点击显示 收起
Section of Infectious Diseases, Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago
A small percentage of women at high risk for human immunodeficiency virus (HIV) exposure remain uninfected for long periods, protected by unknown mechanisms. We hypothesized that one mechanism could be inhibition of interactions between HIV and dendritic cellspecific intercellular adhesion molecule 3grabbing nonintegrin (DC-SIGN) in the genital tract. In an analysis of 95 cervicovaginal lavage samples, we found that 12 (12.6%) strongly inhibited the binding of laboratory-adapted and primary HIV-1 isolates to B-THP-1/DC-SIGN cells in a dose-dependent manner, independently of the donor's risk of exposure. Three of 5 primary isolates were also blocked from binding to primary DCs. The inhibitor has a high molecular weight, is heat stable, and is resistant to trypsin. It is sensitive to pronase and periodate, indicating that it is likely a glycoprotein. Mannosidase digestion and concanavalin A adsorption indicate that the terminal residues of the carbohydrate are not mannose. Mechanistic experiments indicate that the inhibitor acts via binding to DC-SIGN. Further study of such inhibitors may help to elucidate the role played by DC-SIGN in HIV transmission.
A subpopulation of women highly exposed to HIV through sexual contact remain uninfected [1, 2]. Several studies have investigated HIV-specific antibodies, cytotoxic T lymphocyte (CTL) responses, and cytokine levels [38]. Collectively, such studies have suggested that protection could be the result of a complex synergy between multiple factors. Therefore, it is important to identify any factors that could play a role in the interruption of HIV transmission.
Dendritic cellspecific intercellular adhesion molecule 3grabbing nonintegrin (DC-SIGN) is a gp120 receptor expressed on the surface of DCs present in the lamina propria of the vaginal, cervical, and uterine mucosae as well as in all layers of the rectal mucosa [9, 10]. DC-SIGN has been associated with increased uptake of HIV by CD4+ T cells [11]. Also via DC-SIGN, HIV is bound, internalized, and transferred to CD4+ T cells, suggesting a Trojan horse model of viral transmission [9, 12]. The transfer of green fluorescent proteinlabeled virus from DCs to T cells, along with the recruitment of HIV receptors to the site of contact between cells, has been observed [13]. Hu et al. have described the rapid migration of DC-SIGNrich DC-like cells from human cervical explants [14], and the DC-mediated trafficking of virus to lymph nodes has been demonstrated in a mouse model [15]. Despite the absence of expression in epithelial Langerhans cells [10], DC-SIGN may play a role in the sexual transmission of HIV, because subepithelial layers of mucosa become exposed as a result of microabrasions.
On the basis of the potential role this receptor plays in HIV transmission, we hypothesized that some factor that interferes with the binding of HIV to DC-SIGN might contribute to the protection observed in highly exposed, seronegative (HESN) women. We collected cervicovaginal lavage (CVL) samples from 95 women and found that 12 (12.6%) were able to strongly inhibit binding of laboratory-adapted and primary HIV-1 isolates to B-THP-1/DC-SIGN cells in a dose-dependent manner. Preliminary characterization suggests that the CVL inhibitor may be a large glycoprotein.
PARTICIPANTS, MATERIALS, AND METHODS
Participants and sample collection.
Women at high risk for heterosexual acquisition of HIV were recruited as part of a study of behavioral change resulting from participation in a phase 3 preventive vaccine trial (Centers for Disease Control and Prevention Project Vision). Women at low risk were recruited independently. Informed consent was obtained in accordance with institutional review board policies.
All participants were HIV seronegative and had never injected drugs. High-risk participants (n = 63) had to have had at least 1 HIV-seropositive sex partner or at least 2 of the following risk factors: crack cocaine use during the last 6 months; exchange of sex for money, drugs, or shelter during the last 6 months; at least 5 sex partners during the last 6 months; or a history of sexually transmitted diseases (STDs) during the last year. Low-risk participants (n = 32) had to have had no HIV-seropositive sex partners; never used crack cocaine; never exchanged sex for money, drugs, or shelter; had no more than 1 sex partner during the last 6 months and no more than 5 sex partners during the last 5 years; and no history of STDs (tables 1 and 2).
Participants were given physical examinations. Screening and treatment for bacterial vaginosis (BV) and infection with Neisseria gonorrhea, Chlamydia trachomatis, and Trichomonas vaginalis were performed according to approved methods. For CVL sample collection, the os of the cervix was washed with 10 mL of sterile saline. CVL samples were centrifuged, and aliquots of the supernatant were heat inactivated (HI; 15 min at 56°C) and filter sterilized.
Our study group, observed over 31.5 person-years (PY), underwent no seroconversions. In the HIVNET Vaccine Preparedness Study, which used identical inclusion criteria, a seroincidence rate of 1.13 cases/100 PY (95% confidence interval [CI], 0.572.27 cases/100 PY; P = .01) was observed [16].
Cells and viruses.
B-THP-1/DC-SIGN and matched parental B-THP-1 cells were provided by D. R. Littman [17, 18]. Peripheral-blood mononuclear cells (PBMCs) were obtained from blood from healthy donors by use of lymphocyte separation medium (Mediatech) and were maintained in the presence of 20 U/mL recombinant interleukin (rIL)2 (AIDS Research and Reference Reagent Program , Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health; from M. Gately, HoffmanLa Roche) [19]. Monocytes were isolated from PBMCs by use of huCD14 microbeads (Miltenyi Biotec) and were differentiated into monocyte-derived DCs (MDDCs) by culturing them for 7 days in AIM-V medium (Invitrogen) supplemented with 1000 U/mL each granulocyte-macrophage colony-stimulating factor (Immunex) and rIL-4 (R&D Systems). CD14- cells were used as peripheral-blood lymphocytes (PBLs).
HIV-1Ba-L and HIV-1MN were obtained from the ARRRP (HIV-1Ba-L from S. Gartner, M. Popovic, and R. Gallo and HIV-1MN [originally designated "HTLV-IIIMN/H9"] from R. Gallo) [2023]. HIV-1Ba-L was propagated in phytohemagglutinin (PHA)stimulated PBMCs, and HIV-1MN was propagated in CEM-SS cells (ARRRP; from P. L. Nara) [2426]. Primary isolates HIV-101USUIC01, HIV-193USUIC01, HIV-193USUIC02, HIV-193USUIC03, and HIV-193USUIC04 were obtained from acutely infected patients. Patient PBMCs were isolated and cocultured with day 3 PHA-stimulated PBMCs from healthy donors. Viral growth was monitored by p24 ELISA (AIDS Vaccine Program, National Cancer Institute). Viral tropism was determined on the basis of the ability to infect GHOST cells expressing CD4 and either CCR5 or CXCR4 (ARRRP; from V. N. KewalRamani and D. R. Littman) [27].
Inhibition assays.
Inhibition of HIV-1 binding to B-THP-1/DC-SIGN cells and MDDCs was assayed by a protocol described by Geijtenbeek et al. [9]. Briefly, 2 × 106 cells were incubated with 500 L of CVL sample (diluted in Dulbecco's PBS [D-PBS; BioWhittaker]) and HIV-1 (10 ng in 500 L) for 1 h at 37°C. Unbound HIV-1 was removed by 3 washes in RPMI 1640 with 1% HI fetal bovine serum (Mediatech) and 10 mmol/L HEPES buffer (BioWhittaker). Cells were lysed in 1% Triton X-100, and the amount of bound HIV-1 was determined by p24 ELISA. The amount of nonspecific HIV-1 binding was determined by exposing B-THP-1 cells to D-PBS and virus.
CVL samples were screened for inhibition at 2 dilutions with B-THP-1/DC-SIGN cells and HIV-1Ba-L. If binding was strongly inhibited, further dilutions were screened until <50% inhibition was observed. The IC50 is the inverse of the dilution at which 50% of viral binding is inhibited. CVL samples with <50% inhibition at a 1 : 10 dilution were interpreted as having no activity and were assigned an IC50 of 0.
For transinfection assays, MDDCs were incubated with 25 L of CVL sample for 30 min at 37°C and then with HIV-1 (500 L; 30 ng/mL) for 2 h. MDDCs were washed and resuspended in complete RPMI 1620 plus 20 U/mL rIL-2 and were added to wells containing PHA-activated, autologous PBLs at a ratio of 1 MDDC : 20 PBLs. Viral growth was quantified by p24 ELISA of culture supernatants on days 3, 7, and 10.
Molecular characterization.
To determine whether the CVL inhibitor bound to DC-SIGN, 15 L of CVL sample was incubated with 1 × 107 B-THP-1/DC-SIGN cells for 1 h at 37°C. The cells were pelleted, and the incubation was repeated with the supernatants. DC-SIGNbinding substances were also depleted by incubation with immobilized DC-SIGN. Soluble DC-SIGN was produced from plasmid DC-020 (gift from K. Drickamer) by the method described by Mitchell et al. [28] and was coupled to cyanogen bromideactivated agarose (Sigma). The DC-SIGN/agarose matrix was equilibrated with 4 times 4 volumes of D-PBS (with Ca++/Mg++); 0.5 volumes of an HI inhibitory CVL sample was incubated with the matrix for 30 min at room temperature, and the supernatant was carefully collected.
Antibodies were depleted from CVL samples by immunoprecipitation. Samples (50 L) were incubated for 30 min at room temperature with 27.5 L (1 mg/mL) of AffiniPure mouse antihuman IgG, F(ab)2, and 5 L (2 mg/mL) of AffiniPure goat antimouse IgG (heavy and light; Jackson ImmunoResearch). Antibody complexes were precipitated by centrifugation at 2000 g for 10 min. Human anticardiolipin antibody control was provided by M. Teodorescu. Carbohydrates with terminal mannose residues were depleted from CVL samples by use of ConA Sepharose resin (Amersham Pharmacia Biotech), in accordance with the supplier's instructions.
Periodate oxidation was performed as described by Oegema et al. [29]. Periodic acid (0.2 mol/L) and sodium acetate (50 mmol/L; pH 4.75) were diluted 1 : 10 in CVL sample and incubated for 20 h at 4°C in the dark. Oxidation was halted by use of 1.7 L of ethylene glycol, and the sample was recovered by dialysis. Digestion of CVL sample diluted in D-PBS with immobilized L-1-tosylamido-2-phenylethyl chloromethyl ketone (Pierce) was performed overnight in a shaker at 37°C. CVL sample diluted 1 : 10 in 50 mmol/L sodium acetate and 50 mmol/L glacial acetic acid (pH 5.0) was digested by use of -mannosidase Canavalia ensiformis (5 U/mL; Sigma) for 24 h at 37°C; the digestion was then halted by boiling for 5 min. HI (10 min at 95°C) CVL samples were digested by use of 500 U/mL pronase (Calbiochem) for 24 h at 37°C. The digestion was halted by heat (10 min at 95°C), and the products were fractionated in Microcon YM50 centrifugal filter devices (Millipore).
Statistical analysis.
Results were analyzed by 2 tests performed by use of SAS (version 8; SAS Institute).
RESULTS
Because DCs express multiple HIV receptors, we evaluated the effectiveness of CVL samples in inhibiting HIV binding to MDDCs. CVL sample 1 was diluted 1 : 100 in D-PBS, and CVL samples 2 and 3 were diluted 1 : 10 in the same medium. HIV-101USUIC01 binding to both B-THP-1/DC-SIGN cells and MDDCs was inhibited well. However, HIV-1Ba-L binding to MDDCs was unaffected by the CVL samples (figure 2). This is consistent with the high affinity of HIV-1Ba-L for CD4, compared with that of primary isolates [30].
We investigated the ability of CVL samples to block HIV binding to MDDCs further by performing transinfection inhibition assays, to distinguish between HIV binding to DC-SIGN (and similar receptors) and CD4. HIV-193USUIC02 and HIV-193USUIC03 behaved similarly to HIV-101USUIC01that is, they were inhibited from binding to MDDCs (and from subsequently transferring to PBLs) by CVL samples that were inhibitory with respect to B-THP-1/DC-SIGN cells, and they were not inhibited from binding to MDDCs by CVL samples that were noninhibitory with respect to B-THP-1/DC-SIGN cells. HIV-193USUIC04 appeared to transfer only in the presence of noninhibitory CVL samples, and HIV-193USUIC01 did not replicate under any conditions, including in the D-PBS control (figure 3).
We suspected that the CVL inhibitor blocked the gp120 binding site on DC-SIGN, but we also considered the alternative hypotheses: that it had a toxic effect on cells or HIV. To determine whether the CVL inhibitor was cytotoxic, we quantified the viability of B-THP-1/DC-SIGN cells after incubation with CVL samples. Viability was 91.0% before incubation, whereas, after incubation, it was 91.5% (SD, ±2.3%) (n = 12 samples), suggesting that CVLs do not affect cellular viability. To confirm viral infectivity after CVL exposure, an inhibitory CVL was diluted 1 : 100 in viral stock, incubated for 1 h at 37°C, and then added to B-THP-1/DC-SIGN cells or PHA-stimulated PBLs. The B-THP-1/DC-SIGN cells were washed and lysed as described above. PBLs were cultured for 4 days. At this dilution, the sample blocked 87.3% (SD, ±1.9%) of the viral binding to B-THP-1/DC-SIGN cells but inhibited viral replication in PBLs by only 20.2% (figure 4A and 4B). The inhibitory CVL did not appear to diminish viral infectivity or block CD4-mediated infection of lymphocytes.
To determine whether the CVL inhibitor blocked the gp120 binding site on DC-SIGN, we incubated an inhibitory CVL sample with B-THP-1/DC-SIGN cells and compared the ability of the supernatant to inhibit binding of HIV-1Ba-L to B-THP-1/DC-SIGN cells to that of untreated CVL. We saw no change in the inhibitory activity of the supernatant of the CVL sample incubated with parental B-THP-1 cells, but we did see a loss of activity after incubation with B-THP-1/DC-SIGN cells (figure 4C). We also found that the CVL inhibitor could be depleted by immobilized DC-SIGN, resulting in a shift of IC50 from 589 to <100. These data strongly suggest that the CVL inhibitor blocks HIV by binding to DC-SIGN. To distinguish between DC-SIGN binding and endocytosis, we performed parallel binding-inhibition assays at 37°C and on ice and observed no difference in the performance of the CVL inhibitor (data not shown). This result suggests that inhibition of HIV uptake occurs at the level of binding and not endocytosis.
Clinical associations.
Our initial hypothesis was that the CVL inhibitor is associated with risk of exposure to HIV. However, we found no significant difference between high-risk and low-risk women with respect to the number of CVLs with an IC50 100 (high risk, 8/63; low risk, 4/32 [P = .98]), except as described below. This implies that the CVL inhibitor is intrinsic to the host and may not be modulated by sexual behaviors.
We looked for relationships between a high IC50 and individual risk factors (including STDs/BV). An IC50 100 was found to be associated with having at least 1 HIV-seropositive partner (P = .05; odds ratio [OR], 8.50 [95% CI, 1.0568.89]). The large 95% CI is possibly due to the sample sizeonly 4 of 63 women reported having HIV-seropositive partners (table 1). In addition, there was a negative association between an IC50 >0 and white blood cells in the CVL sample (P = .02; OR, 0.49 [95% CI, 0.080.91]), suggesting that the CVL inhibitor may be down-regulated or degraded under inflammatory conditions. See tables 2 and 3 for associations between risk factors and IC50.
Presence of the CVL inhibitor was not stable over time. Follow-up CVL samples from 36 participants were assayed. None of the 7 participants who initially had an IC50 100 maintained that level of inhibition at follow-up. Of 18 participants with no inhibitory activity at baseline, only 10 still had no inhibition at follow-up. There was no difference between the means of the IC50s at baseline and follow-up (P = .57).
Molecular characterization of the CVL inhibitor.
We performed basic assays to discover the molecular nature of the CVL inhibitor. We found that heat did not lower the IC50. For CVL 1, the IC50 was 873 (SD, ±112) before heating (for 10 min at 95°C) and was 1060 (SD, ±94) after heating (n = 3 replicates; data representative of 3 experiments). To estimate the inhibitor's size, CVLs 1 and 2 were filtered in centrifugal filter devices with 10-, 50-, and 100-kDa cutoffs. All retentates inhibited HIV-1Ba-L from binding to B-THP-1/DC-SIGN cells (IC50 of 1521 [SD, ±2699], 919 [SD, ±233], and 1275 [SD, ±1096] for 10, 50, and 100 kDa, respectively). The 100-kDa filtrate had an IC50 of 795 (SD, ±4), suggesting that the CVL inhibitor is present as a high molecular weight compound but may not be of a uniform size (figure 5A).
Inhibition by CVL samples was not affected by trypsin. Untreated CVL sample blocked 99.8% (SD, ±0.3%) of HIV-1Ba-L binding to B-THP-1/DC-SIGN cells, and trypsinized CVL sample blocked 99.4% (SD, ±0.4%) (n = 3). In contrast, pronase digestion followed by fractionation on a 50-kDa molecular weight cutoff centrifugal filtration device shifted the IC50s of CVL samples 1 and 2 from 259 and 595, respectively, to <100, for both. These data suggest that the CVL inhibitor contains protein that is essential for activity. It is resistant to trypsin, because suitable cleavage sites are either lacking or shielded.
To determine whether antibodies played any role in inhibition, we immunoprecipitated all antibodies from CVL 1. At a dilution of 1 : 20, the antibody-depleted sample inhibited 95.8% (SD, ±7.6%) of HIV-1Ba-L from binding to B-THP-1/DC-SIGN cells, whereas the control inhibited 90.0% (SD, ±1.9%) (n = 2 replicates). Antibodies were successfully depleted from a control consisting of a known quantity of human anti-cardiolipin antibody (data not shown). This implies that the CVL inhibitor is not an antibody.
Because DC-SIGN is a lectin, we hypothesized that the CVL inhibitor's active site might be carbohydrate based. Terminal carbohydrate residues in CVL samples were oxidated with 0.2 mol/L periodic acid, and the oxidated CVL samples were no longer able to block HIV-1Ba-L binding to B-THP-1/DC-SIGN cells to the same degree (figure 5B). Because DC-SIGN was initially described as being mannose specific, we hypothesized that the terminal residue of the CVL inhibitor could be mannose. To this end, we depleted terminal mannose-bearing carbohydrates from CVL sample 3 by passing it over concanavalin A sepharose. The flowthrough was still able to inhibit HIV-1Ba-L binding to B-THP-1/DC-SIGN cells (99.4% [SD, ±1.6%], compared with 99.8% [SD, ±0.4%] for untreated CVL sample; n = 2 replicates). To further confirm that the CVL inhibitor does not depend on terminal mannose residues, we digested CVL sample 1 with -mannosidase; no change in the inhibition of binding was observed. Undigested CVL sample blocked 94.4% (SD, ±2.5%), and digested CVL sample blocked 97.8% (SD, ±1.2%) (n = 3 replicates). Together, these data suggest that the CVL inhibitor contains DC-SIGNbinding carbohydrate moieties that are not mannose derived. Furthermore, its sensitivity to pronase suggests that the inhibitor is a high molecular weight glycoprotein.
DISCUSSION
This is the first report of inhibition of HIV binding to DC-SIGN by a factor in the female genital tract. This inhibitor was found in 12.6% of our study population and was present in CVL samples from both low-risk and high-risk women. Inhibitory CVL samples blocked DC-SIGN binding to both X4 and R5 laboratory-adapted strains of HIV, as well as several primary isolates, in a dose-dependent manner.
Subpopulations of women highly exposed to HIV have been reported to remain uninfected for prolonged periods of time [1, 2]. Multiple studies have sought to identify unique immune responses that provide some mechanism of protection. HIV-specific immunoglobulins have been detected in vaginal secretions from HESN women at frequencies ranging from 0%76% for IgG and/or IgA [36]. Evidence of HIV-specific CTL responses has been identified in 55%69% of HESN women [3], compared with 73% of HIV-seropositive women [7, 8]. The production of cytokines and chemokines such as interferon-, IL-6, and RANTES is higher in cervicovaginal secretions of women who also have HIV-specific immunoglobulin [4]. Together, these findings suggest that there are probably multiple mechanisms of protection, acting either singly or in concert. Our present experiments provide evidence of a potential additional innate mechanism of protection that may be important in limiting HIV transmission.
We expected that this inhibitor would be associated with a high risk of exposure to HIV, but it was not. This may be due to the relatively low prevalence of HIV in Chicago (the site of our study), compared with sub-Saharan Africa (the site of many of the studies cited above), or in the seronegative partners of HIV-seropositive men. In our high-risk group, only 4 of 63 women reported having had at least 1 HIV-seropositive partner, a risk factor that was weakly associated with an IC50 100. A larger study focused on addressing this particular question could possibly find a stronger association. The only other relationship we identified was a negative association between an IC50 >0 and white blood cells in the CVL sample. This implies that the CVL inhibitor is less effective in the presence of white blood cells, as in inflammatory conditions. However, no association was seen between IC50 and certain causes of inflammation, such as STDs/BV. It was also not consistent over time; many of the donors provided CVL samples at baseline that demonstrated inhibition and then provided CVL samples during follow-up that did not demonstrate inhibition, and vice versa.
Although the CVL inhibitor was present in a number of HIV-seronegative women, it is difficult to determine whether it actually plays a role in protection against HIV transmission. We observed no seroconversions in our study population, with an expected seroincidence of 0.356 cases over the duration of our study [16]. Further determination of the protective effect conferred by this inhibitor should involve examination of an exposed, uninfected cohort, such as the uninfected female partners of HIV-seropositive men.
In control experiments, we demonstrated that the CVL samples did not diminish the viability of either the cells or virus and did not interfere with CD4-mediated infection of lymphocytes. We found that incubation with DC-SIGNexpressing cells or immobilized DC-SIGN successfully depleted activity from the CVL samples. This finding supports our hypothesis that the CVL inhibitor binds to DC-SIGN and obscures the gp120 binding site.
Inhibitory CVL samples effectively blocked the binding of the primary isolate HIV-101USUIC01 to MDDCs as well as the transinfection of 3 of the 5 primary isolates used in the present study. However, MDDC binding and transinfection of laboratory-adapted HIV-1Ba-L was not inhibited. This supports our observations that the CVL inhibitor is active against DC-SIGN and not gp120 or CD4. In this scenario, viruses that have the strongest tropism for DC-SIGN would be prevented from establishing DC-mediated infections in the presence of the CVL inhibitor, whereas viruses that have a stronger tropism for CD4, such as HIV-1Ba-L, would not be affected by the CVL inhibitor. This observation has important ramifications for research on the role played by DCs in HIV transmission. Different affinities for DC-SIGN among primary isolates suggests that DC-SIGN may, in fact, play an important role in HIV transmission. Future research will include evaluation of the affinity of primary isolates for DC-SIGN.
Basic characterization experiments were undertaken to assess the molecular nature of the CVL inhibitor. Filtration suggests that it is at least 50 kDa, distinguishing this inhibitor from simple sugars that bind to DC-SIGN [9]. Because the CVL inhibitor is inactivated by periodate oxidation, is heat stable, is resistant to trypsin, and is susceptible to digestion with pronase, it is likely to feature an active site comprised of carbohydrate moieties associated with protein. The results of immunoprecipitation, together with heat stability and trypsin resistance, suggest that sialylated antibodies can be eliminated from consideration. The CVL inhibitor failed to bind to concanavalin A, and mannosidase digestion did not eliminate inhibition; this reveals that the carbohydrate sequence is unlikely to terminate with a mannose residue and, therefore, could represent a novel DC-SIGNbinding group. In summary, these experiments describe a naturally occurring glycoprotein in the female genital tract that may provide innate protection against HIV infection. Research is ongoing to further characterize the molecular structure and immunological properties of this glycoprotein.
Acknowledgments
We thank Greg Spear, Marius Teodorescu, and Tom Hope, for their expertise, and Joanna Wang and Hua Yun Chen, for statistical analysis.
References
1. Fowke KR, Nagelkerke NJ, Kimani J, et al. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 1996; 348:134751. First citation in article
2. Butera ST, Pisell TL, Limpakarnjanarat K, et al. Production of a novel viral suppressive activity associated with resistance to infection among female sex workers exposed to HIV type 1. AIDS Res Hum Retroviruses 2001; 17:73544. First citation in article
3. Kaul R, Trabattoni D, Bwayo JJ, et al. HIV-1-specific mucosal IgA in a cohort of HIV-1-resistant Kenyan sex workers. AIDS 1999; 13:239. First citation in article
4. Belec L, Ghys PD, Hocini H, et al. Cervicovaginal secretory antibodies to human immunodeficiency virus type 1 (HIV-1) that block viral transcytosis through tight epithelial barriers in highly exposed HIV-1seronegative African women. J Infect Dis 2001; 184:141222. First citation in article
5. Buchacz K, Parekh BS, Padian NS, et al. HIV-specific IgG in cervicovaginal secretions of exposed HIV-uninfected female sexual partners of HIV-infected men. AIDS Res Hum Retroviruses 2001; 17:168993. First citation in article
6. Dorrell L, Hessell AJ, Wang M, et al. Absence of specific mucosal antibody responses in HIV-exposed uninfected sex workers from the Gambia. AIDS 2000; 14:111722. First citation in article
7. Kaul R, Plummer FA, Kimani J, et al. HIV-1-specific mucosal CD8+ lymphocyte responses in the cervix of HIV-1-resistant prostitutes in Nairobi. J Immunol 2000; 164:160211. First citation in article
8. Kaul R, Rowland-Jones SL, Kimani J, et al. New insights into HIV-1 specific cytotoxic T-lymphocyte responses in exposed, persistently seronegative Kenyan sex workers. Immunol Lett 2001; 79:313. First citation in article
9. Geijtenbeek TB, Kwon DS, Torensma R, et al. DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell 2000; 100:58797. First citation in article
10. Jameson B, Baribaud F, Pohlmann S, et al. Expression of DC-SIGN by dendritic cells of intestinal and genital mucosae in humans and rhesus macaques. J Virol 2002; 76:186675. First citation in article
11. Lee B, Leslie G, Soilleux E, et al. cis Expression of DC-SIGN allows for more efficient entry of human and simian immunodeficiency viruses via CD4 and a coreceptor. J Virol 2001; 75:1202838. First citation in article
12. Kwon DS, Gregorio G, Bitton N, Hendrickson WA, Littman DR. DC-SIGN-mediated internalization of HIV is required for trans-enhancement of T cell infection. Immunity 2002; 16:13544. First citation in article
13. McDonald D, Wu L, Bohks SM, KewalRamani VN, Unutmaz D, Hope TJ. Recruitment of HIV and its receptors to dendritic cell-T cell junctions. Science 2003; 300:12957. First citation in article
14. Hu Q, Frank I, Williams V, et al. Blockade of attachment and fusion receptors inhibits HIV-1 infection of human cervical tissue. J Exp Med 2004; 199:106575. First citation in article
15. Masurier C, Salomon B, Guettari N, et al. Dendritic cells route human immunodeficiency virus to lymph nodes after vaginal or intravenous administration to mice. J Virol 1998; 72:78229. First citation in article
16. Seage GR, Holte SE, Metzger D, et al. Are US populations appropriate for trials of human immunodeficiency virus vaccine The HIVNET Vaccine Preparedness Study. Am J Epidemiol 2001; 153:61927. First citation in article
17. Trumpfheller C, Park CG, Finke J, Steinman RM, Granelli-Piperno A. Cell type-dependent retention and transmission of HIV-1 by DC-SIGN. Int Immunol 2003; 15:28998. First citation in article
18. Wu L, Martin TD, Carrington M, KewalRamani VN. Raji B cells, misidentified as THP-1 cells, stimulate DC-SIGN-mediated HIV transmission. Virology 2004; 318:1723. First citation in article
19. Lahm H, Stein S. Characterization of recombinant human interleukin-2 with micromethods. J Chromatogr 1985; 326:357361. First citation in article
20. Gartner S, Markovits P, Markovitz DM, Kaplan MH, Gallo RC, Popovic M. The role of mononuclear phagocytes in HTLV-III/LAV infection. Science 1986; 233:2159. First citation in article
21. Popovic M, Gartner S, Read-Connole E, Beaver B, Reitz M. Cell tropism and expression of HIV-1 isolates in natural targets. Retroviruses of Human AIDS and Related Animal Diseases. 1989:217. First citation in article
22. Gallo RC, Salahuddin SZ, Popovic M, et al. Frequent detection and isolation of cytopathic retroviruses (HTLV-III) from patients with AIDS and at risk for AIDS. Science 1984; 224:5003. First citation in article
23. Shaw GM, Hahn BH, Arya SK, Groopman JE, Gallo RC, Wong-Staal F. Molecular characterization of human T-cell leukemia (lymphotropic) virus type III in the acquired immune deficiency syndrome. Science 1984; 226:116571. First citation in article
24. Foley GE, Lazarus H, Farber S, Uzman BG, Boone BA, McCarthy RE. Continuous culture of human lymphoblasts from peripheral blood of a child with acute leukemia. Cancer 1965; 18:5229. First citation in article
25. Nara PL, Fischinger PJ. Quantitative infectivity assay for HIV-1 and -2. Nature 1988; 332:46970. First citation in article
26. Nara PL, Hatch WC, Dunlop NM, et al. Simple, rapid, quantitative, syncytium-forming microassay for the detection of human immunodeficiency virus neutralizing antibody. AIDS Res Hum Retroviruses 1987; 3:283302. First citation in article
27. Morner A, Bjorndal A, Albert J, et al. Primary human immunodeficiency virus type 2 (HIV-2) isolates, like HIV-1 isolates, frequently use CCR5 but show promiscuity in coreceptor usage. J Virol 1999; 73:23439. First citation in article
28. Mitchell DA, Fadden AJ, Drickamer K. A novel mechanism of carbohydrate recognition by the C-type lectins DC-SIGN and DC-SIGNR: subunit organization and binding to multivalent ligands. J Biol Chem 2001; 276:2893945. First citation in article
29. Oegema TR Jr, Kraft EL, Jourdain GW, Van Valen TR. Phosphorylation of chondroitin sulfate in proteoglycans from the swarm rat chondrosarcoma. J Biol Chem 1984; 259:17206. First citation in article
30. Kabat D, Kozak SL, Wehrly K, Chesebro B. Differences in CD4 dependence for infectivity of laboratory-adapted and primary patient isolates of human immunodeficiency virus type 1. J Virol 1994; 68:25707. First citation in article