In this study, we have found rs761142 T
G in GCLC
to be significantly associated with SMX-induced hypersensitivity in HIV/AIDS patients, with each copy of the minor G allele increasing risk nearly 2 fold. Consistent with this finding, the rs761142 G allele was also significantly associated with reduced GCLC mRNA expression in livers and B-lymphocytes. In contrast, previously reported promoter SNP rs17883901 and 5’UTR GAG trinucleotide polymorphisms [33
] did not show significant associations. Although reactive metabolites and oxidative stress were proposed to be involved in the pathogenesis of idiosyncratic drug reactions [4
], this is the first study implicating a gene involved in antioxidant defense, affecting risk of idiosyncratic drug-induced cutaneous reactions.
Glutamate-cysteine ligase (GCL), a rate limiting enzyme for biosynthesis of glutathione (GSH) (Figure ), is composed of a catalytic subunit (GCLC) and a modifier subunit (GCLM). GSH is the main cellular antioxidant, scavenging reactive metabolites and preventing tissue damage [11
]. In HIV/AIDS patients, GSH levels are progressively depleted [38
], consistent with the higher incidence of SMX-induced hypersensitivity in HIV/AIDS patients than in non-infected controls [39
]. Moreover, SMX cytotoxicity is suppressed by addition of GSH in vitro [37
], and cells with GCLC knockdown were more sensitive to reactive metabolites induced cytotoxicity [40
]. Given the important role of GCLC in scavenging reactive metabolites, variants that reduce GCLC expression have a plausible role in increasing risk of developing SMX-induced hypersensitivity, especially in HIV/AIDS patients with already compromised GCLC function [38
studies have focused on promoter SNP rs17883901 and 5’UTR GAG trinucleotide polymorphisms [33
]. Promoter SNP rs17883901 was shown to reduce basal and H2
-induced promoter activity [33
], while the GAG trinucleotide repeat variants affect GCLC protein expression through translation [41
]. However, the reported results have been inconsistent. For example, the reference 7 repeat has been associated with either lower or higher GCL activity/GSH levels compared to variant repeats (4, 8, 9 or 10 repeats) in different cell types or disease conditions [35
], indicating tissue/cell or environmental specific regulation of GCLC polymorphisms, or the presence of other unidentified functional polymorphisms in GCLC
. This is consistent with numerous conflicting clinical association studies reported for GCLC
]. Our study failed to reveal significant association between promoter SNP rs17883901 or 5’UTR GAG trinucleotide repeat polymorphisms and SMX-induced hypersensitivity. Instead, the significantly associated rs761142 is located in the middle of intron 1 of GCLC
. Although intronic polymorphisms can affect gene expression by various mechanisms [47
], there is no evidence that rs761142 is functional by itself; instead, the association observed in this study could be caused by other functional polymorphisms in LD with rs761142 responsible for lowering GCLC mRNA expression. Similarly, SNP rs670548, located in intron12 of GCLC
and showing significant association in our study, had also been associated with GCLC mRNA expression previously [32
]. Taken together, the results indicate that a regulatory polymorphism in GCLC
that affects mRNA expression modify risk of developing SMX-induced hypersensitivity in HIV/AIDS patients. This result warrants replication in a larger cohort. Whether the GCLC
polymorphisms are associated with SMX-induced hypersensitivity in non-HIV/AIDS patients will require further investigation.
There are several limitations in this observational clinical association study. First, the CD4 cell counts at the time of SMX administration were not uniformly available, therefore the influence from CD4 cell count cannot be evaluated; Second, patient comorbidity and co-medication information were not available. Since SMX is inactivated and bio-activated by drug metabolizing enzymes, other disease states or concomitant administration of other drugs may affect the balance between bio-activation and bio-inactivation of SMX, influencing the level of toxic metabolites. And finally, evaluation of rs670548 and risk of hypersensitivity in African American may be limited by small sample size. A prospective larger cohort study will needed in the future to fully evaluate the association between SNPs in GCLC and SMX-induced hypersensitivity.
We have previously reported the association between polymorphisms in NAT1
and SMX-induced hypersensitivity, and gene-gene interactions between NAT1
]. Since idiosyncratic adverse drug reactions are thought to be multigenic, it is likely that the risks of developing hypersensitivity are modified by interactions between multiple genes. Before testing the interactions between NAT1/NAT2
, it is important to identify the functional polymorphism(s) and assess the frequency, direction and effect size for each.
Although not reported for drug-induced idiosyncratic cutaneous reaction; previous studies have associated drug induced idiosyncratic liver injury to antioxidant defense genes (SOD2 and GPX1) [48
]. Consistently, SOD2 knockout mice have increased sensitivity to idiosyncratic liver injury induced by troglitazone or acetaminophen [49
]. Similarly, mice deficient in NFE2L2 (NRF2), a transcription factor regulating antioxidant genes expression, also have increased sensitivity to acetaminophen induced liver injury [50
]. In the present study, we observed additional SNPs in antioxidant defense genes CATGSS
to be associated with SMX-induced hypersensitivity at nominal p values less than 0.05 (Table ). These results suggest that multiple polymorphisms in antioxidant defense genes may modify risk of developing idiosyncratic drug reaction in general.