THE ROLE OF PHARMACOGENETICS IN CANCER THERAPY, PREVENTION, AND RISK

Pharmacogenetics is the study of the role of genetics in individual variation in the therapeutic or toxic response to drugs. Often, such individual variation is the result of genetic polymorphisms in enzymes that are involved in the metabolism of drugs. Many of those enzymes are also involved in the metabolism of endogenous chemicals, and therefore, polymorphisms that represent markers for drug response may also represent risk factors for diseases such as cancer. Our laboratory studies the pharmacogenetics of sulfation, desulfation and glucuronidation reactions. Sulfation reactions are catalyzed by members of a superfamily of cytosolic sulfotransferase (SULT) enzymes, while desulfation reactions are catalyzed by the membrane-bound arylsulfatases (ARSs). Glucuronidation reactions are catalyzed by members of a superfamily of microsomal UDP-glucuronosyltransferases (UGTs). The goals of our research include identifying and characterizing common genetic variations in human SULT, ARS and UGT genes that result in altered metabolism of drugs and/or steroid hormones.

The human phenol SULTs represent a subfamily within the superfamily of SULT enzymes. Phenol SULTs catalyze the sulfate conjugation of a number of phenolic drugs, as well as endogenous compounds, such as phenolic steroid hormones and catechol-amines. We have cloned two human phenol SULT genes, 1A1 and 1A2 (1,2), and have recently identified and characterized two functionally significant common polymorphisms in those genes (3,4). Both of those polymorphisms significantly diminished the capacity of those enzymes to sulfate a prototypic phenolic substrate. In a Caucasian population, the allele frequency for each polymorphism was approximately 33% and the two genes were found to be in linkage disequilibrium. Therefore, individuals that were deficient in SULT1A1 activity had a greater than 90% chance of also being deficient in SULT1A2 activity. The 1A1 polymorphism resulted in an Arg213His amino acid substitution while the variant 1A2 allozyme, defined by a double point mutation, resulted in Ile7Thr and Asn235Thr amino acid substitutions.

The functional significance of the SULT1A1 and 1A2 polymorphisms was determined in different experimental settings. The 1A1 polymorphism was evaluated by correlating 1A1 genotype with the capacity of human liver and platelet cell extracts to sulfate 4-nitrophenol, a prototypic substrate. Figure 1 shows the relationship between the level of SULT1A1 activity in human platelet samples and the thermal stability of that activity stratified by SULT1A1 genotype. The samples that are homozygous for His213 are uniformly associated with low level of enzyme activity and low thermal stability. This association between variant 1A1 genotype and low level of 1A1 activity and thermal stability has been confirmed in 61 human liver samples. 1A2 allozymes were evaluated by determining biochemical characteristics of the recombinant wild type

and variant 1A2 proteins. The variant 1A2 enzyme was associated with an almost 50fold higher Km value (lower affinity) for 4-nitrophenol than was the wild type enzyme (Km, 373 m M and 8.7 mM, respectively). Because SULT1A2 is not highly expressed in the human blood platelet or liver, we have not yet evaluated this effect in a human tissue. The next step in our evaluation of these polymorphisms will include determining their effect on the response of individuals to drugs that are sulfated, as well as their role as risk factors in human disease.

THE PHARMACOGENETICS OF TAMOXIFEN AND RALOXIFENE. RAFTOGIANIS, WALTHER

An important goal of the pharmaco-geneticist is to identify functionally significant genetic polymorphisms in selected genes, and then to apply those findings clinically to identify individuals who are at increased risk for toxic drug effects, or altered drug efficacy, or those who may be particularly amenable to specific drug therapy. Identification of genetic variants that affect response to a variety of therapeutics is occurring at a rapid pace and the clinical application of these findings will dramatically change the way that drugs are prescribed. Tamoxifen (TAM) is a mainstay in the treatment of breast cancer and, based on results of the Breast Cancer Prevention Trials (BCPT), the FDA recently approved the use of this drug as a chemopreventive agent in women at high risk for developing breast cancer. Despite its efficacy, TAM has also been associated with increased incidence of uterine cancer. Raloxifene (RAL) is a newer generation selective estrogen receptor modulator (SERM) that exhibits antiestrogenic activity in the breast similar to that of tamoxifen, but does not appear to be associated with increased incidence of uterine cancer. Therefore, the National Cancer Institute has recently undertaken an additional national study, The Study of Tamoxifen and Raloxifene (STAR), to compare the safety and efficacy of TAM versus RAL in preventing breast cancer in women at high risk. In the next decade, potentially hundreds of thousands of women each year will be treated with TAM or RAL. Although these drugs are effective antiestrogens, there is individual variation in the response of women to TAM and RAL. Pharmacogenetic factors likely contribute to this variation.

The biological inactivation of TAM and RAL are dependent upon two conjugative drug metabolism reactions, sulfation and glucuronidation. A critical metabolic modification affecting the biological activity of TAM (and potentially RAL) is sulfation of 4hydroxytamoxifen (OHT), the active metabolite of TAM. Sulfation of OHT renders this molecule biologically inactive. SULT1A1 has been shown to catalyze the sulfation of OHT, and based on the high degree of amino acid identity between SULTs 1A1 and 1A2 (97%), 1A2 would also be predicted to catalyze the sulfation of OHT. Furthermore, SULT1A1, and likely 1A2, are expressed in human breast tumors.

The major metabolic fate of RAL is glucuronidation, catalyzed by the UGTs. Glucuronidation of RAL is known to be subject to wide interindividual variation that may be the result of common genetic polymorphisms. Recently, a common genetic polymorphism in the UGT1A6 isoform was identified and shown to result in a recombinant protein with only approximately 25% of the capacity of the wild type protein to catalyze glucuronidation reactions. It is not currently known which of the UGT isoform(s) is responsible for the glucuronidation of RAL; based on structural considerations, UGT1A6 would be predicted to catalyze that reaction. Therefore, we hypothesize that poly-morphisms in SULT1A1 and 1A2, as well as in UGT1A6, play a role in individual variation in the response of breast tumors to TAM and RAL. Our initial goals include determining the effects of SULT1A1, SULT1A2 and UGT1A6 polymorphisms on the capacity of recombinant enzymes to sulfate OHT and RAL. Subsequently, we will characterize the ability of isogenic human breast tumor cells carrying different SULT and UGT variants to respond to the antiestrogen effects of OHT and RAL. These experiments will generate preliminary data that is essential for determining the value of continued clinical research studying the role of these polymorphisms in the response of women to TAM and RAL.

Sulfation and desulfation are important reactions in the metabolism of many steroid hormones. Estrone, estradiol and dehydro-epiandrosterone (DHEA) circulate predominantly in the sulfated form and as such are not biologically active (i.e., do not bind target receptors). Furthermore, the sulfated forms of many steroid hormones exhibit half-lives up to ten-fold higher than the desulfated form. Biological "cycling" of sulfated/desulfated steroid hormones has been demonstrated. The sulfated moiety represents a readily accessible, yet biologically inactive, "storage" form for many steroid hormones whereby hydrolysis of the sulfate group (desulfation) regenerates the biologically active steroid. These observations suggest that sulfation and desulfation represent important reactions in the regulation of the biological activity of steroid hormones, and this regulatory system has become a target for chemotherapy of steroid hormone dependent tumors. SULT1A1 catalyzes the sulfation of estrone and 17b-estradiol, while SULT2B1 catalyzed the sulfation of DHEA. ARSC, also known as steroid sulfatase (STS), catalyzes the desulfation of estrone-, 17b-estradiol-, and DHEA sulfate. We hypothesize that functionally significant genetic polymorphisms within these SULTs, as well as ARSC, represent risk factors for development and/or progression of steroid hormone dependent tumors. Furthermore, we hypothesize that ARSC polymorphisms may represent markers for response to sulfatase inhibitors, a new class of antiestrogens.

Breast Cancer. Estrogen metabolism is an important factor in the proliferation of estrogen receptor (ER) positive tumors. Antiestrogens, such as TAM, represent first line therapy against breast cancer. Sulfation and desulfation of estrone and 17b-estradiol clearly occurs in both normal and malignant breast tissues. In the benign breast, estrogens are biosynthesized from steroid precursors via the aromatization of androstenedione to estrone by aromatase (CYP19) in adipose tissue or via the hydrolysis of estrone sulfate by ARSC in epithelial tissue (Figure 2). Aromatase inhibitors such as aminoglutethamide have been used widely as second line therapy in breast cancer, and there is now evidence for a chemopreventive role for these agents. However, in many ER positive breast tumors, ARSC activity is 30- to 150fold higher than aromatase activity and, therefore, the ARSC pathway represents the major route for intra-tumoral estrogen synthesis. Furthermore, estrogen SULT activity is significantly higher in ER positive breast tumors than ER negative tumors. It is currently believed that the balance between ARS and SULT activities is an important factor in the progression of ER positive breast tumors. The realization of the importance of the ARSC pathway for intra-tumoral estrogen production has made the inhibition of this enzyme a target for therapeutic control of breast tumor. Danazol, used in the treatment of endometrial tumors, is the first sulfatase inhibitor to be used clinically and has been shown to significantly block the conversion of estrone sulfate to estradiol in several breast cancer cell lines. Potent third generation steroidal and non-steroidal derived sulfatase inhibitors are now being developed and targeted for breast cancer therapy. Functionally significant polymorphisms within this enzyme may represent markers for response of tumors to ARSC inhibitors.

Steroid metabolism can be affected by a number of variables, including genetic polymorphisms. SULT1A1 is known to catalyze the sulfation of estrone and 17b-estradiol, and is also highly expressed in breast tissue. We hypothesize that the SULT1A1 polymorphism that we have described represents a modifying risk factor for the development and/or progression of breast cancer. To test this hypothesis, we will study the allele frequency of the Arg213His polymorphism in women with breast cancer and in a control group of women without the disease. For these studies, we will take advantage of the extensive and well-characterized Family Risk Assessment Program (FRAP) DNA "bank" maintained by Dr. Godwin at Fox Chase. These DNA samples were isolated from women at high familial risk for breast cancer and from "control" family members not at high risk. Individual DNA samples have been genotyped for BRCA1 and BRCA2, as well as a number of other breast cancer-associated genes. These data will be useful for detailed statistical analysis of the potential risk modifying effects of the SULT1A1 polymorphism.

We are also, identifying common genetic polymorphisms in the human ARSC gene (in collaboration with D. Besack and Dr. Yeung). For this project we are utilizing the CEL I DNA endonuclease mismatch detection assay to "scan" 100 random DNA samples for common ARSC polymorphisms. Subsequently, the functional implications for those polymorphisms will be characterized by generating recombinant wild type and variant ARSC allozymes and by determining the ability of variant ARSC allozymes to catalyze the de-sulfation of estrone sulfate. Furthermore, we are phenotyping 100 individuals for the level of blood leukocyte ARSC activity, and we will subsequently correlate that phenotype with ARSC genotype in those same individuals. Functionally significant ARSC polymorphisms will then be further studied as potential breast cancer risk factors in the same fashion as the SULT1A1 polymorphism. We will also study the response of breast cancer cells that have been stably transfected with ARSC variants, to various sulfatase inhibitors; these studies will test the hypothesis that ARSC polymorphisms may represent markers for tumor response to sulfatase inhibitors.

Prostate Cancer. The importance of androgen activity in the prostate has been clearly established in the pathophysiology of prostate cancer. Surgical or chemical androgen ablation therapy is currently first line therapy for inoperable prostate tumors. DHEA sulfate is quantitatively the most predominant steroid hormone in prostate tissue and is the biological precursor for androgen synthesis in the prostate (Figure 3). DHEA sulfate sulfatase activity has been characterized in both malignant and benign prostate tissue and is localized predominantly to the epithelium. ARSC catalyzes the desulfation of DHEA sulfate, and inhibition of ARSC has been suggested as a potential therapy for prostate cancer. Functionally significant ARSC polymorphisms will be tested as risk factors for prostate cancer, and will also be studied as markers for the response of prostate tumors to sulfatase inhibitors.

The human SULT2B1 gene, as well as two alternatively transcribed 2B1 cDNAs (2B1a and 2B1b), have recently been cloned and these transcripts are highly expressed in the human prostate. Recombinant SULT2B1a and 2B1b enzymes both catalyze the sulfation of DHEA (Km, 9 mM and 12 mM, respectively). Polymorphisms within the SULT2B1 gene may represent risk factors for the development and/or progression of androgen dependent prostate tumors. We are in the process of identifying common genetic polymorphisms within the SULT2B1 gene, and characterizing the functional significance of those poly-morphisms. Functionally significant SULT2B1 polymorphisms will be studied as risk modifying genes for prostate cancer.

PUBLICATIONS

The publications listed below represent work carried out prior to R.B.R. joining Fox Chase.

1.   Raftogianis R.B., Her C., Weinshilboum R.M. Human phenol sulfotransferase pharmacogenetics: STP1 gene cloning and structural characterization. Pharmacogenetics 6:473-87, 1996.

2.   Her C., Raftogianis R., Weinshilboum R.M. Human phenol sulfotransferase STP2 gene: molecular cloning, structural characterization, and chromosomal localization. Genomics May 33:409-20, 1996

3.   RAFTOGIANIS, R.B., WOOD, T.C., OTTERNESS, D.M., VAN LOON, J.A., WEINSHILBOUM, R.M. Phenol sulfotransferase pharmacogenetics in humans: association of common SULT1A1 alleles with TS PST phenotype. Biochem. Biophys. Res. Commun. 239:298-304, 1997.

4.   RAFTOGIANIS, R.B., WOOD, T.C., WEINSHILBOUM, R.M. Human phenol sulfotransferases SULT1A2 and SULT1A1: genetic polymorphisms, allozyme properties and human liver genotype-phenotype correlations. Biochem. Pharmacol. (in press).

PREUSS C.V., WOOD T.C., SZUMLANSKI C.L., RAFTOGIANIS R.B., OTTERNESS D.M., GURARD B., SCOTT M.C., WEINSHILBOUM R.M. Human histamine N-methyltransferase pharmacogenetics: common genetic polymorphisms that alter activity. Mol. Pharmacol. 53:708-17, 1998.

Weinshilboum R.M., Otterness D.M., Aksoy I.A., Wood T.C., Her C., Raftogianis R.B. Sulfation and sulfotransferases 1: Sulfotransferase molecular biology: cDNAs and genes. FASEB J. 11:3-14, 1997.

WEINSHILBOUM, R.M., RAFTOGIANIS, R.B. Sulfotransferases and methyltransferases. In Metabolic Drug Interactions, edited by R.H. Levy. Lippincott-Raven, Philadelphia (in press).

^§   Fox Chase researcher

^a   T.C. Wood, R.M. Weinshilboum: Mayo Clinic, Rochester, MN 55901

Illustrations or unpublished data in these reports should not be used without permission of the author.

Fox Chase Cancer Center

Scientific Report 1998