|Home | About | Journals | Submit | Contact Us | Français|
Transmigration of human immunodeficiency virus (HIV)-infected mononuclear cells through the genital mucosa is one of the possible mechanisms of sexual transmission of HIV. Here, we investigated the transmigration of cell-associated R5-tropic HIV type 1 (HIV-1) through a tight monolayer of human epithelial cells in vitro. We show that this process is dependent on an initial interaction between αLβ2 integrin CD11a/CD18 on infected monocytic cells and intercellular adhesion molecule 2 (ICAM-2; CD102) and ICAM-3 (CD50) on the apical membrane of epithelial cells. The CD50 and CD102 ligands were overexpressed on epithelial cells when the cells were activated by proinflammatory cytokines in the cellular microenvironment. An accumulation of proviral DNA was found in the transmigrated cells, clearly reflecting the preferential transepithelial migration of HIV-1-infected cells under proinflammatory conditions. Our observations provide new insights supporting the hypothesis that HIV-infected mononuclear cells contained in genital secretions from infected individuals may cross the epithelial genital mucosa of an exposed receptive sexual partner, particularly under inflammatory conditions of damaged genital tissue. Understanding the fundamental aspects of the initial HIV entry process during sexual transmission remains a critical step for preventing human infection and developing further vaccinal strategies and virucidal agents.
Most cases of human immunodeficiency virus (HIV) infection worldwide occur following heterosexual contact, implying that the virus may breach the protective epithelial barrier lining the genital tract. Several mechanisms may be involved in the penetration of HIV through the mucosal barrier. Viral entry across a tight epithelial barrier may increase the risk for mucosal infection and systemic spread of the virus. Both free and cell-associated HIV is considered to be relevant for interindividual transmission through mucosae (49). Free and cell-associated HIV may directly reach submucosal CD4-positive cells, i.e., T cells, monocytes/macrophages, dendritic cells (DC), and Langerhans’ cells under conditions where the integrity of the mucosa is compromised. The latter hypothesis is supported by the increased risk of HIV transmission in individuals presenting with epithelial lesions (15). Mucosal DC capture HIV-1 by a CD4- and chemokine receptor-independent mechanism involving a C-type lectin DC-specific ligand termed DC-SIGN (17) and transmit the virus to T cells, thus initiating their productive infection and the regional spread of HIV type 1 (HIV-1) infection. Several hypotheses to explain the passage of free and cell-associated virus (i.e., virus present in infected cells) through the epithelial cell layer have been proposed. First, free HIV could directly infect CD4− epithelial cells in an env-independent fashion (34). Alternatively, simple epithelial monolayers, such as the endocervix and rectal mucosae, may support the rapid transcellular transport (or transcytosis) of free HIV-1 from the apical pole to the basolateral pole of the cell and thus provide viral access to the cells of the submucosa (7, 8, 10, 42). Transcytosis of free virus involves interactions between viral envelope glycoproteins and lectins such as d-(+)-mannose and galactosylcerebroside on the apical membrane of epithelial cells (23). Transcytosis is also a mechanism by which HIV, originating from infected cells (cell-associated virus), may pass through epithelial cells after having been delivered to the cells following the adherence of an infected mononuclear cell to the apical membrane of the epithelial cell. Finally, the passage of virus through the mucosa may occur by transmigration of infected mononuclear cells through gaps between epithelial cells. This mechanism was first described in vitro in 1996 by Tan and Phillips (42), who reported that, soon after the adherence of HIV-infected T cells or monocytes to monolayers of CD4-negative epithelial cells, budding of HIV from infected mononuclear cells onto the surface of epithelial cells occurs. Further studies have indicated that HIV-infected monocytes (35) or lymphoid H9 cells (28) may adhere to intact explants of the cervicovaginal epithelium to further advance between epithelial cells while secreting virus in a polarized fashion.
Transmigration of HIV-infected mononuclear cells through the genital mucosa may constitute a thus far underestimated and poorly understood mechanism for penetration of HIV through the genital barrier. Monostratified mucosae, such as the endocervical, urethral, and rectal epithelia, could be target tissues for transmigration of HIV-infected lymphocytes and monocytes contained in cervicovaginal and seminal secretions of HIV-infected individuals (44). The aims of the present study were (i) to reproduce the transmigration of HIV-infected monocytes through a monolayer of endometrial HEC-1 cells in vitro, (ii) to assess the influence of proinflammatory conditions on the transmigration process, and (iii) to evaluate the involvement of αLβ2 integrins that are upregulated on HIV-infected monocytes in the interaction between monocytes and epithelial cells.
Endotoxin-free reagents and plastics were used in all experiments. l-Glutamine-containing RPMI 1640 (BioWhittaker Europe, Verviers, France), saponin, formaldehyde, brefeldin A, and bovine serum albumin (BSA) (Sigma, St. Louis, Mo.), lymphocyte separation medium (Eurobio, Les Ulis, France), and human recombinant macrophage colony-stimulating factor (M-CSF), interleukin-1 (IL-1), tumor necrosis factor alpha (TNF-α), macrophage inflammatory protein 1α (MIP-1α)/CCL3, MIP-1β/CCL4, RANTES/CCL5, and monocyte chemotactic protein 1 (MCP-1)/CCL2 (R&D Systems, Minneapolis, Minn.) were obtained. Lipopolysaccharide (LPS) from Neisseria meningitidis was purified as previously described (45). The following murine monoclonal antibodies (MAb) were purchased from the indicated sources: anti-CD50 (clone HP2/19), anti-CD102 (clone B-T1), anti-CD54 (clone HA58), rhodamine-labeled anti-HIV-1 p24 (RD KC57), Beckman Coulter, Brea, Calif.; anti-CXCR4, anti-CCR5 (clone 2D7), anti-TNF-α, anti-IL-1β, fluorescein-conjugated anti-CD50 (anti-CD50 fluorescein isothiocyanate [FITC]), anti-CD11a FITC (clone H111), anti-CD54 FITC (clone HA58), MAb to human immunoglobulin G (IgG) isotypes, Becton Dickinson, San Diego, Calif., anti-CD102 FITC, Diaclone, Besançon, France. The allo-phycocyanin-labeled murine monoclonal anti-TNF-α was from Becton Dickinson. The following murine phycoerythrin (PE)-labeled anti-human chemokines were purchased as indicated: anti-IL-1α, anti-IL-1β, anti-IL-8, and anti-gamma interferon (IFN-γ), Becton Dickinson; anti-MCP-1 and anti-RANTES, R&D Systems. The rabbit polyclonal antibody and murine MAb (clone M77/3B3) to gp120/HIV-1 were purchased from ABI (Columbia, Md.). Intravenous IgG was provided by Sandoglobulin (Novartis, Basel, Switzerland).
Human monocytes were purified by allowing peripheral blood mononuclear cells isolated from buffy coats of healthy donors by centrifugation on lymphocyte separation medium to adhere to plastic culture dishes (Costar, Cambridge, Mass.) in the absence of serum, as previously described (37). The adherent cells contained more than 98% monocytes as assessed by staining with anti-CD14 MAb and cytofluorometric analysis. To obtain monocyte-derived macrophages (MDM), purified monocytes were cultured in RPMI 1640 containing 20% fetal calf serum (FCS) in the presence of M-CSF (10 ng/ml/106 cells) in 5% CO2 at 37°C. Every second day, a fraction (0.4 ml) of medium was removed and replaced by 0.5 ml of fresh medium containing M-CSF. On day 6, MDM (1 ml at 2 × 106 cells/ml) were then infected with the HIV-1Ba-L primary R5 strain (100 μl of 2 × 104 50% tissue culture infective doses/ml) for 18 h at 37°C. The cells were then extensively washed and cultured in RPMI 1640 containing 20% FCS and 50 ng of M-CSF/ml (2 ml). After 11 days of culture, infected MDM were collected for use in the transmigration assay. Infection of cells was quantified by measuring the amount of p24 antigen released in the culture medium by means of a capture enzyme-linked immunosorbent assay (NEN Life Science Products, Paris, France).
Membrane antigens of the cell subsets were analyzed by flow cytometry. Cells were incubated with decomplemented normal human AB serum (1.5 ml) for 15 min at 4°C to decrease nonspecific binding and centrifuged. The pellet was incubated with the different specific MAb (1 μg) for 30 min at 4°C, and cells were washed twice with phosphate-buffered saline (PBS) containing azide (0.01%) and BSA (0.2%) and then once in PBS and fixed using a 1% formaldehyde-PBS buffer solution. Cells were further analyzed by cytofluorometry. The following specific mouse anti-human antibodies were used: anti-CXCR4-PE and anti-CCR5-PE.
Flow-cytometric determination of intracellular HIV-1 production at the single-cell level was performed as previously described (14). MDM were first surface stained with FITC anti-CD11a MAb by standard procedures. The cells (106 cells/ml) then interacted with MAb RD KC57 (1 μg/106 cells) in 50 μl of PBS containing 0.2% BSA and 0.5% saponin for 20 min at room temperature. Stained cells were extensively washed in saponin buffer and in PBS and were further analyzed by cytofluorometry.
Confluent HEC-1 cells were first preincubated in the presence of brefeldin A (10 μg/ml) for 30 min at room temperature to avoid the release of cytokines in the culture supernatant. Cells were then cultured overnight under normal or proinflammatory conditions as previously described. Cells were then extensively washed in PBS without Ca2+ or Mg2+, trypsinized, and collected for intracellular staining. Flow cytometry determination of intracellular IL-1α, IL-1β, IL-8, IFN-γ, TNF-α, MCP-1, RANTES, MIP-1α, and MIP-1β reduction at the single-cell level was performed as described previously (14).
The human HEC-1 endometrial cell line was used as a model for the mucosal epithelial barrier and was obtained from the American Type Culture Collection (Manassas, Va.) (2). HEC-1 cells were derived from a patient with stage IA endometrial cancer (26). Cells were maintained in RPMI 1640 containing 10% FCS and antibiotics. HEC-1 cells were seeded in 12-mm-diameter, 3-μm-pore-size polycarbonate membrane transwells (Costar) at a density of 105 cells/well and cultured for 6 days until formation of tight junctions and of a well-differentiated monolayer was achieved. The transepithelial electrical resistance across the membrane was measured using a Millicel resistance system (ERS; Millipore, Bedford, Mass.) to ensure that the resistivity of the monolayer was of 200 Ω/cm2 prior to and after performing the transmigration assay (24). For the assay itself, infected MDM (1105 to 5105 per ml) collected at day 11 after infection were added to the apical chamber of the transwell system and allowed to transmigrate for between 30 min and 18 h in humidified 5% CO2 at 37°C. Transmigrated cells were recovered in the basal chamber of the transwell system and counted using Glastic 10 slides.
Infection with HIV of transmigrated MDM and HEC-1 cells was assessed by PCR of DNA extracted using the QIAamp DNA kit (Qiagen AG, Basel, Switzerland). One microgram of extracted DNA in a reaction volume of 50 μl containing Milli-Q water, 1.5 mM MgCl2, 200 μM (each) deoxynucleoside triphosphate, 500 nM (each) primer, and 1.25 IU of Taq polymerase (Promega, Charbonnières, France) was amplified by a single PCR of the pol HIV-1 gene by using the primer set comprising P63 (5′-GCCATTTAAAAATCTGAAAACAGG-3′) and P58 (5′-GACAAACTCCCACTCAGGAATCCA-3′ as previously described (48). The PCR consisted of an initial denaturation at 94°C for 4 min, followed by 38 cycles of amplification (94°C for 45 s, 55°C for 45 s, 72°C for 60 s) and a final elongation for 10 min at 72°C. The final PCR products were visualized under UV on a 2% ethidium bromide-stained agarose gel.
Stained cells were analyzed using a FACSCalibur flow cytometer (Becton Dickinson Immunocytometer Systems, Palo Alto, Calif.) and the Cellquest software. Ten thousand events were collected in list mode files for each test. Fluorescence parameters were collected using a 4-decade logarithmic amplification. Dead cells were excluded by forward and side scatter gating. The area of positivity was determined using isotype-matched MAb.
For immunofluorescence staining and laser confocal examination, cells cultured on sterilized 22- by 22-mm glass coverslips (Cole Parmer, Vernon Hills, Ill.) were stained with appropriate MAb in the presence of normal AB serum at 4°C. Cells were washed twice with PBS containing azide (0.01%) and BSA (0.2%), fixed in PBS containing 3% paraformaldehyde, and washed with PBS. The slides were mounted with 1 drop of mounting solution (Mowiol) on poly-l-lysine-coated slides before examination with an LEICA TCS SP laser confocal imaging system equipped with an Arkr laser (Leica Microsystems, Heidelberg, Germany).
Statistical analysis was performed using Student’s paired t test. Quantitative results are expressed as means ± standard errors of the means (SEM).
The transmigration of HIV-infected MDM through epithelial cells was investigated using MDM that had been infected with HIV-1Ba-L in vitro and a tight monolayer of HEC-1 cells. Transmigration was consistently observed under our experimental conditions, with 0.14 to 1.2% of infected cells being recovered in the lower chamber of the system (mean ± SEM: 0.79% ± 0.33% in four separate experiments) after 18 h of transmigration at 37°C (Fig. (Fig.1).1). Under similar conditions, the transmigration of uninfected MDM was 0.16% ± 0.03% in three separate experiments (data not shown). At 4°C, transmigration of HIV-infected MDM was significantly reduced (to 0.14% ± 0.06%) compared with that observed at 37°C (P < 0.01). Lowering the temperature did not affect the amount of cells recovered when uninfected MDM were used (not shown). The monolayer resistivity measured after the transmigration of infected MDM was similar to that measured prior to transmigration, indicating that the HEC-1 cell monolayer had remained confluent without alteration of permeability during the transmigration process (not shown). Transmigrating cells were infected with HIV-1, as shown by PCR amplification of viral DNA in the cells recovered in the lower chamber (see Fig. Fig.88).
To mimic a proinflammatory microenvironment, N. meningitidis LPS, IL-1β, and TNF-α were added to the culture medium in the transwell system and maintained throughout the transmigration process. Under experimental conditions similar to standard conditions, a threefold increase in transmigration of HIV-infected MDM (2.70% ± 0.80%) was observed (P < 0.0001) (Fig. (Fig.1).1). The presence of LPS, IL-1β, and TNF-α did not, however, affect the amount of cells recovered in the lower chamber when the experiment was performed at 4°C. The kinetics of transmigration of infected MDM at 37°C are shown in Fig. Fig.2.2. A plateau in transmigration was observed after 3 h of incubation of HIV-infected MDM with the epithelial monolayer.
The infection of MDM with HIV-1 in vitro resulted in a moderate increase in the expression of CD11a on the membrane of the cells under conditions where approximately two-thirds of the cells harbored the virus, as assessed by intracellular staining for the p24 antigen (Fig. (Fig.3).3). Preincubation of HIV-infected MDM with anti-CD11a MAb for 30 min at 37°C prior to contact with the epithelial monolayer resulted in a dramatic decrease in transmigration of the cells (0.36% ± 0.32%) compared with the 2.70% ± 1.02% transmigration observed in the absence of anti-CD11a MAb (P < 0.001) (Fig. (Fig.4).4). The amount of transmigration observed in the presence of irrelevant anti-human Ig MAb was similar to that observed in the absence of anti-CD11a (2.70% ± 0.95% versus 2.70% ± 1.02%). Inhibition of transmigration of HIV-infected MDM by anti-CD11a MAb was dose dependent (Fig. (Fig.55).
We further investigated the role of the CD54, CD50, and CD102 ligands of CD11a on epithelial cells in the transmigration of HIV-infected MDM through the HEC-1 monolayer. The presence of LPS, IL-1β, and TNF-α in the culture medium in the transwell system, under the conditions described in the legend for Fig. Fig.1,1, resulted in an enhanced expression of the three molecules on the membrane of HEC-1 cells, as assessed by immunofluorescence staining (Fig. (Fig.6).6). The CD54 molecule appeared constitutively expressed on unstimulated HEC-1 cells, in contrast to CD102 and CD50, whereas CD50 and CD102 were induced upon stimulation with LPS and cytokines.
When the epithelial monolayer was preincubated with anti-CD50 MAb or anti-CD102 MAb for 30 min at room temperature prior to the addition of HIV-infected MDM in the presence of proinflammatory stimuli, transmigration was inhibited by 60% (58.7% ± 21.0%) and 48% (47.6% ± 23.3%), respectively. Preincubation of HEC-1 cells with both anti-CD50 and anti-CD102 MAb inhibited transmigration by 73.5% ± 5.0% in an additive fashion (Fig. (Fig.7).7). Preincubation of HEC-1 with anti-CD54 MAb resulted in a much lesser inhibition of 19.5% ± 2.2% (Fig. (Fig.77).
Evidence that anti-CD102 or anti-CD50 MAb are capable of inhibiting the transmigration of HIV-infected MDM was obtained by PCR amplification of viral DNA in cells recovered in the basal chamber of the transmigration system. Transmigrated cells recovered in the basal chamber in the absence of preincubation of the HEC-1 monolayer with anti-CD50 and anti-CD102 MAb harbored viral DNA, whereas no DNA was detected in the few cells recovered in the lower chamber when transmigration was blocked with anti-CD50 and anti-CD102 MAb (Fig. (Fig.88).
We first assessed the intracellular production of chemokines by epithelial HEC-1 cells, including MIP-1α, MIP-1β, RANTES, and MCP-1, under normal and proinflammatory conditions. We did not observe any significant production of these chemokines by HEC-1 cells under our experimental conditions (data not shown). We further investigated by flow cytometry the intracellular production of IL-1α, IL-1β, IFN-γ, and TNF-α cytokines by HEC-1 cells. These proinflammatory mediators are known to be involved in the leukocyte diapedesis through an endothelial barrier that occurred within an inflammatory environment. The confluent HEC-1 cells were cultured in the presence of LPS, IL-1β, and TNF-α in the culture medium under the conditions described in the legend for Fig. Fig.1.1. We observed intracellular production of IL-1β (12 versus 0% positive cells) and TNF-α (5 versus 0% positive cells) by stimulated HEC-1 cells compared to cells cultured in the absence of proinflammatory mediators (Fig. (Fig.9).9). Treatment with brefeldin A did not affect cell viability (data not shown). Prestaining plasma cell membranes with unlabeled specific MAb directed either against anti-IL-1β or TNF-α did not affect intracellular staining (data not shown).
We evaluated the involvement of CXCR4, CCR5, and gp120 in the transmigration of HIV-1-infected MDM in blocking experiments. HIV-1-infected MDM express high levels of CCR5 antigen, in contrast to the weak expression of CXCR4 (data not shown). Cells were preincubated with antibodies to CXCR4, CCR5, or gp120 for 30 min at room temperature prior to the transmigration assay. Under proinflammatory conditions, MDM transmigration was not affected by antibodies to CXCR4, CCR5, or gp120 (Table (Table1).1). Transmigration under noninflammatory conditions was similar to that observed with noninfected cells (data not shown). Pretreatment of MDM and HEC-1 cells with MCP-1, MIP-1α, MIP-1β, or RANTES did not affect HIV-1 MDM transmigration under both proinflammatory and noninflammatory conditions (Table (Table11).
HEC-1 endometrial cells are not infected after transmigration of HIV-1-infected MDM. The epithelial cells used in the transmigration assays under proinflammatory conditions were extensively washed and recovered from the filter by trypsinization and cultured for 30 days to eliminate contaminating cells. HIV-1 provirus was detected in 106 epithelial cells by PCR of the pol gene region. No proviral DNA was detected in HEC-1 cells that had been exposed to HIV-1-infected MDM (data not shown).
In the present study, transmigration of cell-associated R5-tropic HIV-1 through a tight monolayer of epithelial cells was dependent on the initial interaction between the αLβ2 integrin CD11a/CD18 on infected monocytic cells and the ICAM-2 (CD102) and ICAM-3 (CD50) adhesion molecules on the apical membrane of epithelial cells. To investigate the mechanisms involved in transmucosal passage of HIV-positive cells, we used an in vitro transwell system where human macrophages infected with R5-tropic HIV were allowed to transmigrate through a tight monolayer of endometrial epithelial cells (7, 22). We used the HIV-1Ba-L strain of HIV since HIV-1 CCR5-tropic variants are those which expand in vivo during primary infection (5, 40, 43). The phenotype of genital HIV strains is primarily macrophage-tropic and non-syncytium-inducing (13). We elected to use MDM as target cells for HIV infection and as migrating cells, since genital tract tissues from HIV-infected individuals contain high levels of HIV-infected monocytes (12, 25, 46). Spira et al. demonstrated that the first target cells for infection, observed within 2 days of intravaginal inoculation of simian immunodeficiency virus in macaques, are mononuclear cells in the lamina propria of the cervicovaginal mucosa (41). Another report using female genital mucosal explants indicated that the majority of HIV-infected cells found within the cervical subepithelial mucosa were monomacrophages (19). We used endometrial cell line HEC-1, which originates from the human female genital tract, as a source of epithelial cells. The integrity of the epithelial cell monolayer was carefully monitored throughout the experiments by measuring the resistivity between the apical and basal chambers of the transwell system. We further mimicked the proinflammatory environment that may favor sexually transmitted infection in vivo by adding proinflammatory cytokines TNF-α and IL-1β to the apical chamber of the transwell system together with small amounts of LPS, which promote HIV replication and spread (1, 16, 38). Proinflammatory cytokines are physiologically produced in the genital tract (20). Increased genital levels of IL-1, TNF-α, and IL-6 in HIV-infected women have been reported (4). An overproduction of proinflammatory cytokines in the genital tract is associated with increased risk of HIV transmission, including sexually transmitted infections (27), cervicitis (29), and bacterial vaginosis (36).
Under such experimental conditions and at 37°C, we consistently observed the transmigration of approximately 3% of HIV-infected MDM through the confluent HEC-1 monolayer, starting within 90 min after the initial contact of HIV-positive monocytes with the epithelial cells and peaking after 180 min. These observations are consistent with those reported by Tan and Phillips, who demonstrated the transmigration of HIV-1-infected monocytes through a tight monolayer of human cervical epithelial cell line ME180 (42). In their study, transmigration appeared as an active process, since HIV-infected monocytes extended pseudopods between ME180 cells, from which HIV virions were seen budding (42). We observed that transmigration of HIV-infected MDM through the monolayer of HEC-1 cells was blocked at 4°C. Although low temperature may alter cell membrane fluidity, making it harder for the virus to move between epithelial cells, this finding strongly suggests that transmigration represents an active, temperature-dependent mechanism analogous to cellular diapedesis. In the absence of proinflammatory conditions, migration of infected MDM at 37°C (below 1.0%) did not differ from that of uninfected cells or from that of infected cells at 4°C. Transmigration of mononuclear cells through a tissue barrier has previously been reported. The transmigration of HIV-infected monocytes through the blood-brain barrier is considered to determine the infiltration of brain tissue by HIV in AIDS dementia (33). Human mononuclear blood cells placed into the vaginas of mice have been shown to migrate within 4 h in the connective tissue beneath the vaginal epithelium and the iliac lymph nodes (47). In addition, HIV-infected human blood-derived lymphocytes may cross in vivo the vaginal epithelia of humanized SCID mice, allowing for the migration of lymphocytes infected with R5-tropic strain of HIV to the regional lymph nodes within 24 h of inoculation, followed by a systemic infection with detectable viremia (S. Di Fabio et al., personal communication). Our results suggest that significant transmigration of genital HIV-infected mononuclear cells could occur through the target mucosa of an exposed receptive individual in vivo in the presence of a local inflammatory microenvironment.
To investigate the molecules involved in transmigration through the endometrial cell monolayer, we analyzed the expression of adherence molecules and their ligands on MDM and HEC-1 cells, respectively. We focused on the CD11a/CD18 molecule, whose expression is upregulated on peripheral blood leukocytes upon HIV disease progression (18, 31). CD11a was recently shown to be important for the initial contact of HIV with a target cell and the subsequent viral replication in an in vitro infectivity model using Jurkat-derived Jbeta2.7 T cells (21). The CD11a molecule has further been implicated in the transendothelial migration of lymphocytes in the presence of TNF-α. Antibodies to CD11a were shown to block the adhesive interaction of macrophages with endothelial cells (18). Furthermore, distinct epitopes of ICAM-3 (CD50) have been demonstrated to mediate HIV-1-specific entry into lymphoid and monocytoid cells (39). In the present study, we observed that the CD11a molecule on MDM primarily interacts with its CD50 and CD102 ligands, which are overexpressed on epithelial cells activated within the context of an inflammatory microenvironment. Blocking experiments using specific MAb to CD50 or CD102 demonstrated the involvement of both ligands in the transmigration process. We were not able to demonstrate involvement of chemokines IL-8, MIP-1α, MIP-1β, RANTES, and MCP-1 in the transmigration process of HIV-1-infected MDM. Furthermore, blocking experiments using antibodies to CXCR4, CCR5, or gp120 demonstrated that neither CXCR4, CCR5, nor gp120 appeared to be involved in HIV-1-infected MDM transmigration, suggesting that only interactions of CD50 or CD102 with CD11a are important. Overexpression by stimulated HEC-1 cells of intracellular IL-1β and TNF-α likely emphasizes the role of adhesion molecules in HIV-1-infected MDM transmigration. Determination of whether interactions between CD11a and CD50 or CD102 lead to activation of cellular signals that could subsequently promote cellular migration of activated MDM through the epithelial barrier will require further investigation. HIV-infected cells within tissues may become a “fertile field” for spreading viral particles. Resident infected macrophages could also provide a persistent source of proinflammatory mediators, virus-encoded proteins, and infectious virions. Interactions of CD11a with CD50, CD54, and CD102 molecules play an important part during syncytium formation, participating directly in HIV production (3, 11). Sommerfelt et al. have demonstrated that only antibodies directed against CD50 significantly inhibited HIV-1-induced syncytium formation, entry, and infectivity of lymphoid and monocytoid cells (39). Moreover, Hioe et al. have recently shown that the CD11a molecule expressed on target cells promotes HIV-1 infection and transmission (21). Of particular interest is the fact that inhibition of transmigration with antibodies to the adhesion proteins was preferentially targeted to the transmigration of HIV-infected MDM over that of uninfected cells, as assessed by measuring proviral DNA in cells recovered in the basal chamber of the transwell system in the absence and in the presence of inhibiting antibodies in the experiment. These results are consistent with previous observations demonstrating that infected cells are preferentially found among the transendothelial migratory leukocytes (6). Demonstrating that the transmigration of HIV-infected monocytes/macrophages through endometrial cells is effectively blocked by antibodies to adhesion molecules allows us to hypothesize that such antibodies would also limit the spreading of sexually transmitted HIV-1 in vivo. In the same way, the transmigration model as we described it here may be used to evaluate the functionality of mucosal antibodies directed against HIV, which may block the transmigration process, as observed in a preliminary experiments described by Tan and Phillips (42).
In our model using a monolayer of endometrial HEC-1 cells, the transmigration of HIV-1-infected MDM was not associated with the infection of HEC-1 endometrial cells, suggesting that transmigration may bring HIV-infected mononuclear cells to the submucosa, and was not associated with infection of the epithelial cell line. The relative role of HIV-infected cell transmigration through the genital epithelial cell monolayer in heterosexual virus transmission has not as yet been fully understood. Our observations provide new insights supporting the hypothesis that HIV-infected mononuclear cells contained in genital secretions from infected individuals may cross the epithelial genital mucosae of an exposed receptive sexual partner. Greenhead et al., using an explant of human female genital tissue, did not observe evidence of migration of peripheral blood mononuclear cells infected with T-cell adapted M-tropic or T-tropic HIV-1 strains into either ectocervical, endocervical, or vaginal tissues (19). However, HIV infection is known to preferentially occur when the genital epithelial integrity is altered, particularly when epithelial microulcerations occur during heterosexual intercourse (32), when ulcerations due to sexually transmitted infections appear (15), in the context of ectopy of the endocervical mucosa (30), which may leave the genital tissue more friable, and, in general, under inflammatory conditions in the genital tract. One may speculate that transmigration of HIV-infected cells could be facilitated, for example, through a cervicovaginal mucosa partially altered by sexually transmitted infections, as is frequently the case in the context of HIV transmission in developing countries.
This study was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), the Université Pierre et Marie-Curie (Paris VI), and the French National Agency for AIDS Research (ANRS; grant 01128).
We acknowledge Antonio Cos for providing us the HIV-1Ba-L strain. We also thank Michel Paing and Pierre Kitmatcher for their contribution in the artwork.