The immune milieu of the female genital tract (FGT) is an important determinant of both HIV acquisition and secondary sexual transmission1 and may be altered by a number of local factors, including the indigenous bacterial flora.2 Alterations in vaginal flora constitute a spectrum, from normal to bacterial vaginosis (BV), and this spectrum is characterized by the progressive loss of hydrogen peroxide-producing Lactobacillus species and colonization with a heterogenous flora consisting of Gardnerella vaginalis, Mycoplasma hominis, Ureaplasma species, and various gram-negative anaerobes.3
Genital tract levels of proinflammatory cytokines such as interleukin 1β (IL1β) may be increased in women with both BV and intermediate flora,4 and BV has been associated with increased genital shedding of HIV.5 This might be a direct effect because BV-associated bacteria enhance HIV virus replication in vitro.5 Alternatively, it might be an indirect effect, perhaps, due to the recruitment of activated CD4+ T cells with subsequent increased HIV replication in the genital mucosa or due to interactions between BV and herpes simplex virus type 2 (HSV-2).6,7
We hypothesized that the well-documented association of BV with increased genital HIV shedding might be underpinned by changes in genital immune cell populations, specifically by an increased number of activated CD4+ T cells expressing the HIV coreceptor CCR5 that could enhance local HIV replication. We tested this hypothesis by examining cervical immune cell populations collected by cytobrush in a cohort of Kenyan HIV-infected female sex workers (FSWs) with BV, before and immediately after therapy with oral metronidazole.
Participant Enrollment, Sampling, and Diagnostic Procedures
HIV-infected participants with BV were enrolled through a FSW clinic in Nairobi, Kenya. All clinic attendees completed a behavioral questionnaire and were asked to abstain from sexual intercourse for 24 hours before genital tract sampling. Genital tract sampling and diagnostics were performed as previously described,8 including cervicovaginal lavage, vaginal swabs for Gram stain and Trichomonas vaginalis culture, cervical swab for HIV viral load, endocervical swab for Neisseria gonorrhoeae and Chlamydia trachomatis polymerase chain reaction, and a cervical cytobrush collected into 10 mL of saline. Cervical samples were not collected from women who were menstruating. Nugent scores were reported from Gram stains, with a score of 7-10 defined as BV.9 Serology was performed for syphilis, HIV-1, and HSV-2 as previously described.8 Blood CD4/CD8 T-cell counts were enumerated via flow cytometry. Cervical and blood plasma HIV-1 RNA was measured using the Versant bDNA kit (limit of detection 50 copies/mL; Bayer Diagnostics, Emeryville, CA), and HSV-2 DNA was quantified with the RealArt HSV1/2 LC PCR kit (lower limit of detection 10 copies/mL; Artus Biotech, Hamburg, Germany). All infections were treated according to Kenyan national guidelines. Participants with BV were offered treatment (metronidazole 500 mg twice daily for 7 days) and invited to return within 1 week after completion of treatment. Informed consent was obtained, and the protocol was approved by Research Ethics Boards at the University of Toronto, the University of Manitoba, and Kenyatta National Hospital (Nairobi, Kenya).
Cervical Immune Assays
Cervical samples were filtered through a 100μm filter (Fisher Scientific, Pittsburgh, PA), washed, and divided into 2 equal aliquots for staining with a panel of dendritic cells or T-cell markers.8 Populations were enumerated by flow cytometry (FACSCalibur; Becton-Dickinson Immunocytometry Systems, San Jose, CA).8 Appropriate isotype controls were used in parallel. Cell numbers were multiplied by 2 to determine “cells per cytobrush” and then log10 transformed for analysis and presentation. Cervicovaginal lavage cytokines and chemokines were assayed by cytokine bead array (BD Biosciences, San Diego, CA).8
All analyses were performed with SPSS version 11.0 software (Chicago, IL). The paired samples t test was used to compare immunological changes between paired intraindividual specimens at pre-BV and post-BV treatment visits.
The study population consisted of 15 HIV-infected Kenyan FSWs with BV who had screened negative for classical sexually transmitted infections (C. trachomatis, N. gonorrhoeae, T. vaginalis, or syphilis). All were HSV-2 seropositive, and none had clinically apparent genital ulceration. No participants received an antiherpes medication before or after treatment of BV. Eight participants (53%) were on combination antiretroviral therapy (CART), with an undetectable blood HIV RNA viral load in 6 (75%) and a median CD4 T-cell count of 274 cells per cubic millimeter (range: 123-1204 cells/mm3). In those participants not taking CART, the median blood HIV RNA viral load was 1137 RNA copies per milliliter (range: <50 to 366, 280 RNA copies/mL) and median CD4 count was 498 cells per cubic millimeter (range: 207-593 cells/mm3). The blood viral load exceeded 5000 RNA copies per milliliter in only 1 participant who was therapy naive.
Impact of BV Therapy on Viral Shedding and the Genital Immune Milieu
Oral metronidazole therapy was associated with a reduced Nugent score at follow-up in 14 of 15 participants (93%; mean score: pretherapy 8.1 vs posttherapy 4.5; P < 0.001). HSV-2 shedding was detected in 1 of 15 participants before metronidazole therapy and in 1 (different) participant after treatment. At enrollment, HIV genital shedding was only detected in the single CART-näive individual with a blood viral load exceeding 5000 RNA copies/ml; after metronidazole therapy, no HIV genital shedding was detected.
BV therapy was associated with a significant reduction in the levels of genital proinflammatory cytokines/chemokines (Fig. 1), including IL1b (log10 pg/mL: 2.4-1.7; P = 0.03), IL8 (log10 pg/mL: 3.0-2.4; P = 0.02), and RANTES (log10 pg/mL: 1.4-0.9; P < 0.01). In addition, there was a substantial decline in total cervical CD4+ T-cell numbers after therapy (log10 2.17 vs 1.69; P = 0.006), CD4+ T cells expressing the CD69 activation marker (log10 1.83 vs 1.35; P = 0.005) and the CCR5 HIV coreceptor (log10 1.86 vs 1.32; P = 0.003). Therapy was not associated with any significant changes in the number of cervical CD8+ T cells (log10 2.38 vs 2.27; P = 0.45) or of immature dendritic cells (log10 2.96 vs 3.03; P = 0.66).
Alterations in the normal vaginal flora may be associated with increased HIV susceptibility and secondary transmission. We have confirmed that BV in HIV-infected women was associated with reversible increases in genital levels of proinflammatory cytokines. In addition, for the first time to our knowledge, we define reversible increases in activated CD4+ T cells in the genital mucosa. Unfortunately, our sample size was relatively small, and the majority of participants were taking CART. In addition, genital HIV shedding was infrequent, even in the therapy-naive group, likely because only 1 participant had a blood HIV viral load above 5000 RNA copies per milliliter. This means that our study had a very limited power to directly examine the effects of BV elimination on HIV levels in the genital tract.
Based on these results, we hypothesize that perturbations in microbial flora might increase HIV replication in the genital mucosa via at least 2 mechanisms. First, the BV-associated increases in numbers of cervical CD4+ CCR5+-activated T cells may provide the necessary “fuel” for enhanced mucosal HIV replication, with subsequent increased levels of both cell-free and cell-associated HIV in the genital secretions.10 A second, perhaps related, mechanism is that HIV replication may be driven by the increased levels of IL1β, a proinflammatory cytokine that directly activates HIV long terminal repeat (LTR) and enhances viral replication.11 The fact that BV elimination improved the mucosal immune milieu in the absence of HSV-2 therapy, despite universal HSV-2 coinfection in participants, suggests that the previously described effects of BV on HIV shedding5 are not likely to be mediated through HSV-2.
BV has also been associated with increased HIV susceptibility.12 If the effects of BV on the FGT immune milieu were similar in HIV-infected and HIV-uninfected women, then this would suggest that increased susceptibility might relate to increased cervical CD4+ T cells expressing the CCR5 HIV coreceptor. These activated CD4+ T cells have been shown to be key HIV “target cells” in ex vivo models of HIV transmission.13 However, given the profound impact of HIV infection itself on the FGT immune milieu,1 it will be important to study the immune associations of BV independently in HIV-uninfected women.
Overall, these data demonstrate that BV has a substantial impact on mucosal immune cell populations in the FGT. We postulate that these changes may underpin both the association of BV with increased proinflammatory cytokine levels and the association of BV with increased genital levels of HIV. This provides further rationale for the study of vaginal flora restoration as a potential strategy to reduce the sexual transmission of HIV.
We thank Jane Kamene and the Pumwani clinic nurses; Ann Miangi, Nyakio Chinga, and the laboratory staff at the University of Nairobi, Microbiology Annex; and the women of the Pumwani cohort for their continued support of our studies.
1. Kaul R, Pettengell C, Sheth PM, et al. The genital tract
immune milieu: an important determinant of HIV
susceptibility and secondary transmission
. J Reprod Immunol
2. Schwebke JR. Role of vaginal flora as a barrier to HIV
acquisition. Curr Infect Dis Rep
3. Eschenbach DA, Hillier S, Critchlow C, et al. Diagnosis and clinical manifestations of bacterial vaginosis
. Am J Obstet Gynecol
4. Hedges SR, Barrientes F, Desmond RA, et al. Local and systemic cytokine levels in relation to changes in vaginal flora. J Infect Dis
5. Al-Harthi L, Roebuck KA, Olinger GG, et al. Bacterial vaginosis
-associated microflora isolated from the female genital tract
-1 expression. J Acquir Immune Defic Syndr
6. Cherpes TL, Meyn LA, Krohn MA, et al. Association between acquisition of herpes simplex virus type 2 in women and bacterial vaginosis
. Clin Infect Dis
7. Kaul R, Nagelkerke NJ, Kimani J, et al. Prevalent herpes simplex virus type 2 infection is associated with altered vaginal flora
and an increased susceptibility to multiple sexually transmitted infections. J Infect Dis
8. Rebbapragada A, Wachihi C, Pettengell C, et al. Negative mucosal synergy between herpes simplex type 2 and HIV
in the female genital tract
9. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis
is improved by a standardized method of gram stain interpretation. J Clin Microbiol
10. Haase AT. Perils at mucosal front lines for HIV
and SIV and their hosts. Nat Rev Immunol
11. Spear GT, St John EP, Zariffard MR. Bacterial vaginosis
and human immunodeficiency virus infection. AIDS Res Ther
12. Atashili J, Poole C, Ndumbe PM, Adimora AA, Smith JS. Bacterial vaginosis
acquisition: A meta-analysis of published studies. AIDS
13. Hladik F, Sakchalathorn P, Ballweber L, et al. Initial events in establishing vaginal entry and infection by human immunodeficiency virus type-1. Immunity