Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Feb 1;36(3):346-360.
doi: 10.15252/embj.201694335. Epub 2016 Dec 19.

miRNA Profiling of Human Naive CD4 T Cells Links miR-34c-5p to Cell Activation and HIV Replication

Affiliations
Free PMC article

miRNA Profiling of Human Naive CD4 T Cells Links miR-34c-5p to Cell Activation and HIV Replication

Andreia J Amaral et al. EMBO J. .
Free PMC article

Abstract

Cell activation is a vital step for T-cell memory/effector differentiation as well as for productive HIV infection. To identify novel regulators of this process, we used next-generation sequencing to profile changes in microRNA expression occurring in purified human naive CD4 T cells in response to TCR stimulation and/or HIV infection. Our results demonstrate, for the first time, the transcriptional up-regulation of miR-34c-5p in response to TCR stimulation in naive CD4 T cells. The induction of this miR was further consistently found to be reduced by both HIV-1 and HIV-2 infections. Overexpression of miR-34c-5p led to changes in the expression of several genes involved in TCR signaling and cell activation, confirming its role as a novel regulator of naive CD4 T-cell activation. We additionally show that miR-34c-5p promotes HIV-1 replication, suggesting that its down-regulation during HIV infection may be part of an anti-viral host response.

Keywords: HIV‐1; HIV‐2; T‐cell activation; miR‐34c‐5p; naive CD4 T cells.

Figures

Figure 1
Figure 1. miR expression profile of naive CD4 T cells and impact of HIV infection

Relative abundance of miR species in human naive CD4 T cells. Cell purity was > 97%.

Average miR expression levels in unstimulated naive CD4 T cells non‐infected versus HIV‐1NL4‐3‐(B) or HIV‐2ROD‐(C) infected after 24 h (n = 2, three pooled samples per library). Mean normalized read counts across libraries are shown. Library normalization was performed using the mean normalization method available in the Deseq package of Bioconductor (Anders & Huber, 2010). Lines indicate changes of ± 1log2 fold between samples.

Figure EV1
Figure EV1. HIV infection parameters in sequencing library samples

Cell‐associated proviral DNA (top) and gag mRNA (center) levels upon HIV‐1 and HIV‐2 infections of the naive CD4 T cells used to generate the sequencing libraries analyzed in Fig 1. Each dot represents one individual, with color identifying the individual samples pooled together for subsequent NGS analysis. Whiskers represent the median and interquartile range. Bottom: Illustrative semi‐quantitative PCR for HIV‐1 tat‐rev spliced mRNA from naive CD4 T cells non‐infected (NI) or HIV‐1‐infected (HIV) along with G6PC housekeeping gene.

Hierarchical clustering of the normalized expression counts of all miRNA species in sequencing libraries from unstimulated CD4 T cells, either non‐infected (NI) or infected with HIV (HIV1 or HIV2). Index a and index b identify paired libraries generated by pooling samples from the same set of three donors.

Levels of cell‐associated proviral DNA (top) and gag mRNA (center) upon HIV‐1 and HIV‐2 infections of the TCR‐stimulated naive CD4 T‐cell samples used to generate the sequencing libraries analyzed in Fig 3. Each dot represents one individual, with individual samples pooled together for subsequent NGS analysis identified by the same color. Whiskers represent median and interquartile range. Bottom: Illustrative semi‐quantitative PCR for HIV‐1 tat‐rev spliced mRNA from stimulated naive CD4 T cells non‐infected (NI) or HIV‐1‐infected (HIV), and G6PD housekeeping gene.

Source data are available online for this figure.
Figure 2
Figure 2. miR‐34c‐5p is up‐regulated in naive CD4 T cells in response to TCR‐mediated stimulation

Average miR expression levels in naive CD4 T cells (purity > 97%) before and after 72 h TCR stimulation. Lines indicate changes of ± 1 log2 fold between samples (n = 2, paired pools of three and nine individual samples per library). Library normalization performed using the mean normalization method available in Deseq package of Bioconductor (Anders & Huber, 2010). miRs with significant changes are highlighted (black) and with insets presenting mean read counts in unstimulated samples (mean), log2 fold change values in stimulated samples (log FC), and corresponding adjusted P‐value (P; Fisher's test implemented in Deseq). The two most highly expressed miRs (mir‐21‐5p and miR‐146b‐5p) are also indicated.

Relative miR‐34c‐5p and miR‐155‐5p expression levels (2−ΔCt normalized to RNU48) in unstimulated and stimulated naive CD4 T cells of individual samples pooled to generate the sequencing libraries. Mean ± SEM and P‐value for differences in miR‐155‐5p expression (paired two‐tailed t‐test) are shown. Comparison could not be performed for miR‐34c‐5p as it is undetermined in unstimulated samples.

Time‐course quantification of miR‐34c‐5p expression (top panel, 2−ΔCt values), cell proliferation (middle panel), and cell activation (bottom panel) markers by flow cytometry, in three different donors.

Figure EV2
Figure EV2. Changes in naive CD4 T‐cell miR expression profile in response to TCR stimulation

Relative abundance of miR species after 72 h TCR stimulation.

miR‐34c‐5p detection sensitivity assay. Plot shows Ct values for serial 1:5 dilutions of a miR mimic standard. Number of molecules in each data point is indicated, as well as information regarding the exponential fit and estimate of the number of molecules that correspond to the maximum Ct cycles of standard qPCR assays.

Compiled data for relative miR‐34s‐5p expression (normalized to RNU48) in the 10 individual samples analyzed in Figs 2B and C, and EV2E. To allow for fold change calculation and statistical comparison, undetermined Ct values for non‐stimulated samples were set to 43 (red asterisk). Differences between conditions are significant with P < 0.05 (paired two‐tailed t‐test). Error bars are ± SEM.

Comparative analysis of miR‐34c‐5p and miR‐155 levels. Illustrative fold change in miR abundance for one of the individual samples in the time‐course experiment in Fig 2C. For representation purposes of the RT–qPCR data, miR‐34c‐5p CtT0=40.

Time‐course of pri‐miR‐34c expression. Representative pri‐miR‐34c and G6PD housekeeping gene expression levels assessed by semi‐quantitative RT–PCR in 72 h TCR‐stimulated naive CD4 T cells (top, one out of 3), and levels of mature miR‐34c‐5p quantified in parallel by RT–qPCR (2−ΔCt; bottom) with the statistics for variation along time in the three samples tested shown in the table below (ANOVA test).

Source data are available online for this figure.
Figure 3
Figure 3. Changes in miR expression in response to HIV infection of TCR‐stimulated naive CD4 T cells

Comparison of the mean miR expression level in TCR‐stimulated naive CD4 T‐cell small RNA‐seq libraries from uninfected and HIV‐1NL4‐3‐(A) or HIV‐2ROD‐(B) infected samples (24 h). Only miRNAs with a minimum of 10 normalized read counts. Library normalization was performed using the mean normalization method available in Deseq package of Bioconductor (Anders & Huber, 2010). Lines indicate changes of ± 1log2 fold between samples (n = 3, paired pools of three individual samples per library). miR‐34c‐5p is highlighted in black.

Quantification of miR‐34c‐5p expression levels of the individual samples pooled to generate the sequencing libraries by qPCR (mean ± SEM 2−ΔCt values and P‐value from a paired two‐tailed t‐test; n = 5).

Figure EV3
Figure EV3. miR‐34c‐5p expression levels in Jurkat cell lines and clones and set‐up of antagomiR assays

Ct values for miR‐34c‐5p and miR‐21 in Jurkat cell lines stably transfected with the pcDNA3 empty vector (JØ) or pcDNA3 containing the pri‐miR‐34c‐5p sequence (J34c). JØ and J34c pools used for infection experiments presented in Fig 4 and in clonal populations isolated from each of these pools (JØ‐1 to 3 and J34c‐2, 4, and 5) used in the microarray profiling study presented in Fig 5 are shown. miR‐34c‐5p was undetectable in all JØ samples, in agreement with parental Jurkat cells. Table presents estimate of number of miR‐34c molecules in each sample based on the standard sensitivity curve presented in Fig EV2B and corresponding estimated fold increase.

miR‐34c‐5p and miR‐21 levels in J34c cells treated with antagomiRs after 24 and 72 h (top) or 96 h (bottom) by RT–PCR. miR levels in untreated J34c and JØ are also shown (bottom). qPCR quantification of miR levels in cells treated with antagomiRs compared to untreated J34c cells is shown for one assay (right). Error bars are ± SD of technical replicates.

Source data are available online for this figure.
Figure 4
Figure 4. Impact of miR‐34c‐5p expression on HIV infection and replication

Cell‐associated viral DNA levels (left), tat mRNA levels (center), and RT activity levels (right) in cell culture supernatants in infected J34c cells versus infected JØ cells at 24 and 48 h post‐infection. *≤ 0.05 (paired two‐tailed t‐test).

Representative flow cytometry plot of intracellular HIV‐1 Gag protein staining (KC57) versus side scatter (SSC) in JØ and J34c cells, mock infected or infected with HIV‐1NL4‐3, after 48 h of culture and corresponding fold change (J34c/JØ) plot in four biological replicate assays. Differences to control were found to be significant with a paired two‐tailed t‐test (≤ 0.05).

Fold change levels for miR‐34c‐5p, cell‐associated viral DNA level, tat mRNA, and RT activity levels in J34c cells treated with miR‐34c‐5p antagomiR versus control antagomiR, 48 h post‐infection. Differences to control were found to be significant with a paired two‐tailed t‐test (≤ 0.05).

Data information: All plots show fold change levels observed in a total of three biological replicate experiments performed (four for panel B). Lines depict the corresponding mean and SEM values.
Figure EV4
Figure EV4. Quality control of the transcriptome profiling dataset of J34c and JØ cell lines with Affymetrix Human Gene 2.0 arrays

NUSE plot. Boxes represents interquartile range with median, whiskers represent range.

RLE plot. Boxes represents interquartile range with median, whiskers represent range.

Unsupervised clustering analysis of the transcriptome.

Principal components analysis.

Data information: JØ‐1, JØ‐2, and JØ‐3: control clones; J34c‐2, J34c‐4, and J34c‐5: miR‐34c‐5p‐overexpressing clones (see Fig EV3A).
Figure 5
Figure 5. Impact of miR‐34c‐5p overexpression on Jurkat T cells

Hierarchical clustering of the 82 differentially expressed (DE) genes identified in Jurkat cell lines overexpressing miR‐34c‐5p (J34c‐2 and J34c‐5) versus control (JØ‐1, 2 and 3 clones). Color scale represents mean centered log2 fold expression level across samples. Pie charts represent the relative proportion of predicted miR‐34c‐5p binding sites by the Miranda or PITA algorithms (or both) in down‐regulated (top) and up‐regulated (bottom) gene sets.

qRT–PCR validation of selected genes. Plot represents the average percentage of target gene mRNA (normalized to GAPDH) in J34c versus JØ cells (all differences were found to be significant with a paired two‐tailed t‐test, P < 0.05; = 5). Error bars are ± SEM.

Figure 6
Figure 6. Functional enrichment and network analysis of the miR‐34c‐5p‐dependent transcriptome in Jurkat T cells

Distribution of enriched GO terms (Biological Processes) in J34c DE genes grouped into nine functional categories. Bar chart represents the number of expected and observed hits for GO terms associated with “Immune response”.

Gene interaction network for J34c DE genes (filled nodes) and additional highly connected genes (transparent nodes). Node size reflects the number of network connections.

Figure 7
Figure 7. miR‐34c‐5p targets in the HIV interactome

Interaction network between genes regulated by miR‐34c‐5p and HIV‐1 proteins. Figure depicts the subset of DE genes identified in J34c cells that have reported interactions with HIV‐1 proteins in the HIV interactome database (Ako‐Adjei et al, 2015), as well as their subcellular localization and function (see also Table EV6).

Western blot of PCAF/KAT2B and histone H3 in JØ or J34c (top), confirming decrease in protein levels. Bottom: AntagomiR to miR‐34c‐5p results in increased KAT2B mRNA levels compared to control‐treated cells. Plot presents relative expression values for mRNA in three biological replicates; bars represent mean and SEM. Differences were found to be significant with P < 0.05 (paired two‐tailed t‐test).

Predicted target sites for miR‐34c‐5p in the PCAF/KAT2B mRNA (black bars) and miR‐target hybridization for the two strongest sites (top); Luciferase assay with the pEZX Renilla/firefly‐PCAF/KAT2B 3′UTR reporter in HEK293 cells co‐transfected with either the miR‐34c‐5p overexpression vector or pcDNA3 empty control (bottom). Differences were found to be significant with P < 0.05 (paired two‐tailed t‐test, n = 5). Error bars are ± SEM.

Source data are available online for this figure.
Figure EV5
Figure EV5. miR‐34c‐5p and PCAF/KAT2B expression in naive CD4 T cells

Quantification of PCAF/KAT2B and miR‐34c‐5p expression levels in naive CD4 T cells isolated from two healthy donors by qRT–PCR in response to TCR stimulation.

Same as in (A), except in HIV‐1‐infected versus non‐infected stimulated cells after 72 h TCR stimulation.

Data information: Each symbol represents one individual. Relative expression corresponds to 2−ΔCt values, normalized to RNU48.

Similar articles

See all similar articles

Cited by 8 articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback