DC-specific ICAM-3-grabbing nonintegrin mediates internalization of HIV-1 into human podocytes

Abstract

Human immunodeficiency virus (HIV)-1 has been demonstrated to contribute to the pathogenesis of HIV-associated nephropathy. In renal biopsy studies, podocytes have been reported to be infected by HIV-1. However, the mechanism involved in HIV-1 internalization into podocytes is not clear. In the present study, we evaluated the occurrence of HIV-1 internalization into conditionally immortalized human podocytes and the mechanism involved. Human podocytes rapidly internalized R5 and X4 HIV-1 primary strains via an endocytosis-dependent pathway, without establishing a productive infection. The HIV-1 internalization was dendritic cell-specific ICAM-3-grabbing nonintegrin (DC-SIGN) receptor mediated. The role of DC-SIGN was confirmed by using specific blocking antibodies and transfection with small interfering (si) RNA/DC-SIGN. Since podocyte HIV-1 trafficking was not altered by pH-modulating agents, it appeared that HIV-1 routing occurred through nonacid vesicular compartments. Interestingly, transfection of podocytes with neither siRNA/caveolin-1 nor siRNA/clathrin heavy chain inhibited podocyte viral accumulation. Thus it appears that clathrin-coated vesicles and caveosomes may not be contributing to HIV-1-associated membrane traffic.

visceral epithelial cells (podocytes) are highly specialized and differentiated cells which support the glomerular capillaries, synthesis of glomerular basement membrane, and regulation of glomerular permselectivity (31). The study of podocytes in culture had been controversial because of its terminally differentiated phenotype. However, this problem was taken care of by the development of a constitutively immortalized human podocyte cell line by transfection with the temperature-sensitive simian virus 40 (SV40)-T gene (50). These cells exhibit a regular, cobblestone-like polygonal phenotype and express podocyte-specific O-acetylated ganglioside, WT-1 protein, nephrin, and CD2AP, providing a suitable phenotype for studying pathways involved in human renal disorders including HIV-associated nephropathy (HIVAN).

Many studies confirmed the central role of podocytes in the pathogenesis of HIVAN (29, 66, 67), which undergo characteristic phenotypic alteration-dysregulated phenotype, re-entry of the cell cycle, and proliferation (2). Podocyte dysfunction appears to be the result of the combination of the direct effect of viral gene product expression (66, 68), the specific host, and genetic factors (26, 28, 30, 44). However, much of our understanding of the effect of HIV-1 on podocytes is related to the rodent models, which allow only the study of the postentry phase of the HIV-1 life cycle; therefore, little is known about the interaction between HIV-1 and human podocytes in general, and the mechanism involved in podocyte HIV-1 entry in particular.

In general, HIV-1 enters susceptible cells by fusion of its envelope with the plasma membrane after binding to the CD4 molecule (11) and interaction with the chemokine coreceptors CCR5 or CXCR4 (54). Nevertheless, HIV-1 entry into CD4-negative human cells has been widely reported (1, 5, 15, 57, 59). In the majority of instances, HIV-1 entry into CD4-negative cells has been reported to occur through the endocytic pathways (35, 38, 62, 63) and at times, resulted in the productive infection (13, 63).

Dendritic cell-specific ICAM-3-grabbing nonintegrin (DC-SIGN) is an endocytic receptor, which belongs to the C-type lectin family. This large group of proteins [such as the macrophage mannose receptor (MMR) and DEC-205] are specialized in the recognition of carbohydrate structures present on cellular and viral proteins and are implicated in several biological processes, including cell adhesion and antigen presentation (17, 53). The role of C-type lectins in HIV-1 pathogenesis has been highlighted by studies demonstrating their capability to bind HIV-1 in a CD4-independent manner (10, 23, 41, 58, 65), which may further increase viral transmission to permissive cells (19). DC-SIGN is highly expressed in dendritic cells (56, 60) and was originally cloned for its ability to bind and internalize heavily glycosylated HIV-1 gp120 protein (10). Interestingly, the expression of DC-SIGN has also been demonstrated in other cells, including macrophages (53) and platelets (9).

In the present study, we have shown that human podocytes express the DC-SIGN receptor, which efficiently captures HIV-1. Specific antibody-blocking studies inhibited entry of the virus in podocytes transfected with siRNA/DC-SIGN as well as podocyte transfer of HIV-1 to T cells confirmed the functionality of DC-SIGN in these cells. The interaction between HIV-1 and human podocytes resulted in the activation of podocyte HIV-1 trafficking and degradation. Our results provide insights into the role of the attachment factors on human podocytes involved in HIV-1 internalization that appears to involve the DC-SIGN-dependent endocytic pathway.

MATERIALS AND METHODS

Reagents.

Chloroquine (CLQ), ammonium chloride (NH₄Cl), bafilomycin A1, and aldrithiol-2 (AT-2) were purchased from Sigma-Aldrich (St. Louis, MO). T-20 (Roche, Indianapolis, IN) was obtained from NIH AIDS Research and Reference Reagent Program, (cat. no. 9845). Clathrin heavy chain (HC) siRNA (h), caveolin-1 siRNA (h), clathrin HC (TD.1) mouse monoclonal antibody, and caveolin-1 (N20) rabbit polyclonal antibody were obtained from Santa Cruz Biotechnology (Santa Cruz CA).

Cells and viruses.

Immortalized human podocytes were developed from primary human podocytes by transfection with the temperature-sensitive SV40-T gene and were characterized by their expression of WT-1 and nephrin (50). Human peripheral blood mononuclear cells (PBMCs) were prepared from peripheral blood mononuclear cells of HIV-1-negative donors as described previously (23). PBMCs were isolated from blood obtained from the New York Blood Bank (unidentified donors). Monocytes were obtained from PBMCs that adhered to plastic plates and cultured for 6 days in medium containing 100 U/ml GM-CSF (Sigma-Aldrich) and 10 ng/ml IL-4 (PeproTech, Rocky Hill, NY) to generate immature dendritic cells (IDCs). Several strains of HIV-1 were used, including R5 HIV-1_92US660 and X4 HIV-1_92HT599. Viral inoculums were adjusted according to reverse transcriptase (RT) activity to 20 × 10⁴ cpm/ml. In cocultivation studies, lymphocytes activated with phytohemagglutinin (PHA; 5 μg/ml, Sigma) were added to podocytes in a ratio of 2:1 in a medium supplemented with 20 U/ml of IL-2 (Roche). In blocking studies, cells were pretreated with specific antibodies: anti-DEC-205 (MG38; Serotec, Raleigh, NC); anti-CXCR4 (NIH AIDS Research and Reference Reagent Program, cat. no. 4083), anti-MMR (15-2; Santa Cruz Biotechnology); anti-DC-SIGN (ab13487; Abcam, Cambridge, MA); or a matching isotype control (Serotec) for 30 min before HIV-1 treatment. For inactivation with AT-2, a viral stock solution was treated with 1 mM AT-2 in DMSO or DMSO alone for 1 h at 37°C. Treatment agents were removed by ultrafiltration with a centricon centrifugal filter device with a YM-100 MW membrane (Millipore, Bedford, MA). A control virus preparation was processed in parallel with inactivated samples. Human podocytes were transfected with specific siRNA against DC-SIGN (Invitrogen), clathrin HC (Santa Cruz Biotechnology), caveolin-1 (Santa Cruz Biotechnology), or control siRNA at concentration of 40 nM, using as a transfection reagent Lipofectamine 2000 (Invitrogen). After 6 h of transfection, the cells were grown for another 48 h and then used for Western blot analysis or treatment with HIV-1. 293T cells were transfected with pcDNA3-DC-SIGN-expressing plasmid (NIH AIDS Research and Reference Reagent Program) or control plasmid (Clontech) using FuGENE (Roche) transfection reagent for 3 days.

Detection of HIV-1-specific DNA by PCR.

Before pulsing, viral stocks were treated (1 h at room temperature) with 200 U/ml of RNase-free DNase (Roche). Before preparation of cell lysates, control and HIV-pulsed cells were washed once with PBS, then treated with 0.05% trypsin at 37°C for 10 min to eliminate noninternalized virus, and subsequently washed four times with PBS. Cell lysates were subjected to PCR analysis using HIV-1-specific primers amplifying early transcript LTR RU5, as well as late pol and gag transcripts as described previously (23). Amplified DNA was analyzed by Southern blot hybridization using ³²P-labeled probes. Amplification of the β-tubulin gene was used to control the amount of DNA in each sample. Serial dilutions of 8E5/LAI cells, containing one HIV-1 genome per cell, were included in each amplification reaction to standardize the results. In each set, the number of experiments was carried out in duplicate.

HIV-1 p24 Ag detection.

HIV-1 p24 Ag concentrations in the culture supernatants were determined by ELISA (41). In brief, cell-free culture supernatants were lysed with Empigen BB followed by p24 capture by an immobilized anti-p24 Ag polyclonal antibody (D7320; Aalto Bio Reagents, Dublin, Ireland). Bound p24 Ag was then detected using an alkaline phosphatase-conjugated anti-p24 Ag monoclonal antibody (BC 1071-AP; Aalto Bio Reagents) and the AMPAK ELISA amplification system (DAKO, Glostrup, Denmark).

Real-time RT-PCR.

Total RNA was isolated using an RNeasy Kit (Qiagen, Valencia, CA). To eliminate DNA contamination, extracted RNA was subjected to digestion using RNase-free DNase. Reverse transcription was performed with 1 μg of RNA using Moloney murine leukemia virus reverse transcriptase (MMLV-RT; Life Technologies BRL, Carlsbad, CA) in a reaction mix that contained 5 mM MgCl₂, 50 mM KCl, 10 mM Tris·HCl, 1 mM dNTP, 2.5 μM random hexamers, and 2.5 U/μl MMLV-RT. Quantitative RT-PCR was carried out in an ABI Prism 7900HT sequence detection system using the primers shown in Table 1. SYBR green was used as the detector and ROX as the passive reference gene. The data were analyzed using the comparative Ct method (Δ^Ct method). All data were controlled for quantity of RNA input by performing measurements on an endogenous reference gene, GAPDH. In each set, the number of experiments was carried out in duplicate.

Table 1.

Primer sequences for real-time PCR

Gene	Primer Sequence
CD4	5′ 5′-GGCAGTGTCTGCTGAGTGAC-3′
	3′ 5′-GACCATGTGGGCAGAACCT-3′
CCR5	5′ 5′-AATATGAACGGTGAGCATTGTG-3′
	3′ 5′-AAGGAAAGACATTTCCCTTCG-3′
CXCR4	5′ 5′-AGGATATAATGAAGTCACTATGGGAAA-3′
	3′ 5′-AAGGGCACAAGAGAATTAATGTAGA-3′
DC-SIGN	5′ 5′-CCCAGCTCGTCGTAATCAA-3′
	3′ 5′-GAAGCGGTTACTTCTGGAAGAC-3′
l-SIGN	5′ 5′-CGGGGTAGGGATCTGAACTC-3′
	3′ 5′-CAGGAACCCAAGCCTCAAC-3′
DEC-205	5′ 5′-CACTTTTTAACGGACCAGTTCAG-3′
	3′ 5′-GGGCTCCTTTTATTCAAACCA-3′
MMR	5′ 5′-CACCATCGAGGAATTGGACT-3′
	3′ 5′-ACAATTCGTCATTTGGCTCA-3′
GAPDH	5′ 5′-ACTCCACTCACGGCAAATTC-3′
	3′ 5′-ACTGTGGTCATGAGCCCTTC-3′

DC-SIGN, dendritic cell-specific ICAM-3-grabbing nonintegrin; MMR, macrophage mannose receptor.

Electron microscopy.

Podocytes were pulsed with HIV-1 for 1 h, then washed and fixed with 2% glutaraldehyde, buffered with 0.05 M Na cacodylate (pH 7.3). The cell pellet was postfixed in osmium tetroxide and prepared for electron microscopy (EM) studies by routine procedures.

Western blotting and immunoprecipitation.

Cells were lysed in RIPA buffer (50 mM Tris·HCl, pH 8, 1% Triton X-100, 100 mM NaCl, 1 mM MgCl₂). The protein fraction was analyzed on SDS-PAGE, followed by transfer to a nitrocellulose membrane (Bio-Rad, Hercules, CA), and immunoblotted with specific antibodies [anti-DEC-205 (H-250), anti-CXCR4 (C-20; Santa Cruz Biotechnology), and anti-DC-SIGN (ab5716; Abcam)]. Blots were developed with horseradish peroxidase-conjugated appropriate secondary antibody (Santa Cruz Biotechnology) followed by enhanced chemiluminescence (PerkinElmer, Boston, MA). DC-SIGN immunoprecipitation was done with anti-DC-SIGN (120507; Abcam).

Flow cytometric analysis.

Podocytes were washed with PBS and blocked with PBS containing 20% human serum. Afterward, cells were stained with specific antibodies [anti-DEC-205-FITC (MG38; eBiosciences, San Diego, CA); anti-DC-SIGN-FITC (DCN46; BD Pharmingen, San Jose, CA); and anti-CXCR4 (C-20; Santa Cruz Biotechnology) and corresponding isotype antibody (BD Pharmingen)]. After washing, cells were stained with appropriate secondary FITC-conjugated antibody (BD Pharmingen) and then fixed with 4% buffered formalin. Staining was analyzed with FACS Calibur (Becton Dickinson, San Jose, CA) using Cell Quest software.

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide assay.

Cell viability in the presence of used drug concentrations was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT; Sigma) assay as described previously (39).

Statistical analysis.

Student's t-test was used to calculate P values.

RESULTS

HIV-1 entry into human podocytes.

Human podocytes were exposed to primary strains of HIV-1 with different coreceptor usage, R5 HIV-1_92US660 and X4 HIV-1_92HT599. There was no difference in viability of cells treated under control and experimental conditions (data not shown). As shown in Fig. 1A, human podocytes exposed to either R5 or X4 viral strains were positive for the specific HIV-1 strong-stop DNA (LTR RU5), the HIV-1 early reverse transcript. The adequacy of washing was tested on HIV-1-nonpermissive 293T cells, transfected with either control or DC-SIGN plasmid. Both control (mock) and cells expressing control-plasmid were negative for LTR RU5, whereas, 293T cells expressing DC-SIGN were positive for LTR RU5 (Fig. 1B). The time course effect on podocyte HIV-1-LTR RU5 content was studied by incubating human podocytes with the primary HIV-1_92US660 for variable time periods. HIV-1 strong-stop DNA was detected in podocytes within 30 min, whereas, podocyte viral load peaked at 2 h and decreased by 24 h (Fig. 1C).

Fig. 1.

Internalization of human immunodeficiency virus-1 (HIV-1) in human podocytes. Before incubation, HIV-1 was treated with 200 U/ml of RNase-free DNase (to eliminate contamination with viral DNA) for 1 h at room temperature. A: podocytes were exposed to different primary HIV-1 strains, R5 HIV-1_92US660 and HIV-1_92HT599. Cells were pulsed with the virus for 24 h, followed by washing with PBS, and then treatment with 0.05% trypsin at 37°C for 10 min (to eliminate noninternalized virus). The cells were washed (again) 4 times with PBS and then plated for culture. After 24 h, cells were trypsinized, washed, lysed, and analyzed for HIV-specific LTR RU5 by PCR. One representative experiment of 3 is shown. Mock infection served as a negative control. B: to verify appropriate viral washing, HIV-1-nonpermissive 293T cells (mock) and 293 transfected with DC-specific ICAM-3-grabbing nonintegrin (DC-SIGN) plasmid or control plasmid (control) were incubated with HIV-1 for 24 h and washed as described in A. HIV-1-nonpermissive 293T cells (mock) and 293 transfected with control plasmid (control) were used as washing controls. C: to facilitate synchronized viral entry, cells were preincubated with HIV-1_92US660 at 4°C; subsequently, after 1 h, cells were transferred to 37°C. At the indicated times of pulsing with HIV-1 (0–72 h), cells were trypsinized, washed, and analyzed for LTR RU5 by PCR. Amplification of the α-tubulin gene was used to control for the amount of DNA. Results represent relative LTR RU5 counts/min (cpm; means ± SD) of 3 independent experiments. D: dilutions of 8E5/LAI cells that contained 1 HIV-1 genome/cell were used as PCR standards (HIV-1 copies).

Intracellular viral localization by EM studies.

To confirm HIV-1 internalization, EM studies were carried out. Podocytes showed the presence of intracellular viral-like particles in the vacuoles located in close proximity to endoplasmic reticulum (ER) (Fig. 2).

Fig. 2.

Intracellular localization of viral-like particles by electron microscopy. Podocytes were washed and fixed after 1 h of incubation with HIV-1₆₆₀. Fixed cells were scraped, stained, and embedded. Thin sections were cut and evaluated under electron microscopy. Viral-like particles (white double line arrows) are shown in various vesicular compartments in close proximity to endoplasmic reticulum (with single line arrows). Magnification: ×40,000 (a–c) and ×100,000 (d–f).

Noninfectious entry of HIV-1 into human podocytes.

Since we did not detect the HIV-1 late reverse transcripts (pol, gag) (Fig. 1C) and because the nascent synthesis of HIV-1 strong-stop DNA has been reported to occur also within a viral particle (64), we wanted to ascertain whether the presence of LTR RU5 was the result of reverse transcription that occurred in the host cell or whether it was a part of the incorporated viral particle. In these studies, HIV-1 was inactivated with AT-2, which eliminates infectivity but preserves conformational and functional integrity of viral surface protein (49). AT-2-treated virions bound comparably to native virions (to target cells) but showed arrest of the progression of viral life cycle. Since there was no difference in strong-stop DNA content between AT-2-HIV-1 vs. native-HIV-1-treated podocytes, it appeared that the proviral DNA was part of the virions (Fig. 3).

Fig. 3.

Noninfectious entry of HIV-1 into human podocytes. HIV-1_92US660 was inactivated with aldrithiol-2 (AT-2) for 1 h at 37°C. Control virus was processed in parallel with inactivated samples. Before incubation, viral stock was treated with 200 U/ml of RNase-free DNase. After 2 h of incubation with HIV-1_, podocytes were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5. Mock infection served as a negative control. Amplification of the α-tubulin gene was used to control for the amount of DNA. Relative LTR RU5 cpm represent means ± SD from 2 independent experiments.

Because cell-free HIV-1 transcytosis has been proposed as a potential mechanism for viral crossing to several tissue-specific cells (3, 4, 21, 33), we studied whether the cell-free HIV-1 could cross human podocytes and escape (without the establishment of productive infection). After treatment with HIV-1 (2 h), podocytes were trypsinized, washed, and seeded onto Transwells (0.4-μm membrane separating podocytes from primary human activated PBMCs). During 14 days of cocultivation, no HIV-1 p24 antigen was detected (data not shown). Similarly, the cell-free supernatants harvested at 24 and 48 h (after the cells were exposed to HIV-1 and washed) were not able to establish the productive infection in activated PBMCs (data not shown). These findings indicate that the internalization of captured HIV-1 by human podocytes results in the intracellular hijacking and degradation of viral particles.

Expression of HIV-1 receptors in human podocytes.

The primers used for gene amplification of HIV-1 receptors in CIHPs are described in Table 1. We did not find any detectable levels of gene expression for CD4 and CCR5 receptors (Table 2). Consistent with the published data (25), we found very low CXCR4 gene expression. Moreover, podocytes showed mRNA expression of C-type lectins, including DEC-205 and DC-SIGN, and no expression of MMR. Data on podocytes and PBMCs (RT-PCR) are shown in Fig. 4A. Percentages of human podocytes showing surface expression of different receptors were as follows: DEC-205 (25%), DC-SIGN (3%), and CXCR4 (0%) (Fig. 4B), whereas Western blot analysis showed relatively high expression of DC-SIGN protein, very low expression of DEC-205 protein, and no expression of CXCR4 (Fig. 4C). Because other homologs of DC-SIGN protein, l-SIGN (also called DC-SIGNR) (53), have been reported to be able to bind HIV-1 (48), we also verified its expression by real-time PCR; however, as shown Table 2, we did not detect the expression of l-SIGN transcripts in CIHPs.

Table 2.

Analysis of expression of HIV-1 receptors by real-time PCR

Receptor	PBMCs	Podocytes	IDCs
CD4	34,550 ± 452	00 ± 00	*
CXCR4	1,028.0 ± 120.2	2.0 ± 0.7	*
CCR5	40.0 ± 5.8	0.0 ± 0.0	*
DEC-205	45.0 ± 2.3	10.0 ± 2.5	*
MMR	2.5 ± 0.9	0.0 ± 0.0	*
DC-SIGN	*	8.0 ± 2.9	1,905.0 ± 145.7

Values are means ± SD and represent relative gene expression (2 Δct), normalized to GAPDH, of 2 independent experiments. Total RNA, extracted from podocytes, activated peripheral blood mononuclear cells (PBMCs), and immature dendritic cells (IDCs), was subjected to real-time RT-PCR amplification using primers (Table I) for classic human immunodeficiency virus (HIV)-1 receptors, including CD4, CCR5, CXCR4, and C-type lectin receptors, including DEC-205, MMR, and DC-SIGN.

^*, Not done.

Fig. 4.

Analysis of the expression of HIV-1 receptors by RT-PCR, FACS, and Western blotting. A: expression of CD4, CCR5, CXCR4, DC-SIGN, DEC-205, and MMR was analyzed by RT-PCR. B: expression of DEC-205, DC-SIGN, MMR, and CXCR4 was analyzed by FACS. C: Western blot analysis of podocytes for DEC-205, MMR, DC-SIGN, and CXCR4. PBMC, peripheral blood mononuclear cells; MMR, macrophage mannose receptor.

DC-SIGN-dependent endocytosis of HIV-1 into human podocytes.

To rule out podocyte HIV-1 entry through fusion (conventional HIV-1 receptors), we evaluated the effect of the T-20, a known HIV-1 fusion inhibitor, on podocyte viral entry. Although T-20 completely inhibited the infection of primary R5 and X4 HIV-1 strains in T cells, it had no effect on the internalization of the same primary R5 and X4 HIV-1 strains in podocytes (data not shown). In addition, pretreatment of podocytes with an anti-CXCR4-specific antibody before HIV-1 exposure did not decrease the HIV-1 strong-stop DNA concentration (Fig. 5A). These findings suggest that podocyte viral entry is not dependent on classic fusion-mediated entry.

Fig. 5.

Anti-DC-SIGN-antibody inhibits HIV-1 entry into human podocytes. A: cells were preincubated in media containing either buffer (control), anti-DC-SIGN antibody (10 μg/ml), anti-CXCR4 antibody (10 μg/ml), anti-MMR antibody (10 μg/ml), anti-DEC-205 antibody (10 μg/ml), or matching isotype antibody (10 μg/ml) for 30 min followed by pulsing with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase). After 2 h of viral pulsing, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. The percentage of inhibition represents the ratio of the relative LTR RU5 cpm between samples treated with anti-DC-SIGN-, anti-CXCR4-, anti-MMR-, anti-DEC-205-, isotype antibodies, and control. Values are means ± SD from 3 independent experiments. B: dose-response of anti-DC-SIGN antibody_. Relative LTR RU5 cpm represent means ± SD of 3 independent experiments. One representative experiment is shown for the presence of HIV-1-specific-strong-stop DNA. Mock pulsing served as a negative control. Amplification of the β-tubulin gene was used to control for amount of DNA. *P < 0.0001 compared with control. **P < 0.004 compared with control.

To determine the role of endocytosis, we evaluated the effect of hypertonic media (a nonspecific inhibitor of endocytosis) (12, 24) on podocyte HIV-1 entry. As shown in Fig. 6, there was complete inhibition of HIV-1 entry in podocytes in the media containing 0.45 M sucrose. Because C-type lectins are bound to the high-mannose-containing carbohydrates (16), we evaluated the effect of yeast mannan on podocyte HIV-1 entry. As shown in Fig. 6, mannan (5 mg/ml) blocked the internalization of virus by 80%. These findings supported the role of mannose in podocyte HIV-1 entry.

Fig. 6.

Effect of inhibitors of endocytic pathway on podocyte HIV-1 internalization. Podocytes were preincubated in media containing either buffer (control), sucrose (0.45 M), or mannan (0.2, 1.0, or 5 mg/ml) for 1 h followed by pulsing with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase). After 2 h of incubation, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. Amplification of the α-tubulin gene was used to control for the amount of DNA. Relative LTR RU5 cpm represent means ± SD from 3 independent experiments. One representative experiment of 3 is shown for the presence of HIV-1-specific strong-stop DNA. Mock pulsing served as a negative control. *P < 0.0002 compared with control. **P < 0.0125 compared with control.

To identify the role of the specific lectin involved in podocyte HIV-1 entry, we carried out blocking studies. Podocytes were pretreated with specific antibodies against DEC-205 and DC-SIGN before incubation with HIV-1. As shown in Fig. 5A, preincubation with an anti-DC-SIGN antibody markedly decreased HIV-1 entry in human podocytes. In contrast, no inhibition of podocyte HIV-1 entry was observed with pretreatment with anti-DEC-205 or a correspondent isotype. The effect of the anti-DC-SIGN antibody was dose dependent (with the saturation of inhibition at a concentration of 10 μg/ml) (Fig. 5B). In addition, transfection with specific siRNA/DC-SIGN significantly inhibited the entry of virus in human podocytes compared with the control siRNA (Fig. 7).

Fig. 7.

Downmodulation of DC-SIGN by specific small interfering (si) RNA inhibits HIV-1 internalization in podocytes. A: podocytes nontransfected (NT) or transfected with specific DC-SIGN siRNA (DC-SIGN) or control siRNA (control) were pulsed with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase) for 2 h. After treatment, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. Amplification of the β-tubulin gene was used to control for the amount of DNA. One representative experiment of 2 is shown. Mock pulsing served as a negative control for LTR RU5. B: Western blot analysis of DC-SIGN protein in podocytes NT or transfected with specific DC-SIGN siRNA (DC-SIGN) or control siRNA (control).

Human podocytes express DEC-205/DCL-1 fusion mRNA transcripts.

Since anti-DEC-205 did not inhibit HIV-1 entry into podocytes, we hypothesized that the podocytes may not be expressing classic (wild-type) DEC-205 receptors. Recently, a novel DEC-205/DCL-1 fusion protein has been demonstrated in a Hodgkin's lymphoma cell line and in human dendritic cells (6, 27). The DEC-205/DCL-1 fusion protein is a splice variant of DEC-205 and a novel C-type lectin, DCL-1, that carries the extracellular domains of DEC-205 and a transmembrane and cytoplasmic domain derived from DCL-1 (Fig. 8B). To detect mRNA for DEC-205 (between CTLD10 and the cytoplasmic domain) and DCL-1 (CTLD and the cytoplasmic domain), we have used specific primers for the cytoplasmic tail (6). The DEC-205/DCL-1 fusion transcripts were detected by combining the DEC-205 forward primer and the DCL-1 reverse primer. A very small amount of the wild-type DEC-205 mRNA transcripts was detected with the primers specific for the cytoplasmic tails of this receptor compared with the wild-type of the DCL-1 receptor and DEC-205/DCL-1 fusion transcripts (Fig. 8A). Upon sequence analysis of the DEC-205/DCL-1 fusion protein and wild-type DCL-1, using primers shown in Fig. 8C, the transcripts were found to correspond to the DEC-205/DCL-1 fusion protein variant (Ref. AY184222) and DCL-1 (Ref. AY314007).

Fig. 8.

Detection of DEC-205/DCL-1 fusion mRNA transcripts in human podocytes. A: primers specific for DEC-205, DCL-1, and DEC-205/DCL-1 fusion were used to detect mRNA transcripts in cDNA from human podocytes. One representative experiment of 2 is shown. B: schematic representation of protein structures encoded by detected mRNA transcripts. C: primers used for amplification and sequencing of DEC-205/DCL-1 fusion protein and DCL-1 protein: CP7 (DCL-1), 5′-AACTGTGGTGAAACGAGAATCAG-3′; CP8 (DEC-205), 5′-AGAATAAATTTGTGAGCAGACTG-3′; CP9 (DCL-1), 5′-GGACTGTCCTTCATCTACTTG-3′.

DC-SIGN is a functional receptor for HIV-1 entry into human podocytes.

It has been reported that in dendritic cells, DC-SIGN not only captures HIV-1 but also contributes to cytoplasmic viral sequestration before viral transfer to T cells (18, 32). Moreover, stably DC-SIGN-transfected cells were not only capable of retaining HIV-1 for 3–6 days in culture but were also able to transmit HIV-1 to target cells (T cells) (18, 32). To determinate whether the DC-SIGN receptor expressed in human podocytes is a functional receptor for viral sequestration and/or viral transmission to the target cells, we cocultivated HIV-1-pulsed podocytes with activated PBMCs. HIV-1 could establish productive infection in lymphocytes for up to 6 days (Fig. 9), indicating that DC-SIGN expressed on human podocytes has the potential for viral sequestration. Since we could detect the HIV-1 strong-stop DNA (by PCR) only up to 72 h, it appeared that DC-SIGN could sequester only a minimal (undetectable) number of infectious viral particles in podocytes for a longer time.

Fig. 9.

DC-SIGN in human podocytes mediates transmission of virus to the target cells. Podocytes were pulsed with HIV-1_92US660. Before incubation, viral stock was treated with 200 U/ml of RNase-free DNase. After 4 h of incubation with HIV-1, cells were trypsinized, washed, and cultured for an additional 3 and 6 days. Afterward, cells were washed extensively and activated T cells were added (ratio, 1:2). Viral replication was measured in lymphocyte supernatants (collected at different time points after cocultivation) by p24 assay. Values are means ± SD of 2 independent experiments. Mock pulsing served as a negative control.

HIV-1 trafficking in human podocytes.

To gain insight into viral traffic, we evaluated the effect of vesicular pH-modulating agents on podocyte HIV-1 strong-stop DNA content. Cells were pretreated with CLQ and NH₄Cl, the agents known to block vesicle acidification as well as the endo-lysosomal degradation system. CLQ and NH₄Cl did not alter podocyte HIV-1 accumulation (Fig. 10A). To confirm this further, cells were pretreated with bafilomycin A followed by evaluation for podocyte viral accumulation. As shown in Fig. 10B, bafilomycin A did not modulate podocyte viral accumulation. These findings indicate that podocyte viral trafficking might have occurred through pH-neutral vesicles.

Fig. 10.

HIV-1 trafficking in human podocytes. A: effect of chloroquin (CLQ) and NH₄Cl on HIV-1 trafficking in podocytes. Podocytes were preincubated in media containing either buffer (control), CLQ (250 μM), or NH₄Cl (20 mM) for 1 h followed by pulsing with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase). After 4 h of treatment, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. Amplification of the α-tubulin gene was used to control for the amount of DNA. Relative LTR RU5 cpm represent means ± SD of 2 independent experiments. One representative experiment of 2 is shown for the presence of HIV-1-specific-strong-stop DNA. Mock pulsing served as a negative control. B: effect of bafilomycin A (BAF) on HIV-1 trafficking in podocytes. Podocytes were preincubated in media containing buffer (control) or BAF (100 nM) for 1 h followed by pulsing with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase). After 4 h of treatment, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. Amplification of the α-tubulin gene was used to control for amount of DNA. Two sets of experiments were carried out.

Since clathrin-coated vesicles and caveosomes also contribute to the endocytic pathway (34, 47), we studied the role of clathrin-coated vesicles and caveosomes on HIV-1 membrane traffic in podocytes. Podocytes were transfected with the specific siRNA/clathrin HC or siRNA/caveolin-1 or control siRNA, followed by viral entry protocol. As shown in Fig. 11A, cells transfected with siRNA/caveolin-1 showed diminished expression of caveolin compared with cells transfected with control siRNA and cells transfected with siRNA/clathrin HC. Similarly, cells transfected with siRNA/clathrin HC displayed diminished expression of clathrin compared with control and cells transfected with siRNA/caveolin-1 (Fig. 11B). Cells transfected with either siRNA/caveolin-1 or siRNA/clathrin HC did not modulate viral accumulation in podocytes (Fig. 11C).

Fig. 11.

Role of clathrin-coated vesicles and caveosomes in HIV-1 trafficking in podocytes. A and B: podocytes transfected with specific caveolin-1/siRNA (siRNA/CAV), clathrin/siRNA (CLA), or control siRNA (control) were harvested, followed by protein extraction and preparation of Western blots. The blots were probed for caveolin (A) and clathrin (B). The blots were stripped and reprobed for actin. C: podocytes transfected with siRNA/caveolin-1 (siRNA/CAV), siRNA/clathrin (siRNA/CLA), or control siRNA (control) were pulsed with HIV-1_92US660 (pretreated with 200 U/ml of RNase-free DNase) for 2 h. After treatment, cells were trypsinized, washed, and analyzed for HIV-1-specific LTR RU5 by PCR. Amplification of the α-tubulin gene was used to control for amount of DNA. One representative experiment of 2 is shown.

DISCUSSION

Human podocytes rapidly internalized R5 and X4 HIV-1 primary strains via an endocytosis-dependent pathway. Human podocytes expressed DC-SIGN receptors, which mediated capture of R5 and X4 HIV-1 primary strains. Since the internalized virus did not undergo reverse transcription and the amount of viral components decreased with time, it appears that HIV-1 particles in human podocytes are directed for degradation. Interestingly, a small fraction of the captured replication-competent viral particles survived inside the podocytes and could establish viral replication after transmission to target cells. These results are consistent with previously published studies showing that DC-SIGN not only captured and internalized HIV-1 particles but also enhanced trans-infection to T cells (18, 32).

We have previously reported that DEC-205 mediated HIV-1 entry in renal tubular cells (23, 39). Interestingly, podocytes also expressed DEC-205 receptors. Therefore, we expected a potential role of DEC-205 in podocyte HIV-1 entry. However, the blocking study with the specific anti-DEC-205 antibody did not support this notion. In addition, pH-modulating agents (contrary to renal tubular cells) did not show any effect on virus trafficking in podocytes. These findings indicated that pH-dependent endocytic compartments involved in renal tubular cell HIV-1 trafficking are not involved in podocyte HIV-1 trafficking. Further investigation revealed podocyte-expressed DEC-205/DCL-1 fusion protein instead of wild-type DEC-205 receptors. This hybrid protein carries the outer domain of DEC-205 and transmembrane and cytoplasmic domain of DCL-1 (6). These findings provided an explanation for nonfunctional DEC-205. At present, the role of DCL-1 and the functional significance of DEC-205/DCL-1 fusion protein are not known. However, expression of this protein in dendritic cells was associated with the abrogation of endocytic activity (6). We propose that podocyte expression of DEC-205/DCL-1 may be contributing to the lack of endocytic HIV-1 entry in podocytes.

We have recently demonstrated that HIV-1 entry into human podocytes was inhibited by disruption of specific membrane microdomains by cholesterol depletion (39). However, we did not evaluate any selective pathway in these studies. In dendritic cells, a significant portion of DC-SIGN resides in cellular membrane microdomains (7). Moreover, HIV-1 binding and uptake in IDCs have been reported to occur through a cholesterol-sensitive pathway (20). Many viruses, including HIV-1, target lipid rafts during both cellular entry and viral assembly (14, 37, 55). Recently, new data on the postendocytic trafficking of raft markers have provided insights into postendocytic trafficking. Interestingly, one of the better known lipid raft-mediated trafficking pathways discovered for SV40 involved pH-neutral vesicles (46).

The phagocytic pathway of the DC-SIGN receptor remains controversial. DC-SIGN binding has been shown to many pathogens (8, 51, 56, 64); however, several studies indicate that the internalization pathway of DC-SIGN in dendritic cells is dependent on the type of ligand. The antigen Lewis X and mycobacterial ManLAM are efficiently captured and internalized through DC-SIGN and are targeted to lysosomes (19). On the other hand, DC-SIGN-mediated HIV-1 entry was routed to nonlysosomal compartments and it was associated with sustained virulence in dendritic cells (32). Similarly, the hepatitis C virus entered via DC-SIGN/l-SIGN escape lysosomal degradation (36). Our study in human podocytes suggest the involvement of the neutral pH-compartment in the trafficking pathway of HIV-1; however, further studies are required to delineate the exact route.

In the present study, HIV-1 could establish productive infection in lymphocytes for up to 6 days, indicating that DC-SIGN expressed on human podocytes has the potential for viral sequestration. Since we could detect the HIV-1 strong-stop DNA (by PCR) only up to 72 h, it appeared that DC-SIGN could sequester only a minimal (undetectable) number of infectious viral particles in podocytes for a longer time. An alternative possibility can be considered that a smaller number of viral particles might have persisted by being attached to the cellular surface in the trypsin-resistant compartments. However, EM studies did not show any presence of HIV-1 particles on the surface of the human podocytes. Moreover, EM studies showed the presence of HIV-1 particles inside the cells. Thus it appears that there may be two possible sources for viral transfer to the immune cells.

We conclude that human podocytes have the capability to capture and internalize HIV-1 through a pH-independent endocytic pathway, which utilizes DC-SIGN receptors.

GRANTS

This work was supported by NIDDK Grant RO1 DK 083931. The reagents and viral strains were obtained through the National Institutes of Health (NIH) AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH included T-20 (fusion inhibitor from Roche; no. 9845), pcDNA3-DC-SIGN-expressing plasmid, anti-CXCR4 Ab (no. 4083), and all viral strains.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.