptosis of the host cell Microbes Infect 2004 6882ndash891

41 Castro NDS Barbosa MS Maia ZA Baacuteo SN Felipe MS Santana JM Mendes-Giannini MJS Pereira M Soares CMA Characterization of Paracoccidioidesbrasiliensis PbDfg5p a cell-wall protein implicated in filamentousgrowth Yeast 2008 25141ndash154

42 Pereira LA Bao SN Barbosa MS Silva JL Felipe MS Santana JM Mendes-Giannini MJS Soares CMA Analysis of the Paracoccidioides brasiliensis

de Oliveira et al BMC Microbiology 2013 13107 Page 14 of 14httpwwwbiomedcentralcom1471-218013107

triosephosphate isomerase suggests the potentialfor adhesin functionFEMS Yeast Res 2007 71381ndash1388

43 Donofrio FC Calil AC Miranda ET Almeida AM Benard G Soares CPNogueira SV Soares CMA Mendes-Giannini MJS Enolase fromParacoccidioides brasiliensis isolation and identification as fibronectin-binding protein J Med Microbiol 2009 58706ndash713

44 Coelho Neto J Agero U Oliveira DC Gazzinelli RT Mesquita ON Real-timemeasurements of membrane surface dynamics on macrophages and thephagocytosis of Leishmania parasites Exp Cell Res 2005 303207ndash217

45 Pereanez JA Goacutemez ID Patino AC Relationship between the structureand the enzymatic activity of crotoxin complex and its phospholipaseA2 subunit An in silico approach J Mol Graph and Model 2012 3536ndash42

46 Burger AM Seth AK The ubiquitin-mediated protein degradation pathwayin cancer therapeutic implications Eur J Cancer 2004 402217ndash2229

47 Jeferry CJ Mass spectrometry and the search for moonlighting proteinsMass Spectrom Rev 2005 24772ndash782

48 Borges CL Pereira M Felipe MSS Faria FP Gomez FJ Deepe GS SoaresCMA The antigenic and catalytically active formamidase ofParacoccidioides brasiliensis protein characterization cDNA and genecloning heterologous expression and functional analysis of therecombinant protein Microbes Infect 2005 766ndash77

49 Bradford MM A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dyebinding Anal Biochem 1976 72248ndash254

50 Cell Bank in Rio de Janeiro Brazil httpb200nceufrjbrbcrjindexphpoption=com_contentamptask=viewampid=10ampItemid=30

51 Borges CL Parente JA Barbosa MS Santana JM Baacuteo SN Sousa MV SoaresCMA Detection of a homotetrameric structure and protein-proteininteractions of Paracoccidioides brasiliensis formamidase lead to newfunctional insights FEMS Yeast Res 2010 10104ndash113

52 Breitkreutz BJ Stark C Tyers M Osprey a network visualization systemGenome Biol 2003 422

53 Saccharomyces Genome Database ndash SGD httpwwwyeastgenomeorg54 Structural genome databases of Paracoccidioides brasiliensis httpwww

broadinstituteorgannotationgenomeparacoccidioides_brasiliensis55 Bailatildeo AM Nogueira SV Bonfim SMRC Castro KP da Silva JF Mendes-

Giannini MJS Pereira M Soares CMA Comparative transcriptome analysisof Paracoccidioides brasiliensis during in vitro adhesion to type I collagenand fibronectin identification of potential adhesins Res Microbiol 2012163182ndash191

56 Batista WL Matsuo AL Ganiko L Barros TF Veiga TR Freymuumlller E Puccia RThe PbMDJ1 gene belongs to a conserved MDJ1LON locus inthermodimorphic pathogenic fungi and encodes a heat shock proteinthat localizes to both the mitochondria and cell wall of Paracoccidioidesbrasiliensis Eukaryot Cell 2006 5379ndash390

57 Lenzi HL Pelajo-Machado M Vale BS Panasco MS Microscopia deVarredura Laser Confocal Princiacutepios e Aplicaccedilotildees BiomeacutedicasNewslab 1996 1662ndash71

58 Eswar N John B Mirkovic N Fiser A Ilyin VA Pieper U Stuart AC Marti-Renom MA Madhusudhan MS Yerkovich B Tools for comparative proteinstructure modeling and analysis Nucleic Acids Res 2003 313375ndash3380

59 NIH-MBI laboratory servers httpnihservermbiuclaedu60 Colovos C Yeates TO Verification of protein structures patterns of

nonbonded atomic interactions Protein Sci 1993 21511ndash151961 Lovell SC Davis IW Arendall WB III Bakker PIW Word JM Prisant MG

Richardson JS Richardson DC Structure validation by Calpha geometryphi psi and Cbeta deviation Proteins Struct Funct Genet 2002 50437ndash450

62 Luthy R Bowie JU Eisenberg D Assessment of protein models withthree-dimensional profiles Nature 1992 35683ndash85

63 Kabsch W Sander C Dictionary of protein secondary structure patternrecognition of hydrogen-bonded and geometrical featureBiopolymers 1983 222577ndash2637

64 Helix System httphelixnihgov65 Okimoto N Futatsugi N Fuji H Suenaga A Morimoto G Yanai R Ohno Y

Narumi T Tai M High-performance drug discovery computationalscreening by combining docking and molecular dynamics simulationsPLoS Comput Biol 2009 5e1000528

66 Sakkiah S Thangapandian S Woo-Lee K Pharmacophore modelingmolecular docking and molecular dynamics simulation approaches foridentifying new lead compounds for inhibiting aldose reductase J MolModel 2012 22249ndash2747

67 Darden T York D Pederson L Particle mesh Ewald An Nsdotlog(N) methodfor Ewald sums in large systems J Chem Phys 1993 9810089ndash10092

68 Maiorov VN Crippen GM Size-independent comparison of proteinthree- dimensional structures Proteins Struct Funct Genet 199522273ndash283

69 Tovchigrechko A Vakser IA GRAMM-X public web server for protein-protein docking Nucleic Acids Res 2006 34310ndash314

70 Mashiach E Nussinov R Wolfson HJ FiberDock flexible induced-fitbackbone refinement in molecular docking Proteins 2009 781503ndash1519

doi1011861471-2180-13-107Cite this article as de Oliveira et al Intermolecular interactions of themalate synthase of Paracoccidioides spp BMC Microbiology 2013 13107

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit