HIV-associated sensory neuropathy (HIV-SN) is a common neurological complication of HIV infection and its treatment, and it is frequently painful. The pain causes significant decreases in the health-related quality of life,1 and no commercially available agents have proven efficacy for the treatment of the pain.2,3 The intensity of the pain experienced can differ significantly between individuals,4 and although psychosocial factors probably contribute to this difference,5 recent evidence from other types of peripheral neuropathy indicates that genetic variation also may be a contributing factor.6
Polymorphisms in the genes KCNS1 and GCH1 have been associated with differences in reported pain intensity in previous studies conducted in non-African populations. A particular single nucleotide polymorphism (SNP) in KCNS1 (rs734784), a gene encoding a subunit of voltage-gated potassium ion channels that affects the excitability of sensory neurons, was found to be associated with increased sensitivity to pain in several neuropathic pain states.7 GCH1 encodes guanosine triphosphate cyclohydrolase 1, the rate-limiting enzyme in the biosynthesis of tetrahydrobiopterin, which is a cofactor in the synthesis of several algogenic molecules.8 In a white cohort, individuals with a particular haplotype (15-SNP “pain-protective” haplotype) experienced reduced pain after a discectomy for persistent lumbar radiculopathy.9 A later study showed that this haplotype could be represented by 3 SNPs alone (rs8007267, rs3783641, rs10483639).10
Few studies investigating associations between genes and pain intensity have been performed in non-European populations. The association between the valine encoding allele of rs734784 in KCNS1 and increased pain intensity was detected in an Israeli cohort (Ashkenazi and non-Ashkenazi ethnicities) with postamputation pain,7 but no association was found between polymorphisms in GCH1 and pain sensitivity to experimental stimuli in African Americans, Hispanics, and Asian Americans.11 Our previous study was the first to investigate genetic risk factors for pain in a black Southern African population.12 In that study, we investigated whether 6 SNPs from the 15-SNP GCH1 “pain-protective” haplotype that had been identified in a European cohort were associated with altered pain intensity and pain susceptibility in a black Southern African population with painful HIV-SN. We found that the 3-SNP “pain-protective” haplotype10 was significantly associated with reduced pain intensity in our cohort on univariate analysis. However, the association was weak and did not withstand correction for other factors that may influence pain sensitivity (eg, sex and age). Given the weak association, we detected on univariate analysis and the greater genetic diversity in African populations compared with other populations,13 we suggest that a more detailed study of GCH1 be undertaken employing a more nuanced analysis that is using African-specific markers, rather than only those identified in a European population.
In this study, we reinvestigated the association between polymorphisms in GCH1 and pain intensity in a black Southern African cohort with HIV-SN using population-specific tagging SNPs. Additionally, due to the strong association between rs734784 in KCNS1 in various neuropathic pain conditions in European and Israeli cohorts, we have extended our assessment of “pain genes” in HIV-SN to include KCNS1. Here we investigate whether the previously identified SNP in KCNS1 and population-specific tagging SNPs in KCNS1 associate with pain intensity in a black Southern African population.
Ethical approval for the study was obtained from the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (protocol no M110754), and written informed consent was obtained from all the participants before sampling.
Population and DNA Samples
DNA samples were collected in a previous study from 342 adult (>18 years) black Southern African individuals attending the Virology Clinic at the Charlotte Maxeke Johannesburg Academic Hospital, South Africa.12 The total number of individuals in the current analysis includes only those 158 patients who were deemed to have HIV-SN using the AIDS Clinical Trials Group Brief Peripheral Neuropathy Screen,14 based on the bilateral presence of signs and symptoms of neuropathy, and who had a current pain score.4 The group of 158 participants included 123 females (77.8%). The mean (SD) age of the group was 40.9 (8.1), and the median (interquartile ratio) CD4 T-cell count was 400 (291–557). There were no significant differences in age and CD4 T-cell count between males and females. All the patients had been on antiretroviral therapy for at least 6 months, and all but 8 individuals had been exposed to stavudine. The intensity of the pain experienced at the site of the neuropathy by each individual was rated on an 11-point numeric pain rating scale (“0” = no pain, “10” = worst pain imaginable).
SNPs from KCNS1 and GCH1 that had a previous association with pain intensity were selected and supplemented with tagSNPs appropriate for an African population. tagSNPs were selected based on data from the Yoruba in Ibadan, Nigeria population (HapMap Data Release 27, Phase II + III, February 2009, on NCBI B36 assembly, dbSNP b126).15 tagSNP selection employed a pairwise approach at r2 = 1.0 and with minor allele frequency >0.01. All alleles are designated respective to the forward (+) strand of the genome assembly and were obtained using BioMart (Ensembl release 67, May 2012).16,17 A final SNP list was compiled after assay design tool evaluation by Illumina,18 which eliminated any SNPs that could not be genotyped using the method outlined below.
Genotyping and Data Cleaning
Genotyping was carried out using the GoldenGate genotyping assay with VeraCode microbeads (Illumina) and data were read on an Illumina BeadXpress Reader.19 Data were loaded into BeadStudio Data Analysis Software From Illumina, Inc (Version 184.108.40.206, genotyping module version 3.2.32; Illumina) for a preanalysis data quality control, and SNPs and samples were excluded accordingly.
Data were analyzed using PLINK software for association analysis.20,21 All SNPs were assessed for Hardy–Weinberg equilibrium and were excluded from further analysis if P < 0.01 in all individuals. A minor allele frequency threshold of 0.01 was used throughout the analysis. Univariate analysis involved a t-test for assessment of the association between the alleles present and pain score. Multivariate analysis correcting for age, gender, and CD4 T-cell count was performed using linear regression. Haplotype analysis in KCNS1 involved testing all possible combinations of the SNP alleles across the region, from 2-SNP through to 4-SNP haplotypes. The sliding window approach (specifying a fixed haplotype size of 2 and 3 SNPs across the region) was carried out in GCH1. Univariate and multivariate analyses correcting for age, gender, and CD4 T-cell count were also conducted for haplotype association analysis. The pointwise empirical P value (PEMP1; based on 1000 iterations) was calculated for all statistical comparisons. SNPs were only carried forward to multivariate analysis if PEMP1 < 0.1 and significant associations after multivariate analysis were considered to be those with PEMP1 < 0.05. Familywise correction for multiple comparisons was not performed for these exploratory analyses.
tagSNP selection for KCNS1 produced 3 additional SNPs (in addition to the previously associated SNP rs734784) for investigation—rs4499491, rs6017486, and rs6073643 (Fig. 1A).
In GCH1, 12 of the 15 “pain-protective” haplotype SNPs were genotyped and included in the analysis. The 3 SNPs that were not included (rs2183081, rs7147286, and rs7492600) failed assay design tool evaluation when developing the SNP list. tagSNP selection produced a further 19 SNPs for genotyping and analysis (Fig. 1B).
All SNPs were in Hardy–Weinberg equilibrium (results not shown).
Associations With Pain Intensity
One SNP, rs4499491, which was associated with decreased pain intensity, passed the PEMP1 < 0.1 univariate analysis threshold (P = 0.09). No significant association was detected between the SNP and pain intensity after correcting for age, gender and CD4 T-cell count (P = 0.08).
Five haplotypes associated with pain intensity on univariate and multivariate analyses (Table 1). Three of the haplotypes associated with decreased pain intensity and 2 with increased pain intensity. All haplotypes contained the SNP rs6017486, with the A-allele associating with decreased pain intensity and the G-allele with increased pain intensity. No haplotypes significantly associated with altered pain intensity contained rs734784, which had previously been identified in non-African populations.
In addition to the lack of association we reported previously, no individual SNPs or haplotypes associated with pain intensity after correcting for age, gender or CD4 T-cell count (data not shown).
We investigated whether polymorphisms in KCNS1 and GCH1 were associated with the intensity of pain experienced by individuals with symptomatic HIV-SN. Marker selection was based on SNPs previously described in the literature in non-African populations, and additionally, population-specific tagging SNPs that allow a more nuanced investigation based on the underlying genetic structure of the population being studied. This is the second study to investigate the genetics of pain in a black Southern African population, and the first such study to employ strategies to attempt to tailor the investigation to the population under investigation.
Despite a strong indication for the association of polymorphisms in GCH1 with reduced pain in individuals of European descent who had undergone discectomy for persistent lumbar root pain, a similar association was not detected when we assessed the same polymorphisms in a black Southern African population with HIV-SN.12 In this study, we extended our initial study12 by the inclusion of population-specific tagging SNPs, but nevertheless, the same conclusions were drawn as in our previous investigation, namely, no association between polymorphisms in GCH1 and pain intensity in a black Southern African population with HIV-SN. Our failure to detect an association, even when using a population-tailored approach to SNP selection, could mean that polymorphisms in GCH1 do not have a role in pain sensitivity in black Southern Africans as such, or that the polymorphisms in GCH1 do not have a role in pain intensity in HIV-SN specifically.
Analysis of KCNS1 revealed that no individual SNPs significantly associated with pain intensity after correcting for age, gender, and CD4 T-cell count. Five haplotypes, however, did show a significant association with reported pain intensity on univariate analysis and multivariate analysis. Of the 5 haplotypes, 3 were associated with decreased pain intensity and 2 associated with increased pain intensity. The SNP rs6017486 seems to mark the direction of the relationship, with the presence of the A-allele being associated with the decreased pain intensity and the presence of the G-allele being associated with increased pain intensity, despite the SNP not associating itself, with changes in pain intensity. The fact that haplotypes, but not individual SNPs, associated with pain intensity in our cohort suggests that the actual causative SNP(s) could be incorporated in the region of the gene marked by the haplotypes. It is also noteworthy that the SNP (rs734784) originally described in the literature as being associated with increased pain sensitivity in a variety of neuropathic pain states in non-African populations7 was not part of any of the haplotypes that we identified in the pain intensity analyses. This indicates the importance of using a population-based SNP selection process when assessing genetic associations.
Interpretation of our findings, both positive and negative, needs to occur in the context of the HapMap population used to select tagSNPs. We used the West African Yoruba population data from the HapMap database to select our population-specific tagging SNPs but conducted the study in a Southern African population. Population differences in genetic diversity, and hence linkage disequilibrium, between these 2 African groups may have weakened our power to detect small effects,13 but at the time of the study, the Yoruban data were the best-matched dataset available.
In conclusion, polymorphisms in GCH1 did not associate with pain intensity in HIV-positive individuals of black Southern African ancestry with HIV-SN, even though we used population-specific genetic markers. However, unique, population-specific haplotypes of KCNS1 was associated with differences in pain intensity in the same group of patients. No individual SNPs we assessed in KCNS1 were independently associated with differences in pain intensity suggesting that the haplotypes could incorporate the causative SNP(s). Our findings for KCNS1 provide additional evidence supporting an important role for KCNS1 in neuropathic pain across diverse population groups and neuropathic pain states. We believe that our study illustrates the importance of conducting association analyses in independent ethnic groups and using population-specific SNP selection.
The authors thank the staff and patients of the Virology Clinic in the Charlotte Maxeke Johannesburg Academic Hospital and Florence Mtsweni for acting as the interpreter for the study. In addition, we would like to thank Punita Pitamber for her assistance with genotyping.
1. Kamerman PR, Wadley AL, Cherry CL. HIV-associated sensory neuropathy: risk factors and genetics. Curr Pain Headache Rep. 2012;16:226–236.
2. Phillips TJ, Cherry CL, Cox S, et al.. Pharmacological treatment of painful HIV-associated sensory neuropathy: a systematic review and meta-analysis of randomised controlled trials. PLoS One. 2010;5:e14433.
3. Clifford DB, Simpson DM, Brown S, et al.. A randomized, double-blind, controlled study of NGX-4010, a capsaicin 8% dermal patch, for the treatment of painful HIV-associated distal sensory polyneuropathy. J Acquir Immune Defic Syndr. 2012;59:126–133.
4. Wadley AL, Cherry CL, Price P, et al.. HIV neuropathy risk factors and symptom characterization in stavudine-exposed South Africans. J Pain Symptom Manage. 2011;41:700–706.
5. Kim H, Neubert JK, San Miguel A, et al.. Genetic influence on variability in human acute experimental pain sensitivity associated with gender, ethnicity and psychological temperament. Pain. 2004;109:488–496.
6. Young EE, Lariviere WR, Belfer I. Genetic basis of pain variability: recent advances. J Med Genet. 2012;49:1–9.
7. Costigan M, Belfer I, Griffin RS, et al.. Multiple chronic pain states are associated with a common amino acid-changing allele in KCNS1. Brain. 2010;133:2519–2527.
8. Thony B, Auerbach G, Blau N. Tetrahydrobiopterin biosynthesis, regeneration and functions. Biochem J. 2000;347(pt 1):1–16.
9. Tegeder I, Costigan M, Griffin RS, et al.. GTP cyclohydrolase and tetrahydrobiopterin regulate pain sensitivity and persistence. Nat Med. 2006;12:1269–1277.
10. Lotsch J, Belfer I, Kirchhof A, et al.. Reliable screening for a pain-protective haplotype in the GTP cyclohydrolase 1 gene (GCH1) through the use of 3 or fewer single nucleotide polymorphisms. Clin Chem. 2007;53:1010–1015.
11. Kim H, Dionne RA. Lack of influence of GTP cyclohydrolase gene (GCH1) variations on pain sensitivity in humans. Mol Pain. 2007;3:6.
12. Wadley AL, Lombard Z, Cherry CL, et al.. Analysis of a previously identified pain-protective haplotype and individual polymorphisms in the GCH1 gene in Africans with HIV-associated sensory neuropathy: a genetic association study. J Acquir Immune Defic Syndr. 2012;60:20–23.
13. Tishkoff SA, Reed FA, Friedlaender FR, et al.. The genetic structure and history of Africans and African Americans. Science. 2009;324:1035–1044.
14. Cherry CL, Wesselingh SL, Lal L, et al.. Evaluation of a clinical screening tool for HIV-associated sensory neuropathies. Neurology. 2005;65:1778–1781.
15. The international HapMap project. Nature. 2003;426:789–796.
16. Flicek P, Amode MR, Barrell D, et al.. Ensembl 2012. Nucleic Acids Res. 2012;40(database issue):D84–D90.
17. Kinsella RJ, Kahari A, Haider S, et al.. Ensembl BioMarts: a hub for data retrieval across taxonomic space. Database (Oxford). 2011;2011:bar030.
18. Illumina I. Designing custom GoldenGate genotyping assays. Available at: www.illumina.com
. Accessed October 13, 2012.
19. Lin CH, Yeakley JM, McDaniel TK, et al.. Medium- to high-throughput SNP genotyping using VeraCode microbeads. Methods Mol Biol. 2009;496:129–142.
20. Purcell S, Neale B, Todd-Brown K, et al.. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–575.