Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The decreased production of ovarian steroids, which occurs after the climacteric, has been linked to a number of postmenopausal pathologies, in particular osteoporosis and coronary heart disease (Lindsay, 1988; Ross et al., 1990). Estrogen replacement therapy is effective in reducing the risks associated with these pathologies (Cumming, 1991). However, there are several undesirable side effects associated with chronic estrogen therapy that create difficulties in compliance. In this respect, estrogens, when administered without progestin, substantially increase the incidence but not the mortality of endometrial cancer (Ziel and Finkle, 1975; Vesey, 1984); furthermore, concerns about the increased risk of breast cancer associated with estrogen replacement therapy have been raised (Cauley et al., 1999; Jacobs, 2000). Accordingly, an ideal therapy should prevent bone loss and improve the serum lipid profile as estrogen does, without inducing proliferate effects on reproductive tissues. It is this need that has provided the impetus to the pharmaceutical industry to search for new estrogens that may modulate estrogen receptors in a tissue-selective manner.

Tamoxifen was the first estrogen agonist/antagonist shown to inhibit bone loss in castrate female rats (Jordan et al., 1987) and in postmenopausal women (Love et al., 1994). However, its estrogen-like effect on human endometrial carcinoma (Satyaswaroop et al., 1984; Kedar et al., 1994) and its strong ephatocarcinogenic effects in rats (Greaves et al., 1993) prevented the chronic use of tamoxifen to mimic some of the helpful actions of estrogen after the menopause. Subsequent endeavors have seen the first approval of a selective estrogen receptor modulator (SERM) for the prevention and treatment of osteoporosis (raloxifene; Black et al., 1994) and the emergence of several new synthetic compounds, including triphenylethylene, naphthalene, and benzopyrans derivatives, with this possible spectrum of activities (Sato et al., 1999). However, the development of many of these compounds as drugs is problematic because of their excessive stimulation of uterine tissue (Sato et al., 1996; Grese et al., 1997).

With this in mind, we wanted to design a series of novel SERMs demonstrating significant protective effects on nonreproductive tissue (bone and cholesterol levels) and antagonistic effects associated with a low degree of intrinsic estrogenicity on reproductive tissue. A compound characterized by this pharmacological profile may have potential utility in the prevention and treatment of a number of postmenopausal pathologies such as osteoporosis, coronary heart disease, and estrogen-dependent human cancer.

Our chemical plan involved surveying the various structural classes known to interact with the estrogen receptor, including in particular the benzopyrans exemplified by levormeloxifene, which is an estrogen agonist/antagonist that has been shown to inhibit bone loss in postmenopausal women (Bjarnason et al., 1997) and in the ovariectomized rat model (Novak et al., 1997). However, the partial estrogenic effects on uteri in women raised issues of safety of levormeloxifene as a compound for use in the prevention and treatment of postmenopausal women. Its development as antiosteoporotic drug was discontinued when the compound was in advanced clinical studies (Mitlak and Cohen, 1999).

We performed a wide variety of synthetic modifications of the benzopyran moiety at various sites to obtain a SERM characterized by a lower estrogen agonist effect on reproductive tissue as compared with levormeloxifene. Several candidate compounds were prepared and tested for the following activities: 1) binding affinity to human ER- and -; 2) antiestrogenic and estrogenic effect on uterine growth in an immature female rat model, which is widely accepted as the model for studying in vivo agonist and antagonist estrogenic effects (Eppenberger et al., 1991); 3) estrogenic effects on bone, total serum cholesterol, and uterus in an OVX rat model of postmenopausal bone loss (Kalu, 1991). As a result of our screening, we have discovered a new, orally active, potent estrogen agonist/antagonist named CHF 4056 (Fig. 1) with potential advantages over estrogen in the uterus. Herein, the pharmacological properties of CHF 4056 observed in the above-mentioned experimental models are reported. In addition, its in vivo effects are compared with those of EE2 and levormeloxifene.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Chemical structure for CHF 4056. CHF 4056 is a new benzopyran derivative that is structurally related to the well known SERM levormeloxifene.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

All animals studies were performed in strict accordance with the decreto Legislativo sulla sperimentazione animale (Italian law on rules for animal experimentation, Decree 116, January 27, 1992) and the "European Directive for the protection of vertebrate animals used for experimental and other scientific purposes" (European Union Directive 86/606/CEE).

Chemicals. CHF 4056 (3-phenyl-4-[[4-[2-(1-piperidinyl)ethoxy]phenyl]methyl]-2H-1-benzopyran-7-ol), M_r = 441.59; levormeloxifene (1-[2-[4-[(3R,4R)-3,4-dihydro-7-methoxy-2,2-dimethyl-3-phenyl-2H-1-benzopyran-4-yl]phenoxy]ethyl]-pyrrolidine hydrochloride), M_r = 494.08; and raloxifene ([6-hydroxy-2-(4-hydroxyphenyl)benzo[b] thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]-methanone hydrochloride), M_r = 510.06, were synthesized by Chiesi Farmaceutici S.p.A. (Parma, Italy). EE2, M_r = 296.4; 17-estradiol, M_r = 272.4; and diethylstilbestrol, M_r = 268.4 were obtained from Sigma (St. Louis, MO). Kits for radioimmunoassay of osteocalcin were supplied by Biomedical Technologies Inc. (Stoughton, MA). All other reagents were purchased from Sigma.

Human ER- and ER- Binding. ER- and - binding analysis was performed as previously described (Obourn et al., 1993). Briefly, the standard assay was performed in a volume of 100 µl, containing a final concentration of 0.5 nM ³H-estradiol (New England Nuclear, Boston, MA), increasing concentration of unlabeled CHF 4056 or reference compounds (0.01-100 nM), 5 µl of diluted (1:100 in binding buffer) human recombinant ER- or - (insect Sf9 cells), and 95 µl of binding buffer (10 mM Tris, pH 7.5, 10% glycerol, 1 mM dithiothreitol, and 1 mg/ml bovine serum albumin). The incubation was carried out at room temperature for 3 h. After incubation, 100 µl of 50% hydroxylapatite slurry (equilibrated in 50 mM Tris, pH 7.4, and 1 mM EDTA) was added to each tube and vortexed three times over 15 min. One milliliter of wash buffer (40 mM Tris, pH 7.4, 1 mM EGTA, 1 mM EDTA, and 100 mM KCl) was added to each reaction, the reactions were centrifuged at 10,000g for 5 min, and the supernatant was aspirated. The wash step was repeated two additional times, and then the hydroxyapatite pellet was resuspended in 400 µl of ethanol, transferred to a scintillation vial, and counted. Nonspecific binding was defined as that which occurred in the presence of 1 µM diethylstilbestrol and represented 10 to 15% of the total binding. K_i values were calculated using the equation of Cheng and Prusoff (1973) using the observed half-maximal inhibition concentration of the tested compound, the concentration of radioligand employed in the assay, and the dissociation constant value of the ligand. The data were also fitted by an iterative program (RECEPT) for nonlinear regression analysis (Benfenati and Guardabasso, 1984) both to a one-site and to a two-site model. The one-site model was then chosen when it yielded the best correlation coefficient and when the improvement of goodness-of-fit for the two-site model was not statistically significant (P < 0.05) according to the F test on the sums of squared errors.

Immature Female Rat Study. Twenty-one-day-old female Sprague-Dawley rats, weighing approximately 40 to 50 g (Charles River, Calco Italy) were treated by oral gavage with either vehicle (0.5% methylcellulose, 3 ml/kg), CHF 4056 (0.01-10 mg/kg·day), levormeloxifene (0.01-10 mg/kg·day), or EE2 at 0.05 mg/kg·day for 3 days. The compounds under investigation were also administered 15 min before the EE2 gavage, used as estrogenic stimulus to increase uterine weight. Nonestrogenic controls were given vehicle alone.

Four-Day OVX Rat Study. Virgin Sprague-Dawley rats (90 days old) were obtained from Harlan Nossan (Correzzana, Italy) and group housed on a 12-h light/dark cycle. The animals had ad libitum access to both food and tap water. Animals were randomized into experimental treatment groups, with six animals per treatment group. Compound administration was initiated 14 days after ovariectomy, to insure clearance of endogenous estrogen and to allow for acclimatization to the home cage. Compounds were dissolved in 0.5% methylcellulose and given by daily oral gavage in a volume of 3 ml/kg of body weight. Animals were dosed for 4 consecutive days and fasted in the evening after the final dose. On the next morning the animals were sacrificed by exsanguination from the abdominal aorta under anesthesia with ketamine and xilazine. The uteri were collected and weighed. Blood samples were allowed to clot at 4°C for 2 h and then centrifuged at 2000g for 10 min. Serum samples were collected and stored at 80°C; serum cholesterol was assayed using a high performance colorimetric assay (Roche, Mannheim, Germany). One horn of the uterus was removed, weighed, and transferred into a Tris buffer for analysis of eosinophil uterine peroxidase activity (see below).

Four-Week OVX Rat Study. Nine- to 10-month-old virgin Sprague-Dawley rats, weighing approximately 280 to 300 g (Harlan Nossan) were used in this study. The animals were acclimatized to the local vivarium conditions (22 ± 2°C; 12-h light/dark cycle) for 2 weeks and housed individually during the experimental period.

Bone Densitometry. Bone mineral density (BMD) was measured by dual energy X-ray absorptiometry (DEXA) using a Hologic QDR-1000 plus instrument equipped with dedicated software for small animal measurements. An ultrahigh-resolution mode (0.0254-cm line spacing and 0.0127-cm resolution) was used with a collimator of 0.63-mm diameter. This technique provides an integrated measure of both cortical and trabecular bone.

Uterine Histology. Formalin-fixed uteri were processed for conventional paraffin embedding. Sections of about 5-µm thickness were obtained from each block. Slides were stained with hematoxylin and eosin before undergoing image analysis for the measurement of endometrium epithelia and myometrial thickness. The measurements were performed using an Ibas20 computerized imaging system (Kontron/Zeiss, Welwyn Garden City, UK) run on a Kontron 386 personal computer. The images were acquired with a JVC (Yokohama, Japan) black and white camera fitted with a 50-mm macro lens (for myometrial thickness) or an Axioscope microscope (for endometrium epithelia). A black and white camera was used as it is more sensitive than a color one.

Uterine Eosinophil Peroxidase Activity. The test protocol was based on the method described by White et al. (1991). The assay is based on the oxidation of o-phenylenediamine by uterine eosinophil peroxidase in the presence of hydrogen peroxide (H₂O₂).

Statistical Analysis. Results are expressed as mean ± S.E.M. Significance was determined by analysis of variance and, when analysis of variance was significant, by the Newman-Keuls test for posthoc multiple comparisons. Probability values of < 0.05 were considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Human ER- and ER- Binding Effects. CHF 4056 binds with high affinity to the human ER- and ER-. Binding K_i were 0.041 ± 0.011 and 0.157 ± 0.028 nM, respectively (Tab 1). This competitive binding assay showed that CHF 4056 competes for a single binding site on both ER- and ER-. For comparison, K_i values for some SERMs, 17 estradiol, and diethylstilbestrol were evaluated and reported in Table 1.

View this table: [in this window] [in a new window]   TABLE 1 Human ER- and ER- binding affinity Competitive binding analysis of CHF 4056 and a series of reference compounds for ER- and ER-. Binding analysis was performed as detailed under Materials and Methods. Each value is the mean ± S.E.M. of two experiments.

Immature Female Rat Assay. In immature female rats the ovaries do not produce estradiol; however, the uterus is fully responsive to exogenous estrogen and hence this model permits a ready measure of an agonist or antagonist action. In these rats, treatment with EE2 at 0.05 mg/kg p.o. for 3 days significantly increased uterine wet weight (150-200%) compared with vehicle-treated controls. This concentration of EE2 was the lowest producing near-maximal effect and was chosen on the basis of preliminary dose-response experiments.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A, Effect of CHF 4056 () and levormeloxifene () on EE2-induced uterine weight increase in immature rats. B, Effect of CHF 4056 (), levormeloxifene () and EE2 () on uterine weight in immature rats. The estrogenic and antiestrogenic activity was expressed as uterine weight/body weight ratios. Each point is the mean ± S.E.M. of one to three experiments. For each experiment n = 8. Statistical significance relative either to EE2-treated control (A) or vehicle-treated control (B) is denoted by * (P < 0.05) and ** (P < 0.01).

Four-Day OVX Rat Assay. Differential tissue selectivities of CHF 4056 have been demonstrated in OVX rats treated for 4 days with endpoints of serum cholesterol lowering and uterine stimulation (uterine weight and uterine peroxidase activity).

Four-Week OVX Rat Assay. CHF 4056 effects were evaluated in 9- to 10-month-old OVX rats that were dosed for 4 weeks postsurgery and compared with OVX and sham controls.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of CHF 4056 (), levormeloxifene (), and EE2 () on bone mineral density of lumbar vertebrae L1-4 (A) and serum osteocalcin (B) in OVX rats treated for 4 weeks. Bone mineral density values are reported as percent change in BMD from the baseline. Each point is the mean ± S.E.M. of two experiments. Statistical significance relative to vehicle-treated OVX rats is denoted by * (P < 0.05) an ** (P < 0.01). For each experiment n = 7-9.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effect of CHF 4056 (), levormeloxifene (), and EE2 () on serum cholesterol levels in OVX rats treated for 4 weeks. Each point is the mean ± S.E.M. of two experiments. Statistical significance relative to vehicle-treated OVX rats is denoted by * (P < 0.05) and ** (P < 0.01). For each experiment n = 7-9.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effect of CHF 4056 (), levormeloxifene (), and EE2 () on uterine weight (uterine weight/body weight ratio) (A) and uterine peroxidase activity (B) in OVX rats treated for 4 weeks. Uterine peroxidase activity is reported as the mean Vmax. Each point is the mean ± S.E.M. of two experiments. Statistical significance relative to vehicle-treated OVX rats is denoted by ** (P < 0.01). For each experiment n = 7-9.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Effect of CHF 4056 (), levormeloxifene (), and EE2 () on endometrium ephitelia thickness (A) and myometrial thickness (B) in OVX rats treated for 4 weeks. For each parameter the data were expressed as pixel ± S.E.M. of two experiments. Statistical significance relative to vehicle-treated OVX rats is denoted by ** (P < 0.01). For each experiment n = 7-9.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Estrogens can inhibit bone resorption and consequently they can be used for the treatment and prevention of postmenopausal osteoporosis; in this respect, it is well established that long-term hormone replacement therapy reduced fractures rate in postmenopausal women (Lindsay et al., 1980; Weiss et al., 1980). Moreover, estrogens have beneficial activities in the cardiovascular system (Zumoff, 1993; Grodstein et al., 1997) and may also have beneficial effects in the central nervous system, as several studies have linked them to improvements in cognitive function and in delaying the onset of Alzheimer's disease (Grady et al., 1992).

Despite the overall health benefits of estrogens, relatively few women actually considered hormone replacement therapy or the therapy was discontinued within a year (Lobo, 1995). This reduced compliance with therapy is because of the potential link between estrogen replacement therapy and breast/endometrial cancer and some other side effects including breakthrough bleeding (Jacobs, 2000; Evans and Turner, 1995). Thus, a therapeutic agent that has estrogen-agonist effects on the skeleton and cardiovascular system and estrogen-antagonist activities on the uterus and breast would be highly desirable for postmenopausal women.

CHF 4056 is a new nonsteroidal estrogen agonist/antagonist that binds with high affinity to the human ER- and ER- (K_i were 0.041 and 0.157 nM, respectively). Compared with raloxifenethe only SERM currently approved for the prevention and treatment of postmenopausal osteoporosis with possible additional protective effects on breast cancer, cardiovascular diseases, and cognitive functionsthe ER- affinity of CHF 4056 was similar, whereas the affinity for ER- was 10-fold higher (Tab 1). It appears quite clear today that ER- has biological roles that are distinct from those of ER- (Gustafsson, 1999). Knock-out mice deficient in ER- show a distinct phenotype when compared with that of ER- / mice. In this respect, because it appears that ER- ligands could prove to be useful therapeutically in targeting the neuroprotective (Kuiper et al., 1998) and cardioprotective actions of oestrogens (Mahela et al., 1999), the higher affinity for this receptor may qualify CHF 4056 as a SERM with more potential on brain and/or cardiovascular system than raloxifene.

However, it should be noted that in light of recent studies the action of estrogens on cardiovascular system is still a matter of debate (Clemett and Spencer, 2000) and that to date we have no experimental evidence that CHF 4056 could have a different pharmacological profile compared with raloxifene either in the central nervous system (e.g., cognitive function, hot flushes) or in the cardiovascular apparatus.

In immature rats, the compound antagonized estrogen stimulation of the uterus down to the level of vehicle-treated controls with no estrogenicity. In the same experimental conditions, the structurally related compound levormeloxifene only partially blocked estrogen-induced uterine weight gain (~ 40% inhibition) because of its evident agonist activity (~65% that of EE2). Thus, although CHF 4056 and levormeloxifene share the same benzopyran core, the pharmacological profile of CHF 4056 in immature rats appears strictly different. Studies are in progress to clarify the structural features of CHF 4056 that are relevant in reducing the uterine-stimulating effects associated with levormeloxifene in immature rats. Preliminary data indicate the importance of the 4-methylenic hinge between the aminoalchoxyphenyl side chain and the benzopyran moiety of CHF 4056, the deletion of which results in increases in estrogenic activity (data not shown).

Differential tissue selectivities of CHF 4056 have been studied in OVX rats treated for 4 days/weeks with endpoints of body weight, serum cholesterol lowering, bone tissue, and uterine stimulation. Ovariectomy resulted in significant osteopenic responses after 4 weeks in the lumbar spine L1-4 as measured by DEXA densitometric techniques. CHF 4056 was able to reduce the decreases in BMD of L1-4 lumbar vertebrae at doses of 0.1 to 1 mg/kg·day; maximal efficacy (50% protection at 1 mg/kg·day) was indistinguishable from that of 0.1 mg/kg·day EE2. Moreover, the OVX-induced serum osteocalcin gain was completely prevented by treatment with 1 mg/kg·day CHF 4056. These results showed that CHF 4056 maintained lumbar spine BMD and bone turnover at levels comparable with those seen with EE2. Whereas the results of these studies show that CHF 4056 will provide protection against OVX-induced bone loss after 4 weeks, a longer term study will be performed to show that these effects will be maintained and that CHF 4056 are not simply delaying the eventual loss of BMD due to estrogen deficiency.

EE2 produced a marked hypocholesterolemic effect in OVX rats; this effect is attributed to up-regulation of hepatic LDL receptors, resulting in enhanced clearance of circulating LDL (Brown and Goldstein, 1980). Under the same experimental conditions, CHF 4056 significantly decreased total serum cholesterol in a dose-dependent manner, indicating that the compound acts as an estrogen receptor agonist on serum cholesterol in rat models of postmenopausal osteoporosis. This inhibitory effect on serum cholesterol levels was similar after 4 days and 4 weeks of treatment (Table 2; Fig. 4), indicating that no tolerance occurs toward CHF 4056's estrogen agonist effect on cholesterol metabolism.

In line with the results observed in the immature rats assay, also in OVX rats CHF 4056 has minimal stimulatory effects on the uterus. In fact, after 4 weeks of treatment, histological analysis of the uterine tissue showed that CHF 4056 has nonsignificant effects on endometrium epithelia thickness. In contrast, EE2 and levormeloxifene increased epithelial thickness (120 and 124%, respectively), demonstrating significant uterine hypertrophic effects. This trend was reproduced in analysis of uterine eosinophil peroxidase activity to show that CHF 4056 is less stimulatory in the uterus than estrogen or levormeloxifene. CHF 4056 caused, after 4 weeks of treatment, a statistically significant increase in uterine weight relative to the OVX controls, although it was much less pronounced than that observed in EE2-treated animals. However, this marginal effect on uterine weight was coupled with the lack of a stimulatory activity on endometrium epithelia and uterine peroxidase, indicating that it may not be clinically relevant. In favor of this consideration, CHF 4056's profile on uterine tissue in OVX rats is superimposable to the one previously observed for raloxifene (Black et al., 1994), which, in clinical studies with postmenopausal women, did not show stimulatory effects on the uterus (Delmas et al., 1997). In contrast, the clinical development of levormeloxifene, which significantly affected the epithelia thickness (Fig. 6A) and the uterine eosinophil peroxidase activity (Fig. 6B), was discontinued after reports of endometrial thickening side effects in postmenopausal women. Thus, CHF 4056 is expected to have a better tissue selectivity profile in postmenopausal women compared with the structurally related compound levormeloxifene. The fact that this tissue selectivity is not simply the result of selective tissue distribution is confirmed by the ability of CHF 4056 to completely antagonize the effects of estrogen on the immature rat uterus.

In conclusion, the new benzopyran derivative CHF 4056 has promise as an agent with beneficial bone and cardiovascular effects. Our data showed that CHF 4056, EE2, and levormeloxifene have overlapping patterns of effects on bone, cholesterol, and adipose tissue but distinct activities on uteri. The minimal uterine stimulation suggests a potential therapeutic advantage of CHF 4056 over EE2. Moreover CHF 4056, being characterized by marked estrogen antagonist activity, may be of interest in the prevention and treatment of estrogen-dependent human tumors. Concerning this, investigations in in vitro assays (MCF-7 proliferation) and in animal model of postmenopausal breast cancer with CHF 4056 are ongoing.