Comparative progestational activity of norgestimate, levonorgestrel-oxime and levonorgestrel in the rat and binding of these compounds to the progesterone receptor

Comparative Progestational Activity of Norgestimate, Levonorgestrel-Oxime and Levonorgestrel in the Rat and Binding of these Compounds to the Progesterone Receptor W. Kuhnz, K.H. Fritzemeier, C. Hegele-Hartung,

and R. Krattenmacher

The progestational activity of norgestimate (NORG), levonorgestrel-oxime (LNG-oxime) and levonorgestrel (LNG) were compared in a pregnancy maintenance study in rats. The compounds were administered subcutaneously to pregnant rats at several doses, blood samples were collected repeatedly, and the concentration of LNG was measured in these samples. It could be demonstrated that following the administration of NORG and LNG-oxime, LNG was a major metabolite present in the serum. The pharmacological response in rats treated with NORG and LNGoxime could be related to the systemic exposure of these animals to metabolically derived LNG. Thus, both NORG and LNG-oxime can be regarded as pro-drugs of LNG, the latter being almost exclusively responsible for the pharmacological activity of both pro-drugs. This notion was further supported by studies on the comparative binding affinity of these compounds to rabbit and human progesterone receptor (PR). LNG exhibited the highest binding affinity of the Icompounds studied. Relative binding affinity (RBA) values of LNG using progesterone as reference (100%) were found to be 125% for rabbit PR (rPR), 143% for human uterine PR (hPR) and 125% for recombinant hPR, respectively. In contrast to LNG, NORG exhibited only a low affinity to the PR, which is documented by RBA values of 1.2% for rPR, 3.2% for uterine hPR and 9% for recombinant hPR. The corresponding values of LNGoxime were 30% (rPR), 20% (uterine hPR) and 18% (recombinant hPR), respectively. Thus, the combined experimental evidence of the present study does not support the view of NORG being a progestogen on its own as has been suggested by others. CONTRACEPTION 1995;51:131-139

Introduction

pharmacokinetics, levonorgestrel, norgestilevonorgestrel-oxime, bioavailability, pregnancy maintenance, rat, receptor binding

KEY

WORDS:

mate,

Research

Laboratories,

fur Pharmakokinetik

Schering

Aktiengesellschaft

Schering

0 1995 Elsevier Science Inc. 655 Avenue of the Americas, New York,

NY 10010

13342 Berlin,

Germany

orgestimate (NORG) is a progestogen which is metabolized in vivo after oral administration to animals and humans13 and also in vitro4j5 to several metabolites. Levonorgestrel (LNG) has been identified as one of the major metabolites which was present in the serum of rats and humans.6,7 It is still an open question whether NORG is a new progestogen on its own. Alternatively, the observed pharmacological activity in vivo might be

N

caused predominantly

by the metabolite

LNG and, if

that was the case, NORG was simply a pro-drug of LNG, without offering additional advantages over a treatment with LNG itself. Recently, levonorgestrel oxime (LNG-oxime), another metabolite of NORG, has been claimed to contribute significantly to the pharmacological activity. lsr9 This metabolite, however, might also be just a precursor of LNG (Figure 1). The aim of the present study was to compare the progestational activity of the three progestogens, LNG, NORG and LNG-oxime, in a pregnancymaintenance test with castrated pregnant rats. For that purpose, the ability of LNG, NORG and LNGoxime to maintain an established pregnancy was examined at several doses. The results were correlated with the concentrations of LNG measured in the serum of the animals. In addition, the relative binding affinities of the three compounds to the progesterone receptor were examined. This study was aimed to supplement a previously performed, similar comparative investigation which, however, did not include either LNG-oxime or receptor binding data.6

Material and Methods Steroids

[3H]-Progesterone (specific activity: 1.92 TBq/mmol/ 1) was obtained from Arnersham Buchler (Braunschweig, Germany). Progesterone, cortisol, levonorgestrel (17a-ethinyl-13-ethyl-17P-hydroxy-4,15ISSN OOlO-7824/95/$960 SSDI 001 O-7824(94)0001 9-S

132

Kuhnz

et al.

Contraception 1995:51:131-139

OH

Norgestimate

Levonorgestrel-3-oxime

Figure

1. Proposed metabolic

conversions

of NORG.

Levonorgestrel-I

7-acetate

Levonorgestrel

gonadien-3-one) and all the other steroids investigated were synthesized by Schering AG (Berlin, Germany). Due to the presence of an oxime group in position 3, norgestimate exists as an equilibrium mixture of two geometric isomers (E and Z). The mixture ((E/Z)- 17J$acetoxy- 17a-ethynyl18-methyl-4estren-3-one oxime), as well as the isolated isomers of norgestimate (Z-17p-acetoxy-17a-ethynyl-18methyl-4-estren-3-one oxime) and (E- 17@acetoxy17a-ethinyl- 18-methyl-4-estren-3-one oxime) were included in the present study. In addition, the following metabolites of norgestimate were synthesized: levonorgestrel-3-oxime (E/Z-l 7a-ethynyl17J3hydroxy- 18-methyl-4-estren-3-one oxime), levonorgestrel-acetate ( 17J3-acetoxy-17a-ethynyl18-methyl-4estren-3-one). Buffers The following buffer solutions were used: buffer 1: 0.01 mol/l TRIS-buffer (pH 7.5), containing 1.5 mmol/l EDTA and 10% (v/v) glycerol; buffer 2: 0.01 mol/l TRIS-buffer (pH 7.5), containing 0.02 mol/l molybdate, 0.002 mol/l dithiothreitol (DTT) and 10% (v/v) glycerol; buffer 3: 0.02 mmol/l TRIS-buffer (pH 7.5), 0.5 mmol/l EDTA, 2 mmol/l DTT, 20% (v/v) glycerol, 20 mmol/l molybdate, 0.3 mmol/l phenylmethanesulfonic acid fluoride (PMSF), 0.3 mmol/l aprotinin, 1 mol/l pepstatin, 10 mol/l leupeptin. Dextran-coated charcoal (DCC) was prepared from 0.5% Norit A and 0.05% dextran T400 in buffer 1. Preparation of Progesterone Receptor-Containing Cytosol For the preparation of progesterone receptors, rabbit uteri were isolated from mature rabbits treated for 7

days with 10 p,g estradiol valerate/animal/day, and human uterine tissue was obtained from a local hospital. All the following steps were carried out in the cold room at 4°C. One g of tissue was homogenized with 1 ml of buffer 2 using an Ultra-Turrax (Janke und Kunkel, GERMANY). The homogenate was centrifuged at 105,OOOxg for 90 min, and the supernatant was used as progesterone receptor-containing cytosol. Using the recombinant baculovirus BV-hPR (provided by Prof. P. Chambon, Strasbourgh, France), human progesterone receptor was produced in SF9 insect cells. Cells were harvested 48 h after infection with BV-hPR and centrifuged for 10 min at 150 g. The resultant cell pellet was resuspended in buffer 3 (equal volumes of buffer and cell pellet). The cell suspension was treated by repeated thawing and freezing (at least 4 times), and was afterwards transferred to an ice bath. The homogenization was done by repeated suction and expelling with a pipette, and afterwards with a cannula (Braun Sterican, 0,45 x 23 m, 26 g X 7/8). The suspension was then centrifuged as described above. The cytosol fraction was used in competition experiments. Competition Experiments Rabbit and human uterine cytosol were pre-incubated with cortisol by addition of 10 ~1 of cortisol ( = 10e4 mol/l) to 0.4 ml of cytosol and incubation for 15 min at 4°C. Cytosol obtained from insect cells was used without further treatment. Ten ~1 of [3H]progesterone (final concentration 5 x 10 - 9 mol/l) and 10 ~1 unlabeled progesterone for the standard curve or 10 ~1 of the test substance in appropriate concentrations were added to 20 ~1 of cytosol. The steroids were dissolved in ethanol such that the final concentration did not exceed 3%. Each sample was incu-

Contraception 1995x51:131-139

bated for 120 min at 4°C. After incubation, unbound steroids were adsorbed by incubating with 0.5 ml of a suspension of dextran-coated charcoal in buffer 1 for 10 min at 4°C. After centrifugation for 5 min at 15,OOOxg, an aliquot of the supematant was withdrawn and counted for radioactivity. For determination of the relative affinity of compounds, the displacement of [3H]-progesterone with unlabeled compound was plotted as percent binding versus log molar concentration. The activity of a compound was expressed as relative binding affinity (RBA) which is defined as the IC50 ligandlIC50 test compound (IC = inhibitory concentration). Progestational Activity: Pregnancy Maintenance in Rats The studies were performed in female rats (Wistar Han), 20&220 g body weight. The animals were kept in temperature controlled rooms (20 + 2°C) with an artificial light (14 h)-dark (10 h) rhythm. The animals had free access to food (AltrominR-pellets, Altromin Ltd., Lage, FRG) and tap water. In the first study (study I), the rats were randomly allocated to 14 groups. Except for the non-castrated vehicle controls (n = 12), each group comprised 6 animals, divided in two blocks (A and B) with 3 animals (controls: n= 6) per block. Rats in pro-estrus were mated overnight. Mating was confirmed in the morning of the following day (estrus) by vaginal smear inspection. The presence of sperm was regarded as the beginning of pregnancy (day 1 p.c.). On day 8 p.c., animals were ovariectomized in ether narcosis. Two hours prior to castration, as well as on the subsequent 13 days, the animals received subcutaneously different doses of either LNG (0.01, 0.03, 0.1 and 0.3 mg/ animal/day), NORG or LNG-oxime (each 0.1, 0.3, 1 .O and 3.0 mg/animal/day) or vehicle. The daily dose of each compound was administered subcutaneously in a total volume of 0.2 ml of castor oil/benzyl benzoate (4/l). In addition, all animals, except the noncastrated vehicle controls, received an injection (s.c.) of 1 pg estrone/day in a volume of 0.1 ml of the same vehicle. One day after the last treatment (day 22 of pregnancy), the animals were sacrificed and the number of living fetuses per animal was counted. Pregnancy maintenance was calculated as the number of living fetuses divided by the total number of implantation sites and is given in percent. The absence of implantation sites (castrated controls) was defined as 0 % pregnancy main enance. Blood samples (ca. .5 ml) were taken from each of the animals on day 8 k rior to drug administration and at the following time! points post administration: day 9 (1 h), day 10 (2 h), day 12 (4 h), day 14 (6 h), day 16 (8 h) and day 18 (24 h).

Progestational

Activity

of Norgestimate

133

Blood samples were centrifuged, the serum was separated, and equal aliquots of each sample obtained at corresponding times post administration were pooled within each dose group. Per dose group, seven serum pools were thus obtained. The pooled samples were kept deep-frozen at - 18°C until analysis. LNG was determined in the serum of the animals belonging to treatment groups 1 to 12. In a second experiment (study II), ovariectomized rats received subcutaneously different doses of either LNG (0.003 0.01, 0.03 and 0.1 mg/animal/day), NORG (0.03,0.1,0.3 and 1.0 mg/animal/day) or LNGoxime (0.01, 0.03, 0.1 and 0.3 mg/animal/day) two hours prior to castration, as well as on the subsequent 7 days. Each group comprised 6 animals except for the non-castrated control group which comprised 12 animals. No blood samples were collected during this part of the study for the analysis of LNG. Evaluation of the results was the same as in the first study. Determination of LNG in the Serum by GUMS The analytical determination of LNG in serum samples obtained during study I was performed by means of a GC/MS method (LAB, Neu-Ulm, Germany). Briefly, to 0.25 ml of the serum sample, the internal standard (2H,-NET) was added and subsequently, the sample was extracted with 5 ml of cyclohexane/2butanol (98.5: 1.5; v/v). The organic layer was transferred into a new vial and stored refrigerated until analysis. Prior to analysis, the extract was dried at 40-50°C under a stream of nitrogen and the residue was reconstituted in 20 ~1 of ethyl acetate. An aliquot of this solution was injected into a GC/MS system (TRIO 1000, Fisons Instr.) operated in the positive chemical ionization mode with selected ion monitoring. The samples were analyzed in three batches. Each batch consisted of rat serum samples, one zero standard, one set of calibration standards ranging from 0.1 rig/ml to 100 rig/ml and 9 quality control samples (3 sets of: 16.0, 4.0 and 0.8 rig/ml). For the preparation of the standard curve, a methanolic solution of LNG (100 pg/ml) was diluted with a pool of blank serum to yield final concentrations between 0.1 and 100 rig/ml. The lower limit of quantification (LLQ) was 0.1 rig/ml. The recovery was between 95 and 99%. Individual batches were accepted when at least six non-zero standards remained in the curve and additionally at least 75% of the quality control samples were within a range of 15% of their nominal concentrations. Pharmacokinetic Evaluation The concentrations of LNG measured in the serum samples obtained from the animals of each treatment

134

Kuhnz et al.

Contraception 1995;51:131-139

group were used to calculate the area under the serum level-time curve AUC(O-24 h) according to the linear trapezoidal rule. The systemic availability (fr) of norgestimate-derived LNG relative to an administered dose of LNG was calculated according to: AUC(O-24 f, = AUC(O-24 AUC(O-24 AUC(O-24

h)N x DLNG h)LNo X DN

h) of NORG- or hlN = AUC(S24 LNG-oxime-derived LNG h)rNc D LNG

= AUC(O-24 h) of LNG after the administration LNG =

dose of LNG administered

DN = dose of NORG- or LNG-oximederived LNG (corrected for the molecular weight of NORG or LNG-oxime, respectively) which could be released at most from the dose of NORG or LNG-oxime administered.

Results Pregnancy Maintenance

1. Efficacy of NORG and LNG to maintain pregnancy, when administered subcutaneously over a period of 14 days to ovariectomized rats (study I)

Table

Group Number 1 2 3 4 5 6 7 8 9 10 11 12 Controls Crzols C.C.

DW LNG LNG LNG LNG

Number of Animals

NORG NORG NORG LNG-oxime LNG-oxime LNG-oxime LNG-oxime

: 3 3 3 3 3 3 3 3 3

0.01 0.03 0.1 0.3 0.1 0.3 1.0 3.0 0.1 0.3 1.0 3.0

Vehicle

6

-

100

Vehicle

3

-

0

NORG

3

Dose [mg/day]

Degree of Pregnancy Maintenance WI 0 37.5 92.9 40.0 72.7 81.8 91.7 91.7 92.9 100 100 0

Groups 1 to 12 received an additional administration of 1 pg estronelday. Controls included castrated and intact animals. Pregnancy maintenance is presented as median values. n.c. = non-castrated controls; C.C. = castrated controls.

Test

Study I In block B, a massive virus infection with the Kilham rat virus (KRV) occurred. It is known that KRV can disturb fetal development via a transplacental infection, leading to an increased number of uterine resorption sites. Therefore, the evaluation of pregnancy maintenance was only possible for the animals of block A, which were free of KRV. The results are presented in Table 1. At doses between 0.01 and 0.1 mg/animal/day, LNG showed a dose-dependent increase in pregnancy maintenance up to 92.9%. At a dose of 0.3 mg/animal/day, however, a dramatic decrease of pregnancy maintenance to a value of 40% was observed (Figure 2). For NORG, a pregnancy maintaining effect of 73% to 92% was achieved at doses between 0.1 and 3 mg/ animal/day. There was no difference in the pharmacologic response between the two highest doses of 1 and 3 mg (Figure 2). A similar pregnancy maintaining effect was seen in animals which received LNG-oxime at doses between 0.1 and 1 .O mg/animal/day. With the highest dose of 3.0 mg/animal/day, however, no pregnancy maintenance could be achieved (Figure 2). Obviously, even with the lowest doses of NORG and LNG-oxime administered, an almost complete pregnancy maintenance was achieved. Thus, at least in the case of LNG-oxime and to a somewhat lesser extent with NORG, there was almost no discrimina-

tion between the lowest and the highest dose group in the pharmacological efficacy. Therefore, a second study was performed where several lower doses of the progestogens were included in order to establish a clear dose-response relationship (see below). Determination of LNG in Pooled Serum Samples The LNG concentrations that were determined in the pooled serum samples of rats belonging to treatment groups 1-12 of study I are presented in Table 2. In animals receiving subcutaneous administrations of LNG, the serum levels of LNG reached maximum values about 1 h post dose and decreased thereafter. In those animals that were treated either with NORG or with LNG-oxime, serum levels of LNG increased up to about 8 h post administration and remained almost unchanged or decreased slightly thereafter up to 24 h post dose. A linear relation was found between the dose of LNG administered (groups l-4) and the AUC(O-24 h) of LNG. The same was true for the dose of NORG administered (groups 5-8) and the AUC(O24 h) of LNG calculated. A linear relation was also observed between the dose of LNG-oxime administered (groups 9-12) and the AUC(O-24 h) of LNG calculated. The mean bioavailability of LNG derived from NORG and LNG-oxime over the dose range of 0.1-3.0 mg was 14.4 + 1.4% and 22.8 + 2.8%, respectively.

Progestational

Contraception 1995;51:131-139

120 loo

i

Pregnancy , i.c. x

maintenance

0.001

,,‘,,,”

1 ~~~,,,,~

0.01

0.1

,Y,‘,,,,, 1

Pregnancy I

maintenence

Receptor

[%]

1

10

Studies

Discussion

Dose [mg/day]

0.001

Binding

The relative binding affinities of NORG and some of its metabolites to the rabbit and the human progesterone receptor are presented in Table 4. The strongest relative binding affinity was observed for LNG, followed by LNG-acetate and LNG-oxime. The affinities of the two separated isomers of NORG to hPR were also determined and were found to be similar to those of the mixture of the two isomers. The E-isomer exhibited a slightly higher affinity than the Z-isomer. 0.1

AUC(rS24h) ,

135

10

~0% b-Wdayl

loo

of Norgestimate

did not result in an increased pharmacological response (Table 3). For LNG-oxime, a dose-dependent increase in the pregnancy maintenance was observed between 0.01 and 0.3 mg/animal/day [Table 3 and Figure 2). Thus, the lowest doses of LNG, NORG and LNGoxime that caused effects comparable to the pregnancy maintenance in non-castrated vehicle controls were 0.03, 0.3 and 0.1 mg/animal/day, respectively.

[%] I

cc. I*:‘*-,-

0,

Activity

of LNG [ngxWml]

0.01

I

0.1

1

10

Dose [mg/day] c +-LNG

+NORG

+LNG

oxime’

2. Pregnancy maintaining potential of NORG, LNG and LNG-oxime when administered subcutaneously to ovariectomized pregnant rats together with 1.0 pg estrone over a treatment period of 14 days (top, study I) or 7 days (middle, study II). Pregnancy maintenance relative to the intact (i.c.) and the castrated (c.c.) controls is presented together with the corresponding LNG levels (bottom, study I) in the serum of the treated animals. Figure

In the present study, we have examined the progestational activity of the progestogen NORG and two of its metabolites, LNG and LNG-oxime, using the pregnancy maintenance test in the rat as an established animal model. It has been stated that only NORG itself and LNG-oxime contributed significantly to the pharmacological activity of NORG both in animals and humans.‘!8#9 It has further been claimed that both NORG and LNG-oxime shared the same pharmacological characteristics.’ Recently, however, we have been able to show that LNG, which is a potent and well known progestogen, is present to a considerable extent in the serum of rats and also in the serum of women who have been treated with NORG.6~7 These studies suggested that NORG can indeed be regarded as a pro-drug of LNG, although it could not be excluded that NORG itself and/or LNG-oxime might also contribute to the pharmacological activity. Since NORG seems to be only a minor component in the serum of women who received oral doses of NORG, whereas LNG-oxime has been found to reach about 40 times higher

Study II

There was a dose-dependent increase in the pregnancy maintenance following subcutaneous administration of LNG over a dose range of 0.01 to 0.1 mg/ animal/day (Table ‘3). At the two lowest doses administered (0.003 and 0.01 mg/animal/day), no pregnancy maintenance could be achieved. For NORG, a dose-response relation was observed within a dose range of 0.03 and 0.3 mg/animal/day, whereas a further increase in the dose up to 1 .O mg/animal/day

concentrations,’

it seemed manda-

tory to examine in particular the role of this metabolite as another possible precursor of LNG. To this end, the concentration of LNG was measured in the serum of pregnant rats which were treated with different doses of LNG, NORG and LNG-oxime. The dose-dependent exposure of the animals to LNG, expressed by the corresponding AUC value, was related to the observed biological activity, indicated by the achieved degree of pregnancy maintenance.

136

Kuhnz

et al.

Contraception 1995;51:131-139

Table 2. Concentration of LNG (rig/ml) in pooled serum samples of rats that were treated (s.c.) with either LNG (groups l-4), NORG (groups 5-8) or LNG-oxime (groups 9-12) at different doses; the calculated AUC (O-24 h) values of LNG are also presented Group

Time [day/h1

1

8/O 9/l

0 0.41

10/2 1214 14/6 1618 18124

0.29 0.27 0.31 0.15 0

3.4

15.9

AUC (O-24 h] [ng x ml-’ Values

below

x

h]

the lower

limit

2

3

0

0

1.63

4.07

1.0

2.9

0.66 0.74 0.79 0.36

1.96 2.35 2.83 1.53

of quantification

54.8 (0.1 @ml)

4

5

0 13.78 8.77 6.75 8.04 9.07 5.54 182.5

Group Number 1

Drug

12

LNG LNG LNG LNG NORG NORG NORG NORG LNG-oxime LNG-oxime LNG-oxime LNG-oxime

Controls n.c.

Vehicle

32 4 i 87 9

10 11

5 5 5 5 6 5 6 5 6 6 6 6 12

Dose [mg/day]

0.03

83.3

0.1

91.7

0.03

0 25.0 88.5 83.3

12

0

n.e.

0

0

0

0.91 1.44 2.58 2.43 4.03 2.07

2.66 4.80 8.67 7.85 11.34 7.90

0.41 0.52 0.64 0.57 0.52 0.33

0.94 1.32 1.74 1.35 1.76 1.92

6.57 5.50 5.86 6.51 6.57 5.04

0

0.3

16.7 84.7 92.0

-

92.9

0.1

11

0.25 0.42 0.51 0.79 1.29 0.72

WI 0

0.03

10

0

Degree of Pregnancy Maintenance

0

1.0 0.01

9

0

7.7

0.003

0.3

8

20.9

65.9

208.2

10.9

40.3

0

139.0

14 13 14 18 15 15 363

were set to zero; me. = not evaluable.

0.01

0.1

7

0.21 0 0.33 0.37 0.42 0.29

Table 3. Efficacy of NORG and LNG to maintain pregnancy, when administered subcutaneously over a period of 7 days to ovariectomized rats (study II) Number of Animals

6

Groups 1 to 12 received an additional administration of 1 kg estrone/day. Non-castrated animals were included as controls. Pregnancy maintenance is presented as median values. n.c. = non-castrated controls.

Not only following treatment with LNG, but also after administration of NORG and LNG-oxime, LNG was present in the serum of rats. Within each dose group, a linear relation was observed between the administered dose of the progestogen and the AUC of LNG. Almost twice as much LNG was found in the serum after treatment with LNG-oxime as compared to equivalent doses of NORG. This seems plausible, since additional metabolic pathways can contribute to the biotransformation of NORG, whereas not all of these metabolic steps may be of importance for LNGoxime. In the rat, a relevant fraction of the dose of NORG and LNG-oxime administered is obviously metabolized to LNG. This is also in good agreement

with our previous findings.6 time

profiles

were observed

Different for LNG

serum level vs. following

the

administration of LNG on the one hand and the administration of NORG and LNG-oxime on the other hand. Following the administration of LNG, maximum drug levels were reached already one hour post dose, whereas after the administration of NORG and LNG-oxime, maximum serum levels of LNG were only reached at about 8 h post dose. Since it can be assumed that the cleavage of the acetate group in NORG by ubiquitous esterases is a rapid process as compared to the metabolism of the oxime group, the latter conversion probably represents the ratelimiting step in the formation of LNG. This would not only explain the late appearance of the maximum serum levels of LNG after the administration of NORG and LNG-oxime, but also the fact that there was no difference in the serum level vs. time profiles of LNG after the administration of these two progestogens. Up to a dose of 0.1 mg, the increase in AUC following the administration of LNG was accompanied by a concomitant increase in the pregnancy maintenance rate, which reached almost 100%. At the highest dose of 0.3 mg LNG, pregnancy maintenance was impaired and dropped to values of about

40%.

Similarly,

at doses of 0.1 to 3 mg of

NORG and LNG-oxime, respectively, the metabolically derived LNG increased with the dose and a concomitant increase in the pregnancy maintenance rate could

also be seen, except for the highest

dose of

LNG-oxime, where no pregnancy maintenance could be achieved. The reduction in the pregnancy maintenance potential of progestogens at high doses can be rationalized, bearing in mind that during normal pregnancy there is a balance in the estrogen/progesterone

ratio that stays within

a narrow range. In the experi-

mental model, a constant dose of 1 kg estrone was administered each day together with doses of the

Progestational Activity of Norgestimate

Contraception 1995;51:131-139

137

Table 4. Relative binding affinities (RBA) of NORG and some of its metabolites to the progesterone receptor; sources for receptor isolation were rabbit uterus, human uterus and SF9 insect cells expressing human progesterone receptor Progesterone Receptor RBA [%] Species Tissue Compound

Progesterone NORG

LNG-3-oxime LNG-acetate E-isomer of NORG Z-isomer of NORG LNG

Rabbit Uterus 100 1.2 30 34 <2 <2 125

Rabbit Uterus Phillips et a1.9 100 124 94 521 n.d.

n.d. 541

Human Uterus

Human Uterus Juchem et al.”

100

100

3.2 20 76.9 2.5 1.2 143

Human Recombinant Receptor 100

9 3F.5 625

18 91

n.d.

n.d.

n.d. 375

n.d. 125

The results of the present study are compared with corresponding data published by others. The RBA values are expressed in % relative to the reference compound progesterone which was set to 100%; n.d. = not determined. Data from Juchem et al.” were re-calculated assuming the RBA of progesterone to be 40% relative to R5020, which was used as reference by these authors.

progestogens, covering a range of two orders of magnitude which eventually shifted this ratio extremely towards the progestogen. It has been shown before in similar experiments that this imbalance can negatively affect the outcome of pregnancy.” Unfortunately, even the lowest doses of both NORG and LNG-oxime chosen in this experiment resulted already in about 70 to 90% pregnancy maintenance. Therefore, a differentiation to the efficacy of higher doses was not feasible, although there was a correspondence between the difference in the amounts of LNG metabolically derived from NORG and LNG-oxime, respectively, and the differences in the pharmacological response observed after the administration of both progestogens. Since the interpretation of the pharmacological effects was somewhat limited due to the fact that the doses administered were too high altogether to establish a clear dose-response relationship, the animal experiment was repeated and the doses were adjusted accordingly. This time, the selected dose range proved to be adequate to obtain complete dose-response curves for each of th.e three progestogens examined. The curves were almost parallel and it could be shown that an about equal efficacy in the pregnancy maintenance test was achieved at doses of NORG and LNG-oxime which were about lo- and 3-times higher, respectively, than the corresponding dose of LNG. This difference in the progestational activity between the three progestogens was in excellent agreement with the difference in the exposure of the animals to LNG, w ich had been determined before in study I. About 14/o 7 and 23% of the administered doses of NORG andLNG-oxime, respectively, had been metabolically ‘converted to LNG. Thus, the pharmacological response observed in the NORGand LNG-oxime-treated rats can be quantitatively re-

lated to the exposure of the animals to LNG. These results strongly suggest that not only NORG but also LNG-oxime are basically only pro-drugs of LNG and that the pharmacological response is mainly, if not totally, caused by LNG. This notion is further supported if one looks at the relative binding affinity of NORG and its metabolites to the progesterone receptor (PR). We determined the relative binding affinities (RBA) of NORG and its metabolites LNG, LNG-oxime and levonorgestrel acetate (LNG-acetate) in comparison to the reference progesterone using PR-containing cytosol preparations from rabbit and human uterus. Additionally, biotechnologically produced human PR was included in the study. LNG exhibited the highest binding affinity of the compounds studied. RBA values were found to be 125% for rabbit PR (rPR), 143% and 125% for human progesterone receptor (hPR) isolated from uterus and transfected insect cells, respectively. In contrast to LNG, NORG exhibited only low affinity to rPR, uterine hPR and recombinant hPR which is documented by the RBA values 1.2%, 3.2% and 9%, respectively. Our results obtained with the uterine hPR are in good agreement with those published by Juchem et al.,” when their data were re-calculated using progesterone instead of the synthetic progestin R5020 as reference. The apparent differences in the RBA values observed for uterine and recombinant hPR in our study could be due to differences in the protein composition of cytosol prepared from human uterus and SF9 insect cells, respectively. We were, however, unable to reproduce data published by Phillips et a1.,9 who found similar binding affinities of NORG (124%) and progesterone (100%) to rPR. The reason for the difference in the RBA values of NORG of two orders of magnitude as compared to our data cannot be ex-

138

Kuhnz et al.

plained by differences in the experimental design, such as the incubation time, which was 2 h in our study as compared to 5 h in the study of Phillips et al. However, one possible explanation for the high binding affinity observed for NORG could be either the conversion of NORG to a metabolite such as LNG, with high receptor affinity, or even a contamination of the parent drug with a metabolite. We found LNG-oxime to exhibit moderate affinity to rabbit and human PRs, with RBA values ranging from 18% to 30% as compared to progesterone. Again, our data are in agreement with the study of Juchem et al.” as regards hPR. With respect to the affinity to rPR, our results differ by a factor of 3 from the data presented by Phillips et a1.9 Again, this difference cannot be attributed to differences in assay conditions but again raises the question of purity or metabolic stability of the compound under the assay conditions used by these authors. Our experiments revealed a considerable affinity of LNG-acetate to rPR and hPR, respectively. The RBA values were determined to be 34% for rPR and 77% and 91% for uterine and recombinant hPR, respectively. According to Juchem et al.,” LNG-acetate exhibited a higher affinity than even LNG. This finding could not be confirmed by our studies using hPR from different sources. The RBA value of LNG-acetate for rPR reported by Phillips et a1.9 is similar to the RBA of LNG determined in the same study, but is again tenfold higher compared to the corresponding value obtained in our study. The reason for this discrepancy is unknown, but might be attributed to the instabilility of this compound. NORG has been suggested to exhibit progestogenic activity in vivo on its own.‘~8~‘2-‘8 The results of the present study on pregnancy maintenance in rats revealed an about lo-fold higher potency of LNG as compared to NORG, when the drugs were administered subcutaneously. Assuming an about 14% conversion of NORG to LNG during this experiment, and considering the 45 to loo-fold lower binding affinity of NORG to the PR as determined for hPR and rPR, respectively, the lo-fold lower progestogenic potency of NORG can be fully ascribed to its metabolite, LNG, whereas only a marginal contribution of the parent drug is to be expected. LNG-oxime too, was assumed to contribute to the progestogenic activity of NORG.’ In the present study, LNG-oxime was found to be 3- to lo-fold less potent in rats than LNG. The affinity of LNG-oxime to the PR was 3- to s-fold lower as compared to LNG in studies with rPR and hPR, respectively. Assuming a 23% conversion of LNG-oxime to LNG in rats, equivalent progestogenic effects of the pro-drug observed at about 3-fold higher doses of the LNG-oxime

Contraception 1995;51:131-139

compared to LNG can almost fully be attributed to LNG, whereas a contribution of LNG-oxime as such is most likely only marginal. In conclusion, the present study demonstrated that in the rat, both NORG and LNG-oxime can be regarded as pro-drugs of LNG. Furthermore, it could be shown that the progestogenic activity of NORG and LNG-oxime can be directly related to the exposure of the animals to metabolically derived LNG. Neither of the two pro-drugs seem to contribute significantly to the pharmacologic activity. This was further supported by the results of binding studies to human and rabbit progesterone receptor, where LNG-oxime exhibited only a moderate and NORG hardly any affinity. Thus, the experimental evidence does not support the view of NORG being a progestogen on its own as has been suggested by others.

Acknowledgments We wish to express our thanks to Mrs. B. Brandt and Mr. L.-Q. Cam for their excellent assistance in the performance of the animal experiments. We are also grateful to Dr. T. Petri for the expression of BV-hPR in SF9 insect cells and to S. Walter and I. Fuchs for technical assistance.

References 1. McGuire JL, Phillips A, Hahn DW, Tolman EL, Flor S,

Kafrissen ME. Pharmacologic and pharmacokinetic characteristics of norgestimate and its metabolites. Am J Obstet Gynecol 1990; 163: 2127-31. 2. AIton KB, Hetyei NS, Shaw C, Patrick JE. Biotransfor3.

4. 5.

6. 7.

8.

9.

mation of norgestimate in women. Contraception 1984; 29: 19-29. Sisenwine SF, Liu AL, Kimmel HB, Ruelius HW. The conversion of d-norgestrel-3-oxime-17-acetate to d-norgestrel in female rhesus monkeys. Contraception 1977; 15: 25-37. Madden S, Back DJ. Metabolism of norgestimate by human gastrointestinal mucosa and liver microsomes in vitro. J Steroid Biochem Molec Biol 1991; 38: 497-503. Wild MJ, Rudland PS, Back DJ. Metabolism of the oral contraceptive steroids ethynylestradiol and norgestimate by normal (HUMA 7) and malignant (MCF-7 and ZR-75-1) human breast cells in culture. J Steroid Biothem Molec Biol 1991; 39: 53543. Kuhnz W, Beier S. Comparative progestational and androgenic activity of norgestimate and levonorgestrel in the rat. Contraception 1994; 49: 275-89. Kuhnz W, Blode H, Mahler M. Systemic availability of levonorgestrel after single oral administration of a norgestimate-containing combination oral contraceptive to 12 young women. Contraception 1994; 49: 255-63. Anderson F. Selectivity and minimal androgenicity of norgestimate in monophasic and triphasic oral contraceptives. Acta Obstet Gynecol Stand Suppl 1992; 156: 15-21. Phillips A, Demarest K, Hahn DW, Wong F, McGuire JL. Progestational and androgenic binding affinities and

Contraception 1995:51:131-139

10.

11.

12. 13.

in vitro activities of norgestimate and other progestins. Contraception 1990; 41: 399-410. Neumann F, Elger W. Critical considerations of the biological basis of toxicity studies with steroid (sex) hormones. In: Plotz EJ, Hazier J, eds. Methods in Steroid Toxicology. Los Altos, California: Geron-X, Insurance, 1972: 10-91. Juchem M, Pollow K, Elger W, Hoffmann G, Mobus V. Receptor binding of norgestimate-a new orally active synthetic progestational compound. Contraception 1993; 47: 283-94. Hahn DW, Allen GO, McGuire JL. The pharmacological profile of norgestimate, a new orally active progestin. Contraception 1977; 16: 541-53. Phillips A, Hahn DW, Klimek S, McGuire JL. A comparison of the potencies and activities of progestogens used in contraceptives. Contraception 1987; 36: 18192.

Progestational Activity of Norgestimate

139

14. Chapdelaine A, Desmarais JL, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int J Fertil 1989; 334: 347-52. 15. Becker H. Supportive European data on a new oral contraceptive containing norgestimate. Acta Obstet Gyneco1 Stand Suppl 1990; 1.52: 33-9. 16. Kafrissen ME. A norgestimate-containing oral contraceptive: review of clinical studies. Am J Obstet Gyne~01 1992; 167; 119~202. 17. London RS, Chapdelaine A, Upmalis D, Olson W, Smith J. Comparative contraceptive efficacy and mechanism of action of the norgestimate-containing triphasic oral contraceptive. Acta Obstet Gynecol Stand S~ppl 19923 156: 9-14. 18. Phillips A, Hahn DW, McGuire JL. Preclinical evaluation of norgestimate, a progestin with minimal androgenie activity. Am J Obstet Gynecol 1992; 167: 119196.