VDR rs2228570 Polymorphism Is Related to Non-Progression to AIDS in Antiretroviral Therapy Naïve HIV-Infected Patients

Background: Vitamin D is a fundamental regulator of host defenses by activating genes related to innate and adaptive immunity. In this study, we analyzed the association among single nucleotide polymorphisms (SNPs) in the vitamin D receptor (VDR) gene, with clinical patterns of AIDS progression in antiretroviral treatment (ART)-naïve HIV-infected patients. Methods: We conducted a retrospective study in 667 HIV-infected patients, who were classified within three groups according to their AIDS progression pattern (183 long-term non-progressors (LTNPs), 334 moderate progressors (MPs), and 150 rapid progressors (RPs)). Five VDR SNPs (rs11568820, rs4516035, rs2228570, rs1544410, and rs7975232) were genotyped using Agena Bioscience’s MassARRAY platform. Results: Significant association results were found for rs2228570. Within all HIV patients, the presence of T allele at VDR rs2228570 SNP was protective against AIDS progression (ordinal outcome) under additive (adjusted odds ratio (aOR) = 0.75; p = 0.009), dominant (aOR = 0.69; p = 0.015), and codominant (aOR = 0.56; p = 0.017) inheritance models. In addition, the same allele was protective under additive and codominant inheritance models when we compared with LTNPs vs. RPs [aOR = 0.64 (p = 0.019) and aOR = 0.37 (p = 0.018), respectively] and when we compared MPs vs. RPs [aOR = 0.72 (p = 0.035) and aOR = 0.45 (p = 0.028), respectively]. Conclusions: The VDR rs2228570 T allele was related to a lower AIDS progression pattern in ART-naïve HIV-infected patients. These findings expand upon the knowledge about HIV pathogenesis in untreated HIV-infected patients with different clinical outcomes.

that these SNPs could play an important role in the clinical progression of HIV-infected patients, and could be associated with clinical patterns of AIDS progression.

Objective
We aimed to analyze the association among five VDR SNPs and the clinical patterns of progression, categorized as LTNPs, MPs, and RPs, for the natural history of HIV infection in European ART-naïve HIV-infected patients from two large Spanish cohorts (Cohort of the Spanish AIDS Research Network (CoRIS) and Cohort of LTNPs).

Patients
We conducted a retrospective study in 667 European ART-naïve HIV-infected patients from both CoRIS and LTNPs cohorts, whose samples were available at the Spanish HIV HGM Biobank [19]. The methodological and organizational aspects, and the characteristics of the subjects enrolled in the present study, have been previously described [20]. Moreover, 113 healthy blood donors (HIV, HCV, and HBV negative subjects) from "Centro de Transfusión de la Comunidad de Madrid" were used as a Control-group. The programs were approved by the institutional review boards of each participating center. In addition, the Research Ethics Committee of the Instituto de Salud Carlos III (ISCIII) approved this study (CEI PI_2010-v3), which was conducted under the ethical principles of the Declaration of Helsinki. All patients signed an informed consent form.
Patients infected with HIV were classified within three groups depending on their drop in CD4 T-cell counts and AIDS disease progression rates in the absence of ART [21,22]: (a) 183 LTNPs, who had asymptomatic HIV infection over ten years after seroconversion, and always showing CD4+ ≥500 cells/mm 3 and RNA-viral load ≤10,000 copies/mL. (b) 334 MPs who had an average decrease of 50-100 CD4+/mm 3 per year for at least two years after HIV diagnosis. (c) 150 RPs, who had two or more CD4+ T-cell values below 350/mm 3 (recommended threshold for initiating ART) and/or who had AIDS or AIDS-related death within three years after HIV seroconversion (seroincidents). These are three well-defined groups with available sample in the Spanish HIV Biobank, which show the three main patterns of progression to AIDS. All patients had a regular follow up of CD4+ counts and plasma HIV-RNA during the study period.

DNA Genotyping
We selected five VDR SNPs, which had been previously related to the risk of acquisition and progression of HIV infection [16][17][18]23,24]. These SNPs are located upstream (rs11568820 and rs4516035), within the coding region (rs2228570 [missense variant]), and at two intronic region (rs1544410 and rs7975232).
Samples from patients were kindly provided by the HIV BioBank integrated into the Spanish AIDS Research Network (RIS). The samples were immediately processed and frozen after their reception. The total DNA isolation was performed from peripheral blood mononuclear cells with Qiagen columns (QIAamp DNA Blood Midi/Maxi; Qiagen, Hilden, Germany). DNA genotyping was carried out at the Spanish National Genotyping Center (CeGen; http://www.cegen.org/) by Agena Bioscience's MassARRAY platform (San Diego, CA, USA) with the iPLEX ® Gold assay design system.

Statistical Analysis
For the descriptive study, categorical data and proportions were analyzed using the chi-squared test or Fisher's exact test. Kruskal-Wallis and Mann-Whitney U tests were used to compare continuous variables among independent groups.
The genetic association analysis was carried out according to dominant, recessive, over-dominant, codominant, and additive models. The Mantel-Haenszel linear-by-linear association test (values between −1 and +1), a non-parametric test for measuring the association between ordinal data, was used to explore the distribution of VDR genotypes among the three clinical patterns of AIDS progression. We also used Generalized Linear Models (GLMs) adjusted by age at HIV diagnosis, gender, and risk group (men who have sex with men (MSM) versus others). Firstly, a GLM model, with a multinomial distribution (cumlogit-link), was used to evaluate the association of VDR SNPs with the three clinical patterns of AIDS progression (ordinal variable: 0, LTNP; 1, MP; 2, RP). The reference category was the LTNP-group. The goodness of fit of each model was evaluated by the Akaike information criterion (AIC) value and Bayesian information criterion (BIC). In order to exclude spurious associations, multiple testing correction was carried out by the false discovery rate (FDR) with the Benjamini and Hochberg procedure. Secondly, a GLM model with a binomial distribution (logit-link) was used for estimating frequency differences of VDR SNPs between clinical patterns of AIDS progression (dichotomous variable: LTNPs vs MPs; LTNPs versus RPs; MPs versus RPs). In this case, only SNPs with an adjusted p-value less than 0.05 (after an FDR correction) from the ordinal regression (Step 1) were analyzed. All statistical tests were performed with the Statistical Package for the Social Sciences (SPSS) 22.0 software (IBM Corp., Armonk, CA, USA). Statistical significance was defined as p < 0.05. All p-values were two-tailed.  Table 1 shows the epidemiological and clinical characteristics of HIV infected patients (183 LTNPs, 334 MPs, and 150 RPs). LTNP-patients had the highest values of age at the moment of HIV diagnosis, age at the study inclusion, percentage of intravenous drug users (IDU) (p < 0.001), and the lowest proportion of male patients (p < 0.001). Besides, MP and RP groups were diagnosed with HIV infection in the 2000s and had higher percentages of men and homosexual HIV transmission than LTNP group.

Characteristics of VDR SNPs
All VDR SNPs (rs11568820, rs1544410, rs2228570, rs4516035, and rs7975232) were in HWE (p > 0.05), the values for the minor allelic frequency (MAF) were higher than 5%, and the genotyping call-rate success was over 95% (Table 2). Similar values of genotypic frequencies for VDR SNPs were found in healthy-controls and HIV-infected patients (Table 2). Furthermore, the VDR genotype frequencies were in accordance with the NCBI SNP database (http://www.ncbi.nlm.nih.gov/projects/ SNP). Moreover, LD values were very high for rs7975232 and rs1544410 (D' = 1.0), and rs11568820 and rs4516035 (D' = 1.0); but the r 2 statistic was very low among all VDR SNPs (Figure 1), which means that each SNP provides different information.

VDR SNPs and AIDS Progression
We explored the genetic association between VDR SNPs (rs11568820, rs1544410, rs2228570, rs4516035, and rs7975232) and the three clinical patterns of AIDS progression (LTNP, MP, and RP) by univariate analysis (Figure 2). We found a significant association for only rs2228570 SNP under an additive (p = 0.010), dominant (p = 0.031), recessive (p = 0.040), and codominant (p = 0.014) inheritance models. However, after the FDR (Benjamini & Hochberg) controlling procedure, no significant p-values were obtained. Despite this, we do not necessarily think that this association between VDR rs2228570 SNP, and the clinical patterns of AIDS progression, could be a false positive. Correcting for multiple testing is generally a very strict methodology to maintain the statistical significance for certain SNPs that could be associated with the outcome variable. Furthermore, we also observed that the distribution of patients in each of the three categories of progression to AIDS, according to genotypes, showed a prominent "effect size" of the T allele for rs2228570.
statistical significance for certain SNPs that could be associated with the outcome variable. Furthermore, we also observed that the distribution of patients in each of the three categories of progression to AIDS, according to genotypes, showed a prominent "effect size" of the T allele for rs2228570. Afterwards, the GLM tests adjusted for age, gender, and risk category largely confirmed these findings ( Figure 3). Therefore, when analyzing all 667 HIV patients, the presence of T allele at VDR rs2228570 SNP protected against AIDS progression (ordinal outcome) under additive (adjusted odds ratio (aOR) = 0.75; p = 0.009), dominant (aOR = 0.69; p = 0.015), and codominant (aOR = 0.56; p = 0.017) inheritance models. In this case, p-values were also corrected by multiple comparison (FDR -Benjamini & Hochberg), and remained significant.

Discussion
In this study, we evaluated the distribution of VDR SNPs (rs11568820, rs1544410, rs2228570, rs4516035, and rs7975232) among European ART-naïve HIV-infected patients, from two large Spanish cohorts (CoRIS and LTNPs cohorts), grouped by their clinical progression pattern (LTNPs, MPs, RPs). Our most significant finding was that, HIV-infected patients carrying the VDR rs2228570 T allele, had lower odds of progression to AIDS.
VitD induces an antiviral effect in the body, that promotes the expression of antiviral genes, reduces the expression of CCR5 on CD4+ T-cells, promotes an HIV-1-restrictive CD38+HLA-DR+ phenotype, and reduces the ability of TNFα to upregulate HIV replication in latently infected CD4+ T-cells [25,26]. The variability that the genetic background of patients may confer differential

Discussion
In this study, we evaluated the distribution of VDR SNPs (rs11568820, rs1544410, rs2228570, rs4516035, and rs7975232) among European ART-naïve HIV-infected patients, from two large Spanish cohorts (CoRIS and LTNPs cohorts), grouped by their clinical progression pattern (LTNPs, MPs, RPs). Our most significant finding was that, HIV-infected patients carrying the VDR rs2228570 T allele, had lower odds of progression to AIDS.
VitD induces an antiviral effect in the body, that promotes the expression of antiviral genes, reduces the expression of CCR5 on CD4+ T-cells, promotes an HIV-1-restrictive CD38+HLA-DR+ phenotype, and reduces the ability of TNFα to upregulate HIV replication in latently infected CD4+ T-cells [25,26]. The variability that the genetic background of patients may confer differential susceptibility to HIV infection and progression to AIDS [5]. The five VDR SNPs analyzed in our study (rs11568820, rs1544410, rs2228570, rs4516035, and rs7975232) were previously involved in protection against HIV infection [18,23,24]. Four of them (rs11568820, rs1544410, rs2228570, and rs4516035) have been related to the AIDS progression in Spanish and Indian cohorts [15][16][17][18]. However, we have only confirmed the significant association of rs2228570 with AIDS progression, in accordance with the findings of Nieto et al. [17] and Laplana et al. [16]. In contrast, Barber et al. [15] and Torres et al. [18] did not analyze the VDR rs2228570 SNP. Therefore, rs2228570 polymorphism seem to have a relevant impact in the natural history of our ART-naïve HIV-infected patients, although we should not rule out that other SNPs in high LD, with rs2228570, could be the causal polymorphism.
Previous retrospective studies in very similar populations [16,17] found a weak association in the VDR rs2228570 polymorphism. Our data has the key advantage of a much larger sample size and confirms the locus, although the most significant associations came from a slightly different genetic model. We therefore believe that the VDR rs2228570 polymorphism might be relevant for HIV progression, although the exact model (additive, recessive, dominant, over-dominant or codominant) remains to be determined. Even without a specific model, this study warrants further investigation on the potential role of VDR SNPs, and this one in particular as a partial determinant of HIV progression.
The rs2228570 C->T polymorphism (also called FokI because may be detected by Restriction Fragment Length Polymorphism (RFLP) using the endonuclease FokI is located in the 5 untranslated region (UTR), three codons before the start codon (ATG) [27,28]. Thus, the C allele (F variant) does not originate any new start codon (ATG), whereas the T allele (f variant) creates a new start codon and a VDR protein, with three extra amino acids [28]. However, discrepant results have been published around the putative implication of rs2228570 polymorphism on the VDR expression and vitamin D signaling pathway. On the one hand, the T allele has been related to a less efficient interaction between VDR protein and transcription factor IIB (TFIIB) [29], and to a lower VDR mRNA copy number [30]. This might be due to the three extra amino acids of the final protein, which could be disrupting a correct interaction. On the other hand, Oskooei et al. [31] described that the TT genotype was associated with a higher VDR expression levels in breast cancer patients. Additionally, there are also studies that have described a lack of association between the VDR rs2228570 polymorphism and VDR expression in immune cells of European, and South African Populations [32,33], respectively. We did not analyze the effect of rs2228570 SNP at the functional level. However, we hypothesize that the effect of the T allele might cause a decrease in the level of expression of VDR and the impairment of its function, which would induce the inhibition of HIV replication and block the destruction of CD4+ T cells.

Limitations of the Study
Firstly, this work is a retrospective study that included three groups of patients (defined by changes in CD4 T-cell counts and AIDS disease progression rates) with substantial differences in demographic and clinical characteristics. Thus, LTNPs acquired the HIV infection, mainly by IDU, and were diagnosed at the beginning of the 90s; while MPs and RPs were mainly MSM that were diagnosed of HIV infection more recently (2000s), these patients were predominantly men, with a younger age at the time of HIV diagnosis. These differences could introduce some bias in the study. Besides, the sample size was limited in some sub-groups of patients, which may influence the statistical significance level when the effects of the variables studied are not significant.
Secondly, our study was completely performed in Spanish Caucasians individuals. Thus, since the frequency of the alleles varies among races, an independent study with different ethnic groups would be interesting in clarifying the association between different genetic backgrounds. We should not rule out the influence of other factors, such as the exposure to sunlight, since vitamin D levels depend on the conversion rate of pro-vitamin D3 to pre-vitamin D3 by sunlight, which is affected by seasonal characteristics, latitude, radiation, and urban living. In our study, HIV-infected patients came from all different regions of Spain.
Thirdly, the HCV coinfection data were available for 75% (500/677) of patients, but the data of HBV coinfection, opportunistic infections, and other non-AIDS-related co-morbidities (e.g., cancer, liver disease, renal impairment, diabetes, obesity, endocrine disease, and cardiovascular disease) were not available. Thus, these co-factors could not be included in the multivariate analysis to account for these effects.

Conclusions
In conclusion, the VDR rs2228570 T allele was related to a lower AIDS progression rate in European ART-naïve HIV-infected patients. These findings expand the knowledge about HIV pathogenesis in untreated HIV-infected patients with different clinical outcomes. Funding: This work has been (partially) funded by the RD16/0025/0019 and RD16CIII/0002/0002, projects as part of Acción Estratégica en Salud, Plan Nacional de Investigación Científica, Desarrollo e Innovación Tecnológica (2013-2016) and cofinanced by Instituto de Salud Carlos III (Subdirección General de Evaluación) and Fondo Europeo de Desarrollo Regional (FEDER), RETIC PT17/0015/0042, Fondo de Investigación Sanitaria (FIS) (grant number PI16/01863, PI17/01115, PI17CIII/00003), EPIICAL Project and Comunidad de Madrid B2017/BMD-3703. Programa de Investigación de la Consejería de Sanidad de la Comunidad de Madrid to JLJ. CIBER-BBN is an initiative funded by the VI National R&D&i Plan 2008-2011, Iniciativa Ingenio 2010, the Consolider Program, and CIBER Actions and financed by the Instituto de Salud Carlos III with assistance from the European Regional Development Fund. COST CA17140 Cancer Nanomedicine-Front the Bench to Bebside. AFR are supported and funded by "Instituto de Salud Carlos III" [grant number CP14/0010].