1521-0111/89/3/376–387$25.00 http://dx.doi.org/10.1124/mol.115.100610MOLECULAR PHARMACOLOGY Mol Pharmacol 89:376–387, March 2016Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

Verbascoside Alleviates Pneumococcal Pneumonia by ReducingPneumolysin Oligomers s

Xiaoran Zhao, Hongen Li, Jianfeng Wang, Yan Guo, Bowen Liu, Xuming Deng,and Xiaodi NiuKey Laboratory of Zoonosis, Ministry of Education, College of Veterinary Medicine, Jilin University, Changchun, China (X.Z., H.L.,J.F., Y.G., B.L., X.D.); and Key Laboratory of Zoonosis, Ministry of Education, Department of Food Quality and Safety, JilinUniversity, Changchun, China (X.N.).

Received June 30, 2015; accepted December 18, 2015

ABSTRACTPneumolysin (PLY), an essential virulence factor of Streptococ-cus pneumoniae (pneumococcus), can penetrate the physicaldefenses of the host and possesses inflammatory properties.The vital role PLY plays in pneumococcus pathogenesis makesthis virulence factor one of the most promising targets for thetreatment of pneumococcal infection. Verbascoside (VBS) is anagent that does not exhibit bacteriostatic activity but has beenshown to inhibit PLY-mediated cytotoxicity. The results frommolecular dynamics simulations and mutational analysis

indicated that VBS binds to the cleft between domains 3 and 4of PLY, thereby blocking PLY’s oligomerization and counter-acting its hemolytic activity. Moreover, VBS can effectivelyalleviate PLY-mediated human alveolar epithelial (A549) cellinjury, and treatment with VBS provides significant protectionagainst lung damage and reduces mortality in a pneumococcalpneumonia murine model. Our results demonstrate that VBS is astrong candidate as a novel therapeutic in the treatment ofStreptococcus pneumoniae infection.

IntroductionStreptococcus pneumoniae (pneumococcus) infection

causes significant morbidity and mortality around the world.S. pneumoniae causes a variety of diseases, ranging from otitismedia to pneumonia, meningitis, and bacteremia, which havebecome increasingly difficult to cure. Drug-resistant strainsexhibit resistance against b-lactam antibiotics and even novelantibiotics, such as vancomycin (Campbell and Silberman,1998; Novak et al., 1999; Hidalgo et al., 2003). Pneumococcalpneumonia is a major cause of hospitalization. Pneumococcusis also one of the leading worldwide causes of deaths inchildren younger than 5 years, giving rise to an estimated820,000 deaths in 2000 (Black et al., 2010).In the articles of Paton et al. (1983, 1993), PLY, which is an

indispensable virulence factor generated in the streptococcuspneumonia clinic isolation process, is a 53-kDa protein thatbelongs to the family of cholesterol-binding cytolysins(Palmer, 2001). Unlike other cytolysins, this hemolysin islocated in the cytoplasm during the early period of bacterialgrowth and is released, due to growth and lysis, into theextracellular environment (Balachandran et al., 2001) where

it localizes to the cell wall (Price and Camilli, 2009). Once thisprotein has localized to the cell wall, it binds to cholesterol inthe host cell’s cytoplasmic membrane, infiltrates this mem-brane, and oligomerizes to form relatively large pores. Duringthis process, the targeted cell is lysed as these pores areformed. In this respect, oligomerization is a key aspect ofPLY’s mode of action. Similar to autolysin and surface proteinA, vaccines against which have been proven to provideprotection against S. pneumoniae infection (Musher et al.,2001; Wu et al., 2010), PLY is regarded as a pneumococcalprotein vaccine candidate. PLY’s interaction with the cells ofthe respiratory system is the likely cause of alveolar edemaand hemorrhage during the disease course. PLY also attenu-ates phagocyte and immune cell function, thus repressing hostinflammatory and immune responses (Boulnois et al., 1991).In 2001, Jedrzejas (2001) pointed out that one importantfactor of S. pneumonia boarding on its host is the activity,especially the activity at the initially infected phase. There-fore, the discovery of inhibitors that target PLY could opennovel therapeutic avenues for treating pneumonia infections.Verbascoside (VBS) is a phenylpropanoid glycoside, and

plants containing VBS are widely used in Chinese herbalmedicine (Pu et al., 2003). Previous studies have shown thatVBS has various pharmacological activities, including antiox-idant activity (Chiou et al., 2004), hepatoprotective activity(Xiong et al., 1998), and anti-inflammatory and antinocicep-tive activities (Schapoval et al., 1998). VBS’s ability to affect

Our study was supported by the National Basic Research Program of China[Grant 2013CB127205]; the National Nature Science Foundation of China[Grant 31130053]; and the National 863 program [Grant 2012AA020303].

dx.doi.org/10.1124/mol.115.100610.s This article has supplemental material available at molpharm.

aspetjournals.org.

ABBREVIATIONS: A549, human alveolar epithelial; BAL, bronchoalveolar lavages; CDC, cholesterol-dependent cytolysin; CHO, cholesterol; IL-1b,interleukin 1b; LDH, lactate dehydrogenase; MD, molecular dynamics; MIC, minimal inhibitory concentration; PBS, phosphate-buffered saline; PCA,principal component analysis; PLY, pneumolysin; RMSF, root mean square fluctuation; rPLY, recombinant PLY; strain PLN, the isogenicpneumolysin-deficient S. pneumoniae; TNF-a, tumor necrosis factor a; VBS, verbascoside.

376

http://molpharm.aspetjournals.org/content/suppl/2015/12/23/mol.115.100610.DC1Supplemental material to this article can be found at:

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

hemolytic activity was identified via drug screening. Ourpresent study showed that VBS, a natural product with noanti-S. pneumoniae activity, can lower the virulence ofS. pneumoniae in vivo and in vitro by influencing the lyticactivity of PLY, which indicates that VBS may interactdirectly with PLY. To determine the inhibition mechanism,standard molecular dynamics simulation, binding energyprofiles, protein mutants, and principal component analysiswere performed for the PLY-VBS complex. The results in-dicated that VBS can bind to the cleft between domains 3 and 4of PLY, thereby blocking its oligomerization and inhibiting itslytic activity.The results of this study may be useful for drug design

against pneumococcal infection. Indeed, an antivirulence ther-apeutic strategy may be beneficial as an adjuvant to antibac-terial therapy, because it is unlikely to accelerate antibacterialresistance (Wang et al., 2015).

Materials and MethodsBacterial Strains and Culture. S. pneumoniae strain D39

(NCTC 7466) and strain PLN (an isogenic pneumolysin-deficientS. pneumoniae) were selected for this study (Berry et al., 1989). TheS. pneumoniae strains were cultured at 37°C using Todd-Hewittbroth. Calculation of minimal inhibitory concentration (MIC) of VBSfor S. pneumonia is available when broth microdilution method isadopted and upon confirmation of the stipulations of the Clinical andLaboratory Standards Institute (CLSI). The minimal inhibitoryconcentrations (MICs) of VBS for S. pneumoniae were identifiedusing the broth microdilution method based on the Clinical andLaboratory Standards Institute (CLSI) guidelines.

Chemicals. VBS (purity .98.8%) was purchased from ChengduHerbpurify Co., Ltd. (Chengdu, Sichuan, China). VBSwas dissolved atvarious concentrations in sterile phosphate-buffered saline (PBS). TheVBS solution was filter sterilized through a 0.22-mm microfiltrationmembrane.

Construction, Expression, and Purification of PLY. Primerswere designed to amplify the S. pneumoniae gene encoding ply usingS. pneumoniae strain D39 as the template. The expressions of theprimers are listed in the following:

Forward 59-CGCGGATCCGCGATGGCAAATAAAGCAGTAAA-39 as well as the reverse 59-CCGCTCGAGCGGCTAGTCATTTTC-TACCTTAT-39 (restriction enzyme sites are underlined). BamHIand XholI enzymes were used to digest the amplified ply gene, andthe digested genes were then cloned into a pET28a prokaryoticexpression vector. To overexpress the recombinant protein, thepET28a-PLY vector was transformed into Escherichia coli BL21(DE3).

The cells were cultured until absorbance at 600 nm reached 0.6–0.8at 37°C, induced with 0.3 mM IPTG, and harvested after growing foran additional 12 hours at 16°C. The harvested cells were resuspendedin sterile PBS and crushed by a high-pressure homogenizer(EmulsiFlex-C3, AVESTIN, Ottawa, Ontario, Canada). The cell lysatewas centrifuged at 12,000 g for 30 minutes, and the supernatant wasloaded onto a Ni-NTA agarose column. The recombinant protein wasbound to a Nickel-affinity chromatography column, and the nickelcolumn was then flushed with washing buffer consisting of 20 mMTris, 20 mM imidazole, and 300 mM NaCl (pH 8.0) The His-taggedprotein was eluted in elution buffer containing 20 mM Tris, 300 mMimidazole, and 300 mM NaCl (pH 8.0). The eluted solution wasconcentrated using a Millipore Amicon filter (30 kDa molecularweight cutoff; Bedford, MA) for desalting. The purified product wasidentified by sodium dodecyl sulfate polyacrylamide gel electropho-resis (SDS-PAGE).

Hemolytic Test. In 2010, Ragle and other researchers said thatthe assessment method of the hemolytic activities is as same as that

used in other places (Ragle et al., 2010). Briefly, 10 ml of purified PLYwas preincubated in a 96-well plate with a series of differentconcentrations of VBS in PBS at 37°C for 10 minutes. Defibrinatedsheep blood erythrocytes (50 ml at 5 � 106 cells/ml) were added to thewells, and the final volume of the wells was then increased to 200 mlwith PBS. Then, the samples were subject to incubation at thetemperature of 37°C for 10 minutes. After centrifugation at 3000 gfor 5 minutes, the supernatants were removed into a cuvette tomeasure their absorption at 543 nm.

Immunoblot Analysis. S. pneumoniae were cultured at 37°C inTodd-Hewitt brothwith various concentrations of VBS until the OD600

reached approximately 1.0. Equal volumes were boiled in Laemmlisample buffer. After being loaded into a 12% SDS-PAGE gel, thecentrifuged supernatant samples were delivered to the polyvinylidenefluoride membranes through a half-dry carrier cell. The polyvinyli-dene fluoride membrane was then cut and sealed off for 2 hours with5% skim milk at room temperature. The primary PLY antibody wasadded at a 1:1000 dilution and incubated overnight at 4°C. Horserad-ish peroxidase-conjugated anti-mouse antiserum (secondary anti-body) was diluted to 1:2000. The testing reagent for Amersham(Pittsburgh, PA) ECL immunoblotting was applied to develop theblots.

Site-Directed Mutagenesis. Three amino-acid mutations(S254A, W278A, L447A) were created using a QuikChange site-directed mutagenesis kit. The template plasmid was the pET28a-PLY plasmid noted above. The primers used to mutate the relevantamino acids are shown in Table 1.

Oligomerization Analysis. rPLY (5 nM at 10 mg/ml) was mixedwith 10 nMVBS or 10 nM cholesterol liposome (CHO), 5 nM of rPLY atthe same volume with neither agent was used as a control, and CHO(10 nM):VBS (20 nM) 5 1:2 mixed with rPLY was used to attest theinterference of VBS. After incubating at 37°C for 1 hour, the mixtureand control sample were subjected to high-performance liquid chro-matography (DGU-20A5, Shimadzu Corporation, Kyoto, Japan) usingNanofilm SEC-250 (Sepax Technologies, Inc., Newark, DE), wherein0.5 ml the Nanofilm SEC-250 was filled per minute.

Live/Dead and Cytotoxic Testing. The experimental cells andthe alveolar epidermal cells, which were bought from ATCC (Mana-ssas, VA), were cultured in the Dulbecco’s modified Eagle medium, anexpression vector according to Chemical Abstracts of the UnitedStates, and the content of fetal calf serum in the Dulbecco’s modifiedEagle medium is 10%. The cells were delivered to a 96-well plate,wherein each well can deliver 1.5 � 104 cells. The A549 cell wassubjected to isolated incubation under assistance of S254A, WT-PLY,W278A, and L447A (0.5 ml, at the same concentration) and increasingconcentrations of VBS. These test samples as well as positive control(PBS) were placed in a temperature-controlled box for 6 hours at 37°C.The LDH, namely the Cytotoxic Detection Kit (Roche Mannheim.Germany), can be adopted to measure releasing capacity of LDH,and the cellular activity could be determined. Live/dead assays wereperformed according to the methods recommended by the manufac-turer (Invitrogen, Carlsbad, CA). LDH activity was measured on amicroplate reader (TECAN, Salzburg, Austria). Microscopic images ofstained cells were captured using a confocal laser scanningmicroscope(Olympus, Tokyo, Japan). The results are representative of a mini-mum of three independent experiments.

Pharmacokinetics Study. Employing a protocol based on pre-vious pharmacokinetic research methods (Li et al., 2014), analyteswere separated on an Agilent Zorbax C18 column (Santa Clara, CA).As to the mobile phase, it contained solvent A (methyl alcohol) andsolvent B (the content of methane acid in ddH2O being 0.1%). The flowrate was 0.8 ml/min. The injection volume was 10 ml, the detectionwavelength was at 330 nm; meanwhile the blood temperature waskept at 25°C. Mouse blood samples were collected from veins intoheparinized tubes at 5, 15, 30, 45, 60, 120, 240, and 360 minutes afterdosing. Blank blood was collected before dosing.

Murine Model of Endonasal Pulmonary Infection. The fe-male C57BL/6Jmice, which have grown for 8 weeks andweigh 206 2 g,

Verbascoside Inhibits the Oligomerization of Pneumolysin 377

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

were bought in the Experimental Animal Centre of Jilin Universitylocated inChangchunCity, JilinProvince.Themice used for tests restedfor 1 week to acclimatize before use. The whole processes gainedauthorization and were implemented according to the stipulations ofthe ACUC (Animal Care and Use Committee) affiliated to the JilinUniversity.

The experimental methods used for inducing pneumonia in micewere presented previously (Dessing et al., 2008). Themicewere lightlyanesthetized by inhalation of isoflurane and then inoculated with20 ml of suspension containing 5 � 107 colony-forming units (cfu) ofstrain D39 or strain PLN in the left nare. To test the effect of VBStreatment, several injections of 100 ml of VBS were subcutaneouslyadministered to the mice 2 hours after infection with strain D39.Measurements were taken every 3 hours for a period of 120 hours.About 100 ml of aseptic poly butylenes succinate was filled in each ofthe mice under control.

Bronchoalveolar Lavages. Bronchoalveolar lavage (BAL) fluidcollections were administered twice by intratracheal instillationof 500 ml of sterile PBS. The clear surface liquid obtained aftercentrifugation was collected, and the lavage fluid supernatants wereused for cytokine detection (Qiu et al., 2012).

Measurement of Bacterial Loads. The bacterial numbers wereidentified by colony counts of lung tissue smears as previouslydescribed (Dessing et al., 2007). Lungs were collected from euthanizedmice, and lung tissue homogenates were prepared in 1 ml of sterilePBS at 4°C and used to calculate the bacterial colony counts throughthe serial dilution method and smearing on solid media.

Homology Modeling Study. The monomeric 3D structure of thepneumolysin (PLY) has not yet been reported. Therefore, a homologysimulation study of PLY was performed using the structure ofPerfringolysin O as the template (Protein Data Bank code 1M3I). TheAlign Sequence to Templates tool in Discovery Studio2.5 (DiscoveryStudio, BIOVIA, San Diego, CA) was used to carry out the sequencealignment between PLY and the template (Laskowski et al., 1993).The 3D structure of PLY was then constructed using the BuildHomology Models protocol implemented in Discovery Studio 2.5.Subsequently, to obtain the equilibrium structure, a 500-ns moleculardynamics simulation was performed with the 3D structure of PLYusing the Gromacs 4.5.5 software package (Hess et al., 2008). Toachieve docking aswell as dynamic analogy ofmolecules, theGaussian09 programwas used to optimize the structure of VBS at theB3LYP/6-31G* level (Frisch et al., 2009).

Molecular Docking. In this work, verbascoside (VBS), consideredto be the flexible ligand, was docked into PLY using the dockingprogram AutoDock 4.0 (AutoDock program, The Scripps ResearchInstitute, La Jolla, CA) (Morris et al., 1996, 2009; Hu et al., 2009). Thedetailed docking process docking was reported previously (Dong et al.,2013). The docking model of VBS with PLY has been provided as theSupplemental Material.

Molecular Dynamics Simulation. A molecular dynamics simu-lation of the complex of PLY with VBS was carried out using theGromacs 4.5.5 package to explore the binding mode of complex. TheAmberff99sb force field and TIP3Pwater model were applied (Ryckaertet al., 1977; Jorgensen et al., 1983). The specific parameters of the

molecular dynamics (MD) simulation were consistent with those de-scribed previously. The force field parameters of VBS were estimatedbased on one portion of AM1-REST atom charges, which belongs to theprogram packages of Amber (Wang et al., 2006).

Calculation of the Binding Free Energy. The Molecularmechanics Poisson-Boltzmann surface area (MM-PBSA) approach(Swanson et al., 2004; Jogalekar et al., 2010; Rungrotmongkol et al.,2010; Vorontsov and Miyashita, 2011) supplied by the Amber 10package was used to calculate the binding free energy between theprotein and ligand. The detailed calculation was based on previousreports in the literature (Dong et al., 2013; Niu et al., 2013).

Principal Component Analysis. Principal component analysis(PCA) is the simplest of the multivariate techniques that are used toreduce or simplify large, complex datasets (Zhou et al., 2001; Barrettand Noble, 2005; Liu et al., 2008). In this work, to define the dominantmotion over an MD simulation, the collective motions of the complexwere addressed by using the positional covariance matrix, Ca, of theatomic coordinates and its eigenvectors based on PCA. The Gromacs4.5.5 module was used to perform PCA, and the trajectories wereobtained from the previous MD simulations. The detailed process ofprincipal component analysis was based on previous reports in theliterature (Qiu et al., 2013).

Fluorescence-quenching Assay. Verbascoside to the bindingsites of wild-type and mutant PLY, namely the binding constants (KA)could be calculated in a fluorimetric quenching way. The bindingenergy of the complexes were converted from the binding constantsusing DGbind 5 RTlnKA (Qiu et al., 2012; Wang et al., 2015).

Statistics. The 13.0 version Statistic Package for Social Science(SPSS Inc., Chicago, IL) was used to analyze the empirical data.Fisher type precision experiments were used to calculate the mean-ings of the mortality research statistically. Differences were consid-ered statistically significant when P , 0.05.

ResultsInhibition of PLY-Induced Hemolytic Activity by

Verbascoside. Previous studies have shown that someextracts from traditional Chinese medicine have bacterio-static activity against S. pneumoniae. Verbascoside (Fig.1A) is a compound that occurs relatively widely in tradi-tional Chinese medicine and has various pharmacologicalactivities. Studies examining the MIC found that even aconcentration of VBS as high as 2048 mg/ml was still notcapable of impeding the growth of the bacteria, indicatingthat VBS exhibits no antimicrobial activity against S.pneumoniae. However, VBS was found to have a negativeinfluence on the hemolytic activities of PLY on the basis ofdifferent concentrations via drug screening (Fig. 1, B and C).Furthermore, VBS does not affect the production of PLY in S.pneumoniae (Fig. 1D). The above results implied that VBSmay directly interact with PLY.

TABLE 1Primers used for site-directed mutagenesis

Primer SequenceAminoAcid

Change

1fwd 59 CTCAAGTTGGAAACCACGGCGAAGCGTGATGAAGTAGAGG 39 S254A1rev 59 CCTCTACTTCATCACTCTTCGCCGTGGTTTCCAACTTGAG 392fwd 59 GCTCCTCAGACAGAGGCGAAGCAGATTTTGG 39 W278A2rev 59 CCAAAATCTGCTTCGCCTCTGTCTGAGGAGC 393fwd 59 AACCGATTTGCCAGCGGTGCGTAAGCGGAC 39 L447A3rev 59 GTCCGCTTACGCACCGCTGGCAAATCGGTT 39

fwd, forward; rev, reverse.

378 Zhao et al.

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

Inhibition of PLY Oligomerization by VBS. To findwhether VBS weakens the hemolytic activity of PLY byinhibiting its oligomerization, we used size exclusion chroma-tography to separate PLY monomers and oligomers andmonitored them by high-performance liquid chromatography.We already know that PLY in a solution that lacks cholesterolor membranes can self-associate to form oligomers frommonomers (Gilbert et al., 1998). We mixed 10 nM of VBS intoan rPLY sample and incubated themixture at 37°C for 1 hour.Another rPLY sample without VBS was used as the control.Peak 1 represents PLY oligomers, whereas Peak 2 and Peak 3represent PLY monomers. The rPLY incubated with VBScould not form oligomers and exhibited only a monomer peak(Peak 3) (Fig. 1E); thus, the PLY in this sample was incapable

of self-association. Cholesterol (CHO) is the natural ligand ofPLY and it induces oligomerization of the protein. We mixed10 nM CHO into an rPLY sample, and mixed CHO (10 nM):VBS (20 nM)with rPLY to attest the interference of VBS. Peak4 represents PLY oligomers, Peak 5 and Peak 6 represent PLYmonomers (Fig. 1F). The results indicate that VBS coulddetectably interfere with CHO-induced Ply oligomeriza-tion. From what has been shown above, we could safelydraw the conclusion that VBS molecule could inhibit PLYoligomerization.Identification of the Binding Mode of PLY with

VBS. To obtain the stable structure of PLY in complex withVBS, standard MD simulations were performed for thecomplexes, and the root-mean-square deviations of backbone

Fig. 1. Suppressing effect of verbascoside (VBS) on hemolysis caused by pneumolysin (PLY). (A) Chemical structural formula of VBS. (B and C) Asdetermined through hemolysis tests using purified rPLY and sheep blood erythrocytes in PBS and measuring the absorbance values of the centrifugalsupernatants at 543 nm, the addition of VBS decreased the hemolysis rates. The column diagrams stand for the average values of the assays (n = 3). Theerror bars stand for standard errors. **When contrasting to the matched group, P is smaller than 0.01. (D) Western blot analysis of PLY expression.Bacteria lysate of S. pneumoniae D39 cultured in increasing concentrations of VBS. VBS is not able to influence PLY from being generated inS. pneumoniae. (E) Inhibition of PLY oligomerization byVBS. Purified rPLY that was able to self-associate was incubatedwith or without VBS at 37°C for1 hour. Peak 1 presents PLY oligomers. Peak 2 and Peak 3 present PLY monomers. The results indicate that VBS interacts with PLY and makes thecytotoxin incapable of self-association. F, Peak 4 represents PLY oligomers, Peak 5 and Peak 6 represent PLY monomers, indicating that VBS coulddetectably interfere with CHO-induced Ply oligomerization.

Verbascoside Inhibits the Oligomerization of Pneumolysin 379

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

Ca atoms were used provide the stability and conformationaldrift of PLY in the simulation. As shown in Fig. 2A, the proteinin the complex system could reach the plateau at 20 ns ata value of ∼0.35 nm with small fluctuations at the value of∼0.1 nm.However, the protein in the free protein system couldalso reach the plateau at 20 ns with the bigger fluctuationsaround the value of∼0.2 nmdue to the binding of PLYwithVBS.The stable structure of PLY with VBS was obtained via

the 200-ns MD simulation, as shown in Fig. 3. VBS could beprecisely embedded into the cleft between domains 3 and 4 inPLY through the H-bonding as well as hydrophobic interac-tions. The hydroxyl group of the benzene ring on the right sideof VBS can form a strong hydrogen bond with Asp471, whichhas a crucial significance in making the right side of VBS to bestable. Moreover, two .O atoms in the central section of VBS

can form three hydrogen bonds with Asn470, and the hydroxylgroup can form a hydrogen bond with Glu277, indicating thatthe central section of VBS can be anchored by Asn470 andGlu277. The quantity of H-bonds that were found betweenPLY and VBS during the last 160-ns simulation was displayedin Fig. 2B, fluctuating from 4 to 6 within the simulation time,which is consistent with the above results. To ensure theeffectiveness of hydrogen bonds between VBS and PLY, thestability of the hydrogen bonding was calculated in Table 2,which confirms that the hydrogen bonds can exist in a stablemanner. In addition, the plane of the benzene ring on the leftside of VBS and the plane of benzene rings of the remainedTyr358 are level to each other. Thus, there is likely a strongp-p interaction between this residue and VBS, stabilizing theleft side of VBS in complex with PLY.To further validate the binding sites of PLYwith VBS and to

study the flexible property of the remained Tyr358, the rootmean square fluctuation (RMSF) value, namely the root meansquare fluctuation value of the remained Tyr358 thatenclosed the bonding positions of PLY was computed. Asshown in Fig. 2C, the flexibilities of the residues surroundingthe binding sites in the free protein and complex are different.The residues (250–471) of PLY that bind VBS show weakerflexibility when bound to VBS, with a RMSF value of less than0.40 nm, compared with the corresponding residues in the freeprotein. This finding indicates that these residues are morerigid due to binding with VBS. Given the above information,the stabilization at the binding site of PLY with VBS waslargely due to residues Asp471, Asn470, Glu277, Tyr358, andArg359, as shown in Fig. 3.Binding Energy Calculation of VBS to PLY. The

initial binding mode of PLY with VBS could be identifiedby the 3D structure based on the MD simulation. However,the information obtained for the binding site residues in thecomplex is not sufficient. To address this problem, theimportance of remained binding sites for the bonding energybetween PLY and verbascoside could be assessed in an MM-PBSA way, and the sums of the bonding energy of everyremained binding spot could be partitioned into the electro-static (DEele), the salvation (DEsol), the Van der Waals (DEvdw)and the total contribution (DEtotal), as shown in Fig. 4. Asshown in Fig. 4A, although the value of Glu277 is smaller than22.0 kcal/mol, its electrostatic (DEele) contribution is great,indicating a strong electrostatic interaction between PLY andthe middle part of VBS. However, because of the unfavorablecontribution of solvation (DEsol), with a value of ∼2.0 kcal/mol,the total contribution (DEtotal) of Glu277 is low, with a valueof ∼20.8 kcal/mol. Consistent with the results of the aboveanalysis, Asn470 and Asp471 have strong electrostatic inter-actions with VBS, with values of ∼22.31 and 20.96 kcal/mol,respectively, because of the formation of hydrogen bonds.Moreover, as shown in Fig. 4A, Tyr358 and Arg359 havestrong Van der Waals terms of ∼22.0 and ∼2.3 kcal/mol(DEtotal), respectively. In summary, this analysis confirms thatthe key residues of the binding site are Asp471, Asn470,Glu277, Tyr358, and Arg359.To confirm the binding site in the PLY-VBS complex,

similar MD simulations were performed for the complexescontaining S254A-PLY, W278A-PLY, and L447A-PLY mu-tants bound to VBS, and then the binding free energies of thethree complexes were calculated using theMM-PBSAmethod.Then, a method of fluorescence spectrum quenching was used

Fig. 2. The dynamic properties of PLY with VBS complex. (A) The rootmean square deviation values of PLY with VBS complex (black line) andunliganded PLY (red line) via simulation times; (B) the number of H-bondsin the complex system fluctuates within the simulation time; (C) the rootmean square fluctuation (RMSF) of the residues of PLY in the complex andunliganded protein was calculated during the last 100-ns simulation.

380 Zhao et al.

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

to calculate the binding free energies of VBS with the threemutants. As shown in Fig. 4, B, C, and D, the key residues inthe binding site of complexes are similar to those ofWT-PLY incomplex with VBS, except for the mutant residues. The totalbinding free energy of WT-PLY, S254A-PLY, W278A-PLY,and L447A-PLY in complex with VBS and their energycontributions are summarized in Table 3. It can be seen fromthe calculation of the binding free energy of the complex thatin comparison with the WT-PLY/VBS complex, the value ofthe binding free energy was reduced by about 4–5 kcal/mol.According to the results based on the fluorescence spectros-copy quenching method, the binding free energy between VBSand PLY decreases in the following order: WT . S254A-PLY . W278A-PLY . L447A-PLY, well consisting with theresults of theMD simulation. These analyses provide furtherevidence that the binding of PLY to VBS is due to theresidues Asp471, Asn470, Glu277, Tyr358, and Arg359.Principal Component Analysis of the Motion for PLY

from the Complex. According to previous reports, thehemolytic activity of PLY can be achieved by monomericoligomerization resulting from a conformational change ofthe PLYmonomer.MD simulation and hemolysis experimentsshow that as VBS directly interacts with protein, it has anegative effect in the hemolytic activities of PLY. Therefore,this paper discusses themost importantmotions of PLYwith aligand or without it so as to seek the inhibitory mechanism ofVBS by principal component analysis (PCA) on the basis ofthe molecular dynamics tracks of PLY-VBS complexes andunbound PLY. As shown in Fig. 5, there is an extendedmotionbetween domains 3 and 4, which has an overall influence onthe structure of the isolated protein in a first element (PC1),wherein the dash line in Fig. 5 stands for PC1. In addition, the

extendedmotion is sufficiently large that PLY can successfullychange from the monomer structure to the oligomeric struc-ture. A slight vibration of the backbone of the protein wasshown in the second principal component (PC2), as shown inFig. 5. Interestingly, VBS binds in the cleft between domains 3and 4 of PLY based on the MD simulation, indicating that themotion of domains 3 and 4 can be influenced by the binding ofVBS. As expected, the extended motion between domains 3and 4 is clearly weaker in the PLY/VBS complex comparedwith that of the unliganded PLY, as shown in Fig. 5. Thus,this analysis confirmed that the motion of the conformationtransition for PLY from the monomeric to the oligomeric formis restricted by the binding of VBS with PLY.Moreover, to further confirm the above mechanism, the

distances between domains 3 and 4 were calculated in PLYwith VBS and without VBS system, as shown in Fig. 6. Theaverage distances for the PLY/VBS complex and free PLYwere 4.25 and 4.50 nm, respectively. These results indicatethat the conformation of domains 3 and 4 in PLY wasrestrained due to the binding of VBS, which is consistent withthe above results.In summary, based on these findings, the inhibition mech-

anism can be hypothesized: the conformational change fromthemonomeric to oligomeric form is blocked due to the bindingof VBS to the cleft between domains 3 and 4 in PLY, leading toa decrease in the lytic activity of PLY.VBS Hinders PLY-Mediated Human Alveolar Epithe-

lial (A549) Cell Damage. It has long been known that PLY,a pore-forming cytotoxin of S. pneumoniae, directly contrib-utes to alveolar epithelial and pulmonary endothelial injuries(Jedrzejas, 2001). In the past, how S. pneumoniae injures lungcells was determined under assistance of the lung alveolarepithelium cell of people, represented by A549 (Schmeck et al.,2004), and PLY has been found to be the major factor involvedin their apoptosis and necrosis (Jedrzejas, 2001). As shown inFig. 7, it can be seen that the intact A549 cells were detectedretained a green fluorophore (Fig. 7A), whereas A549 alveolarepithelial cells, which were incubated together with PLY,contained the red fluorophore, indicating their dead status(Fig. 7B). In contrast, treatment with 8 mg/ml VBS protectedA549 cells from death (Fig. 7C). The addition of 32 mg/ml VBSprovided nearly complete protection to A549 cells (Fig. 7D).Moreover, a LDH release assay was performed to evaluate the

Fig. 3. The binding mode of VBS with PLY on thebasis of MD simulation. During the standard 200-nsMD simulation, VBS could strongly bind at the cleftbetween domain 3 and domain 4 in PLY. ResiduesAsp471, Asn470, Glu277, Tyr358, and Arg359 playkey roles in VBS binding with PLY.

TABLE 2The hydrogen bonds of PLY with VBS based on the MD simulations

Acceptor Donor Presence Distance

% nm

Glu277 . O VBS O-H 76.3 0.24 6 0.1Asn470 = O VBS O-H 81.6 0.31 6 0.3VBS . O Asn470 N-H 77.1 0.28 6 0.2VBS . O 85.6 0.33 6 0.5Asp471 = O VBS O-H 72.9 0.34 6 0.3

Verbascoside Inhibits the Oligomerization of Pneumolysin 381

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

effect of VBS on the PLY-mediated lysis of A549 cells. Asexpected, the addition of 2 to 32 mg/ml of VBS to samplesreduced the cytotoxicity of WT-PLY in a dose-dependentmanner (Fig. 7E). We created PLY variants with amino acidmutations at Ser254, Trp278, and Leu447; purified thesemutants; and examined their effects in LDH release assays.The mutants did not have significantly altered cytotoxicityrelative to WT-PLY. However, VBS did not produce conspic-uous cytoprotective effects when samples containing mutant

PLY were treated with various concentrations of VBS (Fig. 7,F-H). These dramatic results confirm our hypothesis that VBSexhibits a potential therapeutic effect that merits furtherinvestigation and indicate that as we predicted, mutations inthe aforementioned residues impair the effectiveness of VBS.VBS Protects Mice from S. pneumoniae Pneumonia.

Based on the above-presented experimental data, we sought toanalyze whether similar protection would occur in vivo inmouse affected with pneumonia caused byS. pneumoniae. The

Fig. 4. The binding free energy contribution of residues in the PLY binding sites. The binding energies decomposition based on a per-residue at thebinding sites among WT-PLY (A), S254A-PLY (B), W278A-PLY (C), and L447A-PLY (D) and VBS. The bar graph illustrates the van der Waals (white),electrostatic (blue), salvation (yellow), and total (black) contributions for the complexes.

382 Zhao et al.

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

first step was to analyze the pharmacokinetic features of VBSin mice. After subcutaneous injection of a single VBS dose of100 mg/kg, we detected the drug concentrations in the mice

plasma. The maximum concentration (Cmax) of VBS in plasmawas 112.22 mg/ml 30 minutes after administration (Fig. 8A),suggesting that VBS effectively enters the systemiccirculation.Mortality due to pneumonia caused by S. pneumoniae D39

was monitored over a 120-hour time course. Approximately75% of C57BL/6J mice infected with S. pneumoniaewere deadwithin 120 hours. Through a test where the mice were treatedwith 100 mg/kg VBS, it can be seen that an obvious protectionplayed a role, with a mortality rate 120 hours after infection ofonly 25%. As expected, none of the mice infected with strainPLN (the isogenic pneumolysin-deficient S. pneumoniae) diedwithin the time course (Fig. 8B). Obviously, the survival timeof the VBS-treated group was longer than that of the un-treated group (P , 0.01). Additionally, the bacterial count inthe lungs was significantly lower in the VBS-treated groupthan in the PBS treated group (P, 0.01), indicating that VBSinfluences S. pneumoniae survival within the lungs (Fig. 8C).Through themacroscopic inspection, it can be seen that the

untreated mice had dark red lungs that exhibited severecongestion, whereas the lung tissue of mice treated withVBS and infected with the PLN strain was pink and fungous(Fig. 8D). Our examination of the pathologic manifestationsrevealed that the S. pneumoniae-infected mice in the PBSgroup exhibited severe tissue injury and an accumulation ofinflammatory cells in the alveolar space. In contrast, thepathologic tissue sections of the mice in the treatment groupwere similar to those of the mice infected with the PLNstrain. The histopathological features corresponded to arelief of pulmonary inflammation, as suggested by areduction of inflammatory exudates (Fig. 8E). Furthermore,we analyzed the levels of various cytokines, including tumornecrosis factor a (TNF-a) and interleukin 1b (IL-1b), in thebronchoalveolar lavage (BAL) fluid of the experimental mice.Consistent with our previous observations, the concentra-tions of TNF-a and IL-1b were significantly lower in VBS-treated mice than those observed in the control mice (P ,0.05) (Fig. 8, F and G). Taken as a whole, these resultsdemonstrated that VBS treats S. pneumoniae pneumonia in amurine model of infection.

DiscussionConventional antibiotics are always recommended as

the primary therapy for S. pneumoniae-infected patients.

Fig. 5. Principal component analysis of the PLY motion. The first andsecond principal components (PC1 andPC2) in free protein (A) and the firstand second principal components (PC1 and PC2) in complex (B) obtainedby PCA are depicted by cones on the alpha carbon atoms. The length of thecones represents the magnitude of the motion. The dotted line rangerepresents the binding region of VBS with PLY.

TABLE 3The binding free energy of VBS binding with WT-PLY and mutants based on the calculations, and thebinding constants (KA) (1 � 105) l×mol21 of the PLY-VBS complexes systems gained from the fluorescence-quenching method

Energy components (kcal/mol) WT-PLY S254A-PLY W278A-PLY L447A-PLY

DEele 245.35 6 8.71 236.06 6 6.85 239.58 6 9.26 242.38 6 9.93DEvdw 243.61 6 3.06 238.08 6 2.65 238.43 6 4.03 242.51 6 3.42DEMM 288.96 6 8.75 274.14 6 6.69 278.02 6 7.26 284.89 6 9.64DGele,sol 20.00 6 0.00 20.00 6 0.00 20.00 6 0.00 20.00 6 0.00DGnonpolar,sol 67.15 6 7.63 57.95 6 5.31 59.73 6 5.85 67.50 6 7.93DGsol 74.93 6 7.75 64.51 6 5.35 62.80 6 5.98 79.93 6 7.92DGele,sol + DEele 245.35 6 6.50 236.06 6 2.81 239.58 6 5.69 242.38 6 6.94DGnonpolar +DEvdw 23.54 6 1.23 19.87 6 1.08 21.30 6 2.31 24.99 6 3.54DGtotal 221.81 6 4.28 216.19 6 3.86 218.29 6 3.25 217.39 6 3.67-TDS 5.62 6 1.43 5.22 6 1.81 6.05 6 2.09 5.99 6 2.31DGbind 216.19 6 2.03 210.97 6 2.36 212.24 6 1.98 211.40 6 2.33KA 4.57 3.54 3.67 4.01

Verbascoside Inhibits the Oligomerization of Pneumolysin 383

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

Although these traditional approaches are effective, they havegenerated evolutionary stress on the target bacterium, whichhas thus been selected for resistant subpopulations (Werneret al., 2008). The emergence of antibiotic resistance and eventhe appearance of multidrug-resistant S. pneumoniae havemade infectious diseases more difficult to cure and haveincreased the mortality rate (Thummeepak et al., 2015).Increasing understanding of and research into antivirulencehas been pursued and has revealed a potential strategy todevelop novel drugs to treat bacteria-mediated disease(Rasko and Sperandio, 2010). Instead of applying conven-tional antibiotics to exterminate microbes, this approachmight present a moderate selection pressure because itweakens the pathogen’s virulence and does not increasesurvival pressure.Certain proteins trigger a pathogenic mechanism and

participate in the disease progress induced by these organ-isms. The usability of PLY, one of these proteins, has beenproposed (Jedrzejas, 2001). Here, we provide evidence thatVBS has no antimicrobial activity against S. pneumoniae,but was found to antagonize formation of PLY’s oligomeri-zation, thus weakening the hemolytic activity of PLY in aconcentration-dependent way. Additionally, VBS hinderedinjury of pulmonary epithelial cells by PLY in an LDHrelease assay. PLY mutants produce stable cytotoxic effectsthat cannot be weakened by VBS, indicating that themutated residues affect binding between PLY and VBSand thereby providing evidence for our molecular modelingresults. Moreover, the agent obviously exerts protection inmouse model of pulmonary infection by alleviating lungpathologic damage and reducing mortality. In addition, VBSexhibited anti-inflammatory properties in VBS-treatedmice, and this biologic effect may also play a role in alleviatinglung injury.For the protein with its ligand system, the mechanism of

interaction can be explored by the dynamics analysis.However, this complex is not a receptor-ligand complexin the strict sense, as we cannot calculate the dissociationconstant (KD) using a surface plasmon resonance mea-surement. Therefore, MD simulation is an appropriateapproach to study the interactions of proteins and theirligands.

In this paper, standard MD simulations were performed toexplore the interaction of PLY with VBS. On the basis of theMD simulation and the un-bond free energies, the bondingmethods of VBS with PLY could be understood more deeply.Interestingly, VBS could bind to the cleft between domains 3and 4 in PLY by making strong contact with Asp471, Asn470,Glu277, Tyr358, andArg359. The results of the binding sites ofVBS with PLY were confirmed by ligand-residue interactiondecomposition using the MM-PBSA method, point mutationsof residues, and a fluorescence-quenching assay. According tothe above results, the results of the theoretical calculationsare in good agreement with those of the experiments, and wecan conclude that the binding mode of VBS with PLY can becalculated by MD simulation. The mechanism analysis of theinhibition activity depends on the dynamic analysis of theprotein in the PLY/VBS complex. To our knowledge, PCA isthe best tool for exploring the dynamic characteristics of aprotein in a complex or unliganded state. Thus, based on ouranalysis of the dynamic trajectory, we can predict that thebinding of VBS to PLY can cause a conformational change ofdomains 3 and 4 of PLY, directly leading to a decrease in thedistance between domains 3 and 4. In free PLY, extendedmotion between domains 3 and 4 can meet the requirement ofthe conformational transition for PLY from the monomeric tooligomeric form. However, in the complex, themotion betweendomains 3 and 4 was restricted, blocking the conformationaltransition from the monomeric to oligomeric form. Because ofthat block, the lytic activity of PLY in complex is lower thanthat of the free protein.The limitations of special polyvalent vaccines that prevent

S. pneumoniae infection lie in their immune persistence, localreactions postinoculation, and serious reactions to vaccination(Stansfield, 1987). Based on individual contributions to path-ogenesis, the most appropriate candidates for vaccines arebiologic proteins that diminish the pathogenicity of themicrobes, including pneumolysin (PLY), pneumococcalsurface protein A, and autolysin (Jedrzejas, 2001). PLYstimulates host cell apoptosis (Braun et al., 2002), inspirescomplement (Paton et al., 1984), and combines with choles-terol to insert into the cell membrane and form a pore that is350–450 Å in diameter (Gilbert et al., 1999). S. pneumoniaedeleted strains of ply gene have severe attenuation in mouse

Fig. 6. The distance between the domain3 and domain 4 as a function of time. Thered and black lines represent the distancebetween domain 3 and domain 4 in unli-ganded PLY and the complex of PLY-VBS,respectively.

384 Zhao et al.

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

models of colonization and infection (Paton et al., 1993).Particular agents have been reported to be beneficial for thetreatment of diseases caused by bacterial infection, indicatingthat virulence factors are required for full activity andfunction.High molecular weight complexes formed with PLY are

indispensable for its membrane insertion, and our resultsdemonstrate that VBS intervenes in the formation of thesecomplexes. VBS may represent a novel PLY-specific anti-virulence compound and may be beneficial in the treatmentof S. pneumoniae infections when combined with regular

administration of antibiotics. Although antivirulence agentshave many advantageous characteristics, a great deal of addi-tional work is required before such agents are suitable forpractical clinical applications. The extensive use of these agentsmay lead to the development of mutations at PLY sites thateliminate VBS sensitivity; however, we do not believe that suchmutations will be likely or occur at a high rate becauseantivirulence agents neither disrupt essential bacterial cellu-lar functions nor exert selective pressure on pathogens.The nucleotide sequence of PLY is highly homologous to the

nucleotide sequences of other CDCs; for instance, PLY is 67%

Fig. 7. VBS alleviates PLY-mediated A549 cell injury. When VBS was treated in the environment of pneumolysin for 6 hours, the live pictures in greencolor or the dead pictures in red color of A549 cells could be acquired with help of the confocal laser scan microscope. (A) Untreated experimental cells. (B)A549 cells treated with PLY and without VBS treatment. (C and D) Cells cultured in the presence of 8 and 32 mg/ml VBS, respectively. The data in (A–D)are representative of three independent experiments. (E) VBS hinders the cytotoxicity of PLY. The LDH release by the experimental cells was measuredafter treatment with PLY in the presence of the indicated VBS concentrations. (F–H)PLY variants withmutations at Ser254, Trp278, and Leu447 did notexhibit marked changes in activity relative to WT-PLY; however, the addition of different concentrations of VBS to samples with PLY mutants did notproduce conspicuous cytoprotective effects. *P , 0.05 and **P , 0.01, according to two-tailed Student’s t tests.

Verbascoside Inhibits the Oligomerization of Pneumolysin 385

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

homologous to perfringolysin O ofClostridium perfringens and76% homologous to listeriolysin O of Listeria monocytogenes.PFO, the template used for homology modeling, sharesapproximately 40% overall identity with the other members ofthe CDC family. Thus, VBS should produce similar effects onother CDCs in theory, although further research is required to

elucidate the interactions of VBS with these similar proteinsin practice.

Acknowledgments

Dr. David E. Briles (Departments of Microbiology, University ofAlabama at Birmingham) is thanked for the generous providing of the

Fig. 8. Verbascoside (VBS) protects mice against Streptococcus pneumoniae pneumonia. (A) Pharmacokinetics of VBS. The female C57BL/6J micereceived a single subcutaneous dose of 100mg/kg. Mouse eyeball blood samples were collected at designated times to determine the serum concentrationsof VBS. (B) The influence on the mortality of the mice that were infected. Approximately 5 � 107 cfu of strain D39 were suspended in the left nare ofC57BL/6J mice, and the mortality of mice was supervised for 120 hours. D39-infected mice were treated with VBS 2 hours after infection. The other miceinjected with sterile PBS were influenced with D39 and subjected to PLN formed the matched groups. (C) The influence of VBS exerted on the lungbacteria burden of infected mice. C57BL/6J mice were administered 5 � 107 bacteria of the above-mentioned strains. The infected extent could beestimated by accounting the quantity of the lung bacteria of the mice that had been influenced for 48 hours. (D–E) Pathologic (D) and histopathologicalchanges (E) in the lung tissue from S. pneumoniae-infected mice (48 hours after infection).The lung tissues were dyed with hematoxylin and eosin(original magnification, �400). (F–G) Influence of VBS on inflammatory factors in infected mice. Tumor necrosis factor a (TNF-a; F) and interleukin1b (IL-1b; G) were assessed in the BAL fluid of mice (48 hours after infection). The experimental results shown in (B–G) are from three separate tests,wherein *P, 0.5 and **P, 0.01. Kaplan-Meier testswere used to assess the survivorship curves, and two-tailed Student’s t testswere adopted to analyzethe remaining experimental data.

386 Zhao et al.

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from

S. pneumoniae strains D39 and strain PLN (an isogenic pneumolysin-deficient S. pneumoniae).

Authorship Contributions

Participated in research design: Zhao, Li, Wang, Deng, and Niu.Conducted experiments: Zhao, Li, Wang, Deng, and Niu.Performed data analysis: Zhao, Li, Guo, Liu, Deng, and Niu.Wrote or contributed to the writing of the manuscript: Zhao, Li,

Deng, and Niu.

References

Balachandran P, Hollingshead SK, Paton JC, and Briles DE (2001) The autolyticenzyme LytA of Streptococcus pneumoniae is not responsible for releasing pneu-molysin. J Bacteriol 183:3108–3116.

Barrett CP and Noble ME (2005) Molecular motions of human cyclin-dependent ki-nase 2. J Biol Chem 280:13993–14005.

Berry AM, Yother J, Briles DE, Hansman D, and Paton JC (1989) Reduced virulenceof a defined pneumolysin-negative mutant of Streptococcus pneumoniae. InfectImmun 57:2037–2042.

Black RE, Cousens S, Johnson HL, Lawn JE, Rudan I, Bassani DG, Jha P, CampbellH, Walker CF, and Cibulskis R et al.; Child Health Epidemiology Reference Groupof WHO and UNICEF (2010) Global, regional, and national causes of child mor-tality in 2008: a systematic analysis. Lancet 375:1969–1987.

Boulnois GJ, Paton JC, Mitchell TJ, and Andrew PW (1991) Structure and function ofpneumolysin, the multifunctional, thiol-activated toxin of Streptococcus pneumo-niae. Mol Microbiol 5:2611–2616.

Braun JS, Sublett JE, Freyer D, Mitchell TJ, Cleveland JL, Tuomanen EI, and WeberJR (2002) Pneumococcal pneumolysin and H(2)O(2) mediate brain cell apoptosisduring meningitis. J Clin Invest 109:19–27.

Campbell GD, Jr and Silberman R (1998) Drug-resistant Streptococcus pneumoniae.Clin Infect Dis 26:1188–1195.

Chiou WF, Lin LC, and Chen CF (2004) Acteoside protects endothelial cells againstfree radical-induced oxidative stress. J Pharm Pharmacol 56:743–748.

Dessing MC, Florquin S, Paton JC, and van der Poll T (2008) Toll-like receptor 2contributes to antibacterial defence against pneumolysin-deficient pneumococci.Cell Microbiol 10:237–246.

Dessing MC, Knapp S, Florquin S, de Vos AF, and van der Poll T (2007) CD14facilitates invasive respiratory tract infection by Streptococcus pneumoniae. Am JRespir Crit Care Med 175:604–611.

Dong J, Qiu J, Zhang Y, Lu C, Dai X, Wang J, Li H, Wang X, Tan W, and Luo M et al.(2013) Oroxylin A inhibits hemolysis via hindering the self-assembly of a-hemolysinheptameric transmembrane pore. PLOS Comput Biol 9:e1002869.

Gilbert RJ, Jiménez JL, Chen S, Tickle IJ, Rossjohn J, Parker M, Andrew PW,and Saibil HR (1999) Two structural transitions in membrane pore formation bypneumolysin, the pore-forming toxin of Streptococcus pneumoniae. Cell 97:647–655.

Gilbert RJ, Rossjohn J, Parker MW, Tweten RK, Morgan PJ, Mitchell TJ, ErringtonN, Rowe AJ, Andrew PW, and Byron O (1998) Self-interaction of pneumolysin, thepore-forming protein toxin of Streptococcus pneumoniae. J Mol Biol 284:1223–1237.

Hess B, Kutzner C, van der Spoel D, and Lindahl E (2008) GROMACS 4: Algorithmsfor highly efficient, load-balanced, and scalable molecular simulation. J ChemTheory Comput 4:435–447.

Hidalgo M, Castaneda E, and Arias CA (2003) Tolerance to vancomycin in a multi-resistant, Colombian isolate of Streptococcus pneumoniae. J Antimicrob Chemo-ther 52:300–302.

Hu R, Barbault F, Delamar M, and Zhang R (2009) Receptor- and ligand-based 3D-QSAR study for a series of non-nucleoside HIV-1 reverse transcriptase inhibitors.Bioorg Med Chem 17:2400–2409.

Jedrzejas MJ (2001) Pneumococcal virulence factors: structure and function. Micro-biol Mol Biol Rev 65:187–207 first page, table of contents.

Jogalekar AS, Reiling S, and Vaz RJ (2010) Identification of optimum computationalprotocols for modeling the aryl hydrocarbon receptor (AHR) and its interactionwith ligands. Bioorg Med Chem Lett 20:6616–6619.

Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW, and Klein ML (1983)Comparison Of Simple Potential Functions for Simulating Liquid Water. J ChemPhys 79:926–935.

Laskowski RA, Moss DS, and Thornton JM (1993) Main-chain bond lengths and bondangles in protein structures. J Mol Biol 231:1049–1067.

Li Y, Gan L, Li GQ, Deng L, Zhang X, and Deng Y (2014) Pharmacokineticsof plantamajoside and acteoside from Plantago asiatica in rats by liquidchromatography-mass spectrometry. J Pharm Biomed Anal 89:251–256.

Liu M, Su JG, Kong R, Sun TG, Tan JJ, Chen WZ, and Wang CX (2008) Moleculardynamics simulations of the bacterial periplasmic heme binding proteins ShuT andPhuT. Biophys Chem 138:42–49.

Morris GM, Goodsell DS, Huey R, and Olson AJ (1996) Distributed automateddocking of flexible ligands to proteins: parallel applications of AutoDock 2.4. JComput Aided Mol Des 10:293–304.

Morris GM, Huey R, Lindstrom W, Sanner MF, Belew RK, Goodsell DS, and OlsonAJ (2009) AutoDock4 and AutoDockTools4: Automated docking with selective re-ceptor flexibility. J Comput Chem 30:2785–2791.

Musher DM, Phan HM, and Baughn RE (2001) Protection against bacteremicpneumococcal infection by antibody to pneumolysin. J Infect Dis 183:827–830.

Niu X, Qiu J, Wang X, Gao X, Dong J, Wang J, Li H, Zhang Y, Dai X, and Lu C et al.(2013) Molecular insight into the inhibition mechanism of cyrtominetin to a-hemolysinby molecular dynamics simulation. Eur J Med Chem 62:320–328.

Novak R, Henriques B, Charpentier E, Normark S, and Tuomanen E (1999)Emergence of vancomycin tolerance in Streptococcus pneumoniae. Nature 399:590–593.

Palmer M (2001) The family of thiol-activated, cholesterol-binding cytolysins. Toxicon39: 1681–1689.

Paton JC, Andrew PW, Boulnois GJ, and Mitchell TJ (1993) Molecular analysis of thepathogenicity of Streptococcus pneumoniae: the role of pneumococcal proteins.Annu Rev Microbiol 47:89–115.

Paton JC, Lock RA, and Hansman DJ (1983) Effect of immunization with pneumo-lysin on survival time of mice challenged with Streptococcus pneumoniae. InfectImmun 40:548–552.

Paton JC, Rowan-Kelly B, and Ferrante A (1984) Activation of human complement bythe pneumococcal toxin pneumolysin. Infect Immun 43:1085–1087.

Price KE and Camilli A (2009) Pneumolysin localizes to the cell wall of Streptococcuspneumoniae. J Bacteriol 191:2163–2168.

Pu X, Song Z, Li Y, Tu P, and Li H (2003) Acteoside from Cistanche salsa inhibitsapoptosis by 1-methyl-4-phenylpyridinium ion in cerebellar granule neurons.Planta Med 69:65–66.

Qiu J, Niu X, Dong J, Wang D, Wang J, Li H, Luo M, Li S, Feng H, and Deng X (2012)Baicalin protects mice from Staphylococcus aureus pneumonia via inhibition of thecytolytic activity of a-hemolysin. J Infect Dis 206:292–301.

Qiu J, Wang D, Zhang Y, Dong J, Wang J, and Niu X (2013) Molecular modelingreveals the novel inhibition mechanism and binding mode of three natural com-pounds to staphylococcal a-hemolysin. PLoS One 8:e80197.

Ragle BE, Karginov VA, and Bubeck Wardenburg J (2010) Prevention and treatmentof Staphylococcus aureus pneumonia with a beta-cyclodextrin derivative. Anti-microb Agents Chemother 54:298–304.

Rasko DA and Sperandio V (2010) Anti-virulence strategies to combat bacteria-mediated disease. Nat Rev Drug Discov 9:117–128.

Rungrotmongkol T, Nunthaboot N, Malaisree M, Kaiyawet N, Yotmanee P, Mee-prasert A, and Hannongbua S (2010) Molecular insight into the specific binding ofADP-ribose to the nsP3 macro domains of chikungunya and Venezuelan equineencephalitis viruses: molecular dynamics simulations and free energy calculations.J Mol Graph Model 29:347–353.

Ryckaert JP, Ciccotti G, and Berendsen HJC (1977) Numerical-integration of carte-sian equations of notion if a system with constraints - molecular-dynamics ofN-alkanes. J Comput Phys 23:327–341.

Schapoval EE, Vargas MR, Chaves CG, Bridi R, Zuanazzi JA, and Henriques AT(1998) Antiinflammatory and antinociceptive activities of extracts and isolatedcompounds from Stachytarpheta cayennensis. J Ethnopharmacol 60:53–59.

Schmeck B, Gross R, N’Guessan PD, Hocke AC, Hammerschmidt S, Mitchell TJ,Rosseau S, Suttorp N, and Hippenstiel S (2004) Streptococcus pneumoniae-inducedcaspase 6-dependent apoptosis in lung epithelium. Infect Immun 72:4940–4947.

Stansfield SK (1987) Acute respiratory infections in the developing world: strategiesfor prevention, treatment and control. Pediatr Infect Dis J 6:622–629.

Swanson JMJ, Henchman RH, and McCammon JA (2004) Revisiting free energycalculations: a theoretical connection to MM/PBSA and direct calculation of theassociation free energy. Biophys J 86:67–74.

Thummeepak R, Leerach N, Kunthalert D, Tangchaisuriya U, Thanwisai A,and Sitthisak S (2015) High prevalence of multi-drug resistant Streptococcuspneumoniae among healthy children in Thailand. J Infect Public Health 8:274–281.

Vorontsov II and Miyashita O (2011) Crystal molecular dynamics simulations tospeed up MM/PB(GB)SA evaluation of binding free energies of di-mannose deoxyanalogs with P51G-m4-Cyanovirin-N. J Comput Chem 32:1043–1053.

Wang J, Qiu J, Tan W, Zhang Y, Wang H, Zhou X, Liu S, Feng H, Li W, and Niu Xet al. (2015) Fisetin inhibits Listeria monocytogenes virulence by interfering withthe oligomerization of listeriolysin O. J Infect Dis 211:1376–1387.

Wang J, Wang W, Kollman PA, and Case DA (2006) Automatic atom type and bondtype perception in molecular mechanical calculations. J Mol Graph Model 25:247–260.

Werner G, Strommenger B, and Witte W (2008) Acquired vancomycin resistance inclinically relevant pathogens. Future Microbiol 3:547–562.

Wu K, Zhang X, Shi J, Li N, Li D, Luo M, Cao J, Yin N, Wang H, and Xu W et al.(2010) Immunization with a combination of three pneumococcal proteins confersadditive and broad protection against Streptococcus pneumoniae Infections inMice. Infect Immun 78:1276–1283.

Xiong Q, Hase K, Tezuka Y, Tani T, Namba T, and Kadota S (1998) Hepatoprotectiveactivity of phenylethanoids from Cistanche deserticola. Planta Med 64:120–125.

Zhou R, Berne BJ, and Germain R (2001) The free energy landscape for beta hairpinfolding in explicit water. Proc Natl Acad Sci USA 98:14931–14936.

Address correspondence to: Xiaodi Niu, Key Laboratory of Zoonosis,Ministry of Education, Department of Food Quality and Safety, JilinUniversity, Xi’an Rd 5333, Changchun 130062, China. E-mail: niuxd@jlu.edu.cn or Xuming Deng, Key Laboratory of Zoonosis, Ministry of Education, Collegeof Veterinary Medicine, Jilin University, Xi’an Rd 5333, Changchun 130062,China. E-mail: dengxm@jlu.edu.cn

Verbascoside Inhibits the Oligomerization of Pneumolysin 387

at ASPE

T Journals on June 22, 2020

molpharm

.aspetjournals.orgD

ownloaded from