Polymorphism in the interleukin-10 gene is associated with overactive bladder phenotype associated with HTLV-1 infection

Human T-cell lymphotropic virus type 1 (HTLV-1) is a retrovirus with two main clinical manifestations: Adult T-cell leukemia/lymphoma (ATLL) and HTLV-1 associated Myelopathy/Tropical Spastic Paraparesis (HAM/TSP), a demyelinating disease of the spinal cord, in addition to other inflammatory diseases that impact the morbidity of people infected with this virus¹. It is estimated that about 10 million people are infected with HTLV-1 globally². The virus is endemic to Japan, the Caribbean Basin, and in some Latin American countries. In Brazil, HTLV-1 infection is most prevalent in the Northeast region of the country². The vast majority of people infected with HTLV-1 remain asymptomatic, with approximately 5-10% developing clinical symptoms². Host factors such as genetics, proviral load, and immune response may contribute to the development of ATLL, HAM/TSP, or other clinical manifestations such as urinary complaints in HTLV-1 positive individuals who develop an overactive bladder^3,4. Overactive bladder is considered an early sign of HAM/TSP and therefore these patients are identified as probable HAM/TSP cases according to the criteria proposed by De Castro-Costa⁵. The HTLV-1 virus has a tropism for CD4⁺and CD8⁺ T cells, but other cells can also be infected⁶. Cultures of peripheral blood mononuclear cells (PBMC) from HAM/TSP patients show increased spontaneous production of pro-inflammatory cytokines, such as IFN-γ and TNF, lower production of IL-10, and a high proviral load compared to asymptomatic carriers⁶. Nevertheless, despite this knowledge of HTLV-1 immunopathogenesis, there are few studies describing the role of host genetic factors in HTLV-1 infection. Of the documented genes, IL28B, IL10, and IL6, are used as biomarkers across different populations^7–10. In this context, the study of immune response genes may reveal new markers of susceptibility or resistance to infection and disease. Here, we investigate whether polymorphisms in the IL10 gene are associated with genetic susceptibility to HTLV-1 infection in a population from the city of Salvador, Northeast Brazil.

HTLV-1 infected patients (n =298) of both sexes, aged 18 to 78 years, were enrolled at Complexo Hospitalar Universitário Professor Edgard Santos (ComHUPES), an outpatient referral center for the treatment of HTLV-1 in the city of Salvador-Bahia, Brazil. The participants were divided into three groups according to their clinical form: Group 1 – 164 asymptomatic HTLV-1 carriers; Group 2 – 63 HTLV-1 infected individuals with symptoms of overactive bladder; and Group 3 – 71 HTLV-1 infected patients diagnosed with HAM/TSP. A group of 380 volunteer blood donors recruited at Fundação de Hematologia e Hemoterapia da Bahia (HEMOBA) were used as healthy controls for comparison with the HTLV-1 infected group. This control group comprised 264 males and 116 females, with a mean age of 34.80 ± 10.24 years. This study was approved by the ethical com­mittee of the Maternidade Climério de Oliveira – Universidade Federal da Bahia (N° 35/2013). Demographic data (age, sex) by phenotype are provided in Table 1. Three single nucleotide variants (SNVs) of the IL10 gene, rs3024496 (+4976 A/G), rs1800871 (-819bp A/G), and rs1800896 (-1082bp A/G), were genotyped by TaqMan RT-PCR (Thermo Fisher®) using previously designed genotyping assays. Table 2 shows the allele frequencies for the SNVs in HTLV-1 infected cases and blood bank controls. Tests to check for Hardy-Weinberg equilibrium, and an unconditional logistic regression analysis to determine allele-wise (1 df test) and genotype-wise (2 df test) associations between IL10 SNPs and HTLV-1 clinical phenotypes, were carried out in STATA 8.2 using the GenAssoc package. For stratified analyses, we corrected for multiple testing by multiplying our p-values by 9 to take the 3 SNVs x 3 clinical phenotypes into account. When comparing HTLV-1 infected individuals with blood bank controls, we found no associations between the SNVs at IL10 +4976 A/G, -819bp A/G, or -1082bp A/G and HTLV-1 per se, as shown in Table 3. However, when we stratified patients according to neurological impairment (i.e. HTLV-1 overactive bladder or HAM/TSP) and compared these groups to infected individuals without neurological disease symptoms (i.e. HTLV-1 carriers) we observed an allele-wise association between the Aallele of the -1082bp A/G polymorphism (rs1800896) and protection against overactive bladder (OR = 0.447, CI 0.28-0.70, p = 0.001, p_corrected = 0.009). In the genotype-wise analysis, we observed significant protection against overactive bladder by both A/A homozygotes (OR = 0.17; CI: 0.06-0.45; p = 0.0001, p_corrected = 0.0009) and A/G heterozygotes (OR = 0.30, CI: 0.12-0.74, p = 0.0001, p_corrected = 0.0009) compared to G/G homozygotes. There were no associations with the HAM/TSP group. No significant associations were observed for the other IL10 SNVs. The association between the A allele and HTLV-1 overactive bladder protection is robust to multiple testing correction for 3 variants x 3 clinical phenotypes (p_corrected = p multiplied by 9).

HTLV-1 infection is a public health problem as it impacts the quality of life of HTLV-1 infected individuals. Although asymptomatic in most carriers, a significant proportion of HTLV-1 infected individuals will develop virus-related syndromes, culminating in HAM/TSP. In this study we identified an association between the allele A at the SNV rs1800896 located 1082bp upstream of the IL10 gene and protection against the development of overactive bladder associated with HTLV-1 infection, which is considered to be a precursor to HAM/TSP⁵. Interestingly, we did not find an association with HAM/TSP. Others have looked for associations between IL10 promoter region variants and clinical manifestations of HTLV-1 infection^11,12. In a study undertaken in Japan,¹¹ protection against HAM/TSP was associated with the A allele for the SNV at position -592bp when compared to HTLV-1 carriers. Similar to our study, these studies did not identify an association between promoter variants at -819bp and -1082bp, or at -2763bp, -2849bp or -3575bp, and HAM/TSP. Using reporter gene studies. these researchers demonstrated that the allele A at -592bp was associated with lower transcriptional activation compared to the allele C in HTLV-1 Tax-induced Jurkat T cells¹¹. A second study carried out in Iran¹² focused on comparisons of HTLV-1 carriers, or HAM/TSP positive individuals, with healthy controls. In this case they demonstrated that the allele T at -819bp, and the allele A at -592bp, were associated with HTLV-1 carriage and with HAM/TSP, respectively, as was a haplotype that included the -1082bp allele A. They concluded that the A/T/A haplotype was a risk factor for HTLV-1 infection per se, but did not confer additional risk of developing HAM/TSP. Neither the Japanese nor the Iranian study looked at the phenotype of overactive bladder. The different results obtained in these studies may be due to differences in genetic background, as well as due to the demographic and environmental variations and sample size. This could include the presence of haplotypes carrying different variations of the functional promoter region upstream of the IL10 gene, and/or to differences in functional interactions between variants. Further work is required to determine whether the -1082bp variant, which was associated with overactive bladder in our study, is the functional etiological variant or is in linkage disequilibrium with alternative functional variants affecting HTLV-1 disease. Additional research is required to determine whether there are differences in the pathogenesis of overactive bladder and HAM/TSP that could account for the association with the IL10 -1082bp variant with the overactive bladder phenotype but not with HAM/TSP. Others have shown¹³ that while patients with HTLV-1 overactive bladder and HAM/TSP phenotypes share some immunological features and similar proviral loads, patients with the overactive bladder phenotype are better able to modulate their inflammatory immune response. Further analysis of the association between clinical phenotypes and genetic variants influencing the balance between pro- and anti-inflammatory responses could help us develop prognostic biomarkers for use in HTLV-1 infected individuals.