A GCH1 Haplotype Associated with Susceptibility to Vasoocclusive Pain and Impaired Vascular Function in Sickle Cell Anemia.

Abstract 575

Sickle cell anemia (SCA) is characterized by polymerization of sickle hemoglobin (HbS), erythrocyte deformity, and hemolytic anemia. HbS induced red cell injury is believed to be critical for initiating acute vasoocclusive episodes, although the precise mechanism underlying this has not been elucidated. Our prior work has established a SCA sub-phenotype characterized by hyper-hemolysis, decreased nitric oxide (NO) bioavailability, increase tricuspid jet velocity (TRV), and early mortality. Surprisingly, it was also characterized by fewer acute pain episodes. Recently, a haplotype of GTP cyclohydrolase (GCH1), the rate-limiting enzyme for tetrahydrobiopterin (BH4) synthesis, was associated with pain sensitivity in both animal and human experimental models. BH4 is a cofactor required for catecholamine, serotonin and NO production. Together, these findings led us to hypothesize that genetic variation resulting in uncoupled BH4 dependent enzymes might also modulate complications of SCA. We examined 7 markers of the GCH1 locus against interrelated traits, including hemolysis (LDH), NO bioavailability (arginine/ornithine ratios), TRV, painful episodes and survival among 188 contemporary SCA subjects followed at NIH. There was no association with LDH, arginine:ornithine ratios, TRV or survival at a median follow-up of 3.4 years. However, a GCH1 haplotype defined by 3 linked SNPs (rs8007267, rs2878172, rs7147286) was more prevalent in patients with more frequent pain compared to 73 SCA subjects with very infrequent painful episodes (odds ratio 2.13, 95% CI 1.21-3.78, P=0.007). Further evaluation in a replication study using the Cooperative Study of Sickle Cell Disease (CSSCD) database showed no association between pain and 2 of the GCH1 haplotype markers tested (rs8007267 P=0.59; rs2878172 P=0.35). However, analysis of 9 additional GCH1 SNPs in CSSCD cases indicated that an infrequent variant (rs17738966) is associated with pain (odds ratio 2.50, P=0.008). Further analysis for differences in population stratification, age, demographics and ascertainment bias is needed to better determine if GCH1 polymorphisms are associated with acute SCA pain. To study the possible functional significance of GCH1 polymorphisms, lymphoblastoid cell lines homozygous for these haplotype markers from the Yoruban Hapmap population were studied in vitro and had significantly lower GCH1 mRNA expression after stimulation to high levels of intracellular cAMP compared to homozygotes for the variant haplotype (P=0.004). In vivo, the GCH1 pain haplotype was also associated with diminished forearm blood vessel vasodilation in response to acetylcholine (ACh) infusion in 21 subjects with SCA (P=0.03). ACh induced vasodilation is an expected response to BH4 dependent endothelial NO synthesis and the variable SCA response suggests that those with a diminished vaso-relaxation have a relative BH4 deficiency. As a negative control, there was no association with nitroprusside induced vasodilation. Furthermore, examination of a congenic cross of Brown Norway chromosome 15 (site of rat Gch1) onto a Dahl SS background in rats demonstrated a chromosome dosage effect for ACh response (EC₅₀, P#x003D;0.007). Taken together, we demonstrate an association between pain and a GCH1 haplotype in a contemporary SCA population, which also appears to involve a different haplotype effect than those previously described in non-SCA studies of pain. BH4 therapy has been studied in a Phase 2a clinical trial in SCD. The GCH1 association study with ACh dependent vasodilation in humans and animals suggests the need for additional studies to determine if polymorphisms in GCH1 and variability in BH4 synthesis modulates SCA pain, and if polymorphisms of this gene should be accounted for in clinical trials focused on pain prevention.