Heat shock enhances the susceptibility of BHK cells to rotavirus infectionthrough the facilitation of entry and post-entry virus replication steps
Tomas Lopez, Susana Lopez, Carlos F. Arias ∗Departamento de Genetica del Desarrollo y Fisiologıa Molecular, Instituto de Biotecnologıa, Universidad Nacional
Autonoma de Mexico/UNAM, Av. Universidad 2001, Colonia Chamilpa, Cuernavaca, Morelos 62210, Mexico
Received 10 January 2006; received in revised form 20 April 2006; accepted 21 April 2006Available online 5 June 2006
odified their susceptibility upon heat stress, suggesting that heat shock induces factors that are rate-limiting the replication of rotaviruses in BHKut not in MA104 cells. The heat treatment was shown to facilitate the rotavirus infection of BHK cells at the penetration and post-penetrationevels, and each of these stages seems to contribute comparably to the overall observed 100-fold increase in infectivity. Since the binding of theirus to the cell surface was not affected, the caloric stress probably facilitates the penetration and/or uncoating of the virus. The pathway of virusntry into heat-shocked BHK cells seems to be similar to that used in MA104 cells, since treatments that affect MA104 cell infection also affectedotavirus infectivity in heat-treated BHK cells.
Rotaviruses are the major cause of severe infantile gastroen-eritis worldwide. These viruses are non-enveloped, and have aapsid formed by three concentric layers of protein that enclosedouble-strand RNA genome (Estes, 2001). The outer layer is
omposed of proteins VP4 and VP7.Rotaviruses have a specific cell tropism in vivo, infecting
rimarily the mature enterocytes of the villi of the small intes-ine, suggesting that these cells bear specific receptors for theirus. However, recent reports suggest extra-intestinal spread ofotavirus during infection (Mossel and Ramig, 2003), indicatingwider host tissue range than previously appreciated. In vitro,
otaviruses bind to a wide variety of cell lines, although only aubset of these are efficiently infected (Bass et al., 1992; Ciarlett al., 2002). The molecular basis of this tropism are not fully
understood either in vivo or in vitro, although it has been shownthat poorly susceptible cell lines become susceptible if the virusparticles are introduced by transfection, suggesting that the per-missivity barrier is at the virus entry level (Bass et al., 1992;Ciarlet et al., 2002).
Several cell molecules have been reported to be involved inrotavirus cell binding and penetration, including gangliosides(Guo et al., 1999; Rolsma et al., 1998), integrins �2�1, �v�3,�x�2, and �4�1 (Ciarlet et al., 2002; Coulson et al., 1997;Guerrero et al., 2000a; Hewish et al., 2000; Zarate et al., 2000a)and the heat shock cognate hsc70 protein (Guerrero et al., 2002).Hsc70 has been proposed to be involved during virus entry, at apost-binding step, probably penetration (Lopez and Arias, 2004;Perez-Vargas et al., 2006).
Also, it has been reported that rotavirus infection of kidney(MA104) and intestinal (Caco-2) epithelial cell lines inducesan increase in the expression of stress genes (Cuadras et al.,2002; Xu et al., 1998), including members of the major classesof general chaperons, such as hsp40, hsp70, and hsp90. Giventhese observations, and the proposed role of hsc70 during virus
T. Lopez et al. / Virus Research 121 (2006) 74–83 75
entry, we hypothesized that some of the stress-induced proteinscould have a role in facilitating rotavirus replication, as it hasbeen shown for other viruses (Burch and Weller, 2005; Chromyet al., 2003; Glotzer et al., 2000; Vasconcelos et al., 1998). To testthis hypothesis we explored the effect of heat shock on rotaviruscell infection. We found that the infectivity of rotaviruses isincreased in heat-treated BHK cells during cell entry, as well asat a post-entry stage.
2. Materials and methods
2.1. Cells and viruses
MA104 (monkey kidney epithelium) and BHK (baby ham-ster kidney) cells were cultured in Eagle’s minimal essentialmedium (MEM) supplemented with 10% fetal calf serum (FCS).Rotavirus strains RRV and Wa were originally obtained fromH.B. Greenberg, Stanford University, Stanford, CA; rotavirusnar3 has been described previously (Mendez et al., 1993). RRV,Wa, and nar3 viruses, as well as reovirus serotype 1, were prop-agated as described (Cuadras et al., 1997; Espejo et al., 1980b;Mendez et al., 1993). Double-layered (DLPs) and triple-layered(TLPs) rotavirus particles were purified by CsCl density gradi-ents, as reported (Espejo et al., 1980a).
2.2. Antibodies
h0Mwabhm
2
aita4au
2
salAct
and the infected cells were detected by an immunoperoxidaseinfectious focus detection assay, using rabbit polyclonal sera toeither rotavirus or reovirus, as described previously (Lizano etal., 1991). Wells containing around 1000 focus forming units(FFUs) were counted using a semiautomatic Visiolab 1000 sta-tion (Biocom, France) (Guerrero et al., 2000b). Since one fifthof the cell monolayer of each well was screened for viral foci,then the counts obtained per well were around 200 FFUs. Insome experiments the cells were infected with rotavirus for 1 hat 37 ◦C prior to the heat shock treatment.
2.5. Infectivity blocking assays
To evaluate the blocking activity of methyl-�-cyclodextrin(m�CD) (Aldrich) and of antibodies to hsc70 on rotavirus infec-tivity, confluent MA104 and BHK cell monolayers in 96-wellplates were heat-shocked as described above, and then incubatedwith either 10 mM m�CD in MEM for 1 h at 37 ◦C, or with anti-bodies to hsc70 for 90 min at 37 ◦C. After the incubation stepthe cells were washed twice with PBS, and infected with a fixedamount of infectious virus (1000 FFUs) per well, as describedabove, except for the blocking experiments with antibodies tohsc70, in which the virus was adsorbed for 45 min at 4 ◦C insteadof 60 min at 37 ◦C.
2.6. Permeabilization assay
(owMCiatplbwacortvwa
2
23ocP
Monoclonal antibodies (MAbs) MA3-007 and MA3-014 tosc70 were obtained from Affinity Bioreagents Inc. MAb MA3-07 crossreacts with GRP78, hsc70 and hsp70, while MAbA3-014 recognizes specifically hsc70. MAb B6 to hsc70as from Santa Cruz Biotechnology, and MAb W27 to hsp70
nd hsc70 was from NeoMarkers. Anti-peptide polyclonal anti-ody SPA-812 and MAb SPA-810, which are both specific forsp70, were from StressGen. The rabbit polyclonal hyperim-une serum to hsc70 has been described (Guerrero et al., 2002).
.3. Heat shock conditions
Confluent cell monolayers in 96-well plates were incubatedt 43–45 ◦C during 20 min in MEM-10% FCS, or at 37 ◦C dur-ng the same time for control cells. To reach and maintain theemperature of 43–45 ◦C inside the wells, medium pre-warmedt 48 ◦C was added to the cells and then placed in a water bath at7–48 ◦C for the 20-min incubation period. The cells were thenllowed to recover from the heat shock during 30 min at 37 ◦C,nless otherwise indicated, and infected as described below.
.4. Infectivity assay
After the 30-min period of recovery at 37 ◦C, control and heat-hocked cells, in 96-well plates, were washed twice with MEM,nd then 50 �l of two-fold dilutions of rotavirus or reovirus cellysates were added per well, and adsorbed for 60 min at 37 ◦C.fter the adsorption period, the virus inoculum was removed, the
ells were washed twice with MEM, and the cultures were main-ained for 14 h at 37 ◦C. The cell monolayers were then fixed,
The �-sarcin co-entry assay was carried out as describedCuadras et al., 1997). Briefly, confluent monolayers of MA104r BHK cells in 96-well tissue culture plates were infectedith rotavirus at a multiplicity of infection (MOI) of 10 forA104 cells, in the presence of 20 �g/ml of �-sarcin (Sigmahemical Co.). One hour after adsorption at 37 ◦C, the viral
noculum and the toxin were removed, and fresh MEM wasdded for an additional period of 30 min at 37 ◦C. After thisime, the media was replaced by methionine-free MEM sup-lemented with 25 �Ci/ml of [35S] Easy Tag Express-proteinabelling mix (1175 Ci/mmol, NEN), and the cells were incu-ated for 1 h at 37 ◦C. After the labelling period, the cells wereashed with PBS, treated with 5% trichloroacetic acid for 5 min
t room temperature, and washed three times with ethanol. Theell monolayer was allowed to dry before the addition of 50 �lf 0.1% sodium dodecyl sulphate (SDS) in 0.1 N NaOH. Totaladioactivity in the sample was determined by liquid scintilla-ion counting. As controls, �-sarcin in the absence of virus, orirus alone without toxin, were added to the cells and labelledith [35S] Easy Tag Express-protein labelling mix as described
bove.
.7. Binding assay
The binding assay was carried out as described (Zarate et al.,000b). Briefly, a suspension of 5 × 104 cells were mixed with00 ng of purified RRV TLPs in MEM-1% BSA in a final volumef 200 �l, and incubated for 1 h at 4 ◦C with gentle mixing. Theell–virus complexes were washed three times with ice coldBS containing 0.5% BSA. In the final wash, the cells were
76 T. Lopez et al. / Virus Research 121 (2006) 74–83
transferred to a fresh tube, and then treated with 50 �l of lysisbuffer (50 mM Tris [pH 7.5], 150 mM NaCl, 0.1% Triton X-100).The virus present in the lysates was quantified by an enzyme-linked immunosorbent assay, as described.
2.8. Lactic dehydrogenase assay
BHK cells in 96-well plates were heat-shocked as describedabove. After the 30-min recovery period, the cells were washedtwice with MEM, and MEM without serum was added. At theindicated times, the cell medium was collected and the activityof lactic dehydrogenase was determined with a commercial kit(Sigma Chemical Co.) according to manufacturer instructions.
2.9. Lipofection of DLPs
Ten-fold dilutions of DLPs (from 4 �g/ml to 4 ng/ml) wereincubated with 3 �l of lipofectamine (GIBCO-BRL) in a finalvolume of 30 �l, for 15 min at room temperature, 230 �l of MEMwere subsequently added, and 50 �l of this mixture was used perwell of a 96-well plate. The DLP-lipofectamine complex wasleft on the cells for 1 h at 37 ◦C, and then washed twice withMEM. The cells were then heat-shocked or not, washed twicewith MEM, and the cultures were maintained for 14 h at 37 ◦Cin MEM. The cell monolayers were fixed, and the infected cellswere detected by an immunoperoxidase focus-forming assay((i
2
attb(arceu
2
sasaIMMaiw
with fluorescein-conjugated anti-mouse IgM (8 �g/ml; SigmaChemical Co.) or IgG (12 �g/ml; Zymed) antibodies for 1 h at4 ◦C. After fixation with 2% paraformaldehyde in PBS, the anti-body binding was analyzed using a FACScan flow cytometer andCellquest software (Becton Dickinson) with appropriate gatingparameters.
2.12. Metabolic labelling of viral proteins in heat-shockedBHK cells
Cells were heat-shocked, or not, as described above. Thevirus, at an MOI of 1, was adsorbed for 1 h before or afterthe heat shock. At 12 h post infection the medium was replacedwith MEM without methionine for 30 min, and fresh MEM con-taining 50 �Ci/ml of [35S] Easy Tag Express-protein labellingmix was then added, and the cells were further incubated for1 h at 37 ◦C. Finally, the cells were washed twice with PBSand lysed in Laemmli sample buffer, and the labelled pro-teins in the lysates were analyzed by SDS-polyacrylamide gelelectrophoresis.
2.13. Analysis of the synthesis de novo of viral proteins inheat-shocked BHK cells
Confluent monolayers of BHK cells in 96-well plates wereheat-shocked as described above and, at 0, 1, 2, 4, 6, or 8 hamptd
3
3r
dcutBMAtnaRtciecrsw
Lizano et al., 1991). A concentration of 20 ng/ml of DLPs1 ng/well of a 96-well plate) produced around 1000 FFUs/welln heat-shocked BHK cells.
.10. Immunoblot analysis
Cells grown in 24-well plates were heat-shocked, or not, andt the indicated times the cells were lysed, and the proteins inhe lysate separated in an 11% SDS-polyacrylamide gel andransferred to nitrocellulose. The transferred proteins were incu-ated with antibodies SPA-812 (1:2500) directed to hsp70, or B61:2000) directed to hsc70, in PBS-0.1% Tween 20. The boundntibodies were detected by incubation with either a goat anti-abbit immunoglobulin or rabbit anti-mouse immunoglobulinonjugated to horseradish peroxidase, and developed with annhanced chemiluminiscence system (NEN Life Science Prod-cts Inc.).
.11. Flow cytometry (FACS)
Confluent cells were washed and brought into a single celluspension by incubation with 0.5 mM EDTA in PBS at 37 ◦Cnd dispersed by gentle pipetting. Cells were collected by low-peed centrifugation (200 × g), resuspended in ice-cold MEM,nd the cell concentration determined with a hemocytometer.n each experiment, 2.5 × 105 cells were incubated with either
Ab MA3-014 (IgM, diluted 1:25 in PBS-2% FCS) to hsc70, orAb SPA-810 (IgG, 10 �g/ml in PBS-2% FCS) to hsp70 for 1 h
t 4 ◦C. Mouse IgM (11 �g/ml) or IgG (10 �g/ml) (Sigma Chem-cal Co.) antibodies were used as negative controls. The cellsere washed twice with 2% FCS in PBS, and then incubated
fter the 30-min period of recovery at 37 ◦C, the cells wereetabolically labelled for 1 h with [35S] Easy Tag Express-
rotein labelling mix, and the radioactivity incorporated intorichloroacetic acid-precipitable material was determined asescribed above.
. Results
.1. Heat shock increases the susceptibility of BHK cells tootavirus infection
To test the effect of heat shock on rotavirus infection, weetermined the virus focus forming units (FFUs) produced inells that had been incubated at 43–45 ◦C for 20 min. We eval-ated two different cell lines that differ in their susceptibilityo rotaviruses, MA104 cells, which are highly susceptible, andHK cells, which are about 1000-fold less susceptible thanA104 to rotavirus infection (Fig. 1A, Ciarlet et al., 2002).fter heat shock, the susceptibility of MA104 cells to infec-
ion by the sialic acid-dependent rhesus rotavirus (RRV) wasot significantly modified (1.4-fold), while BHK cells becameround 100-fold more susceptible (Fig. 1A). The production ofRV progeny in heat-shocked BHK cells, as determined by titra-
ion of the infectious virus particles produced in heat-treatedells, was also enhanced to a level similar to that observedn the focus-forming assay (not shown). We then tested if theffect of heat shock-induced increased susceptibility of BHKells was limited to RRV or if it was also observed for otherotavirus strains. Heat-treated cells were about 4-fold moreusceptible to nar3, a sialic acid-independent mutant of RRV,hile they were 63 times more infectable by the natural sialic
T. Lopez et al. / Virus Research 121 (2006) 74–83 77
Fig. 1. Heat shock increases the susceptibility of BHK cells to rotavirus infection. (A) MA104 and BHK cells, grown in 96-well plates, were heat-shocked for 20 minat 43–45 ◦C (closed bars), or maintained at 37 ◦C for the same period of time, as a control (open bars). The cells were then allowed to recover for 30 min at 37 ◦C andwere infected with RRV as described in Section 2. (B) Heat-shocked BHK cells (closed bars) were infected with rotavirus strains RRV, Nar3 or Wa, or with reovirusas control, as indicated. The open bars represent the infectivity of the various viruses in control, non-treated cells. In both panels, (A) and (B), data are expressedas the number of virus focus forming units/ml, and the arithmetic mean and standard error of three independent experiments performed in duplicate are shown. (*)P ≤ 0.05; (**) P ≤ 0.0143, in a Mann-Whitney test. (C) BHK cells were heat-shocked or not (control), and were then infected with RRV (Rota) or reovirus (Reo) at aMOI of 1, or left uninfected (UI), as control. At 12 h post-infection the cells were labelled with [35S] Easy Tag Express-protein labelling mix for 1 h, and the proteinswere resolved by SDS-polyacrylamide gel electrophoresis and detected by autoradiography. The rotavirus and reovirus proteins are identified at the right and leftsides of the panel, respectively; the reovirus proteins are further indicated by vertical lines in the two gels; four heat-induced cellular proteins are also indicated ( ).
acid-independent human rotavirus strain Wa, as compared tountreated cells (Fig. 1B). Since the most pronounced heat-induced increase in infectivity was observed for RRV, this virusstrain was used in all subsequent experiments. The enhanced sus-ceptibility of BHK cells to rotavirus infection was also clearlydetected by metabolic labelling of viral proteins synthesizedin control versus heat-shocked cells. When control BHK cellswere infected, and the cells were pulse-labelled with an [35S]Easy Tag Express-protein labelling mix for 1 h, starting at 12 hpost-infection, rotaviral proteins were non distinguishable fromthe background of cellular proteins (Fig. 1C, left panel). Onthe other hand, in heat-treated cells the rotavirus proteins wereclearly identifiable (Fig. 1C, right panel). At least four BHK heatstress proteins are also prominent in uninfected and infectedcells after the heat treatment (indicated by dots in Fig. 1C,right panel). We also tested the effect of heat shock on theinfectivity of BHK cells by reovirus. As shown in Fig. 1B,as opposed to what was observed for rotaviruses, the suscep-tibility to reovirus infection was slightly reduced by the heattreatment, indicating that the increased infectivity of rotavirusesunder these conditions is not shared by reoviruses. The reducedgene expression of reovirus in BHK-treated cells could also beappreciated by the marked decrement in the synthesis of viralproteins, as compared to untreated cells (Fig. 1C, indicated byvertical bars).
Rotavirus double-layered particles (DLPs) are not infectious,ai1c
a general modification of the cell membrane permeability thatwould allow rotavirus triple-layered particles (TLPs) to enterthe cell in a non-specific manner, we tested if rotavirus DLPswould become infectious in heat-treated cells. As expected, bothuntreated and heat-shocked BHK cells were refractory to infec-tion by DLPs (data not shown). In addition, the integrity ofthe cell membrane seemed not to be affected, at least appre-ciably, as judged by the fact that the release of LDH activityinto the cell culture medium was maintained in heat-shockedcells at control levels even 12 h after the treatment (Fig. 2A).We also evaluated the protein synthesis capacity of the cellsafter the heat shock. For this, the cells were pulse-labeled for1 h with [35S] Easy Tag Express-protein labelling mix at dif-ferent times post-treatment. When the cells were labelled for1 h immediately after the heat shock, and a recovery period of30 min at 37 ◦C (time 0 h), the protein synthesis was decreasedto about 15% that of untreated cells (Fig. 2B). The proteinsynthesis machinery of these cells recovered gradually, suchthat by 6 h post-treatment, it had been recovered essentially tothe same level observed in untreated, control cells. Of inter-est, the window of increased susceptibility of BHK cells torotavirus infection induced by heat shock remained in the samelevel for at least 6 h post-treatment (the infectivity of the virusin control, non heated cells was 5.6 × 104, while the virustiter at 0, 1, 2, 4, 6, and 8 h post-treatment was 5, 5.1, 5.7,5.9, 5.5, and 3.5 × 106, respectively.), indicating that whenteti
lthough they can initiate a virus replication cycle if introducednto cells through physical means, like lipofection (Bass et al.,992). Thus, to rule out the possibility that the increased sus-eptibility of BHK cells to rotavirus infection was the result of
he acute effects of heat-treatment (as judged for their recov-red protein synthesis capacity; Fig. 2B) had been overcome,he cells maintained their enhanced susceptibility to rotavirusnfection.
78 T. Lopez et al. / Virus Research 121 (2006) 74–83
Fig. 2. Effect of heat shock on viability of BHK cells. (A) Cells were grown in 12-well plates, heat-shocked, and the LDH activity released into the cell culturemedium was measured at the indicated times post-treatment. The 100% activity of LDH was determined by lysis of the cells with Triton X-100, as described inSection 2. Open and closed bars represent control and heat-shocked cells, respectively. The arithmetic mean and standard error of three independent experimentsperformed in duplicate are shown. (B) Cell monolayers in 96-well plates were heat-shocked for 20 min at 43–45 ◦C and, at the indicated times after a 30-min period ofrecovery at 37 ◦C, the cells were metabolically labelled for 1 h with [35S] Easy Tag Express-protein labelling mix, and the incorporated radioactivity was determinedas described in Section 2. Data is expressed as percent of the radioactivity incorporated in control, non heated cells. The arithmetic mean and standard error of fourindependent experiments performed in duplicate are shown.
3.2. Heat shock enhances rotavirus infectivity at twodifferent stages of the virus replication cycle
To determine if the increased susceptibility of BHK cells wasat the level of virus entry or at a post-entry step, we determinethe infectivity of RRV when the heat shock was given after thevirus had been allowed to enter the cells for 1 h at 37 ◦C. Toprevent the entry, upon cell heat shock, of the viral particles thatremained attached to the cell surface after the 1-h incubation
period, the cell bound viruses were neutralized with MAb 159before heat treatment. A 10-fold increase in virus infectivity wasobserved under these conditions (Fig. 3A). The fact that a 100-fold increase in infectivity was obtained when the heat shockwas given before virus infection (Fig. 1A) indicates that the heatstress increases the BHK susceptibility to rotavirus infection atthe level of cell entry, as well as at a post-entry step of replication.The analysis of metabolically labelled viral proteins showed thatthe synthesis of rotaviral proteins became only evident when
F icatiow icateda the vic e arithdAeD(w
ig. 3. The heat shock-induced enhancement of infectivity is at two virus replere heat-treated before (HS/Virus) or after (Virus/HS) virus infection, as ind
fter adsorption of the virus for 1 h at 37 ◦C, to prevent further penetration ofells were immunostained as described in Section 2 and the FFUs counted. Th
uplicate are shown. (B) BHK cells were heat-shocked (HS) or not (C) and infectedt 12 h post-infection the cells were labelled with [35S] Easy Tag Express-protein lab
lectrophoresis and detected by autoradiography. The viral proteins are indicated. (C)LPs (200–0.2 ng/well) were incubated with lipofectamine, and the virus-lipid mixtu
control), and were immunostained as described in Section 2 and the FFUs counted.ith 1 ng/well of DLPs. The arithmetic mean and standard error of four independent
n levels. (A) RRV was used to infect BHK cells grown in 96-well plates that, or not heat-shocked (control). In all cases, MAb 159 was added for 15 minrus that remained attached to the cell surface after the incubation period. Themetic mean and standard error of four independent experiments performed in
with RRV at an MOI of 1, after (HS/Virus) or before (Virus/HS) heat shock.elling mix for 1 h, and the proteins were resolved by SDS-polyacrylamide gelCells in 96-well plates were lipofected with RRV DLPs. Different amounts of
res were added to cells for 1 h at 37 ◦C. The cells were then heat-shocked or notAround 200 FFUs were counted in heat-treated cell that had been transfectedexperiments performed in duplicate are shown.
T. Lopez et al. / Virus Research 121 (2006) 74–83 79
Fig. 4. Heat shock facilitates virus penetration. (A) Confluent BHK cells in F25 flasks were heat-shocked, or not, and after a 30-min period of recovery at 37 ◦C thecells were detached with EDTA and used for the binding assay. The cells in suspension were incubated with 300 ng of purified TLPs for 1 h at 4 ◦C. The amountof virus bound to cells was determined by an ELISA, as described (Zarate et al., 2000b). Data are expressed as ng of bound virus. (B) Cells in 96-well plates wereinfected, or mock-infected, with 10 FFU/cell of purified TLPs (as titrated in MA104 cells), in the presence or absence of �-sarcin, as indicated. After 1 h of adsorptionat 37 ◦C, the virus inoculum was removed and fresh serum-free MEM was added for an additional period of 30 min. The cells were then labelled with [35S] EasyTag Express-protein labelling mix for 1 h at 37 ◦C, and the acid-precipitable material was estimated. Data are expressed as percentage of the [35S]-incorporationof mock-infected control cells, in the absence of toxin. MA104 cells were included as reference. The arithmetic mean and standard error of two (A) and four (B)independent experiments performed in duplicate are shown.
the heat shock was done before virus entry (Fig. 3B). Finally,the introduction of DLPs by lipofection, bypassing the normalentry route of the virus, confirmed the increase susceptibility ofBHK cells to rotavirus replication at a post-entry step. As shownin Fig. 3C, a 10-fold increase in the number of infected cellswas detected when DLP-lipofected cells were heat-shocked ascompared to control, non-heated cells; this level of enhancementof virus infectivity was observed regardless whether the heatshock was performed before or after lipofection of DLPs (notshown). To rule out the possibility that heat-shocked cells weremore susceptible to lipofection than untreated cells, a plasmidvector expressing GFP was lipofected into heated and controlcells. Very similar transfection efficiencies were obtained in bothuntreated and heat-treated cells (not shown).
To determine if the heat treatment facilitated virus entry atthe cell binding or penetration steps, the attachment of viralparticles to the cell surface, and the co-entry of the toxin �-sarcin,which has been previously shown to be a good indicator of viruspenetration (Cuadras et al., 1997), were measured. The bindingof purified RRV TLPs to heat-treated or untreated BHK cellswas essentially the same, and this binding was very similar tothat observed for MA104 cells (Fig. 4A). On the other hand, theentry of �-sarcin was augmented into heat-shocked BHK cells(Fig. 4B), suggesting that the treatment facilitates virus entryat a post-binding step, most probably penetration or uncoatingof the virus particle. The finding that the entry of toxin intoBiaioo
3r
i(t
The blocking activity profile of the antibodies tested was verysimilar in MA104 and heat-shocked BHK cells (Fig. 5A). Wewere unable to determine unequivocally if antibodies to hsc70block the infection of untreated BHK cells since, as shown inFig. 1A, the basal infectivity of rotaviruses in this cell line isvery low.
It has been proposed that in MA104 cells the receptors forrotavirus are organized in lipid rafts, and the treatment of cellswith methyl-�-cyclodextrin (m�CD) has been shown to reducevirus infectivity by about 90%, apparently as the result of raftdisruption (Guerrero et al., 2000b; Isa et al., 2004). To evaluate iflipid rafts are important for rotavirus infection of heat-shockedBHK cells, heat-treated cells were incubated with 10 mM m�CDfor 1 h at 37 ◦C, and then infected with RRV. The infectivity ofthe virus was significantly reduced in these cells, as observedin MA104 cells, which were used as reference (Fig. 5B). Alto-gether, the similar effect of antibodies to hsc70 and the m�CDtreatment on virus infectivity of MA104 and heat-shocked BHKcells, suggest that rotaviruses enter these two cell lines througha similar pathway.
3.4. Hsc70 is not involved in the increase of susceptibilityof heat-shocked BHK cells to rotavirus infection
Given the proposed role of hsc70 as a post-attachment recep-tor for rotaviruses, we tested if the increased synthesis of thisproIb(Bwhh
esr
HK-treated cells is not as evident as in untreated MA104 cellss consistent with the observation that heat-treated BHK cellsre about 20-fold less susceptible than MA104 cells to RRVnfection (see Fig. 1A), and with the fact that the level of co-entryf �-sarcin into cells is in direct correlation with the multiplicityf infection (Cuadras et al., 1997).
.3. Antibodies to hsc70 and methyl-β-cyclodextrin reduceotavirus infectivity in heat-shocked BHK cells
Since hsc70 mediates rotavirus infection of MA104 cells, andt has also been shown to be present on the surface of BHK cellsGuerrero et al., 2002), we evaluated the effect of antibodieso hsc70 on the infectivity of RRV in heat-shocked BHK cells.
rotein and/or its exposure on the cell surface, or of the closelyelated heat-inducible protein hsp70, could be responsible for thebserved enhanced rotavirus entry in heat-treated BHK cells.mmunoblot analysis showed that the abundance of hsc70 inoth MA104 and BHK cells was not modified by heat shockFig. 6). On the other hand, hsp70 was not present in untreatedHK cells, although it was detected 8 h after the heat shock,hile in MA104 cells there was a basal constitutive level ofsp70 in untreated cells, which increased sharply 2 h after theeat treatment (Fig. 6).
We then tested if the heat shock increased the cell surfacexposure of these proteins. The analysis by flow cytometryhowed that the amount of hsc70 present on the cell membraneemained essentially the same in BHK treated and untreated cells
80 T. Lopez et al. / Virus Research 121 (2006) 74–83
Fig. 5. Rotavirus infection of heat-shocked BHK cells is inhibited by antibodies to hsc70, and by treatment with methyl-�-cyclodextrin (m�CD). (A) Preimmune (PI)and hyperimmune (HI) rabbit antibodies to hsc70 (80 �g/ml), and MAbs 014 (dil 1:100), W27 (2 �g/ml), B6 (2 �g/ml), and 007 (dil 1:100) to hsc70 were added for90 min at 37 ◦C to untreated MA104, or heat-shocked BHK, cells in 96-well plates. The cells were then infected with 1000 FFUs/well of RRV. Data are expressed aspercentage of the virus infectivity obtained when the cells were pre-incubated with MEM as control. Open and closed bars represent control and heat-shocked cells,respectively. MAbs 014 and B6 are specific for hsc70, MAbs W27 and 007, as well as the rabbit hyperimmune serum to hsc70, cross react with hsp70. (B) MA104cells or heat-treated BHK cells in 96-well plates were incubated with 10 mM m�CD for 1 h at 37 ◦C. The cells were washed twice and infected with 1000 FFUs/wellof RRV, and 14 h later the infected cells were detected by immunostaining as described in Section 2. The arithmetic mean and standard error of six (A) and three (B)independent experiments performed in duplicate are shown.
Fig. 6. Immunoblot analysis of hsc70 and hsp70 in heat-shocked and controlcells. MA104 and BHK cells were heat-shocked, or not, and at the indicated timespost-treatment the cellular proteins were separated by SDS-polyacrylamide gelelectrophoresis and transferred to nitrocellulose. MAb B6 was used to detecthsc70, and the monospecific rabbit polyclonal antibody SPA-812 to recognizehsp70. The bound antibodies were developed by incubation with specific sec-ondary antibodies conjugated to horseradish peroxidase, and developed with anenhanced chemiluminescence system.
(Fig. 7). The level of cell surface expression in untreated MA104cells is shown as reference. On the other hand, hsp70 was notdetected on the surface of either BHK or MA104 cells, regard-less whether they had been heat-treated or not (Fig. 7). Thisobservation agrees with the fact that antibodies to hsp70 did notblock rotavirus infectivity (data not shown). These results indi-cate that an increased cell surface expression of hsc70 or hsp70is not responsible for the enhancement of rotavirus entry intoheat-shocked BHK cells.
4. Discussion
We have shown that a caloric stress increases the susceptibil-ity of BHK cells to rotavirus infection by more than 100-fold.This enhanced susceptibility seems to be specific for rotavirus,and not the result of a general increase in permeability of theplasma membrane due to the heat treatment, since: (a) the sus-ceptibility of heat-treated BHK cells to reovirus infection wasnot augmented; (b) rotavirus DLPs were not able to enter heat-treated BHK cells and; (c) the heat treatment did not disrupt the
cells’ plasma membrane integrity, as judged by the release ofcytoplasmic LDH into the medium. Altogether, these observa-tions support the specificity of the heat treatment on the increasedsusceptibility of BHK cells to rotavirus infection.
The heat-stress facilitates the replication of rotaviruses inBHK cells during cell entry, as well as at a post-entry stage. Eachof these stages seems to contribute comparably to the overallobserved 2-logs increase in infectivity, since if the heat-shockwas carried out after the virus had been allowed to enter thecells, the number of detected FFUs increased 1-log. A similarresult was obtained when the entry step was bypassed by lipo-fection (using DLPs) before the caloric stress was applied. Ofinterest, the infectivity of rotaviruses was not, or only marginally,enhanced in the highly susceptible MA104 cells, suggesting thatheat shock induces factors that are rate-limiting the replicationof rotaviruses in BHK but not in MA104 cells.
Since the immunocytochemistry assay used in this work todetect the viral FFUs depends on the detection of viral anti-gen, the 10-fold increase of foci observed when the cells wereheat-shocked after the virus had been allowed to enter the cellsshould be the result of an augmented viral transcription or trans-lation of viral mRNAs. In this regard, it has been reported thatduring continuous heat shock, the translation of viral mRNAscontaining internal ribosomal entry sites, like those of hepati-tis C and encephalomyocarditis viruses, is increased (Kim andJang, 2002), and the enhancement of viral protein synthesis byhavVtaavfe
c
eat shock has also been reported for morbilliviruses (measlesnd canine distemper viruses), and for human T-lymphotropicirus type I (Andrews et al., 1995, 1998; Oglesbee et al., 1993;asconcelos et al., 1998). In the case of morbilliviruses, the heat
reatment increased the production of viral transcripts, associ-ted to an increased viral membrane glycoprotein expressionnd cytopathic effect. In the case of human T-lymphotropicirus type I, the cellular stress induced an increased cell sur-ace expression of the viral Env protein gp46, resulting in annhanced virus-mediated syncytia formation.
Heat shock, however, does not always increase the repli-ation of viruses, since it has been shown that it inhibits the
T. Lopez et al. / Virus Research 121 (2006) 74–83 81
Fig. 7. Flow cytometry analysis of the surface expression of hsc70 and hsp70 in MA104 and heat-shocked or not, BHK cells. MAbs MA3-014 (diluted 1:25) specificfor hsc70, and SPA-810 (10 �g/ml) specific for hsp70 (dashed lines), were used. Control antibodies of isotypes IgM (for MA3-014) and IgG (for SPA-810) (solidlines) were used at 11 and 10 �g/ml, respectively. The cell line and antibody used are indicated.
replication of Mayaro virus, SV40, and rhinoviruses (Andradeand Carvalho, 1993; Angelidis et al., 1988; Conti et al., 1999;Virgilio et al., 1997). The effect of heat shock on the replica-tion of Mayaro virus is at the level of protein translation, sincethis treatment reduces the amount of viral mRNA associated toheavy polysomes (Rosas et al., 1997).
The step facilitated by heat shock during the early stepsof infection seems to occur after the initial interaction of thevirus with the cell surface, since the binding of RRV TLPs toBHK cells was not augmented by heat treatment. In addition,the fact that the level of virus/�-sarcin co-entry into BHK cellsincreased as a consequence of heat shock suggests that this treat-ment increases the penetration and/or uncoating of the virus,however, this enhancement is not related to an increased syn-thesis or cell surface exposure of hsc70 or hsp70. It cannot bediscarded however, that the increased cell surface expressionof ER molecular chaperons, like grp78 and grp94, which areknown to interact with rotavirus during virus morphogenesis(Xu et al., 1998), and which have been identified as recep-tors for dengue and Coxsakie viruses (Jindadamrongwech etal., 2004; Triantafilou et al., 2002) could contribute to theobserved effect. This is the first report that describes thatheat shock can facilitate virus infectivity at the cell entrylevel.
The entry of rotaviruses into epithelial cells is thought to bea multistep process during which several contacts between thevt�aiii
2004). In addition, it has been suggested that sphingolipid-and cholesterol-enriched membrane lipid microdomains, usu-ally referred to as “lipid rafts” (Simons and Ikonen, 1997), mightbe involved in rotavirus cell entry, since: (a) disorganization ofthese domains by depletion of cell cholesterol inhibits rotavirusinfection (Guerrero et al., 2000b; Sanchez-San Martın et al.,2004); (b) ganglioside GM1, integrin subunits �2 and �3, andhsc70, cell molecules implicated in rotavirus infection, are asso-ciated with these lipid microdomains (Isa et al., 2004) and (c)infectious rotavirus particles associate with lipid rafts during theearly interactions of the virus with the MA104 cells (Isa et al.,2004). The need for various cell molecules organized in a precisefashion in lipid microdomains has been suggested to contributeto the selective cell and tissue tropism of these viruses (Lopezand Arias, 2004). The entry pathway enhanced by heat shock inBHK cells seems to be similar, if not the same, used in MA104cells, since antibodies to hsc70 and cholesterol depletion withcyclodextrin had the same effect on the virus infectivity of bothcell lines.
Profound changes in the patterns of gene expression,cytoskeleton organization, and membrane fluidity are inducedby heat shock (Pouchelet et al., 1983; Vigh et al., 1998; Wanget al., 1998), including the synthesis of a characteristic setof proteins referred to, collectively, as heat shock proteins(Morimoto, 1993). In principle, any of these alterations could beresponsible for the increased susceptibility of heat-treated BHKcseelil
irus and cell receptors take place. After their initial attachmento cells via either a sialic acid containing receptor or integrin2�1 (Hewish et al 2000; Lopez and Arias, 2004; Zarate etl., 2000a, 2004), rotavirus particles have been proposed tonteract with integrins �x�2, �v�3, �4�1, and protein hsc70,nteractions that might be important for the virus to penetratento the cell’s interior (Graham et al., 2003; Lopez and Arias,
ells to rotavirus infection. The potential redistribution of cellurface rotavirus receptors into cholesterol- and sphingolipid-nriched microdomains could also be involved in the observedffect. Further studies are required to determine the cellu-ar factors involved in the facilitation of rotavirus infectionnto heat-shocked BHK cells both at the entry and post-entryevels.
82 T. Lopez et al. / Virus Research 121 (2006) 74–83
Acknowledgments
We thank Pedro Romero for his technical assistance in thepreparation of rotavirus DLPs and TLPs. This work was partiallysupported by grants 55003662 and 55000613 from the HowardHughes Medical Institute, and G37621N from the NationalCouncil for Science and Technology-Mexico.
References
Andrade, A.F., Carvalho, M.G.C., 1993. Metabolic and morphologicalchanges in Aedes albopictus cells infected with Mayaro virus under heat-shock conditions. Arch. Virol. 131, 101–114.
Andrews, J.M., Oglesbee, M.J., Lairmore, M.D., 1998. The effect of thecellular stress response on human T-lymphotropic virus type I envelopeprotein expression. J. Gen. Virol. 79, 2905–2908.
Andrews, J.M., Oglesbee, M.J., Trevino, A.L., Guyot, D.J., Newbound, G.C.,Lairmore, M.D., 1995. Enhanced human T-cell lymphotropic virus TypeI expression following induction of the cellular stress response. Virology208, 816–820.
Angelidis, C.E., Lazaridis, I., Pagoulatos, G.N., 1988. Specific inhibition ofsimian virus 40 protein synthesis by heat and arsenite treatment. Eur. J.Biochem. 172, 27–34.
Bass, D.M., Baylor, M.R., Chen, C., Mackow, E.M., Bremont, M., Green-berg, H.B., 1992. Liposome-mediated transfection of intact viral particlesreveals that plasma membrane penetration determines permissivity of tis-sue culture cells to rotavirus. J. Clin. Invest. 90, 2313–2320.
Burch, A.D., Weller, S.K., 2005. Herpes simplex virus type 1 DNA poly-
C
C
C
C
C
C
E
E
E
G
G
Guerrero, C.A., Bouyssounade, D., Zarate, S., Isa, P., Lopez, T., Espinosa,R., Romero, P., Mendez, E., Lopez, S., Arias, C.F., 2002. Heat shockcognate protein 70 is involved in rotavirus cell entry. J. Virol. 76, 4096–4102.
Guerrero, C.A., Mendez, E., Zarate, S., Isa, P., Lopez, S., Arias, C.F., 2000a.Integrin avb3 mediates rotavirus cell entry. Proc. Natl. Acad. Sci. U.S.A.97, 14644–14649.
Guerrero, C.A., Zarate, S., Corkidi, G., Lopez, S., Arias, C.F., 2000b. Bio-chemical characterization of rotavirus receptors in MA104 cells. J. Virol.74, 9362–9371.
Guo, C.T., Nakagomi, O., Mochizuki, M., Ishida, H., Kiso, M., Ohta, Y.,Suzuki, T., Miyamoto, D., Hidari, K.I., Suzuki, Y., 1999. GangliosideGM(1a) on the cell surface is involved in the infection by humanrotaviruses KUN and MO strains. Biochem. J. 126, 683–688.
Hewish, M., Takada, Y., Coulson, B.S., 2000. Integrins a2b1 and a4b1 canmediate SA11 rotavirus attachment and entry into cells. J. Virol. 74,228–236.
Isa, P., Realpe, M., Romero, P., Lopez, S., Arias, C.F., 2004. Rotavirus RRVassociates with lipid membrane microdomains during cell entry. Virology322, 370–381.
Jindadamrongwech, S., Thepparit, C., Smith, D.R., 2004. Identification ofGRP 78 (BiP) as a liver cell expressed receptor element for dengue virusserotype 2. Arch. Virol. 149, 915–927.
Lizano, M., Lopez, S., Arias, C.F., 1991. The amino-terminal half of rotavirusSA114fM VP4 protein contains a hemagglutination domain and primesfor neutralizing antibodies to the virus. J. Virol. 65, 1383–1391.
Lopez, S., Arias, C.F., 2004. Multistep entry of rotavirus into cells: a Ver-saillesque dance. Trends Microbiol. 12, 271–278.
M
M
M
O
P
P
R
R
S
S
T
V
V
merase requires the mammalian chaperone hsp90 for proper localizationto the nucleus. J. Virol. 79, 10740–10749.
hromy, L.R., Pipas, J.M., Garcea, R.L., 2003. Chaperone-mediated in vitroassembly of Polyomavirus capsids. Proc. Natl. Acad. Sci. U.S.A. 100,10477–10482.
iarlet, M., Crawford, S.E., Cheng, E., Blutt, S.E., Rice, D.A., Bergelson,J., Estes, M.K., 2002. VLA-2 (a2b1) integrin promotes rotavirus entryinto cells but is not necessary for rotavirus attachment. J. Virol. 76,1109–1123.
onti, C., De Marco, A., Mastromarino, P., Tomao, P., Santoro, M.G., 1999.Antiviral effect of hyperthermic treatment in rhinovirus infection. Antimi-crob. Agents Chemother. 43, 822–829.
oulson, B.S., Londrigan, S.L., Lee, D., 1997. Rotavirus contains integrinligand sequences and a disintegrin-like domain that are implicated in virusentry into cells. Proc. Natl. Acad. Sci. U.S.A. 94, 5389–5394.
uadras, M.A., Arias, C.F., Lopez, S., 1997. Rotaviruses induce an earlymembrane permeabilization of MA104 cells and do not require a lowintracellular Ca2+ concentration to initiate their replication cycle. J. Gen.Virol. 71, 9065–9074.
uadras, M.A., Feigelstock, D.A., Greenberg, H.B., 2002. Gene expres-sion pattern in Caco-2 cells following rotavirus infection. J. Virol. 76,4467–4482.
spejo, R.T., Lopez, S., Arias, C.F., 1980a. Structural polypeptides of simianrotavirus SA11 and the effect of trypsin. J. Virol. 37, 156–160.
spejo, R.T., Martınez, E., Lopez, S., Munoz, O., 1980b. Different polypep-tide composition of two human rotavirus types. Infect. Immun. 28,230–237.
stes, M.K., 2001. Rotaviruses and their replication. In: Knipe, D.N., Howle,P.M. (Eds.), Virology. Lippincott Williams & Wilkins, New York, pp.1747–1785.
lotzer, J.B., Saltik, M., Chiocca, S., Michou, A.I., Moseley, P., Cotten, M.,2000. Activation of heat-shock response by an adenovirus is essential forvirus replication. Nature 407, 207–211.
raham, K.L., Halasz, P., Tan, Y., Hewish, M.J., Takada, Y., Mackow, E.R.,Robinson, M.K., Coulson, B.S., 2003. Integrin-using rotaviruses bindalpha2beta1 integrin alpha2 I domain via VP4 DGE sequence and rec-ognize alphaXbeta2 and alphaVbeta3 by using VP7 during cell entry. J.Virol. 77, 9969–9978.
endez, E., Arias, C.F., Lopez, S., 1993. Binding to sialic acids is not anessential step for the entry of animal rotaviruses to epithelial cells inculture. J. Virol. 67, 5253–5259.
orimoto, R.I., 1993. Cells in stress: transcriptional activation of heat shockgenes. Science 259, 1409–1410.
ossel, E.C., Ramig, R.F., 2003. A lymphatic mechanism of rotavirusextraintestinal spread in the neonatal mouse. J. Virol. 77, 12352–12356.
glesbee, M.J., Kenney, H., Kenney, T., Krakowka, S., 1993. Enhancedproduction of morbillivirus gene-specific RNAs following induction ofthe cellular response in stable persistent infection. Virology 192, 556–567.
erez-Vargas, J., Romero, P., Lopez, S., Arias, C.F., 2006. The peptide-binding and ATPase domains of recombinant hsc70 are required to interactwith rotavirus and reduce its infectivity. J. Virol. 80, 3322–3331.
ouchelet, M., St-Pierre, E., Bibor-Hardy, V., Simard, R., 1983. Localizationof the 70,000 Dalton heat-induced protein in the nuclear matrix of BHKcells. Exp. Cell. Res. 149, 451–459.
olsma, M.D., Kuhlenschmind, T.B., Gelberg, H.B., Kuhlenschmidt, M.S.,1998. Structure and function of a ganglioside receptor for porcinerotavirus. J. Virol. 72, 9079–9091.
osas, S.L., Herculano, S., Carvalho Mda, G., 1997. Distribution of mayarovirus RNA in polysomes during heat shock. Virus. Res. 48, 133–141.
anchez-San Martın, C., Lopez, T., Arias, C.F., Lopez, S., 2004. Characteri-zation of rotavirus cell entry. J. Virol. 78, 2310–2318.
imons, K., Ikonen, E., 1997. Functional rafts in cell membranes. Nature387, 569–572.
riantafilou, K., Fradelizi, D., Wilson, K., Triantafilou, M., 2002. GRP78, acoreceptor for coxsackievirus A9, interacts with major histocompatibilitycomplex class I molecules which mediate virus internalization. J. Virol.76, 633–643.
asconcelos, D., Norrby, E., Oglesbee, M., 1998. The cellular stress responseincreases measles virus-induced cytopathic effect. J. Gen. Virol. 79,1769–1773.
igh, L., Maresca, B., Harwood, J.L., 1998. Does the membrane’s physi-cal state control the expression of heat shock and other genes? TrendsBiochem. Sci. 23, 369–374.
T. Lopez et al. / Virus Research 121 (2006) 74–83 83
Wang, T.T., Chiang, A.S., Chu, J.J., Cheng, T.J., Chen, T.M., Lai, Y.K.,1998. Concomitant alterations in distribution of 70 kDa heat shock protein,cytoskeleton and organelles in heat shocked 9L cells. Int. J. Biochem.Cell. Biol. 30, 745–759.
Xu, A., Bellamy, A.R., Taylor, J.A., 1998. BiP (GRP78) and endoplasmin(GRP94) are induced following rotavirus infection and bind transientlyto an endoplasmic reticulum-localized virion component. J. Virol. 72,9865–9872.
Zarate, S., Espinosa, R., Romero, P., Mendez, E., Arias, C.F., Lopez, S.,2000b. The VP5 domain of VP4 can mediate attachment of rotavirusesto cells. J. Virol. 74, 593–599.
Zarate, S., Romero, P., Espinosa, R., Arias, C.F., Lopez, S., 2004. VP7 medi-ates the interaction of rotaviruses with integrin alphavbeta3 through anovel integrin-binding site. J. Virol. 78, 10839–10847.
