PharmGKB summary - very important pharmacogene information for GSTT1
Abstract
This PharmGKB summary briefly discusses the very important pharmacogene GSTT1 and its variants that can influence drug responses. A fully interactive version of this short review, with links to individual paper annotations and population descriptions can be found at http://www.pharmgkb.org/vip/PA183.
Background
GSTT1 encodes the phase II metabolizing enzyme glutathione s-transferase theta. It is located in a gene cluster with GSTT2 and GSTT2B paralogues on chromosome 22 [1]. Molecular evolution studies suggest that GSTT or theta is the oldest of the families of GSTs since it is present in lower organisms: other GST families may have arisen from duplication and diversification of an ancestral GSTT gene [1].
The GSTT1 protein catalyzes the conjugation of reduced glutathione to electrophilic moeities of xenobiotics, drugs and endogenous compounds such as peroxidized lipids [2]. The conjugated products – often after further metabolism – are more soluble, allowing them to be more readily eliminated from the body. Important pharmacological substrates include etoposide, busulfan, platinum anticancer drugs, as well as the anti-tuberculosis drugs isoniazid, rifampicin and pyrazinamide [3-7] and several important industrial chemicals including butadiene, dichloromethane and trichloroethylene [8-10].
Human GSTT1 is constitutively expressed in the liver and can be induced by the consumption of cruciferous vegetables [1, 11]. GSTT1 is also expressed in the gastrointestinal tract [12], erythroid cells [13], kidney [14] and lung [15]. Most studies of inducers of GSTT1 has been performed in animal models where NSAIDs, phenobarbital, alpha-angelicalactone, alpha-tocopherol, coumarin and oltipraz induced expression of GSTT1 in the liver or gastrointestinal tract of rodents [1, 16-18]. In rats, some inducers were gender specific, with indole-3-carbinol and phenobarbital significant inducers of GSTT1 in males and coumarin the most potent inducer in females [18].
GSTT1 variation
The most studied variant of GSTT1 is the null variant, also referred to as GSTT1*0 or GSTT1 negative, that results from the complete or partial deletion of the gene (see below for more details). A large number of studies have examined the possible role of GSTT1 null in the etiology of cancer (colorectal [19], oral [20], prostate [21], hepatocellular [22], lung [23][20729793], ovarian [24], head and neck [25], gastric [16886896] and others). Individuals lacking GSTT1 may be less able to detoxify environmental xenobiotics and thus be at elevated risk for cellular damage and resultant cancer. However for many cancers, studies have shown conflicting results. For example, a large meta-analysis combining forty-five studies and over twelve thousand cases, suggested that evidence for GSTT1 involvement in lung cancer was weak [20729793]. In contrast, the GSTT1 null variant may be protective against certain cancers including bladder cancer, because GSTT1-dependent conjugation of xenobiotics such as the industrial chemical trichloroethylene produces compounds with increased toxicity [26]. Other epidemiological studies suggest a role for GSTT1 in inflammatory airway diseases such as asthma and emphesema [27, 28]. The pharmacogenomics (PGx) of GSTT1 null and how it may influence drug response, particularly for anti-neoplastic drug and drug-induced liver injury (DILI) is therefore of great interest (discussed further below).
The related glutathione s-transferase mu gene, GSTM1 also has a null variant [18056202]. Since GSTM1 is also involved in drug and xenobiotic detoxification various studies have examined the effects of deletion of GSTM1 null alone and in combination with GSTT1 null for both disease risk and drug response [16886896, 18666253, 20214588].
Relatively few studies have reported single nucleotide variants in GSTT1 [29-32], and only one study has reported a PGx association [33] (see table 1). More variants are listed in dbSNP (103 as of 10/2011). However, most of these variants have not been studied,. The majority of GSTT1 SNPs are present at very low frequencies, and may be population specific; some population studies have not seen polymorphisms [29, 30, 32, 34, 35]. In a pharmacogenetic study of multiple myeloma patients treated with thalidomide, heterozygotes for the 5’UTR −182C>T variant (rs4630) had lower neurotoxicity [33]. Another study of head and neck cancer patients treated with paclitaxel found no association of this variant with toxicity or survival [36], although patients with two or more variants, including GSTT1 rs4620 as well as variants in CYP2C8, ABCB1, GSTP1 and ERCC1, did have significantly higher overall survival [36].
Table 1
Single nucleotide variants in GSTT1 reported in the literature.
| Identifier | Common name | Phenotype [PMID] | Frequency [PMID] |
|---|---|---|---|
|
| |||
| 5’ FR (-714)C>T | 0.01 T Coriell AA (Moyer et al, 2007)[18056202][27] | ||
| 0.015 T Coriell CA (Moyer et al, 2007)[18056202] | |||
|
| |||
| 5’ FR (-476)G>A | 0.008 A Coriell CA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| 5’ FR (-295)C>G | 0.01 G Coriell AA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| 5’ FR (-56)G>C | 0.32 C Coriell AA (Moyer et al, 2007)[18056202][27] | ||
| 0.008 C Coriell MA (Moyer et al, 2007)[18056202] | |||
|
| |||
| rs4630 C>T | 5’ UTR (−182)C>T | CT had lower neurotoxicity with thalidomide in MM [21435719]. Not assoc with survival in head and neck cancer patients treated with paclitaxel [19504558] | 0.03 C (Grau et al) [19504558][36] |
|
| |||
| rs1130990 C>T | Leu5Leu | 0.01 T Coriell AA (Moyer et al, 2007)[18056202][27] | |
| 0.023 T Coriell CA (Moyer et al, 2007)[18056202] | |||
| 0.008 T Coriell MA (Moyer et al, 2007)[18056202] | |||
| 0/100 Coriell HCA (Moyer et al, 2007) [18056202] | |||
|
| |||
| IVS 1 (-178)C>A | 0.01 A Coriell AA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| IVS 1 (-151)G>A | 0.01 A Coriell AA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| IVS 1 (-24)C>T | 0.029 T Coriell AA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| IVS 1 (-14)T>A | 0.008 A Coriell MA (Moyer et al, 2007)[18056202] [27] | ||
|
| |||
| rs2266635 | Ala21Thr | 0/400 (Agundez et al, 2008)[18303971][30] | |
|
| |||
| rs11550606 | Leu30Pro | 0/400 (Agundez et al, 2008)[18303971][30] | |
|
| |||
| Asp43Asn | Decreased protein [18303971] | 0.01 A Coriell AA (Moyer et al, 2007)[18056202] [27] | |
|
| |||
| rs17856199 | Phe45Cys | 0/400 (Agundez et al, 2008)[18303971][30] | |
|
| |||
| Exon 2 (177)C>T | 0.008 T Coriell CA (Moyer et al, 2007)[18056202][27] | ||
|
| |||
| Thr65Met | Decreased protein [18303971] | 0.008 T Coriell MA (Moyer et al, 2007)[18056202][27] | |
| 0/300 Coriell AA, CA, HCA (Moyer et al, 2007)[18056202] | |||
|
| |||
| IVS 2 (-13)C>T | 0.136T Coriell AA (Moyer et al, 2007)[18056202][27] | ||
| 0.008 T Coriell MA (Moyer et al, 2007)[18056202] | |||
|
| |||
| Exon 3(225)G>A | 0.01 A Coriell AA (Moyer et al, 2007)[18056202][27] | ||
| 0.008 A Coriell MA (Moyer et al, 2007)[18056202] | |||
|
| |||
| rs11550605 A>C | GSTT1*B, Thr104Pro | Decreased protein [18303971] | 0.01 (Alexandrie et al, 2002)[12439221][34] |
| 0/400 (Moyer et al, 2007)[18056202] [27] | |||
| 0/147 (Matsuno et al, 2004)[15202795][35] | |||
| 0/317 (Piacentini et al, 2011) [20563854][32] | |||
| 0/400 (Agundez et al, 2008)[18303971][30] | |||
|
| |||
| Exon 4 G>x | Decreased protein [18303971] | 0.008 T Coriell MA (Moyer et al, 2007)[18056202][27] | |
|
| |||
| rs2266633 | Asp141Asn | 0/400 (Agundez et al, 2008)[18303971][30] | |
|
| |||
| rs2266637 G>A | Val169Ile | 0.126 A Coriell AA (Moyer et al, 2007)[18056202][27] | |
| 0.008 A Coriell MA (Moyer et al, 2007)[18056202] | |||
|
| |||
| rs2234953 G>A | Exon 4 Glu173Lys | 0/86 (Piacentini et al, 2011) [20563854][32] | |
| 0/400 (Agundez et al, 2008)[18303971][30] | |||
|
| |||
| IVS 4 (-87) del CCT | 0.029 del Coriell AA (Moyer et al, 2007)[18056202] [27] | ||
|
| |||
| Exon 5 (573) G>A | 0.008 A Coriell MA (Moyer et al, 2007)[18056202] [27] | ||
|
| |||
| rs17850155 G>A | Exon 5 Lys228Lys | 0/86 (Piacentini et al, 2011) [20563854][32] | |
numbering from Moyer et al, 2007)[18056202]
Abbreviations: MM, multiple myeloma; AA, African American; CA, Caucasian; HCA, Han Chinese; MA, Mexican American.
Important variants
GSTT1 null
The GSTT1*0 or null variant represents the complete or partial deletion of the GSTT1 gene. Alignment of the Genbank reference sequence of the GSTT1 deletion/junction region sequence AF240785.1 [37] shows the deletion between chr22:24,343,276 - chr22:24,397,528 on the GRCh37/hg19 build of the human genome. The frequency of GSTT1 null varies widely in different populations: approximately 50-60% in Asians, 15% in White populations, 15-20% in Blacks or African Americans, and less than 10% in Hispanic populations (reviewed in [1]).
Possible associations with various cancers have been investigated with mixed results. There are several HuGE reviews (colorectal [19], oral [20], prostate [21], hepatocellular [22], lung [23], ovarian [24], head and neck [25] and bladder cancer [26]).
The GSTT1 null variant is a risk factor for coronary artery disease in Type 2 diabetic patients, especially among smokers [38]. Other epidemiological studies have suggested a role in inflammatory airway diseases such as asthma and emphesema [27, 28].
In a study of samples from White subjects from the Coriell collection, GSTT1 null was found to be in linkage disequilibrium with a deletion of GSTT2B [39]. Cells with the GSTT2B deletion also had lower expression of GSTT2 and a reduced capacity for glutathione conjugation [39]. This linkage may account for some of the discord between studies of cancer risk.
GSTT1 null and drug toxicity
Since GSTs are expressed in the liver and since glutathione is involved in the detoxification of many drugs, several studies have examined the effect of the GSTT1 null genotype as well as other GST family members, GSTM1 and GSTP1, on adverse drug reactions including DILI and allergic responses in the skin.
The GSTT1 null variant is associated with an increased risk of allergic skin reactions to a variety of drugs, including NSAIDs and antibiotics [40]. Although not significant for the GSTT1 null variant alone, the double null of GSTT1 and GSTM1 is also associated with increased risk for DILI particularly with NSAIDs and cardiovascular drugs [18666253, 20214588]. Neither GSTT1 null nor GSTM1 null genotype alone could predict susceptibility to tacrine-induced hepatotoxicity. However, the combination of GSTT1 null and GSTM1 null variant is associated with increased hepatotoxicity in patients with Alzheimer disease taking tacrine [41]. The double null mutations of GSTT1 and GSTM1 are also associated with troglitazone-induced [42] and alcohol-induced liver injury [43]. However, GSTT1 null is not associated with DILI after treatment with anti-tuberculosis drugs [44, 45].
GSTT1 null and response to antineoplastic agents in cancer treatment
Several antineoplastic drugs are metabolized by GSTs, including platinum drugs, anthracyclines, vinca alkaloids, cyclophosphamide and epipodophylotoxins [5, 6, 46-50]. Cytotoxic antineoplastic drugs may contribute to cell death by depleting cellular glutathione, allowing toxic products to build up and damage DNA. A few studies have examined the role of GSTT1 in antineoplastic drug toxicity. In Hodgkin lymphoma patients treated with anthracyline regimens, the GSTT1 null variant was associated with increased toxicity [51]. A study of large B-cell lymphoma patients treated with rituximab plus cyclophosphamide/doxorubicin/vincristine/prednisone showed increased toxicity in individuals with the GSTT1 null variant [52]. The GSTT1 null variant is associated with increased likelihood of adverse events, including cognitive impairment in pediatric medulloblastoma patients treated with cisplatin, cyclophosphamide, and vincristine [53]. The null genotype of GSTT1 is also associated with rate of early death after the initiation of chemotherapy in Japanese AML patients treated with cytarabine, mercaptopurine, prednisone and daunorubicin [54]. However GSTT1 null is not associated with cardiac damage after anthracycline exposure (pediatric ALL) [55], platinum-based toxicity (mesothelioma) [56], or toxicity with fluorouracil-based regimens (colorectal cancer) [57].
The GSTT1 and GSTM1 double null variant has also been associated with decreased event free survival in NHL, adult AML and R-CHOP–treated B-cell lymphoma and increased relapse in pediatric ALL [17454600, 12351375, 20303013, 14607752].
Conclusions
Studies of various cancers and treatments and the GSTT1 null variant have shown mixed and contradictory results (see Table 2). While the lack of GSTT1 may increase the efficacy of some cytotoxic drugs towards cancer cells due to a reduction in their elimination from the cell, it could also be hypothesized that GSTT1 null individuals might have a poorer prognosis. Further studies of focused populations with defined treatments may aid in elucidating any risk or benefit from identifying GSTT1 variants prior to treatment.
Table 2
GSTT1 null and response to antineoplastic agents.
| Allele/genotype | Phenotype | Drugs | Population (study size, race/ethnicity) | PMID |
|---|---|---|---|---|
| GSTT1 null | decreased survival | cytarabine, mercaptopurine, prednisone and daunorubicin | Adult AML (n=193, Asian) | 11840286 [54] |
| GSTT1 null | decreased event free survival | Not specified | NHL/ follicular lymphoma subtype (n=89, unknown race) | 17454600 |
| GSTT1 null | decreased event free survival | cyclophosphamide, doxorubicin and fluorouracil | Breast neoplasms (n=152, Multiple) | 20459744 |
| GSTT1 null | decreased overall survival | paclitaxel or docetaxel and carboplatin or cisplatin | Ovarian cancer (n=118, Asian) | 19203783 |
| GSTT1 null | decreased overall survival, progression free survival | fluorouracil and platinum compounds | Gastric cancer (n=134, unknown race) | 19332728 |
| GSTT1 null | increased toxicity | cisplatin, cyclophosphamide, and vincristine | Pediatric Medulloblastoma (n=42, Multiple) | 18952980 |
| GSTT1 null | increased toxicity | rituximab plus cyclophosphamide/doxorubicin/vincristine/prednisone | de novo diffuse large B-cell lymphoma (n=94, Asian) | 20303013 |
| GSTT1 null | increased toxicity, increased overall survival | anthracylines | Hodgkin lymphoma (n=125, unknown) | 20977336 |
| GSTT1 null | increased survival | Not specified | NHL/ follicular lymphoma subtype (n=112, multiple) | 20029944 |
| GSTT1 null | increased overall survival | platinum compounds, doxorubicin and ifosfamide | Osteosarcoma (n=30, Multiple races) | 20577141 |
| GSTT1 null | increased event free survival, increased progression free survival | mitomycin | Bladder cancer (n=282, Asian) | 21045267 |
| GSTT1 null | increased overall survival | Not specified | Gastric cancer (n=130, Asian) | 21378360 |
| GSTT1 present | increased likelihood of complete response | homoharringtonine, cytarabine, daunorubicin | AML (n=254, Asian) | 18035413 |
| GSTT1 present | decreased survival | platinum compounds | Non-small cell lung cancer (n=973, White) | 20200426 |
| GSTT1 null | is not associated with toxicity (cardiac damage) | anthracylines | Pediatric ALL (n=76, White) | 19863340 |
| GSTT1 null | is not associated with toxicity | fluorouracil-based regimens (FOLFOX/FOLFIRI) | Colorectal neoplasms (n=346, unknown race) | 20385995 |
| GSTT1 null | is not associated with toxicity | platinum compounds | Mesothelioma (n=133, White) | 21765044 |
| GSTT1 null | is not associated with toxic death* | doxorubicin and cytosine arabinoside | AML (n=98, White) | 17671537 |
| GSTT1 null | Is not associated with survival | daunorubicin, cytosine arabinoside, mitoxantrone | AML (n=139, White) | 18207572 |
| GSTT1 null | Is not associated with survival | Not specified | Colorectal neoplasms (n=315, Multiple races) | 19748847 |
| GSTT1 null | Is not associated with survival | oxaliplatin | Colorectal neoplasms (n=65, Multiple races) | 20017670 |
| GSTT1 null | Is not associated with survival | paclitaxel/docetaxel, cisplatin | Gastric cancer (n=200, Asian) | 20331623 |
| GSTT1 null | Is not associated with survival | cisplatin-based regimens | Gastric cancer (n=138, unknown race) | 18443805 |
| GSTT1 null | Is not associated with survival | fluorouracil, oxaliplatin | Colorectal neoplasms (n=107, Multiple races) | 12072547 |
| GSTT1 null | Is not associated with survival | cisplatin, paclitaxel | Ovarian Cancer (n=24, unknown race) | 12851839 |
| GSTT1 null | Is not associated with treatment associated secondary neoplasms | etoposide/teniposide | treatment-related AML or myelodysplastic syndrome following pediatric ALL (n=302, Multiple races) | 10673738 |
| GSTT1 and GSTM1 double null | decreased survival | Not specified | Adult AML (n=106, unknown race) | 12351375 |
| GSTT1 and GSTM1 double null | increased likelihood of relapse | doxorubicin, vincristine, prednisone, cyclophosphamide, asparaginase, mercaptopurine, methotrexate, etoposide, cytarabine | Pediatric ALL (n=82, Asian) | 14607752 |
| GSTT1 and GSTM1 double null | decreased event free survival | Not specified | NHL/ follicular lymphoma subtype (n=89, unknown race) | 17454600 |
| GSTT1 and GSTM1 double null | decreased event free survival | rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone | de novo diffuse large B-cell lymphoma (n=60, Asian) | 20303013 |
Abbreviations: AML, acute myeloid leukemia; NHL, non-hodgkin lymphoma; ALL, precursor T-Cell lymphoblastic leukemia-lymphoma; FOLFOX, fluorouracil leucovorin and oxaliplatin; FOLFIRI, fluorouracil leucovorin and irinotecan.
Acknowledgments
This work is supported by the NIH/NIGMS (R24 GM61374).
