Progestational and androgenic receptor binding affinities and in vivo activities of norgestimate and other progestins

CONTRACEPTION

PROGESTATIONAL AND

A.

IN VIVO

Phillips, R. W.

AND

ANDROGENIC

ACTIVITIES

K. Demarest,

RECEPTOR

OF NORGESTIMATE

D.W.

Hahn,

BINDING AND

F. Wong

OTHER

AFFINITIES PROGESTINS

and J.L.

McGuire

Johnson Pharmaceutical Research Institute at Ortho Pharmaceutical Corporation USA Raritan, N.J. 08869

ABSTRACT The progestational and androgenic in vitro receptor binding affinity and the in vivo activity of norgestimate was compared with that of its metabolites and other progestins. The relative binding affinities (RBAs) of norgestimate and its 17deacetylatedmetabolite for rabbit uterine progestin receptors were similar to that of progesterone (P); those of 3-keto norgestimate and levonorgestrel were about five times that of P; those of gestodene and 3-keto desogestrel were about nine times that of P. The RBAs of norgestimate, P, and 3-keto norgestimate for rat prostatic androgen receptors were from 0.003 to 0.025 times that of dihydrotestosterone (DHT); those of 3-keto desogestrel, gestodene, and levonorgestrel were from 0.118 to 0.220 times that of DHT. The order of receptor level selectivity represented by the ratio of androgen:progestin (with a greater ratio value reflecting a better IC,, values selectivity) was norgestimate z= P = 3-keto norgestimate > 17deacetylated norgestimate > 3-keto desogestrel > gestodene > levonorgestrel. Invivo studies demonstrated similar profiles for norgestimate and its 17-deacetylated metabolite. These latter two steroids were equally potent as progestins in stimulating rabbit endometrium, and compared with the other progestins, both steroids exhibited minimal androgenicity as measured by the stimulation of rat prostate growth. In conclusion, these studies, as well as previous preclinical and clinical studies, provide evidence of the selectivity of norgestimate based on minimal androgenicity, indicating an improvement over other progestins used in oral contraceptives.

Submitted Accepted

for publication for publication

APRIL 1990 VOL. 41 NO. 4

October December

10, 21,

1989 1989

3!39

CONTRACEPTION

INTRODUCTION Progestins contained in oral contraceptives have exhibited androgenic activity. Androgenic stimulation results in decreased blood levels of high density lipoprotein cholesterol (HDL).'*2'3 Therefore, research has been directed toward the development of a progestin with improved selectivity based on reduced androgenicity. Preclinical evaluation of norgestimate indicated that it is an effective progestational agent with minimal androgenicity.4s5*6 Subsequent clinical studies confirmed the minimal androgenicity of this progestin by demonstrating that norgestimate in combination with ethinyl estradiol (EE) induces a significant elevation in blood levels of HDL, decreases the ratio of low density lipoprotein cholesterol (LDL) to HDL (LDL/HDL), and increases blood levels of sex hormone binding globulin (SHBG).' Further evidence of norgestimate's selectivity is demonstrated by its comparatively low relative binding affinity for human sex hormone binding globulin.8 Preclinical and clinical investigations have shown clear differences in the pharmacologic responses of norgestimate and levonorgestrel. Unlike levonorgestrel, norgestimate exhibits no greater androgenicity in laboratory animals than does the natural hormone, progesterone.4 Clinical studies have contrasted the minimal androgenic response to norgestimate to the marked androgenicity of levonorgestrel.' These studies demonstrated the inability of norgestimate (0.25 mg) to appreciably inhibit the EE-induced elevation in blood levels of HDL and SHBG, whereas levonorgestrel (0.15 mg) clearly reversed these effects. The studies reported here expand our previous investigations by further characterizing the progestational and androgenic activities of norgestimate compared with that of its metabolites and other new progestins.

MATERIALS

AND METHODS

COMPOUNDS Norgestimate [(+)-13-ethyl-17-acetoxy-18,19-dinor-17&pregnnorgestimate, 3-keto 4-en-20-yn-3-one oxime], 17-deacetylated levonorgestrel, gestodene, 3-keto desogestrel norgestimate, (the active metabolite of desogestrel), dihydrotestosterone were either synthesized at the R. W. (DHT) I and progesterone Johnson Pharmaceutical Research Institute or obtained commercially. For in vitro studies, the steroids were dissolved in ethanol and diluted in Tris-EDTA buffer to yield a final 5%

400

APRIL

1990 VOL. 41 NO. 4

CONTRACEPTION

ethanol concentration. For in vivo studies, dissolved and administered in sesame oil. IN VITRO

BIRDING

the steroids

were

STUDIES

Prosestin Receptor Studies cytosolic The relative affinities of the steroids for progestin receptors were assessed by their ability to displace 3H-R5020, a highly specific synthetic progestin, from rabbit uterine receptors. Immature female rabbits were primed with 17 B-estradiol (5 mcg SC) daily for five days and then sacrificed on day six. The excised uteri were washed and homogenized in ice cold Tris-EDTA buffer (1:5 w/v; pH 7.4). The homogenate was centrifuged at 105,000 x g at 4°C for one hour and the supernatant was used as the cytosolic progestin Aliguots of the cytosol preparation receptor preparation. were incubated with 0.4 nM 3H-R5020 (86.2 Ci/mmole; New and unlabelled test steroids in Tris-EDTA England Nuclear) buffer for 5 hours at 4%. After incubation, free 3H-R5020 was extracted with dextran-coated charcoal from the bound ligand, and the amount of 3H-R5020 bound to the receptor was The condetermined by liquid scintillation spectrometry. centration of the test progestin corresponding to 50% inhib3H-R5020 binding (I&,) was used as the measure ition of total of binding affinity for the progestin receptor. The relative was calculated by binding affinity (RBA) of each steroid dividing the- IC,, of progesterone by the IC,, of the test steroid. Androuen Recerkor Studies The relative affinities of the test steroids for cytosolic androgen receptors were assessed by their relative ability to 3H-dihydrotestosterone (H-DHT) from rat prostatic displace receptors. Adult male rats were sacrificed 24 hours after bilateral castration. The excised ventral prostates were washed and homogenized in ice cold Tris-EDTA buffer (1:3 w/v). The homogenate was centrifuged at 105,000 x g for one hour and the supernatant was used as the cytosolic androgen receptor preparation. Aliguots of the cytosol preparation were incubated with 1.0 nM 3H-DHT (50.6-58.4 Ci/mmole; New England Nuclear) and the unlabelled test steroid in Tris-EDTA buffer at 4°C for 24 hours. The free 3H-DHT was extracted with dextran-coated charcoal from the bound ligand and the to the receptor was determined by amount of 'H-DHT bound liquid scintillation spectrometry. The data were calculated and expressed as described above for the progestin receptor studies. Stability of 3H-Noruestimate The concentrations of norgestimate and its metabolites, L7deacetylated norgestimate, 3-keto norgestimate and levonorgestrel, in the cytosolic receptor preparations were measured by HPLC using an external standard method employing 3H-Norgestimate (Roussel, radioactivity flow detection.

APRIL

1990 VOL. 41 NO. 4

401

CONTRACEPTION was incubated under the experimental conditions France) described above for the binding assays. Following incubation, aliguots of the buffer/cytosol preparation were injected directly on an HPLC column (u-Bondapak C,, column, 30 cm x 3.9 a 68~32 mm i.d., 10 u; Waters Assoc., Milford, MA), utilizing methanol:water mobile phase, a flow rate of 1.4 ml/min, and an operating pressure of 150-160 bar at ambient temperature. Radioactivity was quantified and analyzed using a FLO-ONE model HS radioactive flow monitor (Radiomatics Instrument and Chemical Co., Inc., Tampa, FL). The limit of quantification of this method is 1980 dpm of 3H-norgestimate per injection. HNUOCRINE

BIOASSAYS

Procrestational Activitv in Rabbits Progestational activity was measured by the Clauberg bioassay with modifications previously described.' The assay is based on the ability of orally administered progestins to stimulate endometrial an response in the estrogen-primed immature rabbit. The progestational activity was also measured by the method of McGinty,g which determines the endometrial response of a compound injected directly into the lumen of one uterine horn of an estrogen-primed rabbit. Androqenic Activity Androgenic activity was measured by the method of Hershberger with modifications previously described.‘ The assay is based on the ability of androgens to stimulate the growth of the ventral prostate of immature castrated rats after either subcutaneous or oral administration. STATISTICAL ANALYSIS Relative progestational mated using the method

and andrqgenic of Finney.

potencies

were

esti-

RESULTS PROGESTATIONAL

RESPONSES

Relative Bindin o Affinity for Proa e&in R ecerkors The binding affinities of norgestimate, its 17-deacetylated and 3-keto-metabolites, levonorgestrel, gestodene and 3-keto desogestrel for the progestin receptor were compared to that of progesterone, which was assigned a relative binding The order of relaaffinity (RBA) value of 1.00 (Table I). tive binding affinities of the steroids was gestodene 2 3keto desogestrel > levonorgestrel, 3-keto norgestimate > norgestimate 2 17-deacetylated norgestimate 1 progesterone.

402

APRlL1990VOL.41 NO.4

Q

0.47(0.11) 0.51(0.13) 127.15(27.0)

Gestodene

3-Keto Desogestrel

Dihydrotestosterone

0.03

8.49

9.21

5.41

5.21

0.94

1.24

the mean (SE) of 3-4 experiments.

O.gO(O.14)

Levonorgestrel

* Values represent

0.83(0.11)

4.61(0.98)

17-Deacetylated

3-Keto Norgestimate

3.48(0.47)

Norgestimate

4.33(1.11)

Progesterone

RBA

2CO.2)

17 (3)

13 (2)

9 (1)

77 (12)

222 (53)

764 (45)

401 (41)

1.000

0.118

0.154

0.220

0.025

0.013

0.003

0.005

RBA

Binding

IC50,nM mean (SE)*

Androgen

RECEPTORS

1.00

Binding

IC50,nM mean (SE)*

Progestin

ANDROGEN

Norgestimate

Compound

AND RAT PROSTATIC

RELATIVE BINDING AFFINITY FOR RABBIT UTERINE PROGESTIN

TABLE I

0.02

33

28

11

92

48

219

93

IC50

A/P

CONTRACEPTION

TABLE

II

STIMULATION OF ENDOMETRIUM IN IMMATURE ESTROGEN-PRIMED RABBITS AFTER EITHER ORAL OR DIRECT UTERINE ADMINISTRATION: PROGESTATIONAL RELATIVE POTENCIES

COHPOUND

Study I (Oral 1

Study IIb IOral I

Norgestimate 17-Deacetylated Norgestimate 3-Keto Norgestimate Levonorgestrel

1.00 1.38 3 *3ga 4.92a

Norgestimate

1.00

Levonorqestrel Desogestrel

Study III (Oral 1

Norgestimate Gestodene

Study IV (Direct 1

Norqestimate Levonorgestrel

:Respon-EesigniEicantly The results

of

RELATIVE POTENCY

this

(0.98-1.88) (2.45-4.64) (3.50-6.75)

4.23’ 7.1ga

(2.80-6.41) (5.00-10.56)

1.00 LO.54a

(7.81-13.99)

1.00 0.92

(0.61-1.341

different from that of norqegtimate study were reported previously.

The progestin binding affinities of deacetylated norgestimate were similar one; those of levonorgestrel and 3-keto about five times that of progesterone, and gestodene were about desogestrel progesterone.

95% FIDUCIAL LIMITS

(p = 0.05).

norgestimate and 17to that of progesternorgestimate were both while those of 3-keto nine times that of

Evaluation of the stability of norgestimate under the conditions of the receptor binding studies showed that 'Hincubated alone or in the presence of pronorgestimate, only was stable for the incubation period. gestin receptor, 1.1% of the radioactivity was not associated with norgestimate and this was attributable entirely to 17-deacetylated norgestimate. The progestational potencies of orally administered 17norgestimate and 3-keto norgestimate, deacetylated that of norlevonorgestrel were 1.4, 3.4 and 4.9 times in the estrogen-primed rabbit (Table gestimate, respectively, In separate experiments using the same test II; Study I).

404

APRIL 1990 VOL. 41 NO. 4

CONTRACEPTION

system, the potencies of levonorgestrel, desogestrel, and gestodene were 4.2, 7.2 and 10.5 times that of norgestimate, After direct respectively (Table II; Studies II, III). norgestimate and levonorgestrel intrauterine administration, stimulated the endometrium to a similar degree (Table II; study IV). ANDROGENIC

RESPONSES

Relative Bindinc Affinity for Androaen Receptors The binding affinities of the progestins for the androgen receptor were compared to that of DKT, which was assigned an Norgestimate RBA of 1.00 (Table I). and progesterone exhibited the weakest affinity for the androgen receptor with RBAs 0.003 and 0.005 times that of DHT, respectively. The 17-deacetylated norgestimate and 3norgestimate metabolites, also bound relatively weakly, with RBAs keto norgestimate, 0.013 and 0.025 times that of DHT, respectively. Stronger affinities for the androgen receptor were demonstrated by 3keto desogestrel, gestodene and levonorgestrel, which had RBAs 0.118, 0.154 and 0.220 times that of DHT, respectively. The receptor-level selectivity of each progestin is demonstrated by the androgen to progestin ratio of IC,, values with a larger value reflecting a greater selectivity (Table I). The order selectivity is norgestimate > progesterone of = 3-keto norgestimate > 17-deacetylated norgestimate > 3-keto desogestrel > gestodene > levonorgestrel. Androuenic Activity in Rats The androgenic potencies of the various progestins after subcutaneous administration were compared to that of testosterone propionate (TP). The ability of norgestimate and 17-deacetylated norgestimate to stimulate ventral prostate growth in immature rats was very poor and similar to that of progesterone (Table III; Studies I, II). Levonorgestrel and 3-keto norgestimate exhibited significantly greater androgenic potencies which were 0.10 and 0.09 times that of TP respectively (Table III; Study II). Gestodene was 2.4 times more potent than norgestimate (or 0.1 times as potent as TP) when given subcutaneously (Table III: Study III). The order of oral potencies for norgestimate, 17-deacetylated norgestimate, levonorgestrel and 3-keto norgestimate were similar to that observed after subcutaneous administration. The oral potency of desogestrel was not significantly different than that of levonorgestrel (Table III: Study V), and the ability of gestodene to stimulate ventral prostate growth after oral administration was similar to that of norgestimate.

APRIL

1990 VOL. 41 NO. 4

CONTRACEPTION

TABLE III EITHER

STIMJLATION OF VENTRAL PROSTATE GROWTH IN IMMATURE RATS AFTER SUBCUTANEOUS OR ORAL ADMINISTRATION: ANDROCENIC RELATIVE POTENCIES Compound

Study Ia (S.C.)

Relative Potency

95% Fiducial Limits

Testosterone Propionate Norgestimate Progesterone Levonorgestrel

1.000

Study II (S.C.)

Testosterone Propionate Norgestimate 17-Deacetylated Norgestimate 3-Keto Norgestimate Levonorgestrel

1.000 0.010 0.007 0.090 0.100

10.005-0.020) (0.003-0.012) (0.060-0.130) (0.070-0.150)

Study III (S.C.)

Testosterone Propionate Norgestimate Levonorgeitrel Gestodene

1.000 0.031 0.112 0.101

(0.017-0.074) (0.068-0.203) (0.036-0.2321

Study IV (oral)

Methyl Testosterone Norgestimate 17-Deacetylated Norgestimate 3-Keto Norgestimate Levonorgestrel

1.000 0.012 0.010 0.019 0.022

(0.005-0.024) (0.004-0.0211 (0.008-0.037) (0.010-0.0431

Study va (oral)

Hethyl Testosoterone Norgestimate Desogestrel Levonorgestrel

1.000 0.026 0.081 0.113

0.007-0.060) 0.036-0.140) 0.051-0.200)

Study VI (oral)

Norgestimate Gestodene

1.00 1.02

0.81-1.22)

0.003

0.004

0.120

i The results of this study were reported previously.' The relative potency of gestodene in this study vas 2.43 (1.34-4.46) times that of norgestimate.

APRIL

1990 VOL. 41 NO. 4

CONTRACEPTION

DISCUSSION

These studies confirm the previous preclinica14 and clinical' investigations of norgestimate demonstrating that it is an effective progestin with minimal androgenicity. It binds to uterine progestin receptors with an affinity equal to that of progesterone and has an affinity for androgen receptors that is even weaker than that of progesterone. This profile of relative receptor binding affinity indicates that norgestimate's receptor-level selectivity appears superior to that of levonorgestrel, gestodene, and 3-keto desogestrel, The excellent the active metabolite of desogestrel. progestational and weak androgenic potencies of norgestimate are also demonstrated by the in vivo studies in laboratory animals reported here. The stability of norgestimate under the conditions of these in vitro binding studies and its progestational stimulation when injected directly into the lumen of the uterus, provide evidence that norgestimate possesses inherent progestational activity. The results of the studies reported here showing that the affinity of norgestimate for the progestin receptor is similar to that of progesterone are in contrast to a previous report.ll This difference is attributable to the binding of norgestimate to the glass assay tubes in earlier experiments, a condition which was later minimized by using ethanol instead of dimethyl sulfoxide as the solvent. There are previous reports of studies using various methods and receptor sources to examine the relative binding affinity of levonorgestrel, gestodene and 3-keto desogestrel for progestin and androgen receptors. Kloosterboer et al.,l'using either Org 3236 or Org 2058 as the ligand in intact MCF-7 cells, ranked the affinity of these agents for progestin receptors as: gestodene 2 3-keto desogestrel > levonorgestrel. Then, using DHT as the ligand in either intact MCF-7 cells or MCF-7 cytosol, they ranked the affinity of the progestins for the androgen receptor as: levonorgestrel 2 gestodene > 3-keto desogestrel. These relative binding affinities are in agreement with those of the present experiments. The relative binding affinities for these three progestins reported here were also similar to those reported by Spona and Huber13 and by Bergink et al.“ The relative binding affinities of these progestins for androgen receptors reported here were similar to those reported by Bergink,14whereas the results of Spona15 differ from ours in that 3-keto desogestrel exhibited a greater affinity for the mouse kidney cytosol androgen receptor than did levonorgestrel. The minimal androgenic potential of norgestimate compared with other progestins is also demonstrated by its ability to stimulate ventral prostate growth in rats only to the same degree as progesterone. This is true after either oral or subcutaneous administration. The evaluation of parenterally

APRIL 1990 VOL. 41 NO. 4

407

CONTRACEPTION

administered progestins in this bioassay is particularly important since several progestins have been shown to exhibit relatively poor oral bioavailability and progestational potency in the rat.4s'6 Norgestimate, however, maintains antiovulatory potency after oral administration in the rat.4 It is important to note that orally administered levonorgestrel and desogestrel are significantly more androgenic than norgestimate in the rat, even in the face of their relatively poor oral antiovulatory potency (less than that of norgestimate) in this species.4 It has been suggested that norgestimate must be metabolized to levonorgestrel to exert an effect. However, the results of preclinical studies reported here and previously,4 as well as the results from clinical studies,' indicate that this is The studies reported here establish that not the case. unchanged norgestimate possesses inherent progestational activity as shown by its affinity for progestin receptors and its direct progestational effects on target organs. Although this inherent activity does not preclude norgestimate from acting via a metabolite, it bhows that the parent drug has the potential to contribute to the response. The dramatic pharmacologic differences between norgestimate and levonorgestrel that have been demonstrated here and in previous studies both in laboratory animals4 and in humans' indicate that it is unlikely that norgestimate could be acting mainly via levonorgestrel. In multi-center clinical evaluations, combination oral contraceptives containing EE with either norgestimate or levonorgestrel, which were equally effective in inhibiting pregnancy , resulted in very different responses in androgen-sensitive parameters such as levels of SHBG, HDL, and LDL, demonstrating the minimal clinical androgenicity of norgestimate in contrast to the significant androgenic effects of levonorgestrel.' As with any steroid, norgestimate is extensively metabolized and metabolites may participate together with the parent drug in the biological response. The profile of urinary metabolites of norgestimate in humans" indicates that norgestimate is metabolized mainly by the loss of the 17-acetate and 3oxime groups, hydroxylation at various positions, and subsequent conjugation. Although levonorgestrel appears to be one of many metabolites of norgestimate, the weight of evidence demonstrating that the pharmacology of these two drugs is very different, indicate that levonorgestrel does not play a major role in the response to norgestimate. However, studies reported here showing that anothermetabolite of norgestimate, 17-deacetylatednorgestimate, exhibits pharmacology consistent with that of norgestimate, indicate that it may contribute significantly to the response. studies are ongoing investigating the pharmacokinetics of norgestimate and its metabolites, and the possibility of the existence of other active metabolites.

APRIL 1990 VOL. 41 NO. 4

CONTRACEPTION

In conclusion, these preclinical studies demonstrate that norgestimate is a selective progestin based on minimal androgenicity; this selectivity is consistent with reported clinical observations.

REFERENCES 1.

Fotherby K. Effect of oral contraceptives on serum lipid and cardiovascular disease. Br J Family Planning 1985; 11: 8691.

2.

Silfverstolpe G, Gustafson A, Samsoie G, Svanborg A. Lipid metabolic studies in oophorectomized women. Effects of three different progestagens. Acta Obstet Gynecol Stand 1979; Suppl 88: 89-95.

3.

Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein as a protective factor against coronary heart disease. Am J Med 1977; 62: 707-714.

4.

Phillips A, Hahn DW, Klimek S, McGuire JL. A comparison of the potencies and activities of progestogens used in contraceptives. Contraception 1987; 36: 181-192.

5.

Hahn DW, Allen GO, McGuire JL. Norgestimate: the progestational component of a new combination oral contraceptive. Pharmacologist 1976; 18: 250.

6.

Hahn DW, Lai MT. The possible mechanism of antiovulatory action of norgestimate (ORF 10131). Fed Proc 1979; 38: 1253.

7.

Chapdelaine A, Desmarais J-L, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int J Fert 1989; 34: 347-352.

8.

Hahn DW, Foldesy RG, McGuire JL, Anderson F, Murphy RJ. Influence of norgestimate and levonorgestrel on SHGB. Arch Gynecol 1985; 237: 332.

9.

McGinty DA, Anderson CP, McCullough NB. Effect of local application of progesterone on the rabbit uterus. Endocrinology 1939; 24: 829-832.

10.

Finney DJ: Assays based on quanta1 responses. In: Finney DJ, methods in biological assay. 3rd ed. London: ed. Statistical Charles Griffin & Co. Ltd, 1978: 370.

APRIL 1990 VOL. 41 NO. 4

409

CONTRACEPTION 11.

Killinger J, Hahn DW, Phillips A, Hetyei NS, McGuire JL. Ths affinity of norgestimate for uterine progestogen receptors and its direct action on the uterus. Contraception 1985; 32: 311-319.

12.

Kloosterboer HJ, Vonk-Noordegraaf CA, Turpijn EW. Selectivity in progesterone and androgen receptor binding of progestagens used in oral contraceptives. Contraception 1988; 38: 325-332.

13.

Spona J, Huber J. Pharmacological and endocrine profiles of gestodene. Int J Fertil 1987; 32 (Suppl): 6-14.

14.

Bergink EW, Hamburger AD, de Jager E, van der Vies J. Binding of a contraceptive progestogen ORG 2969 and its metabolites to receptor proteins and human sex hormone binding globulin. J Steroid Biochem 1981; 14: 175-183.

15.

Spona J. Androgenic properties of progestagens used in oral contraceptives. In: Rolland R, ed. Advances in fertilitycontrol and treatment of sterility. Boston: MTP Press Ltd, 1984: 89-98.

16.

Dusterberg B, Humpel M, Speck U. Terminal half-lives in plasma and bioavailability of norethisterone, levonorgestrel, cyproterone acetate and gestodene in rats, beagles and rhesus monkeys. Contraception 1981; 24: 673-683.

17.

Alton KB, Hetyei NS, Shaw C, Patrick JE. Biotransformation of norgestimate in women. Contraception 1984; 29: 19-29.

410

APRIL 1990 VOL. 41 NO. 4