Polymorphisms in GSTM1, GSTT-1, and GSTP1

Read Also

The best understood polymorphisms in GST enzymes exist in the mu (GSTM-1), theta (GSTT-1) and pi (GSTP-1) classes (Wormhoudt, et al., 1999); thus, these are the focus of the current analysis.  The particular enzyme class(es) responsible for catalyzing acrylamide and glycidamide binding with glutathione are not known.  However, these enzyme classes are known to have overlapping substrate specificity and thus may act in concert to conjugate acrylamide.  Therefore, it is possible that multiple polymorphisms in multiple GSTs occurring in a single individual may affect risk in a way that could not be predicted by evaluating a single polymorphism.  Therefore, using the approach delineated in an earlier report (Ginsberg, et al., 2003), we first summarize the enzyme function and population distributions of each polymorphism separately for GSTM1, GSTT1, and GSTP1. 

In our PBTK analysis we simulate the role of GSTs polymorphisms in two ways. We use data on GSTM1, the most variable of the GST enzymes, to represent potential variability in GSH conjugation assuming only one GST isoform is responsible for binding to acrylamide and/or glycidamide.  We also simulate the role of all three GSTs in an individual by integrating across the individual population distributions of enzyme function.

            The wild type form of GSTM1 encodes the active enzyme, GSTM1*A, while a polymorphism involving a mutation at a single base in exon 7 yields the variant termed GSTM1*B (Eaton, 2000).  This mutation does not appear to affect enzyme function as genotyping and phenotyping probes commonly used cannot distinguish between this variant and the wild type (Taningher, et al., 1999).  However, a second variant, GSTM1*0 is of major consequence.  It represents deletion of the major portion of the GSTM1 gene rendering the product inactive.  Homozygotes for the deleted gene are termed GSTM1 null or GSTM1 (-/-) as distinguished from heterozygotes (GSTM1 (+/-) and the homozygous wild type, GSTM1(+/+). 

            The wild type form of GSTT1-1 is fully active.  The primary variant of interest is GSTT1-0 that has a substantial part of the gene deleted and is devoid of enzyme activity (Landi, 2000).  Homozygotes are GSTT1 null (-/-) while heterozygotes (GSTT1(+/-) have intermediate activity demonstrating a gene dosage effect (Thier et al., 1998).  In addition to catalyzing GSH conjugation with electrophiles, GSTT1 has peroxidase activity towards a variety of organic peroxides (e.g., phospholipid hydroperoxides).   Another isozyme, GSTT2, is very similar in sequence but has been little studied and the phenotypic consequences of this polymorphism have not been delineated (Landi, 2000). 

            The GST pi subclass is the major fetal isoform for GSTs, though its levels in liver decrease after birth as other GST levels increase (Strange, et al., 1989).  However, GST pi can remain quantitatively important in kidney, lung and other extrahepatic tissues well after birth (Beckett, et al., 1990; Vos, and van Bladeren, 1990).  In addition, elevated expression of this GST class in liver is an indication of pre-neoplastic transformation.   GSTP1-A is the wild type and generally but not always the most active form of this class.  Three allelic variants have been identified as follows:  GSTP1-B results from a mutation at codon 105 involving isoleucine (ile) to valine (val) substitution; GSTP1-C has the codon 105 mutation plus a codon 114 mutation that changes alanine (ala) to val; GSTP1-D is different from the wild type only with respect to the codon 114 mutation. Note: some authors state the polymorphisms are at codons 104 and 113 depending on the nucleotide considered as the starting point for transcription. The effect of polymorphisms on GSTP1 conjugating activity is substrate-specific.

            Table 4-3 summarizes the relative activity levels of the individual GSTs and their polymorphisms compared to wild type activity. They are expressed relative to wildtype activity in order to be able to compare activity levels across substrates and different measures of activity. The analysis supporting this table was taken from a previous evaluation of GST genetic polymorphisms (Ginsberg et al., 2003)  

            In order to explore the impact of having multiple GSTs involved in acrylamide and glycidamide metabolism,  the joint activity of the possible GSTM1, T1, and P1 genotype crosses on acrylamide clearance was assessed.   Table 4-4, also taken from Ginsberg et al.  (2003), shows the relative activity levels for possible GSTM1, T1 and P1 crosses, for two different GSTP1 substrates, 1-chloro-2,4-dinitrobenzene (CNDB) and BPDE-type substrates (e.g. benzo(a)pyrene diol-epoxide).  The total relative activity of a particular combination of genotypes is the sum of the relative activity level for each individual GST genotype from Table 4-3.  For example, in case number 1, where the wildtype GSTM1, T1 and P1 are each represented, the individual activity levels sum to 3 activity units when CNDB is the substrate (Table 4-4).  Ultimately for modeling purposes, activities were normalized relative to a total activity of 100% with a maximum of 33% contributed by each of the specific GST isoforms.

            Population variability in GST activity results not just from the presence of particular polymorphisms but from their relative frequency in different ethnic groups in the population.  The frequency of the GSTM1, T1 and P1 polymorphisms vary by ethnic group.[1]  For example, 53 percent of Caucasian populations are estimated to be GSTM1 null (-/-) compared to 21 percent of African Americans (Appendix Table A-1).  Given the importance of GST in catalyzing the conjugation of particular chemicals with glutathione, these populations could have reduced GSH binding capabilities. The relative frequency of the GSTM1, T1, and P1 genotype crosses in three ethnic groups (Caucasians, African Americans and Asians), also taken from Ginsberg et al. (2003), are provided in Appendix Table A-2.

---

[1] Ginsberg et al. (2003 unpublished) estimated the frequency of individual GSTM1, T1 and P1 polymorphisms in various ethnic groups using available data on the frequency of null genotypes and Hardy-Weinberg equations.  See that report for details.