(CANCER RESEARCH 49. 6551-6555. December I, 1989] Lack of Evidence for Aromatase in Human Prostatic Tissues: Effects of 4-Hydroxyandrostenedione and Other Inhibitors on Metabolism1

A. M. H. Brodie,2 C. Son, D. A. King, K. M. Meyer, and S. E. Inkster

Department of Pharmacology and Experimental Therapeutics, School of Medicine, University of Maryland, Baltimore, Maryland 21201

ABSTRACT enzyme in vitro (7). Our studies in rats treated with 4-OHA demonstrated reduction in ovarian vein concentrations The effects of 4-hydroxyandrostenedione (4-OHA) and other aroma- tase inhibitors, 10-propargylestr-4-ene-3,17-dione and imidazo|l,5-a|- and inhibition of ovarian aromatase, concomitant with regres sion of carcinogen (dimethylbenzanthracene)-induced mam 3,4,5,6-tetrahydropyrin-6-yl-(4-benzonitrile), as well as 5a-reductase in hibitors /V./V-diethyW-methyl-S-oxo^-aza-Sa--n/J-carboxy- mary tumors (8). We have also shown that 4-OHA treatment amide and 4-methyl-3-oxo-4-aza-androsta-5-ene-17-ol were investigated inhibits peripheral aromatization in male Rhesus monkeys (9). in prostatic tissue from six patients with benign prostatic hypertrophy This compound is now being evaluated for the treatment of and seven patients with prostatic cancer, and from normal men at autopsy. breast cancer. Plasma estradiol and levels were found We attempted to measure aromatase activity in the tissue incubations by to be significantly reduced in postmenopausal patients with quantitating 'IM) released from or labeled advanced disease treated with 4-OHA. Partial or complete at the C-l position. High performance liquid Chromatograph} and thin tumor regression occurred in 30% of 4-OHA-treated patients layer chromatography were used to isolate products. Although (10, 11). Recently, a small group of men with prostatic cancer the amount of 3H2Oreleased was at least twice that of the heat-inactivated have been treated with 4-OHA. A subjective response to treat tissue samples, no estrone or estradiol was detected on high performance liquid chromatography. The 'I !..() release was significantly inhibited by ment was observed in a high proportion of these patients (12). 4-OHA and jV,/V-diethyl-4-metnyl-3-oxo-4-aza-5a-androstane-17/3-car- It is therefore of interest to determine whether aromatase is boxyamide, but not by the other aromatase inhibitors. 4-OHA also present in prostatic tissue and/or whether 4-OHA has an effect inhibited 5or-reductase in both benign prostatic hypertrophy and cancer on androgen metabolism in primary carcinoma tissue, as well tissue, although to a lesser extent than A'tA'-diethyl-4-methyl-3-oxo-4- as in BPH, which might be associated with the response. aza-5a-androstane-17/3-carboxyamide. The other aromatase inhibitors In this study, we attempted to use the 3H2O assay as an were without effect on 5a-reductase. Our results indicate that 'IM) indirect assessment of production. This method de released from [l/3-'H]androstenedione and [l,2,6,7-3H]androstenedione pends upon the loss of hydrogen atoms at the C-l/3 and C-2ß does not correlate with estrogen formation and may be the result of other positions of the androgen substrate during aromatization (13). metabolic reactions. Although it appears that the prostate lacks aroma Thus, when substrates are labeled with tritium at these posi tase, 4-OHA may be of benefit in patients with benign prostatic hyper tions, 3H2O is formed equivalent to each mole of estrogen trophy or prostatic cancer by inhibiting this enzyme in peripheral tissue. produced. The method has been widely used by ourselves and others to measure aromatization in the human placenta and rat ovary (8). We have also employed the product isolation proce INTRODUCTION dure to quantitate directly, as well as to identify other products of androgen metabolism. We found that the two DHT3 is of primary importance in the growth of normal and methods do not provide equivalent results of estrogen produc pathological prostatic tissue (1,2). Although estrogen has been tion in the prostate. shown to be involved in controlling androgen biosynthesis in the testis, the importance of this hormone in the prostate, to either regulate steroid biosynthesis or to affect growth, is un MATERIALS AND METHODS clear. However, estrogen receptors have been identified (3, 4) Chemicals. [l,2,6,7-3H]Androstenedione (92.7 Ci/mmol), [l/3-3H]an- and greater than normal levels of estradiol have been detected drostenedione (27.8 Ci/mmol), [7-3H]testosterone, -androstenedione, in stroma of BPH (5), which suggests that estrogens may have -estradiol, and -estrone (24.5 Ci/mmol), [4-uC]estrone, -estradiol, a role in this disease. -testosterone, - (57 Ci/mmol), and -androstenedi During the past few years, we have been developing selective one (59 Ci/mmol) were obtained from NEN Research Products (Du inhibitors of aromatase (estrogen synthetase) (6). If estrogen pont) (Boston, MA). The distribution of tritium labeling on [1,2,6,7- does mediate growth of prostatic tissue, these inhibitors might 3H]androstenedione was: la, 19.9%; 1/3,6.7%; 2a + 6a, 32.4%; Iß+ be of value in treating diseases of the prostate, such as BPH 6/3, 13.1%; la, 21.2%; and 7/3, 6.7%; and on [l/J-3H]androstenedione and cancer. We have determined that 4-OHA is a potent was: 1«,24%; and 1/3, 76%. The scintillation fluid used was 3a70B competitive inhibitor which also causes inactivation of the from Research Products Inc. (Mt. Prospect, IL). Organic solvents were HPLC grade from J. T. Baker. Ether (USP) was purchased from Received 3/27/89; revised 8/7/89; accepted 8/14/89. Mallinkrodt. Unlabeled , glucose-6-phosphate. glucose-6-phos- The costs of publication of this article were defrayed in part by the payment phate dehydrogenase, and NADPH were purchased from Sigma Chem of page charges. This article must therefore be hereby marked advertisement in ical Co. (St. Louis, MO). accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1Supported by N1H grant CA-27440. Presented in part at the 8th International Inhibitors. Preparation of 4-OHA was carried out as previously Congress of Endocrinology. Kyoto. Japan, 1988, and the American Association described (8). Another steroidal aromatase inhibitor, PED (14), was for Cancer Research, New Orleans. Louisiana, 1988. 2To whom requests for reprints should be addressed, at Department of kindly provided by Dr. J. O. Johnston (Merrell Research Labs., Cincin Pharmacology and Experimental Therapeutics. School of Medicine, University nati, OH). The nonsteroidal aromatase inhibitor CGS 16949A (15) was of Maryland, 655 W. Baltimore St.. Baltimore. MD 21201. donated by Ciba-Geigy Research Laboratory (Summit, NJ). The 5a- 3The abbreviations used are: DHT, dihydrolestosterone; 4-OHA, 4-hydroxy- reductase inhibitors 4-MA (16) and L659110 were gifts from Dr. G. androstene-3,17-dione; BPH. benign prostatic hypertrophy; HPLC, high perform Rasmusson (Merck, Sharp and Dohme Research Laboratories, Rah- ance liquid chromatography; PED. 10-propargylestr-4-ene-3,17-one;4-MA, jV,/V- diethyl-4-methyl-3-oxo-4-aza-5«-androstane-17fi-carboxyamide; TLC, thin layer way, NJ). was also used to inhibit 5a-reductase in some chromatography: CGS 16949A, imidazo[1.5-«]3.4.5.6-tetrahydropyrm-6-yi-(4- experiments. benzonitrile); L659110. 4-methyl-3-oxo-4-aza-androsta-5-ene-17-ol. Tissues. BPH and primary prostatic carcinoma tissues were obtained 6551

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research. AROMATASE AND 5jCi[1.2.6,7-3H]androstenedione.[4-MC|steroids( (were solution (0.5 mg NADPH, 2 mg glucose-6-phosphate, and 2.5 ID added to the incubates to serve as markers for estradiol (E2), estrone (A-i), glucose-6-phosphate dchydrogcnase per ml) and was terminated after testosterone (7"). androstenedione (A), and DHT. The major peaks of tritium 2 h at 37°Cbythe addition of 4 ml organic solvent. Unlabeled steroids. activity ( ) were identified as 5iï-androstane-3a,17^-diol (28 ml). 5«-andros- 2-4 ng each, and 4-l4C-labclcd steroids (3000 dpm), androstenedione, tanedionc (.11 ml), and 5

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research. AROMATASE AND 5n-REDUCTASE INHIBITOR STUDIES IN THE PROSTATE profile of the steroid metabolites from incubation with 13.5 /¿Ci tive in hyperplastic tissue. In BPH tissue, 4-OHA inhibited 5tv- [l,2,6,7-'H]androstenedione is shown in Fig. 1. Contrary to the reductase in a concentration-dependent manner. Tissue was incubated with 5 /¿M4-OHA and I0~8 to 10~5 M testosterone. implications of the above results, no tritiated estrone or estra dici could be detected by HPLC. To determine whether testos The apparent K¡of4-OHA on the 5o-reductase was determined terone was a preferential substrate for aromatization, we incu to be 0.69 n\t, while the apparent Kmwas 0.19 /¿M.5«-Reductase bated 12.1 pCi [7-3H]testosterone. The limit of detection was was also inhibited by 4-MA, L651190 (Table 4), as well as 0.001%. This experiment was carried out in the presence (or progesterone (Fig. 2). However, other aromatase inhibitors, absence, for controls) of 5 n\i progesterone in order to inhibit CGS-16949 or PED, were without effect (Table 4). Siv-reduction. However, even under these conditions to maxi The reason for the release of tritium from labeled androstene- mize substrate availability for the aromatase pathway, no estro dione in absence of estrogen formation was investigated by gen could be detected (Fig. 2). Tritium (dpm) associated with comparing [1,2,6,7-'H]- and [l/i-'H]androstenedione in incu the [4-'4C]estrogens was not significantly different from that of bations of normal prostatic or testicular tissue. The testis me the inactivated samples, in most cases. In a few samples, meas tabolized androstenedione to testosterone with little 5«-reduc- urable tritium was detected in the estrogen fractions but was tion (0.3% converted to DHT). Small amounts of estradiol were removed by further chromatography on TLC. also detected on HPLC, as shown in Table 5. When the values Androstenedione was mainly converted by BPH tissue to 5«- for 3H2O production from [1/Õ-3H]-and[l^oJ-'Hjandrostene- androstanedione and with little formation of dione were corrected for the percentage of tritium at C-lff and DHT (7%) (Table 2). The extent of conversion of androstene- C-2/3, there is reasonable agreement with values for estrone and dione to 5«-androstanedione was greater in BPH tissue than in estradiol measured by product isolation using HPLC. In con cancer tissue, which produced more androsterone. The conver trast to the testis, the tritium content of the aqueous phase was sion of testosterone to DHT was similar to that of androstene- increased 3-fold when [l,2,6,7-'H]androstenedione (2 x IO6 dione to 5«-androstanedione in cancer tissue (Table 3). For dpm) was incubated with prostatic tissue, compared to incuba mation of 5«-reducedmetabolites in both tissues was inhibited tions with [l/i-3H]androstenedione (Table 5). When values were by 4-OHA, although the compound appeared to be more effec- adjusted for the assumed loss of tritium from C-\ßand C-2ß during aromatization, there was an even greater discrepancy between the results, as the mean value for incubations with 3.0-oX2 DHTnE, TA [l,2,6,7-'H]androstenedione was increased approximately 20- fold. Confirmation that the tritium was due to 'H:O was nir'' So.Q-1000 determined by lyophilization. Despite the production of 'H2O from [l,2,6,7-3H]androstenedione by prostatic tissue, neither }1 „-800 2.0-CLQih;§ [3H]estrone nor ['Hjestradiol could be detected when steroids 'i¡' '11 Tf Õ2-600 lli in the chloroform extract were separated by HPLC and subse 1 , If'ií'! ?i, Q§-400 quent TLC (Table 5). 1.0-¡X 'íV\ The amount of 'H2O released from control tissue incubated ^î £-200 1iii .\1 Table 4 Inhibition of5a-reduction in human hypertrophiedproslatic tissue *ti1 '\VX 7jn OE2 ' w-1400-1200 Tissue was incubated with [1.2.6.7-3H]androstenedione and cofactors for 2 h at 37"C in the absence or presence of 5 /JMcompound: 5«-reductaseinhibitors 4 10 15 20 25 MA and L651190. aromatase inhibitors 4-OHA. PED, and CGS 16949A. or FRACTION TIME (minutes) DHA (5-androsten-3(i-ol-17-one). Values are expressed as percentage of inhibi tion of control. Fig. 2. HPLC elution profile of steroid metabolism by BPH tissue incubated with 12.1 /jCi |7-'H]testosterone in the presence of progesterone (5 J/M). , Inhibition (%) [4-'4C]Steroid markers; , tritium from the substrate. 4-MA 98.4 4-OHA 74.5 Table 2 Conversion of /1,2,6.7-*H/androslenedione to 5a-androstanedione and L651190 35.4 PED 0 DHT in tissue from normal and hyperplastic human prostates CGS 0 Prostatic tissue (30 mg homogenate) from a patient with BPH or from a DHA (I normal male at autopsy was incubated with [1.2,6,7-3H]androstanedione and cofactors for 2 h at 37"C. DHT and 5«-androstanedione (nmol/g tissue/h) were separated by HLPC followed by TLC. Table 5 Production of'H2Ofrom ///<-'///- or [1,2,6,7-3HJandrostenedione by Normal BPH human prostatic and testicular tissue Homogenates (30 mg) were incubated with cofactors for 2 h at 37'C. The 5a-Androstanedione (nmol/g tissue/h) 0.23 0.759 value for the heat-inactivated sample from each tissue was subtracted from all DHT (nmol/g tissus/h) 0.067 0.116 results. Estradiol (E2) and estrone (E,) were purified to constant 3H/"C ratio by HPLC and TLC. followed by acetylation and TLC of the acetates. Values were corrected for 3H distribution assuming loss of the \ßand 2ßtritium from the Table 3 Effect of 4-OHA on the metabolism of testosterone or androstenedione to substrate during aromatization and are expressed as percentage of conversion. in-reduced steroids by human prostatic tissue Triplicate samples of tissue (30 mg homogenate) from patients with BPH or Conversion (%) prostatic cancer were incubated for 2 h at 37'C with cofactors and |1-3H]- (M-'H]Androstenedi- androstenedione in the absence or presence of 5 ¡IM4-OHA. Malignant tissue [l,2,6,7-JH|Androstenedione samples were also incubated with [l-3HJtestosterone. Products (nmol/g tissue/h) one 5tt-androstanedione from androstenedione and DHT from testosterone were E,ProstateControl4OHATestisControl4-OHA0.590.070.610.120.453H2O Lyophilization0 3H2O Lyophilization0 E2 isolated by HPLC. tissue/h)[3H]AndroslencdioneControlProduct formation (nmol/g 11.630.991.170.7911.56 000.46

]Testosterone, Cancer0.53 0 0.071 ±0.010 ±0.052 4-OHABPH1.44±0.09 ±0.053Cancer0.850.21 ±0.003[3H 0.15 ±0.001 ' 3H2O collected by lyophilization.

6553

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research. AROMATASE AND 5<,-REDUCTASE INHIBITOR STUDIES IN THE PROSTATE with NADPH was more than 4 times the value for tissue that (HPLC) for measuring low level estrogen production. Table 5 was inactivated and/or incubated without cofactors (Table 6). shows that the product isolation method was sensitive enough This finding indicates that an NADPH-dependent enzymatic to detect low levels of estrogens. The amounts of 'H;O gener process is involved in the 'H^O release. The further possibility ated by both the testis and prostate were quite similar and the investigated was that production of estradiol or estrone may equivalent amount of estrogen in the testis was easily detected not be detected by HPLC if these steroids were rapidly con by HPLC. We therefore conclude from our results with HPLC, verted by the prostate to other metabolites. However, when which effectively separates DHT from the estrogens, that nei cancer tissue was incubated with [4-'4C]estradiol and -estrone, ther testosterone nor androstenedione is metabolized to estro more than 85% of the radioactivity was accounted for as estra gens in significant amounts by normal, hypertrophied, or ma diol or estrone, suggesting that neither estrogen was metabo lignant prostatic tissues. Nevertheless, it appears that andro lized in significant amounts (results not shown). stenedione is metabolized in the prostate with release of tritium to form 3H2O. We suggest that the greater release of'H2O from incubation of [1,2,6,7-'H]- than of [l/i-'Hjandrostenedione DISCUSSION could be due to the replacement of the 'H at C-6 and C-7 by The 'H2O assay has been routinely used by us to measure hydroxyl groups during metabolism of DHT to the 6«-and la- triols (19). Thus, our values appear to increase dramatically aromatase activity in human placenta! and rat ovarian incuba when calculated for the assumed loss of tritium from C-l/i and tions (7). We have confirmed that the assay provides results C-2/3 (Table 5). Although the cause of the loss of tritium from equivalent to those of the product isolation method for meas C-l/3 to produce 'H2O is unknown, it may also be due to a uring aromatization of to estrogens in term placenta! metabolic process, possibly hydroxylation, involving the 5a- microsomes (17). There are several reports in which tritium released from [l/j-'Hjandrostenedione and [l,2,6,7-'H]andro- reductase pathway. Inhibition of the release of tritium by 4- OHA as well as 4-MA but not by other aromatase inhibitors is stenedione was used in attempts to measure estrogen produc consistent with the idea that metabolism via the 5«-rcduction tion during incubation with prostatic tissue. Our findings are pathway, rather than aromatization, is involved in production quite similar to those reported by Stone et al. ( \ 8) but we differ of 'H2O. However, further studies are required to determine in our interpretation of the results. the validity of this hypothesis. The present study demonstrates that, during incubation with 4-OHA was found to inhibit 5

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research. AROMATASE AND 5<«-REDUCTASEINHIBITOR STUDIES IN THE PROSTATE

REFERENCES 12. Shearer, R. J.. Cunningham. D., Malone, P. R., Nicholls. S., Coombes. R. C., and Dowsett. M. Aromatase inhibition in prostatic cancer. A phase I 1. Bruchovsky, N., and Wilson, J. D. The conversion of testosterone to 5«- study. Br. Med. J.. in press, 1989. androstan-l7fi-ol-3-one by rat prostate in vivo and in vitro. J. Biol. C'hem., 13. Brodie, H. J., Kripaluni. K. J., and Possanza, G. Studies on the mechanism 243: 2012-2021. 1968. of estrogen biosynthesis. VI. The stereochemistry of hydrogen elimination of 2. Muggins. C., Stevens, R. E.. and Hodges, C. V. Studies on prostatic cancer. C-2 during aromatization. J. Am. Chem. Soc.. 91: 1241-1243. 1969. II. Effects of castration on advanced carcinoma of the prostate gland. Arch. 14. Melcalf. B. W.. Wright. C. L.. Burkhart. J. P., and Johnston. J. O. Substrate- Surg.. 43: 209-223, 1941. based inactivation of aromatase by allenic and acetylenic steroids. J. Am. 3. Krieg, M.. Kloetzl. G.. Kaufmann. J., and Voigt. K. D. Stroma of benign Chem. Soc., 103: 3221-3222. 198Õ. prostatic hypcrplasia: preferential tissue for androgen and oestrogen binding. 15. Steele, R. E., Mellor. L. B.. Sawyer. W. K.. Wasvary, J. M., and Browne, L. Acta Endocri. (Copenh.). 96:422-432. 1981. J. In vitro and in vivo studies demonstrating potent and selective estrogen 4. Bashirelahi. N.. Young. J. D.. Sidh. S. M., and Sanefuji. H. Androgen. inhibition with the nonsteroidal aromatase inhibitor CGS 16949A. Steroids. oestrogen and and their distribution in epithelial and stromal 50: 147-161. 1987. cells of human prostate. In: F. H. Schroder and H. J. dc Voogt (eds.). Steroid 16. Brookes. J. R.. Baptista. E. M.. Berman. C.. Ham. E. A.. Hichens, M., Receptors. Metabolism and Prostatic Cancer, pp. 240-255. Amsterdam: Johnston, D. B. R., Primka, R. L.. Rasmussen, G. H.. Reynolds, G. F., Excerpta Medica, 1980. Schmitt, S. M., and Arth, G. E. Response of rat ventral prostate to a new 5. Kozak. I.. Bartsch. W.. Krieg. M., and Voigt. K. D. Nuclei of stroma: site of and novel Sn-reductase inhibitor. Endocrinology, 109: 830-836, 1981. the highest estrogen concentration in human benign prostatic hyperplasia. 17. Inkster. S. E.. and Brodie, A. M. H. Immunocytochemical studies of aro Prostate. 3:433-438. 1982. matase in early and full term human placenta! tissue: comparison with 6. Brodie. A. M. H. Commentary: aromatase inhibition and its pharmacologie biochemical assays. Biol. Reprod.. in press. 1989. implications. Biochem. Pharmacol.. 34: 3213-3219. 1985. 18. Stone, N. N., Fair. W. R., and Fishman. J. Estrogen formation in human 7. Brodie. A. M. H., Garret!. W. M. Hendrickson, J. R.. Tsai-Morris, C-H., prostatic tissue from patients with and without benign prostatic hyperplasia. Marcotte, P. A., and Robinson. C. H. Inattivatici) of aromalase in vitro by Prostate, 9: 311-318, 1986. 4-hydroxyandrostenedionc and 4-acetoxyandrostenedione and their sus 19. Issacs, J. T.. Brendler, C. B.. and Walsh. P. C. Changes in the metabolism tained effects in vivo. Steroids, 38:693-701. 1981. of dihydrotestosterone in the hyperplastic human prostate. J. Clin. Endocri- 8. Brodie. A. M. H.. Schwarze!. W. C.. Shaikh. A. A., and Brodie. H. J. The nol. Metab., 56: 139-146. 1983. effects of an aromatase inhibitor 4-hydroxyandrostenedione on estrogen 20. Zoppi. S., Cocconi, M., Natali. A., Serio, M., Martini, L., and Motta. M. In dependent processes in reproduction and breast cancer. Endocrinology. 100: vitro effects of an aromatase inhibitor on So-reductasc activity in human 1684-1695, 1977. hypertrophie prostatic tissue. J. Clin. Endocrino!. Metab.. 63:269-271. 1986. 9. Brodie, A. M. H., and Longcopc, C. Inhibition of peripheral aromatization 21. Houston. B., and Habib. F. K. A kinetic analysis of the inhibition of human by aromatase inhibitors. 4-acctoxy- and 4-hydroxyandrostenedione. Endocri prostatic 5a-reductase by 4-hydroxyandrostenedione and related steroids. nology. 106: 19-21, 1980. Steroids, 52: 237-247, 1988. 10. Coombes. R. C., Goss, P.. Dowsett, M., Gazet, J. C., and Brodie, A. M. H. 22. Bartsch, W., Klein, H., Sturenburg. H., and Voigt, K. Metabolism of andro- 4-Hydroxyandrostenedione treatment of postmenopausal patients with ad gens in human benign prostatic hyperplasia: aromatase and its inhibition. J. vanced breast cancer. Lancet. 2: 1237-1239. 1984. Steroid Biochem.. 27: 557-564. 1987. 11. Goss, P. E.. Coombes. R. C.. Powles, T. J., Dowsett. M.. and Brodie. A. M. 23. Cunningham. D.. Powles. T. J.. Dowsett. M., Hutchinson. G.. Ford. H. T.. Brodie. A. M. H.. Gazet, J. C.. and Coombes. R. C'. Oral 4-hydroxyandros H. Treatment of advanced postmenopausal breast cancer with aromatase inhibitor 4-hydroxyandrostenedione: Phase 2 report. Cancer Res.. 46:4823- tenedione is a new active endocrine treatment in breast cancer. Cancer Res.. 4826. 1986. 47: 1957-1961, 1987.

6555

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research. Lack of Evidence for Aromatase in Human Prostatic Tissues: Effects of 4-Hydroxyandrostenedione and Other Inhibitors on Androgen Metabolism

A. M. H. Brodie, C. Son, D. A. King, et al.

Cancer Res 1989;49:6551-6555.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/49/23/6551

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/49/23/6551. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1989 American Association for Cancer Research.