Journal of Biochemistry & Molecular Biology 96 (2005) 201–215

Effects of , Premarin and Acolbifene on histomorphology and sex steroid receptors in the rat vagina L. Berger, M. El-Alfy, C. Martel, F. Labrie ∗

Oncology and Molecular Endocrinology Research Center, Laval University Medical Center (CHUL), 2705 Laurier Boulevard, Quebec City, Que., Canada G1V 4G2 Received 14 June 2004; accepted 4 February 2005

Abstract

To assess the specific estrogenic and/or androgenic effects of a potential novel hormone replacement therapy, we have examined the morphology of the rat vagina 9 months after ovariectomy (OVX) and treatment of OVX animals with dehydroepiandrosterone (DHEA), conjugated Premarin and the selective receptor modulator Acolbifene. OVX led to atrophy and inflammatory changes while Acolbifene reduced the inflammation incidence and induced mucification of the vaginal epithelium. Premarin induced a typical keratinized stratified squamous epithelium while DHEA induced stimulation of the vaginal epithelium, with mucous cells typical of an androgenic effect, combined with increased collagen fiber compactness of the lamina propria. On the other hand, after OVX, the vaginal muscle layer decreased by 46%, an effect which was 41 and 100% reversed by DHEA and Premarin, respectively. The present data show particularly interesting effects of DHEA on the three layers of the vaginal wall, namely a highly mucified epithelium, an increased muscularis thickness and increased collagen fiber compactness in the lamina propria. DHEA exerts both androgenic and estrogenic effects on the vaginal mucosa, thus providing a more physiological replacement therapy. © 2005 Elsevier Ltd. All rights reserved.

Keywords: receptor; DHEA; ; SERM; Acolbifene; EM-652; ; Hormone replacement therapy; Intracrinology; Menopause; Progesterone receptor; Vagina; Sexual dysfunction

1. Introduction There is an increasing interest in the potential of combined estrogen–androgen replacement therapy [8,9], although the Vaginal dryness affects about 50% of postmenopausal use of the estrogenic component is limited by the potential women at the age of 50–60 years and 72% after 70 years [1]. negative impact on and cardiovascular events Of these women, about 80% experience urogenital disorders, [10–12]. Based upon recent advances in our understanding of especially vaginitis and dyspareunia [2]. Since these prob- human sex steroid physiology, especially in postmenopausal lems are believed to be at least partially related to the depri- women [3,5], the use of DHEA becomes, in this case, a pos- vation of sex , appropriate local hormonal replace- sibility to provide with the appropriate levels of ment therapy should be considered. In fact, postmenopausal and estrogens synthesized in specific tissues by intracrine women do not only lack all ovarian estrogens, but they are also mechanisms, while avoiding systemic effects [5,13–15]. progressively deprived of the androgens originating from the The selective estrogen receptor modulator (SERM) Acolb- peripheral intracrine transformation of dehydroepiandros- ifene (EM-652) is a benzopyran derivative originally devel- terone (DHEA) into both androgens and estrogens [3–5].In oped for the prevention and treatment of breast cancer [16]. fact, serum DHEA and DHEA-S progressively decrease from Acolbifene is the compound having the highest affinity of all the age of 30–50 years [5–7]. known compounds for the estrogen receptor (ER) [16–18]. This compound displays a pure and highly potent antie- ∗ Corresponding author. Tel.: +1 418 654 2704; fax: +1 418 654 2735. strogenic activity in the mammary gland and endometrium E-mail address: [email protected] (F. Labrie). while decreasing serum cholesterol and triglycerides and

0960-0760/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsbmb.2005.02.018 202 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 preventing bone loss, at least in the rat [17]. On the other (9) OVX + DHEA + Acolbifene + Premarin. On the first day hand, the inhibitory effect of DHEA on the growth of human of the study, the animals of all groups (except group one) were breast cancer xenografts in nude mice provides further sup- bilaterally ovariectomized under isoflurane-induced anesthe- port for its use in hormone replacement therapy [19,20].In sia. Premarin and Acolbifene were administered by oral fact, combined treatment of DHEA and Acolbifene has been gavage (0.5 mL/rat) as suspensions in 0.4% methylcellulose proposed as a beneficial chemopreventive and therapeutic while DHEA was solubilized in 50% ethanol–50% propylene approach in breast cancer [14]. It is possible that the combi- glycol and was topically applied (0.5 mL/rat) on a shaved area nation of DHEA, Acolbifene and possibly an estrogen could of 2 cm × 2 cm of the dorsal skin. Acolbifene was synthe- provide optimal benefits for women at menopause. Recently, sized in the medicinal chemistry division of our laboratory, we found that the high potency of Acolbifene completely as described [16]. The compound had a purity of 99.2%. blocks the stimulatory effect of (E2) on the mam- DHEA was obtained from Scheweizerhall Inc. at a purity mary gland and uterus in the rat and could thus avoid the risk of 100% while Premarin was a product of Wyeth Pharma- of breast and uterine cancer [21]. ceuticals for intravenous human use containing 24.8 mg/vial In the present study, the ovariectomized (OVX) rat model (claimed for 25 mg) of total of which was used to examine the effects of DHEA, Premarin and 54.4% were sodium sulfate and 28.7% sodium equi- Acolbifene, alone or in combination, on the morphology and lin sulfate. At the DHEA administered dose, the DHEA blood on the distribution of ER␣, PR and AR after 9 months of level ranged between 70 and 100 nmol/L [7, and unpublished treatment. Since the rat adrenal does not secrete DHEA or data]. Dosage selection for Premarin corresponds to the min- DHEA-S [3,22], the percutaneous administration of DHEA imal dose sufficient to reverse OVX-induced uterine atrophy, in OVX animals, while preventing first pass of the orally while Acolbifene was administered at a dose sufficient to administered steroid through the liver, is the only source of cause uterine atrophy similar to OVX after its administration sex steroids in the model used, thus facilitating the interpre- to Premarin-treated OVX animals. Treatments were initiated tation of the data. As an estrogen, Premarin was chosen in on day 2 of the study and the compounds were administered this protocol because it is the most commonly used estro- once daily for 36 weeks. Animals from the intact and OVX gen preparation in North America as hormone replacement control groups received the vehicle alone by oral gavage and therapy. In a previous study performed in our laboratory, the topical application. morphology observed following treatment of OVX rats with Twenty-four hours after the last dosing, overnight fasted 17␤-estradiol was identical to that seen with the use of Pre- animals were sacrificed under isoflurane anesthesia by exsan- marin. guination at the abdominal aorta (nine animals per group) or by intracardiac perfusion with 10% neutral buffered formalin (five animals per group). Vaginae from non-perfused animals 2. Materials and methods were collected and weighed, while the vaginae collected from perfused animals were marked with black ink on the ventral 2.1. Animals and treatments side and then trimmed as described below.

Ten to 12-week-old female Sprague–Dawley rats (Crl: 2.2. Histological procedures CD®(SD)Br VAF/PlusTM) (Charles River Laboratory, St- Constant, Canada) weighing approximately 220–270 g at The entire vagina of each perfused animal was postfixed in start of the experiment were used. The animals were 10% neutral buffered formalin. Each vagina was then divided acclimatized to the environmental conditions (temperature: into seven equal cross-segments as illustrated in Fig. 1, rou- 22 ± 3 ◦C; humidity: 50 ± 20%; 12-h light/12-h dark cycles, tinely processed and embedded all together in the same paraf- lights on at 07:15 h) for at least 1 week before starting the fin block. Within the paraffin block, the seven vaginal cylin- experiment. The animals were housed individually and were drical segments were positioned in a sequence corresponding allowed free access to water and rodent food (Lab Diet 5002, to their original anatomical position and oriented perpendic- Ralston Purina, St. Louis, MO). The experiment was con- ular to the surface of the block, thus allowing the segments to ducted in accordance with the CCAC Guide for Care and be cut in cross-sections. For each animal, a 4 ␮m-thick paraf- Use of Experimental Animals in an animal facility approved fin section was cut and stained with haematoxylin–eosin for by the Canadian Council on Animal Care (CCAC) and the morphological examination. Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). 2.3. Histomorphometry A total of 126 female rats were randomly distributed into 9 groups of 14 animals each as follows: (1) intact Measurements of the different vaginal layers were per- control; (2) ovariectomized control; (3) OVX + Acolbifene formed on the fifth segment (Fig. 1), which is approximately (2.5 mg/kg); (4) OVX + Premarin (0.5 mg/kg); (5) OVX + halfway between the middle region and the portio vaginalis Premarin + Acolbifene; (6) OVX + DHEA (80 mg/kg); (7) uteri (segment 7). This fifth segment was found to display OVX + DHEA + Acolbifene; (8) OVX + DHEA + Premarin; a representative epithelial surface and a sufficient thickness L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 203

ies, at 1:200, 1:250 and 1:250, respectively. After washing in PBS buffer, sections were incubated with biotinylated antirabbit secondary antibody for 10 min and thereafter with streptavidin–peroxidase for another 10 min (Zymed, CA). Diaminobenzidine was used as the chromogen to visual- ize the biotin/streptavidin–peroxidase complex, under micro- scope monitoring. Counterstaining was performed using #2 Gill’s hematoxylin for 30 s. For controls, immunoabsorption with an excess of the peptide used to raise the antibody, or substitution with non-immune rabbit IgG, was performed. Semi-quantitative evaluation of the number and intensity of immunostained nuclei was performed as indicated in Table 2.

2.5. Statistical analysis

Data are presented as means ± S.E.M. of eight to nine ani- mals per group for vaginal weight or five animals per group for vaginal layer thickness determinations. Statistical signif- icance was determined according to the multiple-range test of Duncan–Kramer [23].

3. Results

3.1. Morphology and thickness of the different layers of Fig. 1. Division along the longitudinal axis of the rat vagina into seven cross- segments, from the external orifice (ostium) (segment 1) to the cervix level rat vagina (segment 7) (portio vaginalis uteri) (modified from Popesko et al. [62]). Each segment is about 3–4 mm long. To examine with precision the three layers of the rat vaginal wall, namely the epithelium, lamina propria and mus- cularis, seven segments obtained along the longitudinal axis of smooth muscle. Images were captured with a DC-330 (Fig. 1) were first examined. While important morphologi- 3CCD color camera (Dage-MTI, Michigan City, IN, USA) cal differences were observed between the different groups, and quantified using Image-Pro Plus 3.0 software (Media in general, morphology was uniform in all animals of the Cybernetics, Silver Spring, MD, USA). Thus, using a 5× same group and segments 2–7 have shown similar epithelial objective (Leica Microsystems, Willowdale, Ont., Canada), histological features. The few exceptions observed will be three to four thickness measurements per layer were obtained mentioned later. Segment 5 was thus used to illustrate the from representative artifact free areas of the epithelium and effect of the various treatments on the vaginal epithelium. muscularis, as well as for the three vaginal layers together. The thickness of the lamina propria was obtained by subtract- 3.1.1. Epithelium ing the thickness of the epithelium and muscularis from total In intact animals at estrus, which represents the typ- vaginal thickness. ical estrogenic pattern, a keratinized stratified squamous epithelium was observed in all segments (Table 1; Fig. 2A). 2.4. Immunohistochemistry A similar epithelium characterized all segments of the OVX + Premarin group (Fig. 2E), with the exception of some Immunostaining was performed using Zymed SP kits areas of large mucous cells in segments 2–5 (Table 1), a reac- (San Francisco, CA). Paraffin sections (4 ␮m) were deparaf- tion seen following long-term administration of an estrogenic finized in toluene and rehydrated through ethanol. Endoge- compound [24]. On the other hand, cycling rats at proestrus, nous peroxidase activity was eliminated by preincubation which are under a mixed estrogenic-progestational influ- with 3% H2O2 in methanol for 30 min. A microwave ence, show the presence of epithelial mucification (Fig. 2B). retrieval technique using 0.01 M citrate buffer for 15 min This effect is illustrated by a stratified squamous epithe- was applied. After cooling the slides, non-specific bind- lium covered by layers of mucified cells lining segments 2–7 ing was blocked using 10% goat serum for 20 min. Sec- (Table 1). tions were then incubated for 1.5 h at room temperature In the OVX group, the absence of ovarian stimulation led with ER␣ (AB-1, Calbiochem, CA), AR (N-20, Santa Cruz to an atrophy of the vaginal epithelium, which characterized Biotechnology, CA) or PR (Ab-4, NeoMarkers, CA) antibod- all segments (Table 1; Fig. 2C). Thus, this poorly stratified 204 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215

Table 1 Histological evaluation of the epithelium, lamina propria and muscularis in the seven rat vaginal segments Group Vaginal Segment 1 Segment 2 Segment 3 Segment 4 Segment 5 Segment 6 Segment 7 walls Intact Estrus E KS KS KS KS KS KS KS LTa–Lb t–L t–L t–M M–M M–M M–M Ms s t MTTT Proestrus E KS MSM–KS MSM MSM MSM–KS MSM–KS MSM–KS L T–M t–M t–M t–M t–M M–M M–M Ms s t MMT T OVX Control E A A AAAAA L t–H t–M t–H t–H t–H M–M M–H Mss ttttM Acolbifene E KS SCM–A SCM–A SCM–A SCM–A SCM–A SCM–A L M–M t–M t–H t–H t–H t–H M–H Mss ssttt Premarin E KS KS–MSM KS–MSM KS–MSM KS–MSM KS KS L M–L t–M t –M M–M M–M M–H T–H Ms s t MTTT Premarin + Acolbifene E KS A–SCM SCM–A SCM–A SCM–A SCM–A SCM–A L M–M t–M t–H t –H M–M M–M M–M Mss ttttM DHEA E KS–LHM LHM LHM LHM LHM LHM–SCM LHM–SCM L M–H t–H t –H t–H t–H M–H M–H Ms s s t MMM DHEA + Acolbifene E KS–LHM–SCM SCM–LHM SCM–LHM SCM–LHM SCM–LHM SCM–LHM LHM–SCM L M–M t–M t –H t –H M–H M–H M–H Mss sttMM DHEA + Premarin E KS MSM–KS MSM–KS MSM–KS MSM–KS KS KS L M–M t–L t–M t–M M–H M–M M–H Ms s t MTTT DHEA + Premarin + Acolbifene E KS SCM –LHM–KS SCM–LHM SCM–LHM SCM–LHM SCM–LHM LHM–SCM L M–M M–H t–H M–H t–H M–H M–M Ms s s t MMM E = epithelium morphology: KS (keratinized stratified squamous), LHM (large hypertrophied mucous cells), SCM (small aligned columnar or cuboidal mucous cells), MSM (mixed stratified squamous overlaid by mucous cells), A (atrophy: small cuboidal cells). Two different abbreviations for a given segment indicate two different patterns and that the first one is the predominant. a L = lamina propria thickness: T (thick), MT (moderately thick), t (thin). b Compactness of collagen fibers: H (high), M (moderate), L (low), M (muscularis thickness), T (thick), MT (moderately thick), t (thin), s (scarce). epithelium consisted of one to two atrophied cuboidal or flat- 1 was thus higher than that of OVX rats, which comprised tened squamous cell layer(s) with some small mucous cell only four to six cell layers. In the other segments (Table 1), the areas, overlying a basal cell layer. Most importantly, moderate basal layer was covered by a layer of low columnar mucous inflammation with many foci of intraepithelial microabcesses cells (Fig. 2D), which were more developed than in OVX ani- (small areas of agglomerated leucocytes within the epithe- mals. In the OVX + Acolbifene group, only three out of five lium) was a frequent finding. These changes were often animals showed signs of minimal inflammation while mod- accompanied by focal erosion (zones where the epithelial erate inflammatory changes were general findings in all OVX layer partially disappears) and ulceration (complete disap- animals. When the combination of Premarin and Acolbifene pearance of the epithelial layer) (Fig. 3). was used, the morphology was superimposable to that found In OVX animals treated with Acolbifene, segment 1 in the OVX + Acolbifene group, namely the typical pattern of showed a keratinized stratified squamous epithelium simi- cylindrical mucous cells, more rarely seen in OVX animals lar to that of the intact group, except that it included 9–11 (Table 1; Fig. 2F). No inflammation occurred in the group of cell layers compared to the 10–15 layers found in animals at animals receiving both Premarin and Acolbifene. estrus or when OVX animals received Premarin (Fig. 4). The In all groups, segment 1 showed a comparable strat- epithelial thickness of Acolbifene-treated animals in segment ified squamous epithelium, except in the animals of the L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 205

Fig. 2. Vaginal epithelium histomorphology of segment 5 in the nine groups of rats. Bar in (J) 40 ␮m. (A) As representative of the estrogenic effect, the stratified squamous epithelium of cycling rats at estrus consists of four main layers: one cell layer of stratum basale (b), six to seven cell layers of stratum spinosum (s) and a stratum granulosum of five to six layers (g) overlaid by the tightly packed flattened cornified cells that form the stratum corneum (c). (B) Cycling rats at proestrus, which are under an estrogenic–progestational influence, are used to illustrate mucification. In most segments 2–7, one basal cell layer (b) was overlaid by five to six cell layers of stratum spinosum (s) and a stratum mucification (m) consisting of three to four layers of mucous cells. (C) In the OVX control, a basal cell layer (b) was overlaid by one to two layer(s) of atrophic cuboidal or flattened cells (a). (D) The vaginal epithelium of OVX rats treated with a daily oral dose of Acolbifene (2.5 mg/kg) shows atrophy, but with an outer layer of low columnar mucous cells (m) overlying the basal cell layer (b). (E) Vaginal epithelium of OVX rats treated with a daily oral dose of Premarin (0.5 mg/kg). The OVX-induced atrophy was replaced by an estrogenic pattern similar to that found at estrus. (F) In OVX animals, which received Premarin + Acolbifene, atrophy predominated with a morphology similar to that of Acolbifene-treated animals, although larger mucous cells were seen. (G) Following treatment of OVX animals with a once daily cutaneous application of DHEA (80 mg/kg) on an area of 2 cm × 2 cm of the dorsal skin, an hypertrophic epithelium which consisted of three to five layers of mucous cells (m) was seen overlying a basal layer (b). Several invaginations characterized this epithelium (arrow). (H) In most areas of the vaginal epithelium of DHEA + Acolbifene-treated animals, a layer of mucous cells (m) rested on a basal cell layer (b), while in some areas, many layers of mucous cells overlaid the basal cell layer. Several invaginations characterized this epithelium (arrows). (I) Treatment with DHEA + Premarin led in three animals to a mixed epithelium composed of three to seven cell layer-thick stratified squamous epithelium (s) overlaid by three to five layers of mucous cells (m). In other two animals, areas of stratified squamous epithelium were predominant (insert). Bar (in insert) 30 ␮m. (J) When DHEA, Premarin and Acolbifene were combined, the epithelium was similar to that of the DHEA + Acolbifene group. 206 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215

Fig. 3. Vaginal mucosa of OVX animals showing: (A) moderate inflammatory changes characterized by focal leukocyte infiltration with intraepithelial microabscess (IM) and (B) focal erosion (E) characterized by reduced epithelial thickness and ulceration (U) visualized as a complete disappearanceofthe epithelium. Bar in (B) 30 ␮m.

OVX group, where atrophy was a predominant char- the DHEA + Acolbifene group, thus indicating a blockade of acteristic (Fig. 4). Moreover, in the OVX + DHEA and the estrogen-induced squamous cell proliferation by Acolb- OVX + DHEA + Acolbifene groups, mucous cells frequently ifene (Table 1; Fig. 2J). accompanied the dominant stratified squamous epithelium of As illustrated in Fig. 5A, OVX markedly reduced the segment 1 (not shown). thickness of the vaginal epithelium (15 ± 1 ␮m) in compari- The seven segments of the vaginal epithelium of OVX son with that of the intact group (58 ± 5 ␮m). When animals animals which received DHEA were composed of large of the OVX group received Acolbifene, a similar low value multilayered columnar mucous cells with distended cyto- of 16 ± 1 ␮m was obtained. The epithelium thickness was plasmic vacuoles, a feature typical of an androgenic effect restored to 76 ± 4 ␮m in the group of animals which received [25,26] (Table 1; Fig. 2G). Several large invaginations char- Premarin, a value significantly higher than that of the intact acterized the epithelium after DHEA treatment. A simi- group, which included animals at different stages of the estrus lar epithelial morphology was found in all segments of cycle. When Acolbifene was added to Premarin, the epithelial the OVX + DHEA + Acolbifene group, except that the num- thickness was reduced to 17 ± 1 ␮m, thus showing a complete ber of cell layers decreased (Table 1; Fig. 2H). In the blockade of the effect of the estrogen. DHEA + Premarin-treated animals, a thick stratified epithe- The administration of DHEA to OVX animals increased lium of a “mixed” type composed of different ratios of squa- thickness of the vaginal epithelium to 38 ± 4 ␮m, while the mous epithelium covered by layers of mucous cells (Table 1; addition of Acolbifene to DHEA did not change the epithe- Fig. 2I) was observed from the second to the fifth segment, lial thickness which remained at 37 ± 5 ␮m. When Premarin thus revealing combined estrogenic and androgenic effects. and DHEA were co-administered to OVX animals, a high In this group, three animals displayed mucification over the thickness of 84 ± 6 ␮m was observed, a value not signif- above-mentioned segments, while in the other two animals, icantly different from that of the group of animals which a stratified squamous epithelium was observed in all seg- received Premarin alone. Finally, the combination of Pre- ments (insert Fig. 2I). When Premarin was combined with marin, DHEA and Acolbifene resulted in an epithelium DHEA and Acolbifene, the epithelium was similar to that of of 32 ± 3 ␮m-thick, a value similar to that of the groups,

Fig. 4. To indicate the estrogen-like effect of Acolbifene on the vaginal epithelium lining segment 1 (external opening), comparisons with relevant groups are illustrated. In intact rats at estrus (A) and in Premarin-treated (D) animals, the epithelium is 10–15 layer-thick and keratinized. In OVX controls (B), the epithelium is keratinized and its thickness is reduced to four to six layers and is generally not keratinized (see the absence of stratum granulosum). Meanwhile, in Acolbifene-treated animals (C), the thickness is restored to 9–11 layers and keratinized. Bar in (D) 30 ␮m. L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 207

Fig. 5. Thickness (␮m) of the three different vaginal compartments, as well as of the total rat vaginal wall, at the level of the fifth segment: (A) epithelium; (B) lamina propria; (C) muscularis; (D) total thickness, after 36 week-treatment of OVX animals with DHEA, Premarin and Acolbifene, alone or in combination. Measurements in intact animals at estrus, proestrus and diestrus are averaged and presented as reference controls. Groups sharing the same letter are not statistically different at p < 0.05. Dosages are described in legend of Fig. 2. which received DHEA alone or DHEA combined with in rats at estrus and in Premarin-treated OVX animals or Acolbifene. high in OVX and DHEA-treated OVX rats), while increas- ing in compactness in segments 2 and 3 to reach a plateau 3.1.2. Lamina propria that generally remains constant until segment 7. Thus, along The degree of compactness of collagen fibers in the vagina the longitudinal axis of the vagina in intact rats at estrus, was categorized as low, moderate and high (Fig. 6). In all the compactness of the collagen fibers was low in segments 1–3 animals, there were relatively few fibrocytes in proportion and moderate in segments 4–7. Atrophy was often associ- to the amount of collagen. Careful examination of each ani- ated with increased compactness of collagen fibers in the mal (Table 1) reveals that, in general, the compactness of OVX and OVX + Acolbifene groups (Table 1). In Premarin- the collagen fibers in segment 1 is moderate (except low treated OVX animals, compactness in segment 1 was low

Fig. 6. The compactness of collagen fibers of the lamina propria is illustrated in segment 5 as low (A), moderate (B) or high (C), as observed in the sub-epithelial lamina propria. The two former categories were associated with the presence of coarse collagen fibers loosely or less loosely aggregated together, respectively, and the latter was used to describe tightly packed fine collagen fibers, displaying a smooth, textured appearance. Bar in (C) 30 ␮m. 208 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 and moderate in segments 2 and 3, while in the other seg- Premarin and DHEA resulted in a notable thickness increase ments, compactness of the collagen fibers increased gradually (62 ± 3 ␮m), when compared to animals treated with DHEA to become high in segments 6 and 7. alone. When Acolbifene was added to Premarin and DHEA, In all segments, except the orifice area of the two estro- muscularis thickness decreased (46 ± 5 ␮g) to a value com- gen predominant groups as described earlier, compactness parable to DHEA and Acolbifene. of the collagen fibers could be described as moderate When total vaginal wall thickness was measured (Fig. 5D), in intact rats at proestrus, in OVX animals treated with the outer layer of connective tissue composing the adven- Premarin + Acolbifene or with Premarin + DHEA. On the titia was not included. OVX led to a marked (51%) vagi- other hand, most of the segments displayed highly com- nal wall atrophy (128 ± 3 ␮m versus 262 ± 39 ␮minthe pact collagen fibers in OVX animals treated with DHEA, intact) and the addition of Acolbifene to OVX had no effect DHEA + Acolbifene or DHEA + Premarin + Acolbifene. (108 ± 8 ␮m) (Fig. 7A–C). Premarin treatment maintained Light microscope examination of the lamina propria thick- the total thickness to a value (253 ± 10 ␮m) similar to that of ness along the vagina generally revealed that it was mod- intact animals while the addition of Acolbifene to Premarin erately thick in segment 1 while thinner in segments 2–4. reversed the effect of Premarin (150 ± 4 ␮m) (Fig. 7D and In segments 5–7, the thickness increased progressively to a E, respectively). Total vaginal thickness achieved by DHEA level similar to segment 1. The corresponding mean values treatment (184 ± 21 ␮m) was about 25% lower than that of of lamina propria thickness measured in segment 5 (Fig. 5B) the Premarin alone-treated group (Fig. 7F). On the other indicate that OVX led to a significant decrease in lamina pro- hand, the addition of Acolbifene to DHEA led to a non- pria thickness (76 ± 2 ␮m versus 135 ± 28 ␮m in the intact significant thickness increase (213 ± 20 ␮m), which became group) with a non-significant decrease induced by Acolbifene non-significantly different from the intact group (Fig. 7G). (60 ± 8 ␮m). The increase observed after Premarin or Pre- Finally, co-administration of Premarin and DHEA to OVX marin + Acolbifene administration in OVX animals remained animals markedly increased the thickness (290 ± 13 ␮m) to below the intact group (100 ± 9 and 90 ± 3 ␮m, respectively, a value similar to that of intact animals (Fig. 7H). The addition versus 135 ± 28 ␮m). The administration of DHEA led also of Acolbifene to DHEA and Premarin reversed the effect to a to a statistically non-significant increase in lamina propria value (176 ± 11 ␮m) not significatively different from DHEA thickness (96 ± 20 ␮m), and the addition of Acolbifene led alone (Fig. 7I). to further increased thickness (136 ± 17 ␮m), a value sim- ilar to that of intact animals. Treatment of OVX animals 3.2. Vaginal weight with Premarin + DHEA significantly increased the thickness (144 ± 14 ␮m) when compared to DHEA alone. Finally, After 9 months of treatment, the changes observed in the animals treated with Premarin + DHEA + Acolbifene dis- vaginal weight between the different groups (Fig. 8) gener- played a thickness (99 ± 9 ␮m) similar to that of the group ally follow the above-described morphological observations. treated with DHEA alone and lower than that of the Pre- Indeed, vaginal weight after OVX decreased by about 50% marin + DHEA group, although the difference was not statis- (101 ± 5 mg) compared to the intact group (205 ± 11 mg) tically significant. while treatment of OVX animals with Acolbifene alone had no effect on vaginal weight. On the other hand, adminis- 3.1.3. Muscularis tration of Premarin led to a vaginal weight increase that In all groups, only a few smooth muscle fibers of the did not reach the value of intact animals (170 ± 9 mg) muscularis were found in segment 1, along with few irregu- while the addition of Acolbifene to Premarin reversed the larly arranged striated muscle fibers. Gradually, the thickness estrogen-induced weight gain to a value similar to that of of the muscularis increased, remaining thin until segment OVX animals (96 ± 4 mg). Conversely, when DHEA was 5, where an oblique smooth muscle layer was seen above given to OVX animals, vaginal weight increased to a value the longitudinal fibers, thus resulting in a further thickness (171 ± 12 mg), similar to that of the Premarin-treated group, increase in segments 6 and 7 (Table 1). Measurement of mus- and the combination of Acolbifene and DHEA resulted in cularis thickness in segment 5 (Fig. 5C) revealed a value of a decrease in weight (135 ± 9 mg), which remained above 69 ± 7 ␮m in the intact group at estrus while a 46% signifi- that of the OVX group. Finally, co-administration of Pre- cant decrease was observed 9 months after OVX (37 ± 2 ␮m). marin and DHEA in OVX animals resulted in a vaginal Acolbifene treatment did not change the muscularis thick- weight gain (179 ± 10 mg) reaching a value similar to those of ness, which remained at 33 ± 3 ␮m. the OVX + DHEA and OVX + Premarin groups. The addition On the other hand, Premarin treatment of OVX animals of Acolbifene to this combination had no significant effect restored a thickness of 77 ± 5 ␮m, a value similar to that (194 ± 12 mg). of the intact group while Acolbifene added to Premarin reversed this effect, thus resulting in a thickness of 43 ± 2 ␮m. 3.3. Immunohistochemistry of steroid receptors DHEA induced a moderate increase in muscularis thickness (50 ± 2 ␮m), which was slightly decreased by the addition of Examination of the seven vaginal cross-segments from Acolbifene (41 ± 3 ␮m). Finally, the combined treatment of individual animals revealed that, in most groups, the intensity L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 209

Fig. 7. Microphotographs of the three vaginal compartments: (E) epithelium, (L) lamina propria and (M) muscularis, at the level of the fifth segment of the rat vagina, with emphasis on the relative muscularis thickness in the different groups. Separation of the three vaginal wall layers with bars is indicated to best estimate the thickness distribution between the different groups: (A) intact; (B) OVX and OVX treated with (C) Acolbifene; (D) Premarin; (E) Premarin + Acolbifene; (F) DHEA; (G) DHEA + Acolbifene; (H) DHEA + Premarin; (I) DHEA + Premarin + Acolbifene. Bar in (I) 50 ␮m.

of AR, ER␣ and PR immunolabeling in the lamina propria and muscularis generally increased from segment 1 to the plateau of segments 5–7. Since segment 5 was previously chosen to compare morphology, the immunostaining results were semi-quantitatively evaluated in this segment, as shown in Fig. 9 and summarized in Table 2.

3.3.1. Androgen receptor As seen in the upper row of Fig. 9 and summarized in Table 2, the vaginal wall of intact rats at estrus showed a moderate nuclear labeling in the epithelium, with the basal and the three suprabasal layers stained. A moderate number of strongly stained nuclei were seen in the lamina propria Fig. 8. Vaginal weight measured 36 weeks after OVX and treatment of OVX while a few weakly stained nuclei were observed in the mus- animals with Acolbifene, Premarin and DHEA alone or in combination. cularis. At diestrus, the basal and one suprabasal epithelial Intact animals are added as controls. layer intensely stained for AR and a large number of nuclei in 210 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215

Fig. 9. Immunohistochemical localization of AR, ER␣ and PR in the (E) epithelium, (L) lamina propria and (M) muscularis at the level of the fifth vaginal segment of the different groups. The thin bars indicate the approximate separation between the three vaginal compartments. Bar in lower row (right) 30 ␮m. L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 211

Table 2 Semi-quantitative evaluation of the number and intensity of immunostained nuclei for AR, ER alpha and PR in the epithelium, lamina propria and muscularis in the fifth segment of the rat vagina Group AR ER␣ PR

ELMELMELM Intact Estrus reference of low ER␣ and high PR 2 ++ 2 +++ 1 + 1 + 1 + 3 + 2 +++ 3 +++ 3 +++ Diestrus reference of high ER␣ and low PR 3 +++ 3 +++ 3 +++ 3 +++ 1 ++ 1 + 0 2 ++ 2 ++ OVX Control 1 ++ 1 +++ 1 +++ 2 ++ 1 +++ 1 +++ 0 0 0 Acolbifene 2 ++ 1 +++ 1 +++ 000000 Premarin 2 ++ 2 +++ 1 + 2 ++ 3 ++ 3 ++ 2 ++ 2 +++ 3 +++ Premarin + Acolbifene 2 ++ 1 + 1 + 000000 DHEA 3 +++ 3 +++ 3 +++ 2 ++ 2 +++ 3 +++ 0 0 1 ++ DHEA + Acolbifene 3 +++ 3 +++ 3 +++ 000000 DHEA + Premarin 2 +++ 3 +++ 3 +++ 2 + 2 +++ 3 +++ 2 ++ 2 +++ 3 +++ DHEA + Premarin + Acolbifene 3 +++ 3 +++ 3 +++ 000000 Sex steroid receptors: AR, androgen receptor; ER␣, ; PR, progesterone receptor. Vaginal layers: E, epithelium; L, lamina propria; M, muscularis. Numbers represent the semi-quantitative evaluation of the number of labeled nuclei: 0, none, 1, low; 2, moderate; 3, high. Labeling intensity is indicated as (+) low; (++) moderate; (+++) high. the lamina propria and muscularis were also strongly stained. few nuclei were moderately stained in the lamina propria and In the OVX group, the basal epithelial layer showed few mod- weakly labeled in the muscularis. erately stained nuclei, while a few strongly stained nuclei The epithelium of OVX animals revealed a moderate num- were seen in the lamina propria and muscularis. On the other ber of variably stained nuclei, with weakly to strongly labeled hand, in the Acolbifene-treated group, the basal and super- areas and few strongly labeled nuclei in the lamina propria ficial epithelial layers were moderately labeled while few and muscularis. Treatment of OVX animals with Acolbifene nuclei were strongly stained in the lamina propria and mus- eliminated ER␣ labeling from the three tissue layers. Mean- cularis. while, treatment with Premarin led to strong to moderate With Premarin, the epithelial basal and the three to four staining of the basal cells, while the three suprabasal epithe- suprabasal cell layers were moderately stained, while a mod- lial layers were weakly stained. Many moderately stained erate number of strongly labeled nuclei were found in the nuclei were found in the lamina propria and muscularis fol- lamina propria and a few weakly labeled nuclei were seen lowing Premarin. No staining could be detected when the in the muscularis. After treatment with Premarin and Acolb- combination of Premarin and Acolbifene was used. ifene, a moderate to strong, but scattered labeling was found The vaginal epithelium of DHEA-treated animals showed in the basal and in the superficial epithelial layers, while the many strongly stained nuclei for ER␣ in the basal layer lamina propria and muscularis displayed few weakly labeled and a moderate scattered labeling in the superficial layers. nuclei. Following DHEA treatment, the lamina propria and muscu- DHEA treatment of OVX animals led to the strong laris displayed a moderate number and many strongly stained labeling of many nuclei in the three vaginal layers. nuclei, respectively. No staining was detected when the com- The same pattern was found when the combinations of bination of DHEA and Acolbifene was used. In animals DHEA + Acolbifene and DHEA + Premarin were used, with treated with DHEA and Premarin, the epithelial basal layer the exception of a smaller number of labeled nuclei in was moderately stained while a moderate number of nuclei the superficial layers of the epithelium in the latter group. in the suprabasal layers and some mucous cell nuclei were The combination of DHEA, Premarin and Acolbifene also weakly labeled. Following this treatment, the lamina propria resulted in a strong staining of the majority of the nuclei in and muscularis showed a moderate and a high number of the three vaginal wall layers. strongly stained nuclei, respectively. Combining Acolbifene to DHEA and Premarin led to the complete disappearance of 3.3.2. Estrogen receptor alpha ER␣ staining. As presented in Table 2 and the middle row of Fig. 9, in intact rats at estrus, the epithelium revealed few weakly 3.3.3. Progesterone receptor labeled nuclei in the basal cell layer while some weakly Immunolabeling for PR (Table 2; lower row of Fig. 9) stained nuclei were detected in the lamina propria and many showed many strongly stained nuclei in all tissue compart- weakly stained nuclei were observed in the muscularis. On ments of the vagina in intact rats at estrus while such labeling the other hand, at diestrus, basal and two suprabasal epithe- was absent in the epithelium at diestrus, with the exception of lial layers showed many strongly stained nuclei while very cytoplasmic background. A moderate number of moderately 212 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 stained nuclei were seen in the other two vaginal compart- ment has been reported to lead to a significant decrease in ments in rats at diestrus. In the vagina of OVX animals and all total vaginal collagen [37]. groups of animals which received Acolbifene, there was no In the potential triple combination, the detectable PR labeling. Premarin-treated animals displayed a Premarin is aimed at acting in the brain to relieve the vaso- variable degree of staining with a moderate number of mod- motor symptoms. In fact, the benefits of co-administration erately to strongly stained nuclei in the epithelium and lamina of the pure selective antiestrogen Acolbifene with an estro- propria and a high number of strongly labeled nuclei in the gen in order to neutralize the unwanted systemic effects of muscularis. The DHEA + Premarin group showed a staining the estrogen have been well described by Labrie et al. [21]. pattern similar to that of Premarin alone, while DHEA-treated In the present study, when OVX animals were treated with animals revealed few moderately stained nuclei only in the Acolbifene, the most typical morphological feature induced muscularis. by the SERM was the appearance of a superficial layer of well-aligned small mucous cells overlying a basal cell layer. This pattern was slightly more pronounced in the Pre- 4. Discussion marin + Acolbifene group and remained preponderant in all groups treated with Acolbifene, including when the SERM Vaginal dryness or atrophic vaginitis, also referred to as was combined with DHEA. Vaginal epithelium mucifica- urogenital atrophy, with sexual dysfunction is a common tion has been reported in immature [38] and adult [39] problem in postmenopausal women [27]. The most common rats under antiestrogen treatment. Although this morpholog- symptoms are dryness, burning, pruritus, irritation and dys- ical pattern has been compared to the progesterone-induced pareunia, thus leading to decreased libido and quality of life mucification [38], the molecular mechanism by which [28]. an antiestrogen induces epithelial mucification remains The rat vaginal wall provides an excellent model to study unknown. the inductive/suppressive effects of estrogens, androgens and The most spectacular effects of Acolbifene in OVX + estrogen antagonists. The overall hormonal [29] or antihor- Premarin rat vagina reside in a weight decrease to the level monal [30] effects on the human vagina are usually deter- seen in OVX animals and an important thickness reduction mined by vaginal cytology, although the observations are lim- of the epithelial and muscularis layers, as well as decreased ited to the external layers of the epithelium [13]. In rodents, total vaginal wall thickness, when compared to intact and specific histological patterns of keratinization and mucifica- Premarin-treated groups (Figs. 5 and 8). Such effects support tion of the vaginal epithelium provide the hallmarks of overall the presence of a pure antiestrogenic action of Acolbifene on estrogenic [31], progestative [32] or androgenic [25] effects, these parameters. respectively. To the best of our knowledge, this is the first Contrary to its antiestrogenic actions mentioned above, in study that examines morphology over the entire longitudinal the rat vaginal orifice (segment 1), Acolbifene surprisingly axis of the vagina, especially in the OVX rat treated with displays an estrogen-like effect of epithelial proliferation and DHEA, an estrogen and an antiestrogen. keratinization. Compared to the thin 4–6 layer-thick epithe- In the present study, a typical estrogenic effect reflected lium lining the same segment in OVX animals, the 9–11 lay- by a keratinized stratified squamous epithelium was observed ers of the cornified Acolbifene-induced epithelium is likely in the rat vagina following treatment with the conjugated to provide a better resistance, as well as reduction in inflam- estrogen Premarin. In a previous study performed in our mation incidence. The latter is supported by the observation laboratory, the same morphology was observed following of a reduced incidence of inflammatory changes, not only in treatment with 17␤-estradiol instead of Premarin. Moreover, the first segment but also throughout the vaginal mucosa of the increased thickness and PR expression in the muscu- Acolbifene-treated OVX rats. At the internal vaginal level, laris observed after Premarin treatment represent estrogenic the mucification induced by the SERM is likely to account effects, similar to the findings in intact animals at estrus. for the reduced inflammation incidence. Interestingly, in the uterus of OVX mice treated with the Many beneficial effects of DHEA have been reported in antiestrogen Acolbifene, the atrophic changes were more pro- postmenopausal women [5,13]. The morphological changes nounced in the myometrium [33], thus supporting that, in the observed in the rat vagina after DHEA treatment reflect its absence of estrogen, cell proliferation in the smooth muscle intracrine conversion into active sex steroids having andro- is inhibited in the female rat reproductive tract. genic and/or estrogenic action through intracrine mecha- In the vaginal lamina propria of intact rats at estrus, nisms [3]. Those changes comprise marked epithelial muci- Premarin-treated rats and, to a lesser extent, Premarin + fication, high compactness of delicate, finely woven lamina DHEA-treated animals, low compactness of collagen fibers propria collagen fibers and moderate muscularis thickness in the first segments could also be ascribed to an estrogenic increase when compared to OVX animals. The first and effect. These data are in agreement with previous observa- the second morphological changes are typical of androgenic tions in rats and rabbits at estrus and after estradiol treatment effects while the third shows an estrogen-like activity, which following OVX [26,34–36]. Moreover, in postmenopausal is further supported by a concomitant increase in proges- women with stress incontinence, a 6-month estrogen treat- terone receptor expression in the muscularis layer. L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 213

In addition to the well-described epithelial squamous pro- relaxation (reviewed in [53]). It is noteworthy that Acolbifene liferation and maturation observed under estrogen treatment, and DHEA have been found to induce eNOS in human and it is well known that vaginal epithelial mucification occurs rat endothelial cells [54,55]. in the OVX rat in response to treatment with Finally, the combination of Acolbifene, DHEA and Pre- (Testo) or 5␣- (DHT) [40]. The vagi- marin induced mucification of the vaginal epithelium through nal morphology under Testo treatment has been described androgenic and antiestrogenic effects. These effects were as “cuboidal to columnar mucified cells exhibiting a pseu- reflected by invaginations of multilayered hypertrophied dostratified appearance at places, superficial vacuolation and mucous cells in alternance with well-aligned mucous cells. occasional cornification in the top layer, with compact lam- The high compactness of fine collagen fibers found in the lam- ina propria and atrophic changes in the muscular layer and ina propria is indicative of the androgenic action of DHEA, weight similar to that of intact rats” [26]. Since DHEA is while the significant thickness reduction of the epithelial and transformed into either or both androgens and estrogens in muscularis layers, when compared to the Premarin + DHEA peripheral tissues, the thick mucified multilayered epithelium group, is indicative of the reversal by Acolbifene of the major observed in the present study after treatment of OVX ani- and minor estrogenic actions of both Premarin and DHEA in mals with DHEA, suggests a predominant androgenic effect these two vaginal layers, respectively. of mucification in the rat vagina, an effect which could mask While the antiestrogenic action of Acolbifene should pro- any potential co-existing minor estrogenic effect at the epithe- tect the uterus and mammary gland against estrogen-induced lial level. A previous study has shown the same mucification cancer, and leads, in this study, to the formation of a con- effect in the rat vagina and the intravaginal application of sistent superficial mucous cell layer lining the rat vaginal DHEA achieved a significant effect at a dose 10× lower than lumen, the agonist-like action of Acolbifene gives rise to that found to be active following application of DHEA on the the inexpected formation of a keratinized squamous stratified dorsal skin [41]. epithelium at the vaginal ostium (estrogen-like effect). These The present data indicate co-existing major androgenic two actions of Acolbifene were likely to bring protection and minor estrogenic actions of DHEA in the rat vagina. against the observed inflammation in the OVX group. The Other studies using co-administration of DHEA and the antie- mucifying action of Acolbifene is further increased by the strogen Acolbifene or the FLU, respectively, mucification effect of the androgenic component of DHEA, have demonstrated that the action of DHEA in the rat mam- thus potentially leading to an improved vaginal function, mary gland [42], skin sebaceous glands [43] and bone mineral which is supported by the stimulatory effect of DHEA on density [44] is almost exclusively androgenic. Nevertheless, the lamina propria. the presence of an estrogenic action of DHEA in the rat vagina The beneficial morphological changes, observed con- has been previously demonstrated, either by the formation comitantly with the strong modulation of rat vaginal AR by of a stratified epithelium in the vagina of rats treated with androgens, suggest that decreased serum levels of DHEA- FLU and DHEA (unpublished data), or through induction of derived androgens in postmenopausal women could con- vaginal opening and precocious ovulation in immature rats tribute to the decreased vaginal health and eventually to the treated with this compound, while DHT, an androgen not loss of libido and sexual enjoyment observed in this age aromatizable to estrogens, did not produce such effects [45]. group. Decreased serum total Testo, free Testo and DHEA-S The capacity of rat vaginal tissue to aromatize androgens, were indeed found in women who consulted for decreased especially, Testo, is likely to account for the major part of sexual desire [56]. the estrogenic effect of DHEA in this organ [46]. Androst- The present data show low levels of ER␣ in the vaginal 5-ene-3␤,17␤-diol (5-diol), a DHEA metabolite known to epithelium and lamina propria of rats at estrus, which are bind the estrogen receptor [47–50], could also contribute to under high estrogen influence, and increased ER␣ levels in the estrogenic effect [51]. The proposed combination of the the epithelium of intact rats at diestrus, when estrogen lev- antiestrogen Acolbifene with DHEA would thus prevent any els are low. In addition, during the estrus cycle, ER␣ levels unwanted stimulatory effect of 5-diol. In addition, preven- fluctuated more in the epithelium than in the lamina propria. tion of bone loss would represent an additional benefit of These results are in agreement with those of other investi- Acolbifene [52]. gators [57], except that, in the lamina propria, they detected To the best of our knowledge, no previous study has shown more (nearly 60%) ER␣-immunolabeled nuclei throughout the stimulatory effects of DHEA on the compactness and mor- the rat estrus cycle than in the present study. Moreover, under phology of the lamina propria’s collagen fibers and, to a lesser the different hormonal treatments, ER␣ labeling displayed extent, on the muscularis. Such actions of DHEA-derived more variations in the number and intensity of stained nuclei androgens and estrogens could have beneficial effects on in the epithelium than in the other two tissue layers. While in vaginal function in postmenopausal women and possibly pro- the Premarin, and DHEA + Premarin groups, the presence of vide the substrate required for the action of inhibitors of type 5 estrogens appeared to up-regulate ER␣ expression, as already phosphodiesterase, such as sildenafil or tadalafil, possibly via demonstrated in the mouse uterus [58], the existence of an androgen- or estrogen-induced endothelial nitric oxide syn- ER␣ up-regulation in the absence or presence of low levels of thase (eNOS)-mediated facilitation of vaginal smooth muscle estrogens was observed in OVX- and DHEA-treated animals, 214 L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 respectively. As expected, treatment with Acolbifene com- [11] R.K. Ross, A. Paganini-Hill, P.C. Wan, M.C. Pike, Effect of hormone pletely eliminated ER␣ immunostaining in the whole vaginal replacement therapy on breast cancer risk: estrogen versus estrogen wall and co-treatment with Premarin and/or DHEA did not plus progestin, J. Natl. Cancer Inst. 92 (2000) 328–332. [12] W.G.f.t.W.s.H.I. Investigators, Risks and benefits of estrogen plus change this response. As discussed earlier, a potential bene- progestin in healthy postmenopausal women: principal results from ficial effect of the combination of Acolbifene and DHEA for the Women’s Health Initiative randomized controlled trial, JAMA postmenopausal women is that the binding of the antiestrogen 288 (2002) 321–333. to ER␣ would prevent estrogens produced from aromatiza- [13] F. Labrie, P. Diamond, L. Cusan, J.L. Gomez, A. Belanger,´ Effect tion of DHEA-derived androgens and the metabolite 5-DIOL of 12-month DHEA replacement therapy on bone, vagina, and ␣ endometrium in postmenopausal women, J. Clin. Endocrinol. Metab. to bind ER , thus potentially preventing proliferation of 82 (1997) 3498–3505. ER␣-sensitive breast cancers cell [59]. [14] F. Labrie, Dehydroepiandrosterone (DHEA) as potential hormone Various attempts have been made to solve the problem replacement therapy, Ref´ erence´ en Gynecologie´ Obstetrique´ 8 (2001) of vaginal dryness, often linked to dyspareunia and loss of 317–322. sexual enjoyment. As an example, local [15] A. Lasco, N. Frisina, N. Morabito, A. Gaudio, E. Morini, A. Tri- filetti, G. Basile, V. Nicita-Mauro, D. Cucinotta, Metabolic effects preparations are often prescribed to provide relief but the of dehydroepiandrosterone replacement therapy in postmenopausal endometrium may be stimulated by the unopposed estrogen women, Eur. J. Endocrinol. 145 (2002) 457–461. [60]. In order to achieve a more physiological and tissue- [16] S. Gauthier, B. Caron, J. Cloutier, Y.L. Dory, A. Favre, D. specific HRT, DHEA combined with an estrogen and with Larouche, J. Mailhot, C. Ouellet, A. Schwerdtfeger, G. Leblanc, a SERM, such as Acolbifene, which displays antiestrogenic C. Martel, J. Simard, Y. Merand,´ A. Belanger,´ C. Labrie, F. Labrie, (S)-(+)-4-[7-(2,2-dimethyl-1-oxopropoxy)-4-methyl-2-[4-[2- activity in the mammary gland and uterus, could potentially (1-piperidinyl)-ethoxy]phenyl]-2H-1-benzopyran-3-yl]-phenyl 2,2- meet the important needs of postmenopausal women. In fact, dimethylpropanoate (EM-800): a highly potent, specific, and control of vasomotor symptoms, relief of vaginal dryness orally active antiestrogen, J. Med. Chem. 40 (1997) with potential benefits on sexual function could be achieved, 2117–2122. concomitantly with prevention of breast, uterine and ovarian [17] F. Labrie, C. Labrie, A. Belanger,´ J. Simard, S. Gauthier, V. Luu-The, Y. Merand,´ V. Giguere,` B. Candas, S. Luo, C. Martel, S.M. Singh, cancer as well as osteoporosis and potential improvement M. Fournier, A. Coquet, V. Richard, R. Charbonneau, G. Charpenet, of cognitive functions and memory as well as prevention of A. Tremblay, G. Tremblay, L. Cusan, R. Veilleux, EM-652 (SCH Alzheimer’s disease [61]. 57068): a third generation SERM acting as pure antiestrogen in the mammary gland and endometrium, J. Steroid Biochem. Mol. Biol. 69 (1999) 51–84. [18] G.B. Tremblay, A. Tremblay, N.G. Copeland, D.J. Gilbert, N.A. References Jenkins, F. Labrie, V. Giguere, Cloning, chromosomal localization and functional analysis of the murine estrogen receptor ␤, Mol. [1] B. Rossin-Amar, Vaginal dryness in the menopausal woman (physio- Endocrinol. 11 (1997) 353–365. logical and psychological aspects), Gynecol. Obstet. Fertil. 28 (2000) [19] S. Dauvois, C.S. Geng, C. Levesque,´ Y. Merand,´ F. Labrie, Addi- 245–249. tive inhibitory effects of an androgen and the antiestrogen EM-170 [2] L. Pandit, J.G. Ouslander, Postmenopausal vaginal atrophy and on estradiol-stimulated growth of human ZR-75-1 breast tumors in atrophic vaginitis, Am. J. Med. Sci. 314 (1997) 228–231. athymic mice, Cancer Res. 51 (1991) 3131–3135. [3] F. Labrie, Intracrinology, Mol. Cell. Endocrinol. 78 (1991) [20] S. Couillard, C. Labrie, A. Belanger,´ B. Candas, F. Pouliot, F. Labrie, C113–C118. Effect of dehydroepiandrosterone and the antiestrogen EM-800 on [4] F. Labrie, A. Belanger,´ J. Simard, V. Luu-The, C. Labrie, DHEA and the growth of human ZR-75-1 breast cancer xenografts, J. Natl. peripheral androgen and estrogen formation: intracrinology, Ann. N. Cancer Inst. 90 (1998) 772–778. Y. Acad. Sci. 774 (1995) 16–28. [21] F. Labrie, M. El-Alfy, L. Berger, C. Labrie, C. Martel, A. Belanger,´ [5] F. Labrie, V. Luu-The, C. Labrie, A. Belanger,´ J. Simard, S.-X. B. Candas, G. Pelletier, The combination of a novel SERM with an Lin, G. Pelletier, Endocrine and intracrine sources of androgens in estrogen protects the mammary gland and uterus in a rodent model: women: inhibition of breast cancer and other roles of androgens the future of postmenopausal women’s health? Endocrinology 144 and their precursor dehydroepiandrosterone, Endocr. Rev. 24 (2003) (2003) 4700–4706. 152–182. [22] W.M. van Weerden, H.G. Bierings, G.J. van Steenbrugge, F.H. de [6] N. Orentreich, J.L. Brind, R.L. Rizer, J.H. Vogelman, Age changes Jong, F.H. Schroder, Adrenal glands of mouse and rat do not syn- and sex differences in serum dehydroepiandrosterone sulfate concen- thesize androgens, Life Sci. 50 (1992) 857–861. trations throughout adulthood, J. Clin. Endocrinol. Metab. 59 (1984) [23] C.Y. Kramer, Extension of multiple range tests to group means 551–555. with unique numbers of replications, Biometrics 12 (1956) 307– [7] F. Labrie, A. Belanger,´ L. Cusan, J.L. Gomez, B. Candas, Marked 310. decline in serum concentrations of adrenal C19 sex steroid pre- [24] Y.D. Yuan, G.L. Foley, Female reproductive system, in: M.A. Wal- cursors and conjugated androgen metabolites during aging, J. Clin. lig (Ed.), Handbook of Toxicologic Pathology, second ed., Elsevier Endocrinol. Metab. 82 (1997) 2396–2402. Science, USA, 2002. [8] M.J. Rosenberg, T.D. King, M.C. Timmons, Estrogen–androgen for [25] J.L. Ladinsky, H.W. Gruchow, B.M. Peckham, Cellular behaviour hormone replacement: a review, J. Reprod. Med. 42 (1997) 394–404. of the vaginal epithelium treated with alone [9] I.D. Burd, G.A. Bachmann, Androgen replacement in menopause, and in combination with diethylstilboestrol, J. Endocrinol. 41 (1968) Curr. Womens Health Rep. 1 (2001) 202–205. 161–169. [10] G.A. Colditz, S.E. Hankinson, D.J. Hunter, W.C. Willett, J.E. Man- [26] P.K. Mehrotra, J.N. Karkun, Studies on physiology and biochemistry son, M.J. Stampfer, C. Hennekens, B. Rosner, F.E. Speizer, The use of female genital tract: response of cervix and vagina of albino rats of estrogens and progestins and the risk of breast cancer in post- to estrogen, progesterone and testosterone, Indian J. Exp. Biol. 11 menopausal women, N. Engl. J. Med. 332 (1995) 1589–1593. (1973) 1–6. L. Berger et al. / Journal of Steroid Biochemistry & Molecular Biology 96 (2005) 201–215 215

[27] M. Notelovitz, Urogenital atrophy and low-dose vaginal estrogen [46] E.D. Lephart, D. Mathews, J.F. Noble, S.R. Ojeda, The vagi- therapy, Menopause 7 (2000) 140–142. nal epithelium of immature rats metabolizes androgens through an [28] J.R. Berman, L.A. Berman, T.J. Werbin, I. Goldstein, Female sexual aromatase-like reaction: changes during the time of puberty, Biol. dysfunction: anatomy, physiology, evaluation and treatment options, Reprod. 40 (1989) 259–267. Curr. Opin. Urol. 9 (1999) 563–568. [47] T.C. Shao, E. Castaneda, R.L. Rosenfield, S. Liao, Selective reten- [29] G.N. Papanicolaou, E. Shorr, The action of ovarian follicular hor- tion and formation of a delta5-–receptor complex in mones in the menopause, as indicated by vaginal smears, Am. J. cell nuclei of the rat vagina, J. Biol. Chem. 250 (1950) 3095– Obstet. Gynecol. 31 (1936) 806–831. 3100. [30] M. Friedrich, D. Mink, C. Villena-Heinsen, A. Woll-Hermann, S. [48] J. Poortman, J.A. Prenen, F. Schwarz, J.H. Thijssen, Interaction Wagner, W. Schmidt, The influence of on the maturation of delta-5-androstene-3beta, 17beta-diol with estradiol and dihy- index of vaginal epithelium, Clin. Exp. Obstet. Gynecol. 25 (1998) drotestosterone receptors in human myometrial and mammary cancer 121–124. tissue, J. Clin. Endocrinol. Metab. 40 (1975) 373–379. [31] B. Peckham, W. Kiekhofer, Cellular behavior in the vaginal epithe- [49] L.G. Van Doorn, J. Poortman, J.H. Thijssen, F. Schwarz, Actions lium of estrogen-treated rats, Am. J. Obstet. Gynecol. 83 (1962) and interactions of 5-androstene-3␤,17␤-diol and estradiol 17␤ 1021–1027. in the immature rat uterus, Endocrinology 108 (1981) 1587– [32] H. Selye, J.S.L. Browne, J.B. Collip, Effect of combined administra- 1593. tion of oestrone and progesterone in adult ovariectomized rats, Proc. [50] J. Adams, M. Garcia, H. Rochefort, Estrogenic effects of physio- Soc. Exp. Biol. Med. 34 (1936) 198–200. logical concentrations of 5-androstene-3 beta, 17 beta-diol and its [33] S. Luo, A. Sourla, C. Labrie, S. Gauthier, Y. Merand, A. Belanger, metabolism in MCF7 human breast cancer cells, Cancer Res. 41 F. Labrie, Effect of twenty-four-week treatment with the antiestro- (1981) 4720–4726. gen EM-800 on estrogen-sensitive parameters in intact and ovariec- [51] R. Poulin, F. Labrie, Stimulation of cell proliferation and estrogenic tomized mice, Endocrinology 139 (1998) 2645–2656. response by adrenal C19-delta 5-steroids in the ZR-75-1 human [34] M.G. Sartori, M.J. Girao, M. de Jesus Simoes, J.P. Sartori, E.C. breast cancer cell line, Cancer Res. 46 (1986) 4933–4937. Baracat, G. Rodrigues de Lima, Quantitative evaluation of collagen [52] C. Martel, S. Picard, V. Richard, A. Belanger, C. Labrie, F. and muscle fibers in the lower urinary tract of castrated and under- Labrie, Prevention of bone loss by EM-800 and in the hormone replacement female rats, Clin. Exp. Obstet. Gynecol. 28 ovariectomized rat, J. Steroid Biochem. Mol. Biol. 74 (2000) 45– (2001) 92–96. 56. [35] N. Fleischmann, G. Christ, T. Sclafani, A. Melman, The effect of [53] R. Munarriz, S.W. Kim, N.N. Kim, A. Traish, I. Goldstein, A review ovariectomy and long-term estrogen replacement on bladder structure of the physiology and pharmacology of peripheral (vaginal and cli- and function in the rat, J. Urol. 168 (2002) 1265–1268. toral) female genital arousal in the animal model, J. Urol. 170 (2003) [36] H.N. Yoon, W.S. Chung, Y.Y. Park, B.S. Shim, W.S. Han, S.W. S40–S44 (discussion S44–S45). Kwon, Effects of estrogen on nitric oxide synthase and histological [54] T. Simoncini, G. Varone, L. Fornari, P. Mannella, M. Luisi, F. Labrie, composition in the rabbit clitoris and vagina, Int. J. Impot. Res. 13 A.R. Genazzani, Genomic and nongenomic mechanisms of nitric (2001) 205–211. oxide synthesis induction in human endothelial cells by a fourth- [37] S. Jackson, M. James, P. Abrams, The effect of oestradiol on vaginal generation selective estrogen receptor modulator, Endocrinology 143 collagen metabolism in postmenopausal women with genuine stress (2002) 2052–2061. incontinence, BJOG 109 (2002) 339–344. [55] T. Simoncini, P. Mannella, L. Fornari, G. Varone, A. Caruso, A.R. [38] W.A. Anderson, Y.H. Kang, Estrogen and antagonist-induced struc- Genazzani, Dehydroepiandrosterone modulates endothelial nitric tural changes in the cervico-vaginal epithelium of immature rats, oxide synthesis via direct genomic and nongenomic mechanisms, Am. J. Anat. 144 (1975) 197–207. Endocrinology 144 (2003) 3449–3455. [39] M. Yoshida, K. Kudoh, S. Katsuda, M. Takahashi, J. Ando, A. [56] A.T. Guay, J. Jacobson, Decreased free testosterone and dehy- Maekawa, Inhibitory effects of uterine endometrial carcinogenesis droepiandrosterone (DHEA-S) levels in women with decreased in Donryu rats by tamoxifen, Cancer Lett. 134 (1998) 43–51. libido, J. Sex Marit. Ther. 28 (2002) 129–142. [40] T.G. Kennedy, D.T. Armstrong, Induction of vaginal mucification in [57] H. Wang, H. Eriksson, L. Sahlin, Estrogen receptors alpha and beta rats with testosterone and 17beta-hydroxy-5alpha-androstan-3-one, in the female reproductive tract of the rat during the estrous cycle, Steroids 27 (1976) 423–430. Biol. Reprod. 63 (2000) 1331–1340. [41] A. Sourla, M. Flamand, A. Belanger, F. Labrie, Effect of dehy- [58] M.D. Mergman, B.S. Schachter, K. Karelus, E.P. Combatsiaris, T. droepiandrosterone on vaginal and uterine histomorphology in the Garcia, J.F. Nelson, Up-regulation of the uterine estrogen receptor rat, J. Steroid Biochem. Mol. Biol. 66 (1998) 137–149. and its messager ribonucleic acid during the mouse estrous cycle: [42] A. Sourla, C. Martel, C. Labrie, F. Labrie, Almost exclusive andro- the role of estradiol, Endocrinology 130 (1992) 1923–1930. genic action of dehydroepiandrosterone in the rat mammary gland, [59] M. Maggiolini, O. Donze,´ E. Jeannin, S. Ando, D. Picard, Adrenal Endocrinology 139 (1998) 753–764. androgens stimulate the proliferation of breast cancer cells as direct [43] A. Sourla, V. Richard, F. Labrie, C. Labrie, Exclusive androgenic activators of estrogen receptor a, Cancer Res. 59 (1999) 4864– effect of dehydroepiandrosterone in sebaceous glands of rat skin, J. 4869. Endocrinol. 166 (2000) 455–462. [60] L.A. Mattson, G. Culberg, O. Eriksson, F. Kruitsson, Vaginal admin- [44] C. Martel, A. Sourla, G. Pelletier, C. Labrie, M. Fournier, S. Picard, istration of low dose estradiol: effects on endometrium and vaginal S. Li, M. Stojanovic, F. Labrie, Predominant androgenic component cytology, Maturitas (1989) 217–222. in the stimulatory effect of dehydroepiandrosterone on bone mineral [61] F. Labrie, Extragonadal synthesis of sex steroids: intracrinology, density in the rat, J. Endocrinol. 157 (1998) 433–442. Ann. Endocrinol. (Paris) 64 (2003) 95–107. [45] J.F. Knudsen, V.B. Mahesh, Initiation of precocious sexual mat- [62] P. Popesko, V. Rajtova, J. Horak, Female reproductive organs—a uration in the immature rat treated with dehydroepiandrosterone, colour atlas of anatomy of small laboratory animals, volume two: Endocrinology 97 (1975) 458–468. rat-mouse-hamster, Elsevier Sci. (2003) 85.