Expression of the theta class GST isozyme, YdfYdf, in low GST dogs

Glutathione-S-transferase-theta genotypes and the risk of cyclophosphamide toxicity in dogs

The antineoplastic agent cyclophosphamide (CP) has dose-limiting side effects including sterile haemorrhagic cystitis (SHC), bone marrow (BM) suppression and gastrointestinal (GI) toxicity in dogs. The metabolites acrolein and phosphoramide that mediate these toxicities are glutathione-S-transferase (GST) substrates, and low functioning GST alleles are associated with CP toxicity in humans. The aim of this study was to determine whether variants in 2 canine GST genes, GSTT1 and GSTT5, were over-represented in dogs that developed CP toxicity. Dogs undergoing pulse or metronomic CP chemotherapy were recruited (n = 101) and genotyped for 6 GSTT1 polymorphisms and 1 GSTT5 6-bp deletion that leads to non-functional enzyme. Median cumulative CP dosages for dogs with SHC (1350 mg/m² ) were significantly higher than for dogs with GI/BM toxicity (871 mg/m² ) or no toxicity (991 mg/m² ; P = .0012). Dogs with SHC were more likely to have had metronomic (84.2%, 16 of 19 SHC cases) vs pulse (15.8%, 3 of 19 SHC cases) CP dosing (P < .0001). All dogs with BM or GI toxicity (n = 30) had pulse chemotherapy. GSTT1 and GSTT5 variant allele frequencies were not significantly different in CP-treated dogs with SHC or GI/BM toxicity compared to dogs without documented adverse effects. Work is underway to identify which canine GSTs detoxify acrolein and phosphoramide, so that better tools are available to predict the risk of CP toxicity in dogs.

A 6-bp Deletion Variant in a Novel Canine Glutathione-S-Transferase Gene (GSTT5) Leads to Loss of Enzyme Function

OBJECTIVES

Glutathione-S-transferases (GSTs) detoxify reactive xenobiotics, and defective GST gene polymorphisms increase cancer risk in humans. A low activity GST-theta variant was previously found in research beagles. The purpose of our study was to determine the molecular basis for this phenotype and its allele frequency in pet dogs.

METHODS

Banked livers from 45 dogs of various breeds were screened for low GST-theta activity by the substrate 1,2-dichloro-4-nitrobenzene (DCNB), and were genotyped for variants in a novel canine GST gene, GSTT5. Whole-genome sequences from 266 dogs were genotyped at one discovered variant GSTT5 locus.

RESULTS

Canine livers ranged 190-fold in GST-theta activities, and a GSTT5 exon coding variant 385_390delGACCAG (Asp129_Gln130del) was significantly associated with low activity (P < 0.0001) and a marked decrease in hepatic protein expression (P = 0.0026). Recombinant expression of variant GSTT5 led to a 92% decrease in Vmax for DCNB (P = 0.0095). The minor allele frequency (MAF) for 385_390delGACCAG was 0.144 in 45 dog livers, but was significantly higher in beagles (0.444) versus nonbeagles (0.007; P = 0.0004). The homozygous genotype was significantly over-represented in Pembroke Welsh corgis (P < 0.0001) based on available whole-genome sequence data.

CONCLUSIONS

An Asp129_Gln130del variant in canine GSTT5 is responsible for marked loss of GST-theta enzyme activity. This variant is significantly over-represented in purpose-bred laboratory beagles and in Pembroke Welsh corgis. Additional work will determine the prevalence of this variant among other purebred dogs, and will establish the substrate range of this polymorphic canine enzyme with respect to common environmental carcinogens.

Characterization of a low expression haplotype in canine glutathione S-transferase (GSTT1) and its prevalence in golden retrievers

Glutathione S-transferase-theta (GSTT1) is a carcinogen detoxification enzyme, and low activity variants are associated with lymphoma in humans. We recently found a variant in the 3' untranslated region (UTR) of canine GSTT1, *101_102insT, which was predicted to change miRNA binding and was found in 5 of 17 golden retriever (GR) dogs with lymphoma but none of 14 healthy GRs. The aim of this study was to determine whether this variant led to decreased GSTT1 expression and was a discernible risk factor for lymphoma within the GR breed. On resequencing, *101_102insT appeared to be in complete linkage disequilibrium with 3 additional 3'UTR variants, leading to the inferred haplotype *3T>C; *101_102insT; *190C>A; *203T>C. In canine livers that were heterozygous for this variant haplotype, GSTT1 protein expression was significantly lower compared to the reference haplotype (densitometry .40 vs .64, P = .022), and GSTT1 transcript levels by qPCR were also significantly lower (fold difference .52, P = .012), without evidence of substantial allelic expression imbalance. The variant haplotype led to >50% decrease in expression in vitro (.31 ± .07 vs .64 ± .19; P = .019). We found no significant difference in minor allele frequencies between 71 GR dogs with lymphoma (MAF .162) and 33 healthy age-matched controls (MAF .136, P = .69). Our results indicate that the variant GSTT1 3'UTR haplotype containing *101_102insT reduces gene expression, which could lead to impaired carcinogen detoxification, but was not a detectable risk factor for lymphoma in GR dogs.