TLR9 polymorphism correlates with immune activation, CD4 decline and plasma IP10 levels in HIV patients

The study was reviewed and approved by the Texas Tech University Health Sciences Center’s regional Institutional Review Board (IRB). All methods were performed in accordance with the relevant guidelines and regulations. The study design was cross-sectional and the study number recorded as IRB# E12092, approval date 07/31/2012. All participants were provided with written and oral information about the study. Written informed consent of all study participants in accordance with the Institutional policy was documented. All participants were identified by coded numbers to assure anonymity and all patient records kept confidential.

Fifty HIV-infected individuals and 29 healthy controls were recruited from the outpatient HIV clinic at the Texas Tech University Health Sciences Center at El Paso. The mean age of the HIV+ patient population was 37.9 ± 11.9. The HIV group comprised of 9 females (18.0%) and 41 males (82.0%). The study was cross sectional consisting of patients at different stages of the disease. An age matched HIV- population control group (n = 29) was also recruited from the same geographical location and comprised of 10 females (34%) and 19 males (65.5%). The mean age of the healthy control group was 34.7 ± 9.8.

Each patient provided a 20 ml blood sample that was separated into plasma and cellular components using Ficoll based separation. Genomic DNA was extracted from whole blood samples using the QIAamp DNA Blood Mini kit (Qiagen). All components from the sample including plasma, cells and DNA were aliquoted and stored at − 70 °C till further analysis.

PCR-RFLP was performed to determine the TLR9 1635A/G polymorphism in the TLR9 gene at position 1635 to determine genotypes of all patient samples. For this, the TLR9 gene was PCR amplified surrounding position 1635 (nucleotide bases 1219–1830, numbering from the ORF of the TLR9 gene) using forward primer 5’-CAG CTC GGC ATC TTC AGG GCC TTC-3′ and reverse primer 5’-CAG TGC ATT GCC GCT GAA GTC CAG-3′. The resulting PCR product was digested with BstUI enzyme resulting in 2 bands at 417 and 195 bp for the homozygous GG genotype; one band at 612 bp for homozygous AA and three bands at 612, 417, and 195 bp for heterozygous GA. For TLR9 1486 C/T polymorphism, the TLR 9 gene surrounding the position 1486 was PCR amplified using the forward primer 5′ - CTA TGG AGC CTG CCT GCC ATG ATA CC - 3′ and the reverse primer 5′ - CTG GTC ACA TTC AGC CCC TAG AG - 3′. The resulting 755 bp PCR fragment was digested with AflII resulting in one band at 755 bp for homozygous CC, three bands at 755, 505 and 250 bp for heterozygous CT and two bands at 505 and 250 bp for homozygous TT.

Staining of cells for different immune markers has been described previously [2, 43]. Briefly, lymphocytes isolated from the blood samples obtained from HIV-infected or normal patients were stained for cell surface markers using specific antibodies: CD3-Cy7, CD4-Tx red, CD8-APC (Beckman Coulter), CD38 PE, HLA-DR FITC, CCR5 PE (BD Pharmingen) and CaspACE FITC-VAD-FMK (Promega). Stained cells were fixed using the Cytofix reagent (Beckman Coulter) and run on a 10 color Beckman Coulter Gallios Flow Cytometer. At least 20,000 events for each sample were acquired. Data was analyzed using the FlowJo software (Tree Star). Cells were first gated on the CD3+ population and CD4+ and CD8+ T cell subsets determined.

IL-6, IP10 and sCD14 levels in the plasma samples from HIV infected patients and healthy controls were determined via specific ELISAs. For IL-6 determination, the Quantikine HS ELISA kit (R&D Systems) was used following the manufacturer’s protocol. The kit is designed to measure human IL-6 in serum, plasma and urine and has a sensitivity of 0.016–0.110 pg/mL. IP10 and sCD14 levels were determined using the Quantikine ELISA human IP10 and CD14 Immunoassay (R & D Systems) following the manufacturer’s protocol. The sensitivity of IP10 detection using the kit is 0.41–4.46 pg/mL and the minimum detectable dose (MDD) of human CD14 is less than 125 pg/mL. Plasma LPS levels were determined using the endpoint chromogenic Limulus Amebocyte Lysate (LAL) assay (Lonza) using the manufacturer’s recommendations. The kit shows linear absorbance at 405–410 nm in the concentration range of 0.1–1.0EU/mL endotoxin.

Data were analyzed using GraphPad Prism (GraphPad Software, Inc., La Jolla, CA). Differences between groups were assessed using the students one tail t-test; and p values were considered significant at p < 0.05. Comparison between multiple groups was done using the Kruskal-Wallis test. Spearman’s correlation with linear regression was used for all correlation determination using the GraphPad Prism Software.

Selective depletion of CD4+ T cells via apoptosis followed by reduction of CD4:CD8 ratio is a hallmark of HIV disease progression [44]. The course of HIV infection is highly variable in different individuals. This variability is manifested not only as differential rates of disease progression [45] but also in recovery after HAART administration [13]. Differences in the genetic make-up of the host like TLR polymorphisms can in part explain some of the variability in HIV disease progression [34]. We hence looked at TLR9 1635A/G polymorphism and its association with HIV clinical parameters in a cross sectional patient cohort. The frequency of TLR9 1635A/G polymorphism was determined via PCR-RFLP and distribution of this polymorphism between HIV infected and healthy controls was found to be similar (Table 1). Further analysis showed that patients with TLR9 1635AA genotype had significantly lower (p = 0.025) CD4 counts than subjects with the 1635GG/AG genotype (Fig. 1a). Interestingly, although not significant, a similar trend was seen with CD4:CD8 ratio with 1635AA genotype being associated lower CD4:CD8 ratio than 1635AG/GG (Fig. 1b). These data are in agreement with previous findings [36] and indicate an association of TLR9 1635A/G polymorphism with CD4 levels in HIV infection. In order to better understand the involvement of 1635AA, AG or GG genotype in CD4 decline, we further stratified the CD4 decline data according to the respective genotypes. As demonstrated in Fig. 1c, the 1635AA genotype was associated with the lowest CD4 counts as well as CD4:CD8 ratio (Fig. 1d). This was followed by intermediate CD4 counts for the AG and highest CD4 counts for the GG polymorphism both in terms of CD4 counts and the CD4:CD8 ratio. These data demonstrate that the TLR9 1635AA genotype is associated with lower CD4 counts in HIV infected individuals.

Non-specific activation of both CD4+ and CD8+ T cells characterized by up regulation of markers like CD38 and HLA-DR is a key characteristic of HIV infection [10]. Although the precise mechanism of HIV induced immune activation remains controversial, several hypotheses have been put forth to explain the phenomenon including a role of TLRs [5, 46]. We looked at the association of TLR9 1635A/G polymorphism with immune activation in CD4+ and CD8+ T cells, measured as %CD38 + HLA-DR+ cells, in each cell populations. As demonstrated in Fig. 2, the 1635AA genotype was associated with significantly higher immune activation in both CD8+ (p = 0.017) (Fig. 2a) and CD4+ (p = 0.012) (Fig. 2b) T cells. As HIV infection is characterized by specific depletion of CD4 T cells, we also looked at the association of TLR91635A/G polymorphism with CD4+ T cell apoptosis. As shown in Fig. 2c, although not significant (p = 0.079), TLR9 1635AA polymorphism showed a trend towards higher CD4+ T cell apoptosis when compared to 1635AG/GG. Further analysis of individual contributions of the TLR91635AA, AG or GG genotype in CD8 immune activation showed that the 1635AA genotype was associated with highest immune activation followed by AG and GG genotypes (Fig. 2d) consistent with the CD4 count data. However, we did not find a significant correlation of TLR9 1635A/G polymorphism with plasma viremia (Fig. 2e). This might be because not all the patients in our study population were on HAART at the time of sample collection. These data suggest that TLR9 1635A/G polymorphism may influence HIV disease by regulating immune activation in both CD4 and CD8 T cells.

LPS, sCD14, IL6 and IP10 are some of the plasma biomarkers that have been shown to be elevated in HIV patients [7, 47]. In fact, these inflammatory biomarkers are reliable predictors of HIV disease progression correlating consistently with CD4 decline and plasma viral load [3, 48]. To determine if these markers of inflammation were also elevated in our patient population compared to HIV- controls, we tested plasma samples for LPS using the Limulus Amebocyte Lysate (LAL) assay and quantitative ELISAs for IL6, sCD14 and IP10. Interestingly, in our patient population, LPS (Fig. 3a), sCD14 (Fig. 3b) and IP10 (Fig. 3c) levels were significantly elevated (p < 0.0001) when compared to HIV- controls. Although IL-6 levels were also higher in HIV+ patients compared to HIV- controls (Fig. 3d), the difference did not reach statistical significance in our study. These data confirm previous findings and emphasize the importance of nonspecific inflammatory markers in HIV disease. Furthermore, these findings demonstrate that the upregulation of inflammatory markers in our patient population are consistent with HIV disease.

Our data above showed elevated LPS, sCD14, IL-6 and IP10 levels in HIV infected patients and a significant correlation of the TLR9 1635AA polymorphism with both CD4 and CD8 immune activation. We next asked whether TLR9 1635A/G polymorphism correlated with any of the above inflammatory biomarkers. Although LPS (Fig. 4a), sCD14 (Fig. 4b), and IL-6 (Fig. 4d) levels were higher in the TLR9 1635AA genotype compared to AG/GG, these differences did not reach statistical significance. Interestingly, IP10 levels were significantly elevated (p = 0.021) in patients with the TLR91635AA genotype compared to AG/GG genotype (Fig. 4c). These findings are interesting because TLR9 stimulation has been shown to induce CD38 and HLA-DR upregulation [19] as well as mediate IP10 release from various cells [49, 50]. Taken together, our data suggest a role for TLR9 1635A/G polymorphism in HIV disease by altering HIV mediated immune activation.

There is direct evidence for a key role of plasma viremia in HIV induced immune activation as in most instances, markers of inflammation correlate with viral load [2] and suppressing the virus via ART also reduces immune activation [51]. We hence sought out to determine if in our HIV+ patient population, the levels of plasma biomarkers (LPS, sCD14, IP10 and IL6) were higher in viremic patients (defined as viral load > 100 copies/ml) compared to those that had their plasma viremia well controlled (≤100 copies). In our study population, LPS (Fig. 5a), sCD14 (Fig. 5b), or IL6 levels (Fig. 5d), did not show a significant difference in viremic versus non-viremic patients. Interestingly, IP10 levels were significantly elevated (p = 0.0381) in viremic patients (Fig. 5c) compared to those that had their viremia controlled. The elevated levels of IP10 in viremic patients is consistent with a recent report showing IP10 to be a highly sensitive marker of HIV replication [52]. Moreover, these data are suggestive of a role of plasma viremia in elevating IP10 levels presumably via TLR9 signaling.

Our data above showed that amongst the plasma biomarkers studied, IP10 levels correlate the strongest with viremia and immune activation. We hence conducted a correlation analysis of plasma IP10 levels with CD4 decline and T cell immune activation. Interestingly, plasma IP10 levels correlated strongly (p = 0.0104) with CD4:CD8 ratio (Fig. 6a), CD4 counts (p = 0.0366) (Fig. 6b), CD8 immune activation (p < 0.0001) (Fig. 6c), and CD4 immune activation (p < 0.0001) (Fig. 6d). This suggests that IP10 levels may be strong predictors of HIV induced CD4 decline, inversion of CD4:CD8 ratio and more importantly, CD4 and CD8 immune activation. This is consistent with the study by Pastor et al [52] where assessment of 49 inflammatory biomarkers in a cohort of HIV seronegative individuals showed that IP10 had the highest accuracy in identifying individuals with acute HIV infection.

Several other polymorphisms in the TLR9 region like 1486C/T in the TLR9 promoter have been shown to be associated with diseases like lupus [53] and rheumatoid arthritis [40]. However, there are no reports on the association of TLR9 1486C/T polymorphism with HIV infection or disease progression. We hence looked at the association of TLR9 1486C/T polymorphism with HIV clinical parameters in our cross sectional patient cohort. Interestingly, 1486C/T polymorphism was found to be in linkage disequilibrium with TLR9 1635A/G in our HIV patient population (D’ 0.943). As shown in Fig. 7a, presence of TLR9 1486CC genotype was associated with lower CD4 counts compared to 1486CT/TT, although the difference was not statistically significant. Similarly, CD8 T cell immune activation, a hallmark of HIV disease was also higher for the TLR9 1486CC group (Fig. 7b). Further stratification of data into 1486CC, CT or TT genotypes showed that CD8 immune activation was again highest for 1486CC followed by CT and least for the TT genotype (Fig. 7c). These HIV disease progression markers that were significantly different for the 1635A/G polymorphism only showed a trend but did not reach statistical significance for the 1486C/T analysis. However, the exception was plasma IP10 levels that were significantly higher (p = 0.031) for TLR9 1486CC compared to CT/TT genotypes (Fig. 7d). These data suggest a possible association of TLR9 1486C/T polymorphism with HIV induced CD4 T cell decline, T cell immune activation and increased plasma IP10 levels, although this association was weaker than that seen with 1635AA genotype.

Our data above demonstrated that in our patient population, both TLR9 1635A/G and 1486C/T polymorphisms were likely determinants of CD4 decline, T cell immune activation and elevated plasma IP10 levels. As 1486C/T polymorphism was found to be in linkage disequilibrium with TLR9 1635A/G, we investigated the combined effect of these polymorphisms on some key parameters of HIV disease. As demonstrated in Fig. 8a, the 1635AA genotype combined with 1486CC was associated with the lowest CD4 counts followed by AA-CT, AG-CT, AG-TT and GG-TT. This correlated inversely with CD8 immune activation with the highest immune activation seen for 1635AA-1486CC followed by AA-CT, AG-CT, AG-TT and GG-TT (Fig. 8b). Furthermore, levels of plasma IP10 followed the same pattern as CD8 immune activation with highest IP10 levels seen for the 1635AA-1486CC genotype which was significantly different from AG-CT genotype (p < 0.05) (Fig. 8c). Interestingly, in this stratification of data the AA-CT genotype was not different from AG-CT genotype suggesting a combined role of both 1635A and 1486C TLR9 polymorphisms in HIV mediated CD4 decline and immune activation.