Gene expression profiling of CD4⁺T cells in treatment-naive HIV, HCV mono- or co-infected Chinese

Microarray analysis was performed to identify the altered transcripts in 3 study groups with HIV, HCV mono-infection and HCV/HIV co-infections, and the raw data has been deposited in the ArrayExpress database (access number: E-MEXP-3601). Pairwise comparisons from the three study groups (HCV-mono versus HIV-mono, HCV/HIV co-infected versus HCV-mono, HCV/HIV co-infected versus HIV-mono) were carried out. Differentially expressed (DE) transcript identifiers (DETIs) with >2 fold change and p < 0.05 were identified for each comparison. The number of DETIs identified in each comparison is listed in Table 1. The detailed information of each DETI in each comparison was listed in Additional file 1: Table S2.

To identify the important categories from the DE genes, functions of each DE gene were inspected. Of 54 DE genes identified in comparison between HCV and HIV mono-infected groups, 24 showed significant matches with the known genes. Functional analysis revealed that genes involved in immune system development, immune response (CXCL10, OAS3, CD38, SERPING1, CCR1, FCGR3A and IFI44L) and G-protein coupled receptor (GPCR) signaling pathway (P2RY13, GPR56, CX3CR1 and FFAR2) were differently affected by these two viruses. In the case of HCV/HIV co-infection versus HCV mono-infection, 72 genes with known function were identified. Many of them (16/72) played important roles in stimulus response and immune system process, including: RSAD2, IL6, C1QC, FCGR1C, LILRA3, IGSF6, IL1RN, OAS3. Besides, genes involved in regulating locomotory behavior such as CCRL2, CX3CR1, FPR2 and IL8RA were also shown different expression profiles between these two groups. In comparison between the HCV/HIV co-infections and the HIV mono-infections, only 7 were identified to be known genes. 3 were small nucleolar RNAs, and the remaining 4 were TNFAIP6, AQP9, IL6 and PTX3. The later four genes play important role in immune response and involve in many human diseases, such as cancers and diseases caused by a variety of virus infections.

To unravel the biological mechanisms differentiating between the HCV/HIV mono- and the co-infected groups, pairwise comparisons using GSEA were performed for all the three groups (HCV-mono versus HIV-mono, HCV/HIV co-infected versus HCV-mono, HCV/HIV co-infected versus HIV-mono). Rather than single DE genes, GSEA evaluates microarray data at the biological pathway level by performing unbiased global searches for genes that are coordinately regulated in predefined gene sets. The number of significantly enriched gene sets (FDR < 0.01) in each pairwise comparison is listed in Table 1. And pathways identified in gene set enrichment analysis (GSEA, FDR < 0.05) were shown in Additional file 2: (Table S3).

We first applied GSEA to identify functional gene sets distinct between the HIV mono- and the HCV/HIV co-infections. In total, 13 gene sets were down-regulated in the HCV/HIV co-infected group (FDR < 0.05), 12 of which were gene sets related to cell cycle check point and mitosis. The most up-regulated pathway was platelet degranulation pathway with a significant change (FDR < 0.05).

We next compared the global gene expression profiles between the HCV and HIV mono-infected individuals, and between the HCV mono- and HCV/HIV co-infections. It is interesting to note that the majority of gene sets up-regulated in the HIV mono-infections also show high enrichment in the HCV/HIV co-infections when compared with the HCV mono-infections. According to their biological function, these gene sets could be mainly divided into (1) cell cycle; (2) immune response; and (3) gene expression (regulation). In the cell cycle category, 32/88 and 18/67 gene sets with an FDR < 0.05 were respectively identified to be significantly up-regulated in the HIV-infected group and in the HCV/HIV co-infected group, respectively. The leading edge analysis revealed that the majority of these genes appear to engage in G1/S and G2/M transitions (Figure 1/ Figure 2). In relation to immune response, genes involved in the innate immune response, particularly, in pathogen-associated molecular patterns (PAMPs) recognition, were shown an increased expressions in the HIV mono-infections and contributed most to the enrich score (Figure 1B). In the HCV/HIV co-infected group, the enrichment plot and heat map of the genes involved in this pathway were shown as representatives in Figure 3. Again, the most generally up-regulated gene sets identified in the HCV/HIV co-infected individuals were innate immunity signaling. These included natural killer cell mediated cytotoxicity, toll like receptor signaling pathways, NOD-like receptor signaling pathways and complement activation. While analyzing gene sets related to gene expression, a little difference existed between these two pairwise comparisons. Although both the HIV mono-infected and HCV/HIV co-infected individuals displayed up-regulation of many genes that involved in mRNA maturation (Figure 1C), genes involved in ribosome formation were significantly down-regulated in the HCV/HIV co-infections when compared with HCV mono-infections. Besides, genes involved in the metabolic pathways were also differently regulated. Gene sets including carbohydrate, lipid, amino acid, nucleotide and even vitamins metabolism were all increased in the HCV/HIV co-infected group. On the contrary, only genes function in amino acid metabolism was detected in the HIV mono-infected group when compared with the HCV mono-infected group.

Besides the above mentioned gene sets, there still existed some other pathways that showed high overlap between these two parewise comparisons (HCV-mono versus HIV-mono, HCV/HIV co-infected versus HCV-mono). Among them, the gene sets involved in signal transduction were top listed. Contrary to the cell cycle category, these gene sets were up-regulated in HCV mono-infected individuals. Further analysis revealed that nearly these entire gene sets could be directly or indirectly associated with GPCR signaling pathway.

To confirm the DE genes from the microarray analysis, the mRNA levels from each paired comparison were selected and quantitated by qRT-PCR. These genes included P2RY13, Mx1, IL6, PTX3, GPR56, OAS1, CX3CR1, USP18 (Figure 4) and FCGR3A, CCR1, TLR4, IFI44L, CD38 (Supplemental Figure 1). The RNA isolated from the CD4⁺ T cells of each individual (16 individuals in each group) was used. And the blood samples from un-infected males with marched ages were collected as control. Expression levels of these genes in four groups were all evaluated (Figure 4) and compared with the results obtained by microarray (Table 2). The qRT-PCR results showed that the expressing profiles of 6 genes (PTX3, IL6, P2RY13, OAS1, CX3CR1 and USP18) exactly matched with the observation in microarray assays; another 2 genes (Mx1 and GPR56) displayed partial similarity at least in one express pattern out of three pairwise comparisons. These results largely confirmed the data from the microarray assays. Besides, the mRNA levels of all the selected genes in un-infected persons with matched ages were also analyzed. As shown in Figure 4 and Table 2, the expression levels of CX3CR1, PTX3, MX1 and P2RY13 decreased significantly in infected groups as compared to the uninfected healthy group; while the expression level of OAS1 elevated in the infected groups. There is one gene (USP18) showing altered expressions only in HCV-infections. For IL6, lower expression levels were observed both in the HCV and HIV infections. Other immune related genes like FCGR3A and CCR1 were significantly upregulated while CD38 were decreased in co-infected group as compared to the mono infected or uninfected groups (Additional file 3: Figure 1).

A male population of Chinese was recruited from an ongoing voluntary-based HIV/AIDS surveillance study in Shenzhen, from September 2009 to December 2010. Written content was obtained from the participants. Ethics approval was obtained from The Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee. All participants have been screened for HIV (Beijing Wantai Biological Pharmacy Enterprise CO., LTD, Beijing), HCV (Beijing Wantai Biological Pharmacy Enterprise CO., LTD, Beijing) and HBV (Beijing Wantai Biological Pharmacy Enterprise CO., LTD, Beijing) antigens and antibodies by standard ELISA analysis recommended by The Center of Disease Control and Prevention of China (guidelines of CDC, China). HIV infection was further confirmed by western blot (HIV Blot 2.2 WB; Genelabs Diagnostics, Singapore) following the National Guideline for Detection of HIV/AIDS of China. In order to rule out the confounding factors including treatment responses and drug resistant bias, all HIV + samples were tested for CD4 counts with flow cytometry, and drug resistant mutations by genotyping analysis. All enrolled HIV + individuals are antiretroviral-naïve with CD4 counts ≥300 cells/μl and without baseline drug resistant mutation. Based on the stringent criteria described above, 24 samples from each infected group (HIV mono-infected, HCV mono-infected and HIV/HCV co-infected group) were selected in this study. For microarray analysis, each group contained 3 biological replicates with 8 individuals in each replicates. To validate our findings by quantitative real-time PCR (qRT-PCR), 24 healthy individuals were further enrolled. Their demographic and epidemiological data have been documented in Additional file 4: Table S1. The average age was similar in all the study groups (29.5, 33.0, 31.4 and 27.5 years for HIV mono-infection group, HCV mono-infection group, HIV/HCV co-infection group and non-infected group, respectively). The CD4 counts were 433 ± 101, 839 ± 234, 524 ± 219 and 651 ± 210 (mean ± SD, cells/μl) for HIV infected, HCV infected, HIV/HCV co-infected and un-infected group, respectively.

Peripheral blood mononuclear cells (PBMCs) were separated and purified immediately after obtaining blood samples (30 ml whole blood from each individual) by Ficoll-Hypaque separation. Fresh CD4⁺ T cells were then obtained by positive isolation with use of microbead immunoselection (Miltenyi Biotec, Oslo, Norway). To maximize the RNA yields and minimize possible variations in gene expression profiles, the experiments were strictly followed as described in a previous study [44]. The purity of isolated CD4⁺ T cells was measured by Flow-cytometric analysis of cell markers (CD4). And it demonstrated that 98.1% ± 0.013 (mean ± SD), 98.5% ± 0.072, 97.6% ± 0.021 and 98.5% ± 0.012 of purified CD4⁺ cells were single positive for the CD4 marker in HIV infected, HCV infected, HIV/HCV co-infected and un-infected group, respectively. For microarray analysis, RNA pooling method was used to reduce the noise and enhance reliability when many subjects were pooled without loss of individual specific information [45]. In each infection group, samples were divided into 3 subgroups for independent replicates. In each subgroup, equal amount of CD4⁺ T cells from 8 individuals were pooled together for RNA isolation and the followed gene expression profile studies. The left CD4⁺ T cells were stored at -80°C freezer for validation. RNA was isolated using RNAeasy Total RNA Isolation Kit (Qiagen, Germany) and applied for microarray assays.

To further understand the biological meanings, Gene set enrichment analysis (GSEA) analysis (version 2.0, Broad Institute http://​www.​broad.​mit.​edu/​gsea) was used. First, a ranked list was obtained by ranking all genes according to the correlation between their expression levels and the group distinctions using the metric signal to noise ratio. Then the association between a given gene set and the group was measured by the non-parametric running sum statistic termed the enrichment score (ES). To estimate the statistical significance of the ES, a nominal p value was calculated by permuting the genes 1,000 times. To adjust for multiple hypotheses testing, the maximum ES was normalized to account for the gene set size (NES) and the false discovery rate (FDR). The gene sets used were from Molecular Signatures Database (MsigDB), catalog C2 (version 3.0) functional sets, subcatalog canonical pathways, which include 880 gene sets from pathway databases. These gene sets were collected from online databases such as Bio-Carta, React, and KEGG (Kyoto Encyclopedia of Genes and Genomes).

RNA isolated from the CD4⁺ T cells of each participant was reverse-transcribed into cDNA using Reverse Transcription System (Promega, USA). qRT-PCR was carried out by using ABI 7500 Real-Time PCR system with Power SYBR Green Master Mix (Applied Biosystems, USA). Primer sequences are listed in Additional file 5: Table S4. The relative mRNA level of each gene was calculated as following formula: R = 2^{CT (GAPDH-X)} x10³, x represent CT value of each gene. Each RT-PCR experiment was performed three times.