Previously Unidentified Single Nucleotide Polymorphisms in HIV/AIDS Cases Associate with Clinical Parameters and Disease Progression

The genetic background of an individual plays an important role in the progression of HIV infection to AIDS. Identifying previously unknown or uncharacterized single nucleotide polymorphisms (SNPs) that associate with disease progression may reveal important therapeutic targets and provide a greater understanding of disease pathogenesis. In the present study, we employed ultra-high multiplex PCR on an Ion Torrent next-generation sequencing platform to sequence 23 innate immune genes from 94 individuals with HIV/AIDS. This data was used to identify potential associations of SNPs with clinical parameters and disease progression. SNPs that associated with an increased viral load were identified in the genes for the interleukin 15 receptor (IL15RA), toll-like receptor 7 (TLR7), tripartite motif-containing protein 5 (TRIM5), and two killer-cell immunoglobulin-like receptors (KIR2DL1 and KIR2DL3). Additionally, SNPs that associated with progression from HIV infection to AIDS were identified in two 2′-5′-oligoadenylate synthetase genes (OAS2 and OAS3). In contrast, other SNPs identified in OAS2 and OAS3 genes, as well as in the TRIM5 and KIR2DS4 genes, were associated with a slower progression of disease. Taken together, our data demonstrates the utility of ultra-high multiplex PCR in identifying polymorphisms of potential clinical significance and further,identifies SNPs that may play a role in HIV pathogenesis.


Introduction
Acquired immunodeficiency syndrome (AIDS) is a devastating disease caused by the human immunodeficiency virus (HIV). AIDS is defined as having a CD4 + T cell count below 200 cells per L or displaying specific clinical presentations in association with an HIV infection [1]. On average, most individuals infected with HIV will progress to AIDS within 10 years if antiretroviral therapy is not administered.
It is evident that the clinical manifestations of HIV infection vary in different individuals. Previous studies have reported that these differences are not dependent on gender, mode of infection, or age [2]. Therefore, genetic variations in genes that control immune responses have been suggested to play a role in the progression of HIV or lack thereof [3,4]. Indeed, innate immunity plays an important role in controlling HIV replication [5]. Upregulation of type I interferon (IFN) is essential to restrain HIV replication, 2 BioMed Research International especially at the early stages of infection [6]. The activation of IFN expression occurs primarily through the engagement of toll-like receptors (TLR) 7 and 9 by single stranded RNA or CpG DNA, respectively (reviewed by [7]). It has been demonstrated that single nucleotide polymorphisms (SNPs) in TLRs are associated with an increased susceptibility to infection [8][9][10][11] and are involved in different progressions of HIV/AIDS, although a consensus has not been reached. For example, Papadopoulos et al. demonstrated an association between a TLR4 polymorphism and increased risk of opportunistic infection in subjects with advanced HIV-1 infection [8]. However, a subsequent study by Soriano-Sarabia et al. could not confirm these results, although these investigators did identify a polymorphism in TLR9 that was associated with a lower CD4 count and a higher viral load in HIV-positive cases [12].
Numerous reports describe impaired immunomodulatory and cytotoxic activity of natural killer (NK) cells in HIV-1-infected individuals [5,[13][14][15]. Additionally, HIV infections are often characterized by increased expression of inhibitory receptors and downregulation of activating receptors [16,17]. Killer-cell immunoglobulin-like receptors (KIRs) regulate NK activation status. Interestingly, NK cells expressing the activating KIR3DS1 inhibit HIV-1 replication in target cells [18]. Therefore, it could be suggested that differential expression of various KIRs may impact progression and outcome of HIV infection. Although the presence of some SNPs can alter the expression and function of KIRs, our knowledge regarding polymorphism of KIRs in HIV subjects remains limited.
A significant body of evidence suggests that the innate immune system plays an important role in susceptibility to HIV infection and disease progression. For this reason, genetic polymorphisms in genes of innate immune pathways may also influence the course of the disease. In the present study, we focused our attention on several key genes involved in regulation of the innate immune response. We employed ultra-high multiplex PCR on an Ion Torrent next-generation sequencing platform to sequence 23 innate immune genes of 94 individuals with HIV/AIDS. These data were then used to identify potential associations of SNPs with clinical parameters and disease progression. Of the 649 SNPs identified in this study, SNPs which associated with viral load were identified in the following genes at the respective chromosomal positions: IL15RA (rs2229135; chr10:5995052), TLR7 (rs179008, chrX:12903659), TRIM5 (rs11601507; chr11:5701074), KIR2DL1 (rs77397437; chr19:55286864), and KIR2DL3 (chr19:55251049). Further analysis revealed that the C/T genotype of IL15RA (rs2229135; chr10:5995052), the A/T genotype of TLR7 (rs179008; chrX:12903659), the A/C genotype of KIR2DL1 (rs77397437; chr19:55286864), the A/G genotype of KIR2DL3 (chr19:55251049), and the C/A genotype of TRIM5 (rs11601507; chr11:5701074) were associated with increased viral load in HIV cases. Furthermore, two SNPs; one in OAS2 (rs2072137; chr12:113440921) and the other in OAS3 (chr12:113376388), were identified that associated with progression of HIV infection. Additionally, three SNPs associated with disease progression included TRIM5 (rs11038628; chr11:5688940;), KIR2DS4 (chr19:55358734) as well as the SNP in IL15RA (rs2229135; chr10:5995052). These data further support the role of the innate immune response in maintaining viral persistence and disease progression in HIV-infected individuals.   Table 2. The library after purification, equalization, and the template prep was processed for the Ion 318 Chip on the Ion OneTouch system (Thermo Fisher Scientific, Waltham, MA). Final sequencing was conducted using the Ion PGM Next-Generation Sequencing Platform (Thermo Fisher Scientific, Waltham, MA). Reads from Ampliseq amplicon sequencing were aligned to the reference human genome (NCBI37/hg19) using the Bowtie2 alignment tool [19]. Single nucleotide variants were called at each amplicon using SAMtools and BCFtools [20]. Possible genotypes and their likelihoods were computed at each position with SAMtools mpileup, from which variants and genotypes were called using BCFtools [19].

Statistical Analysis
Six hundred forty-nine polymorphisms were identified across genotypes of 94 of the HIV samples using Ampliseq sequencing technology. Analyses of identified polymorphisms revealed 261 SNPs exhibited a minor allele frequency of 3.5% [94 HIV samples + CF1 + CFS2 + USCTR = 160]. Next, 261 SNPs were examined for association with clinical parameters including sCD14, viral load, and progressor status.
To identify association between genotype and sCD14 or viral load measures, a simple linear model approach was performed between each clinical parameter and the genotypes of the HIV samples. These values were then adjusted using a multiple testing correction (false discovery rate) [21], and any adjusted values with < 0.05 were considered statistically significant.
Simple Chi-squared tests were performed to establish any relation between the genotypes of the 94 HIV samples and their progressor status. The values of the 261 Chi-squared tests were not adjusted.

Clinical Presentation of HIV Cases. Diagnosis of HIV infection was established by detection of anti-HIV antibodies
using ELISA and western blot methods. The following criteria were used for patient enrollment: (a) detection of anti-HIV antibody; (b) CD4 counts ≥ 350 cell/ L at the time of enrollment; and (c) absence of opportunistic infections at the time of enrollment. Subjects were followed up for five years after enrollment to determine progression of the disease.
A total of 94 subjects were enrolled in this study (  nonnucleoside reverse transcriptase inhibitors, and protease inhibitors. Average HIV RNA viral load was 39925.9 copies/mL; 55 cases (58.5%) had a viral load higher than 250 copies/mL and the remaining 39 cases (41.5%) had an undetectable viral load or less than 250 copies/mL. Based on CD4 counts and the rate of their decline, all cases were separated into two groups: group 1 "typical progressors" and group 2 "slow progressors." Group 1, typical progressors, was characterized by a decline of CD4 counts >50 cells/ L annually with overall CD4 counts < 350 cells/ L during the five years of observation [22][23][24]. During this fiveyear period, subjects in group 1 lost on average 250 cells/ L CD4 cells. Group 2, slow progressors, was characterized by <50 cells/ L annual decline in CD4 counts with overall CD4 counts > 350 cells/ L. Group 1 included 55 cases (58.5%), while group 2 contained 39 cases (41.5%).

Ampliseq Analysis.
Twenty-three genes were selected for Ampliseq analysis ( Table 2). The selected genes were chosen as mediators of innate immune responses and their corresponding intracellular pathways. Using the Ampliseq approach, complete sequences of each gene of interest were obtained allowing the identification of known, as well as previously unidentified SNPs. A total of 649 SNPs were identified, which were not present in the reference genome or in any of the 48 non-HIV controls. Correlation analyses were performed between each of the identified SNPs and antibody titer to LPS and LPS glycolipid, CD4 counts, viral load, and sCD14 level (Table 3). Data on serum level of sCD14 was available for 59 HIV cases.

Discussion
Without antiretroviral therapy, the majority of those infected with HIV develop uncontrolled viremia and undergo progressive immune impairment ultimately leading to AIDS within three years of their diagnosis [24,25]. However, a small number of HIV-infected individuals are characterized by a slow progression of HIV infection where CD4 counts are maintained above 350 cells/ L without the development of AIDS for decades [26][27][28]. Several mechanisms have been suggested to explain this phenomenon including viral mutation, immune response, and host genetic factors. Since the virus relies heavily on the host's cellular machinery for replication, initial studies have focused on the host's genetic factors in an effort to explain these observations. For instance, screening of genetic variants in selected genes led to the discovery of a 32 bp deletion (CCR5Δ32) in the C-C chemokine receptor type 5 (CCR5) gene, which codes for a coreceptor used by the virus to enter the host cells [29,30]. It has been demonstrated that the CCR5Δ32 genotype delays HIV progression in heterozygous individuals and prevents infection in homozygous individuals [31,32]. Discovery of the CCR5Δ32 genotype provided compelling evidence to support the role of host genetic factors in the pathogenesis of HIV and AIDS. Since then, many studies have been conducted to identify additional genetic markers that influence disease progression.
Using a combination of analyses including genome-wide association studies (GWAS), many SNPs have been identified that associate with different forms of HIV progression. For example, it has been shown that long-term nonprogressors have a higher frequency of the HLA B * 5701, B5703, and HLA B * 2705 alleles [28,33,34]. Furthermore, Pereyra et al. proposed that the B14, B52, Bw4, and Cw1402 haplotypes were protective [33,35]. It has been suggested that the HLA-B allele is more protective against HIV-1 when compared to the HLA-A allele. Studies of HLA and killer-cell immunoglobulin-like receptors revealed that individuals expressing KIR2DS2 have a more rapidly declining CD4 population and progression to AIDS than those who do not [36]. Conversely, the genotype KIR3DS1/HLA-B Bw4-80I, which stimulates NK cell activation, has been shown to promote virus containment and delay progression to AIDS [37,38].
In this study, we have analyzed the complete sequence of 23 innate immune genes from 94 individuals with HIV/AIDS. The genes studied were divided into three groups: one group pertained to NK cell function and included genes of KIR family as well as IL15RA. The second group encompassed genes coding for the intrinsic antiviral factors OAS, RNASEL, TLRs, and TRIMs. Finally, the third group included the downstream type I interferon response transcription factors TRIF, MYD88, TRAF6, and IRF7. The Ampliseq method employed in this study utilizes a high throughput sequencing approach to identify known as well as novel SNPs. Correlation analyses were performed between the 649 identified SNPs and clinical laboratory findings such as CD4 counts, anti-LPS antibody, anti-LPS glycolipid antibody, and viral load. Selection of clinical parameters was based on their role in disease progression and outcome. Changes in CD4 counts determine the course of HIV infection and establish a starting point for HAART [39]. Presence of anti-LPS and anti-LPS glycolipid antibodies reflect the integrity of gut epithelium, which is affected as HIV infection progresses [40]. Viral load is a diagnostic marker for HIV infection as well as being used to monitor disease progression and efficacy of HAART [41,42].
We observed the following SNPs, at the respective chromosomal positions, to be associated with viral load: IL15RA (rs2229135; chr10:5995052), TLR7 (rs179008; chrX:12903659), TRIM5 (rs11601507; chr11:5701074), KIR2DL1 (rs77397437; chr19:55286864), and KIR2DL3 (chr19:55251049). The SNP rs2229135 of IL15RA occurs in the three prime untranslated region (3 -UTR) of the gene and thus does not result in an amino acid change. However, SNPs that occur in this region can influence translation efficiency, polyadenylation, localization, and mRNA stability of their respective gene [43,44]. Given the role of IL15 signaling in NK cell function, altered translation of IL15RA may broadly impact NK-cellmediated antiviral immunity. In contrast to the IL15RA polymorphism, the SNPs observed in TLR7 and TRIM5 both result in missense substitutions. The SNP rs179008 in TRL7 leads to a nonsynonymous Gln11Leu substitution within exon of the gene. Previous studies have reported this polymorphism is associated with chronic hepatitis C virus infection and systemic lupus erythematosus [45,46]. Additionally, in support of a role for this polymorphism in HIV and in further support of our observations, Beima-Sofie et al. reported an association between a the TLR7 32T (rs179008) allele and time to mortality in female infants infected [47]. The SNP rs11601507 in TRIM5 leads to a nonsynonymous Val112Phe missense substitution; however, previous studies addressing the potential role for this SNP and HIV found no association [48,49]. However, other SNPs identified in the TRIM5 report that some SNPs in this gene may be protective [50]. Notwithstanding, our data support the notion that cellular mechanisms involving TRIM5 and TLR7 are important in controlling HIV infection. Two significant SNPs were identified in two KIR genes that also associated with viral load-KIR2DL1 (rs77397437; chr19:55286864) and KIR2DL3 (chr19:55251049). The rs77397437 SNP in KIR2DL1 leads to the synonymous substitution Pro206, while the previously nonannotated SNP KIR2DL3 (chr19:55251049) occurs in an intron. Although other studies have identified KIR associations with HIV infection, to the best of our knowledge, our data represents the first association with these specific polymorphisms and HIV viral load.
Also occurring within introns, two additional SNPs showed significance when analyzed in relation to disease progression-OAS2 (rs2072137; chr12:113440921) and OAS3 (chr12:113376388). We observed the OAS2 T/C and C/C genotypes (rs2072137; chr12:113440921) to be associated with "typical" disease progression; however, the T/T genotype presented with higher frequency in the slow progressors. It is well documented that SNPs residing within introns, or those upstream or downstream of genes, also have the capacity to be causal [51][52][53][54]. In fact, in a recent study, Farh and colleagues utilized a fine-mapping algorithm to analyze GWAS data for 21 autoimmune diseases and reported that approximately 90% of all causal variants map to noncoding regions [55]. They further reported that only 10-20% of causal SNPs directly alter recognizable transcription factor binding motifs.
Our results also provide evidence of an association between various KIRs and HIV progression. A large body of evidence suggests a role for KIR3DS1 and its HLA ligand Bw4 in protection from HIV infection and delayed HIV progression [56,57]. For instance, Alter et al. demonstrated that KIR3DS1+ NK cells are more potent in killing HIV-infected Bw4-80I+ CD4 T cells [18]. Additionally, an epidemiological study by Martín et al. revealed an association between slower HIV progression and the expression of KIR3DL1 [58]. It has been suggested that KIR3DL1 could play a role in "educating" NK cells, with higher expression of this KIR, promoting the generation of a larger pool of functionally competent NK cells, thus establishing a more vigorous response towards infection [59,60]. Previous data pertaining to SNPs in KIRs has been limited to demonstrating that the variations in KIR2DS3 were associated with sustained viral response to pegylated-IFN and ribavirin treatment in HIV cases coinfected with HCV [61]. In the present study, we identified SNPs in KIR2DL1, KIR2DL3, and KIR2DS4 that associated with maintenance of viral load and disease progression. However, further studies will be required to firmly establish the relationship these SNPs and the course of the disease.
The role of TRIM5 in controlling HIV replication is well established [62][63][64]. Previous studies have shown that the presence of SNPs in TRIM5 exon2 and linker regions attenuates its ability to control viral replication [65,66]. Our data provides further support for a TRIM5 involvement in controlling HIV replication and disease progression. We identified two SNPs, (rs11601507; chr11:5701074) and (rs11038628; chr11:5688940), which have a greater frequency in HIV cases with high viral load and "typical" progression of the disease. Furthers studies of HIV cases with these SNPs may provide additional support for their potential application as biological markers of disease progression.
Perhaps our most intriguing observation was the association between viral load, disease progression, and IL15RA. IL15RA binds specifically to IL-15, which is required for the differentiation of NK cells, CD8+ lymphocytes, and memory CD8+ T cells [67,68]. Importantly, Naora and Gougeon demonstrated that when used in vitro, IL-15 stimulates the proliferation of CD56+, CD16+, CD4+, and CD8+ T cells from HIV-infected individuals [69]. They also showed that IL-15 was more potent than IL-2 as a survival factor for CD56+ cells, and this effect was associated with upregulation of Bcl-2 expression. Taken together, these findings indicated that IL-15 plays a pivotal role in survival and proliferation of NK cells in control of HIV maintenance and disease propagation. Our SNP data with regard to the IL15RA identifies a potential mechanism for the regulation of NK cell activity in HIV-infected individuals, in addition to presence of KIR SNPs. However, these previously unreported observations would require further in vitro and clinical evaluation to confirm their involvement in HIV disease progression.
Our data on the association of SNPs in genes coding for OAS2 and OAS3 with disease progression represents an additional and potentially novel finding. The OAS enzymes are interferon-inducible proteins required for the activation of RNase L [70]. While they become activated by the transactivation responsive region of HIV-1 mRNA, OAS activation is prevented by viral Tat binding to TAR [71][72][73]. Furthermore, RNase L inhibitor is activated in HIV-1-infected cells downregulating the OAS/RNase L pathway [74]. Therefore, it is not surprising that OAS enzymes are activated in HIV-infected cells as part of innate antiviral response and the efficacy of antiviral defense will, in part, depend on the ability of infected cells to prevent OAS inhibition. Our data on association between HIV disease progression and SNPs in OAS genes provides additional evidence of a role for these genes in HIV progression; however, additional studies are needed to firmly establish importance of these SNPs in HIV pathogenesis.
The prevalence of a given allele, whether it is protective or causative, may contribute to the prevalence of a disease. The Global Minor Allele Frequency (MAF) of rs2229135 for IL15RA is 0.09 [75]. In contrast, the same allele has a frequency of 0.11 in African populations, 0.05 in East Asian populations, and 0.08 in European populations [75]. The MAP of rs2072137 for OAS2 has minor allele frequency in America, East Asia, and Europe of 0.48, 0.38, and 0.42, respectively, but only 0.06 in African populations [75]. Even more striking, the MAF of rs179008 for TLR7 is 0.12 and a minor allele frequency of 0.23 in European populations but is only 0.05 in Southern Asian populations and is virtually zero in East Asian populations [75]. Larger studies that investigate clinical presentations, such as disease progression, in concert with genetic screening, may provide further support for the role of these genes in the pathogenesis of HIV/AIDS.

Conclusions
In conclusion, we identified association between previously unreported SNPs in IL15RA, TLR7, TRIM5, KIR2DL1, and KIR2DL3 and an increased viral load. Additionally, novel SNPs that associated with progression from HIV infection to AIDS were identified in OAS2 and OAS3 genes. In contrast, other SNPs identified in OAS2 and OAS3 genes, as well as in the TRIM5 and KIR2DS4 genes, were associated with a slower progression of disease. Further studies will determine the significance of identified SNPs in pathogenesis of HIV infection and AIDS.