Cancer Research CTRC-AACR San Antonio Breast Cancer Symposium Susan G. Komen for the Cure-AACR Outstanding Investigator Award for Breast Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Cell Growth & Differentiation
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Coles, B. F.
Right arrow Articles by Kadlubar, F. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Coles, B. F.
Right arrow Articles by Kadlubar, F. F.
[Cancer Research 60, 573-579, February 1, 2000]
© 2000 American Association for Cancer Research

Biochemistry

Quantitative Analysis of Interindividual Variation of Glutathione S-Transferase Expression in Human Pancreas and the Ambiguity of Correlating Genotype with Phenotype1

Brian F. Coles2, Kristin E. Anderson, Daniel R. Doerge, Mona I. Churchwell, Nicholas P. Lang and Fred F. Kadlubar

Divisions of Molecular Epidemiology [B. F. C., F. F. K.], and Biochemical Toxicology [D. R. D., M. I. C.], National Center for Toxicological Research, Jefferson, Arkansas 72079; Division of Epidemiology, University of Minnesota, Minneapolis, Minnesota 55454 [K. E. A.]; Department of Surgery, University of Arkansas for Medical Sciences and Arkansas Cancer Research Center, Little Rock, Arkansas 72205 [N.P.L.]: and Surgery Services, Central Arkansas Veterans Health Care System, Little Rock, Arkansas 72205 [N. P. L.]


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Analysis of glutathione S-transferases (GSTs) of the alpha, mu, and pi classes by reverse-phase high-performance liquid chromatography and electrospray-ionization mass spectrometry in 43 samples of normal human pancreas demonstrated a wide variation in expression of subunits P1, A1, A2, A4, M1, M2, and M3 and the presence of a novel form designated GST "A5." GSTA2 consisted of three forms that were differentially expressed between individuals in a manner consistent with allelic polymorphism at the hGSTA2 locus. Expression, in terms of µg GST subunit/mg cytosolic protein, varied by 6–15-fold for subunits P1, A2, and M3 and 17–30-fold in the case of GSTs A1 and M2. Less consistently expressed were GSTs M1a, M1b, A4, and A5. Among these, GSTM1 expression (excluding M1-null samples) varied 12-fold between samples, whereas GST A4 and A5 expression varied ~50–100-fold between samples, well beyond the range of other subunits, suggesting that their expression is highly inducible. Linear correlations (P < 0.001–0.003) existed between levels of the most consistently expressed GST, GSTP1, and total GSTs, GSTA2 and M3, and in GSTM1-positive samples, between GSTM1, M3, and P1. The correlation between GST subunits P1 and M3 was bimodal according to M1 genotype, reflecting the presence of the regulatory element in hGSTM3*B that is linked with the hGSTM1*A genotype. It is concluded that although a degree of regulation of expression of GSTs occurs in human pancreas, the variability of phenotype is high and might obscure the effects of genetic polymorphisms on individual cancer susceptibility. Interindividual variation of GST expression is, therefore, a factor that should be taken account of in epidemiological studies.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The GSTs3 (EC 2.5.1.18) of the alpha, mu, and pi classes are regarded as enzymes of detoxification (1 , 2) . It has been argued, on the basis of isoenzyme substrate specificity, that GST content and composition within an organ may, in part, determine susceptibility to chemical carcinogenesis and/or chemotherapeutic response (1 , 2) . Recent investigations have identified polymorphisms in some of the genes responsible for the GSTs (2, 3, 4) , and these genetic changes can be reflected in changes in enzyme activity that result from altered expression (2 , 3) or differences in specific activity (4, 5, 6, 7, 8) . Thus, the investigation of these genetic polymorphisms in relationship to cancer susceptibility is of considerable interest in this area of research. Only with a few isoenzymes has a genetic polymorphism been consistently related to cancer risk. Notable is the case of the GSTM1 null polymorphism, where individuals with the homozygous null allele, who are unable to express the GSTM1–1 protein, are at an increased risk of several cancers (9, 10, 11) . In most cases, however, changes in GST expression or changes in enzyme-specific activity that result from genetic polymorphisms are moderate (4, 5, 6, 7, 8) .

In previous studies of human pancreas, we have analyzed levels of exogenous (12) and endogenous (13) DNA adducts in relation to genetic polymorphisms of the GSTs. Here, we present data on the variation of expression of GSTs of the alpha, mu, and pi classes in normal pancreatic tissue to illustrate how interindividual GST protein expression can vary 6- to over 100-fold, depending on the subunit. That is, phenotypic expression of GSTs in a single organ may present greater variation in enzyme activity than that predicted from genetic polymorphisms and should be taken into account in epidemiological studies when considering the effect of genetic polymorphisms on cancer susceptibility.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Isolation of GSTs.
Samples of grossly normal human pancreas tissue were obtained from organs donated for transplant or research. After procurement, the organs were kept in cold storage solution on ice for 5–30 h. Organs preserved in this manner for up to 30 h are considered suitable for transplantation (12) . Samples of tissue were snap frozen in liquid nitrogen and stored at -80°C until use. Information on age, race, gender, smoking status, and body mass index were recorded; however, details were not available for all donors (see "Results"). Cytosols and DNA were prepared from 0.5 to 5 g of tissue as described previously (12 , 14) with the inclusion of the protease inhibitors, 4-(2-aminoethyl)benzenesulfonylfluoride (0.01 mM) and pepstatin A (5.0 mg/l; Sigma Chemical Co., St. Louis, MO), and stored at -80°C until use. A GST mixture was prepared from a known aliquot of each cytosol (0.2–0.4 ml, corresponding to 0.1–5.0 mg of protein) using an S-linked glutathione-agarose (Sigma) column (5-mm diameter, 10-mm length), essentially as described previously (15) . Bound GSTs were eluted from the column in 100 mM Tris, 40 mM sodium phosphate, and 50 mM glutathione (pH 9.6) buffer. Under these conditions, 98% of the bound GSTs elute in 1.3 ml.

Analysis of GSTs.
A portion of the GST eluate (0.250 ml) was subjected to HPLC analysis within 2 days of storage (at 4°C), essentially by the method of Ostlund-Farrants et al. (15) . Significant change from this method is the use of a Phenomenex "Jupiter" 5 µm, C18 column (4.6-mm inside diameter x 250 mm; Phenomenex, Torrance, CA), which provided for better GST subunit separation. GST subunits were identified by their HPLC elution profile (16) and electrospray mass spectrometry, using a Platform single quadrupole mass spectrometer (Micromass, Altrincham, United Kingdom). For this purpose, GST subunits purified by HPLC were dried with a stream of argon and dissolved in 50% aqueous acetonitrile/0.5% formic acid. Positive ion data were acquired over the range of m/z 500-2000 in MCA mode using infusion at 5:l/min. The instrument was calibrated using the multiple-charged ion peaks from a separate introduction of horse heart myoglobin (Sigma; M-1882; 16,951.5 Da), and data were acquired under the same conditions as the GST samples. Subtracted data were processed by maximal entropy calculation to convergence using 1 Da/channel resolution. The data were centroided to accurately determine the masses.

Quantitation of GSTs.
Subunits were quantitated by absorption at 214 nm and by reference to known quantities of authentic GSTP1–1, M1b-1b, and A1-1 standards (PanVera, Madison, WI). Those peaks for which no standards were available were quantitated using data established for peaks of similar retention time and peak width. Approximately one-third of analyses were repeated to establish the reliability of quantitation. Correlations between levels of subunit expression within each tissue sample were examined using the program SigmaStat (Jandel Corp., San Rafael, CA). Protein concentrations were determined by the dye-binding method using Bio-Rad reagent (Hercules, CA), according to the supplier’s instructions and using BSA as standard.

Genotyping.
The hGSTM1/M1*0 polymorphism was determined by PCR using albumin as positive control, essentially as reported previously (17) .


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Forty-three, grossly normal pancreas samples were examined from donors of age 8–59 years (mean, 26.3). Age was not available for 11 donors. Eighteen donors were male, 14 female, and 11 had no data; 16 were nonsmokers, 4 were ex-smokers, 15 were current smokers, and 8 were of unknown status. Seventeen donors were Caucasian, and four were African American; race was not available for 22 donors.

Confirmation of Subunit Identity.
On the basis of established HPLC elution profiles for GSTs (15 , 16) , subunits present in pancreas were identified as: P1, M1a, M1b, M2, M3, A1, and A2 (Fig. 1)Citation . Identity was confirmed by mass spectrometry (Table 1)Citation for subunits P1, M1a, M2, M3, and A1. The masses of subunits A1 and M3 corresponded to the posttranslationally modified NH2-terminal acetylated derivatives (16) . The P1 subunit subjected to mass spectrometry was derived from a homozygous P1*B donor (data not shown), and the predicted mass agrees with that of the Val104 form (4) .



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 1. Separation of GST subunits by reverse-phase HPLC. Approximately 1–10 µg of protein from a GST pool prepared by glutathione-agarose affinity chromatography was subjected to reverse-phase HPLC on a Phenomenex "Jupiter" column (4.6 x 250 mm) and eluted with a 60-min linear gradient of 35–65% acetonitrile (plus 0.04% trifluoroacetic acid)-water (plus 0.06% trifluoroacetic acid). Absorption was monitored at 214 nm. A, profile typical of most samples. B, a GSTM1b-positive sample illustrating the partial resolution of GSTA2 subunit forms. C, a GSTM1a-positive sample illustrating a high level of expression of GST subunit {alpha} "y"(GST subunit A4, see "Results"). D, a GSTM1-negative sample illustrating a high level of expression of GST subunit {alpha} "x"(designated subunit A5, see "Results"). Peaks marked * are the apparent mixed GS-protein disulfides that elute slightly earlier than the parent subunits. The peak labeled A2 of A is the Thr111 Glu209 form; those of C and D are the Ser111Glu209 form. Those of B include both the Thr111, Glu209 (later eluting) and the Ser111, Ala209 (earlier eluting) forms. The peak labeled M is a probable mu class minor component.

 

View this table:
[in this window]
[in a new window]
  		Table 1 Observed and predicted masses (Da) of GST subunits

 
The HPLC elution pattern of GST subunit A2 was complexed with the peak appearing as a "doublet" in ~50% of the analyses (Fig. 1B)Citation . Detailed examination of 13 samples showed that the later eluting material contained only a protein of mass corresponding to the reported mass of N-acetylated GSTA2-Thr111 (16) , and the earlier eluting material contained a protein of mass corresponding to that of the known N-acetylated A2-Ser111 variant (16 , 18) . In addition, the earlier eluting material of several samples contained a protein of mass ~25,520 Da, corresponding to the mass of the N-acetylated product of the hGSTA2 clone reported by Röhrdanz et al. (19) , which codes for a GSTA2-Ser111, Ala209 variant.

Two subunits that were not readily identifiable on the basis of known elution profiles of human GSTs were present as major components in several samples (Fig. 1, C and DCitation , peaks x and y). These were determined to be isoforms within the alpha class by their cross-reaction with antibodies raised to the GST alpha class but not to those of the mu or pi classes (results not shown). Automated Edman degradation of the HPLC-purified subunits demonstrated that they possessed blocked NH2 termini; GSTP1 retained a free NH2 terminus after an identical purification procedure. Their masses (Table 1)Citation do not agree with the deduced masses for any known human alpha class GST (16 , 20, 21, 22) . However, the later eluting peak "y" apparently represents the GSTA4 subunit because its mass corresponds to that of putative NH2 terminally acetylated hGSTA4 gene product (20 , 21) . The mass of subunit "x" does not correspond to that of any known alpha class GST nor to likely posttranslationally modified products and appears to be a novel form designated here as GSTA5.

Minor components eluting in the region of subunits M4 and M5 (16) were inconsistently expressed in most samples (labeled "M" in Fig. 1Citation ). The peak from one sample that contained the greatest amount of such material gave a mass corresponding to that of the deduced M5 sequence quoted by Rowe et al. ( Ref. 16 ; although not that of the observed mass for testicular GSTM5). This region of the HPLC profile was variable between samples, and because of the small amounts of protein present, we have not been able to investigate the identity of these minor components further. Two further minor components (* of Fig. 1Citation ) appear to represent anomalous modification of subunits P1 and M3 after storage of GST pools, because during repeat analysis, they increased in height as the P1 and M3 peaks decreased in height. Their masses (i.e., mass of adjacent peak plus ~300 Da, results not shown) are consistent with the formation of mixed protein-SG disulfides, as suggested by Rowe et al. (16) .

Variation of Expression of GST Subunits.
Distributions of data for GST subunit expression are presented in Fig. 2Citation , and variation of expression is summarized in Table 2Citation . Total GST expression among the samples varied by 7.3-fold. GST subunits identified in all samples were P1, M2, M3, and A1 (Figs. 1Citation and 2)Citation . All samples expressed a form of GSTA2. Of these A2 forms, A2-Thr111 alone was expressed in 13 samples; a mixture of A2-Thr111 plus either A2-Ser111or A2-Ser111/Ala209 was expressed in 20 samples, and either A2-Ser111 or A2-Ser111/Ala209 was expressed in 10 samples. Although unambiguous assignment to subunits A2-Ser111 or A2-Ser111/Ala209 was not made in all cases, tissues were identified that expressed only one of the three A2 forms or any two of them. Because of this complexity, GST subunit A2 forms have been grouped together for analysis.



View larger version (50K):
[in this window]
[in a new window]
  		Fig. 2. Expression of GST subunits in human pancreas. For each subunit, samples have been arranged in order of increasing expression. Note, therefore, that the order of samples is different within each panel. For subunits A4, A5, and M2, the lowest expression has been set at the lower limit of detection.

 

View this table:
[in this window]
[in a new window]
  		Table 2 Range and fold variation of expression of GST subunits in human pancreas

 
Levels of expression varied 6–100-fold, according to subunit (Table 2)Citation . GST subunits P1 and A2 form the bulk of GSTs present (52–92% of total GSTs) and were the least variable, both in terms of absolute expression and percentage of total GSTs (Table 2)Citation . Superimposed on this pattern were subunits M1a, M1b (Fig. 1, B and C)Citation , A4 (Fig. 1C)Citation and A5 (Fig. 1D)Citation . Twelve donors were determined to be of the GSTM1-positive genotype, and 17 were GSTM1-null (genotype not available for 14 tissues). An additional nine donors, for which genotype was not available, were determined to be GSTM1-positive on the basis of presence of the M1 subunit. No GSTM1 subunit was detected in tissue from the GSTM1-null individuals; and, within the GSTM1-positive tissues, expression varied 12.5-fold, from 0.04 to 0.5 µg/mg cytosolic protein (Table 2)Citation . Thirteen samples showed expression of only subunit M1a, one M1a and M1b (in approximately equal amounts), and six M1b only. Subunits A4 and A5 are highly variable in expression (Fig. 2Citation ; Table 2Citation ). Subunit A4 was apparently absent in 34 samples (set at the lower limit of detection in Fig. 2Citation ), expressed at low levels (0.1–0.7 µg/mg cytosolic protein) in eight samples and highly expressed in only one sample, in which it represented a major component accounting for 18% of the total (Table 1Citation and Fig. 1CCitation ). A5 was expressed at low levels of 0.02–0.25 µg/mg cytosolic protein in all but four samples (Fig. 2)Citation in which it formed a major component accounting for 9–20% of the total (e.g., Fig. 1DCitation ).

The lower limit of quantitation of GSTP1 and mu class GSTs was 0.01 µg/mg cytosolic protein, and for GSTs of the alpha class, 0.02 µg/mg cytosolic protein. Repetition of analyses established that quantitation was reliable to within ± 8%; HPLC analyses were reproducible to within ± 3%.

Correlations between GST Subunit Expression.
For each sample, the total amount of GST subunits did not correlate with age, smoking status, body mass index, or gender (results not shown). Nor was there any significant correlation between amounts of any one GST subunit and these variables (results not shown).

Comparisons between levels of the least variable major component, the GSTP1 subunit, with total GSTs, GSTA2 or GSTM3, for each of the samples established a correlation between expression of the subunit pairs (Table 3Citation and Figs. 3Citation and 4Citation ). Excluding M1-null individuals, there was also a significant correlation between GSTs P1 and M1 (Table 2Citation ; Fig. 3BCitation ) and M3 and M1 (Fig. 3D)Citation . In addition, there was a trend for higher expression of total GSTs and of subunits A1 or A2 to be associated with higher expression of the minor GST components; subunits M1 (in M1-positive samples) and M2 (r < 0.5; P = 0.01–0.03). Those samples that expressed GSTs A4 and A5 at high levels (Fig. 1, C and DCitation ; Fig. 2Citation ) did not, otherwise, exhibit an anomalous pattern of GST expression. Expression of these subunits was not correlated with that of other alpha class GSTs nor with each other.


View this table:
[in this window]
[in a new window]
  		Table 3 Spearman rank order correlation coefficients and associated Ps between levels of expression of GSTP1 and other GST subunits

 


View larger version (35K):
[in this window]
[in a new window]
  		Fig. 3. Examples of correlations between GST subunit expression: A, GSTA1 against GSTP1, showing lack of correlation. B, GSTM1 against GSTP1 (i.e., GSTM1a and/or M1b but excluding M1-null samples). C, GSTA2 against GSTP1. D, GSTM1 (omitting GSTM1 null samples) against GSTM3. Lines represent the regressions.

 


View larger version (37K):
[in this window]
[in a new window]
  		Fig. 4. Correlation of expression of GSTM3 with GSTP1 according to GSTM1 status. A, all samples; B, GSTM1b or M1-null samples. C, GSTM1-null samples. D, GSTM1a samples, Lines represent the regressions.

 
The mean levels of M3 expression in the two groups "M1a phenotype" and "M1-null or M1b phenotype" were not significantly different in terms of µg M3/mg cytosolic protein (0.37 ± 0.21 and 0.32 ± 0.14, respectively) nor in percentage composition (3.83 ± 1.7% and 3.90 ± 1.4%, respectively). However, when normalized to GSTP1 expression, there was a relatively higher GSTM3 expression in the GSTM1a-positive individuals than the GSTM1b or M1-null individuals (Fig. 4)Citation .


   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Definition and Variation of the GST Phenotype in Pancreas.
This study of GSTs in normal human pancreas shows that, despite the wide variation in expression, the general pattern of expression of GST subunits P1, A1, A2, M2, and M3 is conserved. Total GST content varied 7-fold, and levels of expression among these five subunits varied 6–30-fold. Within this variation, there is a "core" of GSTs, subunits P1, A2, and M3, that makes up the bulk of GST content of pancreas and in which variation, both in absolute terms and composition, is less, i.e., 6–15-fold in amounts and 3–7-fold in composition. These appear to be constitutive elements of the pancreas. Thus, the pancreatic GST phenotype is defined by an ~7-fold variation in GST subunits P1and M3 and 15–30-fold variation in subunits A2 and A1. Although tissue-specific expression of GSTs in pancreas has been described on the basis of a few tissue samples (16 , 23) , the physiological and genetic basis for this is still unknown. Additionally, the phenotype appears to be independent of age (8–59 years), gender, and smoking status. However, it should be noted that GST composition is highly variable within different cell types of the pancreas (24) .

Superimposed on this pattern is the expression of GST subunits M1a, M1b, A4, and A5. Although the expression of GSTM1 depends on the genotype, within the group of GSTM1-positive individuals, low level of M1 expression is also part of the pancreas phenotype.

Complexity of the Alpha Class GSTs in Pancreas.
Particularly notable and complex is the high level of expression of GSTs of the alpha class in pancreas, where four alpha class GSTs (subunits A1, A2, A4, and A5) are expressed. Concerning GSTA2, three variants are known, two on the basis of the amino acid substitution Thr-111->Ser (18 , 25) and the third on the basis of the GSTA2 gene isolated by Röhrdanz et al. (19) , in which the base substitutions at 401 and 695 bp code for the amino acid changes Thr-111->Ser, Glu-209->Ala. Our assignment to these forms is based on the observed masses of the HPLC-purified subunits, which agree with the predicted amino acid substitutions, taking into account the known posttranslational modification of human alpha class GSTs (removal of the NH2-terminal methionine and acetylation of the resultant alanine; Ref. 16 ). The genomic GSTA2 sequence of Röhrdanz et al. (19) apparently differs from that of Klöne et al. (25) in the approximate lengths of introns 1 and 6, and it has been suggested, on this basis, that the two GSTA2 clones represent distinct genes. The gene for the A2-Ser111 variant has not been reported. It is not clear, therefore, whether the three A2 proteins are the products of separate gene loci or whether they represent allelic variation within the A2 locus. Despite the reported differences in intron length between the known A2 genes, they are identical in the 5' upstream region for 310 bp (26) and may, therefore, be subject to coregulation. Thus they may behave as allelic forms, even if they are distinct genes. Because of these ambiguities, we have regarded the A2 proteins as the products of allelic variation, which is consistent with their pattern of expression in pancreas. Whatever their nature, GSTA2 protein is clearly subject to polymorphism, and there are individuals who appear to be homozygous or heterozygous within the three forms.

In addition to GSTs A1 and A2, DNA sequences are available for GSTA3 and GSTA4. As mentioned above, the late-eluting HPLC peak "y" appears to represent the NH2 terminally acetylated protein product of the GSTA4 gene. GSTA4 is known to be expressed in pancreas on the basis of the presence of its mRNA (20) but has not, previously, been observed as a native protein. The available clones for GSTA3 are incomplete cDNAs (21 , 22) , and an accurate, deduced mass is not available. However, the mass of subunit A5 does not correspond to the amino acid sequence deduced from the purported full-length GSTA3 cDNA (21) , nor to likely posttranslationally modified products. On this basis, GSTA5 appears to be a novel gene product. Whether subunits A4 and A5 are related to the basic GSTs of human skin (27) or lung (28) is unclear. However, their highly variable pattern of expression, independent of that of subunits A1 and A2, suggests that they are products of highly inducible gene expression.

All known human class alpha GSTs have peroxidase activity (2 , 20 , 21 , 28 , 29) . Additionally, GSTA4 has high activity toward the genotoxic hydroxyalkenal products of lipid peroxidation (21 , 29) . Because a high-energy diet, possibly resulting in high levels of lipid peroxidation products, is a risk factor in pancreatic cancer (30 , 31) , variation between individuals in type and levels of alpha class GSTs may be of considerable importance in pancreatic disease.

Regulation of GSTs.
The genes for the alpha mu and pi classes of GSTs are located on separate chromosomes (2 , 32) , and a molecular basis for the apparent coregulation of expression of GSTs P1, A2, and M3 in the pancreas is not clear. In the case of GSTP1, the gene is known to contain a complex of regulatory elements between -95 to 55 bp [summarized in Hayes and Pulford (2) ]. These include an nuclear factor-{kappa}B-like (silencer) element and ARE/TRE (AP-1 recognition site) consensus motifs contained in the "C1-region" (33) . This last is absolutely required for GSTP1 transcription. Activation of the C1 region (e.g., by binding of the Fos-Jun complex) is apparently cell type specific (34 , 35) . GSTP1 expression is also enhanced by insulin (36) . Alpha class GST mRNA and protein levels are selectively inducible in human hepatocytes in primary culture by dithiolethiones, phenobarbital and 3-methylcholathrene (37) . In addition, a daily intake of Brussels sprouts also leads to selective GST alpha class protein induction in humans (38) . Variation in diet may, therefore, be the cause of the wide variation of expression of GST subunits A1, A4, and A5 in the pancreas. Although potential AP1, AP2, GRE, and HNF1 regulatory elements have been identified in the 5'-upstream regulatory region of human GST alpha class genes (26 , 39 , 40) , no elements corresponding to ARE or XRE consensus sequences have been revealed, and the mechanism by which inducers act is not known. Also, there are no similarities between the 5'-upstream regions of the human GST alpha and Pi genes (26) , nor between those of GSTM1 and M3 (41) . It has been suggested that SP1 elements in the GSTM3 gene are responsible for low levels of expression of M3 in most tissues (41) . Whether such sites offer a mechanism for coinduction is not clear.

GSTs and Cancer Susceptibility.
Much attention has been focused on GSTs of the mu and pi classes because they have activity toward the carcinogenic polycyclic aromatic hydrocarbon metabolites that are a consequence of smoking and dietary exposure (1 , 2) , and there are also genetic polymorphisms in these classes that are reflected in levels of enzyme expression or specific activity (2, 3, 4, 5, 6, 7, 8) . The most marked of these is the M1-null polymorphism, where GSTM1-null individuals cannot express the GSTM1 protein. However, even in this case, correlation between genotype and cancer susceptibility is highly dependent on the organ under study (11) . In the case of pancreas, GSTM1-null genotype does not appear to be a risk factor for pancreatitis or pancreatic cancer, although possession of the GSTM1*B allele may confer risk for pancreatic disease (42) . Nevertheless, our studies illustrate several factors concerning the complexity of the GSTM1-positive phenotype in the pancreas. Although the M1-positive genotype is well defined, the M1-positive phenotype is not, because protein levels vary by 12.5-fold and may consist of the M1a and/or Mlb subunits. Further complexity is introduced by the M1*A allele being in linkage with the M3*B allele and the suggestion that a regulatory element in the latter could cause reduced expression of the M3 subunit in M1*A individuals compared with M1*0 or M1*B individuals (3) . However, in the lung, M1-positive individuals were shown to have higher levels of expression of GSTM3 than M1-negative individuals (43) . This is also seen in the pancreas in the case of GSTM1a-positive samples, but only when GSTM3 expression is normalized to GSTP1 levels. Thus, the effect of GSTM1-deletion is compounded by the tendency toward lower levels of expression of GSTM3.

In GSTP1, at least three alleles are known (4 , 5) in which mutations in exonic sequence lead to proteins that differ in their catalytic properties. However, these differences appear to be moderate. For example, the kcat of the P1a-1a and P1b-1b homodimers toward the carcinogenic benzo(a)pyrene metabolite, (+)-anti-benzo(a)pyrene-7,8-diol-9,10-oxide, determined using cell expression protein, are 1.3 and 4.4 s-1, respectively (5) . This 3.4-fold variation could be masked by variation in expression, such as the 6-fold variation of GSTP1 seen in this study. For example, two samples that were homozygous for the P1*A allele exhibited levels of expression of P1a-1a of 4.1 and 5.3 µg/mg cytosolic protein, and two that were heterozygous, P1*A/P1*B, express P1a and P1b subunits (presumably as a mix of homo- and heterodimers) at 1.2 and 1.7 µg/mg cytosolic protein. Assuming subunit activity to be independent of protein dimer (i.e., a 2.7-fold difference in activity between P1a-1a and P1a-1b), the catalytic activity of the cytosols would be the reverse of that predicted from genotype.

Concluding Remarks.
The data presented here show that although pancreas-specific expression of GSTs is highly regulated, variability of phenotypic expression of GSTs in the pancreas should be taken into account when assessing the possible effects of genetic polymorphisms in the GSTs. The pancreas is not unique in this respect. Similar ranges of variation of GST expression is known to be a property of many human organs and of both normal and tumor tissue, e.g., breast (44) , colon (45, 46, 47) , endometrium (48) , liver (49 , 50) , lung (43 , 47) , kidney (47 , 49) , stomach (46 , 49) , and hair follicles (51) . In addition, the wide variation in expression of GSTs A1, A4, and A5 suggest that they are subject to inducibility beyond the ranges of normal phenotypic expression. Our results also show the complexity of alpha class GSTs in pancreas, a complexity that also exists in other human tissues (27 , 28) and the significance of which remains to be determined. Despite the wide range of variation of expression of GSTs, the effect of minor regulatory elements on expression (e.g., that of GSTM3) are discernible by "normalizing" protein levels to those of the least variable component of the phenotype, GSTP1. This may be a useful approach to examine other factors responsible for GST induction in vivo.


   ACKNOWLEDGMENTS
 
We thank the following individuals for their assistance in obtaining tissue samples: Dr. D. Sutherland, Dr. D. Sharp, Dr. P. Lacy, Dr. P. Gores, W. Minor, J. Field, J. J. Loveras, B. Olack, J. Massengill, M. Dunning, E. Van Bergen, and the staff of LifeSource (Upper Midwest Organ Procurement Organization, Inc., Minneapolis, MN). The authors are grateful to Dr. P. Thompson for helpful comments.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by Grant RO1-CA58697 from the National Cancer Institute/NIH. Back

2 To whom requests for reprints should be addressed, at Division of Molecular Epidemiology, HFT 100, National Center for Toxicological Research, Jefferson, AR 72079-9502. Phone: (870) 543-7596; Fax: (870) 543-7773; E-mail: bcoles{at}nctr.fda.gov Back

3 The abbreviations used are: GST, glutathione S-transferase; HPLC, high-pressure liquid chromatography. Back

Received 7/19/99. Accepted 12/ 2/99.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Coles B., Ketterer B. The role of glutathione and glutathione transferases in chemical carcinogenesis. Crit. Rev. Biochem. Mol. Biol., 25: 47-70, 1990.[Medline]
  2. Hayes J. D., Pulford D. J. The glutathione S-transferase supergene family: regulation of GST and the contribution of the isoenzymes to cancer, chemoprotection, and drug resistance. Crit. Rev. Biochem. Mol. Biol., 30: 445-600, 1995.[Medline]
  3. Inskipp A., Elexperu-Camiruaga J., Buxton N., Dias P. S., MacIntosh J., Campbell D., Jones P. W., Yengi L., Talbot J. A., Strange R. C., Fryer A. A. Identification of polymorphism at the glutathione S-transferase, GSTM3 locus: evidence for linkage with GSTM1*A.. Biochem. J., 312: 713-716, 1995.
  4. Ali-Osman F., Akande O., Antoun G., Mao J-X., Buolamwini J. Molecular cloning, characterization, and expression in Escherichia coli of full length cDNAs of three human glutathione S-transferase Pi gene variants. J. Biol. Chem., 272: 10004-10012, 1997.[Abstract/Free Full Text]
  5. Hu X., Xia H., Srivastava S., Herzog C., Awasthi Y. C., Ji X., Zimniak P., Singh S. V. Activity of four allelic forms of glutathione S-transferase hGSTP1–1 for diol epoxides of polycyclic aromatic hydrocarbons. Biochem. Biophys. Res. Commun., 238: 397-402, 1997.[Medline]
  6. Hu X., Pal A., Krzeminski J., Amin S., Awasthi Y. C., Zimniak P., Singh S. V. Specificities of human glutathione S-transferase isozymes toward anti-diol epoxides of methylchrysenes. Carcinogenesis (Lond.), 19: 1685-1689, 1998.[Abstract/Free Full Text]
  7. Sundberg K., Johansson A-S., Stenberg G., Widersten M., Seidel A., Mannervik B., Jernström B. Differences in the catalytic efficiencies of allelic variants of glutathione transferase P1–1 towards carcinogenic diol epoxides of polycyclic aromatic hydrocarbons. Carcinogenesis (Lond.), 19: 433-466, 1998.[Abstract/Free Full Text]
  8. Sundberg K., Seidel A., Mannervik B., Jernström B. Detoxication of carcinogenic fjord-region diol epoxides of polycyclic aromatic hydrocarbons by glutathione transferase P1–1 variants and glutathione. FEBS Lett., 438: 206-210, 1998.[Medline]
  9. Raunio H., Husgafvel-Pursiainen K., Anttila S., Hietanen E., Hirvonen A., Pelkonen O. Diagnosis of polymorphisms in carcinogen-activating and inactivating enzymes and cancer susceptibility–a review. Gene (Amst.), 159: 113-121, 1995.[Medline]
  10. Nebert D. W., McKinnon R. A., Puga A. Human drug-metabolizing enzyme polymorphisms: effects on risk of toxicity and cancer. DNA Cell Biol., 15: 273-280, 1996.[Medline]
  11. Rebbeck T. R. Molecular epidemiology of the human glutathione S-transferase genotypes GSTM1 and GSTT1 in cancer susceptibility. Cancer Epidemiol. Biomark. Prev., 6: 733-743, 1997.[Abstract]
  12. Anderson K. E., Hammons G. J., Kadlubar F. F., Potter J. D., Kaderlik K. R., Ilett K. F., Minchin R. F., Teitel C. H., Chou H-C., Martin M. V., Guengerich F. P., Barone G. W., Lang N. P., Peterson L. A. Metabolic activation of aromatic amines by human pancreas. Carcinogenesis (Lond.), 18: 1085-1092, 1997.[Abstract/Free Full Text]
  13. Kadlubar F. F., Anderson K. E., Häussermann S., Lang N. P., Barone G. W., Thompson P. A., MacLeod S. L., Chou M. W., Mikhailova M., Plastaras J., Marnett L. J., Nair J., Velic I., Bartsch H. Comparison of DNA adduct levels associated with oxidative stress in human pancreas. Mutat. Res., 405: 125-133, 1998.[Medline]
  14. Culp S. J., Roberts D. W., Talaska G., Lang N. P., Fu P. P., Lay J. O., Jr., Teitel C. H., Snawder J. E., Von Tungeln L. S., Kadlubar F. F. Immunochemical, 32P-postlabeling, and GC/MS detection of 4-aminobiphenyl-DNA adducts in human peripheral lung in relation to metabolic activation pathways involving pulmonary N-oxidation, conjugation, and peroxidation. Mutat. Res., 378: 97-112, 1997.[Medline]
  15. Ostlund-Farrants A-K., Meyer D. J., Coles B., Southan C., Aitken A., Johnson P. J., Ketterer B. The separation of glutathione transferase subunits using reverse-phase high-pressure liquid chromatography. Biochem. J., 245: 423-428, 1987.[Medline]
  16. Rowe J. D., Nieves E., Listowsky I. Subunit diversity and tissue distribution of human glutathione S-transferases: interpretations based on electrospray ionization-MS and peptide-specific antisera. Biochem. J., 325: 481-486, 1997.
  17. Arand M., Mühlbauer R., Hengstler J., Jäger E., Fuchs J., Winkler L., Oesch F. A multiplex polymerase chain reaction protocol for the simultaneous analysis of glutathione S-transferase GSTM1 and GSTT1 polymorphisms. Anal. Biochem., 236: 184-186, 1996.[Medline]
  18. Hayes J. D., Kerr L. A., Cronshaw A. D. Evidence that glutathione S-transferases B1B1 and B2B2 are the products of separate genes and that their expression in human liver is subject to inter-individual variation. Molecular relationship between the B1 and B2 subunits and other alpha class glutathione S-transferases. Biochem. J., 264: 437-445, 1989.[Medline]
  19. Röhrdanz E., Nguyen T., Picket C. B. Isolation and characterization of the human glutathione S-transferase A2 subunit gene. Arch. Biochem. Biophys., 298: 747-752, 1992.[Medline]
  20. Liu S., Stoesz S. P., Pickett C. B. Identification of a novel human glutathione S-transferase using bioinformatics. Arch. Biochem. Biophys., 352: 306-313, 1998.[Medline]
  21. Board P. G. Identification of cDNAs encoding two human alpha class glutathione transferases (GSTA3 and GSTA4) and the heterologous expression of GSTA4–4. Biochem. J., 330: 827-831, 1998.
  22. Suzuki T., Johnston P. N., Board P. G. Structure and organization of the human alpha class glutathione S-transferase genes and related pseudogenes. Genomics, 18: 680-686, 1993.[Medline]
  23. Awasthi Y. C., Sharma R., Singhal S. S. Human glutathione S-transferases. Int. J. Biochem., 26: 295-308, 1994.[Medline]
  24. Collier J. D., Bennett M. K., Hall A., Cattan A. R., Lendrum R., Bassendine M. F. Expression of glutathione S-transferases in normal and malignant pancreas: an immunohistochemical study. Gut, 35: 266-269, 1994.[Abstract/Free Full Text]
  25. Klöne A., Hussnätter R., Sies H. Cloning, sequencing and characterization of the human alpha class glutathione S-transferase gene corresponding to the cDNA clone pGTH2. Biochem. J., 285: 925-928, 1992.
  26. Lörper M., Schulz W. A., Morel F., Warskulat U., Sies H. Positive and negative regulatory regions in promoters of human glutathione transferase alpha genes. Biol. Chem. Hoppe-Seyler, 377: 39-46, 1996.[Medline]
  27. Del Boccio G., Di Ilio C., Ålin P., Jörnvall H., Mannervik B. Identification of a novel glutathione transferase in human skin homologous with class alpha glutathione transferase 2-2 in the rat. Biochem. J., 244: 21-25, 1987.[Medline]
  28. Singhal S. S., Saxena M., Ahmad H., Awasthi S., Haque A. K., Awasthi Y. C. Glutathione S-transferases of human lung: characterization and evaluation of the protective role of the {alpha}-class isoenzymes against lipid peroxidation. Arch. Biochem. Biophys., 299: 232-241, 1992.[Medline]
  29. Habatsch I., Ridderström M., Mannervik B. Human glutathione transferase A4–4: an alpha class enzyme with high catalytic efficiency in the conjugation of 4-hydroxynonenal and other genotoxic products of lipid peroxidation. Biochem. J., 330: 175-179, 1998.
  30. Pancreas. In: Food, Nutrition and the Prevention of Cancer: a Global Perspective, pp. 176–197. Washington, DC: World Cancer Research Fund and American Institute for Cancer Research, 1997.
  31. Anderson K. E., Potter J. D., Mack T. Pancreatic Cancer Ed. 2 Schottenfeld D. Fraumeni J. F., Jr. eds. . Cancer Epidemiology and Prevention, : 725-771, Oxford University Press New York 1997.
  32. Islam M. Q., Platz A., Szpirer J., Szpirer C., Levan G., Mannervik B. Chromosomal localization of human glutathione transferase genes of classes alpha, mu, and pi. Hum. Genet., 82: 338-342, 1989.[Medline]
  33. Morrow C. S., Goldsmith M. E., Cowan K. H. Regulation of human glutathione S-transferase pi gene transcription: influence of 5'-flanking sequences and trans-activating factors which recognise AP-1 binding sites. Gene (Amst.), 88: 215-225, 1990.[Medline]
  34. Moffatt G. J., McLaren A. W., Wolf C. R. Involvement of Jun and Fos proteins in regulating transcriptional activation of human pi class glutathione S-transferase gene in multidrug-resistant MCF7 breast cancer cells. J. Biol. Chem., 269: 16397-16402, 1994.[Abstract/Free Full Text]
  35. Henderson C. J., McLaren A. W., Moffatt G. J., Bacon E. J., Wolf C. R. Pi class glutathione S-transferases: regulation and function. Chem.-Biol. Interact., 111–112: 69-82, 1998.
  36. Xia C., Taylor J. B., Spencer S. R., Ketterer B. The human glutathione S-transferase P1-1 gene: modulation of expression by retinoic acid and insulin. Biochem. J., 292: 845-850, 1993.
  37. Morel F., Fardel O., Meyer D. J., Langouët S., Gilmore K. S., Meunier B., Tu C-P. D., Kensler T. W., Ketterer B., Guillouzo A. Preferential increase of glutathione S-transferase class alpha transcripts in cultured human hepatocytes by phenobarbital, 3-methylcholanthrene, and dithiolethiones. Cancer Res., 53: 231-234, 1993.[Abstract/Free Full Text]
  38. Bogaards J. J. P., Verhagen H., Willems M. I., van Poppel G., van Bladeren P. J. Consumption of Brussels sprouts results in elevated {alpha}-class glutathione S-transferase levels in human blood plasma. Carcinogenesis (Lond.), 15: 1073-1075, 1994.[Abstract/Free Full Text]
  39. Morel F., Schulz W. A., Sies H. Gene structure and regulation of expression of human glutathione S-transferases alpha. Biol. Chem. Hoppe-Seyler, 375: 641-649, 1994.[Medline]
  40. Lörper M., Clairmont A., Carlberg C., Sies H. Identification of two activating elements in the proximal promoter region of the human glutathione transferase-A1 and -A2 genes. Arch. Biochem. Biophys., 359: 122-127, 1998.[Medline]
  41. Patskovsky Y. E., Huang M-Q., Takayama T., Listowsky I., Pearson W. R. Distinctive structure of the human GSTM3 gene-inverted orientation relative to the Mu class glutathione transferase gene cluster. Arch. Biochem. Biophys., 361: 85-93, 1999.[Medline]
  42. Bartsch H., Malaveille C., Lowenfels A. B., Maisonneuve P., Hautefeuille A., Boyle P. Genetic polymorphisms of N-acetyltransferases, glutathione S-transferase M1 and NAD(P)H quinone oxidoreductase in relation to malignant and benign pancreatic disease risk. The International Pancreatic Disease Study Group. Eur. J. Cancer Prev., 7: 215-223, 1998.[Medline]
  43. Nakajima T., Elovaara E., Anttila S., Hirvonen A., Camus A-M., Hayes J. D., Ketterer B., Vainio H. Expression and polymorphism of glutathione S-transferase in human lungs: risk factors in smoking related cancer. Carcinogenesis (Lond.), 16: 707-711, 1995.[Abstract/Free Full Text]
  44. Kelley M. K., Engqvist-Goldstein A., Montali J. B., Schmidt D. E., Jr., Kauvar L. M. Variability of glutathione S-transferase isoenzyme patterns in matched normal and cancer human breast tissue. Biochem. J., 304: 843-848, 1994.
  45. Hengstler J. G., Böttger T., Tanner B., Dietrich B., Henrich M., Knapstein P. G., Junginger T., Oesch F. Resistance factors in colon cancer tissue and the adjacent normal tissue: glutathione S-transferases {alpha} and {pi}, glutathione, and aldehyde dehydrogenase. Cancer Lett., 128: 105-112, 1998.[Medline]
  46. deBruin W. C. C., Wagenmanns M. J. M., Board P. G., Peters W. H. M. Expression of glutathione S-transferase {theta} class isoenzymes in human colorectal and gastric cancers. Carcinogenesis (Lond.), 20: 1453-1457, 1999.[Abstract/Free Full Text]
  47. Temellini A., Castiglioni M., Giuliani L., Mussi A., Giulianotti P. C., Pietrabissa A., Angeletti C. A., Mosca F., Pacifici G. M. Glutathione conjugation with 1-chloro-2,4-dinitrobenzene (CDNB): interindividual variability in human liver, lung, kidney, and intestine. Int. J. Clin. Pharmacol. Ther., 33: 498-503, 1995.[Medline]
  48. Barnette K. G., Sarkar M. A., Glover D. D., Li P., Boyd C., Lalka D. Glutathione S-transferase in human endometrium: quantitation and interindividual variability in isoform content. Gynecol. Obstet. Investig., 47: 114-119, 1999.[Medline]
  49. Howie A. F., Forrester L. M., Glancey M. J., Schlager J. J., Powis G., Beckett G. J., Hayes J. D., Wolf C. R. Glutathione S-transferase and glutathione peroxidase expression in normal and tumour human tissues. Carcinogenesis (Lond.), 11: 451-458, 1990.[Abstract/Free Full Text]
  50. Takamatsu Y., Inaba T. Inter-individual variability of human hepatic glutathione S-transferase isozymes assessed by inhibitory capacity. Toxicology, 88: 191-200, 1994.[Medline]
  51. Puerto A. M., Nogues M. R., Cervello I., Giralt M., Ortin F., Mallol J. Glutathione transferase in normal human anagen hair follicles. Rev. Esp. Fisol., 50: 103-108, 1994.



This article has been cited by other articles:


Home page
 CarcinogenesisHome page
F. Gemignani, S. Landi, N. Szeszenia-Dabrowska, D. Zaridze, J. Lissowska, P. Rudnai, E. Fabianova, D. Mates, L. Foretova, V. Janout, et al.
Development of lung cancer before the age of 50: the role of xenobiotic metabolizing genes
Carcinogenesis, June 1, 2007; 28(6): 1287 - 1293.
[Abstract] [Full Text] [PDF]

Home page
 BloodHome page
R. K. Dasgupta, P. J. Adamson, F. E. Davies, S. Rollinson, P. L. Roddam, A. J. Ashcroft, A. M. Dring, J. A. L. Fenton, J. A. Child, J. M. Allan, et al.
Polymorphic variation in GSTP1 modulates outcome following therapy for multiple myeloma
Blood, October 1, 2003; 102(7): 2345 - 2350.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
H. Ahsan and A. G. Rundle
Measures of genotype versus gene products: promise and pitfalls in cancer prevention
Carcinogenesis, September 1, 2003; 24(9): 1429 - 1434.
[Abstract] [Full Text] [PDF]

Home page
 JNCI J Natl Cancer InstHome page
E. J. Duell, E. A. Holly, P. M. Bracci, M. Liu, J. K. Wiencke, and K. T. Kelsey
A Population-Based, Case-Control Study of Polymorphisms in Carcinogen-Metabolizing Genes, Smoking, and Pancreatic Adenocarcinoma Risk
J Natl Cancer Inst, February 20, 2002; 94(4): 297 - 306.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
A. Loktionov, M. A. Watson, M. Gunter, W. S.L. Stebbings, C. T.M. Speakman, and S. A. Bingham
Glutathione-S-transferase gene polymorphisms in colorectal cancer patients: interaction between GSTM1 and GSTM3 allele variants as a risk-modulating factor
Carcinogenesis, July 1, 2001; 22(7): 1053 - 1060.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Coles, B. F.
Right arrow Articles by Kadlubar, F. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Coles, B. F.
Right arrow Articles by Kadlubar, F. F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Cell Growth & Differentiation