Macrophages Transmit Human Immunodeficiency Virus Type 1 ...

1 downloads 0 Views 970KB Size Report
Mar 29, 2007 - lovirus (17), measles virus (22), and human hepatitis C virus. (58). The human T-lymphotropic virus type 1 retrovirus was found to propagate ...
JOURNAL OF VIROLOGY, Sept. 2007, p. 9078–9087 0022-538X/07/$08.00⫹0 doi:10.1128/JVI.00675-07 Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vol. 81, No. 17

Macrophages Transmit Human Immunodeficiency Virus Type 1 Products to CD4-Negative Cells: Involvement of Matrix Metalloproteinase 9䌤 Claudia Muratori,1 Antonella Sistigu,1 Eliana Ruggiero,1 Mario Falchi,1 Ilaria Bacigalupo,2 Clelia Palladino,2 Elena Toschi,2 and Maurizio Federico1* Division of Pathogenesis of Retroviruses1 and of Virus-Host Interaction and Core-Laboratory of Immunology,2 National AIDS Center, Istituto Superiore di Sanita `, Rome, Italy Received 29 March 2007/Accepted 11 June 2007

It was previously reported that human immunodeficiency virus type 1 (HIV-1) spreads in CD4 lymphocytes through cell-to-cell transmission. Here we report that HIV-1-infected macrophages, but not lymphocytes, transmit HIV-1 products to CD4-negative cells of either epithelial, neuronal, or endothelial origin in the absence of overt HIV-1 infection. This phenomenon was detectable as early as 1 h after the start of cocultivation and depended on cell-to-cell contact but not on the release of viral particles from donor cells. Transfer of HIV-1 products occurred upon their polarization and colocalization within zones of cell-to-cell contact similar to virological synapses. Neither HIV-1 Env nor Nef expression was required but, interestingly, we found that an HIV-1-dependent increase in matrix metalloproteinase 9 production from donor cells significantly contributed to the cell-to-cell transmission of the viral products. The macrophage-driven transfer of HIV-1 products to diverse CD4-negative cell types may have a significant role in AIDS pathogenesis. system counterpart of macrophagic cells, i.e., microglia cells (41). Macrophages are good producers of matrix metalloproteinases (MMPs), i.e., zinc-dependent extracellular proteases that function at a neutral pH to cleave a wide variety of substrates (61). These include basement membrane and extracellular matrix components, growth and death factors, cytokines, and cell and matrix adhesion molecules. The broad range of substrate specificities and expression patterns of MMPs results in their involvement in many different processes, both normal and pathological (54). MMPs can be either secreted or cell-membrane-associated enzymes. Both MMP types are produced in an inactive form and undergo full activation upon propeptide processing induced by plasmin or autocatalysis or through the action of other active MMPs. Here we report that macrophages transmit HIV-1 products to CD4-negative epithelial or endothelial cells or astrocytes by cell-to-cell contact through a mechanism likely involving MMP-9. These findings could be relevant to a better delineation of the role of infected macrophages in AIDS-related immune dysregulation.

It is widely documented that viruses can spread through mechanisms alternative to receptor-mediated cell internalization. These include transcytosis (31), transinfection (57), and cell-to-cell infection. This latter way of infection has been documented for human herpesviruses (13), human cytomegalovirus (17), measles virus (22), and human hepatitis C virus (58). The human T-lymphotropic virus type 1 retrovirus was found to propagate exclusively through the formation of zones of tight cell-to-cell adhesion (virological synapses) very similar to the contact between antigen-presenting cells and lymphocytes (immunological synapses) (32). Human immunodeficiency virus type 1 (HIV-1) was found to propagate very efficiently by cell-to-cell transmission (33, 48, 53) through a mechanism requiring expression of HIV Env receptors in donor cells and of CD4 and CXCR4 or CCR5 cell receptors in target cells, although coreceptor-independent HIV-1 transfer to peripheral blood mononuclear cells was also described (5). HIV infection can lead either to cell death, mostly in activated CD4 lymphocytes, or to persistent infection in cells controlling HIV gene expression and/or resisting its cytopathic effects, as is the case with macrophages (55). These cells play a critical role in AIDS pathogenesis, both as viral reservoirs during highly active anti-retroviral therapy (36) and by affecting the pattern of released soluble factors involved in both innate and adaptive immunity. In addition, the nature of macrophages as migratory blood cells strongly favors their interaction with cells of different types, e.g., epithelial, stromal, or endothelial cells. This is also the case with the central nervous

MATERIALS AND METHODS Cell cultivation, cocultivation, and purification. U937, U937/HIV-1, D10 (25), 8E5 (26), CEMss, and C8166 cells were grown in RPMI media supplemented with 10% decomplemented fetal calf serum (dFCS). Human embryonic kidney epithelial 293 (HEK293)/CD8T, 293T, human astrocytoma U87 (43), and human endothelial EA-hy 926 cells (21) were grown in Dulbecco’s modified Eagle’s medium plus 10% dFCS. 293/CD8T cells were recovered upon transfection of 293 cells with a pcDNA3 vector (Invitrogen, Carlsbad, CA) expressing the human CD8 receptor truncated in its cytoplasmic domain and G418 selection. Human primary monocytes were isolated and cultivated as previously described (24). Human primary CD4 and CD8 lymphocytes were negatively selected from peripheral blood mononuclear cells, using an appropriate immunomagnet-based selection kit from Miltenyi Biotec, and cultivated in RPMI containing 20% dFCS.

* Corresponding author. Mailing address: National AIDS Center, Istituto Superiore di Sanita`, Viale Regina Elena, 299, 00161 Rome, Italy. Phone: 39-06-49903248. Fax: 39-06-49903002. E-mail: maurizio [email protected]. 䌤 Published ahead of print on 20 June 2007. 9078

VOL. 81, 2007

MMP-9 AND CELL-TO-CELL TRANSFER OF HIV-1 PRODUCTS

Cocultivations were typically set up in 1 ml of RPMI–10% dFCS in 12-well plates by seeding 5 ⫻ 105 target cells with 106 donor cells. In some instances, target cells were purified by negative anti-CD14 immunoselection followed by positive immunoselection, using anti-CD8 microbeads (Miltenyi Biotec, Auburn, CA). Transwell cocultures were carried out in 6-well plates (Becton Dickinson, San Jose, CA) and cell culture insert Falcon membranes (25 mm diameter, 0.4-␮m pore size). Azidothymidine (AZT), ritonavir, indinavir, and saquinavir were obtained from the NIH AIDS Research and Reference Reagent Program. The MMP inhibitors GM6001 and MMP IV were from Chemicon (Temecula, CA) and Calbiochem (San Diego, CA), respectively. Sodium orthovanadate was obtained from Sigma-Aldrich (St. Louis, MO). HIV-1 preparations and titration. Preparations of R5-tropic ADA HIV-1 were obtained by transfecting the infectious molecular clone in 293T cells with Lipofectamine 2000 (Invitrogen). Supernatants recovered 48 and 72 h later were clarified and concentrated by ultracentrifugation on a 20% sucrose cushion. Recovery of ⌬env, ⌬nef, and NL4-3/NefF12 HIV-1 molecular clones was described previously (24). Preparations of HIV-1 strains pseudotyped with vesicular stomatitis virus glycoprotein G (VSV-G) were obtained from the supernatants of 293T cells cotransfected with the respective HIV-1 molecular clone and an immediate-early cytomegalovirus promoter-regulated VSV-G-expressing vector in a 5:1 molar ratio. Virus preparations were titrated by measuring the HIV-1 CAp24 content by quantitative enzyme-linked immunosorbent assays (ELISA; Innogenetics, Gent, Belgium) and by reverse transcriptase assays. Cells were infected by spinoculation at 400 ⫻ g for 30 min at room temperature (RT) using 50 ng/105 cells of VSV-G-pseudotyped HIV-1 or 150 ng/105 cells of either HIV-1 strain X4 NL4-3 or R5 ADA. Virus adsorption was prolonged for an additional 2 h at 37°C, and finally, the cells were washed and refed with the complete medium. For the coculture experiments, the next day the infected cells were extensively washed, and cocultures were set up at a 3:1 cell ratio, in some cases in the presence of 20 ␮M AZT to avoid infection of target cells by residual virions. After an additional 16 h, the cocultures were analyzed by fluorescenceactivated cell sorter (FACS) analysis. FACS analysis and ELISA. Cocultures were incubated with trypsin for 15 min at 37°C and then incubated with Permeafix (Ortho Diagnostic, Raritan, NJ) for 30 min at RT. After two extensive washes, the cells were labeled for 1 h at RT with 1:100 dilutions of KC57-RD1 phycoerythrin (PE)-conjugated anti HIV-1 Gag CAp24 KC-57 MAb (Coulter Corp., Hialeah, FL) and fluorescein isothiocyanate (FITC)-conjugated anti-CD8 DK25 MAb (Dako, Glostrup, Denmark) or with species-specific isotypes. For detection of HIV-1 Env gp120 or intercellular adhesion molecule 1 (ICAM-1), the cells were incubated with trypsin, permeabilized, and labeled with a 1:100 dilution of 4G10 anti-gp120 MAb or PE-conjugated anti-ICAM-1 1H4 MAb (EuroBioSciences, Friesoythe, Germany), respectively, for 1 h at RT. Carboxyfluorescein succinimidyl ester (CFSE; Biocompare, San Francisco, CA) cell labeling was carried out according to the manufacturer’s recommendations. The intracellular concentrations of HIV-1 CAp24, Env gp120, and ICAM-1 were measured by lysing the cells with phosphate-buffered saline (PBS)–0.1% Triton X-100 for 15 min at 4°C. Thereafter, both nucleus and cell debris was pelleted, and the supernatants were assayed by quantitative ELISA (Innogenetics, Advanced BioScience Laboratories, Kensington, MD, and R&D Systems, Minneapolis, MN, for CAp24, Env gp120, and ICAM-1, respectively) according to the manufacturers’ recommendations. Fluorescence and confocal microscope analyses. For analysis by fluorescence microscope, the cells were placed on poly-L-lysine (Sigma-Aldrich)-coated cover glass. The cells were then fixed in 4% paraformaldehyde (Sigma-Aldrich) in PBS and quenched for 10 min with 0.1 M glycine in PBS. After that, antibody labeling for intracellular proteins was carried out upon cell permeabilization with 0.1% Triton X-100 (Sigma-Aldrich) in PBS, and coverslips were mounted to slides, using an antifade mounting medium. Finally, the cells were observed with a Zeiss Axioskop 2 Plus fluorescence microscope. For confocal microscope analyses, cells were fixed with 2% (vol/vol) formaldehyde in PBS, and both phase-contrast and fluorescence images were taken by an Olympus IX-81 device. The following antibodies/reagents were used for both analyses: FITC-conjugated anti-HIV-1 Gag CAp24 KC-57 mouse MAb (Coulter); anti-HIV-1 Env gp120 2G12 human MAb; FITC- or PE-conjugated anti-human CD8 mouse MAb clone DK25 (Dako); Alexa Fluor secondary antibodies conjugated with Texas Red (568 nm) or FITC (488 nm) from Molecular Probes (Invitrogen); and 4⬘,6⬘-diamidino-2phenylindole (DAPI) from Vector Laboratories (Burlingame, CA). MMP-9 detection. Detection of MMP-9 in the supernatants was performed either by quantitative ELISA (R&D Systems) or by using zymograms as described previously (35). Briefly, we analyzed the gelatinase activity on 40 ␮l of supernatants from 5 ⫻ 105 cells cultivated for 16 h in 300 ␮l of serum-free medium or on equal volumes of supernatants from 5 ⫻ 105 HIV-1-infected donor cells cocultured in the same conditions with 2 ⫻ 105 293/CD8T cells. The

9079

supernatants were loaded onto a 10% acrylamide–1% sodium dodecyl sulfate gel embedded with 0.12 mg/ml of gelatin. After the run was completed, the gels were soaked for 1 h in 2.5% Triton X-100 and then incubated overnight in a collagenase buffer (50 mM Tris HCl [pH 7.6], 0.2 M NaCl, 5 mM CaCl2, and 0.2% Brij-35). Finally, the gels were stained with 0.1% Coomassie brilliant blue R250 and destained in 30% methanol–10% acetic acid. The MAb 7-11C (Oncogene Research Products, Cambridge, MA) was used to neutralize the MMP-9 activity (46).

RESULTS Cell-to-cell transfer of HIV-1 products from infected macrophages to CD4-negative cells. Retro- and lentiviruses spread by cell-to-cell contact more efficiently than cell-free viruses and through alternative mechanisms (8, 18, 52, 53). Thus, we were interested in assessing whether and how HIV-1 can be transmitted through cell-to-cell contact from infected macrophages or lymphocytes to CD4-negative cells. We set up cocultures of macrophagic (i.e., U937/HIV-1) or lymphocytic (i.e., D10 and 8E5) cell lines chronically infected with X4-tropic HIV-1 (donor cells) and human epithelial HEK293 cells (target cells) engineered for expression of the human CD8 receptor truncated in its intracytoplasmic domain (293/CD8T). After 16 h, floating cells were removed and cocultures were incubated with trypsin for 15 min at 37°C to remove virus particles not specifically bound to the cell membrane (5, 40). FACS analysis showed the presence of HIV-1 Gag products in the majority of CD8-positive target cells after cocultivation with infected macrophages but not with lymphocytes (Fig. 1A). This phenomenon did not relate to CD8 expression on target cells, since the same results were obtained when parental 293 cells were prelabeled with CFSE (data not shown). Env gp120 was also transmitted in target cells upon cocultivation with HIV-1-infected macrophages (Fig. 1B). The transfer of HIV-1 Gag products became clearly detectable as early as 1 h after the coculture was started, while the highest percentage of Gagpositive cells was observed after 16 h (Fig. 1C). No further significant variations were detected when the coculture time was prolonged (data not shown). These results were obtained with a 2:1 donor cell-to-target cell ratio, but no relevant differences were observed in the range of 4:1 to 1:1 donor cellto-target cell ratios. Furthermore, and of major importance, internalization of both Gag and Env products was formally demonstrated by confocal microscope analysis (Fig. 1D). The HIV-1 transfer we observed strictly depended on cellto-cell contact, since no Gag-related products were detectable in cocultures carried out in microporous-membrane transwell chambers for 16 h (Fig. 1E) and until 96 h (data not shown). This suggested that receptor-independent HIV-1 endocytosis (23, 27, 49) is not involved in the transfer of viral products. To test this hypothesis, U937 cells previously infected with VSVG-pseudotyped NL4-3/NefF12, i.e., an HIV-1 variant unable to release viral particles in the supernatant (14, 39), were cocultured with 293/CD8T cells in the presence of 20 ␮M azidothymidine. We noticed that the transfer of HIV-1 Gag products to target cells occurred at levels similar to those detected using U937 cells infected with wild-type (wt) HIV-1 as donor cells (Fig. 1F). Next, we were interested in establishing whether macrophages also transfer cell proteins expressed at levels comparable to those of the viral products. We found that ICAM-1 was

9080

MURATORI ET AL.

expressed by U937/HIV-1 but not by 293/CD8T cells and at levels similar to those for Env gp120 (Fig. 1G). No ICAM-1 transfer in 293/CD8T cells was observed after 16 h of coculture with U937/HIV-1 cells in the presence of a readily detectable transfer activity of Env gp120 (Fig. 1G). No ICAM-1 passage was noticed with uninfected U937 as the donor cells (data not

J. VIROL.

FIG. 1. Macrophage but not lymphocyte cell lines efficiently transfer HIV-1 products to cocultured epithelial cells. (A) FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8 T-cells cocultured with infected macrophages or lymphocytes. U937/HIV-1, D10, or 8E5 cells were cocultured with 293/CD8T cells for 16 h and then analyzed by FACS. As controls, the different cell types were analyzed alone (upper panels). (B) FACS analyses of HIV-1 Env gp120 and CD8 levels in 293/CD8T cells cocultured for 16 h with U937/HIV-1. (C) FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured for indicated times with U937/HIV-1. (D) 293/CD8T cells cocultured for 16 h with U937/HIV-1, separated from donor cells, and labeled with both anti-Gag and anti-Env gp120 MAbs upon cell permeabilization. Two sets of representative confocal microscope images are shown. The images in the lower panels are the results of computer-assisted analysis carried out by overlapping the images obtained with visible and fluorescent lights. Bars represent 10 ␮M. (E) FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured with U937/HIV-1, separated (transwell coculture) or not separated by a 0.4-␮m microporous filter. (F) FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured with U937 cells previously infected with the VSV-G NL43/NefF12 HIV-1 variant or, as a control, with the wt counterpart. As additional controls, the different cell types were analyzed alone. (G) FACS analyses of either HIV-1 Env gp120 or ICAM-1 and CD8 levels in 293/CD8T cells cocultured for 16 h with U937/HIV-1. The amounts of cell-associated Env gp120 and ICAM-1 as measured by ELISA are also reported. The results are representative of three (A and G), four (B and F), two (C and E), or five (D) independent experiments. For single-cell-type cultures, the percentage of total events in each quadrant is reported, whereas each coculture plot shows the percentage of Gag-positive cells in the total number of CD8-positive cells. ␣, anti.

shown). Similar results were achieved by assaying two additional highly expressed membrane proteins not detectable in target cells, i.e., lymphocyte function-associated antigen 1 (LFA-1) and CD45 (data not shown). These data suggest that

VOL. 81, 2007

MMP-9 AND CELL-TO-CELL TRANSFER OF HIV-1 PRODUCTS

FIG. 2. Analyses of additional donor/target cell types. (A) Primary macrophages efficiently act as donor cells. FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured with primary human macrophages infected with R5 ADA HIV-1 and in single-cell-type cultures. Results are representative of two independent experiments. (B) Primary activated CD4 lymphocytes do not transmit HIV-1 products to epithelial cells. FACS analyses of levels of HIV-1 Gag products in HIV-1 infected CD4 lymphocytes cocultivated with CFSE-labeled 293/CD8T cells or activated CD4 lymphocytes in the presence of 20 ␮M AZT for 16 h. Results are representative of experiments carried out with cells from two donors. (C) Both U87 astrocytes and EA-hy 926 endothelial cells but not primary CD8 lymphocytes act efficiently as target cells. FACS analyses for detection of HIV-1 Gag products and either CD8 (for CD8 lymphocytes) or CFSE (for both U87 and EA-hy 926 cells) in U937/HIV-1-based cocultures. 293/CD8T cells were used as target cell controls. Results are representative of five (for CD8 lymphocytes) or two (for both U87 and EA-hy 926 cells) independent experiments. For single-cell-type cultures, the percentage of total events in each quadrant is reported, whereas each coculture plot shows the percentage of Gag-positive cells in the total number of CD8- or CFSE-positive cells. PBL, peripheral blood lymphocyte; ␣, anti.

the observed transfer of viral products is not part of a more generalized phenomenon also involving highly expressed cell proteins. In summary, we provided evidence that infected macrophages but not lymphocytes efficiently and specifically transfer

9081

FIG. 3. The presence of HIV-1 products in target cells does not rely on infection events. (A) AZT treatment does not influence HIV-1 Gag accumulation in target cells. FACS analyses of levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured for 16 h with U937/HIV-1 in the presence of 10 or 100 ␮M AZT (upper panels). As controls (Ctrl), human lymphoblastoid HIV-1-infected CEMss cells were treated with the same concentrations of AZT and analyzed for HIV-1 Gag expression 2 days later (lower panels). FSC, forward scatter. (B) Kinetics of HIV-1 Gag stability in target cells after separation from HIV-1-infected donor cells. FACS analyses of the levels of HIV-1 Gag/CD8 in 293/CD8T cells cocultured with U937/HIV-1 for 16 h and then separated from the donor cells. The target cells were analyzed immediately (time zero) or at the indicated times in culture after immunoselection. The results reported in both panels A and B are representative of two independent experiments. The percentages of Gag-positive cells in the total numbers of CD8-positive cells are shown in the coculture plots, whereas the percentages of HIV-1 Gag-expressing cells are shown in the lower plots of panel A. ␣, anti.

HIV-1 products to CD4-negative cells through a mechanism requiring cell-to-cell contact but not the presence of cell-free viral particles. Screening analyses for donor and target cell activities. To broaden the significance of our findings, we next tested additional cell types for their ability to act as donor or target cells. Among donor cell candidates, we focused our studies on macrophages and activated lymphocytes, i.e., the cell types that best replicate HIV-1. To this end, we infected human primary monocyte-derived macrophages (MDMs) with 150 ng/105 cells of the R5-tropic ADA HIV-1 strain, and 7 days later, the HIV-1 Env gp120- expressing subpopulation was selected. Thereafter, 293/CD8T cells were added at a 2:1 cell ratio, and 16 h later, the coculture was analyzed by FACS for both HIV-1 Gag products and CD8. We also infected phytohemagglutinin (PHA)-activated CD4 lymphocytes with 50 ng of HIV-1 strain

9082

MURATORI ET AL.

FIG. 4. Gag and Env HIV-1 products accumulate and colocalize in donor cell-target cell contact zones. (A) Confocal microscope images of cocultures of 293/CD8T cells with U937/HIV-1 cells after labeling with both anti-CD8 (red fluorescence) and anti-Gag CAp24 (green fluorescence) MAbs. The same field was scanned at three section points. The arrow indicates a zone of high levels of accumulation of Gag products, seemingly corresponding to the contact area between two 293/CD8T cells and a U937/HIV-1 cell. (B) Fluorescence microscope images of 293/CD8T cells purified after 16 h of cocultivation with U937/HIV-1 cells and labeled for cell membrane-associated Env and intracellular Gag products in the presence of DAPI (4⬘,6⬘-diamidino-2-phenylindole). Arrows indicate the most evident zones of Gag-Env colocalization. (C and D) Fluorescence microscope images of 293/CD8T-U937/HIV-1 cocultures 16 h after the cultures were set. Cells were first labeled with Texas Red (C)- or FITC (D)-conjugated anti-CD8 MAbs, then permeabilized, and finally stained with DAPI and either FITC-conjugated anti-CAp24 MAb (C) or anti-Env gp120

J. VIROL.

VSV-G NL4-3 per 105 cells to ensure the simultaneous infection of the majority of cells, and 24 h later, cocultured the cells with either 293/CD8T cells or uninfected, activated CD4 lymphocytes for an additional 16 h in the presence of 20 ␮M AZT. We detected HIV-1 Gag products in 293/CD8T cells cocultured with infected macrophages (Fig. 2A) but not with infected CD4 lymphocytes, which readily transferred HIV-1 products to autologous uninfected CD4 lymphocytes (Fig. 2B). Overall, these results confirmed the data we obtained from the cell lines. Among the nonepithelial CD4-negative target cells we tested, both PHA-activated (Fig. 2C) and quiescent (data not shown) CD8 lymphocytes appeared resistant to the transmission of HIV-1 products. Conversely, both human astrocyte U87 cells and human endothelial EA-hy 926 cells efficiently internalized HIV-1 Gag products (Fig. 2C). These results suggest that the transfer of HIV-1 products to CD4-negative cells may be operative within cells of the central nervous system as well as in endothelial compartments, thus adding relevance to our findings. The presence of HIV-1 products in target cells is not the consequence of viral replication events. We next sought to establish whether the accumulation of HIV-1 products in target cells was at least in part the consequence of viral expression. To this end, we treated 293/CD8T-U937/HIV-1 cocultures with different concentrations of the reverse-transcription inhibitor AZT. We observed that even high doses of AZT did not influence the efficiency of the transmission of HIV-1 products (Fig. 3A), strongly suggesting that at least most of the HIV-1 proteins we found accumulated in target cells did not originate from the transcription of the neosynthesized viral genome. Very low amounts of infectious HIV-1 particles were consistently detected in the supernatants of 293/CD8T cells purified after cocultivation (data not shown). Furthermore, the almost complete disappearance of Gag-related products 16 h after the end of the coculture (Fig. 3B) was not consistent with an authentic HIV-1 infection of target cells. We concluded that target cells accumulate HIV-1 products upon cocultivation with infected macrophagic cells without becoming infected. HIV-1 products accumulate in the zones of cell-to-cell contact. Cell-to-cell transmission of HIV-1 among CD4-positive cells occurs by means of the formation of zones of cell-to-cell adhesion (virological synapses), where viral products preferentially accumulate and colocalize as the consequence of a phenomenon of polarization (20, 34, 42). By confocal microscope analysis of the permeabilized cocultures, we found areas of strong accumulation of HIV-1 Gag products in the zones of contact between donor (marked by the strong Gag-related signal) and target (labeled by anti-CD8 MAb) cells strongly reminiscent of virological synapses (Fig. 4A). Furthermore, we also detected intense signals of Gag-Env colocalization by fluorescence microscope analysis of 293/CD8T cells purified after cocultivation and labeled for cell membrane-associated Env

human MAb followed by Texas Red-conjugated anti-human immunoglobulin G (D). Arrows indicate the zones of accumulation of HIV-1 products overlapping the cell-to-cell adhesion areas. For all panels, the bars represent 10 ␮M. ␣, anti.

VOL. 81, 2007

MMP-9 AND CELL-TO-CELL TRANSFER OF HIV-1 PRODUCTS

9083

FIG. 5. Neither Env nor Nef is required for the transfer of HIV-1 products. FACS analyses of the levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured with U937 cells (upper panels) or 7-day-old MDMs (lower panels) previously infected with HIV-1 strain VSV-G wt, ⌬env, or ⌬nef are shown. As negative-control target cells, PHA-activated CD8 lymphocytes from the same MDM donors were used. The percentage of Gag-expressing cells in the total number of CD8-positive cells is reported in each plot. Results are representative of two (for U937 cells) or four (for MDMs) independent experiments. PBLs, peripheral blood lymphocytes; ␣, anti.

gp120 and intracellular Gag products (Fig. 4B). Finally, the polarization of HIV-1 products at the cell-to-cell contact zones was assessed by permeabilizing and labeling the cocultures with either anti-Gag CAp24 (Fig. 4C) or anti-Env gp120 (Fig. 4D) MAbs together with anti-CD8 MAb for the identification of target cells. The reproducible accumulation of Gag or Env we found in the zones of cell-to-cell contact supports the idea that cell-to-cell passage of HIV-1 products indeed occurs upon the polarization process. In summary, our imaging data indicate that HIV-1 products preferentially accumulate and colocalize in cell-to-cell contact areas highly reminiscent of virological synapses. Cell-to-cell transmission of HIV-1 products does not require expression of Env or Nef. Although the transmission of HIV-1 products we described here did not depend on the Env-CD4 interaction, a role for HIV-1 Env could not be formally excluded, since it can bind additional cell membrane molecules such as galactosylceramide (1, 4) and sulfatide (37). Meanwhile, we also investigated the involvement of Nef, which, considering its effects on a wide array of cell membrane molecules (47), may influence the cell-to-cell contact process. We challenged U937 cells or MDMs with either wt VSV-G HIV-1 or VSV-G HIV-1 with either env or nef deleted, a procedure leading to the infection of more than 80% of cells within 24 to 48 h (24, 38). After 24 h, 293/CD8T cells were added in the presence of 20 ␮M AZT. No major quantitative differences in the transfer efficiency of HIV-1 among the different infected cultures were noticed (Fig. 5), indicating that the expression of neither Env nor Nef is critical for the cell-to-cell transmission of HIV-1 products. HIV-1 protease inhibitors or inhibitors of metalloproteinases block cell-to-cell transmission of HIV-1 products. We found that treatment with the HIV protease inhibitors (PIs) ritonavir, saquinavir, and indinavir potently inhibited the cell-

to-cell transmission of HIV-1 products to epithelial cells (Fig. 6A). A finding of major importance is that this phenomenon did not depend on the effects of PIs on Gag processing in donor cells, since the same outcome was obtained by testing the transfer of Env gp120 (Fig. 6B). Besides blocking virus maturation by inhibiting Gag p55 processing, PIs induce a number of cellular effects, including inhibition of proteasome functions (3, 29) and metalloproteinase activity (6, 51). To establish whether MMPs were indeed involved in the inhibition of transmission of HIV-1 products, we treated the cocultures for 6 h with different concentrations of two broad spectrum inhibitors of MMPs, i.e., GM6001 and MMP IV. Interestingly, we found that both MMP inhibitors significantly reduced the transmission of HIV-1 products in a dose-dependent manner and at levels comparable to those found in PI-treated cocultures (Fig. 6C). This result suggests that the PI inhibitory effect we observed was at least in part mediated by MMPs. More importantly, our data support the idea that the activity of MMPs is involved in the cell-to-cell transmission of HIV-1 products. MMP-9 is involved in transmission of HIV-1 to target cells. Among the 24 human MMPs identified so far, MMP-9 was found to be abundantly produced by macrophages (56) and can be specifically induced by HIV-1 infection or HIV-1 products (16, 59). To investigate the possible involvement of MMP-9 in the transmission of HIV-1 products, its production was first measured in the supernatants of either uninfected or HIV-1infected macrophagic or lymphocytic cells in the presence of epithelial cells or without epithelial cells. To this end, we harvested the serum-free supernatants from U937, U937/HIV-1, D10, or 293/CD8T cells in the absence or in the presence of 293/CD8T cells after overnight cultivation. The amounts of 92-kDa MMP-9 were measured by zymograms, using gelatin as the activity substrate, which is also recognized by the 72-kDa

9084

MURATORI ET AL.

J. VIROL.

MMP-9 production, which was already detectable from the same number of uninfected cells (Fig. 7C) and was confirmed by the zymograms carried out on serial dilutions of the supernatants (Fig. 7D); and (ii) the amount of MMP-9 released by lymphocytes appeared low and was not influenced by HIV-1 expression (Fig. 7C). As with the cell line cocultures, the presence of the target cells did not alter the overall levels of MMP-9 release (data not shown). Thus, we found a correlation between HIV-1 expression levels in macrophages and the levels of MMP-9 released. However, to support the idea that the MMP-9 activity was indeed involved in the transmission of HIV-1 products to target cells, we evaluated HIV-1 Gag transfer in 293/CD8T cells in the presence of different concentrations of either anti-MMP-9neutralizing antibodies (Fig. 7E) or sodium orthovanadate (Fig. 7F). The latter was previously shown to inhibit the Tatdependent stimulatory effect on MMP-9 secretion observed in HIV-1-infected macrophages (35). Interestingly, in both experimental settings we observed a significant inhibition of HIV-1 Gag transfer efficiency without detectable effects on HIV-1 expression levels in donor cells (Fig. 7E and F) and in the absence of significant loss of cell viability (data not shown). As expected, reduced levels of MMP-9 were detected by ELISA in the supernatants of the orthovanadate-treated cocultures (data not shown). In summary, our results strongly suggest that the increased production of MMP-9 induced by HIV-1 in macrophages could be a significant participant in the molecular events leading to the transmission of HIV-1 products to epithelial cells. FIG. 6. Treatment with inhibitors of either HIV protease or MMPs blocks the transfer of HIV-1 products. FACS analyses of the levels of HIV-1 Gag (A) or Env gp120 (B) products in 293/CD8T cells cocultured for 16 h with U937/HIV-1 cells in the presence or absence (Ctrl) of 2 ␮M of the indicated PIs are shown. The percentage of Gagpositive cells in the total number of CD8 positive cells is indicated in each plot. Results are representative of six (for ritonavir) or three (for other PIs) independent experiments. (C) Percentages of HIV-1 Gagpositive 293/CD8T cells, as measured by FACS analysis, after 6 h of coculture with U937/HIV-1 cells in the presence or absence (Ctrl) of the indicated concentrations of either GM6001 or MMP IV are shown. Cocultures treated with 2 ␮M indinavir (IDV) were included as controls. The percentage of HIV-1-expressing U937/HIV-1 cells was 70 to 75% for all experiments. The means ⫾ the standard deviations of the results of three independent experiments are shown. ␣, anti.

MMP-2. As shown in Fig. 7A, HIV-1 expression in U937 cells was correlated with increased MMP-9 release, while the level of MMP-2 remained basically unchanged. On the other hand, MMP-2 was barely detectable in both epithelial and HIV-1infected lymphocytic cells. Note that the overall outcome did not change significantly in the coculture supernatants, suggesting that the increase in MMP-9 was not the result of the donor cell-to-target cell interaction. To analyze the levels of MMP-9 production from primary macrophages and CD4 lymphocytes, 7-day-old MDMs or PHA-stimulated CD4 lymphocytes were tested before and upon infection with VSV-G NL4-3 HIV-1. Two days after the challenge, the cells were cultivated for 16 h in serum-free medium alone or in the presence of epithelial target cells. We observed that when more than 80% of the cells were infected (Fig. 7B), (i) HIV-1 infection of MDMs further increased

DISCUSSION Here we showed that contact between HIV-1-infected macrophages and CD4-negative epithelial or endothelial cells or astrocytes can lead to transfer of HIV-1 proteins in the absence of viral infection. This phenomenon may at least in part be the reason for the remarkable amounts of HIV-1 products frequently detected in CD4-negative cells of specimens from AIDS patients (11, 45, 50). The transmission of HIV-1 products we described here might be reminiscent of trogocytosis, i.e., cell-to-cell passage of cell material mediated by transfer of membrane patches and associated proteins from the surface of one cell to another following synapse formation (9). However, the fact that we did not detect the transfer of cell membraneanchored products expressed by macrophages but not by 293/ CD8T cells (i.e., CD45, ICAM-1, LFA-1) was not consistent with the idea that HIV-1 products are transmitted through trogocytosis. Furthermore, the lack of transfer of these cell markers that are expressed at high levels in macrophages strongly suggests that the protein transfer is not dictated simply by the high amounts of the protein expressed in donor cells. Rather, it appears to be a process basically involving the viral proteins that, conversely, also undergo cell transfer when expressed at relatively low levels, as occurred in R5 ADA HIV1-infected MDMs. ICAM-1 and LFA-1 adhesion molecules were previously found to be important for the formation of virological synapses among CD4 lymphocytes (33). Although virological synapselike formations were also detected in our system, we did not find an obvious correlation between expression levels of

VOL. 81, 2007

MMP-9 AND CELL-TO-CELL TRANSFER OF HIV-1 PRODUCTS

9085

ICAM-1 and LFA-1 and the efficiency of different cell types to act as donor/target cells (data not shown). This may imply that additional adhesion molecules are involved in the macrophage-induced transmission of HIV-1 products. We found that HIV-1 proteins were rapidly transferred to target cells upon their accumulation around zones of cell-tocell contact and apparently in the absence of cell-to-cell fusion events, as suggested by the lack of detectable CFSE diffusion from donor to target cells and vice versa. Possibly, cell-to-cell contact drives a rapid polarization of viral products toward the contact zones by means of a cytoskeleton rearrangement, as previously described for both human T-cell lymphotropic virus type 1 and HIV-1 cell-to-cell transmission. Stabilization of cell-to-cell contact may then favor the passage of viral products through not-yet-characterized selective mechanisms. The relevance of our findings was enforced by the fact that primary macrophages infected with an R5-tropic HIV-1 isolate were also found acting efficiently as donor cells. On the other hand, evidence that astrocytes can internalize valuable amounts of HIV-1 products as well suggests the possibility that such a mechanism could contribute to AIDS-associated neurological disorders. HIV-1 products, including Env gp120 (44, 63), Nef (2), and Tat (10, 64), have consistently been shown to have relevant effects on the functions of CD4-negative nerve cells. Furthermore, whether endothelial cells activate HIVspecific lymphocytes upon contact with infected macrophages by acting as antigen-presenting cells deserves further investigation. We noticed that HIV-1 infection significantly enhanced the secretion of MMP-9 in the supernatants of both U937 cells and primary macrophages, but discordant results can be found in the literature. In fact, both increased (16, 59) and reduced (12, 28) MMP-9 production as a consequence of HIV-1 infection of MDMs has been reported. Possibly, different culture and/or infection conditions account for such inconsistencies. In our study, however, it should be noted that challenge of MDMs with VSV-G HIV-1 ensured rapid infection of the majority of the cells. This finding may be relevant in the overall evaluation of the response of MDMs to HIV-1 infection. In addition, the fact that we routinely purified HIV-1 preparations on 20% sucrose cushions formally excludes biases produced by contamination with soluble factors (i.e., cytokines and chemokines) that are good MMP inducers. In any case, the relevant role played by MMP-9 on HIV-1 protein transfer was highlighted by the specific inhibition we observed by impairing its secretion

FIG. 7. MMP-9 is involved in the transfer of HIV-1 products to target cells. (A) Zymograms of supernatants of either uninfected or HIV-1infected cell lines. Supernatants from the indicated cells cultivated alone or in the presence of 293/CD8T cells (co-c) were analyzed for gelatinase activity. Ten microliters of serum-free supernatant from U937 cells treated for 2 days with 50 ng/ml of tetradecanoyl phorbol acetate was used as a positive control. The data are representative of two independent experiments. (B) FACS analyses of the levels of HIV-1 Gag in both CD4 lymphocyte and MDM-infected cultures. The percentages of HIV-1 Gagpositive cells are indicated. (C) Zymograms of the supernatants of uninfected or HIV-1-infected primary cells. Analyses were carried out on the supernatants from uninfected or HIV-1-infected CD4 primary lymphocytes or MDMs cultivated alone or in the presence of 293/CD8T cells (co-c). The data are representative of experiments carried out on cells

from two healthy donors. (D) Semiquantitative analyses of the levels of MMP-9 released in the supernatants from equal numbers of either uninfected or HIV-1-infected MDMs. The data are representative of experiments carried out on cells from two healthy donors. In panels A, B, and D, MMP-2 and/or MMP-9 migration is indicated on the left, and molecular marker sizes are shown on the right. (E) FACS analyses of the levels of HIV-1 Gag products and CD8 in 293/CD8T cells cocultured with U937/HIV-1 in the presence of 5 ␮g/ml of an unspecific mouse immunoglobulin G or 2.5 or 5 ␮g/ml of anti-MMP-9neutralizing MAb or (F) the indicated concentrations of sodium orthovanadate are shown. The percentage of Gag-positive cells in the total number of CD8-positive cells is shown in each plot. For both panels E and F, the results are representative of two independent experiments. PBLs, peripheral blood lymphocytes; ␣, anti.

9086

MURATORI ET AL.

or its activity. These results imply that the increase in MMP-9 activity induced by HIV-1 is not counteracted by a parallel increase in the release of natural MMP-9 inhibitors, e.g., tissue inhibitors of metalloproteinases (60). We did not detect significant increases in MMP-9 production from CD4 lymphocytes upon HIV-1 infection, unlike what we observed with macrophages. This was not consistent with the previously reported HIV-1-dependent increase in MMP-9 release (62) that was detected in the supernatants from total peripheral blood lymphocytes. Transfer of HIV-1 appeared to be cell type dependent for both donor and target cells. Evidence that infected CD4 lymphocytes transfer HIV-1 products only to CD4-positive target cells suggests that synapse formation is strongly favored by the Env/CD4 cell-to-cell interaction, but when target cells do not express CD4, it is conceivable that adequate levels of MMP-9 are required for the formation of functional synapses. It is possible that robust MMP-9 activity contributes to the optimal presentation of adhesion molecules required for the formation of cell-to-cell adhesion. Alternatively, or in addition, proteolysis of the extracellular matrix induced by MMP-9 activity may be important in creating optimal conditions for the formation of functional synapses. A critical point deserving additional investigation is the identification of the molecular components of the cell-to-cell adhesion zones through which HIV-1 products undergo cell transfer. A more general question is whether the transmission of HIV-1 products can be considered a defensive mechanism of the infected host or, conversely, a means for the virus to promote spreading. The latter hypothesis seems more plausible. In fact, one can envision that the massive presentation in major histocompatibility complex class I of viral products through mechanisms of cross-presentation already described for HIV-1 (7) could lead to selective destruction of epithelial cells induced by HIV-1-specific cytotoxic lymphocytes, thus facilitating the diffusion of both the virus and virus-infected cells, for instance, across the blood-brain or intestinal barriers. In addition, if transmission of HIV-1 products also takes place in professional (e.g., dendritic cells, B lymphocytes) or nonprofessional (e.g., endothelial cells) antigen-presenting cells, it may also lead to hyperstimulation and replication of HIV-1specific T lymphocytes. This could be a great advantage for the virus, since HIV-1-specific CD4 lymphocytes are the major targets of viral replication (15, 19, 30). In such a context, the overall therapeutic benefits induced by PI treatment would include indirect effects, such as the inhibition of transmission of HIV-1 products described here. ACKNOWLEDGMENTS This work was supported by grants from the AIDS project of the Ministry of Health, Rome, Italy. We are indebted to Emanuela Salvi and Patrizia Leone, National AIDS Center, Istituto Superiore di Sanita`, Rome, Italy, for helpful technical support and to Federica M. Regini for excellent editorial assistance. AZT, ritonavir, indinavir, and saquinavir, as well as 4G10 and 2G12 human anti-HIV-1 Env gp120 MAbs, were obtained from the NIH AIDS Research and Reference Reagent Program. REFERENCES 1. Alfsen, A., P. Iniguez, E. Bouguyon, and M. Bomsel. 2001. Secretory IgA specific for a conserved epitope on gp41 envelope glycoprotein inhibits epithelial transcytosis of HIV-1. J. Immunol. 166:6257–6265.

J. VIROL. 2. Ambrosini, E., N. Slepko, B. Kohleisen, E. Shumay, V. Erfle, F. Aloisi, and G. Levi. 1999. HIV-1 Nef alters the expression of betaII and epsilon isoforms of protein kinase C and the activation of the long terminal repeat promoter in human astrocytoma cells. Glia 27:143–151. 3. Andre, P., M. Groettrup, P. Klenerman, R. de Guili, B. L. Booth, Jr., V. Cerundolo, M. Bonneville, F. Jotereau, R. M. Zinkernagel, and V. Lotteau. 1998. An inhibitor of HIV-1 protease modulates proteasome activity, antigen presentation, and T cell responses. Proc. Natl. Acad. Sci. USA 95:13120– 13124. 4. Bhat, S., S. L. Spitalnik, F. Gonzalez-Scarano, and D. H. Silberberg. 1991. Galactosyl ceramide or a derivative is an essential component of the neural receptor for human immunodeficiency virus type 1 envelope glycoprotein gp120. Proc. Natl. Acad. Sci. USA 88:7131–7134. 5. Blanco, J., B. Bosch, M. T. Fernandez-Figueras, J. Barretina, B. Clotet, and J. A. Este. 2004. High level of coreceptor-independent HIV transfer induced by contacts between primary CD4 T cells. J. Biol. Chem. 279:51305–51314. 6. Bourlier, V., A. Zakaroff-Girard, S. De Barros, C. Pizzacalla, V. D. de Saint Front, M. Lafontan, A. Bouloumie´, and J. Galitzky. 2005. Protease inhibitor treatments reveal specific involvement of matrix metalloproteinase-9 in human adipocyte differentiation. J. Pharmacol. Exp. Ther. 312:1272–1279. 7. Buseyne, F., S. Le Gall, C. Boccaccio, J. P. Abastado, J. D. Lifson, L. O. Arthur, Y. Rivie`re, J. M. Heard, and O. Schwartz. 2001. MHC-I-restricted presentation of HIV-1 virion antigens without viral replication. Nat. Med. 7:344–349. 8. Carr, J. M., H. Hocking, P. Li, and C. J. Burrell. 1999. Rapid and efficient cell-to-cell transmission of human immunodeficiency virus infection from monocyte-derived macrophages to peripheral blood lymphocytes. Virology 265:319–329. 9. Caumartin, J., J. Lemaoult, and E. D. Carosella. 2006. Intercellular exchanges of membrane patches (trogocytosis) highlight the next level of immune plasticity. Transplant Immunol. 17:20–22. 10. Chauhan, A., J. Turchan, C. Pocernich, A. Bruce-Keller, S. Roth, D. A. Butterfield, E. O. Major, and A. Nath. 2003. Intracellular human immunodeficiency virus Tat expression in astrocytes promotes astrocyte survival but induces potent neurotoxicity at distant sites via axonal transport. J. Biol. Chem. 278:13512–13519. 11. Chi, D., J. Henry, J. Kelley, R. Thorpe, J. K. Smith, and G. Krishnaswamy. 2000. The effects of HIV infection on endothelial function. Endothelium 7:223–242. 12. Ciborowski, P., Y. Enose, A. Mack, M. Fladseth, and H. E. Gendelman. 2004. Diminished matrix metalloproteinase 9 secretion in human immunodeficiency virus-infected mononuclear phagocytes: modulation of innate immunity and implications for neurological disease. J. Neuroimmunol. 157:11–16. 13. Cocchi, F., L. Menotti, P. Dubreuil, M. Lopez, and G. Campadelli-Fiume. 2000. Cell-to-cell spread of wild-type herpes simplex virus type 1, but not of syncytial strains, is mediated by the immunoglobulin-like receptors that mediate virion entry, nectin1 (PRR1/HveC/HIgR) and nectin2 (PRR2/HveB). J. Virol. 74:3909–3917. 14. d’Aloja, P., A. C. Santarcangelo, S. Arold, A. Baur, and M. Federico. 2001. Genetic and functional analysis of the human immunodeficiency virus (HIV) type 1-inhibiting F12-HIVnef allele. J. Gen. Virol. 82:2735–2745. 15. Demoustier, A., B. Gubler, O. Lambotte, M. G. de Goer, C. Wallon, C. Goujard, J. F. Delfraissy, and Y. Taoufik. 2002. In patients on prolonged HAART, a significant pool of HIV infected CD4 T cells are HIV-specific. AIDS 16:1749–1754. 16. Dhawan, S., L. M. Wahl, A. Heredia, Y. Zhang, J. S. Epstein, M. S. Meltzer, and I. K. Hewlett. 1995. Interferon-gamma inhibits HIV-induced invasiveness of monocytes. J. Leukoc. Biol. 58:713–716. 17. Digel, M., K. L. Sampaio, G. Jahn, and C. Sinzger. 2006. Evidence for direct transfer of cytoplasmic material from infected to uninfected cells during cell-associated spread of human cytomegalovirus. J. Clin. Virol. 37:10–20. 18. Dimitrov, D. S., R. L. Willey, H. Sato, L. J. Chang, R. Blumenthal, and M. A. Martin. 1993. Quantitation of human immunodeficiency virus type 1 infection kinetics. J. Virol. 67:2182–2190. 19. Douek, D. C., J. M. Brenchley, M. R. Betts, D. R. Ambrozak, B. J. Hill, Y. Okamoto, J. P. Casazza, J. Kuruppu, K. Kunstman, S. Wolinsky, Z. Grossman, M. Dybul, A. Oxenius, D. A. Price, M. Connors, and R. A. Koup. 2002. HIV preferentially infects HIV-specific CD4⫹ T cells. Nature 417:95–98. 20. Dustin, M. 2003. Viral spread through protoplasmic kiss. Nat. Cell Biol. 5:271–272. 21. Edgell, C. J., C. C. McDonald, and J. B. Graham. 1983. Permanent cell line expressing human factor VIII-related antigen established by hybridization. Proc. Natl. Acad. Sci. USA 80:3734–3737. 22. Ehrengruber, M. U., E. Ehler, M. A. Billeter, and H. Y. Naim. 2002. Measles virus spreads in rat hippocampal neurons by cell-to-cell contact and in a polarized fashion. J. Virol. 76:5720–5728. 23. Fackler, O. T., and B. M. Peterlin. 2000. Endocytic entry of HIV-1. Curr. Biol. 10:1005–1008. 24. Federico, M., Z. Percario, E. Olivetta, G. Fiorucci, C. Muratori, A. Micheli, G. Romeo, and E. Affabris. 2001. HIV-1 Nef activates STAT1 in human monocytes/macrophages through the release of soluble factors. Blood 98: 2752–2761.

VOL. 81, 2007

MMP-9 AND CELL-TO-CELL TRANSFER OF HIV-1 PRODUCTS

25. Federico, M., F. Titti, S. Butto `, A. Orecchia, F. Carlini, B. Taddeo, B. Macchi, N. Maggiano, P. Verani, and G. B. Rossi. 1989. Biologic and molecular characterization of producer and nonproducer clones from HUT-78 cells infected with a patient HIV isolate. AIDS Res. Hum. Retrovir. 5:385– 396. 26. Folks, T. M., D. Powell, M. Lightfoote, S. Koenig, A. S. Fauci, S. Benn, A. Rabson, D. Daugherty, H. E. Gendelman, and M. D. Hoggan. 1986. Biological and biochemical characterization of a cloned Leu-3⫺ cell surviving infection with the acquired immune deficiency syndrome retrovirus. J. Exp. Med. 164:280–290. 27. Fredericksen, B. L., B. L. Wei, J. Yao, T. Luo, and J. V. Garcia. 2002. Inhibition of endosomal/lysosomal degradation increases the infectivity of human immunodeficiency virus. J. Virol. 76:11440–11446. 28. Ghorpade, A., R. Persidskaia, R. Suryadevara, M. Che, X. J. Liu, Y. Persidsky, and H. E. Gendelman. 2001. Mononuclear phagocyte differentiation, activation, and viral infection regulate matrix metalloproteinase expression: implications for human immunodeficiency virus type 1-associated dementia. J. Virol. 75: 6572–6583. 29. Goldberg, A. L., and K. Rock. 2002. Not just research tools—proteasome inhibitors offer therapeutic promise. Nat. Med. 8:338–340. 30. Harari, A., G. P. Rizzardi, K. Ellefsen, D. Ciuffreda, P. Champagne, P. A. Bart, D. Kaufmann, A. Telenti, R. Sahli, G. Tambussi, L. Kaiser, A. Lazzarin, L. Perrin, and G. Pantaleo. 2002. Analysis of HIV-1- and CMV-specific memory CD4 T-cell responses during primary and chronic infection. Blood 100:1381– 1387. 31. Hocini, H., and M. Bomsel. 1999. Infectious human immunodeficiency virus can rapidly penetrate a tight human epithelial barrier by transcytosis in a process impaired by mucosal immunoglobulins. J. Infect. Dis. 179:S448– S453. 32. Igakura, T., J. C. Stinchcombe, P. K. Goon, G. P. Taylor, J. N. Weber, G. M. Griffiths, Y. Tanaka, M. Osame, and C. R. Bangham. 2003. Spread of HTLV-I between lymphocytes by virus-induced polarization of the cytoskeleton. Science 299:1713–1716. 33. Jolly, C., K. Kashefi, M. Hollinshead, and Q. J. Sattentau. 2004. HIV-1 cell to cell transfer across an Env-induced, actin-dependent synapse. J. Exp. Med. 199:283–293. 34. Jolly, C., and Q. J. Sattentau. 2004. Retroviral spread by induction of virological synapses. Traffic 5:643–650. 35. Kumar, A., S. Dhawan, A. Mukhopadhyay, and B. B. Aggarwal. 1999. Human immunodeficiency virus-1-tat induces matrix metalloproteinase-9 in monocytes through protein tyrosine phosphatase-mediated activation of nuclear transcription factor NF-␬B. FEBS Lett. 462:140–144. 36. Lambotte, O., Y. Taoufik, M. G. de Goer, C. Wallon, C. Goujard, and J. F. Delfraissy. 2000. Detection of infectious HIV in circulating monocytes from patients on prolonged highly active antiretroviral therapy. J. Acquir. Immune Defic. Syndr. 23:114–119. 37. McAlarney, T., S. Apostolski, S. Lederman, and N. Latov. 1994. Characteristics of HIV-1 gp120 glycoprotein binding to glycolipids. J. Neurosci. Res. 37:453–460. 38. Olivetta, E., and M. Federico. 2006. HIV-1 Nef protects human-monocytederived macrophages from HIV-1-induced apoptosis. Exp. Cell Res. 312: 890–900. 39. Olivetta, E., K. Pugliese, R. Bona, P. d’Aloja, F. Ferrantelli, A. C. Santarcangelo, G. Mattia, P. Verani, and M. Federico. 2000. cis expression of the F12 human immunodeficiency virus (HIV) Nef allele transforms the highly productive NL4-3 HIV type 1 to a replication-defective strain: involvement of both Env gp41 and CD4 intracytoplasmic tails. J. Virol. 74:483–492. 40. Peretti, S., I. Schiavoni, K. Pugliese, and M. Federico. 2005. Cell death induced by the herpes simplex virus-1 thymidine kinase delivered by human immunodeficiency virus-1-based virus-like particles. Mol. Ther. 12:1185– 1196. 41. Persidsky, Y., and H. E. Gendelman. 2003. Mononuclear phagocyte immunity and the neuropathogenesis of HIV-1 infection. J. Leukoc. Biol. 74:691– 701. 42. Piguet, V., and Q. Sattentau. 2004. Dangerous liaisons at the virological synapse. J. Clin. Investig. 114:605–610. 43. Ponten, J., and E. H. Macintyre. 1968. Long term culture of normal and neoplastic human glia. Acta Pathol. Microbiol. Scand. 74:465–486. 44. Power, C., J. C. McArthur, A. Nath, K. Wehrly, M. Mayne, J. Nishio, T. Langelier, R. T. Johnson, and B. Chesebro. 1998. Neuronal death induced by

45.

46.

47.

48.

49.

50.

51.

52.

53.

54. 55. 56.

57.

58.

59.

60.

61.

62.

63.

64.

9087

brain-derived human immunodeficiency virus type 1 envelope genes differs between demented and nondemented AIDS patients. J. Virol. 72:9045–9053. Qiao, X., B. He, A. Chiu, D. M. Knowles, A. Chadburn, and A. Cerutti. 2006. Human immunodeficiency virus 1 Nef suppresses CD40-dependent immunoglobulin class switching in bystander B cells. Nat. Immunol. 7:302–310. Ramos-DeSimone, N., U. M. Moll, J. P. Quigley, and D. L. French. 1993. Inhibition of matrix metalloproteinase 9 activation by a specific monoclonal antibody. Hybridoma 12:349–363. Roeth, J. F., and K. L. Collins. 2006. Human immunodeficiency virus type 1 Nef: adapting to intracellular trafficking pathways. Microbiol. Mol. Biol. Rev. 70:548–563. Sato, H., J. Orenstein, D. Dimitrov, and M. Martin. 1992. Cell-to-cell spread of HIV-1 occurs within minutes and may not involve the participation of virus particles. Virology 186:712–724. Schaeffer, E., V. B. Soros, and W. C. Greene. 2004. Compensatory link between fusion and endocytosis of human immunodeficiency virus type 1 in human CD4 T lymphocytes. J. Virol. 78:1375–1383. Schwartz, L., and E. O. Major. 2006. Neural progenitors and HIV-1-associated central nervous system disease in adults and children. Curr. HIV Res. 4:319–327. Sgadari, C., G. Barillari, E. Toschi, D. Carlei, I. Bacigalupo, S. Baccarini, C. Palladino, P. Leone, R. Bugarini, L. Malavasi, A. Cafaro, M. Falchi, D. Valdembri, G. Rezza, F. Bussolino, P. Monini, and B. Ensoli. 2002. HIV protease inhibitors are potent anti-angiogenic molecules and promote regression of Kaposi sarcoma. Nat. Med. 8:225–232. Sherer, N. M., M. J. Lehmann, L. F. Jimenez-Soto, C. Horensavitz, M. Pypaert, and W. Mothes. 2007. Retroviruses can establish filopodial bridges for efficient cell-to-cell transmission. Nat. Cell Biol. 9:310–315. Sol-Foulon, N., M. Sourisseau, F. Porrot, M. I. Thoulouze, C. Trouillet, C. Nobile, F. Blanchet, V. di Bartolo, N. Noraz, N. Taylor, A. Alcover, C. Hivroz, and O. Schwartz. 2007. ZAP-70 kinase regulates HIV cell-to-cell spread and virological synapse formation. EMBO J. 26:516–526. Sternlicht, M. D., and Z. Werb. 2001. How matrix metalloproteinases regulate cell behavior. Annu. Rev. Cell Dev. Biol. 17:463–516. Stevenson, M. 2003. HIV-1 pathogenesis. Nat. Med. 9:853–860. St.-Pierre, Y., T. C. Van, and P. O. Esteve. 2003. Emerging features in the regulation of MMP-9 gene expression for the development of novel molecular targets and therapeutic strategies. Curr. Drug Targets Inflamm. Allergy 2:206–215. Turville, S., J. Wilkinson, P. Cameron, J. Dable, and A. L. Cunningham. 2003. The role of dendritic cell C-type lectin receptors in HIV pathogenesis. J. Leukoc. Biol. 74:710–718. Valli, M. B., A. Serafino, A. Crema, L. Bertolini, A. Manzin, G. Lanzilli, C. Bosman, S. Iacovacci, S. Giunta, A. Ponzetto, M. Clementi, and G. Carloni. 2006. Transmission in vitro of hepatitis C virus from persistently infected human B-cells to hepatoma cells by cell-to-cell contact. J. Med. Virol. 78: 192–201. Vazquez, N., T. Greenwell-Wild, N. J. Marinos, W. D. Swaim, S. Nares, D. E. Ott, U. Schubert, P. Henklein, J. M. Orenstein, M. B. Sporn, and S. M. Wahl. 2005. Human immunodeficiency virus type 1-induced macrophage gene expression includes the p21 gene, a target for viral regulation. J. Virol. 79:4479–4491. Visse, R., and H. Nagase. 2003. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry. Circ. Res. 92:827–839. Webster, N. L., and S. M. Crowe. 2006. Matrix metalloproteinases, their production by monocytes and macrophages and their potential role in HIVrelated diseases. J. Leukoc. Biol. 80:1052–1066. Weeks, B. S., M. E. Klotman, E. Holloway, W. G. Stetler-Stevenson, H. K. Kleinman, and P. E. Klotman. 1993. HIV-1 infection stimulates T cell invasiveness and synthesis of the 92-kDa type IV collagenase. AIDS Res. Hum. Retrovir. 9:513–518. Zhang, K., F. Rana, C. Silva, J. Ethier, K. Wehrly, B. Chesebro, and C. Power. 2003. Human immunodeficiency virus type 1 envelope-mediated neuronal death: uncoupling of viral replication and neurotoxicity. J. Virol. 77: 6899–6912. Zhou, B. Y., Y. Liu, B. Kim, Y. Xiao, and J. J. He. 2004. Astrocyte activation and dysfunction and neuron death by HIV-1 Tat expression in astrocytes. Mol. Cell. Neurosci. 27:296–305.