Early HIV-1 Target Cells in Human Vaginal and Ectocervical Mucosa


Phillip D. Smith, University of Alabama at Birmingham, 1825 University Boulevard, SHEL 610, Birmingham, AL 35294-2182, USA.
E-mail: pdsmith@uab.edu


Citation Shen R, Richter HE, Smith PD. Early HIV-1 target cells in human vaginal and ectocervical mucosa. Am J Reprod Immunol 2011; 65: 261–267

After translocation through the pleuristratified epithelium of the lower female genital tract, HIV-1 encounters potential target mononuclear cells in the lamina propria of the vagina and ectocervix. Here we show that each major type of genital mononuclear cells, including dendritic cells (DCs), macrophages and lymphocytes, are susceptible to HIV-1 in vitro. Among suspensions of vaginal and ectocervical mononuclear cells, DCs were the first cells to take up virus, containing GFP-tagged virions as early as 15 min after exposure. At 2 hr after exposure, DCs still contained the largest proportion of HIV-1+ cells compared to lamina propria macrophages and lymphocytes from the same mucosal compartment. By 4 days, however, lymphocytes from both vaginal and ectocervical mucosa supported the highest level of HIV-1 replication. Genital macrophages from the same mucosal tissues also were permissive to HIV-1, in sharp contrast to intestinal macrophages, which we have shown previously do not support HIV-1 replication. Thus, among human vaginal and ectocervical mononuclear target cells, DCs are the first to take up HIV-1 and T cells support the most robust viral replication. Further characterization of the parameters of HIV-1 infection in genital mononuclear cells will enhance our understanding of HIV-1 infection in the female genital tract.


Despite remarkable scientific advancements during the past three decades, the AIDS pandemic continues to claim millions of lives, particularly in sub-Saharan Africa. In this region of Africa, more than 22 million people, nearly 60% of whom are women, are infected with HIV-1, causing 1.5 million deaths in 2007 alone.1 In the United States, more than 25% of new HIV-1 infections occur in women, particularly young black and Hispanic women.2 The latter is especially alarming, as the prevalence of HIV-1 infection in some US populations is comparable to that of some sub-Saharan countries. As pointed out by El-Sadr,2 the HIV-1 prevalence of 1 in 30 adults in Washington, DC exceeds the country-wide prevalence in Ethiopia, Nigeria, and Rwanda. Consequently, the development of an effective HIV-1 vaccine and novel agents to prevent HIV-1 transmission remains an urgent and pressing need for both resource-limited nations and the United States.

Worldwide, the heterosexual route is the predominant mode of HIV-1 transmission,3,4 underscoring the need for measures to prevent the sexual transmission of HIV-1. To be effective, such measures must interrupt one or more of the early events in HIV-1 mucosal transmission and infection, including HIV-1 translocation across the epithelial barrier, entry into subepithelial target mononuclear cells, and mucosal and systemic dissemination. Although some antibodies and topically applied microbicides are reportedly capable of preventing HIV-1 infection in model systems,5–9 including simian immunodeficiency virus (SIV) infection in macaques,10–16 successful microbicide intervention in humans has been limited or incomplete.17 In this connection, the recent detection of potent anti-HIV-1 activity in the cervicovaginal fluids of women not infected with HIV-1 18 suggests the possibility that endogenous anti-viral defense mechanisms in the female genital tract could be exploited for protective purposes. However, critical issues pertaining to HIV-1 infection in the female genital mucosa remain unresolved, including characterization of the initial target cells, parameters of local viral replication, pathways of viral dissemination, and the mechanism of R5 selection. Thus, greater understanding of the mechanisms involved in the earliest stages of HIV-1 infection, especially the biological parameters of establishing infection in genital mucosa, must be achieved in order to devise effective strategies for prevention.

HIV-1 entry in the female genital tract

In female genital mucosa, HIV-1 infection involves three major events: (i) entry through the mucosal epithelium; (ii) infection and subsequent replication in subepithelial mononuclear cells; and (iii) delivery to lymph nodes to initiate systemic infection. Entry may occur via the vagina, ectocervix, or endocervix. The epithelial architecture is variable in these regions. The epithelium of the vagina and ectocervix is composed of multi-layered, pluristratified epithelial cells that do not have polarized plasma membranes or tight junctions. In contrast, the epithelium of the endocervix is a single layer of polarized, columnar epithelial cells with a plasma membrane that is separated by tight junctions, dividing the epithelium into apical and basolateral domains. Key features of these two architecturally distinct epithelia relevant to HIV-1 entry include (i) the polarity of the endocervical columnar epithelium, which influences processing and sorting in endosomal transcytosis, and (ii) the ‘leakiness’ of ectocervical and vaginal epithelium, which likely allows CD4+ T cells and dendritic cells (DCs) to migrate into the vaginal epithelium. After migration into the epithelium, DCs extend their processes between the epithelial cells. Thus, distinct translocation processes may participate in HIV-1 entry in different regions of the genital tract.

The extensive surface area of the vagina and ectocervix, estimated to be 15 times greater than that of the endocervix,19 is the likely region in which HIV-1 enters the lower female genital mucosa. Several routes of entry may be involved in the translocation of virus across the non-keratinized squamous epithelium that lines these tissues. Free virus may enter at sites of physical abrasion from microtrauma incurred during intercourse, inflammatory lesions associated with vaginosis and cervicitis, and mucosal disruption from ulcer-inducing infections. Free virus also could penetrate between squamous cells of the stratified epithelial barrier, at least in the upper layer. Although transcytosis of virions through squamous epithelial cells has been suggested,19 classic transcytosis occurs in polarized, columnar epithelial cells rather than non-polarized, pluristratified squamous epithelium.20 The identification of CD4+ T cells in the human vaginal epithelium and documentation of HIV-1 penetration into lymphocytes in the epithelial sheets obtained by suction blister21 suggest that CD4+ T cells also may be involved in HIV-1 entry into vaginal mucosal. Using the macaque model, earlier investigations identified CD4+ T cells in genital mucosa as important, possibly the predominant, target cell for productive SIV) infection.22 In addition to CD4+ T cells, Hladik and colleagues21 showed that HVI-1 rapidly penetrates intraepithelial CD1a+ Langerhans cells, which reside in the genital tract epithelium. Although Langerhans cells do not express DC-SIGN or CCR5, they may participate in early HIV-1 uptake, as shown in macaques inoculated intravaginally with SIV.23 The location of Langerhans cells in the upper layer of the stratified epithelium positions these cells for the uptake of free virions that have penetrated this region of the squamous epithelial barrier.

Dendritic cells also likely contribute to the array of cells potentially involved in HIV-1 entry into the vaginal and ectocervical mucosae. Dendritic cells efficiently capture, disseminate, and transmit virus to mononuclear target cells without productively infecting the DCs themselves, a process termed ‘trans-infection’.5,23–33 Our understanding of DCs in HIV-1 transmission is derived from studies of monocyte-derived DCs, blood DCs, Langerhans cells, and DCs in the SIV/rhesus macaque non-human primate model.23,25–28 However, DCs are a heterogeneous population of cells that display distinct phenotypes and functional profiles in different tissues and mucosal compartments,34,35 precluding simple extrapolation from the fore-mentioned DCs to human mucosal DCs. Due to the limited availability of human vaginal, rectal, and intestinal mucosae, the tissues through which HIV-1 is acquired in the vast majority of infections, and the difficulty in isolating mucosal DCs, the role of mucosal DCs in HIV-1 entry and mucosal spread is poorly understood. Vaginal DCs consist of myeloid DCs, plasmacytoid DCs, and Langerhans cells. Langerhans cells have been shown to take up HIV-1 in human vaginal epithelial sheets, as discussed above,21,36 and in human skin explants37 and epidermal cells isolated from human skin.38 In addition, DC-SIGN+ cells from human rectal mucosa have been shown to bind and transfer HIV-1 to peripheral blood CD4+ T cells,39 and human intestinal DCs also rapidly take up HIV-1, transport the virus through the mucosa, and transmit virions in trans to peripheral blood and intestinal lymphocytes, as we recently showed.40 In contrast, surprisingly little is known about the role of DCs in HIV-1 transmission in the human vagina and ectocervix, but the role of these cells in HIV-1 entry is currently under investigation in several laboratories.

Kinetics of HIV-1 uptake by vaginal and ectocervical mononuclear cells

To begin to elucidate the early events in human cervicovaginal infection by HIV-1, we determined the kinetics of vaginal and ectocervical DC, macrophage and lymphocyte uptake of HIV-1. Vaginal and ectocervical mononuclear leukocytes (MNLs) were isolated using our previously described protocol42 from human vaginal and ectocervical tissues provided by healthy women undergoing reconstructive pelvic surgery in accordance with Institutional Review Board approved protocols. To assure a high level of HIV-1 exposure, suspensions of 5 × 105 MNLs were inoculated with 3.75 × 108 GFP-tagged YU2 viral-like particles (VLPs) and cultured in RPMI plus 10% human AB serum at 37°C. YU2 envelope (Env)-pseudotyped GFP-Gag VLPs were produced by transfection of 293T cells, as previously described.42 The cultures were harvested 2 hr after inoculation and stained with CD13-APC, CD11c-APC, or CD3-PE to identify DCs, macrophages, and lymphocytes that contained HIV-1-GFP viral particles by flow cytometry. As shown in Fig. 1, as early as 15 min after virus inoculation, 5.06% of the CD11c+ DCs contained HIV-1-GFP, which increased to 14.5% at 2 hr. In sharp contrast, vaginal macrophages first showed detectable virion uptake (1.2%) at 30 min post-inoculation, increasing to 1.77% at 2 hr, and vaginal lymphocytes first displayed detectable HIV-1 uptake at 2 hr, when 1.92% of the cells contained virions.

Figure 1.

 Vaginal macrophages, dendritic cells, and lymphocytes uptake HIV-1 in isolated mucosal mononuclear cells. Cultures of mononuclear cells isolated from normal human vaginal tissue were inoculated with GFP-tagged YU2 and incubated at 37°C for 2 hr. Cells then were analyzed by flow cytometry using anti-CD13, anti-CD11c, or anti-CD3 antibodies. Results are representative of cells isolated from three separate donors.

Cultures of ectocervical MNLs were inoculated in parallel to the vaginal MNLs and also examined for uptake of GFP-tagged viral particles by flow cytometry. Again, DCs took up virions at 15 min (1.69%), increasing to 5.67% at 2 hr, whereas macrophages and lymphocytes first displayed detectable HIV-1 uptake at 2 hr, the proportion of macrophages increasing from 0.91% to 4.35% and lymphocytes from 0.41% to 1.56%, respectively (Fig. 2). Thus, among vaginal and ectocervical MNLs, DCs were the first cells to take up HIV-1, containing virus at 15 min. In sharp contrast, vaginal and ectocervical macrophages first contained detectable HIV-1 at 30 min and 2 hr, respectively. Vaginal macrophages and lymphocytes contained similar proportions of infected cells at 2 hr, whereas the proportion of infected ectocervical macrophages was twofold more than that of infected lymphocytes.

Figure 2.

 Ectocervical macrophages, dendritic cells, and lymphocytes uptake HIV-1 in isolated mucosal mononuclear cells. Ectocervical mononuclear cells were isolated from normal human ectocervix tissue, exposed to GFP-tagged YU2, and analyzed at 2 hr post-exposure by flow cytometry using anti-CD13, anti-CD11c, or anti-CD3 antibodies. Results are representative of cells isolated from two separate donors.

Vaginal and ectocervical macrophages and lymphocytes support HIV-1 replication

The female genital tract mucosa is a site of HIV-1 entry but replication in vaginal and ectocervical mucosal cells has been difficult to document. Zhang et al.22 showed that in sexually transmitted SIV infection and in early and later stages of HIV-1 infection, the predominant target cells in the genital tract are resting and activated CD4+ T cells. Using an organ culture system, Gupta et al.41 corroborated these results by showing that memory CD4+ T cells are the earliest HIV-1-infected cells in the cervix. In contrast to these findings, Greenhead et al.42 identified macrophages by immunohistochemical analysis as the dominant HIV-1 target cell in the vaginal lamina propria. As noted above, CD4+ T cells may participate in HIV-1 transport across the epithelium. We have reported the presence of low numbers of CD3+ lymphocytes in the basal region and bordering the dermal papillae in the epithelium and scattered throughout the lamina propria of non-inflamed human vaginal mucosa.43. These cells supported HIV-1 replication, as did the vaginal macrophages.43 We next extended these findings by comparing HIV-1 replication in human vaginal and ectocervical macrophages and lymphocytes.

To determine whether HIV-1 uptake was associated with viral replication, suspensions of vaginal and ectocervical MNLs were inoculated with infectious R5 (YU2) HIV-1 (MOI = 1), cultured under standard conditions, and on day 4 culture supernatants were harvested and analyzed for p24 by ELISA. As shown in Fig. 3a, both vaginal and ectocervical MNLs released p24, confirming productive viral infection. To identify the cells that supported HIV-1 replication, suspensions of vaginal and ectocervical MNLs were inoculated with GFP-expressing YU2 (MOI = 1), cultured for 4 days, and analyzed by fluorescence-activated cell sorter. As shown in Fig. 3b, vaginal as well as ectocervical macrophages and lymphocytes supported HIV-1 replication. Macrophages from vaginal and ectocervical mucosa were similar in their capacity to support infection, as were lymphocytes from vaginal and ectocervical mucosa. However, when lymphocytes were compared to macrophages from the same mucosal compartment, lymphocytes supported more robust replication than the macrophages.

Figure 3.

 Vaginal and ectocervical macrophages and lymphocytes support HIV-1 replication. (a) Vaginal and ectocervical MNLs were inoculated with YU2 and 4 days later culture supernatants were harvested and analyzed for p24. (b) Parallel cultures were inoculated with YU-2env pseudotyped GFP-expressing HIV-1, and 4 days post-infection cells were analyzed by flow cytometry to identify the cells that support HIV-1 replication.


HIV-1 infection of the female genital tract mucosa involves translocation of virus across the epithelium, infection and replication in subepithelial mononuclear cells, and local and systemic dissemination. After translocation across the epithelium by CD4+ T cells, Langerhans cells and/or DCs, virus encounters mononuclear target cells in the subepithelial tissue. Vaginal and ectocervical DCs appear to be the first cells to take up virus and do so more efficiently than macrophages or lymphocytes within the first 2 hr of virus exposure. Interestingly, by 4 days after infection, lymphocytes support more robust viral replication than macrophages in both mucosal compartments. Further characterization of the early events in HIV-1 infection of vaginal and ectocervical mucosa should provide important biological insights for devising effective strategies to prevent genital transmission of HIV-1.


This study was supported by the National Institutes of Health (DK-54495, DK-84063, AI-83027, AI-83539, DK-47322, RR-20136, and the Mucosal HIV and Immunobiology Center, DK-64400) and the Research Service of the Veterans Administration.