Variable Expression of Cellular Retinol- and Cellular ... - CiteSeerX

BIOLOGY OF REPRODUCTION 56, 125-132 (1997)

Variable Expression of Cellular Retinol- and Cellular Retinoic Acid-Binding Proteins in the Rat Uterus and Ovary during the Estrous Cycle' Sarah A. Wardlaw, Richard A. Bucco, Wen Li Zheng, and David E. Ong 2 Department of Biochemistry, Vanderbilt University, Nashville, Tennessee 37232-0146 ABSTRACT Ovaries and uteri from normal adult female rats at known stages of the estrous cycle were analyzed for the presence of cellular retinol-binding protein (CRBP) and both types of cellular retinoic acid-binding protein (CRABP and CRABP II). Northern and Western blot analysis of the uteri revealed a peak of CRBP during diestrus and a peak of CRABP during proestrus, whereas CRABP II peaked sharply during estrus. Immunohistochemical studies showed CRABP II localized to the luminal epithelium, while both CRBP and CRABP were observed only in the smooth muscle layers of the uterus. In the ovary, CRABP was not detected, while CRBP levels remained relatively constant throughout the cycle and CRABP II peaked slightly during metestrus. CRBP in the ovary was localized to the oocytes, nearby granulosa cells, and some regions of stroma. CRABP II was found predominantly in the granulosa cells of mature follicles and early corpora lutea, as well as some regions of the stroma. These results suggest a need for further studies to assess the role of retinol and its metabolites in normal uterine function and ovarian follicular development.

calization of these proteins to such cell types as the granulosa/luteal cells in the ovary and the luminal epithelial cells of the uterus, support the fundamental and complex involvement of retinol and retinoic acid in the female reproductive system. Here we have examined the expression of the vitamin A-binding proteins in the uterus and ovary during the estrous cycle of the normal female rat in order to suggest possible trafficking routes and roles for retinoids during the cycle. MATERIALS AND METHODS Animals and Staging Twenty-five adult female Sprague-Dawley rats (6-7 wk old on delivery) were housed in a temperature- and lightcontrolled room (21 1°C, lights-on 0700-1900 h) and provided with chow and water ad libitum. Staging was conducted by daily examination of vaginal cytology [9, 10]. At approximately 10 wk of age, animals exhibiting at least two normal 4- to 5-day estrous cycles were killed immediately after staging (1000-1200 h). Reproductive organs representative of each stage of the cycle were collected and either immediately immersion-fixed for immunohistochemical study or snap-frozen and stored at -70°C for later RNA purification and preparation of cytosols. These studies were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals and with the oversight of veterinarians and our local Institutional Animal Care and Use Committee.

INTRODUCTION In addition to a role in cell growth and differentiation and its role in vision, vitamin A has long been recognized as a necessary factor in the process of reproduction. Vitamin A deficiency in male rats produces germ cell degeneration, while female rats deprived of vitamin A are unable to reproduce, with pregnancies typically ending in fetal death and resorption [1]. Vitamin A deficiency in female rats also produces morphological changes in the luminal and glandular epithelium of the uterus. These changes do not occur with ovariectomy but do appear in vitamin A-deficient ovariectomized rats treated with estrogen [2], suggesting an interplay of vitamin A and ovarian hormones in normal reproductive function. Possible vitamin A-ovarian hormone interactions are also suggested by the fluctuation of vitamin A levels observed in the ovary [3] and serum [4], as well as variation in the amount of cellular retinoid-binding protein (CRBP) mRNAs in the uterus [5] during the human menstrual cycle. Such fluctuations in binding protein mRNA have also been shown to occur in the rat cervix during the estrous cycle [6], as well as in the ovary and uterus of pregnant eCGand hCG-treated prepubertal rats [7, 8]. In the latter case, CRBP expression in the uterus was inhibited by eCG treatment, whereas both types of cellular retinoic acid-binding protein found in the uterus (CRABP and CRABP II) and CRABP II expression in the ovary were up-regulated by gonadotropin treatment. These effects on binding protein levels could also be elicited by administration of estrogen [7, 8]. These data, along with the immunohistochemical lo-

RNA Isolation and Northern Analysis RNA was isolated from individual ovaries and uterine horns by a modification of the method of Chomczynski and Sacchi [11]. The frozen tissues were cut into small pieces and homogenized with a Polytron homogenizer (Brinkman, Luzerne, Switzerland) at 20 000 rpm for 2 x 30 sec. Homogenization took place in 5 ml of a 2 M guanidine thiocyanate, 2.5% (w:v) sarcosyl, 50% phenol solution. After solubilization, 0.75 ml of 49:1 chloroform:isoamyl alcohol was added, followed by vortex mixing and centrifugation at 10 000 x g. The upper layer was removed, and to this was added a 50% volume of the guanidine thiocyanate solution and a 15% volume of chloroform:isoamyl alcohol. After vortex mixing and centrifugation, the upper layer was removed and added to an equal volume of isopropanol. The RNA was allowed to precipitate overnight at -20 0 C and g for then was spun down by centrifugation at 10 000 45 min. The pellet was washed with 70% ethanol and allowed to air-dry. The RNA was resuspended in a small volume of water, and RNA content was determined by spectrophotometric analysis at 260 nm. For Northern blot analysis, equal amounts of total RNA (40 Lg) from each sample were denatured in 1.1 M glyoxal, 50% dimethyl sulfoxide (DMSO), 10 mM sodium phosphate (pH 7.0) at 50°C for 1 h before separation on a 1.2%

Accepted August 23, 1996. Received June 10, 1996. 'Supported by NIH Grants DK 32642 and HD 25206. 2Correspondence. FAX: (615) 343-0704. 125

126

WARDLAW ET AL.

(w:v) agarose gel. The RNA was then transferred to a nylon membrane (Duralon; Stratagene, La Jolla, CA) by the capillary action of 10-strength SSC buffer (single-strength SSC = 15 mM sodium citrate, 0.15 M NaCl). The RNA was fixed to the membrane by UV irradiation and baking at 80°C for I h. The blots were prehybridized in 50 ml of 50% deionized formamide, 5-strength SSC (single-strength SSC is 0.15 M sodium chloride and 0.015 M sodium citrate), 25 mM sodium phosphate (pH 7.0), 0.1% (w:v) BSA (fraction V), 0.1% (w:v) Ficoll (400), 0.1% (w:v) polyvinylpyrrolidone, 0.05% (w:v) salmon sperm, and 0.01% (w:v) yeast RNA for 3-19 h at 420C. 32 P-Labeled probes were generated by the use of random primers (Prime-It II kit; Stratagene) and isolated rat CRBP, CRABP, and CRABP II cDNA as a template. 3 2 P-Labeled cDNA probe (20-40 JLCi) was added to the 50 ml of prehybridization solution, and the blot was incubated with the probe for approximately 24 h at 42 0C. After hybridization, the blots were washed once with 5-strength SSC, 1% (w:v) SDS at 45 0C for 15 min; twice with double-strength SSC, 0.1% SDS at 45°C for 15 min each; and once with 0.1-strength SSC, 0.1% SDS at 60 0C for 15 min. Autoradiographs were generated and quantified using a Molecular Dynamics (Sunnyvale, CA) Phosphoimager and software. In order to control for unequal loading and verify RNA integrity, 3 2P-labeled probes generated to a constitutively expressed Chinese hamster ovary-B (CHO-B) gene, a homolog of the rat 40S ribosomal subunit protein S2 [12], were hybridized and imaged as above, and all quantification of CRBP, CRABP, and CRABP II transcripts were corrected for any variation in CHO-B levels. Corrected values for animals judged to be at the same stage of the estrous cycle were averaged (see Figs. 2 and 3), and the means were then subjected to statistical analysis. Tests of significance included the F-test, which resulted in a p value that indicated the likelihood that all samples were drawn from the same population, and the Student-NewmanKeuls test, or q-test, which provides a procedure for comparing more than one pair of means where the resulting aTvalue indicates the probability of erroneously concluding that a difference exists between the means (see figure legends for Figs. 2 and 3 for actual values) [13]. Preparation of Cytosols and Immune Reagents Frozen tissues were homogenized with a Polytron homogenizer at 12 000 rpm for 2 x 30 sec in 10 mM Tris/ HCI (pH 7.0), 1 mM EDTA, and 1 mM dithiothreitol at a ratio of 5-10 ml buffer:l g tissue. The homogenates were centrifuged at 20 000 g for 15 min, and the supernatant was then centrifuged at 100 000 x g for 1 h. The pellet was discarded, and the supernatant (cytosol) was aliquoted and stored at -70°C. Cytosolic protein was quantified by means of a BCA protein analysis kit (Pierce, Rockford, IL). Antibodies against CRABP II were generated in collaboration with Robert Seitz (Research Genetics, Huntsville, AL), as described previously [8]. The preparation of immune serum against CRBP and CRABP purified from rat has also been previously described [14]. The total IgG fractions of these antisera were isolated by use of a protein A column (Pierce), followed by passage over an affinity column composed of recombinant binding protein bound to Sepharose 4B in order to isolate the binding protein-specific IgG population [15]. The specificity of these antibodies for the binding proteins against which they were generated has been reported elsewhere [7].

Western Analysis and Immunohistochemical Localization Western blots were prepared by separating 50 pxg of total ovarian or uterine protein on a 12% SDS-polyacrylamide gel. The proteins were then transferred to an Immobilon-P membrane (Millipore, Bedford, MA) by means of a tank transfer system run at 0.5 amp with a buffer composed of 25 mM Tris, 192 mM glycine, 20% methanol. The blots were blocked for I h in 20 mM Tris/HCl (pH 7.4), 0.15 M NaCl, 0.05% Tween-20, and 5% (w:v) dry milk powder. Primary antibody incubation was performed overnight at 4°C in blocking solution. Blots were washed 3 times in wash solution (blocking solution without the milk powder). Blots were then incubated for 1 h in blocking solution containing a 1:5000 dilution of donkey anti-rabbit IgG antibody conjugated to horseradish peroxidase (Amersham, Arlington Heights, IL) and washed 3 times. Subsequent visualization steps were performed according to the protocol accompanying Amersham's ECL kit. Tissue samples for immunolocalization were immersionfixed for approximately 24 h in 4% (w:v) paraformaldehyde, 2% (w:v) trichloroacetic acid (TCA), 2% (w:v) zinc chloride, and 20% isopropanol. The tissues were then transferred to 70% ethanol and delivered to the Histopathology Department of Vanderbilt University (Nashville, TN) for embedding in paraffin and preparation of slide sections. After the removal of the paraffin with xylene and the equilibration of the tissue sections in Tris-buffered saline (pH 7.6) (TBS), the slides were blocked for 1 h in TBS containing 3% (w:v) BSA. This blocking agent was also used to dilute the primary antibodies. Incubation with primary antibody was carried out in a humidified chamber overnight at 4°C. The slides were then rinsed and incubated with goat anti-rabbit IgG antibody conjugated to biotin (Jackson ImmunoResearch Labs, West Grove, PA) for I h. After being rinsed, the slides were then incubated for 1 h with an antibiotin antibody linked to alkaline phosphatase (Jackson ImmunoResearch). Staining was produced by incubation for 20-30 min with an alkaline phosphatase substrate (Dimension Labs, Mississauga, ON, Canada). Tissues were counterstained with hematoxylin and preserved with aqueous mountant (Serotec, Washington, DC). Immunohistochemical controls consisted most often of the use of affinity column flow-through (see previous section) in place of primary antibody, but also included incubation of the primary antibody solution with ligand bound to Sepharose 4B before use or the addition of 100-fold or more amounts of pure ligand to the primary antibody solution to compete for tissue binding. RESULTS Cellular Retinoid-Binding Protein mRNA Levels in Staged Organs Adult female Sprague-Dawley rats were staged by daily examination of vaginal cytology. After at least two successive normal estrous cycles, the ovaries and uteri were collected, quick-frozen, and stored at -70C. RNA was extracted from single ovaries and uterine horns representing each stage of the estrous cycle. Three separate RNA samples were obtained for each stage, each from a different animal. Equal amounts of RNA from each sample were separated on an agarose gel, transferred to a nylon membrane, and probed for the mRNA of the binding protein of interest. The resulting Northern blots (Fig. 1) showed significant

VITAMIN A-BINDING PROTEINS IN RAT UTERUS AND OVARY

127

CRBP

40

20

0

0 CRABP I 60 40

E:

4) c D

a

20 0 80

CRABP 11

60 40 20 0

I

E

FIG. 1. Scanned autoradiogram of a total uterine RNA blot (40 CIg RNA/ lane) probed with 32P-labeled CRABP (a)and CRABP II (b) cDNA showing mRNA levels in uterus during estrus (E), metestrus I (M'), metestrus 11(M2), diestrus (D), and proestrus (P) and demonstrating peak of CRABP mRNA during proestrus and peak of CRABP II mRNA during estrus. Each lane represents total RNA isolated from a single uterus; 3 animals examined/ stage, 2 shown.

variation in binding protein mRNAs in the uterus during the estrous cycle, with CRABP and CRABP II mRNA levels showing the most dramatic stage-dependent differences. CRBP mRNA levels appeared to peak slightly during diestrus (data not shown), whereas CRABP mRNA levels appeared to be higher during proestrus, and CRABP II mRNA levels appeared to be significantly higher during estrus. These results were confirmed upon quantification of the binding protein mRNA and normalization to the quantified mRNA values of the constitutively expressed CHO-B gene (Fig. 2). CRBP mRNA levels could then be seen to peak during diestrus, while CRABP mRNA levels peaked during proestrus, falling to a low point during the latter half of metestrus and then rising significantly by diestrus. Like CRABP, CRABP II mRNA levels were lowest during metestrus. In contrast to CRABP, CRABP II mRNA levels did not increase significantly until estrus, when a sharp rise was noted, possibly a result of different functions and/or regulation for these two retinoic acid-binding proteins in the uterus. In the ovary, CRBP mRNA levels appeared to remain relatively constant throughout the estrous cycle (Fig. 3). The slight increase in level during proestrus was not statis-

Mi

M2 Stage

D

P

FIG. 2. Quantified CRBP, CRABP, and CRABP II mRNA in the uterus 2 during estrus (E), metestrus I (M'), metestrus 11(M ), diestrus (D), and proestrus (P) obtained by densitometric analysis of scanned autoradiograms (Fig. 1) and normalization to values obtained for constitutively expressed CHO-B mRNA. Three animals examined/stage; values shown represent average ± SD. Statistical analysis confirmed that CRBP mRNA levels were higher during diestrus, CRABP mRNA levels were higher during proestrus, and CRABP II mRNA levels were higher during estrus. (CRBP: F-test [all stages] = p < 0.05; q-test [D,E = aT < 0.05. CRABP: F-test [all stages] = p < 0.01; q-test [P,M2 ] = aT < 0.01. CRABP II: F-test [all stages] = p

< 0.01; F-test [all stages except El = differences not significant; q-test [E,P] = T < 0.01.)

tically significant. CRABP mRNA was not detected in the ovary during any stage of the estrous cycle (data not shown). CRABP II mRNA levels did appear to reach a high point around the time of metestrus (Fig. 3), in contrast to the sharp peak of CRABP II mRNA observed in the uterus during estrus. Cellular Retinoid-Binding Protein Expression Levels in Staged Organs Cytosols were prepared from frozen ovaries and uteri for analysis of binding protein levels and comparison to variations in mRNA level during the estrous cycle. Ovaries and uteri from rats at the same stage were combined before the preparation of cytosols. Equal amounts of cytosolic protein from each sample were separated on SDS-polyacrylamide gels, transferred to a polyvinyl difluoride (PVDF) membrane, and probed with antibodies previously shown to be specific for the binding proteins against which they were generated [7]. Because the organs were pooled, the results should not be considered definitive but are provided in support of the data resulting from Northern analysis. Western analysis of the uterine cytosols showed peaks

WARDLAW ET AL.

128 CRBP 15 10

Io .e

CRABP II

E 40 § 30

•~ 3Z Xi

20 10 i

I

I

I

M2

D

tn

0 E

M1

P

Stage FIG. 3. Quantified CRBP and CRABP II mRNA in the ovary during estrus (E), metestrus I (M'), metestrus II (M2), diestrus (D), and proestrus (P) ob-

tained by densitometric analysis of scanned autoradiograms and normalization to values obtained for constitutively expressed CHO-B mRNA. Three animals examined/stage; values shown represent average SD. Statistical analysis confirmed that CRBP mRNA levels remained relatively constant while CRABP II mRNA levels were higher during metestrus II. (CRBP: F-test all stages] = differences not significant. CRABP II: F-test [all stages] = p < 0.05; q-test M2', P = c < 0.05.)

FIG. 4. CRBP (a), CRABP (b), and CRABP II (c) protein levels in the uterus during estrus (E), metestrus I (M'), metestrus II (M2 ), diestrus (D), and proestrus (P) obtained by incubation of cytosolic protein blot (50 p.g/ lane) with binding protein-specific antibody and visualization with Amersham ECL kit. Lane 1 (S) contains 20 ng of pure protein. Uteri from 2-3 animals/stage were combined for preparation of cytosols.

FIG. 5. CRBP (a) and CRABP II (b) protein levels in the ovary during estrus (E), metestrus I (M'), metestrus II (M), diestrus (D), and proestrus (P) obtained by incubation of cytosolic protein blot (50 pig/lane) with binding protein-specific antibody and visualization with Amersham ECL kit. Lane 1 (S) contains 20 ng of pure protein. Ovaries from 2-3 animals/ stage were combined for preparation of cytosols.

of binding protein expression that matched those seen for binding protein mRNA (Fig. 4), with CRBP protein levels peaking during diestrus, CRABP levels peaking during proestrus, and CRABP II peaking during estrus. However, the intensities of the peaks in relation to the protein levels at other stages of the cycle appeared to differ slightly. For example, the CRBP level during diestrus was considerably higher than that seen in the preceding stage, while the peak of CRBP mRNA at this time represented a less dramatic increase over the previous stage. The high level of protein during diestrus seemed to persist into the next stage, whereas the mRNA dropped to a level similar to that of the other stages during proestrus. Like CRBP, CRABP levels during proestrus were considerably higher than during the preceding stage, with the protein persisting during the subsequent stage and then declining gradually to a low point during diestrus, rather than during metestrus as was seen with the mRNA. CRABP II also peaked dramatically and then fell to an intermediate level during the next two stages, unlike the mRNA, which dropped to a low level immediately after estrus. The expression of CRBP and CRABP II in the ovary during the estrous cycle also followed the pattern observed for mRNA levels (Fig. 5). As with the mRNA, CRABP itself was not detected in ovarian cytosol at any stage of the cycle (data not shown). CRBP remained relatively constant, while CRABP II expression varied considerably during the cycle. As was seen with the uterus, the peak level of CRABP II was considerably higher than that seen in the preceding stage, with the protein lingering through subsequent stages. Unlike CRABP II in the uterus, the peak of CRABP II in the ovary occurred one stage after the peak of mRNA, possibly the result of differing transcriptional regulation or protein half-life in the two organs. Cell-Specific Localization of Binding Proteins in Staged Organs Several samples of uteri and ovaries representing each stage of the estrous cycle were immersion-fixed in a paraformaldehyde solution and then subjected to immunohistochemical analysis involving incubation with binding pro-

VITAMIN A-BINDING PROTEINS IN RAT UTERUS AND OVARY

129

FIG. 6. Immunohistochemical localization of binding proteins in uterus and ovary using alkaline phosphatase-based staining system (brown) and hematoxylin counter-stain (blue). A-C) Positive staining for CRABP II of the luminal epithelial cells of uterus during proestrus (A), estrus (B), and metestrus (C), with no staining apparent in the glandular epithelium (g) (x100). D) Positive staining for CRBP in smooth muscle of diestrus uterus shown in cross section. Staining of macrophages in stroma (s) is nonspecific (x100). E) Intense CRABP II staining of granulosa cells of mature ovarian follicle during estrus. Some staining of stroma (s) also apparent, as well as speckled staining of mature corpus luteum (cl). Note lack of staining in surrounding thecal cell layer and adjacent less-mature and atretic follicles (x40). F) Positive staining for CRABP II in corpora lutea of diestrus ovary (x40). C) CRBP staining of inner follicular granulosa cells, oocytes, and exterior stroma of proestrous ovary. Note stromal staining associated only with mature, nonatretic follicles and early corpora lutea (x40). H) CRBP staining of mature and primordial oocytes of estrous ovary (200).

130

WARDLAW ET AL.

tein-specific antibody followed by the use of an alkaline phosphatase-based staining system as described in Materials and Methods. Incubation of uterine sections with an antibody raised against CRBP resulted in staining of both the inner and outer muscle layers, with the outer layer staining more strongly (Fig. 6D). The staining was visually similar for each stage, with the staining during diestrus and proestrus perhaps somewhat more intense than during the other stages. No specific staining was detected in the stroma, in sharp contrast to the results obtained with prepubertal rats [8], which showed significant staining of stromal cells immediately underlying the luminal epithelium. This staining disappeared with both eCG and estrogen treatment of the animals [8]. No muscle staining for CRBP was seen with these young animals either before or after treatment with eCG and estrogen [8], suggesting a possible developmental difference in the expression pattern for CRBP in the uterus. Incubation of tissue with antibody raised against CRABP produced staining of the inner and outer muscle layers of the uterus that was virtually indistinguishable from that seen for CRBP (data not shown). Also as with CRBP, no significant differences in the staining pattern for CRABP were observed for the various stages of the cycle. Interestingly, the localized stromal staining for CRABP seen after eCG and estrogen treatment of the prepubertal rat [8] was not observed in these normal cycling adults, although young rats do display similar muscle staining for CRABP both before and after PMSG treatment [8], similar to what was observed in this study. The most dramatic cycle variation was observed with CRABP II staining in the uterus. Significant staining for CRABP II was observed in the luminal epithelium during estrus with no staining apparent in the glandular epithelium, which is continuous with the luminal epithelium (Fig. 6B). Visible, but much less significant, staining was seen before and after estrus (Fig. 6, A and C), while virtually no staining of the luminal epithelium was apparent during diestrus (data not shown). Faint staining of the muscle of the uterus could also be seen, appearing to be most intense during diestrus. The epithelial staining during estrus correlates with a similar dramatic increase in staining of the luminal epithelium following eCG and estrogen treatment of prepubertal rats [8]. The intense epithelial staining seen in the adult rat was not seen in all sections of estrous uteri, suggesting that high levels of CRABP II expression may be specific to particular regions of individual uteri or that expression may be time-dependent and occur in a short pulse during estrus, with not all uteri collected at a time coinciding with this pulse of expression. In the ovary, staining for CRABP II was strongest for the granulosa cells of mature antral follicles (Fig. 6E). This particular staining also seemed to vary with stage, although not as dramatically as seen in the uterus, with the staining appearing to be most intense during estrus and least intense during metestrus and diestrus. A gradient of staining was associated with many follicles, with the granulosa cells nearest the exterior of the ovary appearing darker than those near the interior of the ovary. Significant staining for CRABP II was also displayed by the cells of some corpora lutea (Fig. 6F). Small corpora lutea typically showed uniform staining of all cells, while occasionally displaying a gradient of intensity similar to that described for the follicles. Larger corpora lutea possessed scattered clusters of staining cells, giving them a speckled appearance, while no staining of any luteal cells was observed for the largest,

most mature corpora lutea. Significant staining of the stroma was also often observed at all stages of the cycle, although usually of lower intensity than that displayed by the granulosa or luteal cells. However, the thecal cells surrounding the follicles were never seen to stain for CRABP II. Incubation of ovarian sections with antibody specific for CRBP showed intense staining of oocytes for all stages of oocyte development (Fig. 6H). Fainter staining of some granulosa cells was also observed-generally seen in cells in the interior of large, nonatretic antral follicles, with the granulosa cells adjacent to the basement membrane of the follicle showing little or no staining (Fig. 6G). Stromal cells in the area between the thecal cells of the follicle and the outer cellular layer of the ovary, when the follicle was close to the surface, occasionally showed staining (Fig. 6G). This was seen only with large, antral, nonatretic follicles and some early corpora lutea, and was most apparent during proestrus, diminishing in intensity with distance from the surface in a gradient similar to that observed with CRABP II in the granulosa cells. Because surface tissue adjacent to these areas, but not associated with a mature follicle, did not show such staining, it is unlikely that this edge staining was an artifact; it may instead indicate the involvement of CRBP in the cellular changes that must occur in this area before ovulation. DISCUSSION The expression of CRABP II in the uterus of the normally cycling adult rat was found to increase sharply during estrus, a time when the effects on target tissues of estrogen unopposed by progesterone are reaching a maximum. Such an apparent effect of hormone on CRABP II expression in the uterus was also seen with eCG- and estrogen-treated prepubertal rats, where CRABP II levels were found to be low before treatment but increased significantly 48 h after eCG treatment [8]. Estrogen treatment of these young rats induced CRABP II expression within 1 h, with levels rising steadily for the subsequent 48 h studied. We have observed retinoic acid production in the uterus after eCG treatment, localized to the epithelial cells, the site of CRABP II expression in the uteri of both the treated prepubertal animals and the normal adult animals (unpublished results). One possible explanation for such a colocalization of CRABP II expression and retinoic acid synthesis is the participation of CRABP II in the metabolic production of retinoic acid in these epithelial cells. A sharp peak of CRABP II expression during estrus accompanied by an increase in retinoic acid synthesis within the epithelium at that time could result in altered transcription within the epithelial cells themselves or the neighboring stromal cells, possibly triggering a transition to the next phase of the reproductive cycle. The expression of large amounts of CRABP in the muscle layers of the uterus, which appears to peak just before estrus with the protein levels remaining high through estrus, could then be explained by CRABP serving a role as a protector of the muscle cells from the effects of any retinoic acid produced by the epithelium. CRABP has been shown in other cell types to be excluded from the nucleus [8]. How such exclusion of this relatively small protein occurs is not yet known, but its high capacity for binding retinoic acid combined with such nuclear exclusion suggests an effective defense against the penetration of retinoic acid to the nu-

VITAMIN A-BINDING PROTEINS IN RAT UTERUS AND OVARY cleus of these muscle cells and the altering of gene transcription by its binding to nuclear retinoic acid receptors. The increased expression of CRABP in a localized region of the stroma, seen after eCG and estrogen treatment of prepubertal rats [8], was not observed in the normal adult rats, perhaps pointing out the inherent difficulties in comparing the possibly longer-term effects of hormone treatment of prepubertal animals with the results obtained when studying normal cycling adults, which undergo a more rapid and less well-defined flux of hormones during the 4- to 5-day estrous cycle. The CRBP also shown to be present in the muscle cells of the uterus, rather than serving a protective function, may well be participating in the storage of vitamin A in these cells in the form of retinyl esters. The uptake of retinol and its subsequent esterification have been shown in other tissues to be mediated by CRBP [16]. In addition, the muscle layers of the small intestine have been shown to possess large stores of retinyl esters, CRBP, and the enzyme activity responsible for the esterification of retinol [17]. The uterine muscle may serve a similar role as a local site of storage of retinol in an organ that appears to carry out extensive retinoid processing. Interestingly, CRBP expression in the muscle was not observed in the uteri of prepubertal rats, but expression of both retinol-binding protein (RBP) and CRBP was detected in the stroma of these uteri [8], suggesting that there is transport and processing of retinol in the stroma during organ maturation for a purpose no longer required in the adult animal. This stromal staining for CRBP disappeared upon treatment of the prepubertal rats with eCG and estrogen. Thus, the lack of staining seen in this artificially matured uterus is mirrored by a lack of stromal staining in the uteri of the normal adult. The expression of CRABP II in the ovary is somewhat more complex than in the uterus, in large part because in a normally cycling animal the ovary contains follicles at all stages of development at any one time, and the expression of CRABP II appears to be more dependent upon the stage of development of the individual follicle than upon the overall stage of the estrous cycle. CRABP II staining was found to be very strong in the granulosa cells of the most mature follicles. Such staining was also seen after the rapid follicle maturation induced by eCG and estrogen treatment of prepubertal rats [7]. The granulosa cell staining in the normal cycling adult appeared to be more intense around the time of estrus as opposed to other times during the cycle, but the difference was not as significant as that seen in the uterus. It is possible that in this case, as postulated for the uterus, CRABP II expression is a marker for retinoic acid synthesis. We have recently observed production of retinoic acid in the ovary, localized to the granulosa cells (unpublished results). Here too, the function of the retinoic acid is unclear, perhaps acting on the granulosa cells themselves, triggered once a particular stage of follicle development is reached, or possibly acting in a paracrine manner and diffusing away from the granulosa cells to affect gene transcription in neighboring cells in the ovary. Because no retinoic acid-binding protein staining was observed in the thecal cells surrounding the developing follicles, this suggests that these cells might be a target for the retinoic acid produced by the granulosa cells. Like CRABP, CRABP II has been seen to be excluded from the nucleus of some cell types [7], suggesting a possible protective role for this binding protein in the stromal cells of the ovary similar to that proposed for CRABP in the muscle cells of the uterus. Whether or not such a role

131

is carried out by one of the retinoic acid-binding proteins may be a function of cell type and the presence or absence of a second factor in the cell that binds to and prevents the passage of the retinoic acid-binding protein (and retinoic acid) into the nucleus. However, it is also possible that the presence of CRABP II in the stroma of the ovary, rather than being protective, may be an indicator of retinoic acid synthesis in these cells as well as in the granulosa cells. CRABP II expression was also seen in corpora lutea, particularly the smaller ones, suggesting that the role of CRABP II and possibly retinoic acid in the follicular granulosa cell is one that continues to be necessary for at least a short time after the functional transition of these cells to progesterone-producing luteal cells. The role of CRBP in the ovary may be similar to that proposed in the uterus-participating in the storage of vitamin A in the developing oocyte. The metabolites of retinol are known to play an important role in embryonic development, and both retinol and RBP have been detected in the yolk of chicken oocytes [18]. The presence of RBP suggested the possible internalization by the oocyte of retinol-RBP from the circulation during oocyte development. The CRBP detected here in the rat oocyte may indicate a similar uptake of retinol during development, although possibly via a different mechanism. The presence of CRBP in the cumulus granulosa cells may point to the participation of these metabolically coupled cells in the uptake and/or processing of vitamin A. In a manner similar to that in the testis, retinol may be delivered to the oocyte via these interior granulosa cells after uptake of retinol from RBP. The CRBP staining seen occasionally in the stroma separating mature follicles from the exterior of the ovary may or may not also play a role in the uptake of retinol by the oocyte. The location of this staining suggests a possible alternative role for CRBP in the cellular alterations that must occur in the outer stroma before ovulation. The cellular events occurring in the reproductive tissues of the rat during the estrous cycle obviously represent a complex system, but the cellular localization of CRBP and CRABP II in the ovary, as well as the dramatic variation in binding protein expression in the uterus during the cycle, suggest that these proteins and their ligands are involved in vital processes such as the transition of the uterus from the proliferative to the secretory phase, and follicle development and ovulation in the ovary. The need for further investigation is obvious but should be aided by knowledge of cellular location and variation of level of the binding proteins studied here. ACKNOWLEDGMENTS The authors wish to thank Bharati Kakkad for the purification of recombinant CRABP II and Michael H. Melner, Ph.D. (Dept. of Obstetrics and Gynocology and Dept. of Cell Biology and Anatomy, Vanderbilt University), for his generous offering of advice and equipment. Sarah A. Wardlaw, Ph.D., was supported by USPHS training grant HD 7043 during the course of this work.

REFERENCES 1. Thompson JN, Howell JM, Pitt GAJ. Vitamin A and reproduction in rats. Proc R Soc Lond 1964; B159:510-535. 2. Bo WJ. The relationship between vitamin A deficiency and estrogen in producing uterine metaplasia in the rat. Anat Rec 1956; 124:619628. 3. Ragins AB, Popper H. Variation of vitamin A fluorescence in the cyclic changes of the ovary. Arch Pathol 1942; 34:647-653. 4. Laurence PA, Sobel AE. Changes in serum vitamin A level during

132

5.

6.

7.

8.

9. 10.

WARDLAW ET AL.

the human menstrual cycle. J Clin Endocrinol & Metab 1953; 13: 1192-1198. Loughney AD, Kumarendran MK, Thomas EJ, Redfern CPE Variation in the expression of cellular retinoid binding proteins in human endometrium throughout the menstrual cycle. Mol Hum Reprod 1995; 10:1297-1304. Tannous-Khuri L, Hillemanns P, Rajan N, Wright TC, Talmage DA. Expression of cellular retinol- and cellular retinoic acid-binding proteins in the rat cervical epithelium is regulated by endocrine stimuli during normal squamous metaplasia. Am J Pathol 1994; 144:148-159. Bucco RA, Melner MH, Gordon DS, Leers-Sucheta S, Ong DE. Inducible expression of cellular retinoic acid-binding protein II in rat ovary: gonadotropin regulation during luteal development. Endocrinology 1995; 136:2730-2740. Bucco RA, Zheng WL, Wardlaw SA, Davis JT, Sierra-Rivera E, Osteen KG, Melner MH, Kakkad BP, Ong DE. Regulation and localization of cellular retinol-binding protein, retinol-binding protein, cellular retinoic acid-binding protein (CRABP), and CRABP II1in the uterus of the pseudopregnant rat. Endocrinology 1996; 137:3111-3122. Rugh R. The Mouse: Its Reproduction and Development. Minneapolis, MN: Burgess Publ. Co.; 1968: 38-43. Yuan Y-D, Carlson RG. Structure, cyclic change, and function, vagina and vulva, rat. In: Jones TC, Mohr U, Hunt RD (eds.), Monographs

11. 12.

13. 14. 15. 16. 17. 18.

on Pathology of Laboratory Animals: Genital System. Berlin: Springer-Verlag; 1987: 161-168. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162:156-159. Putowski LT, Coi D, Mordacq J, Scherzer WJ, Mayo KE, Adasi EY, Rohan RM. In vivo hormonal regulation of insulin-like growth factor binding protein-5 mRNA in the immature rat ovary. J Soc Gynecol Invest 1995; 2:735-742. Glantz SA. Primer of Biostatistics, 3rd ed. New York: McGraw-Hill; 1992: 32-109. Ong DE, Crow JA, Chytil FC. Radioimmunochemical determination of cellular retinol- and cellular retinoic acid-binding proteins in cytosols of rat tissue. J Biol Chem 1982; 257:13385-13389. Bok D, Ong DE, Chytil FC. Immunocytochemical localization of cellular retinol binding protein in the rat retina. Invest Opthalmol & Visual Sci 1984; 25:877-883. Ong DE. Cellular transport and metabolism of vitamin A: roles of the cellular retinoid-binding proteins. Nutr Rev 1994; 52:S24-S31. Herr FM, Wardlaw SA, Kakkad B, Albrecht A, Quick TC, Ong DE. Intestinal vitamin A metabolism: coordinate distribution of enzymes and CRBP(II). J Lipid Res 1993; 34:1545-1554. Vieira AV, Schneider WJ. Transport and uptake of retinol during chicken oocyte growth. Biochim Biophys Acta 1993; 1169:250-256.