Short-term oral pre-exposure prophylaxis against HIV-1 modulates the transcriptome of foreskin tissue in young men in Africa

Abstract

Whilst short-term oral pre-exposure prophylaxis (PrEP) with antiretroviral drugs in men who have sex with men has shown protection against HIV-1 infection, the impact of this regimen on the in vivo foreskin transcriptome is unknown. We collected foreskin tissue after voluntary medical male circumcision from 144 young men (72 from Uganda and 72 from South Africa) randomized to one to two doses of either oral tenofovir (TFV) disoproxil fumarate (FTC-TDF) or tenofovir alafenamide (FTC-TAF) or no drug (untreated controls). This novel approach allowed us to examine the impact of short-term oral PrEP on transcriptome of the male genital tract. A single dose of FTC-TDF did not affect the foreskin transcriptome in relation to control arm, however one dose of FTC-TAF induced upregulation of four genes AKAP8, KIAA0141, HSCB and METTL17. Following two doses of either FTC-TDF or FTC-TAF, there was an increase in 34 differentially expressed genes for FTC-TDF and 15 for FTC-TAF, with nine DEGs in common: KIAA0141, SAFB2, CACTIN, FXR2, AKAP8, HSCB, CLNS1A, DDX27 and DCAF15. Functional analysis of differentially expressed genes revealed modulation of biological processes related to mitochondrial stress (KIAA0141, HSCB and METTL17), anti-viral and anti-inflammatory pathways (CACTIN and AKAP8). Our results show that short-course on-demand oral PrEP in men modulates genes in foreskin tissue which are likely unfavorable to HIV acquisition and replication. We also describe an upregulated expression of genes involved in diverse mitochondria biology which may potentially result in worsened mitochondria-related. These results warrant further studies to assess the role of short-course and prolonged oral PrEP on biological processes of the foreskin mucosa.

Keywords: emtricitabine tenofovir; foreskin; inflammation; mitochondrial function; pre-exposure prophylaxis PrEP; transcriptomes.

Figure 1

Flow diagram of participants in the CHAPS trial. The participants are divided according to PrEP drug (FTC-TDF or FTC-TAF), number of PrEP doses (one or two) and interval between last dose and circumcision (5 or 21 hours). The number of included participants from South Africa is shown in the pink boxes and for Uganda in the orange boxes. Arm 1 consists of control individuals who did not receive PrEP. Each arm includes an equal number of participants from South Africa and from Uganda.

Figure 2

Differential gene expression analysis of South African and Ugandan cohorts. Individuals from South Africa and Uganda were stratified according to the type of administered PrEP drug and the expression of transcripts was compared to the corresponding control arm or between drug groups. Only comparisons with statistically significant (p < 0.05) differentially expressed genes (protein-coding, N=15378) are shown for each country.

Figure 3

Similarity of control specimens in Uganda and South Africa. To evaluate the extent of inter-country variability we performed a differential expression analysis between the control arms of South Africa and Uganda. Panel (A) depicts the distribution of control specimens produced by PCA; the overlap between the two countries is shown by corresponding ellipse. Panel (B) shows the detection of 3 DEGs between the controls from the two trial sites.

Figure 4

DEGs in merged PrEP-treatment arms as compared to control arm. All specimens from South Africa and Uganda were merged and stratified by administered PrEP drug. The merged arms were then compared to the control group and the statistically significant DEGs shown according to the administered PrEP drug.

Figure 5

Principal component analysis of Uganda and South Africa trial arms combined together. Specimens from the two trial sites were analyzed based on the type of PrEP and number of administered doses. PCA was then performed using protein coding transcripts (N=15378). The overlap between the trial arms from two countries is shown by corresponding ellipses. A color coding in the figure denotes the different trial arms.

Figure 6

Differential gene expression analysis comparing merged PrEP arms receiving different drugs and doses with control arm. Specimens from both South Africa and Uganda were grouped by the type of PrEP and number of administered doses. The expression of all transcripts was then compared to the control arm and statistically significant DEGs and their number are shown in red.

Figure 7

Clustering of gene ontology terms associated with DEGs in the merged drug and dose arms. GO associations with DEGs were established by using clusterProfiler to extract a single GO term for each DEG. The biological process GO terms associated with each of the identified DEGs were then analyzed using REVIGO and the semantic space clusters visualized for each PrEP regimen. GO terms are represented by circles, whose size indicates the relative number of genes included in the term. Circles are colored according to the significance (p-value) of the gene they represent. Labels correspond to the GO term determined to be representative of the cluster. Shared GO terms are shown in red.