Female Genital Tract Host Factors and Tenofovir and Lamivudine Active Metabolites Alyssa M. Lantz,1, Flavia Kiweewa Matovu,2,3 Reilly Johnson,1 Esther Isingel,2 Rita Nakalega,2 Samuel Kabwigu,2 Mags E. Beksinska,4 and Melanie R. Nicol1, 1Department of Experimental and Clinical Pharmacology, College of Pharmacy, University of Minnesota, Minneapolis; 2Makerere University–John Hopkins University Research Collaboration, Kampala Uganda; 3College of Health Sciences, Makerere University, Kampala, Uganda; and 4Wits MatCH Research Unit, Department of Obstetrics and Gynecology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa Background. We previously reported the effect of contraception on cervical tenofovir concentrations in Ugandan women with human immunodeficiency virus (HIV). Here we explored the role of cervicovaginal cytokines and drug metabolizing enzymes and transporters (DMETs) to elucidate female genital tract (FGT) drug disposition in a Ugandan cohort. Methods. Cervicovaginal fluid and cervical biopsies were collected from Ugandan women with HIV receiving tenofovir/ lamivudine-based therapy and intramuscular depot medroxyprogesterone acetate (n = 25), copper intrauterine device (cuIUD; n = 12), or condoms (n = 13) as contraception. Cytokines were measured in cervicovaginal fluid (CVF). Ectocervical tenofovir diphosphate (TFVdp), lamivudine triphosphate (3TCtp), and deoxyadenosine triphosphate (dATP)/deoxycytidine triphosphate (dCTP) concentrations and immune marker/DMET gene expression were measured in cervical biopsies. Results. Cervical 3TCtp was not correlated with any CVF cytokines. Cervical TFVdp was correlated with IL-10, IL-7, and IL-17 in CVF. CCR5 mRNA expression in cervical biopsies was higher in cuIUD users versus condom users. Using multivariable linear regression, CVF IL-17, tissue dATP, plasma estradiol, and plasma tenofovir were all significant predictors of cervical TFVdp. Tissue dCTP and plasma lamivudine were significant predictors of cervical 3TCtp. Conclusions. TFVdp concentrations in cervix appear to be influenced by local inflammation. In contrast, 3TCtp FGT exposure was not affected by genital inflammation or DMETs. CuIUD users have more immune cells present, which may in turn influence local TFVdp disposition. Main Finding. We investigated changes in tenofovir diphosphate and lamivudine triphosphate due to the microbiome and inflammation. While lamivudine triphosphate was not affected by either, tenofovir diphosphate appeared to be affected by local inflammation. Specifically, Th17 cells may influence tenofovir disposition. Keywords. tenofovir; lamivudine; contraception; inflammation; female genital tract. Received 13 February 2024; editorial decision 18 July 2024; accepted 22 July 2024; published online 24 July 2024 Correspondence: Melanie R. Nicol, PharmD, PhD, Department of Experimental and Clinical Pharmacology, College of Pharmacy, University of Minnesota, 2001 6th St SE, MTRF 4-210, Minneapolis, MN 55455 (mrnicol@umn.edu). The Journal of Infectious Diseases® 2024;230:1434–43 © The Author(s) 2024. Published by Oxford University Press on behalf of Infectious Diseases Society of America. All rights reserved. For commercial re-use, please contact reprints@ oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com. https://doi.org/10.1093/infdis/jiae372 Cisgender women make up half of people with human immu- nodeficiency virus (HIV) globally and yet the factors that reg- ulate drug efficacy in the female genital tract (FGT) are understudied, which has implications for antiretrovirals used as preexposure prophylaxis (PrEP) and treatment (ie, reduce viral shedding in women with HIV). Modeling studies have shown minimal pharmacologic forgiveness in the FGT with 6–7 doses a week required to reach protective concentrations [1]. It is critical to understand factors affecting antiretroviral disposition in the FGT to optimize dosing strategies and ad- vance development of new prevention options. Mucosal antiretroviral pharmacology is complex and in- volves many components of the FGT microenvironment, which may play a role in efficacy of PrEP. Some of these com- ponents include vaginal microbiota, inflammation, and drug metabolizing enzymes and transporters (DMETs). In addition, both endogenous and exogenous hormones regulate the FGT microenvironment and may play a role in efficacy. A healthy vaginal microbiome is mainly comprised of Lactobacillus species that release lactic acid to keep the pH low [2]. Vaginal dysbiosis, clinically known as bacterial vagino- sis, is when there is an increase in microbiota diversity. Bacterial vaginosis in women of reproductive age is associated with the presence of different genera including Prevotella, Sneathia, Dialister, Megasphaera, and Gardnerella [3]. Vaginal dysbiosis has been associated with a decrease in efficacy of topical tenofovir. In a retrospective analysis of Centre for the AIDS Programme of Research in South Africa 004, women with a Lactobacillus-dominated vaginal microbiota had 3-fold higher protection from HIV acquisition using 1% tenofovir gel than women with non-Lactobacillus-dominated vaginal mi- crobiota [4]. There is less known how the vaginal microbiome 1434 • JID 2024:230 (15 December) • Lantz et al The Journal of Infectious Diseases M A J O R A R T I C L E D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 https://orcid.org/0000-0001-5231-0009 https://orcid.org/0000-0001-5513-5509 mailto:mrnicol@umn.edu https://doi.org/10.1093/infdis/jiae372 affects antiretroviral exposure in the FGT. Carlson et al showed that microbiome community types with intermediate diversity had the highest concentration of tenofovir in cervicovaginal fluid, but tenofovir concentrations were similar between high- diversity community types and low-diversity community types [5]. One potential mechanism for decreased tenofovir diphos- phate (TFVdp) is that when the diversity of the vaginal micro- biome increases, there is a probability of increased abundance of proinflammatory taxa, which leads to an increase in genital inflam- mation. Indeed, McKinnon et al found the efficacy of 1% tenofovir gel was significantly reduced in the presence of genital inflam- mation, even in participants with high adherence (>50%) [6]. While the Evidence for Contraceptive Options and HIV Outcomes (ECHO) trial showed no increased HIV risk with 3 methods of contraception, a substudy showed an association be- tween copper intrauterine device (IUD), increased bacterial di- versity, and increased inflammation after 6 months of use [7, 8]. We previously reported, in a observational pharmacokinetic study in 50 Ugandan women with HIV on an antiretroviral reg- imen containing tenofovir, lamivudine, and efavirenz, that cop- per IUD users had lower TFVdp concentrations compared to intramuscular depo-medroxyprogesterone acetate (DMPA-IM) or condom users [9]. We hypothesize that copper IUDs may alter the microbiome and thus create an inflammatory environment in the FGT, causing low TFVdp concentrations. Therefore, we sought to further explore how different components in the FGT influence mucosal pharmacology of TFVdp, lamivudine tri- phosphate (3TCtp), and the endogenous nucleotides deoxyade- nosine triphosphate (dATP) and deoxycytidine triphosphate (dCTP). We conducted an analysis on secondary endpoints, ex- amining the effect of the vaginal microbiome and genital inflam- mation on the active metabolites in cervical tissues by investigating the presence and quantity of cytokines, drug metab- olizing enzymes, transporters, and vaginal microbiota in the FGT. METHODS Study Population The study procedures have been previously described and prima- ry results published [9]. The Uganda Virus Research Institute, the Human Research Ethics Committee of the University of the Witwatersrand, and the Uganda National Council of Science and Technology reviewed and approved the procedures for the study (NCT03377608). The study population was a subset of women who participated in the BONE: Contraception and Anti-Retroviral Effects (BONE- CARE) study [10]. All women were on an antiretroviral therapy regimen that consisted of teno- fovir disoproxil fumarate (TDF), lamivudine, and efavirenz. Women who were enrolled in the BONE-CARE study, stable on their TDF-containing regimen, and virally suppressed (plasma HIV RNA <50 copies/mL) for ≥6 months were offered participa- tion in this substudy. Exclusion criteria included pregnancy, breastfeeding, symptomatic vaginal infection, abnormal vaginal bleeding, history of genital dysplasia or human papillomavirus, or use of oral/vaginal antibiotics or antifungals within 30 days (exception made for sulfamethoxazole-trimethoprim). Written informed consent was obtained prior to study participation [9]. Sample Collection All samples were collected at a single visit and are detailed previ- ously [9]. Relevant to this analysis, 1 vaginal swab for microbiome sequencing and 1 cervical swab for cytokine quantification were collected using polyester swabs. Two cervical biopsies were col- lected using Baby Tischler Biopsy forceps (McKesson) and blood was collected in ethylenediaminetetraacetic acid tubes. Cytokine Quantification Cytokine extraction from cervicovaginal swabs was performed similar to methods previously described [11]. Cervical swabs were centrifuged in 500 μL of phosphate-buffered saline (PBS) and then scraped on the side of the tube. Concentrations of 27 cytokines were measured using the 27-plex Human Cytokine Assay (Bio-Rad) and the MAGPIX Assay Reader. The 27-plex Human Cytokine Assay was chosen based previous literature and the targets being relevant to HIV [8, 12, 13]. Antiretroviral Quantification A detailed description of antiretroviral quantification was pre- viously published [9]. In brief, tenofovir and lamivudine were quantified in blood using a validated assay that had a lower lim- it of 1 ng/mL. TFVdp, 3TCtp, dATP, and dCTP were quantified in cervical tissue with a dynamic range of 0.02–20 ng/mL and peripheral blood mononuclear cells (PBMCs), with a dynamic range of 0.02–100 ng/mL for dATP and TFV-DP, 0.2–600 ng/mL for 3TCtp, and 0.06–100 ng/mL for dCTP. Microbiome Sequencing The vaginal microbiome was analyzed using 16S sequencing as described previously [9]. In brief, DNA was extracted from vagi- nal swabs using the DNeasy Powersoil Kit. The V4 hypervariable region of the 16S ribosomal RNA gene was used for amplification. Western Blot Analysis Nine of the 50 snap-frozen cervical biopsies were used for pro- tein expression with Western blot, while the remaining 41 were used for gene expression. Cervical biopsies for Western blot were chosen based on TFVdp concentrations, selecting samples with low, medium, and high cervical TFVdp. Human liver was collected postmortem as previously described [14] and was used as a positive control. Total proteins from snap-frozen cervical tissue biopsies and liver were lysed using radioimmu- noprecipitation assay (RIPA) buffer (Thermo Scientific). A Bradford assay (Sigma) was used to quantify total protein con- centration. Forty micrograms of protein was separated using Modulators of Drugs in Female Genital Tissue • JID 2024:230 (15 December) • 1435 D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 10% sodium dodecyl sulphate–polyacrylamide gel electrophore- sis and then transferred to a polyvinylidene difluoride membrane. Membranes were incubated overnight with RAR-related orphan receptor gamma (ROR-γ) (Bioss, 1:500, BS-6217R), CD4 (Sigma-Aldrich, 1:500, 104R-1), or glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Bioss, 1:5000, BS-10900R) antibodies. Membranes were then incubated with horseradish peroxidase– conjugated secondary antibody (Prometheus, 1:4000) for 1 hour. Protein bands were visualized using Western Lighting Plus-ECL reagents and imaged on iBright imaging system (Thermo Fisher Scientific). CCR5, CD4, AK2, and MRP4 Expression Total RNA was extracted from snap-frozen cervical biopsies us- ing the RNeasy Fibrous Tissue Kit (Qiagen). Tissues were ho- mogenized in lysis buffer and underwent a 10-minute digestion in Proteinase K prior to RNA isolation and purifica- tion. RNA was quantified and tested for purity using a NanoDrop. First-strand cDNA synthesis was performed from total RNA using Superscript1 Vilo IV cDNA Synthesis kit (Life Technologies, Grand Island, New York) and messenger RNA (mRNA) expression was quantified using quantitative po- lymerase chain reaction (qPCR) with TaqMan Gene Expression Assays. Genes were normalized to GAPDH and relative expres- sion was calculated as 2−ΔCt. CCR5 and CD4 genes were chosen because CD4+ T cells are the target cells of HIV and CCR5 is a coreceptor that facilitates entry into the cell for R5 tropic vari- ants, the most common variant to be transmitted through sex- ual contact [15]. AK2 is a kinase that phosphorylates TFVdp, and we wanted to explore if the copper IUD decreased AK2 ex- pression, thus lowering TFVdp concentrations compared to DMPA-IM and condom users. MRP4 is a transporter that has been implicated in tenofovir disposition [16, 17]. Statistical Analysis Kruskal-Wallis tests were performed to test differences in gene expression between DMPA-IM users, copper IUD users, and condom users. Differences in cytokine concentrations between DMPA-IM users, copper IUD users, and condom users were tested by Dunn test. All statistical comparisons were corrected for multiple comparisons using the Bonferroni method. Correlations with drug active metabolites, endogenous nucleo- tides, and the ratio of metabolite to related nucleotide were assessed using Spearman correlations, specifically looking at correlations with cytokines, bacterial relative abundance, and gene expression. Bacteria taxa with >1% relative abundance in vaginal swabs, which includes Lactobacillus, Prevotella, Dialister, Sneathia, Gardnerella, and Megasphaera were included in the analysis. Associations between antiretrovirals and the mi- crobiome have already been reported [9]. Here, linear regression was used to identify interactions between cytokines and the mi- crobiome when predicting antiretroviral concentrations. Western blot results for ROR-γ and CD4 were quantified us- ing Image J software. To further investigate our hypothesis of the involvement of Th17 cells, we compared ROR-γ and CD4 expression in tissue by TFVdp cervical concentrations separat- ed into tertiles: T1, 1544–10 825 fmol/g; T2, 10 286–18 175 fmol/g; and T3, ≥18 176 fmol/g. Spearman correlations were used to compare interleukin (IL) 17 and CD4/ROR-γ. Stepwise regression in both directions was performed to identify predictors of TFVdp and 3TCtp in cervical tissue. Covariates tested in the models were age, CD4 cervical expres- sion, cervicovaginal IL-17, CCR5 cervical expression, proges- terone, estradiol, respective parent drug in plasma, and respective nucleotide in tissue. Since all cytokines measured were correlated, only 1 cytokine was included in the regression model, which was determined by the best fit in single linear re- gression model with cervical TFVdp. All models were tested us- ing model diagnostics and the final model was the best fit. All statistical analyses were performed in R Studio. RESULTS Participant Characteristics Supplementary Table 1 describes the participant characteristics for the exploratory analysis. Fifty women with HIV were en- rolled for this study. Twenty-five women were on DMPA-IM, 12 had a copper IUD, and 13 were using condoms for primary contraceptive methods. The mean age was 26.2 years (range, 23.9–30.7 years). Two women using nonhormonal contracep- tion and 2 women using DMPA-IM tested positive for gonor- rhea. Three women using nonhormonal contraception tested positive for syphilis. Cervical Cytokine Concentrations and Relative Bacterial Abundance by Contraception Method Condom users had 65% higher granulocyte macrophage– colony-stimulating factor than DMPA-IM (P = .023) and it was 47% higher than copper IUD users (P = .033). RANTES (CCR5-binding chemokine) was 113% lower in copper IUD us- ers compared to DMPA-IM users but was not significant (P = .06). While RANTES was 62% lower in copper IUD users compared to condom users, it was not significant (P = .33). The rest of the cytokines measured were not significantly different between contraceptive methods. Associations Between Cytokines, Antiretrovirals, and the Microbiome 3TCtp was not correlated with any cytokines measured. TFVdp was positively correlated with 3 of the 27 cytokines. Specifically, TFVdp was positively correlated with IL-7 (R = 0.35, P = .016), IL-10 (R = 0.36, P = .01), and IL-17 (R = 0.48, P = .00075) (Figure1A–C). Dialister, Megasphaera, and Prevotella were negatively corre- lated with TFVdp:dATP ratio (Table 1). Prevotella was positive- ly correlated with dATP in tissue, but not correlated with 1436 • JID 2024:230 (15 December) • Lantz et al D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 http://academic.oup.com/jid/article-lookup/doi/10.1093/infdis/jiae372#supplementary-data TFVdp. Dialister trended to have a positive correlation with dATP in cervical tissue, but it was not significant (R = 0.25, P = .092). Mycoplasma was negatively correlated with 3TCtp: dCTP ratio (R = 0.35; P = .017). Both cervical concentrations of 3TCtp and dCTP separately were positively correlated with Lactobacillus (P = .032, P = .044, respectively; Table 2). Two of the 6 most prevalent bacteria were correlated with proinflammatory cytokines (Table 3). Sneathia was positively correlated with 4 proinflammatory cytokines: IL-1β, tumor necrosis factor alpha (TNF-α), IL-8, and IL-17. Lactobacillus was negatively correlated with IL-1β and TNF-α. Shuttleworthia was positively correlated with IL-1β. Gardnerella, Prevotella, and Megasphaera were not correlated with any of the 27 cytokines measured. Gene Expression and ROR-γ in Cervical Tissues Women on copper IUDs had significantly higher gene expres- sion of CCR5 (Figure 2B) and AK2 (Figure 2C) compared to Figure 1. Cytokines were measured in cervicovaginal fluid and tenofovir diphosphate (TFVdp) was measured in cervical biopsies. Correlations between cervical TFVdp concentrations (fmol/g) and all 27 cytokines measured were analyzed using Spearman correlations. Of the 27 measured, only interleukin 7 (IL-7; A), interleukin 10 (IL-10; B), and interleukin 17 (IL-17; C ) were significantly correlated with TFVdp. Modulators of Drugs in Female Genital Tissue • JID 2024:230 (15 December) • 1437 D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 condom users, with CD4 trending to be higher although not significant (Figure 2A). There was no differences in relative mRNA for CD4, CCR5, AK2, or MRP4 between women on DMPA-IM and copper IUD users. There were no correlations between the 4 genes measured and TFVdp, 3TCtp, dATP, dCTP, and any cytokines measured. ROR-γ and CD4 were quantified in 9 cervical biopsies using Western blot to confirm Th17 cell presence in cervical tissue. Neither CD4 nor ROR-γ was correlated with IL-17 (R = −0.37, P = .34 and R = −0.067, P = .067, respectively). Both ROR-γ and CD4 had the highest mean expression in TFVdp tertile 1, which contrasts with our hypothesis that ROR-γ and CD4 would be higher in tertile 3 (Figure 3A and 3B). Stepwise Regression Table 4 includes the data from stepwise regression for both TFVdp and 3TCtp. Cervical TFVdp was significantly predicted by IL-17 (P < .01), dATP in cervical tissue (P < .05), tenofovir in plasma (P < .05), and estradiol (P < .05). Cervical 3TCtp was predicted by dCTP in cervical tissue (P < .01) and lamivu- dine in plasma (P < .01). Other covariates tested were age, CD4 expression, MRP4 expression, and progesterone. DISCUSSION Understanding how the genital tract environment affects anti- retroviral concentrations in mucosal tissues is critical for im- proving antiretroviral efficacy in the FGT. One objective of this study was to further explore the observation that copper IUD users had the highest concentration of cervical TFVdp, which could be due to alterations in the inflammatory milieu or the vaginal microbiome. We did not find an increase in Table 1. Correlations Between Cervical Tenofovir Diphosphate, Deoxyadenosine Triphosphate, Their Ratio, and Bacteria Genera Measured Using 16S Microbiome Sequencing Bacteria R P Value TFVdp:dATP ratio in ectocervix Dialister −0.4 .0063** Prevotella −0.34 .021* Megasphaera −0.39 .0076** Sneathia −0.13 .38 Mycoplasma 0.15 .39 Gardnerella −0.069 .65 Lactobacillus 0.16 .27 dATP in ectocervix Dialister 0.22 .13 Prevotella 0.3 .04* Megasphaera 0.075 .62 Sneathia 0.027 .86 Mycoplasma 0.15 .32 Gardnerella 0.023 .88 Lactobacillus 0.073 .63 TFVdp in ectocervix Dialister −0.16 .28 Prevotella −0.1 .5 Megasphaera −0.28 .056 Sneathia −0.02 .89 Mycoplasma 0.0033 .83 Gardnerella 0.011 .94 Lactobacillus 0.11 .45 Abbreviations: dATP, endogenous deoxyadenosine triphosphate; TFVdp, tenofovir diphosphate. *P < .05. **P < .01. Table 2. Correlations Between Cervical Lamivudine Triphosphate, Deoxycytidine Triphosphate, Their Ratio, and Bacteria Genera Measured Through 16S Microbiome Sequencing Bacteria R P Value 3TCtp:dCTP ratio in ectocervix Dialister −0.25 .09 Prevotella −0.25 .09 Megasphaera −0.12 .42 Sneathia −0.28 .06 Mycoplasma −0.35 .017* Gardnerella −0.12 .44 Lactobacillus 0.25 .098 dCTP in ectocervix Dialister −0.065 .67 Prevotella −0.004 .98 Megasphaera −0.11 .45 Sneathia −0.1 .5 Mycoplasma 0.0058 .97 Gardnerella −0.021 .89 Lactobacillus 0.3 .044* 3TCtp in ectocervix Dialister −0.13 .41 Prevotella −0.12 .41 Megasphaera −0.059 .7 Sneathia −0.13 .39 Mycoplasma −0.11 .45 Gardnerella 0.005 .97 Lactobacillus 0.32 .032* Abbreviations: 3TCtp, lamivudine triphosphate; dCTP, endogenous deoxycytidine triphosphate. *P < .05. Table 3. Significant Spearman Correlations Between Genera of Vaginal Microbiome and Cytokines Measured in Cervicovaginal Fluid Cytokine R P Value Sneathia IL-1β 0.31 .037 TNF-α 0.35 .018 IL-8 0.38 .0088 IL-17 0.33 .024 Lactobacillus IL-1β −0.39 .0078 TNF-α −0.31 .036 Abbreviations: IL-1β, interleukin 1-beta; IL-8, interleukin 8; IL-17, interleukin 17; TNF-α, tumor necrosis factor alpha. 1438 • JID 2024:230 (15 December) • Lantz et al D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 cytokine concentrations for copper IUD users compared to DMPA-IM or condom users, which contrasts with findings from the ECHO trial [8], where the copper IUD was associated with increased cytokine concentrations, including IL-1β, IL-2, IL-17, and TNF-α. However, mean duration with copper IUD in our study was 29.6 months (range, 2–84 months), which is longer than the 6 months in the ECHO trial substudy [7]. It is possible that the difference in cytokine concentrations seen in this study is due to having more time to heal from epithelial damage from IUD insertion, thus decreasing the immune re- sponse to the IUD. Furthermore, this was not a longitudinal study, so we were unable to see changes over time in cytokine concentrations. It is also possible that a change in pH because of the copper IUD alters tenofovir disposition as it is known that tenofovir uptake into cells is decreased as pH increases [18]. Further research into the mechanisms related to lower TFVdp in women with copper IUD is needed. We saw higher AK2 expression in women on copper IUDs compared to women using condoms. This was unexpected giv- en copper IUD users had lower TFVdp compared to women on DMPA-IM or using condoms. Several kinases, including AK2, are critical for tenofovir to become pharmacologically active in PBMCs and vaginal tissue [19]. It is possible that other kinases including pyruvate kinase isozymes (muscle or liver and red blood cell) are decreased instead of AK2. Additionally, we only measured gene expression, not protein, so it is possible that there is lower AK2 protein. Cervical TFVdp was positively correlated with IL-7, IL-10, and IL-17. One previous study found that incubation of IL-7 with cervical explants followed by HIV challenge had a higher Figure 2. RNA was extracted from cervical biopsies and gene expression was quantified using quantitative polymerase chain reaction. CD4 (A), CCR5 (B), AK2 (C ), and MRP4 (D) were normalized to glyceraldehyde 3-phosphate dehydrogenase, a housekeeping gene, and Kruskal-Wallis tests were run to compare expression between 3 con- traception methods: copper intrauterine device (IUD; most left plot), intramuscular depot medroxyprogesterone acetate (Depo; middle plot), and condoms (most right plot). **P < .01. Modulators of Drugs in Female Genital Tissue • JID 2024:230 (15 December) • 1439 D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 number of CD4 cells in the tissue than without IL-7 present, suggesting that IL-7 prevents CD4 apoptosis and induced CD4 proliferation [20]. One possible mechanism for IL-7 being correlated with TFVdp is an increased number of CD4 cells in cervical tissue that phosphorylate tenofovir. A higher number of CD4 T cells present in cervix could lead to a higher produc- tion of TFVdp. In contrast, cervical 3TCtp was not correlated with any of the 27 cytokines measured, nor was it correlated with any of the genes measured, although AK2 and MRP4 were not expected to be associated with 3TC disposition. Cervical 3TCtp was pos- itively correlated with Lactobacillus, but not correlated with any of the other bacterial genera measured. Lactobacillus dominates a “normal,” less diverse microbiome and is associated with a healthy microbiome environment [3]. This suggests that lami- vudine distribution is not negatively affected by genital inflam- mation, nonoptimal vaginal microbiota, or DMETs in the FGT. We also showed that IL-17 was a significant predictor of cer- vical TFVdp in a stepwise regression model. One of the major T-cell subsets in the FGT is T-helper type 17 (Th17) cells. These cells are produced for an immune response against extracellular bacteria/fungus and help maintain mucosal barrier integrity [21, 22]. ROR-γ is a main transcription factor in this cell pop- ulation, and we hypothesized that Th17 cells were the reason for high TFVdp [23]. However, in our samples, expression of ROR-γ was highest when TFVdp concentrations were in T1 and ROR-γ was not correlated with IL-17. This suggests that Th17 cells are not playing a role in the distribution of TFVdp and that other immune cells may be involved. Th17 cells activation and expression are influenced by many factors though, including DMPA-IM initiation and the micro- biome [24, 25]. DMPA-IM has been shown to increase Th17 cell frequency in cervical tissue, but this has only been looked at 1 month postinitiation. In addition, we showed that the Sneathia genus was positively correlated with IL-17 and Lactobacillus negatively correlated with IL-1β and TNF-α, sug- gesting an increase in an immune response to increased diver- sity and proinflammatory bacteria. We did not see a correlation between Th17 and AK2 and we did not measure ROR-γ and AK2 in the same participants as separate biopsies were selected Figure 3. Nine of 50 cervical biopsies were homogenized in radioimmunoprecipitation assay buffer, and expression of RAR-related orphan receptor gamma (A) and CD4 (B) was measured using Western blot. Expression was graphed by tenofovir diphosphate concentration tertiles. The 3 tertiles were divided as follows: T1 (1544–10 285 fmol/g), T2 (10 286–18 175 fmol/g), and T3 (≥18 176 fmol/g). No statistical analyses were done due to the small sample size. Abbreviations: GAPDH, glyceraldehyde 3-phosphate dehydrogenase; ROR-γ, RAR-related orphan receptor gamma; TFVdp, tenofovir diphosphate. Table 4. Final Stepwise Regression Model for Predicting Tenofovir Diphosphate and Lamivudine Triphosphate in Cervical Tissue Covariate Estimate Standard Error TFVdp Log(dATP in cervical tissue)* 0.417 0.158 Log(IL-17)** 0.262 0.061 Estradiol* −0.002 0.001 Age −0.017 0.01 Log(TFV in plasma)* 0.36 0.136 3TCtp Log(dCTP in cervical tissue)** 0.799 0.1 Log(3TC in plasma)** 0.34 0.141 Significant: *P < .05, **P < .01. Abbreviations: 3TC, lamivudine; 3TCtp, lamivudine triphosphate; dATP, deoxyadenosine triphosphate; dCTP, deoxycytidine triphosphate; IL-17, interleukin 17; TFV, tenofovir; TFVdp, tenofovir diphosphate. 1440 • JID 2024:230 (15 December) • Lantz et al D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 for protein versus expression. Other innate immune cells can also produce IL-17, including innate lymphoid cells and natural killer cells [26]. It is possible that other cells are responsible for elevated IL-17 in our population. These data suggest that im- mune cells may be involved either in tissue distribution of TFVdp or conversion of tenofovir to its active metabolite due to an increased immune response to proinflammatory bacteria, but future confirmatory studies are needed to identify cell type(s) that are involved. Estradiol was also a significant predictor of cervical tenofovir in plasma. This contradicts with our previous findings [9]. However, it was a weak association (coefficient = −0.002) and while it was statistically significant, it may not be clinically rel- evant. Moreover, the addition of cytokines, specifically IL-17, as a covariate in the stepwise regression could also alter the results of other covariates. Finally, the Prevotella genus was negatively correlated with TFVdp:dATP ratio and positively correlated with dATP cervical tissue concentrations (Table 1). This sug- gests that Prevotella genus may be increasing endogenous dATP, thus decreasing tenofovir efficacy. Both 3TCtp and dCTP were positively correlated with Lactobacillus genus, but Lactobacillus was not correlated with the 3TCtp:dCTP ratio (Table 2). This may suggest that while dCTP is increasing, it does not affect lamivudine efficacy. The results of this study are limited by a small sample size, especially for subanalyses. This study only had 1 point in time, so further investigations of how these findings translate over time are needed. We only measured ROR-γ and CD4 pro- tein expression in a small subset of our participants (9/50) to elucidate if the correlation between TFVdp and IL-17 was due to the presence of Th17 cells. In addition, the ROR-γ anti- body does not distinguish between isoforms. The study population was women with HIV. Differences in vaginal microbiota between women with and without HIV may limit the translatability for prevention in the HIV-negative population. The community state type of the vaginal microbiome that has anaerobic bacteria associated with bacterial vaginosis is believed to be more prevalent in women with HIV but does not seem to be statistically different [27–29]. In addition, it is possible that TFVdp tissue pharmaco- kinetics in the FGT is different between women with and those without HIV. However, the median peak tenofovir concentra- tions in plasma after 15 days in women with and women with- out HIV were 306 and 357 ng/mL, respectively [30, 31]. This suggests that there seems to be no difference by HIV status, but further work to evaluate the hypotheses generated from this study in women without HIV is needed. In conclusion, 3TCtp is likely not affected by the local micro- environment of the FGT. TFVdp tissue concentrations, on the other hand, seem to be regulated by multiple components of the FGT. Most importantly, immune cells may play a role in teno- fovir tissue distribution and phosphorylation, although we did not find a correlation between Th17 cells and AK2. These data will ultimately help develop hypotheses to improve antiretrovi- ral therapies targeted to FGT mucosa. In addition, these data highlight the importance of the microenvironment in tenofovir distribution. This highlights the importance of involving the whole system in further studies of tenofovir for PrEP, either in a preclinical animal model or an ex vivo tissue model. This will further our understanding of the complex mucosal phar- macology in the FGT and accelerate the development of future prevention options. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author. Notes Author contributions. Sequence processing and analysis was done using the resources of the Minnesota Supercomputing Institute. Drug concentrations were measured at the University of North Carolina Center for AIDS Research Clinical Pharmacology and Analytical Chemistry Laboratory. F. K. M. and M. R. N. designed research. A. L., R. J., E. I., R. N., S. K., and M. E. B. performed research. A. L. did data analysis and wrote the manuscript. All authors participated in editing of the manuscript. Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (grant numbers K08 AI134262 to M. R. N. and R01 AI118332 to F. K. M.). and the National Center for Advancing Translational Sciences at the National Institutes of Health (grant numbers TL1R002493 and UL1TR002494 to A. L.). Potential conflicts of interest. F. K. M. has received funding from Gilead. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors con- sider relevant to the content of the manuscript have been disclosed. References 1. Cottrell ML, Yang KH, Prince HMA, et al. A translational pharmacology approach to predicting outcomes of preex- posure prophylaxis against HIV in men and women using tenofovir disoproxil fumarate with or without emtricita- bine. J Infect Dis 2016; 214:55–64. 2. Greenbaum S, Greenbaum G, Moran-Gilad J, Weintraub AY. Ecological dynamics of the vaginal microbiome in Modulators of Drugs in Female Genital Tissue • JID 2024:230 (15 December) • 1441 D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 http://academic.oup.com/jid/article-lookup/doi/10.1093/infdis/jiae372#supplementary-data http://jid.oxfordjournals.org/ http://academic.oup.com/jid/article-lookup/doi/10.1093/infdis/jiae372#supplementary-data http://academic.oup.com/jid/article-lookup/doi/10.1093/infdis/jiae372#supplementary-data relation to health and disease. Am J Obstet Gynecol 2019; 220:324–35. 3. Ravel J, Gajer P, Abdo Z, et al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A 2011; 108:4680–7. 4. Klatt NR, Cheu R, Birse K, et al. Vaginal bacteria modify HIV tenofovir microbicide efficacy in African women. Science 2017; 356:938–45. 5. Carlson RD, Sheth AN, Read TD, et al. The female genital tract microbiome is associated with vaginal antiretroviral drug concentrations in human immunodeficiency virus– infected women on antiretroviral therapy. J Infect Dis 2017; 216:990–9. 6. McKinnon LR, Liebenberg LJ, Yende-Zuma N, et al. Genital inflammation undermines the effectiveness of te- nofovir gel in preventing HIV acquisition in women. Nat Med 2018; 24:491–6. 7. Evidence for Contraceptive Options and HIV Outcomes (ECHO) Trial Consortium. HIV incidence among women using intramuscular depot medroxyprogesterone acetate, a copper intrauterine device, or a levonorgestrel implant for contraception: a randomised, multicentre, open-label trial. Lancet 2019; 394:303–13. 8. Brown BP, Feng C, Tanko RF, et al. Copper intrauterine device increases vaginal concentrations of inflammatory anaerobes and depletes lactobacilli compared to hormonal options in a randomized trial. Nat Commun 2023; 14:499. 9. Nicol MR, Eneh P, Nakalega R, et al. Depot medroxypro- gesterone acetate and the vaginal microbiome as modifiers of tenofovir diphosphate and lamivudine triphosphate concentrations in the female genital tract of Ugandan women: implications for tenofovir disoproxil fumarate/ lamivudine in preexposure prophylaxis. Clin Infect Dis 2020; 70:1717–24. 10. Kiweewa Matovu F, Kiwanuka N, Nabwana M, et al. Intramuscular depot medroxyprogesterone acetate accen- tuates bone loss associated with tenofovir disoproxil fumarate-containing antiretroviral therapy initiation in young women living with HIV (the BONE: CARE study): a prospective cohort study in Uganda. Lancet Glob Health 2022; 10:e694–704. 11. Jaumdally SZ, Masson L, Jones HE, et al. Lower genital tract cytokine profiles in South African women living with HIV: influence of mucosal sampling. Sci Rep 2018; 8:12203. 12. Dabee S, Tanko RF, Brown BP, et al. Comparison of female genital tract cytokine and microbiota signatures induced by initiation of intramuscular DMPA and NET-EN hor- monal contraceptives—a prospective cohort analysis. Front Immunol 2021; 12:760504. 13. Tanko RF, Bunjun R, Dabee S, et al. The effect of contra- ception on genital cytokines in women randomized to copper intrauterine device, depot medroxyprogesterone acetate, or levonorgestrel implant. J Infect Dis 2022; 226: 907–19. 14. Nicol MR, Pastick KA, Taylor J, et al. Cerebrospinal fluid and brain tissue penetration of tenofovir, lamivudine, and efavirenz in postmortem tissues. Clin Transl Sci 2019; 12:445–9. 15. Wilen CB, Tilton JC, Doms RW. HIV: cell binding and en- try. Cold Spring Harb Perspect Med 2012; 2:a006866. 16. Nicol MR, Fedoriw Y, Mathews M, et al. Expression of six drug transporters in vaginal, cervical, and colorectal tis- sues: implications for drug disposition in HIV prevention. J Clin Pharmacol 2014; 54:574–83. 17. Kiser JJ, Aquilante CL, Anderson PL, King TM, Carten ML, Fletcher CV. Clinical and genetic determinants of intracellu- lar tenofovir diphosphate concentrations in HIV-infected patients. J Acquir Immune Defic Syndr 2008; 47:298–303. 18. Taneva E, Sinclair S, Mesquita PM, et al. Vaginal micro- biome modulates topical antiretroviral drug pharmacoki- netics. JCI Insight 2018; 3:e99545. 19. Lade JM, To EE, Hendrix CW, Bumpus NN. Discovery of genetic variants of the kinases that activate tenofovir in a compartment-specific manner. EBioMedicine 2015; 2: 1145–52. 20. Introini A, Vanpouille C, Lisco A, Grivel JC, Margolis L. Interleukin-7 facilitates HIV-1 transmission to cervico- vaginal tissue ex vivo. PLoS Pathog 2013; 9:e1003148. 21. Rodriguez-Garcia M, Barr FD, Crist SG, Fahey JV, Wira CR. Phenotype and susceptibility to HIV infection of CD4+ Th17 cells in the human female reproductive tract. Mucosal Immunol 2014; 7:1375–85. 22. Dong M, Dong Y, Bai J, et al. Interactions between micro- biota and cervical epithelial, immune, and mucus barrier. Front Cell Infect Microbiol 2023; 13:1124591. 23. Salinas TRW, Zhang Y, Sarnello D, et al. Th17 cell master transcription factor RORC2 regulates HIV-1 gene expres- sion and viral outgrowth. Proc Natl Acad Sci U S A 2021; 118:e2105927118. 24. Bunjun R, Ramla TF, Jaumdally SZ, et al. Initiating intra- muscular depot medroxyprogesterone acetate increases frequencies of Th17-like human immunodeficiency virus target cells in the genital tract of women in South Africa: a randomized trial. Clin Infect Dis 2022; 75:2000–11. 25. Donkor ON, Ravikumar M, Proudfoot O, et al. Cytokine profile and induction of T helper type 17 and regulatory T cells by human peripheral mononuclear cells after micro- bial exposure. Clin Exp Immunol 2012; 167:282–95. 26. Mills KHG. IL-17 and IL-17-producing cells in protection versus pathology. Nat Rev Immunol 2023; 23:38–54. 27. Spear GT, Sikaroodi M, Zariffard MR, Landay AL, French AL, Gillevet PM. Comparison of the diversity of the vaginal microbiota in HIV-infected and HIV-uninfected women 1442 • JID 2024:230 (15 December) • Lantz et al D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 with or without bacterial vaginosis. J Infect Dis 2008; 198: 1131–40. 28. Puran AC, Adler D, Wallace M, et al. Incidental findings of bacterial vaginosis and other infections in Papanicolaou smears of HIV-infected and HIV-uninfected adolescent fe- males in South Africa. J AIDS HIV Res 2014; 6:172–6. 29. Apalata T, Nojaholo S, Seipone ID, Nxasana N. Characterizations of bacterial vaginosis among HIV- positive and HIV-negative women in rural Eastern Cape Province, South Africa. Int J Microbiol 2021; 2021:9913878. 30. Barditch-Crovo P, Deeks SG, Collier A, et al. Phase I/II trial of the pharmacokinetics, safety, and antiretroviral activity of tenofovir disoproxil fumarate in human immunodeficien- cy virus–infected adults. Antimicrob Agents Chemother 2001; 45:2733–9. 31. Thurman AR, Schwartz JL, Cottrell ML, et al. Safety and pharmacokinetics of a tenofovir alafenamide fumarate- emtricitabine based oral antiretroviral regimen for pre- vention of HIV acquisition in women: a randomized controlled trial. EClinicalMedicine 2021; 36:100893. Modulators of Drugs in Female Genital Tissue • JID 2024:230 (15 December) • 1443 D ow nloaded from https://academ ic.oup.com /jid/article/230/6/1434/7719365 by U N IV O F W ITW ATER SR AN D user on 15 January 2025 Female Genital Tract Host Factors and Tenofovir and Lamivudine Active Metabolites METHODS Study Population Sample Collection Cytokine Quantification Antiretroviral Quantification Microbiome Sequencing Western Blot Analysis CCR5, CD4, AK2, and MRP4 Expression Statistical Analysis RESULTS Participant Characteristics Cervical Cytokine Concentrations and Relative Bacterial Abundance by Contraception Method Associations Between Cytokines, Antiretrovirals, and the Microbiome Gene Expression and ROR-Γ in Cervical Tissues Stepwise Regression DISCUSSION Supplementary Data Notes References