Recombinant Extracellular Domains of ... - Journal of Virology

2 downloads 0 Views 781KB Size Report
Dec 5, 2005 - antibodies which bind to the tetraspanin protein CD63, but not by antibodies ... domain (EC2) of human tetraspanins CD9, CD63, CD81, and ...
JOURNAL OF VIROLOGY, July 2006, p. 6487–6496 0022-538X/06/$08.00⫹0 doi:10.1128/JVI.02539-05 Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Vol. 80, No. 13

Recombinant Extracellular Domains of Tetraspanin Proteins Are Potent Inhibitors of the Infection of Macrophages by Human Immunodeficiency Virus Type 1 Siu-Hong Ho,1 Francine Martin,2,3 Adrian Higginbottom,2 Lynda J. Partridge,3 Varadarajan Parthasarathy,3 Gregory W. Moseley,3§ Peter Lopez,1 Cecilia Cheng-Mayer,1*† and Peter N. Monk2† Aaron Diamond AIDS Research Center, The Rockefeller University, 455 First Avenue, 7th Floor, New York, New York 100161; Academic Neurology Unit, University of Sheffield Medical School, Sheffield S10 2RX, United Kingdom2; and Department of Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, United Kingdom3 Received 5 December 2005/Accepted 11 April 2006

Human immunodeficiency virus type 1 (HIV-1) infection of human macrophages can be inhibited by antibodies which bind to the tetraspanin protein CD63, but not by antibodies that bind to other members of the tetraspanin family. This inhibitory response was limited to CCR5 (R5)-tropic virus and was only observed using macrophages, but not T cells. Here, we show that recombinant soluble forms of the large extracellular domain (EC2) of human tetraspanins CD9, CD63, CD81, and CD151 produced as fusion proteins with glutathione S-transferase (GST) can all potently and completely inhibit R5 HIV-1 infection of macrophages with 50% inhibitory concentration values of 0.11 to 1.2 nM. Infection of peripheral blood mononuclear cells could also be partly inhibited, although higher concentrations of EC2 proteins were required. Inhibition was largely coreceptor independent, as macrophage infections by virions pseudotyped with CXCR4 (X4)-tropic HIV-1 or vesicular stomatitis virus (VSV)-G glycoproteins were also inhibited, but was time dependent, since addition prior to or during, but not after, virus inoculation resulted in potent inhibition. Incubation with tetraspanins did not decrease CD4 or HIV-1 coreceptor expression but did block virion uptake. Colocalization of fluorescently labeled tetraspanin EC2 proteins and HIV-1 virions within, and with CD4 and CXCR4 at the cell surfaces of, macrophages could be detected, and internalized tetraspanin EC2 proteins were directed to vesicular compartments that contained internalized dextran and transferrin. Collectively, the data suggest that the mechanism of inhibition of HIV-1 infection by tetraspanins is at the step of virus entry, perhaps via interference with binding and/or the formation of CD4-coreceptor complexes within microdomains that are required for membrane fusion events. The tetraspanins are a family of proteins (33 in Homo sapiens) that have fundamental roles in the cell biology of multicellular organisms, regulating cell activation, proliferation, development, motility, and adhesion (reviewed in reference 5). Tetraspanins have four transmembrane domains linked by short intracellular loops and one small (EC1) and one large (EC2) extracellular domain. The EC2 domain contains characteristic C-containing motifs (CCG, PXSC, and EGC) that identify family members; disulfide linkages formed by these (and up to four additional) C residues are important in the conservation of subloop structure within EC2 (12). The integrity of these structures has been demonstrated to be essential for at least some of the biological activities of tetraspanins. For example, both the binding of hepatitis C virus E2 glycoprotein to CD81 (21) and the inhibition of sperm/egg fusion by recombinant, soluble forms of CD9 EC2 (6) are abrogated by mutation of Cys residues in these motifs. Additionally, the binding of many of the anti-tetraspanin antibodies that can perturb cellular

function is also sensitive to disruption of the disulfide linkages, by either reduction or mutation. However, regions of the EC2 domain outside of the constrained subloops are also important for tetraspanin function, in particular, formation by the N- and C-terminal regions that form a helical stalk probably involved in homo- and heterodimer formation (12, 26). Multimerization of tetraspanins is thought to be a key step in the formation of tetraspanin-enriched microdomains (TEM), in which tetraspanins act as adaptor proteins bringing together a diverse array of other membrane proteins (e.g., immune cell receptors and coreceptors—Fc receptors, MHC proteins, integrins, CD4/ CD8, TCR, and BCR; growth factors and their receptors— EGFR, HB-EGF, and TGF␣; and adhesion molecules—␤1 and ␤2 integrins) into functional clusters (1). Previous work has suggested roles for tetraspanins in the life cycles of human T-cell leukemia virus 1 (9), canine distemper virus (14), and feline leukemia virus (29), and more recently, reports have identified tetraspanin CD63 as possibly playing multiple important roles in human immunodeficiency virus type 1 (HIV-1) infection. CD63-positive multivesicular bodies accumulate Gag (18) and are sites of virion assembly in macrophages (20, 25); multivesicular bodies also contain tetraspanins CD81 and CD82. CD63 is highly enriched in the envelopes of newly budded virus particles (16, 19); in fact, anti-CD63 antibodies could immunoprecipitate nearly 100% of virus,

* Corresponding author. Mailing address: Aaron Diamond AIDS Research Center, The Rockefeller University, 455 First Avenue, 7th Floor, New York, NY 10016. Phone: (212) 448-5080. Fax: (212) 4485159. E-mail: [email protected]. † C. Cheng-Mayer and P. N. Monk contributed equally to this work. § Present address: Department of Biochemistry and Molecular Biology, Monash University, Clayton, Victoria 3080, Australia. 6487

6488

HO ET AL.

whereas other anti-tetraspanin antibodies (CD81, CD82, CD53, and CD151) were much less efficient (20). Treatment of macrophages with anti-CD63 antibody has also been shown to inhibit CCR5 (R5)- but not CXCR4 (X4)-tropic virus infection in a cell type-specific-manner; macrophage, but not peripheral blood mononuclear cell (PBMC), infection was shown to be sensitive (28). Here, we report the use of a new set of tetraspanin tools in HIV-1 research. Recombinant EC2 domains from CD9, CD63, CD81, and CD151, expressed as glutathione S-transferase (GST) fusion proteins, are capable of potently and completely inhibiting both R5 and X4 virus infection of macrophages, but the effect on infection of PBMCs is more moderate. Receptorindependent inhibition by tetraspanins was further demonstrated by the finding that infection mediated by vesicular stomatitis virus (VSV)-G was similarly affected. The mechanism of inhibition by tetraspanins appeared, at least in part, to be mediated through interference with virus entry. MATERIALS AND METHODS Preparation of PBMC and monocyte-derived macrophage cultures. PBMCs were prepared by Ficoll gradient centrifugation, stimulated with phytohemagglutinin (3 ␮g/ml; Sigma, St. Louis, MO) in RPMI 1640 medium containing 10% fetal calf serum and 20 U interleukin-2 (kindly provided by Chiron Corp., Emeryville, CA). Monocytes were enriched by centrifugation of PBMCs through a 46% Percoll cushion. The cells were then resuspended in RPMI 1640 medium supplemented with 10% fetal calf serum and 5% human AB serum. The monocytes were allowed to adhere overnight and were harvested and replated in 96-well plates at 7 ⫻ 104 cells/well for infection. For microscopic studies, the cells were replated in 35-mm glass bottom no. 1.5 poly-D-lysine dishes (Mat-Tek Corp., Ashland, MA) at 7.5 ⫻ 105 cells/plate. The monocytes were allowed to differentiate for 5 to 7 days before use. Generation of viruses. Luciferase reporter viruses pseudotyped with different envelope glycoproteins were generated by transcomplementation, as described previously (4). The reporter viruses were derived from the HIV-1 pNL4-3 proviral DNA, in which the env gene was deleted and a firefly luciferase cassette was inserted in place of the nef gene. The env constructs used were pEnv162P3, expressing a CCR5-using HIV-1 envelope (7); pEnvA2, expressing a CXCR4using HIV-1 envelope (2); and a plasmid expressing VSV-G (pVSV-G; kindly provided by J. McKeating, University of Birmingham, Birmingham, United Kingdom) (23). Because of the lack of a proviral env gene, pseudotyped viruses were capable of only a single round of replication. The viruses were generated by lipofection of 1.5 ␮g each of pNL-LucE-R⫹ plasmid and of a pEnv vector in 293T cells plated at 7 ⫻ 105 cells per well in six-well plates. The lipofection was performed with the DMRIE-C reagent according to the manufacturer’s recommendations (Gibco-BRL, Gaithersburg, Md.). Cell culture supernatants were harvested 72 h posttransfection, centrifuged at 800 ⫻ g, filtered through 0.45␮m-pore-size filters, and stored at ⫺70°C until use. The viral content was quantified by a p24 Gag enzyme-linked immunosorbent assay (Abbott Laboratories, Chicago, Ill.). To generate Vpr-eGFP or Gag-eGFP virus, 293T cells were cotransfected with a full-length NL4-3 proviral genome in which the envelope gene had been replaced with that of R5-SHIVSF162P3 (7) and an expression plasmid for VprEGFP (kindly provided by N. Landau, Salk Institute, La Jolla, CA) or Gag-eGFP (from Paul Bieniasz, ADARC, New York, NY) to generate P3-Vpr-EGFP and P3-Gag-EGFP, respectively. Culture supernatants were harvested 48 h later, centrifuged at 800 ⫻ g, filtered through 0.45-␮m-pore-size filters, and concentrated by centrifugation through a 20% sucrose cushion. The pelleted virions were resuspended in Hanks balanced salt solution, aliquoted, and stored at ⫺70°C until use. The viral content was quantified by p24 Gag enzyme-linked immunosorbent assay. Production and fluorescent labeling of GST fusion proteins. Tetraspanin EC2GST fusion proteins were expressed in BL21 codon-plus Escherichia coli transfected with the appropriate cDNA cloned into pGEX-KG, as previously described (6). The concentration of protein was measured using a Bradford assay and optical density. Protein purity was assessed using Coomassie and silver staining of sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels and was found to be similar to our previous report. The correct conformation of

J. VIROL. purified tetraspanin fusion proteins was assessed by Western blotting performed under nonreducing conditions using conformation-specific antibodies for each tetraspanin, and all EC2 proteins were recognized by relevant antibodies. To remove GST, glutathione-Sepharose beads (Amersham Biosciences, Amersham, United Kingdom) were saturated with GST-CD63 EC2 to prevent nonspecific binding of cleaved CD63 EC2 and treated with 10 units of thrombin protease (Sigma, Poole, United Kingdom) per mg of EC2 for 4 h at room temperature (RT). Thrombin was removed from the reaction buffer by incubation with paminobenzamidine agarose beads. Free GST or GST-CD63 EC2 could not be detected by silver staining of sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels after treatment. For imaging studies, GST and GST-tetraspanin EC2 proteins were labeled with fluorescein isothiocyanate (FITC) or rhodamine using conventional methods (11) and with AlexaFluor 647 (Cambridge Bioscience, United Kingdom) according to the manufacturer’s instructions. Molar dye-protein ratios were as follows: GST-FITC, 4.8:1, GST-CD9 EC2–FITC, 4.2:1; GST-CD63 EC2–FITC, 4.8:1; GST-Alexafluor 647, 3.5:1; GST-CD63 EC2–Alexafluor 647, 1.7:1; GSTrhodamine, 1.1:1; GST-CD63 EC2–rhodamine 1.2:1. Virus infection inhibition assays. Inhibition assays were performed as follows. Macrophages (7 ⫻ 104 cells per well) or PBMCs (106 cells per well) in 96-well plates were pretreated with 50 ␮l of serial dilutions of GST or GST-tetraspanin fusions for 30 to 60 min at 37°C. Control cells received Hanks balanced salt solution alone. CD4-immunoglobulin G2 (a kind gift of William Olson, Progenics Pharmaceuticals, Tarrytwon, NY) was used in some experiments as a positive control. An equal volume containing 5 ng p24 Gag equivalent of each of the pseudoytpe viruses was then added and incubated for 2 h at 37°C. At the end of the incubation period, 100 ␮l of macrophage or PBMC culture medium was added, and the culture was maintained for 72 h before being tested for luciferase activity. Cells were lysed and incubated with the luciferase assay reagents according to the manufacturer’s instructions (Promega, Madison, Wis.). The luciferase activity was measured in a Dynex MLX microtiter plate luminometer (Dynex Technologies, Inc., Chantily, Va.). All infections were performed in duplicate. For time-of-addition studies, virus was inoculated for 1 h at 37°C and then removed. GST or GST-CD63 EC2 fusions were added during the time of inoculation or at 0, 60, and 120 min after virus inoculation. Cultures were maintained for 72 h and processed as described above. Uptake of virus expressing Vpr-eGFP. Macrophages in Mat-Tek plates were treated with 50 ␮l of Hanks balanced salt solution, GST (0.4 ␮M), or GST-CD63 (0.11 or 0.33 ␮M) for 30 min at 37°C; 50 ␮l of Vpr-eGFP virus (containing 200 ng p24 Gag equivalent) was then added and incubated for various periods of time at 37°C. At the end of the incubation period, infected macrophages were washed extensively and fixed with 1% formaldehyde in phosphate-buffered saline (PBS) for 30 min at RT and then left overnight at 4°C. The next day, the cells were stained with concanavalin A-rhodamine (1 ng/ml) at RT for 15 min and washed five times with PBS. DAPI (4⬘,6⬘-diamidino-2-phenylindole) was then added, and images were obtained with Delta Vision (Applied Precision, Issaquah, WA). Fifteen independent experiments were performed using 0.11 ␮M of GST-CD63 and 0.4 ␮M GST as a control. In four of these experiments, 0.33 ␮M of GSTCD63 was also used. For each experiment, 10 random fields, each containing one cell, were selected, followed by particle counting. The average number of green particles in each cell treated with GST-CD63 and GST was then calculated and compared. This manual counting agreed with an automated measure used on a subset of the images. Surface expression of CD4, CCR5, CXCR4, CD14, and CD63. For surface protein staining, macrophages in T25 flasks were treated with 4 ␮g/ml of GST or GST-CD63 for 60 min at 37°C. Cells were collected by scraping and were washed once in PBS with 1% fetal bovine serum and 1 ␮M NaN3. CD4-phycoerythrin (PE) (clone L200), CD14-FITC (clone TuK4), CD63-FITC (clone H5C6), CCR5-PE (clone 2D7), CXCR4-PE (clone 12G5), GST-FITC (clone B-14), and fluorescently labeled isotype control antibodies were added individually to 1 ⫻ 106 pretreated macrophages for 30 min at RT, followed by washing in the same buffer once. Untreated cells were stained in parallel for reference. The stained cells were fixed in 1% formaldehyde and analyzed by FACScalibur. CD14-FITC antibody was purchased from Caltag (Burlingame, CA), GST-FITC from Santa Cruz Biotechnology (Santa Cruz, CA), and all other antibodies from Becton Dickenson. Uptake and localization of fluorescent tetraspanin protein in monocyte-derived macrophages. Confluent (day 7) cultures of adherent monocyte-derived macrophages (MDM) in T25 flasks were treated with 4 ␮g/ml of GST-FITC (0.16 ␮M) or GST-CD63 EC2–FITC (0.11 ␮M) for 1 h at 37°C. After the excess protein was washed off, the cells were harvested by scraping and fixed in 1% formaldehyde (Tousimis, Rockville, MD). Cell-associated fluorescence was measured by flow cytometry (FACScalibur; Becton Dickenson, Franklin Lakes, NJ),

VOL. 80, 2006

TETRASPANINS INHIBIT HIV-1 INFECTION OF MACROPHAGES

6489

FIG. 1. Effects of recombinant human CD63 and CD9 large extracellular domain-GST proteins on infection of macrophages and PBMCs by CCR5- and CXCR4-tropic HIV-1. Macrophages or PBMCs isolated as described in Materials and Methods were treated with different concentrations of recombinant human CD9 or CD63 large extracellular domain-GST fusion proteins, GST alone, or immunoglobulin-CD4 for 30 to 60 min prior to the addition of HIV-1 virions expressing CCR5- or CXCR4-specific Env protein. Infection was measured after 3 days as chemiluminescence from luciferase expressed under the control of the HIV-1 long terminal repeat promoter. The results are shown as percentage inhibition of infection relative to untreated control cells and are the means of at least three separate experiments ⫾ standard errors of the mean.

and the uptake of FITC-labeled protein was determined by counting green pixels associated with cells. To identify intracellular localization of internalized tetraspanin EC2 proteins, GST-CD63 EC2–FITC and tetramethylrhodamine-conjugated dextran or AlexaFluor 647-conjugated transferrin (both from Invitrogen, Carlsbad, CA) was added to cells and incubated for 30 min at 37°C. The cells were then washed and fixed, and images were recorded using DeltaVision microscopy. For tetraspanin-virion colocalization experiments, macrophages were incubated with Gag-eGFP virus and Alexafluor 647-labeled GST-CD63 EC2 for 20 min at RT, washed, and fixed for imaging. To determine whether soluble CD63 colocalizes with HIV-1 receptors at the cell surface, macrophages were incubated with GST-CD63 EC2–FITC and PE-conjugated antibodies to CD4, CXCR4, and CCR5 for 1 h at 4°C. The cells were then washed, fixed, and imaged using Delta Vision microscopy.

RESULTS Tetraspanin EC2 domains strongly inhibit HIV infection of MDM, but not PBMCs. Tetraspanin-dependent functions, such as fertilization (6), leukocyte adhesion (27), and monocyte fusion (27), that can be inhibited by anti-tetraspanin antibodies can also be potently inhibited by soluble forms of tetraspanin EC2 domains. We therefore tested a range of recombinant human tetraspanin EC2 domains produced as GST fusion proteins (GST-CD9 EC2, -CD63 EC2, -CD81 EC2, and -CD151 EC2) for inhibitory activity on R5 and X4 virus infection of MDM and PBMCs, with GST alone and GST-mouse CD9 EC2 serving as negative controls (Fig. 1 and Table 1). Infection of MDM by both R5 and X4 viruses was inhibited by

the fusion proteins, but R5 appeared to be more sensitive. CD63 EC2 was the most potent inhibitor, with CD9 EC2 being only slightly less potent. CD151 and CD81 EC2 proteins, however, were typically 5- to 10-fold less effective than CD63 and CD9, and mouse CD9 EC2 was 10- to 50-fold less potent. GST alone was also inhibitory, but with considerably lower potency than the tetraspanin EC2 fusion proteins (e.g., ⬎400-fold lower than CD63 EC2 for R5 infection of MDM). VSVpseudotyped viruses were also inhibited, with a broadly similar pattern of inhibition by the human tetraspanins, although mCD9 was ineffective. Notably, for R5 and X4 virus, all the human tetraspanin EC2 proteins were more potent inhibitors of macrophage infection than soluble CD4. In contrast, R5 virus infection of PBMCs was only partly inhibited by high concentrations of the tetraspanin EC2 proteins, whereas X4 infection was resistant at concentrations up to 10 ␮M (Fig. 1). Incomplete inhibition of VSV infection of PBMCs was also observed, and similar to R5, high concentrations were required, with all of the tetraspanins displaying approximately equal potencies, whereas mouse CD9 EC2 was ineffective (Table 1). In some experiments, fusion proteins were removed before virus was added; this made little difference to the inhibitory effect (data not shown). To examine the role of the GST fusion partner in the inhibition of infection, GST-CD63 EC2 was treated with thrombin, followed by glutathione affinity chromatography to remove vir-

8 3 4 3 2 5 8.10 ⫾ 0.11*** (7.94) 8.47 ⫾ 0.20*** (3.39) 7.86 ⫾ 0.19NS (13.8) 6.49 ⫾ 0.36NS (324) ND 6.10 ⫾ 0.31NS (794) 6 4 2 7 7 5 pIC50 ⫽ ⫺log of concentration of protein with half-maximal inhibitory effect. n, number of independent experiments performed in duplicate. For MDM, inhibition by extracellular-domain proteins reached 100% in all cases, whereas for PBMCs, maximal inhibition was only ⬃60%. d IC50 in nM. e ND, no effect detected. f Significance of difference from GST control was assessed by a two-tailed t test; ***, P ⬍ 0.001; **, P ⬍ 0.01; *, P ⬍ 0.05; NS, not significant. c

b

a

7 6 2 7 7 5 11 7 2 9 8 5 16 10 3 9 8 6

GST

c

MDM R5 MDM X4 MDM VSV PBMC R5 PBMC X4 PBMC VSV

Cell type

d

7.35 ⫾ 0.07 (61.8) 7.04 ⫾ 0.10 (91.2) 7.92 ⫾ 0.20 (12.0) 6.59 ⫾ 0.11 (257) NDe 7.10 ⫾ 0.11 (79.4)

f

9.46 ⫾ 0.07*** (0.347) 8.89 ⫾ 0.10*** (1.29) 10.0 ⫾ 0.09** (0.100) 7.81 ⫾ 0.17*** (15.5) ND 8.38 ⫾ 0.19*** (4.17)

12 7 2 9 8 5

9.79 ⫾ 0.07*** (0.136) 9.45 ⫾ 0.10*** (0.355) 10.0 ⫾ 0.10** (0.100) 7.89 ⫾ 0.15*** (12.9) ND 7.95 ⫾ 0.20*** (11.2)

8.90 ⫾ 0.12*** (0.962) 8.49 ⫾ 0.11*** (3.24) 9.25 ⫾ 0.18* (0.562) 7.60 ⫾ 0.21*** (25.1) ND 8.14 ⫾ 0.18*** (7.24)

8.86 ⫾ 0.09*** (0.937) 8.67 ⫾ 0.07*** (2.14) 9.44 ⫾ 0.16* (0.363) 7.60 ⫾ 0.22** (25.1) ND 8.12 ⫾ 0.19*** (7.59)

n pIC50 ⫾ SEM

mCD9 CD151

pIC50 ⫾ SEM n

CD81

pIC50 ⫾ SEM n

CD63

pIC50 ⫾ SEM n pIC50 ⫾ SEM nb PIC50a ⫾ SEM

CD9

J. VIROL.

n

HO ET AL.

TABLE 1. Effects of a range of recombinant tetraspanin domains on the infection of macrophages and PBMCs by virions expressing HIV-1 or VSV-G glycoproteins

6490

FIG. 2. The effects of removal of GST from recombinant CD63 EC2 domain on the inhibition of infection of macrophages by CCR5tropic HIV-1. Macrophages were treated with different concentrations of recombinant human CD63 EC2-GST fusion protein (GST-CD63), CD63 EC2 with GST removed (CD63) by thrombin cleavage, or GST alone for 30 to 60 min prior to the addition of virions expressing CCR5-specific HIV-1 Env protein. Infection was measured after 3 days as chemiluminescence from luciferase expressed under the control of the HIV-1 long terminal repeat promoter. (Left) Results are shown as percentage inhibition of infection relative to untreated control cells and are the means of at least three separate experiments ⫾ standard errors of the mean. (Right) Significance of difference was assessed by analysis of variance with a Bonferroni posttest; **, P ⬍ 0.01; ***, P ⬍ 0.001.

tually all of the GST. This treatment did not abrogate the inhibition of R5 infection of MDM by CD63 EC2 (Fig. 2). The cleaved CD63 EC2 was 10-fold more active than GST alone, although 6-fold less active than uncleaved GST-CD63 EC2. This suggests that the inhibitory activity of tetraspanin EC2 proteins is not dependent on GST but that GST can enhance the effect. Exogenous GST has also been found to inhibit sperm binding in mouse fertilization assays (6), perhaps due to the retention of enzymatic activity or through nonspecific binding to the cell surface. Mutation of residues in CD9 EC2 causes incomplete loss of inhibitory activity. Mutation of F176 in the subloop of CD9 EC2 or of the cysteine residues proposed to form disulfide linkages (C152 and C153) completely removed the inhibitory properties of the CD9 EC2 in a mouse fertilization assay (6, 32). To determine if a common mechanism was responsible, we tested whether the same mutations could interfere with the inhibition of infection of MDM (Fig. 3). We found that mutation of either cysteine residue to alanine had a modest but significant effect on infection by R5 virus (a threefold decrease in 50% inhibitory concentraton [IC50] values). Similarly, mutation of C152 and C153 decreased the IC50 values for X4 virus infection by ⬃9- and 4-fold, respectively. Mutation of F176 to alanine had a significant effect on inhibition of R5 virus infection, but this did not reach significance for X4. In contrast, inhibition of infection by VSV was strongly affected by all three mutations (⬃5-fold decrease for cysteine mutants; ⬃7-fold decrease for F176A). Nevertheless, the CD9 EC2 mutants were still significantly more potent inhibitors of infection than GST. The retention of some inhibitory activity by these mutants suggests that while the subloop structure required for ligand binding and specific protein-protein interactions is involved, regions of the EC2 domain outside of the loop constrained by the disulfide linkages are also required. Tetraspanin EC2 domains inhibit the uptake of virus particles. Tetraspanins are known to be involved in cellular adhe-

VOL. 80, 2006

TETRASPANINS INHIBIT HIV-1 INFECTION OF MACROPHAGES

FIG. 3. Effects of mutation of residues in the CD9 EC2 domain on the potency of inhibition of HIV-1 infection of macrophages. Macrophages were treated with different concentrations of wild-type or mutant recombinant human CD9 large extracellular domain-GST or GST alone for 30 to 60 min prior to the addition of virions expressing R5, X4, or VSV-G Env protein. Infection was measured after 3 days as chemiluminescence from luciferase expressed under the control of the HIV-1 long terminal repeat promoter. The results shown are IC50 values and are the means of at least three separate experiments ⫹ standard errors of the mean. The significance of differences from the GST-CD9 wild-type control was assessed by a two-tailed t test; ***, P ⬍ 0.001; **, P ⬍ 0.01; *, P ⬍ 0.05.

sion and uptake events (27, 31), so we investigated whether infection was inhibited at the stage of the binding/uptake of virus or at a later stage. This was first examined with virus fluorescently labeled using a Vpr-eGFP fusion protein. MDM were preincubated with GST (0.4 ␮M) or GST-CD63 EC2 (0.11 or 0.33 ␮M) for 30 min prior to addition of R5 virus, and infection was terminated after 10, 30, or 60 min. CD63 EC2 significantly inhibited the numbers of green particles associated with cells in a concentration-dependent manner at all of these time points, so the data were pooled (Fig. 4A and B). The higher HIV-1 input used in these binding/uptake experiments (200 ng p24 Gag) most likely explained the incomplete inhibition in virion uptake at concentrations (0.33 ␮M) that were sufficient to completely inhibit virus infection initiated with a 5-ng p24 Gag inoculum dose. Next, we treated macrophages with GST-CD63 EC2 or GST during or at various times after virus inoculation to assess the time of inhibitory activity (Fig. 4C). The results showed that the inhibitory effect of CD63 EC2 was significantly reduced even when it was added imme-

6491

diately after virus inoculation, suggesting that the time of action was at the stage of virus entry. In contrast, the time of addition of GST (before or during virus inoculation) made no difference to its inhibitory effect, suggesting that the weak inhibitory activity of GST at high concentrations is a postinfection event that is unrelated to the more potent inhibition of virus entry by tetraspanin EC2 proteins. Incubation with tetraspanin EC2 proteins does not downregulate CD4 or HIV-1 coreceptor expression on MDM. The finding that MDM infection is more susceptible to tetraspanin inhibition than PBMC infection could be due to differences in surface expression of HIV-1 receptors or cell-type-specific proteins that interact with tetraspanins. We therefore measured the expression of cellular proteins at the cell surfaces of macrophages and PBMCs and assessed the effect of addition of GST or GST-CD63 EC2 on their expression (Fig. 5). In agreement with previous reports (17, 30), we found that the levels of CD4 and HIV-1 coreceptor expression, as determined by their mean fluorescence intensities (MFI), were much lower on macrophages than on PBMCs (Fig. 5A). The percentage of cells expressing the coreceptors was also significantly lower for macrophages than PBMCs. In contrast, both the percentage and the level of expression of the tetraspanin CD63 was greater on macrophages than PBMCs. In no case, however, did the addition of GST-CD63 EC2 proteins cause further reduction of expression of these and other receptors on macrophages (Fig. 5B). CXCR4 and CD14 were unchanged, but CD4 and CD63 levels actually increased relative to cells treated with GST (Fig. 5C). CCR5 expression also appeared to be increased on GST-CD63 EC2-treated cells, but the significance of this observation is less clear, since the low-level expression of this receptor on macrophages renders it more sensitive to slight fluctuations in MFI. Interestingly, the increase in CD63 expression appeared to be due to an increase in the expression of endogenous CD63 at the cell surface, as staining with antibodies to GST did not reveal a similar increase in signal on cells treated with GST-CD63 EC2 compared to GST alone. Labeled tetraspanin EC2 proteins are internalized to vesicular compartments. To be able to inhibit infection of MDM at virus entry, tetraspanin EC2 domains must be able to either bind to structures at the cell surface or be taken up into the cells. We tested the uptake of tetraspanins by labeling GST and CD63 EC2 proteins with FITC and incubating MDM with 0.11 ␮M labeled protein at 37°C for 30 to 60 min. Cell-associated fluorescence was then measured by flow cytometry and fluorescence microscopy. CD63 appeared to be internalized to a greater degree than GST (Fig. 6A) and was incorporated into vesicular structures that overlapped with endocytosed dextran and transferrin (Fig. 6B). Tetraspanin EC2 proteins partially colocalize with HIV-1 virions and HIV-1 receptors. To assess whether tetraspanin EC2 and HIV-1 virions are localized to the same structures in MDM, cells were incubated with AlexaFluor 647-labeled GSTCD63 EC2 and Vpr-eGFP virus for 20 min at RT before being imaged. Partial colocalization of labeled CD63-EC2 with HIV-1 virions could be detected (Fig. 7A). To determine whether tetraspanin EC2 colocalizes with HIV-1 receptors at the cell surfaces of macrophages, cells were incubated with labeled tetraspanin EC2 proteins and antibodies to CD4,

6492

HO ET AL.

J. VIROL.

FIG. 4. Effects of treatment with recombinant CD63 EC2 domain on an early stage of CCR5-tropic HIV-1 infection and uptake by macrophages. (A) Virus uptake was quantified by preincubating macrophages with 0.4 ␮M GST or 0.11 ␮M or 0.33 ␮M GST-CD63 EC2 for 30 min prior to addition of Vpr-eGFP R5 virus, and infection was terminated 10, 30, or 60 min after virus addition. Green particles in micrographs of adherent cells were counted as described in Materials and Methods. As CD63 EC2 had an inhibitory effect on the numbers of green particles associated with cells at all time points examined, the data were pooled, and the results shown are the means ⫹ standard errors of the mean of 4 (for 0.33 ␮M GST-CD63 EC2) and 15 (for 0.4 ␮M GST and 0.11 ␮M GST-CD63 EC2) independent experiments. The differences between the doses of CD63 EC2 and the GST-only control (⫽ 100) were significant (**, P ⬍ 0.01) by one sample t test. (B) Examples of micrographs at 30 min post-virus inoculation. (C) Macrophages were incubated with R5 virus for 1 h at 37°C. Serial dilutions of GST alone or GST-CD63 EC2 were either present during the period of virus incubation or added immediately after the virus was removed by washing.

CXCR4, or CCR5 for 1 h at 4°C. A degree of colocalization of FITC-labeled GST-CD63 EC2 proteins with CXCR4 and CD4, but not CCR5, at the cell surface could be observed (Fig. 7B) in support of tetraspanin inhibition at the level of virus entry. DISCUSSION A role for the tetraspanin CD63 in HIV-1 infection was first suggested when it was shown to be upregulated from intracellular vesicles to the surfaces of HIV-1-infected cells and selectively incorporated into budding structures and newly synthesized virus particles (18–20, 25). Consistent with a role in viral release, vesicle fractions in infected H9 T cells contain increased levels of CD63 compared with uninfected cells (15, 16). More recently, it has been shown that HIV-1 infection is inhibited by anti-CD63 antibodies, but not antibodies to tetraspanin CD9 or CD81. This was specific for macrophages and

HIV-1 strains, which use the CCR5 coreceptor, as X4 virus and T-cell infection was unaffected by anti-CD63 antibodies (28). Previously, we and others have demonstrated that recombinant forms of the EC2 domains of tetraspanins can affect cellular functions, such as fertilization and monocyte fusion (6, 27). In this report, we examined the effects of recombinant EC2 domains on infection of monocyte-derived macrophages and PBMCs, using single-cycle luciferase reporter viruses pseudotyped with different HIV-1 and VSV envelope glycoproteins. We found that recombinant CD63 EC2 domain gave consistent inhibition of HIV-1 infection of MDM, although sensitivity varied with the envelope of the virus (Fig. 1 and Table 1). R5 virus was most sensitive (GST-CD63 EC2 IC50 value, 0.14 nM), with 100% inhibition occurring at 1 nM. GST alone also inhibited infection of MDM by R5, but with a much higher IC50 value. Human CD9, -81, and -151 and mouse CD9 EC2 domains were also significantly more inhibitory than GST alone for R5 virus infection. X4 virus infection had a similar

VOL. 80, 2006

TETRASPANINS INHIBIT HIV-1 INFECTION OF MACROPHAGES

FIG. 5. Effects of treatment of macrophages with recombinant CD63 EC2 domain on membrane protein expression. Macrophages were incubated with 0.11 ␮M GST or GST-CD63 EC2 for 1 h at 37°C, and the cells were extensively washed and harvested by scraping. Treated cells were incubated with fluorescent conjugated antibodies against various cellular proteins. Untreated macrophages and PBMCs were stained in parallel for reference, and fluorescence was then quantified by flow cytometry. (A) Percent and MFI of CD4, CD63, CCR5, or CXCR4 expression on untreated macrophages and PBMCs. (B) Expression of CD4, CD14, CD63, CCR5, CXCR4, and GST on macrophages treated with GST (open histograms) or GST-CD63 EC2 (shaded histograms). The histograms shown are representative of three experiments. (C) Results from three separate tetraspanintreated macrophage experiments are shown as percent changes in MFI from GST cells (⫽ 100) and the means and standard errors of the mean. Significance was assessed by one sample t test; *, P ⬍ 0.05.

pattern of inhibition with similar ratios between IC50 values for GST alone and the tetraspanin GST EC2 proteins, but 100% inhibition was rarely achieved even at high concentrations. Interestingly, infection of MDM by virus pseudotyped with VSV glycoprotein was also inhibited by the human (but not mouse) tetraspanin domains. PBMC infection, however, was affected to a much lesser extent than MDM, with R5 and VSV infection inhibited to a maximum of only 60%, even at high concentrations (1 ␮M) of GST-CD63 EC2 and with no significant inhibition of X4 infection. Thus, the tetraspanin EC2mediated inhibition of infection is not restricted to MDM infection by R5 virus, like the reported anti-CD63 antibody-

6493

FIG. 6. Uptake of fluorescently labeled CD63 EC2 in macrophages. (A) Macrophages were incubated with 0.11 ␮M of fluorescently labeled GST or GST-CD63 EC2 protein for 1 h at 37°C, and cell-associated fluorescence was measured by flow cytometry (MFI) or by analysis of images using ImageJ software (green pixels). The data are the means of four separate experiments ⫹ standard errors of the mean, and the significance of differences from GST controls was assessed by a two-tailed t test; *, P ⬍ 0.05. (B) 0.11 ␮M GST-CD63 EC2–FITC proteins were added to macrophages, together with dextran-rhodamine (Dex) or transferrin-AlexaFluor 647 (TF) for 30 min at 37°C. Representative images are shown.

mediated inhibition (28). However, in agreement with the previous report (28), PBMC infection was only partly inhibited by tetraspanin EC2 protein, and infection by X4 was refractory to inhibition. The differential inhibitory effects of tetraspanin EC2 on HIV-1 infection of macrophages and PBMCs suggest that the proteins function on target cells rather than virions. Further support for a target cell effect came from the observation that inhibition occurs when EC2 proteins are added to MDM and then washed out before the addition of virions. Time-of-addition experiments indicate that the tetraspanins act on an early stage of HIV infection, inhibiting the uptake of virions (Fig. 4). Tetraspanin EC2 proteins, similar to the role that the Drosophila tetraspanin sun plays in light-dependent lysosomal accumulation of rhodopsin (31), may block viral entry via modulation of the activity of viral receptors that form complexes with endogenous tetraspanins. Indeed, the fact that multiple tetraspanin EC2 proteins inhibit HIV-1 infection of macrophages suggests that TEM are involved rather than individual tetraspanins. TEM are functional clusters of membrane proteins, akin to lipid rafts, in which dimers or multimers of tetraspanins probably function as organizers or facilitators. Exogenous EC2

6494

HO ET AL.

J. VIROL.

FIG. 7. Partial colocalization of fluorescently labeled CD63 EC2 proteins with HIV-1 virions and with CD4 and CXCR4 at the cell surfaces of macrophages. (A) Macrophages were incubated with 0.11 ␮M of GST-CD63 EC2 labeled with AlexaFluor 647 and P3-Gag-eGFP virus for 20 min at RT. Microphotographs illustrating the patterns of expression were taken after fixation, and representatives are shown. (B) GST-CD63 EC2–FITC proteins (0.11 ␮M) were added, together with PE-conjugated anti-CD4, anti-CCR5, or anti-CXCR4 antibodies, to macrophages for 1 h at 4°C. The cells were then washed and fixed, and images taken. Representative images are shown.

protein may destabilize TEM by inhibiting the multimerization of endogenous tetraspanins and so could inhibit the function of TEM. Binding of tetraspanin EC2 to endogenous tetraspanins on the surfaces of target cells could also exert qualitative effects, such as localization, internalization rates, and/or association with HIV-1 receptors. We demonstrated that expression of endogenous CD63 at the cell surface of macrophages was substantially increased with tetraspanin treatment (Fig. 5C). Conceivably, the increase in endogenous CD63 expression is due to retention at the cell surface via multimerization with GST-CD63 EC2. The finding that mutations of three residues in CD9 EC2 (C152, C153, and F176) (Fig. 3) had only small effects on the inhibition of HIV-1

infections is consistent with the notion that the soluble tetraspanins associate with endogenous tetraspanins, since the structures involved in homo- and heterodimer formation (12, 26) are not perturbed by these mutations. We also showed that while the HIV-1 coreceptor CXCR4 was not quantitatively modified by tetraspanin treatment, the expression of CD4, and perhaps CCR5 as well, was actually increased (Fig. 5B and C). Imaging studies revealed partial colocalization of GST-CD63 EC2 with HIV-1 virions and with CD4 and CXCR4, but not CCR5, on the surfaces of macrophages (Fig. 7). CD4 is located within and outside both lipid rafts and TEM, but CD63 and CCR5 are confined to TEM (3, 5, 24). Further, the tetraspanins CD9, CD81, and CD82 had been reported to associate with

VOL. 80, 2006

TETRASPANINS INHIBIT HIV-1 INFECTION OF MACROPHAGES

and stabilize CD4 (8, 10), and the concentrations of CD4 and CoR required for efficient infections are interdependent, implying a direct or indirect interaction of these receptors in a concentration-dependent manner (13, 22). Thus, it is tempting to speculate that multimerization of endogenous tetrapanins with tetraspanin EC2 proteins acts as a barrier to the formation and/or organization of CD4-CoR complexes within TEM that are required for membrane fusion. The greater effect of the tetraspanin EC2 proteins on MDM than on PBMC infection could be attributed to higher expression of tetraspanins but lower levels of the CD4 and coreceptors on primary macrophages (Fig. 5A), and the increase in susceptibility of R5 compared to X4 virus could be attributed to the inhibitory effects of tetraspanin EC2 as a result of the greater dependence of R5 viruses on CD4-CoR concentrations for efficient entry. Finally, the ability of tetraspanins to complex with diverse cellular partners could explain the broader range of inhibitory activity seen with tetraspanin EC2 proteins compared to that reported for antibodies to CD63 (28). In summary, we have clearly demonstrated that a range of tetraspanin EC2 proteins can inhibit viral infection of MDM. The mechanism of inhibition appears to be at the stage of virus entry by blocking virion binding/uptake via membrane remodeling. Nevertheless, since data show that labeled CD63 EC2 is taken up and translocated via the endosomal-lysosomal route (Fig. 6), the possibility of postentry effects, such as interference with the trafficking of virus and/or fusion within the vesicles, cannot be excluded. Indeed, it could be at these postentry steps that the inhibitory effects on VSV infection are exerted. Further studies are needed to explore these various possibilities. Regardless, our data, together with a previous report of antiCD63 antibody-mediated inhibition (28), support a role for TEM in HIV-1 infection and suggest that tetraspanin-derived agents may have utility in preventing macrophage and, possibly, microglial infection by HIV-1. ACKNOWLEDGMENTS This work was supported by NIH grants RO1 CA72822 and R37 AI 41945. F.M. and V.P. acknowledge financial support from the Humane Research Trust and the Ford Foundation International Fellowship Program, respectively. REFERENCES 1. Boucheix, C., and E. Rubinstein. 2001. Tetraspanins. Cell Mol. Life Sci. 58:1189–1205. 2. Chakrabarti, L. A., T. Ivanovic, and C. Cheng-Mayer. 2002. Properties of the surface envelope glycoprotein associated with virulence of simian-human immunodeficiency virus SHIV(SF33A) molecular clones. J. Virol. 76:1588– 1599. 3. Chen, H., D. Rojo, J. von Lindern, N. Liburd, M. R. Ferguson, and W. A. O’Brien. 2005. Mechanism of cellular resistance to anti-CD63 inhibition of HIV infection. Antivir. Ther. 10:S78. 4. Connor, R. I., K. E. Sheridan, C. Lai, L. Zhang, and D. D. Ho. 1996. Characterization of the functional properties of env genes from long-term survivors of human immunodeficiency virus type 1 infection. J. Virol. 70: 5306–5311. 5. Hemler, M. E. 2003. Tetraspanin proteins mediate cellular penetration, invasion, and fusion events and define a novel type of membrane microdomain. Annu. Rev. Cell Dev. Biol. 19:397–422. 6. Higginbottom, A., Y. Takahashi, L. Bolling, S. A. Coonrod, J. M. White, L. J. Partridge, and P. N. Monk. 2003. Structural requirements for the inhibitory action of the CD9 large extracellular domain in sperm/oocyte binding and fusion. Biochem. Biophys. Res. Commun. 311:208–214. 7. Hsu, M., J. Zhang, M. Flint, C. Logvinoff, C. Cheng-Mayer, C. M. Rice, and J. A. McKeating. 2003. Hepatitis C virus glycoproteins mediate pH-dependent cell entry of pseudotyped retroviral particles. Proc. Natl. Acad. Sci. USA 100:7271–7276.

6495

8. Hu, C. C., F. X. Liang, G. Zhou, L. Tu, C. H. Tang, J. Zhou, G. Kreibich, and T. T. Sun. 2005. Assembly of urothelial plaques: tetraspanin function in membrane protein trafficking. Mol. Biol. Cell. 16:3937–3950. 9. Imai, T., K. Fukudome, S. Takagi, M. Nagira, M. Furuse, N. Fukuhara, M. Nishimura, Y. Hinuma, and O. Yoshie. 1992. C33 antigen recognized by monoclonal antibodies inhibitory to human T cell leukemia virus type 1-induced syncytium formation is a member of a new family of transmembrane proteins including CD9, CD37, CD53, and CD63. J. Immunol. 149:2879– 2886. 10. Imai, T., M. Kakizaki, M. Nishimura, and O. Yoshie. 1995. Molecular analyses of the association of CD4 with two members of the transmembrane 4 superfamily, CD81 and CD82. J. Immunol. 155:1229–1239. 11. Johnson, G. 1990. Immunofluorescence, p. 179–200. In D. Catty (ed.), Antibodies, vol. 2. IRL Press, Oxford, United Kingdom. 12. Kitadokoro, K., D. Bordo, G. Galli, R. Petracca, F. Falugi, S. Abrignani, G. Grandi, and M. Bolognesi. 2001. CD81 extracellular domain 3D structure: insight into the tetraspanin superfamily structural motifs. EMBO J. 20:12–18. 13. Kuhmann, S. E., E. J. Platt, S. L. Kozak, and D. Kabat. 2000. Cooperation of multiple CCR5 coreceptors is required for infections by human immunodeficiency virus type 1. J. Virol. 74:7005–7015. 14. Loffler, S., F. Lottspeich, F. Lanza, D. O. Azorsa, V. ter Meulen, and J. Schneider-Schaulies. 1997. CD9, a tetraspan transmembrane protein, renders cells susceptible to canine distemper virus. J. Virol. 71:42–49. 15. Meerloo, T., H. K. Parmentier, A. D. Osterhaus, J. Goudsmit, and H. J. Schuurman. 1992. Modulation of cell surface molecules during HIV-1 infection of H9 cells. An immunoelectron microscopic study. AIDS 6:1105– 1116. 16. Meerloo, T., M. A. Sheikh, A. C. Bloem, A. de Ronde, M. Schutten, C. A. van Els, P. J. Roholl, P. Joling, J. Goudsmit, and H. J. Schuurman. 1993. Host cell membrane proteins on human immunodeficiency virus type 1 after in vitro infection of H9 cells and blood mononuclear cells. An immuno-electron microscopic study. J. Gen. Virol. 74:129–135. 17. Naif, H. M., S. Li, M. Alali, A. Sloane, L. Wu, M. Kelly, G. Lynch, A. Lloyd, and A. L. Cunningham. 1998. CCR5 expression correlates with susceptibility of maturing monocytes to human immunodeficiency virus type 1 infection. J. Virol. 72:830–836. 18. Ono, A., and E. O. Freed. 2004. Cell-type-dependent targeting of human immunodeficiency virus type 1 assembly to the plasma membrane and the multivesicular body. J. Virol. 78:1552–1563. 19. Orentas, R. J., and J. E. Hildreth. 1993. Association of host cell surface adhesion receptors and other membrane proteins with HIV and SIV. AIDS Res. Hum Retrovir. 9:1157–1165. 20. Pelchen-Matthews, A., B. Kramer, and M. Marsh. 2003. Infectious HIV-1 assembles in late endosomes in primary macrophages. J. Cell Biol. 162:443– 455. 21. Petracca, R., F. Falugi, G. Galli, N. Norais, D. Rosa, S. Campagnoli, V. Burgio, E. Di Stasio, B. Giardina, M. Houghton, S. Abrignani, and G. Grandi. 2000. Structure-function analysis of hepatitis C virus envelope-CD81 binding. J. Virol. 74:4824–4830. 22. Platt, E. J., K. Wehrly, S. E. Kuhmann, B. Chesebro, and D. Kabat. 1998. Effects of CCR5 and CD4 cell surface concentrations on infections by macrophagetropic isolates of human immunodeficiency virus type 1. J. Virol. 72:2855–2864. 23. Pohlmann, S., J. Zhang, F. Baribaud, Z. Chen, G. J. Leslie, G. Lin, A. Granelli-Piperno, R. W. Doms, C. M. Rice, and J. A. McKeating. 2003. Hepatitis C virus glycoproteins interact with DC-SIGN and DC-SIGNR. J. Virol. 77:4070–4080. 24. Popik, W., and T. M. Alce. 2004. CD4 receptor localized to non-raft membrane microdomains supports HIV-1 entry. Identification of a novel raft localization marker in CD4. J. Biol. Chem. 279:704–712. 25. Raposo, G., M. Moore, D. Innes, R. Leijendekker, A. Leigh-Brown, P. Benaroch, and H. Geuze. 2002. Human macrophages accumulate HIV-1 particles in MHC II compartments. Traffic 3:718–729. 26. Seigneuret, M., A. Delaguillaumie, C. Lagaudriere-Gesbert, and H. Conjeaud. 2001. Structure of the tetraspanin main extracellular domain. A partially conserved fold with a structurally variable domain insertion. J. Biol. Chem. 276:40055–40064. 27. Takeda, Y., I. Tachibana, K. Miyado, M. Kobayashi, T. Miyazaki, T. Funakoshi, H. Kimura, H. Yamane, Y. Saito, H. Goto, T. Yoneda, M. Yoshida, T. Kumagai, T. Osaki, S. Hayashi, I. Kawase, and E. Mekada. 2003. Tetraspanins CD9 and CD81 function to prevent the fusion of mononuclear phagocytes. J. Cell Biol. 161:945–956. 28. von Lindern, J. J., D. Rojo, K. Grovit-Ferbas, C. Yeramian, C. Deng, G. Herbein, M. R. Ferguson, T. C. Pappas, J. M. Decker, A. Singh, R. G. Collman, and W. A. O’Brien. 2003. Potential role for CD63 in CCR5mediated human immunodeficiency virus type 1 infection of macrophages. J. Virol. 77:3624–3633. 29. Willett, B. J., M. J. Hosie, O. Jarrett, and J. C. Neil. 1994. Identification of a putative cellular receptor for feline immunodeficiency virus as the feline homologue of CD9. Immunology 81:228–233.

6496

HO ET AL.

30. Wu, L., W. A. Paxton, N. Kassam, N. Ruffing, J. B. Rottman, N. Sullivan, H. Choe, J. Sodroski, W. Newman, R. A. Koup, and C. R. Mackay. 1997. CCR5 levels and expression pattern correlate with infectability by macrophagetropic HIV-1, in vitro. J. Exp. Med. 185:1681–1691. 31. Xu, H., S. J. Lee, E. Suzuki, K. D. Dugan, A. Stoddard, H. S. Li, L. A. Chodosh, and C. Montell. 2004. A lysosomal tetraspanin associated with

J. VIROL. retinal degeneration identified via a genome-wide screen. EMBO J. 23:811– 822. 32. Zhu, G. Z., B. J. Miller, C. Boucheix, E. Rubinstein, C. C. Liu, R. O. Hynes, D. G. Myles, and P. Primakoff. 2002. Residues SFQ (173–175) in the large extracellular loop of CD9 are required for gamete fusion. Development 129:1995–2002.