Reduced eIF3d accelerates HIV disease progression by attenuating CD8+ T cell function

To examine the 13 subunits of eIF3s in early HIV infection (EHI) patients, 15 treatment-naive EHI patients (8 RPs and 7 CPs) and 11 HCs were enrolled. EHI patients, defined as within the first 6 months of infection, were recruited from a large-scale prospective cohort of HIV-negative men who had sex with men, which was established in our institute from 2008 to 2010. Patients in the EHI group were divided into two groups according to CD4+ T cell count during follow-up: RPs (CD4+ T cell counts < 350 cells/μl within 1 year of infection) and CPs (CD4+ T cell counts remained > 500 cells/μl after 1 year of infection). At each visit, blood samples, including PBMCs, were cryopreserved. We detected eIF3 expression in PBMCs from EHI patients at approximately 100 days post-infection (RPs: 99.63 ± 28.04 days; CPs: 94.71 ± 14.23 days). The 11 HCs were randomly selected from among HIV antibody-negative healthy men age matched with those infected with HIV. Table 1 summarizes the relevant characteristics of the 15 patients with EHI and the 11 HCs.

To confirm whether eIF3d expression in CD8+ T cells was altered in HIV-infected patients, 18 treatment-naive patients with chronic HIV-infected patients and 17 matched HCs were enrolled (summarized in Additional file 1: Table S1). Among the 18 patients, 15 received ART during follow-up. Their PBMC samples were preserved in our laboratory from the stages of treatment-naive to 2 years after ART. The Research and Ethics Committee of The First Affiliated Hospital of China Medical University approved the protocol for this study, and each enrolled individual provided their written informed consent for participation in the study.

Real-time polymerase chain reaction (PCR) was used to detect expression of eIF3s in cells. Total mRNA was isolated using the RNeasy Micro kit (Qiagen) and reverse transcribed using the Primpscript^®RT reagent kit (TAKARA) according to the manufacturer’s instructions. Real-time PCR for the eIF3s mRNA was performed using Roche LightCycler480 with SYBR^® Premix Ex Taq™ II (TAKARA). The levels of eIF3 mRNA expression were normalized to those of GAPDH. Relative mRNA expression levels were calculated based on the change in the cycling threshold method as 2^−ΔΔCt. The primers used in the experiment are provided in detail in Additional file 2: Table S2.

Whole blood samples were collected from each subject by venipuncture, and density gradient centrifugation was used to extract PBMCs. CD4+ T cells (CD3+CD4+), CD8+ T cells (CD3+CD8+), monocytes (CD3−CD14+), natural killer (NK) cells (CD3−CD56+), and B cells (CD3−CD19+) of HCs were sorted from PBMCs using a BD FACS Aria flow cytometer. The following monoclonal antibodies (mAbs, Biolegend) were used: PerCP-conjugated anti-CD3, FITC-conjugated anti-CD4, PE-conjugated anti-CD8, PE-CY7-conjugated anti-CD14, PE-CY7-conjugated anti-CD56 and PerCP-conjugated anti-CD19. STEMCELL was used to sort CD8+ T cells from HIV-infected patients, and the sorting purity was detected by flow cytometry. CD8+ T cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 supplemented with 10% fetal bovine serum (FBS) (HyClone), 1% penicillin–streptomycin (Gibco) and IL-2 [30, 31] (30 U/ml, Sigma). Transfection of siRNA and controls (Invitrogen) was performed with Lipofectamine™ RNAiMAX Transfection Reagent (Invitrogen). Briefly, cells were transfected with 100 pmol of eIF3d siRNA or 75 pmol of eIF3d siRNA plus 75 pmol of SOCS-7 siRNA. Transfection efficiency was measured by real-time PCR after 48 h of transfection.

Western blotting to detect the eIF3d protein was performed using standard methods. Total protein was extracted from transfected cells with RIPA buffer, and the protein concentration was measured using a BCA Protein assay kit (Thermo Fisher Scientific, Inc.). The protein samples (30 µg) were subjected to SDS-PAGE (Bio-Rad, USA), transferred onto PVDF membranes and incubated overnight with primary antibodies at 4 °C. On the 2nd day, the membranes were incubated with secondary antibody at room temperature for 2 h. Immunoreactive bands were detected with an ECL western blotting system (Clarity Western ECL Substrate; Bio-Rad). The antibodies used in this study were as follows: rabbit anti-GAPDH (ab119716, Abcam), rabbit anti-eIF3d antibody (ab155419, Abcam), and goat anti rabbit IgG H&L (ab6721, Abcam).

To detect cell proliferation, sorted CD8+ T cells were labelled with CellTrace™ Violet reagent (5 mM; Life Technologies) in PBS and incubated for 15 min at 37 °C after treatment with siRNA for 6 h. Labelled CD8+ T cells were stimulated for 4 days with Phytohemagglutinin (PHA, 5 μg/ml, Sigma) [30]. Proliferation was measured at day 4 using a BD LSRII flow cytometer (Becton–Dickinson). To detect apoptosis, sorted CD8+ T cells from HCs were stimulated for 48 h with PHA after treatment with siRNA for 6 h. The cells were then stained with PE-conjugated-Annexin V and 7-aminoactinomycin D (7-AAD, Biolegend) for 30 min at 4 °C and analysed using an LSRII flow cytometer. An intracellular IFN-γ stimulation assay, was carried out in which freshly isolated CD8+ T cells were transfected with siRNA for 48 h and then stimulated for 6 h with 2 µl/ml Cell Activation Cocktail (each vial of this cocktail contained phorbol-12-myristate 13-acetate 40.5 µM, ionomycin 669.3 µM, and Brefeldin A 2.5 mg/ml in dimethylsulfoxide (DMSO) Biolegend). Finally, the cells were intracellularly stained with APC-conjugated anti-IFN-γ (Biolegend) and analysed using an LSRII flow cytometer.

RNA sequencing technology was provided by Shanghai OE Biotech Co. Ltd. Jurkat cells that had been treated with eIF3d si-RNA or control si-RNA for 48 h were used for RNA extraction using an RNA Isolation Kit (Ambion) following the manufacturer’s protocol. Libraries were constructed using the TruSeq Stranded mRNA LTSample Prep Kit (Illumina) according to the manufacturer’s instructions. Then, these libraries were sequenced on the Illumina sequencing platform (HiSeqTM 2500 or Illumina HiSeq X Ten), generating 125-bp/150-bp paired-end reads. Raw data (raw reads) were processed using the Next-Generation Sequencing (NGS) Quality Control (QC) Toolkit. Clean reads were mapped to a reference genome using hisat2. The Fragments Per kb Per Million Reads (FPKM) value for each gene was calculated using cufflinks, and the read counts for each gene were obtained by htseq-count. IPA pathway enrichment analysis was applied for functional categorization of differences in gene expression (absolute fold change > 1.2 and P value < 0.05).

To generate the pNL4-3-VSV-G virus, 16 µg pNL4-3 with green fluorescent protein (GFP) and 4 µg of VSV-G plasmid were transiently transfected into 293T cells using Lipofectamine 2000 (Invitrogen). The culture supernatant was collected at 48 h and filtered, and P24 was quantified using an ELISA kit (Quantobio). Concentrations of P24 were calculated using a standard curve. STEMCELL was employed to sort CD4+ T cells from HCs. PBMCs and CD4+ T cells were stimulated with PHA at a concentration of 5 μg/ml for 1 day prior to infection. One million cells were infected using the pNL4-3-VSV-G virus supernatant with 5 ng of P24^gag antigen plus 1 µl of polybrene.

SPSS 17.0 and GraphPad Prism were utilized to conduct statistical analyses. Independent-sample t-tests were used to compare eIF3s expression among patients. Correlations between eIF3s, CD4+ T cells, and viral loads were evaluated using Spearman correlation analysis. Kaplan–Meier survival analysis was applied to evaluate the effect of eIF3d expression on the reduction in CD4+ T cell counts. One-way analysis of variance (ANOVA) was employed to investigate eIF3d across the five cell types from HC PBMCs. The paired t-test was used to compare eIF3d expression and apoptosis, proliferation and IFN-γ secretion after siRNA inhibition. P values < 0.05 were considered statistically significant.

It has been reported that eIF3s can act as a predictor of disease prognosis in some tumours [28, 32]. However, the relationship between eIF3 subunits and HIV disease progression has not been reported. We first investigated the levels of 13 eIF3 subunits in PBMCs from the EHI patients enrolled in our study (Additional file 3: Table S3). We found that patients with EHI had reduced levels of eIF3d and eIF3k (P = 0.001, P = 0.001) and increased levels of eIF3b, eIF3c, eIF3l and eIF3m compared to HCs (P < 0.001, P < 0.001, P = 0.004, P = 0.009, Additional file 4: Figure S1). We subsequently compared the expression levels of eIF3b, eIF3c, eIF3d, eIF3k, eIF3l and eIF3m between RPs and CPs and found that among the six eIF3 subunits, only the expression of eIF3d was significantly altered between RPs and CPs, with the former having a lower level (P = 0.002, Fig. 1a). In addition, eIF3d was positively correlated with CD4+ T cell counts (r = 0.804, P < 0.001, Fig. 1b) and negatively correlated with viral load (r = 0.835, P < 0.001, Fig. 1c). These data showed lower levels of eIF3d expression to be related to HIV rapid progression. To confirm this, we performed survival analysis in which EHI patients in the prospective cohort were divided into high and low expressers (above and below the median value, respectively) according to the median value of eIF3d expression; CD4+ T cell counts reaching 350 cells/μl were considered as the end point for follow-up. Kaplan–Meier survival analysis showed that the mean time for CD4+ T cell counts to reach 350 cells/μl in low expressers was significantly shorter than that in high expressers (P < 0.001, Fig. 1d). To the best of our knowledge, our study is the first to suggest that eIF3d is significantly correlated with the rapid progression of HIV infection.

Although eIF3s have been reported to directly influence the proliferation and invasion of tumour cells [24, 33], no relevant research on human T cells can be found in the literature. We suspected that eIF3d may exert effects on CD8+ T cell function in HIV infection. Because PBMCs were used for detecting the 13 eIF3 subunits in patients with EHI, we hoped to identify which primary cell subtype among PBMCs contributes to the observed alterations in eIF3d. Expression of eIF3d in different cell subsets (CD4+ T cells, CD8+ T cells, NK cells, B cells, and monocytes) of PBMCs from five healthy controls was sorted by flow cytometry, and we found that CD8+ T cells displayed the highest level of eIF3d expression (P = 0.026, Fig. 2a). CD8+ T cells are among the most important effector cells in HIV infection, and their function has been shown to be related to faster disease progression [34]. This suggested that the reduction in eIF3d may affect the function of CD8+ T cells. We then examined the expression levels of eIF3d in sorted CD8+ T cells from 18 treatment-naïve patients with chronic HIV-infected patients and 17 HCs. The CD8+ T cell sorting purity was 99.3% (Fig. 2b). Expression of eIF3d mRNA was significantly lower in treatment-naïve patients (0.31 ± 0.13) than in HCs (0.95 ± 0.58, P = 0.008, Fig. 2c). Among the 18 patients referred to in Fig. 2c, 15 received ART during follow-up. CD8+ T cells were sorted from PBMCs, and eIF3d was detected. According to the results, eIF3d mRNA levels in CD8+ T cells (0.65 ± 0.38) were increased after ART (P = 0.013, Fig. 2c) compared with the levels in treatment-naïve patients. We also compared eIF3d expression in the same patients before and 2 years after ART and observed that levels recovered significantly after treatment (P = 0.018, Fig. 2d). In addition, there was a significant positive correlation between eIF3d expression in CD8+ T cells from treatment-naïve patients and CD4 + T cells counts (r = 0.691, P = 0.001, Fig. 2e). This result indicates that eIF3d may affect HIV disease progression by impacting the function of CD8+ T cells. Therefore, we examined the effects of eIF3d on CD8+ T cell proliferation, apoptosis, and IFN-γ secretion. EIF3d siRNA or control siRNA was transfected into CD8+ T cells, which led to a 60% reduction at 48 h after transfection, as determined by RT-PCR (P = 0.012, Fig. 2f). Compared with controls, eIF3d siRNA-treated CD8+ T cells showed a significantly lower level of proliferation (P = 0.046, Fig. 2g, h) and IFN-γ secretion (P = 0.026, Fig. 2k, l) after PHA stimulation. However, the level of apoptosis in CD8+ T cells was significantly increased (P = 0.031, Fig. 2i, j) after eIF3d inhibition. According to our results, eIF3d knockdown directly attenuates CD8+ T cell proliferation and IFN-γ secretion and promotes apoptosis.

Previous studies have reported that eIF3d regulates tumour cell function via different signalling pathways [33, 35, 36]. However, the mechanism by which eIF3d regulates T cells remains unclear. To identify which mRNA in CD8+ T cells eIF3d may be responsible for regulating, we knocked down eIF3d in Jurkat cells and performed comprehensive transcriptome analysis. We obtained 19,975 genes using gene and sample requirement filtering. According to principal component analysis (PCA), the expression profile of individual samples in the eIF3d siRNA group was separate from that of the control siRNA group (Fig. 3a). In all, 211 transcripts were significantly differentially expressed between eIF3d siRNA-treated Jurkat cells and controls, including 76 upregulated and 135 downregulated transcripts (P < 0.05, absolute fold change > 1.2). These genes were sorted by FPKM differences in expression, and hierarchical cluster analysis was employed to illustrate the patterns of these differentially expressed genes (Fig. 3b).

We identified the top ten scoring pathways using ingenuity pathway analysis (IPA) enrichment analysis (Fig. 3c). Of these, the Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway was shown to be involved in the regulation of cell function. We further characterized SOCS-7, which is known to negatively regulate JAK-STAT signalling, and found that it was upregulated in eIF3d siRNA-treated Jurkat cells (P = 0.036, Fig. 3d). Next, we validated these findings in primary CD8+ T cells. After knocking down the eIF3d gene with siRNA in CD8+ T cells from three HCs, we found that the expression level of SOCS-7 was significantly increased, with a fold change of 2.27 (P = 0.028, Fig. 3e), suggesting that eIF3d can negatively regulate expression of SOCS-7.

Transcriptomic analysis suggested that reduced expression of eIF3d may directly result in increased expression of SOCS-7. Nonetheless, it remains unknown whether eIF3d can weaken the function of CD8+ T cells by up-regulating SOCS-7. Therefore, we evaluated the function of human primary CD8+ T cells by simultaneously knocking down eIF3d and SOCS-7 by transiently transfecting eIF3d siRNA alone or plus SOCS-7 siRNA into CD8+ T cells using Lipofectamine, followed by culturing to monitor proliferation, IFN-γ secretion, and apoptosis. In the co-knockdown group, the eIF3d inhibition rate was 67% (P = 0.048), and the SOCS-7 inhibition rate was 76% (P = 0.038) (Fig. 4a). Western blotting verified a knockdown efficiency of eIF3d protein in CD8+ T cells of 57% (Additional file 5: Figure S2). Compared with the single-knockdown eIF3d group, cell proliferation was significantly restored in the co-knockdown group (P = 0.022, Fig. 4c, d), and the rate of apoptosis was reduced (P = 0.036, Fig. 4e, f). Additionally, secretion of IFN-γ was slightly restored (P = 0.064, Fig. 4g, h). However, proliferation and IFN-γ secretion in the co-knockdown group were still lower than those in the HC group (P = 0.043, P = 0.024), and the rate of apoptosis was still higher (P = 0.034). These results suggest that mechanisms involving molecules other than SOCS-7 are likely to exist in the eIF3d-induced regulation of CD8+ T cells.

In addition to its clear association with CD8+ T cells, we postulated that eIF3d may affect HIV disease progression by regulating viral replication. Although previous studies have shown an increase in HIV-1 reverse transcripts in HEK 293 cells in which eIF3d was knocked down [29], it has not been addressed whether eIF3d affects HIV replication in primary T cells. We transfected eIF3d siRNA and control siRNA into Jurkat cells, PBMCs and CD4+ T cells, with knockdown efficiencies of 51% in Jurkat cells (P = 0.001, Fig. 5a), 45% in PBMCs (P = 0.02, Fig. 5c) and 56% in CD4+ T cells (P = 0.009, Fig. 5e) compared to the control group. Western blotting showed a knockdown efficiency for the eIF3d protein in Jurkat and CD4+ T cells of 42% and 48%, respectively (Additional file 5: Figure S2). Next, using a fusion of HIV (pNL4-3-VSV-G), we explored the proportion (%) of Jurkat cells expressing GFP and assessed P24 in the culture supernatant of PBMCs and CD4+ T cells. The levels of GFP in eIF3d siRNA-treated Jurkat cells were significantly higher than the levels in controls after 48 h (P = 0.013, Fig. 5b). In addition, levels of P24 in eIF3d siRNA-treated PBMC and CD4+ T cell culture supernatants were significantly higher than those in controls after 72 h (P = 0.025, Fig. 5d; P = 0.02, Fig. 5f). The findings suggest that reduced levels of eIF3d promote HIV virus replication.