【病毒外文文献】2004 Human Coronavirus 229E Binds to CD13 in Rafts and Enters the Cell through Caveolae

JOURNAL OF VIROLOGY Aug 2004 p 8701 8708 Vol 78 No 16 0022 538X 04 08 00H110010 DOI 10 1128 JVI 78 16 8701 8708 2004 Copyright 2004 American Society for Microbiology All Rights Reserved Human Coronavirus 229E Binds to CD13 in Rafts and Enters the Cell through Caveolae Ryuji Nomura 1 Asuka Kiyota 2 Etsuko Suzaki 2 Katsuko Kataoka 2 Yoshihide Ohe 3 Kaoru Miyamoto 4 Takao Senda 1 and Toyoshi Fujimoto 5 Department of Anatomy I Fujita Health University School of Medicine Toyoake 470 1192 1 Department of Histology and Cell Biology Hiroshima University Graduate School of Biomedical Sciences Hiroshima 734 8551 2 Biosignal Research Center Institute for Molecular and Cellular Regulation Gunma University Maebashi 371 8512 3 Department of Biochemistry Fukui Medical University Matsuoka 910 1193 4 and Department of Anatomy and Molecular Cell Biology Nagoya University Graduate School of Medicine Nagoya 466 8550 5 Japan Received 14 May 2003 Accepted 11 May 2004 CD13 a receptor for human coronavirus 229E HCoV 229E was identified as a major component of the Triton X 100 resistant membrane microdomain in human fibroblasts The incubation of living fibroblasts with an anti CD13 antibody on ice gave punctate labeling that was evenly distributed on the cell surface but raising the temperature to 37 C before fixation caused aggregation of the labeling The aggregated labeling of CD13 colocalized with caveolin 1 in most cells The HCoV 229E virus particle showed a binding and redistribution pattern that was similar to that caused by the anti CD13 antibody the virus bound to the cell evenly when incubated on ice but became colocalized with caveolin 1 at 37 C importantly the virus also caused sequestration of CD13 to the caveolin 1 positive area Electron microscopy confirmed that HCoV 229E was localized near or at the orifice of caveolae after incubation at 37 C The depletion of plasmalemmal cholesterol with methyl H9252 cyclodextrin significantly reduced the HCoV 229E redistribution and subsequent infection A caveolin 1 knockdown by RNA interference also reduced the HCoV 229E infection considerably The results indicate that HCoV 229E first binds to CD13 in the Triton X 100 resistant microdomain then clusters CD13 by cross linking and thereby reaches the caveolar region before entering cells Recent studies have revealed that the plasma membranes of cells contain microdomains with discrete molecular composi tions Rafts are sphingolipid and cholesterol rich membrane microdomains that are thought of as platforms for signal trans duction 39 40 Although there are still many controversies regarding how rafts exist in living cells it is generally agreed that cholesterol is indispensable for their integrity and that the detergent resistant membrane DRM fraction is the in vitro correlate of the raft Because acyl chains of sphingolipids and glycosylphosphatidylinositol GPI anchored proteins en riched in the DRM fraction are more highly saturated than those of glycerolipids in the bulk membrane the raft domain is thought to show less fluidity than nonraft areas of the plasma membrane However it is difficult to capture rafts morpholog ically because their shape and size are likely to change dynam ically 40 On the other hand caveolae were first defined morpholog ically as invaginations of the plasma membrane 49 They are also susceptible to cholesterol depletion 31 Moreover caveolin 1 2 and 3 which were identified as major compo nents of caveolae 31 35 44 47 are highly enriched in the DRM fraction 2 12 14 36 Several results suggest that many molecules are shared by rafts and caveolae but that at least several molecules that are enriched in the DRM fraction are not concentrated in caveolae 11 Thus caveolae are not sim ply a stabilized form of rafts but there should be a regulatory mechanism as yet unknown to control the molecular distri bution between caveolae and rafts It has been shown that cross linked raft molecules such as GPI anchored proteins glycolipids sphingomyelin and cho lesterol are sequestered to caveolae 13 25 The underlying mechanism is not clearly defined but the cross linking may combine unstable small rafts and give rise to stable large rafts 4 the latter may somehow become hooked to caveolae The phenomenon is an indication of the close relationship between rafts and caveolae but its physiological and pathological rele vance has not been clear In searching for a new molecule that is enriched in the DRM fraction we found that CD13 or aminopeptidase N is its predominant component in human fibroblasts CD13 has been known to be a receptor for human coronavirus 229E HCoV 229E 50 Considering the behavior of cross linked raft mol ecules as described above we assumed that the HCoV 229E particle could work as a polyvalent ligand and cause a similar redistribution of CD13 as that induced by anti CD13 antibod ies In the present study we indeed observed that the virus particle initially binding to the cell surface apparently in a random fashion became aggregated to caveolae in a choles terol dependent manner Furthermore cholesterol depletion as well as caveolin 1 knockdown inhibited virus entry into the cells The results indicate that the integrity of the sphingolipid and cholesterol rich microdomain is indispensable for infec tion by HCoV 229E and suggest that manipulation of the membrane lipids could be used as a preventive measure Corresponding author Mailing address Department of Anatomy I Fujita Health University School of Medicine Toyoake Aichi 470 1192 Japan Phone 81 562 93 2648 Fax 81 562 93 5904 E mail rnomura fujita hu ac jp 8701 on April 2 2015 by UNIVERSITY OF STRATHCLYDE http jvi asm org Downloaded from MATERIALS AND METHODS Cells and virus Human fibroblasts were explanted from biopsied normal human adult skin A human embryonic lung cell line L132 and HCoV 229E were kindly donated by Ichiro Matsumoto Iwate Medical University Morioka Japan The cells were grown in Dulbecco s minimum essential medium DMEM Nihonseiyaku Tokyo Japan supplemented with either 10 fetal calf serum for human fibroblasts or 10 bovine serum for L132 cells 50 U of penicillin ml and 0 05 mg of streptomycin ml at 37 C For preparation of a virus solution L132 cells were incubated with HCoV 229E for1hat33 C and cultured at the same temperature for 2 days The total solution was subjected to three cycles of freezing thawing and the supernatant was stored at H1100280 C The virus solution was concentrated by the method used for Mouse hepatitis virus a member of the Coronaviridae 43 Antibodies A rabbit anti HCoV 229E antiserum 24 was also provided by Ichiro Matsumoto A mouse anti coronavirus antibody Chemicon International Inc Temecula Calif mouse anti caveolin 1 antibody clones 2234 and 2297 Transduction Laboratories Lexington Ky rabbit anti caveolin 1 antibody Santa Cruz Biotechnology Santa Cruz Calif mouse anti CD13 antibody clone CLB mon gran 2 Q20 Research Diagnostics Inc Flanders N J rabbit anti CD13 antibody Santa Cruz Biotechnology mouse anti human transferrin receptor TfR or CD71 antibody clone RVS 10 Cymbus Biotechnology Ltd Hants United Kingdom rabbit anti CD71 antibody Santa Cruz Biotechnol ogy fluorescein isothiocyanate and rhodamine conjugated anti rabbit immu noglobulin G IgG and anti mouse IgG antibodies Jackson ImmunoResearch Laboratories West Grove Pa Alexa 488 and Alexa 568 conjugated anti rabbit IgG and anti mouse IgG antibodies Molecular Probes Inc Eugene Oreg and horseradish peroxidase conjugated anti rabbit IgG and anti mouse IgG an tibodies Jackson ImmunoResearch Laboratories were purchased The speci ficities of the anti HCoV 229E and anti CD13 antibodies were confirmed by Western blotting and immunoprecipitation data not shown Isolation of DRMs and peptide sequencing Isolation of the DRM fraction was performed by the method of Hope and Pike 18 with minor modifications Human fibroblasts were treated with 1 Triton X 100 in TNE 25 mM Tris 150 mM NaCl 5 mM EDTA 1 mM phenylmethylsulfonyl fluoride pH 7 5 for 20 min on ice and centrifuged at 200 000 H11003 g for1hat4 C The pellet was suspended in 1 Triton X 100 in TNE and homogenized with a Dounce ho mogenizer The solution was centrifuged suspended again in 1 Triton X 100 in TNE and mixed with an equal volume of 80 sucrose in TNE The mixture was overlaid with 38 sucrose in TNE and 5 sucrose in TNE and centrifuged at 270 000 H11003 g for 15 to 20 h at 4 C A light scattering band seen at the boundary between the 38 and 5 sucrose solutions was diluted four to fivefold with TNE and centrifuged at 200 000 H11003 g for1hat4 C The pellet was dissolved separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis SDS PAGE with a 5 acrylamide gel and electrotransferred to a polyvinylidene difluoride mem brane A 160 kDa band was digested with lysyl endopeptidase and subjected to amino acid sequencing To assess the enrichment of CD13 in the DRM we subjected the total cell lysate in 1 Triton X 100 TNE to sucrose density gradient ultracentrifugation and took six fractions by puncturing the bottom of the tube The proteins were precipitated in cold acetone dissolved in SDS sample buffer and analyzed by Western blotting Immunofluorescence microscopy To observe the localization of CD13 and TfR we treated human fibroblasts cultured on glass coverslips on ice with an anti CD13 or anti TfR antibody for 30 min and then with a fluorochrome conjugated secondary antibody for 30 min rinsed the cells and incubated them for 0 to 60 min at 37 C After being fixed with 3 formaldehyde in 0 1 M sodium phosphate buffer for 10 min the cells were permeabilized with 0 1 Triton X 100 for 5 min treated with 3 bovine serum albumin for 10 min and labeled for caveolin 1 To keep track of HCoV 229E we allowed human fibroblasts to bind to HCoV 229E on ice rinsed the cells and incubated them for 0 to3hat37 C The cells were fixed and treated in the same manner as described above and were doubly labeled for HCoV 229E or CD13 in combination with caveolin 1 To quantify the degree of colocalization we took images of 20 cells with a Zeiss LSM510 laser confocal scanning microscope under the same conditions The colocalization ratio was obtained by dividing the number of pixels with signals for both caveo lin 1 and HCoV 229E above a certain threshold level by the number of pixels with an HCoV 229E signal The experiment was performed three times and representative results are shown below For some experiments to examine the influence of cholesterol depletion we treated cells with 2 mM methyl H9252 cyclodextrin MH9252CD Sigma Chemical Co St Louis Mo in DMEM for 30 min at 37 C before incubating them with HCoV 229E To exclude the possibility of irreversible damage caused by the above treatment we further incubated some cells with an MH9252CD cholesterol complex for 30 min at 37 C to replenish cholesterol MH9252CD cholesterol was prepared as described previously 16 briefly cholesterol in methanol chloroform 1 1 was dried and suspended in MH9252CD in DMEM 0 1H11003 phosphate buffered saline The solution was sonicated and rotated at 37 C overnight After the pH was adjusted to 7 4 the mixture was filtered and diluted with DMEM The final concentrations were about 1 mM MH9252CD and 100 H9262g of cholesterol ml Western blotting Total cell lysates and fractions from sucrose density gradient ultracentrifugation were solubilized in SDS sample buffer electrophoresed in 11 acrylamide gels and transferred to polyvinylidene difluoride membranes The blots were incubated with primary antibodies followed by a horseradish peroxidase conjugated secondary antibody and the reaction was visualized with SuperSignal West Dura Extended Duration substrate Pierce Rockford Ill Electron microscopy Human fibroblasts were incubated with HCoV 229E for 60 min on ice rinsed and kept at 37 C for 0 to 3 h The cells were fixed with 2 5 glutaraldehyde in 0 1 M sodium phosphate buffer for 30 min on ice incubated with osmium tetroxide stained en bloc with uranyl acetate and embedded in Quetol 812 Ultrathin sections were counterstained with uranyl acetate and lead citrate To quantify the accumulation of HCoV 229E in caveolae we measured the proportion of HCoV 229E localized within 150 nm of the orifices of caveolae by using printed electron micrographs Reverse transcriptase PCR RT PCR Human fibroblasts bound to HCoV 229E were treated with 0 25 trypsin and 0 02 EDTA in PBS to remove viruses adhering to the cell surface and the total RNA was extracted by the use of MagExtractor RNA Toyobo Co Ltd Osaka Japan The total RNA was reverse transcribed with ReverTra Ace Toyobo Co Ltd and an oligo dT 20 primer and was amplified with the following primer sets 5H11032 ATGTTCCTTAA GCTAGTGGATGA 3H11032 and 5H11032 TTAGAAATCAATAACTCGTTTAG 3H11032 for the E protein coding region of HCoV 229E RNA and 5H11032 ACCACAGTCCATG CCATCAC 3H11032 and 5H11032 TCCACCACCCTGTTGCTGTA 3H11032 for glyceraldehyde 3 phosphate dehydrogenase G3PDH mRNA To estimate the extent of intracel lular entry of HCoV 229E we performed PCRs with 32 and 35 amplification cycles for the HCoV 229E E protein and with 22 and 25 cycles for G3PDH Knockdown of caveolin 1 by RNA interference A double stranded small in terfering RNA siRNA designed to target the human caveolin 1 mRNA 14 was purchased from Dharmacon Research Lafayette Colo A control IX siRNA Dharmacon Research was used as a control Human fibroblasts were transfected with annealed siRNAs by the use of Oligofectamine Invitrogen Carlsbad Calif and were subcultured 24 h later Forty eight hours after being subcultured the cells were bound with HCoV 229E on ice rinsed and incubated for another 48 h at 37 C The cells were either fixed for immunofluorescence microscopy or processed for Western blotting and the effect of the caveolin 1 knockdown on the expression of the coronavirus N protein was estimated To measure the proportion of cells expressing the N protein by microscopy we incorporated 10 random fields obtained by laser confocal scanning microscopy The experiment was performed three times RESULTS Identification of CD13 in DRM of human fibroblasts Hu man fibroblasts were treated with 1 Triton X 100 in TNE on ice and then subjected to sucrose density gradient ultracentrif ugation Triton X 100 insoluble material floating at the bound ary of the 5 and 38 sucrose solutions was recovered and analyzed by SDS PAGE Silver staining visualized 18 bands Fig 1A A prominent band at 160 kDa was identified as CD13 by amino acid sequencing of two peptides and by data base searching FASTA Genome Net http fasta genome ad jp Fig 1A To examine the extent of CD13 enrichment in the DRM we analyzed the fractions obtained by sucrose density gradient ultracentrifugation by Western blotting Fig 1B CD13 as well as caveolin 1 was enriched in the DRM fraction whereas TfR a nonraft membrane molecule and H9251 tubulin a cytosolic protein were recovered in the bottom fraction Fig 1B This result indicates a marked enrichment of CD13 in the DRM Immunofluorescence microscopy of CD13 When living hu man fibroblasts were incubated with an anti CD13 antibody 8702 NOMURA ET AL J VIROL on April 2 2015 by UNIVERSITY OF STRATHCLYDE http jvi asm org Downloaded from followed by a fluorochrome conjugated secondary antibody on ice and fixed without warming labeling was observed evenly on the cell surfaces Fig 2a in the same cells labeling for caveo lin 1 was seen in patches Fig 2b and c When cells bound with the antibodies were incubated for 10 min at 37 C the labeling of CD13 became aligned along longitudinal lines Fig 2d most of the labeling of CD13 was not correlated with that of caveolin 1 but some labeling of CD13 in the cell periphery colocalized with caveolin 1 Fig 2d to f After incubation for 30 min at 37 C the labeling of CD13 showed clustering and colocalization with caveolin 1 became more frequent Fig 2g to i After 60 min of incubation at 37 C CD13 and caveolin 1 colocalized extensively Fig 2j to l The accumulation of cross linked molecules in caveolin 1 positive areas was similar to the previously reported behavior of GPI anchored proteins and glycosphingolipids 13 25 In contrast TfR did not show colocalization with caveolin 1 even after cross linking TfR was seen evenly on the cell surface without warming Fig 2m and after 60 min of incubation at 37 C although it showed slight clustering it did not colocalize with caveolin 1 Fig 2p to r Immunofluorescence microscopy of HCoV 229E redistribu tion CD13 is known to be a receptor of HCoV 229E 50 Since the virus particle is supposed to be a polyvalent ligand for CD13 we examined whether it could change the distribution of CD13 in a manner similar to that of the anti CD13 antibody When human fibroblasts were incubated with HCoV 229E on ice and fixed without warming HCoV 229E particles were seen as randomly distributed dots and did not show a correla tion with caveolin 1 in most cells Fig 3a to c the proportion FIG 1 Identification of CD13 as a major DRM fraction of human fibroblasts A Total lysates and DRMs of human fibroblasts were subjected to SDS PAGE 7 to 16 gel and silver staining Eighteen bands 160 125 107 97 86 70 58 49 46 41 38 35 31 29 28 22 21 and 16 kDa were detected arrow and arrowheads Two peptides peptides 1 and 2 obtained from the prominent 160 kDa band arrow were subjected to amino acid sequencing In peptide 1 8 of 10 amino acids agreed with amino acids 231 to 239 of CD13 and in peptide 2 all 12 amino acids agreed with amino acids 924 to 935 of CD13 B Frac tions obtained by sucrose density gradient ultracentrifugation of the total cell lysate treated with Triton X 100 were subjected to immuno blotting for CD13 TfR caveolin 1 and H9251 tubulin CD13 and caveo lin 1 were enriched in DRMs but TfR and H9251 tubulin were detected in the bottom fraction FIG 2 Immunofluorescence microscopy Human fibroblasts were treated with an anti CD13 antibody and a fluorochrome conjugated secondary antibody on ice and incubated at 37 C for 0 min a to c 10 min d to f 30 min g to i or 60 min j to l After fixation the cells were labeled with an anti caveolin 1 antibody Labeling for caveolin 1 was seen as patches in all samples b e h and k Labeling for CD13 was observed evenly on the cell surface without warming a after 10 min of incubation at 37 C most of the labeling showed a linear align ment d after 30 min at 37 C CD13 showed clustering and colocal ization with caveolin 1 became apparent g to i arrows and after 60 min at 37 C CD13 and caveolin 1 were colocalized extensively j to l arrows The cells were treated with the anti TfR antibody and the secondary antibody in the same manner as that described above La beling for TfR was observed evenly on the cell surface without warm ing m to o and even after 60 min TfR and caveolin 1 did not show colocalization p to r Bar H11005 50 H9262m VOL 78 2004 HUMAN CORONAVIRUS 229E ENTERS THROUGH CAVEOLAE 8703 on April 2 2015 by UNIVERSITY OF STRATHCLYDE http jvi asm org Downloaded from of cells exhibiting extensive colocalization of HCoV 229E and caveolin 1 was 6 4 When cells bound with HCoV 229E were incubated for1hat37 C the frequency of colocalization of HCoV 229E and caveolin 1 increased markedly Fig 3d to f although less conspicuous than CD13 cross linked by antibod ies HCoV 229E also showed a linear alignment Fig 3d After3hofincubation at 37 C the proportion of cells showing colocalization of HCoV 229E and caveolin 1 was 31 2 Fig 3g to l We also examined the behavior of CD13 when cells were bound with HCoV 229E When cells were incubated with HCoV 229E on ice and fixed without warming labeling of CD13 was observed evenly on the cell surfaces Fig 3m to o Binding of HCoV 229E and incubation at 37 C caused clus tering of CD13 and its colocalization with caveolin 1 These results indicate that CD13 bound by HCoV 229E behaves in the same manner as that cross linked with the anti CD13 an tibody Fig 3p to r Electron microscopy of HCoV 229E redistribution We e