The transcriptome of HIV-1 infected intestinal CD4+ T cells exposed to enteric bacteria

Abstract

Global transcriptome studies can help pinpoint key cellular pathways exploited by viruses to replicate and cause pathogenesis. Previous data showed that laboratory-adapted HIV-1 triggers significant gene expression changes in CD4+ T cell lines and mitogen-activated CD4+ T cells from peripheral blood. However, HIV-1 primarily targets mucosal compartments during acute infection in vivo. Moreover, early HIV-1 infection causes extensive depletion of CD4+ T cells in the gastrointestinal tract that herald persistent inflammation due to the translocation of enteric microbes to the systemic circulation. Here, we profiled the transcriptome of primary intestinal CD4+ T cells infected ex vivo with transmitted/founder (TF) HIV-1. Infections were performed in the presence or absence of Prevotella stercorea, a gut microbe enriched in the mucosa of HIV-1-infected individuals that enhanced both TF HIV-1 replication and CD4+ T cell death ex vivo. In the absence of bacteria, HIV-1 triggered a cellular shutdown response involving the downregulation of HIV-1 reactome genes, while perturbing genes linked to OX40, PPAR and FOXO3 signaling. However, in the presence of bacteria, HIV-1 did not perturb these gene sets or pathways. Instead, HIV-1 enhanced granzyme expression and Th17 cell function, inhibited G1/S cell cycle checkpoint genes and triggered downstream cell death pathways in microbe-exposed gut CD4+ T cells. To gain insights on these differential effects, we profiled the gene expression landscape of HIV-1-uninfected gut CD4+ T cells exposed to bacteria. Microbial exposure upregulated genes involved in cellular proliferation, MAPK activation, Th17 cell differentiation and type I interferon signaling. Our findings reveal that microbial exposure influenced how HIV-1 altered the gut CD4+ T cell transcriptome, with potential consequences for HIV-1 susceptibility, cell survival and inflammation. The HIV-1- and microbe-altered pathways unraveled here may serve as a molecular blueprint to gain basic insights in mucosal HIV-1 pathogenesis.

Fig 1. TF HIV-1 strains deplete CD4+ T cells in the LPAC model.

(A) LPAC model. LPMCs were obtained from jejunum biopsies of HIV-1-uninfected individuals and used to infect with p24-normalized TF HIV-1 strains in the presence or absence of Prevotella stercorea. CD4+ T cells were analyzed at 6 days post-infection (dpi). (B) Depletion values were computed by comparing the mean absolute numbers of CD3+CD8- cells in triplicate HIV-1 infected versus mock cultures. Negative depletion values can be oftained if the CD3+CD8- cells in HIV-1-infected were greater than mock. (C) %CD4+ T cell killing data were compiled from multiple donors. Lines correspond to means and statistical analyses were performed using a one-way ANOVA followed by a Dunnett’s multiple comparison test. **p<0.01; ***p<0.001; ns, not significant.

Fig 2. HIV-1 infection alters the gene expression profile of gut CD4+ T cells.

LPMCs from 4 different donors were infected with CH040.c-eGFP and at 4 dpi, GFP+ and matched mock-infected CD4+ T cells were sorted for microarray analyses. Paired t-statistics were performed to determine altered genes. (A) Volcano plots showing altered genes in gut CD4+ T cells following HIV-1 infection, based on a 1.25-fold change cut-off and p<0.05. (B) Comparison of HIV-1-altered genes between this study and a published study using X4-tropic HIV-1 on mitogen-activated peripheral blood CD4+ T cells [21]. Genes in blue were downregulated; genes in red were upregulated. (C) Top 30 downregulated and upregulated genes due to HIV-1 infection, based on fold-change values. Colored bars/genes were previously shown to be involved in HIV-1 infection, see S2 Table for more details.

Fig 3. Downregulation of HIV-1 reactome genes in HIV-1-infected gut CD4+ T cells.

Gene expression data were subjected to Gene Set Enrichment Analysis (GSEA). For all Panel A, blue bars/panels indicate downregulation, whereas red bars/panels indicate upregulation relative to mock. Color intensities indicate magnitudes of gene expression (log₂ test/reference). (A) Top-ranked gene sets altered in HIV-1-infected gut CD4+ T cells correspond to HIV-1 reactome gene sets. (B) HIV-1 reactome genes (n = 97) from the GSEA in panel A were extracted and a heat map indicating the relative expression in the 4 different LPMC donors are shown. (C) Downregulation of HIV-1 reactome gene sets following HIV-1 infection. Heat maps of genes associated with mRNA processing, transcription, proteasome, TCR-MHC signaling and vesicle transport in HIV-1 infection are shown for each of the 4 LPMC donors analyzed.

Fig 4. HIV-1 alters multiple signaling pathways in gut CD4+ T cells.

Microarray data were subjected to Ingenuity Pathway Analysis (IPA). (A) HIV-1-altered pathways. Bars correspond to log-transformed p-value and colors indicate either pathway activation (orange) or inhibition (blue) based on a Z-score. OX40 and PPAR signaling had the highest Z-scores and p-values. (B) OX40 signaling pathway components and (C) evaluation of OX40 expression in gut CD4+ T cells in 5 LPMC donors by flow cytometry. %OX40+ cells were evaluated in total CD4+ T cells from mock-infected cultures versus HIV-1-eGFP+ cells from HIV-1-infected cultures. (D) PPAR signaling pathway components. For panels B and D, non-gray components correspond to relevant genes in the pathway. This includes genes (black border) or complexes (magenta border) that were upregulated (orange-to-red) or downregulated (green). Blue shapes correspond to components of the pathways being investigated. (E) Upstream regulators, either activated (positive Z-score) or inhibited (negative Z-score) based on IPA. Colored bars correspond to selected upstream regulators discussed in the text.

Fig 5. Microbial exposure enhances TF HIV-1 infection and CD4+ T cell death in the LPAC model.

After spinoculation with the TF HIV-1 CH040.c strain, LPMCs were resuspended in media containing or not containing heat-killed Prevotella stercorea (P.s.) at a 2.5 bacteria: 1 LPMC ratio. Supernatants and cells were analyzed at 6 dpi. (A) Infectious titers. Supernatants were evaluated for infectious HIV-1 titers using TZM-bl reporter cells. Log-transformed firefly luciferase values are shown. (B) CD4+ T cell depletion. The difference in the absolute number of CD3+CD8- T cells between HIV-1 infected and uninfected (mock) LPMC cultures were calculated. Mock controls for CH040.c only was not exposed to P. stercorea, while the mock controls for CH040.c+P.s. were exposed to P. stercorea. For both panels, each connected dot corresponds to a different LPMC donor (n = 9 donors). Data were analyzed using a paired 2-tailed Student’s t test. ***p<0.001.

Fig 6. Microbial exposure alters gut CD4+ T cell gene expression.

HIV-1-uninfected LPMCs were exposed to mock (media) or Prevotella stercorea (2.5 bacteria: 1 LPMC) for 4 days, and sorted gut CD4+ T cells were subjected to microarray analyses. (A) Volcano plots showing differentially regulated genes following microbial exposure. (B) Top 30 ranked down- and upregulated genes in microbe-exposed CD4+ T cells. Colored bars correspond to genes discussed in the text. (C) Ingenuity Pathway Analysis. The pathway with the highest Z-score and p-value is the p38 MAPK pathway.

Fig 7. HIV-1 infection of gut Th17 cells.

(A-B) Relative TF HIV-1 susceptibility of Th17 versus Th1 cells. LPMCs were infected with HIV-1 CH040.c and (A) the percentage of intracellular p24+ cells were evaluated by flow cytometry in IL17 versus IFNγ producing cells at 4 dpi. (B) Data were analyzed using 2-tailed paired Student’s t test; **, p<0.01; *, p<0.05. (C-D) Expression of Th17 differentiation genes. Polarization of blood Th0 cells into Th17 cells was associated with (C) gene upregulation (n = 20 genes) and (D) gene downregulation (n = 18 genes) [42]. These gene lists were extracted from [42] and subjected to GSEA. (C) Upregulated Th17 genes were induced (NOM = 1.54, p = 0.05) in microbe-exposed gut CD4+ T cells. (D) By contrast, downregulated Th17 genes, including RNAses 2, 3 and 6 (in red), were not significantly perturbed (p>0.05). Higher color intensities indicate higher magnitudes of expression changes.

Fig 8. Microbial exposure induces a robust type I IFN signature in gut CD4+ T cells.

(A) ISG expression. Gene set enrichment analyses (GSEA) predicted the upregulation of multiple ISGs in microbe-exposed gut CD4+ T cells. These 62 genes were extracted to generate a heat map detailing individual LPMC donors. Blue bars/panels indicate downregulation, whereas red bars/panels indicate upregulation relative to mock. Higher color intensities indicate magnitudes of gene expression (log₂ test/reference values). Genes highlighted in red are known retrovirus restriction factors. (B) IPA-predicted upstream regulators. Red bars correspond to upstream regulators involved in IFN signaling that were discussed in the text. (C) Microbial induction of tetherin/BST-2 on gut CD4+ T cells, as measured by flow cytometry. A representative flow cytometry plot is shown. Data were analyzed using paired 2-tailed Student’s t-test, *, p<0.05.

Fig 9. HIV-1 infection differentially alters multiple pathways in microbe-exposed gut CD4+ T cells.

Prevotella stercorea were utilized as representative enteric bacteria found in the mucosa of HIV-1-infected individuals. (A) Overlap in HIV-1-altered genes between gut CD4+ T cells exposed or not exposed to bacteria. (B) Comparison of HIV-1-altered pathways with or without microbial exposure using Ingenuity Pathway Analyses (IPA). (C) Comparison of HIV-1-altered upstream regulators based on IPA. (D) HIV-1-altered genes in microbe-exposed CD4+ T cells. Colored bars were discussed in the text. (E) Induction of Granzyme B in HIV-1-infected, microbe-exposed gut CD4+ T cells. The percentages of granzyme B+ cells were evaluated in CD3+CD8- cells from mock-infected LPMCs exposed to Prevotella stercorea versus HIV-1-eGFP+ cells from HIV-1-infected LPMCs exposed to Prevotella stercorea. Data were analyzed using a paired 2-tailed Student’s t-test. (F) IPA-predicted downstream effects, based on differentially altered genes.