Fig 1.
CD127+ memory CD4+ T cells from tonsils are poorly susceptible to productive infection by HIV-1.
A) CD127+ Tm cells are preferentially absent amongst infected tonsillar cells. HLACs were mock-treated or exposed for 3 days to the CCR5-tropic reporter virus F4.HSA, after which the populations of uninfected memory CD4+ T cells (top left) and infected memory CD4+ T cells that have downregulated cell-surface CD4 (top right) were assessed for expression levels of CD57 and CD127 (bottom plots). B) Bar graph comparing the proportions of CD57+, CD127+, and CD57-CD127- Tm cells amongst uninfected or HIV-infected Tm cells from 15 independent donors. ***p<0.001, ****p<0.0001 as determined using a 2-tailed paired parametric t-test. C) CD127+ Tm cells from blood do not restrict productive infection by F4.HSA. Infection conditions were set up as described in panel A but using PBMCs instead of HLACs. Memory CD4+ T cells from the uninfected culture are shown in blue while productively-infected (HSA+) memory CD4+ T cells are shown in red. All results in this figure are pre-gated on live, singlet CD3+CD8-CD45RO+CD45RA- cells.
Fig 2.
Relieving SAMHD1 restriction does not increase permissiveness of CD127+ Tm cells to productive infection by HIV-1.
A) The HLAC Tm subsets express low levels of phospho-SAMHD1. Amounts of SAMHD1 phosphorylated at Thr592 in THP1 cells, total HLACs, Tn cells, and the three Tm subsets, as determined by Western blot. Beta-actin served as a loading control. Both short and long exposures of the phospho-SAMHD1 blots are shown. B) Vpx degrades SAMHD1. Amounts of total SAMHD1 in unstimulated PBMCs that were left uninfected, or infected for 3 days with F4.HSA or F4.HSA harboring Vpx-Vpr fusion protein, as determined by Western blot. 2:1 and 1:2 refer to the ratios of F4.HSA provirus to Vpx-Vpr plasmid in the co-transfections. GAPDH served as a loading control. C) Vpx increases permissiveness of resting PBMCs to productive infection by F4.HSA. Unstimulated PBMCs were mock-treated or exposed for 3 days to F4.HSA lacking Vpx, or harboring different amounts of Vpx. Infection levels were then assessed by flow cytometry. Results are pre-gated on live, singlet CD3+CD8-CD45RO+CD45RA- cells. D) Vpx does not increase HIV infection of CD127+ Tm cells. Representative flow cytometric plots of HLACs mock-treated or exposed for 3 days to F4.HSA lacking or harboring Vpx as indicated. Expression levels of CD57 and CD127 were assessed in uninfected memory CD4+ T cells (plot on left), and HIV-infected (HSA+) memory CD4+ T cells (3 plots on right). Results are pre-gated on live, singlet CD3+CD8-CD45RO+CD45RA- cells. E) Vpx does not increase HIV infection of CD127+ Tm cells. Bar graphs comparing the absolute cell counts of CD57+, CD127+, and CD57-CD127- Tm cells amongst HIV-infected (HSA+) memory CD4+ T cells in two independent donors. Absolute cell counts were determined by normalizing the flow cytometric data to AccuCount beads run for each sample. Error bars correspond to experimental triplicates for each donor. *p<0.05, **p<0.01 as determined using a 2-tailed unpaired student’s t-test.
Fig 3.
CD127+ Tm cells preferentially support latent infection by HIV-1.
A) Schematic of experimental design for quantitating integrated HIV DNA in memory CD4+ T cell subsets from HIV-exposed HLACs. HLACs were mock-treated or infected with F4.HSA and cultured for 3 days. Cells were then sorted using an AriaII instrument for the CD57-CD127-, CD57+, and CD127+ Tm populations. Genomic DNA was extracted from sorted cells, and a two-step Alu-Gag ddPCR was performed to amplify and quantitate HIV DNA from these samples. A second ddPCR reaction designed to detect mitochondrial DNA was performed in parallel for all samples to quantify DNA input, and was used for normalization. B) Gating strategy for sorting of HLAC cultures. Live, singlet CD3+CD8- cells (corresponding to CD4+ T cells) were further gated on memory cells (CD45RO+CD45RA-), and then divided into populations of CD57+, CD127+, and CD57-CD127- Tm cells as shown. These sorted populations were used to quantitate the levels of integrated HIV DNA. C) Flow cytometric plots showing the sorted populations of memory CD4+ T cells from F4.HSA-exposed HLACs, demonstrating the expected low infection rates in the CD127+ Tm cells as compared to the other two Tm subsets. D) The samples shown in panel C were subjected to ddPCR to quantitate the levels of integrated HIV DNA. Infected SupT1-R5 served as a control. Results were normalized to the amount of mitochondrial DNA in each sample. No integrated HIV DNA was detected in uninfected cells subjected to the same protocol. E) The protocol schematized in panel A was conducted on 5 independent donors. The levels of integrated HIV DNA in each population (normalized to mitochondrial content) were divided by the rate of productive infection (as defined by the frequency of HSA+CD4- cells) to demonstrate that the CD127+ Tm cells always harbored a disproportionately high level of HIV DNA relative to their productive infection rate.
Fig 4.
CD127+ Tm cells exhibit a transcriptional profile of quiescence distinct from that of other tonsillar memory CD4+ T cells.
A) Principal Component Analysis (PCA) of RNAseq results from CD127+ (purple), CD57+ (blue), and CD57-CD127- (red) Tm cells displaying the first 3 principal components (left) or the first, second, and fourth components (right). PC3 segregates the samples by donor (each represented by a different shape), while PC1, PC2, and PC4 segregate them by cell type. B) Heatmap of z-scores illustrating the 25 genes with the highest coefficients of variation when comparing CD127+, CD57+, and CD57-CD127- Tm cells. The CD127 transcript is boxed to highlight preferential expression of this gene in the CD127+ Tm cells as expected. Arrows highlight TIGIT and IL2RA, antigens known to be expressed on HIV-infected cells. C) Heatmap of Z-scores illustrating select genes differentially expressed between CD127+, CD57+, and CD57-CD127- Tm cells. These genes were chosen based on known positive associations with HIV infection, and were all expressed at lower levels in CD127+ Tm cells relative to CD57+ Tm cells or CD57-CD127- Tm cells. D) Venn diagram showing the number of overlapping genes differentially expressed between CD127+ and CD57+ Tm cells, and between CD127+ and CD57-CD127- Tm cells. E) Activation z-scores of select IPA pathways as compared between CD127+ and CD57+ Tm cells, or CD127+ and CD57-CD127- Tm cells. Pathways are listed according to z-score, from top to bottom (highest to lowest absolute z-score) with a p-value cutoff filter of log101.3. Blue corresponds to pathways downregulated in CD127+ Tm cells, with the intensity of the blue corresponding to the extent of downregulation. Orange corresponds to pathways upregulated in CD127+ Tm cells. For a full list of significant pathways, see S3 Table. F) Graphical representation of select pathways differentially active in CD127+ Tm cells as compared to CD57+ (left) or CD57-CD127- (right) Tm cells. Biological relationships between the different pathway components were found by IPA and are represented as dashed lines if the relationship is indirect and continuous lines if the relationship is direct. Pathway components are displayed using various shapes that represent the functional class of the gene product, as defined in the figure key. The names of the pathway components correspond to those in the IPA Knowledge Database. Pathways components were colored according to experimental expression values, with the fold-change color code displayed at the bottom of each panel. These data demonstrate that multiple pathways associated with T cell activation and HIV gene expression are downregulated in CD127+ Tm cells.
Fig 5.
Latently-infected CD127+ Tm cells can be reactivated by T cell stimulation.
A) Gating strategy for sorting of HLAC cultures. Live, singlet CD3+CD8- cells (corresponding to CD4+ T cells) were further gated on memory (CD45RO+CD45RA-) or naïve (CD45RO-CD45RA+) cells. The latently-infected CD127+, CD57+, and CD57-CD127- Tm cells were then isolated by gating on the CD4+HSA- cells as shown. B) Latently-infected CD127+ Tm cells can transcribe HIV but are inhibited in HIV splicing. Total (TAR), 5’ elongated (R-U5/pre-Gag “Long LTR”), Pol, polyadenylated (PolyA), and multiply-spliced Tat-Rev (TatRev) HIV RNAs were measured in the total infected culture, or in the HSA- CD127+ Tm cells sorted as described in panel A. Data are normalized to the housekeeping gene TERT (top left) or to the levels of HIV DNA in each sorted population (top right). Bottom: The extent of elongation, mid-transcriptional elongation, transcript completion, and splicing were determined by examining the ratios of the indicated transcripts. The ratio of Tat-Rev/polyA transcripts in latently-infected CD127+ Tm cells was disproportionately low, suggesting a defect in HIV splicing in these cells. Shown are results of one of two representative donors. C) The sorted populations of naïve CD4+ T cells, as well as CD57-CD127-, CD57+, and CD127+ Tm cells defined in panel A from uninfected or infected HLAC cultures were mock-treated or stimulated with anti-CD3/CD28 beads and then assessed levels of reactivation three days later. The proportions of infected (HSA+) cells that have downregulated cell-surface CD4 are indicated. The infected cells in the non-activated samples are expressing HSA due to spontaneous reactivation of the sorted HSA- cells, while those in the activated samples correspond to stimulation-induced reactivation. Shown are results of one of four representative donors, with the average induction in infection rates, in the CD127+ Tm cells, between the non-activated and activated cultures being 6.8-fold (range 4.4–11.2-fold).