3315S.Full.Pdf

[CANCER RESEARCH (SUPPL.) 42, 3315S-3321 S. August 1982] 0008-5472/82/0042-0000$02.00 Basic Studies on Aminoglutethimide1

Hilton A. Salhanick

Department of Population Sciences, Harvard School of Public Health, and Department of Obstetrics and Gynecology. Harvard Medical School, Boston. Massachusetts 02115

Abstract gical attempts to remove the naturally occurring steroidal hor mones on the other. The latter approach includes the admin Aminoglutethimide (AG), a bicyclic substance with two opti istration of tamoxifen, a target-receptor competitor of estro cally active isomers, inhibits the cholesterol side-chain cleav gens. Current thought is that AG acts by inhibiting steroido- age (SCO and aromatase systems by blocking the terminal genesis at the aromatase sites. cytochrome P-450s. AG interacts with these cytochromes to The purpose of this contribution is to review the actions of induce typical type II, low-spin spectra; the amount of spectral AG and attempt to relate them to therapeutic rationale. change correlates directly with the amount of enzymatic inhi bition. AG acts by preventing reduction of the cytochrome, an obligate step preceding oxygÃ©nationof substrate. The apparent Chemical Considerations K, of D(+ )-AG is about 2.5 times less than the K, of L(-)-AG in AG has 2 rings and an ethyl appendage (Chart 1). It resem inhibiting SCC and about 40 times less in inhibiting aromatase. bles glutethimide (Doriden), which lacks the amine. Since glu- The spectral Ks ratios of the D- and L-enantiomers for P-450scc tethimide has little or no antisteroidogenic activity, it follows and P-450a,omataseare2.5 and 5, respectively, in the presence that the amine is necessary for this biological activity. The of substrate. The slope of the substrate concentration versus sedative and other central nervous system-depressive activities inhibition curves is less than 0.1 and extends over 2 logarithmic of AG are most probably related to the piperidinedione ring, as units of substrate concentration. indicated by the common occurrence of the monoureide con By in vivo rabbit and rat assays for SCC, D-AG is 5 and 25 figuration in other sedatives in the glutethimide, barbiturate, times more potent than L-AG. In women, AG rapidly lowers and hydantoin series. These include thalidomide and cyclohex- estrogen and progesterone levels in the luteal phase of the imide, which do not inhibit SCC (4).3 The former is teratogenic menstrual cycle and in early pregnancy, eliciting a reflex and the latter blocks protein synthesis. They raise the suspicion release of luteinizing hormone in the luteal phase. In postmen- that AG may have a direct effect on cell reproduction or opausal, adrenalectomized, ovariectomized women, AG pre function. Studies with AG analogs indicate that few substances, vents the aromatization of administered androstenedione. including those with even minor modifications, are as potent as In inhibition studies in women, plasma steroid concentrations AG in the inhibition of SCC or aromatase. were inversely proportional to AG levels, which were about 20 Because it is likely that the amine interacts with the proto- to 30 /UM.This level of AG, which approximates the apparent porphyrin iron of the cytochrome P-450s involved, we have KÂ¡sof the SCC and aromatase systems, causes a decrease in postulated that the aromatic ring "stacks" on the aromatic-like hormone levels of about 50%. porphyrin structure and that the remainder of the molecule assigns some specificity by its affinity to the apoprotein (9). If Introduction this is true, then the separation of antisteroidogenic properties AG2 has been a drug of promise for more than 2 decades. from central nervous system properties may not be achieved easily. Initially introduced as a promising anticonvulsant agent, it was Of more current significance is the fact that the therapeutic withdrawn from use in 1966 because of its antisteroidogenic drug currently available is a mixture of the D(+ )- and L(â€”)- actions. Casually labeled a substance wh'ch induced "medical adrenalectomy," AG was demonstrated to inhibit steroidogen- enantiomers. These isomers have markedly different biological properties and, in the opinion of this author, merit serious esis at the cholesterol SCC site (1 ). On the basis of the medical consideration for further individual study. adrenalectomy concept, it was tested as a possible therapeutic agent for metastatic breast cancer. This too has been modified Cytochrome P-450-linked Enzyme Systems conceptually; its role in the therapy of breast cancer is now attributed to its action as an aromatase inhibitor. AG inhibits 2 enzyme systems: cholesterol SCC and aroma The understanding of medicinal therapy of breast cancer is complicated by other paradoxes. "Effective" treatments have tase. The SCC system has been thoroughly studied in many species and under many experimental conditions. The side included treatment with large doses of estrogens, androgens, chain of cholesterol is removed by a series of 3 oxidation steps or progestational agents, on one hand, and medical and sur- requiring 3 mol of oxygen and 3 mol of NADPH. The 3 steroidal 1 Presented at the Conference "Aromatase: New Perspectives for Breast steps involve hydroxylations at positions 22 and 20 and sub Cancer." December 6 to 9, 1981. Key Biscayne, Fla. Work from this laboratory sequent cleavage between these carbon atoms with loss of was supported by USPHS Grant 10081 from National Institute of Arthritis, capryl aldehyde. The reaction occurs on the terminal enzyme Metabolism and Digestive Diseases. NIH, Research Contract NIH-70-2319 from the National Institute of Child Health and Human Development, and Grant PDT- of an electron transport chain, cytochrome P-450scc- 147 from the American Cancer Society. The enzyme has been purified, and its amino acid constitu- 2 The abbreviations are: AG, aminoglutethimide [3-<4-aminophenyl)-3-ethyl- 2.6-piperidinedione]; SCC, side-chain cleavage: LH, luteinizing hormone. 3 P. E. Graves, V. I. Uzgiris, and H. A. Salhanick. unpublished observations.

AUGUST 1982 3315s

P-450(RED.)-CHOL.

ICYTOCHROME P-450 â€¢¿if

PREQNENOLONE P-450(OX.) A

Chart 2. Pathway for supplying reducing equivalents to support the conver sion of cholesterol (CHOL.ÃŒto pregnenolone. Electrons (e ) from NADPH are D(+)AMINOGLUTETHIMIDE L(-)AMINOQLUTETHIMIDE transported through flavoprotein reducÃase (FP) to iron sulfur protein (ISP) to cytochrome P-450scc- Cholesterol is bound to cytochrome P-450scc in the oxidized (OX.) state. The complex is reduced (RED), and the substrate is oxidized to pregnenolone in 3 oxidative steps. The other product, capryl aldehyde, is not shown.

elsewhere in this symposium (2), aromatase converts the A ring of a 19-carbon steroid such as testosterone or androstenedione to a phenolic estrogen such as estradiol or estrone. This system is microsomal and does not require a ferridoxin- like protein; electrons are transferred directly by a flavoprotein reducÃase from NADPH to the terminal enzyme, presumably cytochrome P-450a,omataM- QLUTETHIMIDE PHENOBARBITAL Unfortunately, cytochrome P-450aromatasehasnot been puri fied. In fact, the evidence that aromatase is a cytochrome P- 450 is still somewhat circumstantial. Placenta! cytochrome P- 450aromaiase,themost commonly studied enzyme, does not have the affinity for carbon monoxide common to other cytochrome P-450S. Also, placental preparations appear to contain other microsomal P-450s as well as contamination by mitochondrial cytochrome P-450Scc- Therefore, interpretations based upon spectral and electron spin resonance data are not conclusive. Furthermore, since AG can inhibit enzymes other than cyto chrome P-450, this property cannot be utilized as proof of cytochrome P-450 involvement. Nevertheless, it is probable CYCLOHEXIMIDE that the enzyme linked to aromatizaron is a cytochrome P-450, Chart 1. Structure of AG and related substances. and, having expressed the caveat, we will make the assumption that aromatization is achieved by a cytochrome P-450aromatase lion has been determined (5, 6). It probably acts in an aggre system. gate form involving at least 4 units and possibly as many as 16 Using the better studied cytochrome P-450Scc as a model, for optimal activity in vitro. Whether this is the case in vivo is the sequence of molecular events involved in steroid oxidation not clear because it is almost impossible to duplicate, in recon is represented simply in Chart 2. Cholesterol associates with stituted systems, the hydrophobic arrangement of the enzymes cytochrome P-450 in the oxidized state to form the substrate- in the inner membrane of the mitochondrion. Nevertheless, enzyme complex. The heme is reduced by electrons from the under proper conditions, it can be demonstrated to have a electron transport system described above. This complex is turnover rate of about 20 nmol of cholesterol per mol cyto- now sensitive to interaction with available molecular oxygen chrome P-450 per min (5). and oxidation of substrate occurs rapidly. If carbon monoxide The heme group of cytochrome P-450scc is usually in the is present, it binds preferentially to the heme and inhibits the oxidized state. To function, it requires electrons transported by reaction. It should be noted that CO binds to this heme only in a ferridoxin-like protein (iron sulfur protein) which, in turn, the reduced state. The extent of CO-heme interaction can be receives its electrons from a flavoprotein reducÃase. Finally, measured by the generation of a dramatic absorbance peak at the flavoprotein utilizes NADPH as an electron source. NADPH 450 nm. Most likely, the substrate remains at its site and is generated by reversed electron transport through the con undergoes 3 oxidation steps, but it is clear that conditions may ventional respiratory chain. Thus, cholesterol SCC is achieved be modified to favor the collection of intermediates. by a "multienzyme system," the terminal enzyme of which is Aromatase probably acts in a similar fashion although the cytochrome P-450Scc (Chart 2). details have not been described as clearly. Osawa (8) suggests The SCC system is found in the adrenal, ovary, testis, and that there may be 2 aromatase-linked cytochrome P-450s, placenta. Cytochrome P-450s that have been isolated from each specific for a different substrate. There is no evidence these sources are similar in amino acid constitution (6) and yet that each aromatase molecule is composed of several also in interaction with active site-oriented antibodies (7). subunits such that each subunit performs a different oxidation. Therefore, AG inhibits the P-450s in each of these tissues Since aromatase does not bind CO readily, experimentation equivalently under similar experimental conditions (4). utilizing that technique is not helpful. Obviously, purified cyto On the other hand, the aromatase system has not been chrome P-450aromataseÂ¡sneeded for further study to avoid the studied as thoroughly as the SCC system. As demonstrated obfuscations of other cytochrome P-450s.

3316s CANCER RESEARCH VOL. 42

Substrates and inhibitors which bind to cytochrome P-450 cytochrome P-450 chemically. This increment is very small. In of almost any species modify the spectrum of the cytochrome. the lower panel, AG is present in the mixture when NADPH is By difference spectroscopy, a technique used for turbid ma added. The rate of reduction is very slow, requiring about 30 terial which allows subtraction of background, 3 types of spec sec, and only about 50% of the cytochrome is eventually tra can be elicited: type 1 spectra, which correlate with the reduced at this concentration of AG. electron spin resonance high-spin state and are characterized Alternative explanations are possible. For example, the pres by a maximum at about 390 nm and a minimum at about 420 ence of substrate may be necessary before the enzyme is nm; reverse type 1 spectra, which have the opposite maxima reduced and AG may displace some of that substrate. This is and minima; and type II (low-spin) spectra, which have spectral less likely but it does gain support from some of the spectral maxima between 425 and 435 nm and a minimum at about 390 data which indicate that at this concentration of inhibitor, only nm. For mitochondrial cytochrome P-450s, these spectral about 50% of the enzyme is bound to the AG (Chart 4). changes are correlated with the binding of substrate or inhibitor such that substrate binding causes type I changes, steroidal Inhibition of Cholesterol SCC and Aromatization in Vitro inhibitors such as pregnenolone elicit reverse type I changes, and nitrogenous inhibitors such as AG effect type II changes. The inhibition of cholesterol SCC and aromatization by AG in Interpretation of microsomal cytochrome P-450 spectra are vitro is concentration dependent (Charts 4 and 5). From the more complex because substrate analogs produce type I spec curves of spectral change, the apparent Kss can be determined; tra. AG, however, elicits a type II spectrum. For both P-450s, the apparent KÂ¡sare determined from measurement of SCC or the amount of spectral change is concentration dependent and aromatization by corpus luteum mitochondria or placental mi correlates directly with the amount of enzyme inhibition (Charts crosomes by previously published methods (4, 11). Before 4 and 5). considering kinetic results of reactions involving the cyto chrome P-450 systems, it should be clear that these systems may not fulfill the "ideal" criteria for such analyses. For ex Mechanism of Inhibition ample, most kinetic analyses assume that the concentration of The spectral changes induced by AG confirm that it interacts the enzyme is rate limiting, that a "first-order reaction" is with the respective cytochrome P-450s. The molecular inter occurring, and that a single solubility phase exists, to cite a action is also understood. From Chart 2, it is evident that the few conditions. Even highly purified cytochrome P-450 systems first step in the cleavage of cholesterol is the binding of sub strate to cytochrome P~450Scc in the oxidized state. The next step is the reduction of the substrate-bound cytochrome P-450 100 90 by the transfer of an electron from the iron sulfur protein. Chart

3 demonstrates that AG prevents this reduction. In the upper O 75 curve, electrons are introduced by the addition of NADPH. As Aminoglutelhimide 60 the cytochrome P-450 is reduced, it combines with CO to yield OD X 50 the typical increase in absorbance at 450 nm. Almost total OD reduction is achieved in a few sec, as evidenced by the steep Z so slope of the curve. The total amount of reduction possible is demonstrated by the addition of dithionite, which reduces the

P-450-CO 100 1000 (A 450-490 nm) io INHIBITOR (>jM)

Chart 4. Comparison of inhibition of cholesterol SCO by AG with the changes in absorbance of type II spectra induced by various concentrations of AG. Cholesterol SCC was determined enzymatically; absorbance changes were mea DITHIONITE |Â»v~ sured in a dual-beam spectrophotometer (from Ref. 9). NADPH (img) (0.4mM) 100- d - AG CONTROL

AG (82 MM) l 60 -l 005 OD L S

40 10 20 30 40 SO 70 TIME SECONDS Chart 3. Rate study of NADPH-supported reduced CO:P-450 complex for 20 mation. A mixture of purified cytochrome P-450scc. flavoprotein reductase, and iron sulfur protein in phosphate buffer was gassed with CO in a dual-wavelength spectrophotometer. The reaction is initiated by the addition of NADPH. The I IO 100 1000 increase in absorbance (OD) at 450 nm in reference to 490 nm is monitored. In the upper portion of the chart, AG was absent. In the lower portion of the chart, Aminoglutethimide UM) the presence of 62 fiM AG (about half-maximal inhibitory concentration) de Chart 5. Concentration-dependent inhibition of testosterone aromatizaron in creased the rate of reduction by about one-half. human placental microsomes by r>, L-, and DL-AG (from Ref. 10).

AUGUST 1982 3317s

may not fulfill these criteria. For example, certain steroids such Table 1 as cholesterol are poorly soluble in aqueous solutions and Comparison of D-AG, L-AG, and DL-AG: affinity for cytochrome P-450SCc and cytochrome P-450aroms,aseand inhibition of cholesterol SCC and aromatase often exist in micellar arrangements; detergent actions are enzyme systems (from Refs. 2 and 10) unpredictable; the system contains 2 or 3 components such as Aminoglutethimide (JIM) reductase or iron sulfur protein, any of which may be rate limiting; 2 or 3 oxidations may proceed at different rates; the enzyme complex may exist in various aggregate states of Half-maximal inhibition different activities; and, finally, even in the best-reconstituted Aromatase 8 310 14 Cholesterol SCC 16 40 26 systems containing ideal amounts of all enzymes, cofactors, and substrate, oxygen may become rate limiting. Finally, the SpectralK5AromataseWithout transference of in vitro data to in vivo situations is particularly testosteroneWith hazardous for these membrane-bound enzymes. testosteroneCholesterol Nevertheless, certain kinetic constants can be reported. SCC0.112303.662780.31936 Table 1 lists the apparent Ks and apparent KÂ¡data for the enantiomers of AG. The o-enantiomer is about 2.5 times more effective than the L-enantiomer in the inhibition of cholesterol SCC. It should be noted that the SCC determinations have 80 been made in the presence of saturating amounts of substrate, because it is almost impossible to remove cholesterol from the 60 d-Ag enzyme preparations without denaturing the enzyme. On the B 2= (7) other hand, for placental aromatase, the ratio of activities of 2 Â». 40 O) O the D- to L-enantiomers are between about 5 and 40, depending I- (6) (7) upon the presence or absence of substrate. Placental micro- g oj Q- " somes ordinarily do not have large concentrations of substrate, O but this, of course, may not apply to in vivo situations. These findings have important implications for the clinical -20 (8) (7) use of AG. First, if the action of AG on breast cancer cells is l-Ag -40 mediated through its ability to inhibit aromatase, then the use of D-AG would be preferred because of its higher potency. The toxicity data appear to indicate that the L-enantiomer may be E-3.0 somewhat more toxic but certainly not less toxic. Thus, the C O O" CONTROL therapeutic ratio heavily favors the D-enantiomer. .2 w E Second, and perhaps of more serious concern, the slope of 0 o1^ 1.0 the inhibition curve is low. This means that doubling the con centration has little effect. For example, from Chart 5, increas ing the concentration of the DL-mixture from 10 to 100 UM 0.2 0.4 0.8 hardly doubles the amount of inhibition. In clinical practice, Dose (mg) therefore, one can anticipate that even doubling the dosage of AG from the usual level of 1 to 2 g (a dose which is clearly Chart 6. Simultaneous bioassay of D-AG and L-AG in rat plasma and in rat ovaries. Decrease in progesterone concentrations is expressed as percentages intolerable for most women), may not greatly increase its bio of initial plasma level (/PL). Dotted area, S.E. of the mean of the control rats. logical effectiveness unless the higher dose were to result in Numbers in parentheses, number of rats at each point. It should be noted that in markedly higher tissue concentrations. Clearly, more work the upper panel, the negative slope of the regression line indicates an increase in plasma progesterone levels (from Ref. 11 ). needs to be done on the pharmacology of the enantiomers, their tissue distribution, and relative effectiveness in humans. The upper panel of Chart 6 introduces a new concept. D-AG Inhibition of Cholesterol SCC and Aromatization in Vivo effects a dose-responsive decrease in progesterone levels in plasma but, paradoxically, L-AG causes an increase in proges Cholesterol SCC inhibition by AG has been demonstrated to occur in rats, rabbits, and primates in dose-responsive fashion terone levels. We attribute this to the inhibition of hepatic (12). Less work has been published on the inhibition of aro degradative enzymes by the L-enantiomer. Further evidence of matase, but our unpublished data" confirm that the results are this dichotomy of action of the enantiomers is the observation by Schleyer and Cooper5 that the L-form binds to liver micro- similar, although, as expected, the aromatase systems are more sensitive. Chart 6 shows dose-responsive curves for the somal P-450 more than does the D-form. inhibition of cholesterol production by D- and L-AG. Referring If the human responds similarly, this effect introduces an to the lower panel, which measures the depletion of progester other reason why the D-enantiomer would be preferable to the one in the ovary, it is clear that, over a 4-fold dose range, D-AG L-enantiomer. Thus, not only does the use of the DL-mixture induces a 70% decrease in ovarian progesterone concentra not provide optimal inhibition from the aspect of potency, but tion. In contrast, only the highest dose of the L-enantiomer had it also introduces unnecessary toxicity and hepatic inhibition any effect on the progesterone concentration. which are contrary to the desired effect.

' E. N. Mclntosh, H. R. Holtrop. C. A. Whipple, and H. A. Salhanick. unpub 1H. Schleyer and D. Cooper, personal communication. lished observations.

3318s CANCER RESEARCH VOL. 42

Inhibition of Cholesterol SCC and Aromatase in the Human g unless cited otherwise in the charts, was administered p.o. Blood samples were collected through the catheter at progres The inhibition of cholesterol SCC and/or aromatization by sively increasing intervals. In the case of the postmenopausal AG were tested in 3 states in the human: various stages of pregnancy; the luteal phase of the menstrual cycle; and in adrenalectomized, ovariectomized women to whom the precur wo .Â«Q. sor androstenedione was administered. In pregnant women (Charts 7 and 8), it was possible to test the ability of AG to 50 inhibit various rates of production of each of the steroids. The CONTROL 4I-6I DAYS human placenta appears to be autonomous in the production of these steroids, so that feedback mechanisms are not oper 50 20 ative in short-term experiments. In the luteal-phase experiment o U 44 DAYS (Chart 9), we determined the effects on ovarian production of 0 these steroids under the control of pituitary gonadotropins so IOO that a feedback mechanism exists comparable to that of the adrenal. In the studies with postmenopausal women (Chart 10), N- 5I DAYS from whom both glandular sources of precursors had been removed, the inhibition of peripheral aromatization was studied. ISO

The experimental technique was similar in all cases. After a O 58 DAYS suitable period of rest, an indwelling catheter provided control 50 Â« I50 samples over a period of at least 1 hr. A single dose of AG, 1 P 80 DAYS 5O 400

200 Q:84 DAYS

200

IOO 20 IO R 99 DAYS 0 4OO

200 20 K) S: 125Ãˆiboi'Ã´.Å“ DAYS io 2 ' ' 4 12 24 t HOURS AG LOG HOURS Chart 8. Effects of a single p.o. dose of AG on plasma estradici at various times in pregnancy. Conditions are as described in Chart 7. Days refer to days of pregnancy from the last menstrual period.

HOURS LOG HOURS 1 24 8 12 24 48 72 Chart 7. Effects of a single p.o. dose of AG (1 g) on plasma levels of TIME (HOURS) progesterone at various times in pregnancy. Progesterone levels are expressed as actual concentrations and increase as pregnancy progressed. Hatched area, Chart 9. Effects of a single dose of 1.5 g of AG on hormonal levels in the luteal phase. Top, AG levels; bottom, LH, progesterone, and total estrogens. plasma concentrations of AG. O, control values; â€¢¿.levelsobtained after AG was Ordinate is presented as a logarithmic-type scale. At 48 hr, there is a "rebound" administered. Days refer to days of pregnancy from the last menstrual period. The control subjects received glutethimide. elevation of the steroids coincidental with decreases in LH.

AUGUST 1982 3319s

-100 administered i.V., followed by a constant infusion of andro stenedione. From the other arm, blood was withdrawn by an infusion pump at a rate of about 20 ml/hr. Androstenedione and estrone were measured in the plasma at 30-min intervals. Control levels were measured during the first 3 hr to establish baseline concentrations of estrone, and then 1 g of o-AG was taken p.o. Estrone levels dropped rapidly to about half the levels obtained prior to the administration of androstenedione and remained depressed for over 8 hr. In summary, AG has been demonstrated to inhibit cholesterol 1 2 3 4 5 TIME (HOURS) SCC and/or aromatization in all postmenarcheal phases of the Chart 10. Effects of D-AG (1 g) on peripheral aromatizaron of androstenedi- female reproductive cycle. one in an adrenalectomized, ovariectomized woman. A, androstenedione; E, estrone. Conditions are as described in the text. Clinical Correlations women, blood was collected by effusion pump over the indi Chemotherapeutic drugs should be potent, specific, and cated time intervals. Progesterone and estrogen levels were devoid of side effects, but these ideals are not often achieved. determined by radioimmunoassay. Plasma levels of AG were AG is clearly not ideal. It is insufficiently potent, it affects measured chemically. enzymes other than the steroidogenic enzymes, and its side In pregnancy, the standard dose of AG inhibited production effects are unacceptable to some patients. On the other hand, of estrogens and progesterone throughout the first trimester therapies cannot be considered in vacuo. AG is more potent (Charts 7 and 8). Even though the baseline levels increased than most, if not all, other inhibitors of these steroidogenic progressively, reflecting increased tissue enzyme concentra enzymes insofar as systemic human use is concerned. Steroi- tions and increased steroid production, the amount of inhibition dal inhibitors which we have tested in vivo in lower species do remained at about 50%. The steroid concentrations corre- not fulfill their promise based upon in vitro data. Obviously, the sonded inversely to the concentrations of AG. At peak concen "killer substrates" and d9 analogs require testing in animals. trations of AG, the plasma concentrations were approximately Second, AG remains more specific than other inhibitors, until 20 to 30 /IM, which is approximately the KÂ¡determined in vitro these potential alternative inhibitors have been through the (see above). Although the absolute decrements in steroid levels extensive testing with other enzyme systems to which AG has tend to show an increase in absolute terms as pregnancy been subjected over the past decade. Finally, the side effects progresses, nevertheless, the fraction of the initial blood level of AG may not be pleasant but they are probably preferable to remains in the 50% region. Finally, it should be noted that, the effects associated with surgical ablation of the steroido when the AG had been cleared from the blood, there was no genic organs. overshoot in steroid levels. No changes were observed in On the other hand, AG is a model substance for improved plasma human chorionic gonadotropin levels implying lack of inhibitors to be developed. We know now that the antisteroi- feedback control. dogenic activity resides mostly in the D-isomer. We believe that In the luteal-phase experiment (Chart 9), the day of ovulation this specificity should be explored and exploited unless or until was ascertained by the presence of a LH peak. Four days later, other inhibitors can be demonstrated to be preferable. when the plasma progesterone had reached concentrations of about 7 ng/ml, the single-dose experiment was performed. Acknowledgments This woman received 1.5 g of AG. Blood levels of AG rose rapidly and plasma levels of estradiol and progesterone de The author wishes to acknowledge the contributions of his colleagues in the creased inversely to a nadir of about 30 to 40% of the initial published and unpublished work cited herein. They are M. Glode, P. Graves, J. Hourihan, K. Kashiwagi, E. N. Mclntosh, J. Strauss, V. I. Uzgiris, and C. A. blood levels. They remained suppressed for about 12 hr. Within Whipple. about 3 to 4 hr, there was a reflex release of LH, the concen trations of which doubled by the eighth hr. At this time, the References steroid levels began to rise and AG and LH levels began to decrease. Of interest, there was a marked increase in both 1. Dexter, R. N., Fishman, L. M., Ney. R. L., and Liddle, G. W. Inhibition of adrenal corticosteroid synthesis by aminoglutethimide: studies of the mech steroids on the following day and possibly a decrease in LH anism of action. J. Clin. Endocrinol. Metab.. 27: 473, 1967. levels. Not shown in the figure are follicle-stimulating hormone 2. Fishman, J. Biochemical mechanism of aromatization. Cancer Res. (Suppl.), levels which paralleled the LH changes, and cortisol and ad- 42: 3277S-3280S, 1982. 3. Graves, P. E., and Salhanick, H. A. Stereoselective inhibition of aromatase renocorticotropic hormone levels, which showed similar rela by enantiomers of aminoglutethimide. Endocrinology, 105: 52-57, 1979. tionships. Thus, it is clear that AG does affect the premeno- 4. Graves, P. E.. Uzgiris, V. I., and Salhanick, H. A. Modification of enzymatic activity and difference spectra of cytochrome P-450 from various sources pausal ovary but that its effects are overcome by increased by cholesterol side chain cleavage inhibitors. Steroids, 35. 543-559, 1980. gonadotropin secretion. The data of Santen et al. (9) measured 5. Kashiwagi, K., Dafeldecker, W. P., and Salhanick, H. A. Purification and over an entire menstrual cycle tend to confirm this phenomenon characterization of mitochondrial cytochrome P-450 associated with choles terol side chain cleavage from bovine corpus luteum. J. Biol. Chem., 255. but are not as clear-cut. 2606-2611, 1980. The third clinical study tested the effectiveness of AG in 6. Kashiwagi, K., Carraway. R. E., and Salhanick, H. A. Amino acid composition inhibiting the peripheral conversion of androstenedione to es of cytochrome P-450Scc from bovine corpus luteum. Biochem. Biophys. Res. Commun., 705: 110-116, 1982. trone in an adrenalectomized and ovariectomized woman 7. Kashiwagi, K.. MacDonald, A. B., and Salhanick, H. A. Inhibition of choles (Chart 10). A loading dose of 2 mg of androstenedione was terol side chain cleavage by active site directed antibody to corpus luteum

3320s CANCER RESEARCH VOL. 42

cytochrome P-450. J. Biol. Chem., 257; 2212-2217, 1982. corpus luteum mitochondria! cytochrome P-450 spectra and cholesterol 8. Osawa, Y., Tochigi, B., Higashiyama, T.. Yarborough, C., Nakemura, T., and monoxygenation: an assay of enzyme-specific inhibitors. Biochemistry. 16: Yamamoto. T. Multiple forms of aromalase and response of breast cancer 593-600, 1977. aromatase to antiplacental aromatase II antibodies. Cancer Res. (Suppl.), 11. Uzgiris, V. I., Whipple, C. A., and Salhanick. H. A. Stereoselective inhibition 42: 3299S-3306S, 1982. of cholesterol side chain cleavage by enantiomers of aminoglutethimide. 9. Santen, R. J., Samojlik, E.. and Wells. S. A. Resistance of the ovary to Endocrinology, 101: 89-92, 1977. blockade of aromatization with aminoglutethimide. J. Clin. Endocrinol. Me- 12. Whipple, C. A., Colton, T., Strauss. J. M., Hourihan, J., and Salhanick, H. A. tab.. ST. 473-477, 1980. Comparison of luteolytic potencies of aminoglutethimide enantiomers in the 10. Uzgiris, V. I., Graves, P. E., and Salhanick, H. A. Ligand modification of rabbit and rat. Endocrinology, 103: 1605-1610, 1978.

AUGUST 1982 3321s

Hilton A. Salhanick

Cancer Res 1982;42:3315s-3321s.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/42/8_Supplement/3315s

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/42/8_Supplement/3315s. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.