Glutathione Transferase Activity and Isoenzyme Composition in Primary Human Breast Cancers

Thomas C. Shea, Ginger Claflin, Kenine E. Comstock, et al. Cancer Res 1990;50:6848-6853. Published online November 1, 1990.

Updated Version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/50/21/6848

Citing Articles

This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/50/21/6848#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

(CANCER RESEARCH 50, 6848-6853. November 1. 1990]

Glutathione Transferase Breast Cancers1

Activity and Isoenzyme Composition in Primary Human

Thomas C. Shea,2 Ginger Claflin, Renine E. Comstock, Barbara J. S. Sanderson, Nelson A. Burstein, Edward J. Keenan, Bengt Mannervik, and W. David Henner1
Department of Medicine, University of California at San Diego, San Diego, California [T. C. S.]; Division of Hematology and Medical Oncology, Oregon Health Sciences University, Portland, Oregon 97201 [G. C., K. E. C, B. J. S. S., E. J. K., W. D. H.J; Department of Pathology, St. Elizabeth's Hospital, Boston, Massachusetts 02135 [N. A. B.J and Department of Biochemistry, University of Uppsala, Uppsala, Sweden [B. M.]

GST isoenzymes are believed to play an important role in the cellular metabolism and detoxication of electrophilic com The human glutathione transferases (GSTs) are a multigene family of pounds by their conjugation to glutathione (6). Recent investi detoxication enzymes with patterns of expression that are both tissue gations have revealed that these enzymes are capable of utilizing specific and genetically determined. Changes in the levels of one or more several clinically active alkylating agents as substrates for con GST isoenzymes have been associated with the development of anticancer jugation to glutathione (7, 8). drug resistance in cultured cell lines. In this study, total GST activity Increased GST activity has been reported in a number of and GST isoenzyme composition have been determined for 45 primary rodent (8-11) and human (12-16) tumor cell lines selected in human breast carcinomas using a l-chloro-2,4-dinitrobenzene substrate assay and Western blotting, respectively. The GST activity ranged from vitro for resistance to anticancer drugs. There are also reports 5-208 mU/mg protein with a mean of 67 mU/mg protein (44SD). of elevated GST activity in cell lines derived from tumors in GST-T isoenzyme protein was detectable on Western blots in 44 of 45 patients who had developed clinical resistance to either cisplatin samples. MMClass GST protein was detected in 18 of 38 samples and or chlorambucil (17, 18). However, the exact role of the GST undetectable in 20 of the 38 samples tested. By polymerase chain reaction elevations in the resistance of these cells to antineoplastics analysis of genomic DNA, the absence of mu class GST in breast tumors was determined to be due to the deletion of the gene for GST-M in the remains to be established (16). In addition to the roles of GST in drug metabolism, detoxi DNA of those tumors. None of the 43 primary human breast cancer samples tested contained detectable alpha class GST protein. Neither cation, and resistance, there are several reports suggesting that the total GST activity of tumor samples, the quantity of GST-protein, a particular GST isoenzyme, GST-Tr, may serve as a biochemical nor the presence or absence of mu class GST correlated with other factors marker for neoplastic transformation. This was suggested by known to be of prognostic significance including tumor size, nodal status, work using a rat tumor model and by measurement of GST-?r estrogen receptor protein positivity, or progesterone receptor protein in primary human tumor samples (19-21). Recent reports indicate that levels of GST-Tr mRNA (22) and protein (23) are positivity. Substantial differences exist among primary breast carcinomas in both the amount of GST activity and GST isoenzyme composition. negatively correlated with ERP content in primary human However, these are not tightly linked either to tumor stage or to hormone breast cancer specimens, suggesting that increased GST-Tr ac receptor status. Whether the levels of these enzymes are independent tivity is correlated with a marker (i.e., ERP negativity) that predictors of either risk of recurrence or response to anticancer therapy indicates a worse prognosis in this disease. has yet to be tested directly. The current study was undertaken to determine the levels of GST activity and isoenzyme composition of primary human breast cancers from untreated patients. The long-term goal of INTRODUCTION The human GSTs4 are a multigene, isoenzyme family. The these studies is to determine whether the levels of the various GST isozymes might serve as predictors of either prognosis or human cytosolic GST isoenzymes can be classified by their response to therapy. The results obtained have been compared substrate specificities, isoelectric points, amino acid sequence with prognostic variables of known significance for these pa homologies, and immunological relationships into three major tients to determine whether GST activity or isoenzyme content classes termed alpha, mu, and pi (1). Within the alpha and mu might be linked to known prognostic variables (24). The current classes are multiple isoenzymes [e.g., GST,GST-/3,.. .GST-e report demonstrates that primary breast tumors vary widely as (alpha class), GST-/u, GST-i/- (mu class)]. Only one GST isoento both total GST activity and GST isoenzyme content and zyme, GST-Tr, has been identified in the pi class. These isoenhave no strong linkage to known prognostic variables. zymes are expressed in a tissue-specific pattern and one or more GST isoenzymes are present in a variety of normal and malig nant human tissues including breast carcinomas (2-5). The MATERIALS AND METHODS ABSTRACT
Received 3/19/90; accepted 7/30/90. 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. 1A preliminary report of this work appeared in abstract form in Proc. Am. Assoc. Cancer Res., 29: 1208, 1988. Supported by grants from Bristol-Myers Squibb Co. and the Swedish Cancer Society. 2T. C. S. is a recipient of a Faculty Development Award in Clinical Pharma cology from the Pharmaceutical Manufacturers Association Foundation, Wash ington, DC 20005. ' To whom requests for reprints should be addressed, at Division of Hematol ogy and Medical Oncology, Oregon Health Sciences University L586, 3181 SW Sam Jackson Park Road, Portland, OR 97201. ' The abbreviations used are: GST, glutathione transferase; PCR, polymerase chain reaction; CDNB, l-chloro-2,4-dinitrobenzene; ERP, estrogen receptor pro tein; PRP, progesterone receptor protein; HPRT, hypoxanthine-guanine-phosphoribosyltransferase.

Tumor Samples. All tumor samples were selected from a bank of cryopreserved tumors at the hormone receptor laboratories of St. Eliz abeth's Hospital (Boston, MA) or Oregon Health Sciences University

(Portland, OR) which had been frozen within 2 h of surgery and from which samples had been obtained for estrogen and progesterone recep tor analysis. Estrogen and progesterone receptor data were obtained by standard methods utilizing freshly prepared cytosol extracts of tumor specimens using a modification of the dextran/charcoal technique (25). Tumors are defined as ERP or PRP positive if they contain > 10 fmoles/ mg protein or a20 fmoles/mg protein respectively. Histological clas sification, estrogen and progesterone receptor determination, and GST analysis were performed independently without knowledge of the results of the other analyses. All studies were performed with the approval of the Institutional Review Boards of the contributing institutions. 6848

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

GLUTATHIONE

TRANSFERASES

OF HUMAN BREAST CANCER

GST Analysis. Cytosol extracts from tumor samples of approxi mately 100 mg dry weight were obtained as previously described (2). GST activity was based on enzyme-dependent conjugation of reduced glutathione with CDNB and expressed as mU conjugate formed/mg protein (1 mU = 1 nmol conjugate formed/min) (2, 26). All results represent the means of three separate assays for each individual tumor extract. For analysis of samples by Western blotting, aliquots of cytosolic extract containing 150 ig protein were resolved by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose by standard techniques (27). Replicate blots were probed with a 1:500 dilution of antisera raised as previously described (2) in rabbits immunized with purified human GSTs, either GST-T (2), GSTu (Medlabs, Dublin. Ireland), or GST- (28). Second antibody binding and color development were performed as previously described (2). PCR. Primers for amplifying the GST-M gene segment corresponding to exon 4, intron 4, and exon 5 were 5'-CTGCCCTACTTGATTGATGGG-3' and 5'-CTGGATTGTAGCAGATCATGC-3' and were obtained from Midland Certified Reagent Co. (Midland, TX). Primers for amplification of exon 2 of the HPRT gene were 5'-CTGTCCATAATTAGTCCATGAG-3' and 5'-ATTAGTGATGATGAACCAGGTT-3' (29). Genomic target DNA for PCR was prepared from human breast carcinoma samples by standard procedures (30). The polymerase chain reaction was carried out as previously described using 1.0 /jg of target DNA (31). Amplification of exon 2 of the HPRT gene was used as a control for the target DNAs that were negative for the GST-/J gene by PCR. All 18 DNA samples were effective targets for amplification of exon 2 of the HPRT gene (data not shown). PCR products were resolved by electrophoresis on 2.1 % agarose gels, stained with ethidium bromide, and photographed under UV light. Statistical Methods. Correlation coefficients were determined for total GST activity as a function of either estrogen or progesterone receptor content (n = 45). Student's t tests of significance were applied to mean GST activity levels as a function of primary tumor size (Tl, <2 cm; T2, 2-5 cm; T3, >5 cm), lymph node involvement (0-4, >4 nodes involved), age (<50 years or >50 years, the presence or absence of estrogen (>10 fmol receptor/mg protein) and progesterone (>20 fmol receptor/mg protein) receptors, and the presence or absence of mu class GST. x2 analysis was used to compare the presence or absence of mu class GST with the other categorical (nonlinear) variables listed above. The Wilcoxon-Mann-Whitney test for two independent variables was used to compare the GST activity and GST-ir levels of the ERPpositive group versus the ERP-negative group.

Table 1 GST activity and human breast tumor characteristics GST activity"
Characteristic"

(mU/mg)
45

P valuer

Total ER+ (;>10 fmol/mg)'' ER-PR+ (>20 fmol/mg)'' PR-Mu Class + TI M u Class (<2 cm) Til (2-5 cm) cm)<4 Till (>5 nodes nodesAge >4

67 44

49 1117281820614 2564 70 5668 3769 51 2958 60 40 67 36 9IX8162966 5948 60 26 5160 71

<50 yr 5171 Age >50 yr34 410.780.800.670.630.250.41 " Determined as described in "Materials and Methods" or from clinical records. 4 GST activity was measured by enzyme-dependent conjugation of reduced glutathione with CDNB. as described in "Materials and Methods." c Student's t tests of significance for differences between GST activities within each category. ''Not significant (/>> 0.10) by Wilcoxon-Mann-Whitney test.

nodal status was available for 29 and 26 of the patients, respec tively. Six tumors were <2 cm in size (Tl), 14 were 2-5 cm (T2), and 9 were >5 cm (T3). Eighteen tumors were associated with <4 involved lymph nodes and 8 tumors were associated with >4 involved nodes. The median age of these 45 patients was 57 years with 29 patients >50 years and 16 patients <50 years. GST Analysis. Forty-five samples were assayed for GST activity utilizing the universal GST substrate CDNB. The mean GST activity was 67 mU/mg of extract protein (SD, 44 mU/ mg; range, 5-208 mU/mg). We determined the extent to which the variation in GST activity among different tumors was due to variations in GST activity in different portions of the same tumor sample. A subset of the tumors was randomly selected, and three portions of each tumor were separated for individual preparation of cytosolic extracts. Each extract was assayed in triplicate for total GST activity (see "Materials and Methods").

The mean coefficient of variation for individual extracts pre pared from different regions of the same tumor was 36% (Fig. 1). Seven samples of normal breast tissue from reduction mammoplasties were analyzed and found to contain a mean GST activity of 65 mU/mg protein (SD, 27mU/mg; range, 32109 mU/mg protein). Sufficient tumor extract was available to evaluate GST-?r, alpha class GST, and mu class GST content by Western blotting and medical records for these patients, 13 samples were ex in 45, 43, and 38 specimens, respectively (Fig. 2). Although the cluded from this analysis for the following reasons: recurrent classification of samples as positive or negative for each GST disease, 6 samples; non-breast carcinoma, 4 samples; insuffi isozyme class on Western blotting was qualitative, the differ cient tumor in specimen, 3 samples. Forty-five samples with ence between positive and negative samples was always clear invasive primary breast carcinoma were analyzed (Table 1). and the classification of the samples was reproducible among Forty-two specimens were composed of infiltrating ductal car multiple independent observers. Forty-four of 45 samples con tained detectable levels of GST-Tr isoenzyme. The one lacking cinoma. One sample contained predominantly intraductal car cinoma with a small area of focal invasion and two samples GST-7T was found to contain the lowest total GST activity of were diagnosed as infiltrating lobular carcinoma. The small all primary specimens evaluated (5 mU/mg protein). The GST7r-negative tumor did contain mu class GST on Western blot. number of carcinomas with histology other than infiltrating ductal carcinoma precluded subset analysis by histology. Thirty- None of the 43 samples evaluated contained detectable alpha four tumors contained significant estrogen receptor (>10 fmol/ class GST protein despite strongly positive liver control sam mg protein), while 11 did not. Seventeen tumors contained ples. Two of the samples that had been excluded from this significant progesterone receptor (>20 fmol/mg protein) and analysis because they were found on review of histology and 28 did not. Clinical information concerning tumor size and medical records to represent metastatic tumor deposits in the 6849 First Tumor Set: GST Activity and GST Isozyme Composition. Two sets of cryopreserved breast cancer specimens were ob tained from the hormone receptor laboratories of two different institutions. The first set of 58 samples was randomly selected from the bank of cryopreserved tumor specimens that had been assayed for estrogen and progesterone receptor content at St. Elizabeth's Hospital. Following review of histological material

RESULTS

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

GLUTATHIONE

TRANSFERASES

OF HUMAN BREAST CANCER

age. Likewise, the presence of or absence of mu class GST was unrelated to hormone receptor levels or positivity, lymph node exlract 1 involvement, age, or primary tumor size (Table 2). extract 2 100 extract 3 Second Tumor Set: Relationship of GST-ir Protein to Hormone Receptor Levels and Studies of GST-^i Gene Deletion. As de scribed above, no evidence for correlation between total GST 80 activity (as assessed by CDNB conjugation) and ERP level was I i noted in the first set of breast carcinomas. In view of the recent report by Moscow et al. (22) that GST-7r mRNA levels are 60 o negatively correlated with the level of ERP in human breast (0 carcinomas, we tested the specific hypothesis of a GST-ir to J = 10 ERP correlation. A second set of cryopreserved breast tumor specimens (n = 20) was obtained from the hormone receptor laboratory at Oregon Health Science University. Extracts were 20prepared and analyzed for total GST activity (by CDNB con jugation) and levels of GST-?r protein were quantitated by Western blotting. Levels of GST activity with CDNB were similar to the first set of tumor samples (average, 100.9 mU/ 80 100 20 60 mg protein; SD, 60.5;range, 0-254 mU/mg), 10 of 20 were Mean GST Activity estrogen receptor positive, 6 of 20 were progesterone receptor (mU/mg total protein) positive, and 7 of 20 were mu class GST positive on Western Fig. 1. Region to region variation in GST activity measurements of human blotting. All of the samples contained detectable GST-7r on breast tumors. Three different regions of each tumor were prepared and assayed Western blots. GST-vr protein levels varied widely among the separately. Individual extract activities are plotted versus the mean of the three tumors. Levels of GST-;r protein for a particular tumor were measurements for each tumor. reproducible (experiment to experiment coefficient of variation 30%). The levels of GST-r protein in the tumor extracts were correlated with the level of total GST activity determined for liver and skin contained strong signal for alpha class GST (data not shown), presumably from one or more of the alpha class each tumor (r= 0.88, P< 0.01) (Fig. 4). However, no significant GSTs known to be present in liver and skin (28, 32). Of the 38 relationship between the level of GST-?r protein levels and samples evaluated for mu class GST content, 18 contained the quantitative ERP (r = 0.33, P > 0.5) (Fig. 5) or PRP (r = 0.33, P > 0.5) (data not shown) could be demonstrated. Likewise, mu class protein and 20 did not. We determined whether GST activity correlated with known comparison of the mean GST-Tr levels in ERP-positive and negative samples demonstrated no significant difference (P > prognostic factors in breast cancer. There was no correlation 0.2, Student's / test). Use of the Wilcoxon-Mann-Whitney rank between total GST activity and hormone receptor status re gardless of whether the relationship was examined as a function order test for comparison of two groups revealed a trend toward of ERP (r = 0.12, P > 0.5) (Fig. 3) or PRP (r = 0.21, P > 0.4) higher GST-n- levels in ERP-positive samples (the opposite of that previously reported), but this trend did not reach statistical levels (data not shown) or hormone receptor positivity or neg ativity (Table 1). Nor was there significant correlation of tumor significance (P > 0.13). -Mand GST-i/- are both al-eles f the human GST1 locus, o GST content with primary tumor size, lymph node status, or 11 12 13 ia is

120

10

63

Fig. 2. Western blot of human breast tumor and liver extracts developed with antisera against three glutathione transferase soenzymeclasses. Lanes 1-4, 6-9, and 1-14, replicates of four human breast tissue extracts; lanes 5, 10, and 15, contain human liver extract. The extracts were separated by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by Western blotting as described in "Materials and Methods." Lanes 1-5, developed with GST-^ antisera; lanes 6-10, developed with GST- antisera; lanes 11-15, developed with GST-?r antisera; arrows, positions of the proteins of interest: GST-,,, GST-a, or GST-ir; right ordinale, molec ular weight in thousands.

6850

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

GLUTATHIONE

TRANSFERASES

OF HUMAN BREAST CANCER

r = 0.12 ? 300

human breast carcinomas, DNAs from 7 carcinomas positive for mu class GST on Western blot were all positive on the PCR assay and only one of the 11 DNAs from Western blot-negative tumors produced an appropriately sized DNA fragment (P < 0.00l,x2 test) (Fig. 6). DISCUSSION

-2 200 1 a E 100-

The clinical course of carcinoma of the breast is quite vari able. Patients can develop recurrences either soon after their primary therapy or many years later. The risk of relapse, to some extent, is predictable, most notably by clinical and path 0 100 200 300 ological staging and hormone receptors (24). However, even GST Activity within a group of patients of the same stage and hormonal (mU/mg total protein) receptor status, the course of disease and response to therapy Fig. 3. Estrogen receptor protein versus GST activity for 45 primary human breast cancers. ERP levels and GST activities were determined as described in is quite variable. Numerous investigators have sought to resolve "Materials and Methods." the factors involved through the study of additional markers such as assays that estimate the number of cells in S phase (24) Table 2 /i Class GST protein in human breast tumor subsets' or, more valuecNSNSNSNSNS<0.001 recently, the amount of neu gene expression (39). Characteristic*TotalER+ The glutathione transferases were selected for study because
fmol/mg)ER-PR+ (>10 100fmol/mg)PR-T (>20 cm)T(<2 1 cm)T(2-5 2 cm)<4 3 (>5 nodes>4 nodesAge yrAge 50 s yrGST-M >50 +GST-//gene gene n382810122641191681028710MU+1812651337411441460Mu-2016471314554614110P " Class GST protein was determined as present or absent by Western blot 200 240 160 120 40 80 analysis (see "Materials and Methods"). Total GST Activity * Determined as described in "Materials and Methods" or from clinical records. (mU/mg total protein) ' P values represent likelihood of no association between each characteristic and 11class positivity and negativity as measured by x2 analysis. Fig. 4. GST-jr protein versus total GST activity. Preparation of extracts, gel d NS, not significant at the P < 0.05 level. electrophoresis. Western blotting, development with GST-jr antisera, and meas urement of total GST activity were performed as described in "Materials and Methods." Relative GST-Tr protein levels obtained by densitometric scanning of

r = 0.88

--

60-

(/) O

one of several human genes that code for human mu class GSTs (33, 34). However, GST- GST-i/- are expressed in the or lymphocytes or liver of only 31-66% of humans (35-37). In individuals who lack GST-ji and GST-i/s the deficiency appears to be due to homozygous deletion of the GST1 gene that codes for GST-ii or GST-\(38). The frequency of mu class GST in breast carcinomas was similar to that of GST-^/^ in the total human population, suggesting that the mu class GST, detected on Western blotting of breast carcinoma extracts, was GST-ju/ GST-i/'. This, in turn, suggested that the presence or absence of this enzyme in breast carcinomas might be determined by the presence or absence of the gene for GST1 in the particular individual developing the breast carcinoma. To test this possibility, a PCR-based assay for the presence of the GST- in human DNA samples was used (31). In gene this assay, primers hybridizing to the 5' region of exon 4 and the 3' region of exon 5 amplify a 273-base pair product from target DNA derived from lymphocytes of individuals who carry the GST1 gene. This PCR product consists of the DNA se quence of exon 4, intron 4, and exon 5 of GST1 (31, 33, 38). An appropriately sized product was not amplified from DNA of individuals whose lymphocytes do not contain mu class GST by Western blotting (31). When this assay was applied to study

photographic negatives of Western blots to obtain peak areas (coefficient of variation, 30%). GST-r protein levels are expressed as the peak area for the sample divided by the peak area of the sample producing the most intense signal x 100.

100-11

r = 0.33

80-

_UJ

g 1 60 - E

(O

100

200

300

400

Estrogen Receptor Protein (fmoles/mg total protein) Fig. 5. GST-Tr protein versus estrogen receptor protein. Relative GST-jr pro tein levels were obtained as described in the legend to Fig. 4. ERP levels were determined as described in "Materials and Methods."

6851

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

GLUTATHIONE

TRANSFERASES

OF HUMAN BREAST CANCER

234567

Fig. 6. GST-M PCR products resolved by agarose gel electrophoresis. PCR reactions were performed as described in "Materials and Methods." A 10-^1 aliquot of the reaction mixture was resolved by agarose gel electrophoresis. Lane 1 contains DNA standards ( 123-base pair ladder; Bethesda Research Laboratory. Gaithersburg, MD); lanes 2, 3, 6 and lanes 4 and 5 contain PCR products of target DNA from mu class GST-positive and -negative carcinomas, respectively: lane 7 is the product of the PCR reaction in the absence of target DNA; lane 8 contains a 500-base pair product of PCR of a A bacteriophage DNA with X primers. Arrowheads, the expected position for a 273-base pair DNA product.

of evidence from in vitro studies that they might be associated with the development of drug resistance and might be linked to other tumor markers such as estrogen receptor proteins (22). We have found that primary human breast carcinomas are quite heterogeneous with regard to their content of GST activity. This finding is consistent with prior reports, in which smaller numbers of breast carcinoma samples were studied (2-4, 22). However, we observed no significant correlation between the ACKNOWLEDGMENTS level of GST activity in human breast cancer and factors known We thank D. Carden for secretarial support. to be prognostic in these patients including hormonal receptor status and nodal status. Moscow et al. (22) reported that GST7TmRNA levels in primary breast carcinomas are inversely REFERENCES correlated with estrogen receptor content in the same tumor 1. Mannervik, B., Alin, P.. Guthenberg. O, Jensson, H., Tahir, M. K., Warspecimens. Similarly, Howie et al. (23) reported a negative (but holm. M.. and Jrnvall, H. Identification of three classes of cytosolic glutaquantitatively weaker) correlation between GST-n- protein and thione transferase common to several mammalian species: correlation be tween structural data and enzymatic properties. Proc. Nati. Acad. Sci. USA, estrogen receptor content. Our results do not confirm such a 82: 7202-7206, 1985. correlation and suggest that such a relationship is either sub 2. Shea. T. C., Kelley, S. L., and Henner, W. D. Identification of an anionic form of glutathione transferase present in many human tumors and human stantially weaker than previously suggested or does not exist. tumor cell lines. Cancer Res., 48: 527-533, 1988. There are several human GSTs of the mu class which cross3. Lewis, A. D., Forrester, L. M., Hayes, J. D.. Wareing, C. J., Carmichael, J., Harris, A. L., Mooghen, M. and Wolf, C. R. Glutathione S-transferase react immunologically and are coded for by several mu class isoenzymes in human tumours and tumour-derived cell lines. Br. J. Cancer, genes (33, 34, 38, 41). Therefore, it was not possible to deter 60:327-331, 1989. mine which of the mu class GSTs was present in the breast 4. Di Ilio. C.. Sacchetta, P., Del Boccio, G., La Rovere, G., and Federici, G. carcinoma extracts by Western blotting alone. Only about 31Glutathione peroxidase, glutathione S-transferase and glutathione reduc-ase activities in normal and neoplastic human breast tissue. Cancer Lett., 29: 66% of normal individuals have detectable GST-^ activity in 37-42, 1985. peripheral mononuclear leukocytes (37) or liver (40). The pres 5. Di Ilio, C., Del Boccio. G., Massoud, R., and Federici, G. Glutathione transferase of human breast is closely related to transferase of human placenta ence or absence of this enzyme is inherited as an autosomal and erythrocytes. Biochem. Int., 13: 263-293, 1986. dominant marker and the absence of this enzyme has recently 6. Chasseaud, L. F. The role of glutathione and glutathione-S-transferases in been shown to be due to a homozygous gene deletion (38). The the metabolism of chemical carcinogens and other electrophilic agents. Adv. Cancer Res., 29: 175-274. 1979. similar frequency for mu class GST positivity among breast 7. Dulik, D. M., Fenselau, C., and Hilton, J. Characterization of melphalancarcinomas (Table 1) and for GST-^t positivity of liver or glutathione adducts whose formation is catalyzed by glutathione transferases. Biochem. Pharmacol.. 3S: 3405-3409, 1986. mononuclear leukocytes for a normal population suggests that 8. Smith, M. T., Evans, C. G., Doane-Setzer. P., Castro, V. M., Tahir, M. K., the absence of mu class GST in breast carcinomas (and possibly and Mannervik, B. Denitrosation of 1.3-bis(2-chloroethyl)-l-nitrosourea by other neoplasms) may be simply determined by the absence of class mu glutathione transferases and its role in cellular resistance in rat brain tumor cells. Cancer Res., 49: 2621-2625, 1989. a copy of the GST-/J gene in the germ line of the individual 9. Wang, A. L., and Tew, K. D. Increased glutathione-5-transferase activity in rather than being a characteristic of the carcinoma itself. The a cell line with acquired resistance to nitrogen mustards. Cancer Treat. Rep., results of the studies using the polymerase chain reaction are 69:677-682, 1985. Evans, C. K., Doane-Setzer, M. consistent with this hypothesis. We have found that the target 10. Glutathione G.. Bodell, W., Tokuda, rat brain tumor cellP., and Smith. 1,3and related enzymes in resistance to sequence for PCR, a portion of the GST1 gene locus that codes bis(2-chloroethyl)-l-nitrosourea and nitrogen mustard. Cancer Res., 47: for the GST-M and GST-i/- al-eles,is present in all breast 2525-2530, 1987.
6852

carcinomas positive for mu class protein on Western blotting and absent in all but one of the carcinomas negative for mu class protein. The single disparate sample (PCR positive but -Mprotein negative on Western blot) may simply be due to -^tprotein degradation during tissue handling and storage. Alternatively, this result may represent a tumor that contains the GST1 gene but is deficient in one or more of the steps required in production of a stable GST-^i protein. These results demonstrate three different types of variation in GST content for human malignancies. First, the total GST activity and the quantity of GST-?r protein present in primary human breast carcinomas vary substantially. The mechanisms by which this variation in GST-7r occurs have not yet been determined. Second, breast carcinomas can be classified as either GST--positive or GST- negative, and this phenotype appears to be determined by the heredity of the individual who develops the carcinoma. Third, breast carcinomas differ from some other forms of human malignancy, such as renal cell carcinomas, by their lack of detectable alpha class GST (3, 41). The difference in GST-a isoenzyme content among neoplasms of different histology is presumably a function of the differen tiation of the tissue of origin. However, the molecular mecha nism for such tissue-specific regulation has yet to be defined. The lack of correlation with known prognostic factors does not necessarily rule out a role for GST content as an independ ent predictor of prognosis or response to therapy. Such a hypothesis, that GST content is predictive of outcomes, can only be tested through studies of a larger group of patients and their tumors where long-term follow-up information concerning relapse rates and response to therapy is obtained.

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research

GLUTATHIONE

TRANSFERASES

OF HUMAN BREAST CANCER

11. Lewis, A. D., Hickson, I. D., Robson, C. N., Harris, A. L., Hayes, J. D., Griffiths, S. A., Manson, M. M., Hall, A. E., Moss, J. E. and Wolf, C. R. Amplification and increased expression of alpha class glutathione S-transferase-encoding genes associated with resistance to nitrogen mustards. Proc. Nati. Acad. Sci. USA, 85: 8511-8515, 1988. 12. Batist, G., Tulpule, A., Sinha, B. K., Katki, A., Myers, C. E., and Cowan, K. H. Overexpression of a novel anionic glutathione transferase in multidrugresistant human breast cancer cells. J. Biol. Chem., 261:15544-15549, 1986. 13. Mannervik, B., Castro, V. M., Danielson, U. H., Platz, A., Hansson, J., and Ringborg, U. Expression of class pi glutathione transferase from drugresistant human melanoma cells. Proc. Am. Assoc. Cancer Res., 28: 19, 1987. 14. Mannervik, B., Castro, V. M., Danielson, U. H., Tahir, N. K., Hansson, J., and Ringborg, U. Expression of class pi glutathione transferase in human malignant melanoma cells. Carcinogenesis (London), 8: 1929-1932. 1987. 15. Teicher, B. A., Holden, S. A., Kelley, S. L., Shea, T. C.. Cucchi, C. A., Rosowsky, A., Henner, W. D., and Frei, E. Characterization of a human squamous cell line resistant to cis-diamminedichloroplatinum (II). Cancer Res., 47: 3888-3893, 1987. 16. Wang, Y., Teicher, B. A., Shea, T. C., Holden, S. A., Rosbe, K. W., Al-Achi, A., and Henner, W. D. Cross-resistance and glutathione-5-transferase-ir levels among four human melanoma cell lines selected for alkylating agent resistance. Cancer Res., 49: 6185-6192, 1989. 17. Lewis, A. D., Hayes, J. D., and Wolf, C. R. Glutathione and glutathionedependent enzymes in ovarian adenocarcinoma cell lines derived from a patient before and after the onset of drug resistance; intrinsic differences and cell cycle effects. Carcinogenesis (London), 9: 1283-1287, 1988. 18. Nakagawa, K., Yokota, J., Wada, M., Sasaki, Y., Fujiwara, Y., Sakai, M., Muramatsu, M., Terasaki, T., Tsunokawa, Y., Terada, M., and Saijo, N. Levels of glutathione 5-transferase-7r mRNA in human lung cancer cell lines correlate with the resistance to cisplatin and carboplatin. Jpn. J. Cancer Res., 79: 301-304, 1988. 19. Jensson, H., Eriksson, L. C., and Mannervik, B. Selective expression of glutathione transferase isoenzymes in chemically induced preneoplastic rat hepatocyte nodules. FEBS Lett., 187: 115-120, 1985. 20. Rodate, C., Fukushi, A., Narita. T., Kudo, H., Soma. Y., and Sato, K. Human placenta! form of glutathione-5-transferase (GST-jr) as a new immunohistochemical marker for human colonie carcinoma. Jpn. J. Cancer Res., 77:226229, 1986. 21. Shiratori, U., Soma, Y., Maruyama, H.. Sato, S., Takano, A., and Sato, K. Immunohistochemical detection of the placenta! form of glutathione Stransferase in dysplastic and neoplastic human uterine cervix lesions. Cancer Res., 41: 6806-6809, 1987. 22. Moscow, J. A., Townsend, A. J., Goldsmith, M. E., Whang-Peng, J., Vickers, P. J., Poisson, R., Legault-Poisson, S., Myers, C. E., and Cowan, K. H. Isolation of the human anionic glutathione 5-transferase cDNA and the relation of its gene expression to estrogen-receptor content in primary breast cancer. Proc. Nati. Acad. Sci. USA, 5: 518-6522, 1988. 6 23. Howie, A. F.. Miller, W. R., Hawkins, R. A., Hutchinson, A. R.. and Beckett, G. J. Expression of glutathione S-transferase B,, B2, t x in breast cancers and and their relationship to estrogen receptor status. Br. J. Cancer, 60: 834837, 1989. 24. Harris, J. R., and Henderson, I. C. Natural history' and staging of breast cancer. In: J. R. Harris, S. Hellman, I. C. Henderson, and D. W. Kinne,

25. 26. 27. 28. 29. 30. 31. 32.

33. 34.

35. 36. 37.

38.

39.

40. 41.

(eds.), Breast Diseases, pp. 233-258. Philadelphia: J. B. Lippincott Co., 1987. Moseley, H. S., Peetz, M. E., Keenan. E. J.. Awrich, A. E., and Fletcher, W. S. Endocrine ablation for metastatic breast cancer: a reappraisal of hormone receptors. Am. J. Surg., 140: 164-172, 1980. Habig, W. H., Pabst, M. J., and Jakoby, W. B. Glutathione S-transferase. J. Biol. Chem., 249:1130-7139, 1974. Towbin, H., Staehelin. T., and Gordon, L. Electrophoretic transfer of pro teins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nati. Acad. Sci. USA., 76: 4350-43554, 1979. Mannervik, B. The isoenzymes of glutathione transferase. Adv. Enzymol. Relat. Areas Mol. Biol.. 57: 357-417. 1985. Patel, P. I., Framson, P. E., Caskey, C. T.. and Chinault. A. C. Fine structure of the human hypoxanthine phosphoribosyltransferase gene. Mol. Cell. Biol., 6:393-403, 1986. Treco, D. A. Preparation and Analysis of DNA. In: F. M. Ausubel (ed.), Current Protocols in Molecular Biology, pp. 221-222. New York: Greene Publishing Associates and Wiley-Interscience, 1987. Comstock, K. E., Sanderson. B. J., Claflin, G., and Henner, W. D. GST1 gene deletion determined by polymerase chain reaction. Nucleic Acids Res., 18:3670, 1990. Del Boccio, G., Di Ilio, C., Alin, P., Jrnvall, H., and 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: 2125, 1987. Board, P. G., Suzuki, T., and Shaw, D. C. Human muscle glutathione 5transferase (GST-4) shows close homology to human liver GST-1. Biochim. Biophys. Acta, 953: 214-217, 1988. DeJong, J. L., Chang, C-M., Whang-Peng, J., Knutsen, T., and Tu, C-P. D. The human liver glutathione 5-transferase gene superfamily: expression and chromosome mapping of an Hb subunit cDNA. Nucleic Acids Res., 16:85418554, 1988. Board, P. G. Biochemical genetics of glutathione S-transferase in man. Am. J. Hum. Genet., 33: 36-43, 1981. Laisney, V., Cong, N. V., Gross, M. S., and Frezal, J. Human genes for glutathione S-transferases. Hum. Genet., 68: 221-227, 1984. Seidegard, J., DePierre, J. W., and Pero, R. W. Hereditary interindividual differences in the glutathione transferase activity towards frans-stilbene oxide in resting human mononuclear leukocytes are due to a particular isozyme(s). Carcinogenesis (London), 6: 1211-1216, 1985. Seidegard, J., Vorachek. W. R., Pero, R. W., and Pearson, W. R. Hereditary differences in the expression of the human glutathione transferase active on irans-stilbene oxide are due to a gene deletion. Proc. Nati. Acad. Sci. USA, 5:7293-7297, 1988. Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A., and McGuire, W. L. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/wu oncogene. Science (Washington DC), 235:177-182, 1987. Suzuki, T., Coggan, M., Shaw, D. C., and Board, P. G. Electrophoretic and immunological analysis of human glutathione .S-transferase isozymes. Ann. Hum. Genet., 5/: 95-106, 1987. Di Ilio. C., Aceto, A., Zezza, A., Ricci, G., and Federici, G. Electrophoretic and immunological analysis of glutathione transferase isoenzymes of human kidney carcinoma. Biochem. Pharmacol., 38: 1045-1051, 1989.

6853

Downloaded from cancerres.aacrjournals.org on June 5, 2011 Copyright 1990 American Association for Cancer Research