gp340 Promotes Transcytosis of Human Immunodeficiency Virus Type 1 in Genital Tract-Derived Cell Lines and Primary Endocervical Tissue

Abstract

The human scavenger receptor gp340 has been identified as a binding protein for the human immunodeficiency virus type 1 (HIV-1) envelope that is expressed on the cell surface of female genital tract epithelial cells. This interaction allows such epithelial cells to efficiently transmit infective virus to susceptible targets and maintain viral infectivity for several days. Within the context of vaginal transmission, HIV must first traverse a normally protective mucosa containing a cell barrier to reach the underlying T cells and dendritic cells, which propagate and spread the infection. The mechanism by which HIV-1 can bypass an otherwise healthy cellular barrier remains an important area of study. Here, we demonstrate that genital tract-derived cell lines and primary human endocervical tissue can support direct transcytosis of cell-free virus from the apical to basolateral surfaces. Further, this transport of virus can be blocked through the addition of antibodies or peptides that directly block the interaction of gp340 with the HIV-1 envelope, if added prior to viral pulsing on the apical side of the cell or tissue barrier. Our data support a role for the previously described heparan sulfate moieties in mediating this transcytosis but add gp340 as an important facilitator of HIV-1 transcytosis across genital tract tissue. This study demonstrates that HIV-1 actively traverses the protective barriers of the human genital tract and presents a second mechanism whereby gp340 can promote heterosexual transmission.

Through correlative studies with macaques challenged with simian immunodeficiency virus (SIV), the initial targets of infection in nontraumatic vaginal exposure to human immunodeficiency virus type 1 (HIV-1) have been identified as subepithelial T cells and dendritic cells (DCs) (18, 23, 31, 36-38). While human transmission may differ from macaque transmission, the existing models of human transmission remain controversial. For the virus to successfully reach its CD4⁺ targets, HIV must first traverse the columnar mucosal epithelial cell barrier of the endocervix or uterus or the stratified squamous barrier of the vagina or ectocervix, whose normal functions include protection of underlying tissue from pathogens. This portion of the human innate immune defense system represents a significant impediment to transmission. Studies have placed the natural transmission rate of HIV per sexual act between 0.005 and 0.3% (17, 45). Breaks in the epithelial barrier caused by secondary infection with other sexual transmitted diseases or the normal physical trauma often associated with vaginal intercourse represent one potential means for viral exposure to submucosal cells and have been shown to significantly increase transmission (reviewed in reference 11). However, studies of nontraumatic exposure to SIV in macaques demonstrate that these disruptions are not necessary for successful transmission to healthy females. This disparity indicates that multiple mechanisms by which HIV-1 can pass through mucosal epithelium might exist in vivo. Identifying these mechanisms represents an important obstacle to understanding and ultimately preventing HIV transmission.

Several host cellular receptors, including DC-specific intercellular adhesion molecule-grabbing integrin, galactosyl ceramide, mannose receptor, langerin, heparan sulfate proteoglycans (HSPGs), and chondroitin sulfate proteoglycans, have been identified that facilitate disease progression through binding of HIV virions without being required for fusion and infection (2, 3, 12, 14, 16, 25, 29, 30, 43, 46, 50). These host accessory proteins act predominately through glycosylation-based interactions between HIV envelope (Env) and the host cellular receptors. These different host accessory factors can lead to increased infectivity in cis and trans or can serve to concentrate and expose virus at sites relevant to furthering its spread within the body. The direct transcytosis of cell-free virus through primary genital epithelial cells and the human endometrial carcinoma cell line HEC1A has been described (7, 9); this is, in part, mediated by HSPGs (7). Within the HSPG family, the syndecans have been previously shown to facilitate trans infection of HIV in vitro through binding of a specific region of Env that is moderately conserved (7, 8). This report also demonstrates that while HSPGs mediate a portion of the viral transcytosis that occurs in these two cell types, a significant portion of the observed transport occurs through an HSPG-independent mechanism. Other host cell factors likely provide alternatives to HSPGs for HIV-1 to use in subverting the mucosal epithelial barrier.

gp340 is a member of the scavenger receptor cysteine-rich (SRCR) family of innate immune receptors. Its numerous splice variants can be found as a secreted component of human saliva (34, 41, 42) and as a membrane-associated receptor in a large number of epithelial cell lineages (22, 32, 40). Its normal cellular function includes immune surveillance of bacteria (4-6, 44), interaction with influenza A virus (19, 20, 32, 51) and surfactant proteins in the lung (20, 22, 33), and facilitating epithelial cell regeneration at sites of cellular inflammation and damage (27, 32). The secreted form of gp340, salivary agglutinin (SAG), was identified as a component of saliva that inhibits HIV-1 transmission in the oral pharynx through a specific interaction with the viral envelope protein that serves to agglutinate the virus and target it for degradation (34, 35, 41). Interestingly, SAG was demonstrated to form a direct protein-protein interaction with HIV Env (53, 54). Later, a cell surface-associated variant of SAG called gp340 was characterized as a binding partner for HIV-1 in the female genital tract that could facilitate virus transmission to susceptible targets of infection (47) and as a macrophage-expressed enhancer of infection (10).

MATERIALS AND METHODS

Virus, cell lines, antibodies, and inhibitors.

HIV-1 strains BL-2, Ba-L, YU2, JR-FL, 89.6, and IIIB were obtained from the Center for AIDS Research at the University of Pennsylvania. Peripheral blood mononuclear cells (PBMCs) were obtained from the Center for AIDS Research Immunology Core at the University of Pennsylvania under an Institutional Review Board-approved protocol. HEC1A cells were obtained from the American Type Culture Collection (Manassas, VA) and maintained in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum albumin and 2 mM l-glutamine (Invitrogen, Carlsbad, CA) (21, 46). gp340-specific rabbit polyclonal antibody 1527A and the murine polyclonal antibody DAPA were used (10, 47). HIV-1 V3 loop peptide 6284 (CTRPNYNKRKRIHIG), scrambled peptide 6284 (RCIHNRTIKGPYNKR), and control peptide 6220 (KEATTTLFCASDAKA) were used. Virus-like particles (VLPs) were made by transient transfection of 293T cells with Lipofectin (28) by using plasmid pGag, which expresses p55 from codon-optimized NL4-3 sequences (24). Two days later, supernatants were collected and filtered through a 0.45-μm filter. VLPs were then pelleted at 50,000 × g in a Sorvall SS34 rotor. VLPs were then resuspended in phosphate-buffered saline (PBS) at 1,000 times the original concentration and p24 Gag protein content was measured by enzyme-linked immunosorbent assay (ELISA; Beckman-Coulter, Fullerton, CA).

Synthesis of scaffolded peptides.

Scaffolded peptide 6284 and scaffolded scrambled peptide 6284 were synthesized. Poly(maleic anhydride-1-octadecene) (PMAO) was used as the scaffold for scaffolded peptide synthesis. PMAO has an average molecular weight of 40,000, corresponding to an average of 113 maleic anhydride units per polymer chain. The conjugation reaction between PMAO and 113 equivalents of side chain-protected peptides was carried out overnight at 50°C in dimethylformamide with agitation on a rotary shaker catalyzed by diisopropylethylamine. Unreacted maleic anhydride groups were quenched by addition of 150 μl of distilled and deionized water, and then the products were treated with excess trimethylsilyldiazomethane to methylate the resulting carboxylic acid groups. The polymer products were further dialyzed overnight in distilled water. Cleavage of side chain protecting groups from the polymer was performed after drying with trifluoroacetic acid, and the final products were obtained after cold ether precipitation and overnight lyophilization.

Cellular staining.

Cells were grown in six-well plates and removed with PBS without Ca²⁺ or Mg²⁺ with 5 mM EDTA. Optimal staining for gp340 was observed when 4 mM Ca²⁺ was used throughout. Polyclonal antibody 1527A or control rabbit immunoglobulin G (IgG; Sigma, St. Louis, MO) in PBS-2% FCS-4 mM Ca²⁺ was used. Cells were also stained with anti-HSPG antibody 10E4 (Seikagaku Kogyo Co., Tokyo, Japan) or isotype control and goat anti-mouse Ig-phycoerythrin (Sigma, St. Louis, MO). Cells were analyzed on a FACScan flow cytometer with Cell Quest Pro software (Becton-Dickinson, Franklin Lakes, NJ).

Viral binding assay.

HEC1A cells were seeded into 96-well plates and allowed to grow to 80 to 95% confluence. Inhibitors were added to the cells for incubation for 1 h at 37°C. Unless otherwise stated, we used peptides at a concentration of 10 μg/ml, antibodies at 20 μg/ml, heparinase III (Sigma) at 5 U/ml, and chondroitinase ABC (Sigma) at 10 U/ml. The peptide, enzyme, and antibody concentrations used were optimized as previously described (47). After pretreatment, 5 ng of HIV was added to each well and incubated for 30 min at 37°C. Cells were washed four times after viral incubation. Between washes, plates were centrifuged at 1,250 rpm (225 × g) for 7 min. Cells were lysed in 50 μl of cell lysis buffer (Promega), and samples were analyzed for HIV Gag p24 protein by ELISA.

Quantification of HIV Gag by reverse transcription (RT)-PCR.

In the basal chamber of a transwell apparatus, medium samples containing HIV virions were treated with RQ1 RNase-free DNase I (Promega, Madison, WI) for 1 h, heated to 65°C for 10 min to neutralize the DNase, and then added directly (1/20 of the Master Mix volume) to the RT-PCR Master Mix. By directly measuring HIV RNA in medium samples, we avoided error introduced by purification steps that obtain purified RNA from aqueous samples. When added at 1/20 of the volume of the PCR, no inhibition of the RT or PCR steps was observed. RT-PCR was performing with the Titan One Tube RT-PCR system (Roche Applied Science, Indianapolis, IN) with primers Gag-6F (5′-CATGTTTTCAGCATTATCAGAAGGA-3′) and Gag-84R (5′-TGCTTGATGTCCCCCCACT-3′) and probe BHQ1 (5′-FAM-CCACCCCACAAGATTTAAACACCATGCTAA-TAMRA-3′ [FAM, 6-carboxyfluorescein; TAMRA, 6-carboxytetramethylrhodamine]). RT (with avian myeloblastosis virus reverse transcriptase [Invitrogen]) was carried out at 50°C for 30 min and followed by PCR amplification (95°C for 2 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min). Standard curves were developed with ACH2 DNA diluted from 10⁶ to 10° copies. RT-PCR was performed with the ABI 7500 Fast Real-Time PCR System and accompanying ABI software (Applied Biosystems, Foster City, CA).

trans-infection assay.

Phytohemagglutinin (PHA)-induced blast cells (PHA blasts) were prepared by stimulating PBMCs at 10⁶/ml with PHA-P (Sigma) at 4 μg/ml in RPMI 1640 medium supplemented with 10% fetal bovine serum and 2 mM l-glutamine (RPMI medium) for 3 days. Cells were washed and resuspended at 2 × 10⁵/ml in RPMI medium with interleukin-2 (IL-2; 20 U/ml; obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH [human recombinant IL-2 from Maurice Gately, Hoffmann-La Roche Inc.]). HEC1A cells were seeded into 96-well plates to 80 to 95% confluence and incubated with 5 ng of HIV-1 in 100 μl of medium for 30 min at 37°C and then extensively washed. Inhibitors, peptides or antibodies, were added to cells 30 min prior to pulsing with virus and replaced after the cells were washed. Inhibitor concentrations were the same as those used in the viral binding assays described above. Plates were centrifuged at 1,250 rpm (225 × g) for 7 min after each wash. After viral incubation and washing, PHA blasts were cocultured with the pulsed cells for 1 day and then moved to new wells for 6 days at 37°C with IL-2. Infection was monitored by measuring supernatant-associated p24 Gag protein content on day 7 by ELISA.

siRNA treatment.

A small interfering RNA (siRNA) duplex directed against nucleotides 6604 to 6624 of DMBT1 (5′ AAUAUCACCAGUUCGGCG 3′) (26) or a control siRNA (GL3 [5′ CUUACGCUGAGUACUUCGA 3′]) (Applied Biosystems/Ambion, Austin, TX) was transfected into HEC1A cells on day 4 of culture in transwells with TransIT-TKO (Mirus Bio, Madison, WI). Cells were used for transcytosis after knockdown was documented by flow cytometry and resistance criteria were met.

Transcytosis assay.

HEC1A cells were seeded into 12-well transwell plates containing a 0.4-μm-pore-size polycarbonate permeable support (Costar, Cambridge, MA). After ∼7 days of incubation, tight junction formation was monitored daily by measuring resistance across the transwell until the resistivity exceeded 300 Ω/cm² (Millicell ohmmeter; Millipore, Billerica, MA). After reaching sufficient resistivity, the HEC1A cells were incubated with inhibitors and/or heparinase III for 1 h at 37°C at the concentrations described for each inhibitor. After pretreatments, cells were either used or chilled to 4°C. The HEC1A cells were then pulsed with 5 ng of HIV per well at 37 or 4°C. Basal chamber samples were taken at various times subsequent to viral pulsing. Resistivity was confirmed after viral pulsing and supernatant removal to ensure maintenance of tight junctions during experimentation. Samples were measured for HIV content by p24 Gag ELISA or RT-PCR.

Tissue transcytosis assay.

Fresh endocervical tissue was obtained from the Cooperative Human Tissue Network (Eastern Division, Hospital of the University of Pennsylvania, Philadelphia). Upon receipt, tissue was sectioned into ∼0.5-cm-thick epithelial cross sections and seeded into 12-well plates containing RPMI medium-10% FCS. After 48 to 72 h, the sloughed epithelial cells were removed and the tissue was gently washed in PBS before being transferred to fresh 12-well transwell plates. After tissue attachment was allowed for 6 h at 37°C, the apical chamber medium was removed. The tissue was imbedded in a 50:50 RPMI medium plus agarose matrix and allowed to set up for 30 min at room temperature. The edges of the tissue were then sealed to the side walls of the culture dish with cyanoacrylate, and the tissue was dried for 10 min and then covered with medium. Resistivity measurements were taken before and after assaying for transcytosis in each tissue section. Transcytosis assay conditions were the same as those described for HEC1A cells.

Statistics.

Means and standard errors of the means were calculated for three or four replicates. The two-tailed Student t test with equal variance was used.

RESULTS

HEC1A cells express gp340, which facilitates binding and transmission of HIV-1.

We previously demonstrated that gp340 was expressed on a number of cell lines derived from the female human genital tract (47). Despite not supporting active infection, those cell lines were capable of binding, harboring, and transmitting infectious virus to target cells of infection. Another genital tract cell line that has been previously shown to facilitate transmission of HIV-1 is the endometrial carcinoma line HEC1A. Through cell surface staining, we demonstrated that HEC1A cells express gp340 on their extracellular membrane (47). We first sought to analyze the properties of HEC1A cell-expressed gp340 in terms of previously described activities, binding of HIV via Env and transmission to target cells (47, 52). HEC1A cells were incubated with antibodies or peptides that block Env binding to gp340 and/or heparinase III. Heparinase III has been shown to remove HSPG moieties (including syndecans) present in the extracellular domains of a number of host proteins. Removal of HSPG moieties (>85%) by treatment with heparinase III was confirmed by cell surface staining with anti-HSPG monoclonal antibody 10E4. Following pretreatment, the HEC1A cells were incubated with HIV (primary M-tropic strain BL2) for 30 min. After pulsing, the cells were extensively washed and lysed and samples were collected and measured for p24 Gag protein content. Treatment with peptide 6284, anti-gp340 antibodies DAPA and 1527A, and heparinase III, but not control peptide or antibody, showed a significant ability to impair HIV-1 binding to HEC1A cells. Interestingly, combinational treatment of cells with heparinase III and peptide 6284 showed an additive ability to inhibit HIV binding (Fig. 1A and data not shown). HEC1A cells were tested for the ability to transmit infectious virus to PHA- and IL-2-stimulated PBMCs. Pretreatment of cells with or without peptide, antibody, and/or heparinase III, followed by pulsing with virus and extensive washing, was performed. In contrast to measurements of binding, a significantly lower viral input was used and HEC1A cells were not lysed but rather cocultured with stimulated PBMCs. Seven days after coculture, the p24 Gag protein content of the medium was measured. Treatment with peptide 6284, DAPA, and 1527A and heparinase III, but not control peptide or antibody, reduced the effectiveness of HEC1A cells in successfully transmitting virus to healthy targets of infection (Fig. 1B and data not shown). Similar to measurements of binding, cotreatment with peptide 6284 and heparinase III further inhibited trans infection over a single pretreatment. Similar results were seen with other HIV strains with various tropisms (Ba-L, YU2, JR-FL, 89.6, and IIIB). These data demonstrate that gp340 on HEC1A cells derived from an endometrial carcinoma acts similarly to primary genital tract epithelial cell-expressed gp340 (47).

FIG. 1.

HEC1A cells express gp340, which mediates binding and trans infection of stimulated PBMCs. (A) HEC1A cells were pretreated with medium alone, peptide 6220, peptide 6284, peptide 6284 plus heparinase III (Hep), peptide 6220 plus heparinase III, or heparinase III alone. Cells were then pulsed with BL2 virus (5 ng/well). After washing, the cells were lysed and p24 Gag protein content was measured by ELISA. Experiments were performed in three-well replicates. Error bars show the standard error of the mean. An asterisk indicates a P value of 0.001 compared to the appropriate control, peptide 6220 for 6284 and medium treated for heparinase III, and the # symbol shows a P value of 0.004 compared to heparinase III plus peptide 6220. (B) HEC1A cells were pretreated with medium alone, peptide 6284, scrambled peptide 6284 (6284sc), peptide 6284 and heparinase III, scrambled peptide 6284 and heparinase III, or heparinase III alone and then pulsed with 0.33 IU of HIV strain BL2. After viral exposure, the cells were washed four times to remove unbound virus and cocultured with T-cell blasts in the absence of peptides for 24 h. The T-cell blasts were then moved to new wells, and p24 Gag protein content was measured 7 days later. Infectious units were calculated with PHA blasts as target cells in a 50% tissue culture infective dose assay. The experiment was performed in replicates of six. Data are expressed as the percentage of the six wells in which the cells became infected. Error bars represent the standard error of the mean. An asterisk indicates a P value of <0.005 compared to scrambled peptide 6284 or medium, and the # symbol indicates a P value of 0.016 compared to heparinase III and scrambled peptide. The two-tailed Student t test with equal variance was used.

gp340 mediates transcytosis of HIV-1.

Prior studies demonstrated the ability of HEC1A cells to transcytose HIV, which was, in part, mediated by HSPGs (7, 46). gp340 was previously demonstrated to move bound ligand into intracellular compartments (32, 33). We sought to understand the contribution that gp340 might make to the transport of virus through HEC1A cells. The cells were seeded into transwell plates and allowed to grow to a confluence that supported tight junction formation as determined by resistivity measurements across the transwell that exceeded 300 Ω/cm². HEC1A cells were then incubated with or without peptide 6284, control peptide 6220, mouse polyclonal antibody DAPA, rabbit polyclonal antibody 1527A, or control Igs. Following viral pulsing of the upper chamber of the transwell for 3 h, samples were taken from the lower chamber. Throughput virus was measured by RT-PCR for HIV Gag RNA. Pretreatment with peptide 6284 resulted in a 39% reduction in transcytosed virus compared to that in untreated wells or wells pretreated with control peptide 6220. Pretreatment with DAPA led to a 49% reduction, while 1527A reduced transcytosis by 43%. All three inhibitors of HIV Env binding to gp340 significantly inhibited HEC1A transcytosis of HIV-1 (Fig. 2A). Control rabbit or mouse IgG (data not shown) or a control peptide had no impact on viral transcytosis. HIV VLPs were analyzed under the same conditions, and while the amount of VLPs that were transcytosed was low, no inhibition by a specific peptide or antibody was observed (Fig. 2B).

FIG. 2.

HEC1A cells transcytose HIV in a gp340-dependent manner. HEC1A cells were grown in transwells and allowed to form tight junctions as determined by resistance measurement. Cells were pretreated with peptide 6284, peptide 6220, DAPA and 1527A antibodies, or control rabbit IgG. Cells were then pulsed with (A) HIV BL2 (5 ng/well) or (B) VLPs (5 ng/well) for 3 h. Each condition was tested in replicates of three. Lower-chamber medium was taken and measured for (A) HIV Gag RNA content by RT-PCR or (B) p24 Gag protein by ELISA following 3 h of viral pulsing. Error bars show the standard error of the mean. An asterisk indicates a P value of <0.001 compared to the appropriate control, peptide 6220 for 6284 and control rabbit IgG or mouse IgG (not shown) for 1527A and DAPA, respectively.

In experiments with siRNA directed against gp340, we were able to knock down expression by 75% as measured by flow cytometry and reduce transcytosis by 20 to 25%, which was not statistically significant (P > 0.05). We suspect that a more complete inhibition of Env binding to gp340, as observed with peptide or antibody inhibitors, is needed.

HEC1A transcytosis of HIV-1 can be inhibited additively with gp340-Env-inhibiting peptide and heparinase III treatment.

HEC1A transcytosis of HIV has already been reported to be partially dependent on the presence of HSPG moieties on the apical membrane of the cells (7). We sought to determine whether the blocking of gp340 and removal of HSPG moieties with heparinase III in combination enhanced the inhibition of viral transcytosis. HEC1A cells were grown to confluence, and a resistivity of >300 Ω/cm² was confirmed prior to and following transcytosis. Cells were pretreated with peptide 6284, scrambled peptide 6284, and/or heparinase III for 1 h. Following pretreatment, the apical side (upper transwell) was pulsed with HIV-1 BL2 for 3 h. Samples were taken from the lower chamber, and p24 Gag protein content was measured by ELISA. Peptide 6284 alone showed inhibition (48%). A scaffolded peptide 6284 construct designed to enhance stability and activity and given at an approximately threefold lower concentration showed a similar ability to prevent transcytosis (56%). Scrambled peptide 6284 alone produced no significant reduction in viral transcytosis. Heparinase III treatment alone produced a 46% reduction in viral transcytosis. When heparinase III and peptide 6284 treatments were combined, HEC1A viral transcytosis was reduced by 73% (Fig. 3). This suggests that gp340 and HSPG independently mediate transcytosis and inhibition of Env binding to both inhibits the majority of transcytosis by HEC1A cells.

FIG. 3.

HEC1A transyctosis of HIV is mediated by gp340 and HSPG but not chondroitin sulfate. HEC1A cells were grown to confluence in transwells, and resistance was measured to confirm tight junction formation. Cells were treated with heparinase III (Hep), chondroitinase ABC (Chond ABC), peptide 6284, scaffolded (scaf) peptide 6284, scrambled peptide 6284 (6284sc), scaffolded scrambled peptide 6284, scaffolded peptide 6284 and heparinase III, or scaffolded scrambled peptide 6284 and heparinase III. After pretreatment, cells were pulsed with HIV BL2 (5 ng/well) for 3 h. Lower-chamber samples were measured for HIV RNA content by RT-PCR. Replicates of three wells were used for each condition. Error bars show the standard error of the mean. An asterisk indicates a P value of <0.001 compared to the appropriate control, medium, scrambled peptide 6284, scaffolded scrambled peptide 6284, or heparinase III plus scaffolded scrambled 6284.

HEC1A transcytosis of HIV-1 is temperature dependent.

Transcytosis is an active process that requires cellular metabolism. Previous studies have determined that transcytosis across the cytoplasm of cells can be stopped by reducing the cellular temperature (7). We wanted to confirm that the observed transcytosis mediated by gp340 and subsequent inhibition by peptide and antibody were dependent upon transport of virus within HEC1A cells. HEC1A cells were grown in a transwell, and tight junction formation was determined by measuring resistivity. HEC1A cells were then pretreated with peptide 6284, scrambled peptide 6284, or heparinase III for 1 h. After pretreatment, cells were maintained at 37°C or cooled to 4°C for 1 h prior to viral pulsing for 3 h at 37 or 4°C. Temperature had no effect on resistivity measurements of tight junction stability. Regardless of the pretreatment, transcytosis of virus in HEC1A cells was abolished at 4°C (Fig. 4).

FIG. 4.

Transfer of virus through HEC1A cells is temperature dependent. HEC1A cells were grown to confluence in transwells and measured to confirm tight junction formation by resistance. Cells were treated with heparinase III (Hep), peptide 6284, scrambled peptide 6284 (6284sc), scrambled peptide 6284 plus heparinase III, or peptide 6284 plus heparinase III. After pretreatment, cells were pulsed with HIV BL2 (5 ng/well) for 3 h at 4 or 37°C for each pretreatment. Lower-chamber samples were measured for HIV content by p24 Gag protein ELISA. Replicates of three wells were used. Error bars show the standard error of the mean. An asterisk indicates a P value of 0.017 compared to the appropriate controls, scrambled peptide 6284 and medium, and a # symbol indicates a P value of 0.039 compared to heparinase III plus scrambled peptide 6284.

gp340 mediates binding and transcytosis of HIV-1 in primary human cervical tissue.

HEC1A cells, derived from an endometrial carcinoma, form tight junctions and express cell surface gp340, which binds HIV and promotes trans infection, similar to genital tract epithelial cells. In order to understand the role of gp340 in the transmission of HIV-1 in a more physiologic system, we tested its contribution to viral interactions and transcytosis with human primary endocervical genital tract tissue. First, we wanted to observe whether tissue sections would bind virus with specificity for existing host factors shown to modulate binding and trans infection in cell lines. Endocervical tissue was obtained from subjects with no known cervical pathology, which was confirmed by pathological examination. Tissue was sectioned and placed in wells with the apical surfacing facing up for 48 h to allow sloughing of dead epithelial cells. After washing, the tissue was incubated with various inhibitors and then pulsed with HIV. Following further washes, tissue was lysed and the bound HIV-1 was measured for p24 Gag protein content. Both peptide 6284 and heparinase III, but not scrambled peptide 6284, decreased HIV binding to human endocervical tissue (Fig. 5A).

FIG. 5.

Human female genital tract tissue binds and transcytoses HIV in a gp340-dependent manner. (A) Endocervical tissue was sectioned, seeded into 24-well plates, and incubated for 48 h to allow adherence and sloughing of dead cells. Tissue was then pretreated with peptide 6284, scrambled peptide 6284 (6284sc), or heparinase III (Hep). Tissue was then pulsed with HIV BL2 (5 ng/well) for 30 min. Tissue was washed four times and lysed, and HIV p24 Gag protein content was measured by ELISA. Error bars show the standard error of the mean. An asterisk indicates a P value of 0.002 compared to scrambled peptide 6284, and a # symbol indicates a P value of 0.003 compared to medium. (B) Tissue sections were placed in transwells, sealed with agarose and cyanoacrylate, and demonstrated to have a resistance of >300 Ω/cm². Tissue was treated with heparinase III (Hep), peptide 6284, scrambled peptide 6284 (6284sc), peptide 6284 plus heparinase III, or scrambled peptide 6284 plus heparinase III. After pretreatment, cells were pulsed with HIV BL2 (5 ng/well) for 3 h. The HIV RNA contents of lower-chamber samples were measured by RT-PCR. Error bars show the standard error of the mean. An asterisk indicates a P value of <0.001 compared to the appropriate control, scrambled peptide 6284, medium, or heparinase III plus scrambled peptide 6284. For panels A and B, replicates of three were performed for each condition. Experiments with each treatment were repeated with at least four different primary tissue specimens with similar results.

To measure the role of gp340 in the transcytosis of HIV-1 as it might occur during human transmission, healthy endocervical tissue was adapted to the transwell system. The tissue was sectioned and embedded in a soft agar matrix for stability. The edges of the tissue were sealed to the sides of the transwell with cyanoacrylate to form a tight barrier, thus forcing any transported virus to travel through, rather than around, the tissue. Resistivity measurements of the embedded tissue were performed before and after transcytosis, which demonstrated the integrity of the sealed matrix barrier and tissue. In all cases, the resistivity measurements obtained matched or exceeded those obtained with the cellular counterparts in the transcytosis assays. Embedded tissue sections were incubated with or without peptide 6284, scrambled peptide 6284, and/or heparinase III. Pretreatment with specific peptide 6284 or heparinase III, but not scrambled peptide 6284 or medium, resulted in a significant reduction in the transcytosis of HIV BL2 across the tissue barrier (Fig. 5B). The combination of heparinase III and peptide 6284 additively inhibited transcytosis through endocervical tissue. Similar results were obtained with other strains of HIV. In preliminary experiments, we analyzed tissue either immediately after obtaining it or after 24 or 48 h in culture. Similar rates of transcytosis were observed at each time.

DISCUSSION

Macaques remain the principal model system for analyzing HIV transmission. During vaginal exposure to SIV in macaques, the first cells that are infected are Langerhans cells, submucosal CD4⁺ T cells, and DCs (18, 23, 31, 36, 38, 39). A cellular barrier segregates the vaginal, cervical, and uterine lumens from these targets of infection, so the mechanisms for HIV-1 traversing this barrier and the host components contributing to this process remain an important area of study in HIV transmission. In primary genital tract epithelial cells and the endometrial carcinoma line HEC1A, direct transcytosis of virus through the body of the cells has been shown to be partially dependent on HSPG moieties expressed on the apical membrane. More specifically, the syndecan family has been implicated in mediating some of the transcytosis observed in these cell lines (7). Here, we demonstrate that gp340 also directly contributes to the transcytosis observed in HEC1A cells and primary human endocervical tissue, which is distinct from that mediated by syndecans. Initially, the secreted DMBT1 splice variant SAG was identified as a component of human saliva important in inhibiting oral transmission of HIV-1 (33, 42). However, the cell-associated variant gp340 was shown to bind and facilitate trans infection of HIV-1 when expressed on luminal lining cells of the genital tract (47) and promote cis infection of macrophages (10). We also demonstrate that HEC1A cells express gp340 and that expression mediates the ability of these cells to bind, transmit, and transcytose HIV-1. Further, we show that human endocervical tissue can mediate transcytosis of HIV-1 in a manner dependent on the interaction of gp340 with the viral envelope.

Most host cellular factors that interact with HIV Env and do not mediate fusion do so through glycosylation-based interactions with the complex sugars that coat the exterior of the viral envelope protein. The 24 N-glycosylation moieties on the Env protein are highly mutable, and interactions with them are, for the most part, relatively nonspecific. This makes them poor targets for therapeutics and vaccines. The association between HIV Env and gp340, however, is a protein-protein interaction (53, 54). Based on peptide binding studies and subsequent computer modeling of the interaction of Env with gp340, the interaction sites correspond to the base of the Env V3 loop and the SRCR domains of gp340 (54). The amino acid sequence of the V3 loop base is highly conserved. Further, this site contains Arg298, which is critical for interaction with syndecans (13), suggesting that the site may support interactions with multiple host factors important for transcytosis.

gp340 in human alveolar macrophages can be found trafficking between the extracellular membrane and intracellular endosomes after ligand binding (32, 33). Vaginal and cervical cells that express gp340 are capable of binding, harboring, and then transferring virus to susceptible targets of infection (7, 15, 23) up to 4 days post viral pulsing (47). Further, treatment of these and other gp340-expressing cell lines with trypsin following viral pulsing does not ablate their abilities to maintain and transmit virus. Interestingly, the rabbit orthologue of gp340 functions as a modulator of polarity in epithelial cells (48, 49). While a polarity reversal function for human gp340 has yet to be demonstrated, it remains a distinct possibility, given our finding that gp340 mediates transcytosis.

Studies of HIV transmission in the human population demonstrate that transmission is an uncommon event, with a probability of infection per exposure between 0.005 and 0.3% (17, 45). Despite measurable levels of transcytosed virus, our results support the finding that the epithelial cell barrier represents a significant impediment to viral transmission. Even with a viral dose that exceeds that found in human semen by severalfold, ∼0.6% of the virus successfully traverses a single cell layer by transcytosis, but our preliminary results suggest that this small amount of transcytosed virus remains capable of infecting target cells. This was reduced and infection was inhibited by blocking the binding of Env to both gp340 and HSPG moieties. This suggests that blockage of cellular receptors that mediate transcytosis could serve as an effective means of reducing transmission.

Limitations remain in the use of in vitro systems including human genital tract tissue as a model of HIV-1 transmission. Given the inability to demonstrate comparable in vivo functionality for one of the more highly studied HIV host factors, DC-specific intercellular adhesion molecule-grabbing integrin, an in vivo correlation in a physiologically relevant model is necessary to further support the role of gp340 in HIV transmission. In order to understand the function of gp340 in in vivo transcytosis, a macaque model of the interaction of gp340 with virus must be developed. Only recently has the chromosomal locus of the gp340 orthologue in macaques been identified. Interestingly, in macaques, gp340 is linked to a progesterone response element (1). The macaque protein has not been characterized, nor has its tissue distribution been identified, but our preliminary studies demonstrate that the SRCR region of macaque gp340 can bind HIV Env and macaque vaginal and cervical tissue-expressed gp340 mRNA (data not shown). Comparison of the basic cell biology of macaque gp340 and testing of its role in vaginal transmission remain important next steps in understanding human HIV transmission during heterosexual intercourse.

This research presents gp340 as a facilitator of transcytosis in HEC1A cells and normal human endocervical tissue. Through inhibition studies directed at blocking the interaction of gp340 with HIV Env, the contribution of gp340 to viral transcytosis was identified and determined to account for about half. With the help of combinational treatments, the roles of gp340 and HSPG moieties in transcytosis appear distinct based on the ability to inhibit selectively and additively. Blockage of both receptors that mediate transcytosis in HEC1A cells or human tissue nearly blocked their ability to transcytosis virus. This suggests that there might still exist other factors that contribute to transcytosis in these cells or that existing inhibitors are not 100% effective. However, discovery of gp340 as a host cell factor that promotes HIV infection through two mechanisms, trans infection, with promotion of viral infectivity (47), and transcytosis, continues the identification of the earliest events in HIV transmission and provides an additional target for the development of preventive therapeutics.