Drug transporter mRNA expression and genital inflammation in South African women on oral pre-exposure prophylaxis (PrEP)

Abstract

Globally HIV remains a major public health problem. In sub-Saharan Africa most new HIV infections occur in adolescent girls and young women. Previously tested antiretroviral drugs as different pre-exposure prophylaxis (PrEP) formulations have shown inconsistent levels of protection against HIV in African women. Besides adherence, biological factors such as drug transporter proteins are increasingly recognized as key modulators of PrEP levels. Drug transporter mRNA expression levels has been significantly correlated to altered PrEP levels in-vitro in different tissues, with inflammation identified as a further modifier of drug transporters mRNA expression and thus PrEP levels. We therefore, aimed to determine possible concordance between drug transporter mRNA expression in the female genital tract (FGT) and blood of N = 45 South African women taking oral PrEP-Truvada® [TDF/FTC)] over 6 months for HIV prevention. Additionally, we determined associations between drug transporter mRNA expression, genital inflammation, and blood-tenofovir diphosphate (TFV-DP). mRNA-expression of four efflux P-gp; MATE-1; MRP-2; MRP-4 and two influx OAT-1 and OAT-3 drug transporters were determined by qRT-PCR. Multiplexed technology was used to measure 27 cytokines to define genital inflammation. Significant positive correlations of mRNA expression for P-gp, MATE-1, MRP-2, and MRP-4 were observed between the FGT and blood at 3- and 6-months post-PrEP initiation (p < 0.05). For OAT-1 however, significant positive correlations were observed pre- and post-PrEP exposure (p < 0.05). Linear-mixed models showed moderate associations between FGT cytokines and drug transporter mRNA expression, with a direct relationship observed between MIP-1β concentration and MATE-1 mRNA expression. Similarly, PLS-DA showed that in women with genital inflammation, consistently higher mRNA expression of MATE-1 was observed compared to women without genital inflammation. No significant associations were observed between drug transporter mRNA expression and blood TFV-DP. Our results suggest that drug transporters may be similarly expressed in the FGT and blood. Furthermore, genital inflammation may modify PrEP levels by altering drug transporter mRNA expression. Collectively, our data may be used to better understand biological factors that may affect PrEP efficacy in African women who remain vulnerable to HIV.

Fig. 1

A Description of the participants and sample allocation for three experimental and analysis plans in the study. Description of the number of soft-cup, buffycoat, and cytobrush samples included in this cohort study. A description of the analyses conducted, and the total number of samples included in each of the three analyses is shown. B Diagram detailing analyses between drug transporters mRNA expression (buffycoat and cytobrush) and measured TFV-DP drug levels in the blood and cytokine concentrations in the FGT. VS = versus

Fig. 2

Graphical representation of correlations for OAT-1 mRNA expression between the FGT and blood. Moderate and significant correlations are shown at three timepoints A baseline, B 3 months, and C 6 months for OAT-1 in cytobrushes (fold change FGT) and buffycoats (fold change blood). Analyses were done using the Spearman’s rank correlation coefficient (rs) and a two-tailed p-value of ≤ 0.05 was considered significant

Fig. 3

Loadings plot showing the correlation coefficients between drug transporter mRNA expression levels and PC scores. The loading plots for PC1 and PC2 are shown at three timepoints. A Baseline: PC1 positive correlation coefficients were observed for MATE-1 and OAT-1 and for PC2 it was OAT-1, MRP-2 and MPR-4, while negative correlation coefficients for PC1 were observed for OAT-3, MRP-4, P-gp and MRP-2 and for PC2 it was MATE-1, P-gp and OAT-3. B 3 months: PC1 positive correlation coefficients were observed for MATE-1, MRP-4, OAT-1 and OAT-3 and for PC2 it was OAT-3, MRP-4 and OAT-1, negative correlation coefficients for PC1 were observed for MRP-2 and P-gp, and for PC2 it was OAT-3 and MATE-1. C 6 months: PC1 positive correlation coefficients were observed only for MATE-1 and for PC2 it was OAT-3, MRP-4 and MATE-1, negative correlation coefficients for PC1 were observed for OAT-3, P-gp, MRP-2, OAT-1 and MPR-4 and for PC2 it was MRP-2, OAT-1 and P-gp

Fig. 4

Scatter plots of PC1 (X variate1) and PC2 (X variate2) scores from the PLS-DA comparing overall drug transporter mRNA expression levels in inflamed versus uninflamed women. No clear separation was observed at A baseline and B 3 months for drug transporter mRNA expression in inflamed women (orange) versus uninflamed women (blue). At C 6 months a slight separation was observed between inflamed and uninflamed women

Fig. 5

Clustered image map demonstrating drug transporter mRNA expression levels by inflammation status at 6 months. Colour key: colours red to orange represents high drug transporter mRNA expression levels, while green to blue colours represent low drug transporter mRNA expression. On the right axis each row represents an individual participant’s pattern of drug transporter mRNA expression. Dendrograms are used along the axes (top and left) to depict how each row and column clusters based on the hierarchical clustering method