| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Cell Biology, Neurobiology and Anatomy, University of Cincinnati, Cincinnati, Ohio 45267
b Department of Pediatrics and the Reproductive Sciences Program, University of Michigan, Ann Arbor, Michigan 48109
ABSTRACT
Ovarian growth and development are critically dependent upon the influence of endogenous estrogens, and both are highly regulated during the reproductive cycle. The observation that estrogen-receptor-
-deficient mice still exhibit follicular growth and development, together with other evidence, suggests that responsiveness of the ovary to estradiol occurs predominantly through the second estrogen receptor, ERß. We characterized the physiological regulation of ERß expression in ovarian follicles during the follicular phase of sheep that were synchronized for estrus during the breeding season with intravaginal progesterone implants (controlled internal drug release [CIDR] device; InterAg, Hamilton, New Zealand). Ovaries were removed at times corresponding to the early (EF) and late follicular phases (LF) of the ovine estrous cycle (12 h [n = 5] and 32 h [n = 5] after CIDR device removal, respectively). Sections of ovary were then hybridized with a cRNA probe corresponding to the 5' region of ovine ERß. ERß mRNA expression within the granulosa layer of different size follicles (size classes: 
3 mm, 3.1–4.0 mm, 4.1–5.0 mm, >5 mm) was quantified. ERß mRNA expression varied both with follicle size (P < 0.01) and with cycle stage (P < 0.01). In EF ewes, the highest levels of ERß mRNA expression were found in follicles 3 mm in size. ERß mRNA expression declined progressively thereafter among the different size classes with lowest levels expressed in >5-mm follicles. By contrast, expression of ERß mRNA in the 3.1- to 4.0-mm follicles of LF group was significantly higher than in the 3-mm size follicles and declined thereafter progressively to the >5-mm size levels as in the EF group. Furthermore, expression of ERß mRNA in 3-mm size follicles of LF group was significantly lower than the corresponding size class in the EF group. Lower expression of ERß mRNA in >5-mm follicle is suggestive of a down-regulation by the local estrogen milieu.
estradiol, estradiol receptor, follicle, granulosa cells, ovary, ovulatory cycle, steroid hormone receptors
INTRODUCTION
Growth, development and maturation of ovarian follicles are all critically dependent upon the actions of estrogens. A primary action of estradiol (E2) is to synergize with FSH in inducing the maturation of granulosa cells by increasing their proliferation and subsequently increasing expression of gonadotropin receptor levels [1–6]. Estradiol acts through specific binding proteins widely expressed within the reproductive tract. Indeed, many of the physiological actions of E2 occur via a classic nuclear receptor, ER, which is distinct from the newly discovered second receptor subtype, ERß [7]. Surprisingly, follicles of mice lacking functional ER still mature [8] and these mice can be induced to ovulate with exogenous gonadotropins [9, 10]. By contrast, the mice lacking the second receptor, ERß, are infertile or subfertile [11], fail to respond to superovulation or respond with fewer oocytes being shed [10, 11], and exhibit more early atretic follicles [12]. Observations that dual (ER/ß) deficient mice, but not ER knockouts, have antral follicles with few granulosa cells and fail to produce preovulatory follicles [11] lend further support to the hypothesis that many of the actions of E in the ovary are being mediated by the functional ERß.
What role ERß plays in follicular development, maturation, and its regulation during different physiologic states is just beginning to be understood. Given the importance of timing of ovulation and the role of E2 in follicular development, it is essential to elucidate the precise expression pattern of ERß mRNA during the preovulatory period of the estrous cycle, a period when follicular growth and development are both undergoing profound changes.
The sheep estrous cycle provides an excellent model system for understanding the regulation of ERß expression. Many of the reproductive endocrine events associated with ovulation in this model have been elucidated [13–15]. Availability of approaches to synchronize estrous cycles provides a means to time, with precision, events occurring during the cycle. For example, the use of a controlled internal drug release (CIDR) device, an intravaginal progesterone pessary, can routinely produce >90% estrus synchronization in sheep and other domesticated species [16, 17]. The endocrine events associated with this synchronization regimen have also been well characterized [18]. We capitalized on this synchronization approach to determine if changes in expression of ERß mRNA occur during the follicular phase of the ovine estrous cycle.
MATERIALS AND METHODS
Animals
Ovaries were obtained from adult Suffolk ewes (n = 10) as part of an earlier study [18]. Ewes were maintained outdoors and under natural photoperiod at the Sheep Research Facility, Ann Arbor, MI. Animals were fed hay and had free access to water and mineral licks.
Estrous Cycle Synchronization and Tissue Collection
Details of the synchronization approach have been described earlier [18]. Briefly, ewes were treated with prostaglandin (PG)F2 (Lutalyse; Upjohn, Kalamazoo, MI) to induce luteal regression and studied during three consecutive cycles following synchronization of estrus with 1 (cycle 1) or 2 (cycles 2 and 3) CIDR (type G, 9% natural progesterone; InterAg, Hamilton, New Zealand) devices. Endocrine data obtained from cycle 2 (timing cycle) were then used to time the removal of ovaries during the third cycle (experimental cycle). During this cycle, five ewes were killed 12 h after CIDR removal (corresponding to the early follicular phase) and five ewes were killed 32 h after CIDR removal (corresponding to the late follicular phase). Blood samples collected at 2-h intervals from 0 to 12 h prior to sacrifice were utilized to monitor changes in circulating estradiol concentrations. At the time the ewes were killed, ovaries were removed, rapidly frozen in a bath of 2-methylisobutane (cooled to a slush with liquid nitrogen), and stored at -80°C until sectioned. Sections (12 µm) were cut on a cryostat, thaw mounted on SuperFrost plus slides (Fisher Scientific, Pittsburgh, PA) and stored at -80°C until processed for in situ hybridization (ISH). All described experimental procedures were approved by the University of Michigan Committee on the Use and Care of Animals and were performed in accordance with the Guide for the Care and Use of Laboratory Animals.
Estradiol RIA
Estradiol assays were performed on 200-µl duplicates of serum extracted with 2 ml anhydrous ethyl ether (Fisher Scientific), using a modification [19] of a commercially available RIA kit that utilizes magnetic particle separation (Estradiol MAIA, Serono, Italy). Median variance ratio was 0.1. Interassay coefficient of variation (n = 5) based on three quality control sera was 23.4%. Detection limit (2 standard deviations of buffer control) of the E2 assay was 0.3 ± 0.1 pg/ml.
Ovine ERß Cloning
To quantify ERß mRNA expression in the sheep ovary by ISH we first isolated a cDNA fragment of ovine ERß using reverse transcription-polymerase chain reaction (RT-PCR) with primers specific for the 5' region of the human ERß sequence (accession no. AF051427) [20], bases 1086–1323. This region shares little sequence identity with ER [21] but is similar among the various ERß isoforms identified to date [20]. Primers used were upper: 5'-GATGTGGGTACCGCCTTGTGC-3' and lower: 5'-GGCCAACTTGGTCAGGGACA-3'. Total RNA was extracted from two sheep pituitary glands and then 2 µg of RNA was reverse transcribed using avian myeloblastosis virus-RT (30 U; Promega, Madison, WI). This was followed by PCR using Taq polymerase (Life Technologies, Rockville, MD) for 35 cycles (Amplitron II thermocycler) with an annealing temperature of 52°C and a 2-min extension time. A predicted 239-base pair (bp) PCR fragment was isolated from low-melt agarose (Life Technologies), ligated into PGEM-T Easy vector (Promega), and subcloned in JM109 competent cells (Promega). The ERß sequence was determined using T7 and SP6 primers and screened for sequence identity with the BLAST search tool (NCBI).
In Situ Hybridization
Antisense and sense RNA fragments were generated from linearized templates (2 µg) using T7 and SP6 RNA polymerases (Promega), respectively, in a transcription reaction containing: 5x transcription buffer (Promega), 5 mM rNTPs (GTP, CTP, ATP), 100 µM UTP, RNase inhibitor, 100 mM dithiothreitol (DTT), and 100 µCi 33P-UTP (3000 Ci/mmol; NEN) for 2 h at 37°C. Samples were then incubated with DNase-I (Sigma, St. Louis, MO) at 37°C for 15 min to remove template. The reaction was stopped with 0.5 M EDTA. Unincorporated isotope was removed using spin columns (Roche, Indianapolis, IN). Probes were diluted in TE (Tris, EDTA; containing 10 mg/ml Torula RNA; Sigma), denatured at 90°C for 3 min, and combined with hybridization buffer containing 50x Denhardt solution, deionized formamide, 50% dextran sulfate, 5 M NaCl, 1 M Tris, pH 8.0, 0.5 M EDTA, and 1 M DTT. Labeled probes were stored at -80°C until applied to slides.
Just before use, slides containing tissue sections were removed from the freezer, placed in racks, and dried with a hair dryer for 30–45 sec. Sections were then fixed for 5 min in 4% paraformaldehyde in phosphate buffer (PB, pH 7.4) at room temperature and then rinsed twice in PB, with stirring. Then the slides were dipped briefly in water, then triethanolamine (TEA), and finally, soaked in TEA plus acetic anhydride for 10 min, with stirring. Next, the slides were soaked in 2x saline-sodium citrate (SSC: 0.15 M NaCl, 0.015 M sodium citrate), 70%, 95%, and 100% ethanol, respectively, followed by delipidation in chloroform. Slides were then soaked for 3 min each in 100% and 95% ethanol and then air-dried. Once dried, hybridization buffer containing labeled probe (1 x 106 cpm/slide) was applied to the slides in a volume of 100 µl. Slides were then coverslipped with Parafilm, sealed with rubber cement, and hybridized overnight at 56°C in Tupperware chambers containing damp paper towels. On the following day, coverslips were removed, and the slides were reracked and washed immediately for 15 min in 4x SSC, twice, with stirring. Slides were then incubated for 30 min at 37°C in RNase buffer (1 M Tris, pH 8.0, 5 M NaCl, 0.5 M EDTA) containing RNase-A (0.03 mg/ml; Sigma) followed by incubation for 30 min at 37°C in buffer without RNase-A. Next, the slides were washed at room temperature under increasingly stringent conditions (2x SSC, 0.1x SSC [69°C, twice, 30 min each], 0.1x SSC [room temperature, twice]). Finally, slides were dehydrated for 3 min each in 50%, 85%, and 100% ethanol and then air-dried. Slides were then dipped in Kodak NTB3 emulsion, dried, and exposed in the dark for 6 wk at 4°C. The emulsion was developed in Dektol (Kodak), the sections were counterstained with cresyl violet, and then coverslipped with DPX (EM Sciences, Fort Washington, PA).
Messenger RNA Quantification and Granulosa Cell Counts
ERß mRNA was quantified using the National Institutes of Health (NIH) Image program from digital images captured at 50x magnification using a Nikon Microphot microscope equipped with Dage video camera and Scion LG-3 screen-grabber card (Scion Corp., Frederick, MD) in an Apple G3 computer. Each image was integrated over 16 frames, captured, and saved in TIF format. For every follicle within a section, the number of silver grains (expressed as pixels) was determined using an area of 50 x 50 pixels in four selected quadrants perpendicular to one another and overlying the granulosa layer. The size of area to be analyzed (i.e., 50 x 50 pixels) was chosen after a random sampling of large preovulatory follicles with the thinnest granulosa layers indicating that this was sufficient to cover virtually the entire thickness from theca layer to antrum. Follicles were then grouped into classes based on size (3 mm, 3.1–4 mm, 4.1–5 mm, and >5 mm), and the density of silver grains was determined for each follicle within a given size class. Then, for follicles of a particular size the mean level of expression was determined from the four quadrants. For each follicle in which ERß mRNA was quantified, the same four 50 x 50 pixel quadrants were also used to determine the number of granulosa cells. Manual counts were made with the aid of an ocular grid. Group (follicle size and cycle stage) means were calculated for each parameter. Slides were coded and the data were generated without prior knowledge of cycle stage.
Cell counts and mRNA analyses were performed on follicles that were not overtly atretic. That is, any follicle in which gross disruption of the granulosa layer was observed within two to three layers of the basal lamina and in which invagination of the theca layer into the granulosa cell layer occurred (i.e., tertiary atresia) was not included.
Statistical Analysis
Statistical comparisons were made by two-way ANOVA (main effects: cycle stage, follicle size; interaction: cycle stage x follicle size). Posthoc comparison of group means was performed by Fisher least significant difference analysis. Differences were considered statistically significant if P < 0.05.
RESULTS
Accuracy of Predicted Follicular Phase
Based on information from the timing cycle (cycle 2) [18], mean time of collection of early follicular (EF) and late follicular (LF) group ovaries was estimated to occur approximately 34 h before the expected LH surge peak (early follicular) and 14 h before the expected LH surge peak (late follicular), respectively. As expected, none of the 10 ewes in these groups showed an LH surge during this period. However, based on serum estradiol data obtained from samples collected prior to the time the ewes were killed, one of the early follicular phase ewes was found to be in the late follicular phase. This ewe showed the characteristic preovulatory increase in estradiol secretion, and accordingly, she was grouped with the late follicular group for analysis of ERß mRNA expression. Mean serum estradiol concentrations increased progressively in both EF and LF groups (Fig. 1), but the amplitude of the estradiol increase was much greater in LF ewes, reaching preovulatory concentrations (3.51 ± 0.15 versus 1.31 ± 0.21 pg/ml, respectively; P < 0.05).
|
Ovine ERß Sequence
Using primers against human ERß, we isolated a portion of the ovine ERß cDNA from sheep pituitary (pituitary gland expresses ERß mRNA in human [22] and rat [23]). The isolated sequence (accession no. AF110817) shared significant (>80%) identity with human and nonhuman primate, bovine, rat, and mouse sequences (Fig. 2). In addition, two ovine sequences (GenBank accession nos. AF177936, AF257109; not shown) share complete identity with both our nucleotide and predicted amino acid sequences. Even though the cloned sequence used in these studies corresponds to the hypervariable hinge domain (D) to ensure specificity for ERß [21], it cannot be determined precisely which of the five possible ERß isoforms [20] was detected. Isoforms 1, 2, 4, and 5 all share significant sequence homology and are expressed in the ovary [20]. Additionally, observations from our laboratory (unpublished results) of specific ERß expression in sheep brain, a tissue expressing only ERß-1 [20], suggests that our sequence is likely detecting ERß-1; however, other isoforms are probably also being recognized.
|
Expression of ERß mRNA in Ovarian Follicles During the Follicular Phase of the Ovine Estrous Cycle
Utility of the cloned sequence was confirmed by the presence of specific and intense signal over the granulosa layer of growing and mature follicles when using an antisense probe and by the absence of specific labeling using the sense probe (Fig. 3). Specific signal was also present, but to a lower extent, over the theca cell layer (Fig. 4). Examination of corpora lutea also revealed low intensity signal (not shown).
|
|
Analysis of ERß mRNA during the follicular phase revealed that expression varied both with follicle size (P < 0.01) and with stage of the cycle (P < 0.01). In EF ewes, the highest levels of ERß mRNA expression were found in follicles 3 mm in size (Figs. 4 and 5). ERß mRNA expression declined progressively thereafter among the different size classes with the lowest levels expressed in follicles >5 mm in size (Fig. 5).
|
As was observed for the EF group, ERß expression in 3.1–>5-mm follicles of LF ewes declined progressively (Fig. 5). In contrast to EF follicles, expression of ERß mRNA in follicles 3 mm in size of the LF group was significantly (P < 0.01) lower than in the 3.1–4- and 4.1–5-mm size follicles (Figs. 4 and 5). However, ERß mRNA levels in the largest (>5-mm) follicles were not significantly different from the smallest (3-mm) follicles (Fig. 5). Comparisons between cycle stages revealed that ERß mRNA expression in 3-mm follicles of the LF group was significantly lower than the corresponding size class of the EF group (P < 0.01; Fig. 5). ERß mRNA expression in all other size follicles was similar between EF and LF groups.
As shown in Table 1, granulosa cell density was significantly greater in follicles 3 mm when compared to >5-mm follicles. This trend was the same regardless of cycle stage. No other differences in granulosa cell density were observed either within or between groups.
|
DISCUSSION
The present study was performed to examine the expression pattern of ovarian ERß mRNA prior to ovulation in the sheep. Our results reveal highly specific expression of ERß in the granulosa cell layer of small and developing follicles, confirming earlier observations made in cow, mouse, rat, and human [7, 9, 22, 24]. We observed weaker, yet specific, ERß signal overlying the theca layer; a finding consistent with some studies [9, 25, 26] but not others [27, 28]. Expression of ERß in theca cells could represent an important mechanism for the transition between ovulatory and luteal states. Our finding of very low ERß mRNA expression over the granulosa and theca cells of the corpus luteum suggests that these cells have lost their expression in an estrogen-dependent fashion.
In the present study, follicles were classified based on size distribution (<3.0 mm, 3.1–4.0 mm, 4.1–5 mm, >5 mm). Previous reports have shown that follicles up to 4 mm in diameter express FSH receptors in granulosa cells and LH receptor in theca cells [29–31]. Follicles >4 mm in diameter are estrogen-active and follicles >5 mm are presumptive preovulatory follicles, expressing aromatase and LH receptors in granulosa cells. While our experimental protocol does not allow for a definitive conclusion regarding the role of estradiol or LH in causing the decrease in ERß expression, based on the directionality of changes and estrogenic status and low signal in the estrogenic preovulatory follicle, one could conclude that this is an outcome of down-regulation to increasing intraovarian estradiol concentrations. Use of a selective ERß antagonist such as tetrahydrochrysene [32, 33] during the late follicular phase when LH pulse frequency is increasing would help in testing this premise.
In addition to describing the anatomical localization of ERß mRNA in the sheep ovary, two important findings related to estrous cycle regulation were made: 1) ERß expression declines with increasing follicle size during the early follicular phase, and 2) during the late follicular phase, follicles 3 mm in diameter express less ERß as compared to those in the early follicular phase. A similar finding was made recently in rats by Bao et al. [26], who showed that small follicles in ovaries collected during the afternoon of proestrus express less ERß mRNA than medium-sized follicles; this change correlated with changes in thecal LH receptor expression.
Although the precise cause of the cycle stage-dependent down-regulation in follicles 3 mm in diameter remains unclear, several possibilities exist. Because the highest level of ERß mRNA occurs in the small follicles during the EF phase, the existence of an active inhibitory process is a likely possibility. The synchronization protocol itself may have influenced the expression pattern of ERß. A final, and more likely, explanation favors the role of either gonadotropins or gonadal steroids in the apparent downregulation of ERß in follicles in this size class (3 mm) during the late follicular phase. It is known that GnRH/LH pulse frequency, androgen and estrogen production, and aromatase activities are all elevated late in the follicular phase [13, 15, 34–36]. Increased LH pulse frequency that occurs during the late follicular phase and the consequent increase in androgen production by the theca cells may inhibit expression of ERß by facilitating atresia of these smaller follicles. While androgen production is not vital to the normal function of preovulatory follicles, androgens can arrest follicular maturation or induce atresia in smaller ones [25, 37, 38]. Moreover, development of small follicles prior to antrum formation is thought to be gonadotropin independent [5], suggesting that intraovarian rather than extraovarian influences predominate, perhaps via a pronounced reduction in local estrogen synthesis by follicles destined for atresia. The observation that ovaries of mice lacking ERß contain more follicles that are atretic and fewer corpora lutea [12] provides strong evidence that ERß plays an important role in follicular development and maturation.
It is well established that granulosa cells undergo several doublings [30] and follicles destined for atresia show some disaggregation [39, 40]. As such, differences in cell density could account for the observed differences in ERß expression. While significant difference in cell numbers were found between the smallest (3-mm) and largest (>5-mm) follicles, the progressive decline in ERß mRNA expression that we observed in the 3.1–4-, 4.1–5-, and >5-mm follicles could not be accounted for by the loss of granulosa cells. Moreover, despite having nearly identical granulosa cell densities, the smallest follicles of both groups expressed vastly different amounts of ERß mRNA. It should be emphasized that in quantifying ERß expression we excluded all follicles that were clearly atretic. While we may have missed and included in this characterization follicles showing early signs of atresia, considering atresia occurs in all size classes of follicles [40], this would be a constant source of error across different size classes of follicles.
Because ER protein and mRNA is absent, or expressed at exceedingly low levels in granulosa cells [26, 28, 41, 42], elucidating the factors involved in regulation of ERß mRNA in ovary is an important step in understanding follicular development. Recent cloning of the human ERß promoter sequence revealed the presence of several important regulatory regions. An imperfect Alu estrogen response element (ERE) is found upstream of the transcription initiation start site [43]. This ERE appears to confer estrogen-dependent activation in transfected cells [43] and in breast cancer cell lines [44]. Importantly, this ERE is also found in the breast cancer susceptibility gene, BRCA-1 [45]. In addition, binding sites for SP-1, AP-1, Oct-1, and a negative regulatory element (NRE) have also been identified in the ERß promoter [43].
During the early follicular phase, modest increases in local (ovarian) estrogen concentrations may act in an autocrine/paracrine fashion through the ERE to increase ERß mRNA in those small follicles recruited into the next wave of folliculogenesis but not undergoing atresia. The shift to then specifically down-regulate ERß in small follicles later in the follicular phase may be a function of the high intraovarian estradiol concentrations originating from the highly estrogenic preovulatory follicle. Potent down-regulation of ERß mRNA has been shown to occur in granulosa cells of immature rats following exposure to the synthetic estrogen, diethylstilbestrol [27, 46]. It should be cautioned however, that entirely different factors might be involved in the apparent down-regulation of ERß in large follicles when compared to small follicles. For example, the state of differentiation of granulosa cells might also be an important determinant of the ability of individual follicles to become responsive to estrogens or gonadotropins [42, 46]. Thus, whether granulosa cells reside within a follicle or corpus luteum, or whether they belong to a growing or atretic follicle may influence their estrogen-dependence and consequent expression of ERß. Indeed, undifferentiated granulosa cells express higher levels of ERß protein and mRNA when compared to those induced to differentiate with FSH and testosterone [42].
In summary, the present study highlights the dynamic regulation of ERß mRNA in sheep ovary. Like rats and primates, the expression of ERß mRNA over granulosa cells declines with increasing follicle size. However, during the late follicular phase of the estrous cycle, an additional level of regulation seems to occur that influences small follicles specifically. How changes in ERß mRNA expression impact on the mechanisms of follicle recruitment, maturation, or atresia, and how the dynamic regulation of ERß expression is controlled remains to be elucidated.
ACKNOWLEDGMENTS
The authors are indebted to Ms. Judy Van Cleeff for the conduct of the main synchronization study from which the ovaries were obtained. We acknowledge the support of Doug Doop and Gary McCalla for the care and maintenance of animals.
FOOTNOTES
First decision: 12 March 2001.
1 This work was supported by NIH HD23812 to V.P. and U.S. Department of Agriculture 98-35203-6321 to H.T.J., Standards and Reagents, Morphology and Sheep Research Core Facilities of the Center for the Study of Reproduction (NIH P30-HD18528) at the University of Michigan. 
2 Correspondence: Heiko T. Jansen, Dept. VCAPP, Washington State University,
PO Box 646520, Pullman, WA 99164-6520. FAX: 509 335 4650;
heiko{at}vetmed.wsu.edu
3 Current address: Department of Veterinary and Comparative Anatomy,
Pharmacology and Physiology and the Center for Reproductive Biology,
Washington State University, Pullman, WA 99164-6520.
Accepted: April 30, 2001.
Received: February 28, 2001.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  J. L Juengel, D. A Heath, L. D Quirke, and K. P McNatty Oestrogen receptor {alpha} and {beta}, androgen receptor and progesterone receptor mRNA and protein localisation within the developing ovary and in small growing follicles of sheep Reproduction, January 1, 2006; 131(1): 81 - 92. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. Fraley and S. Ritter Immunolesion of Norepinephrine and Epinephrine Afferents to Medial Hypothalamus Alters Basal and 2-Deoxy-D-Glucose-Induced Neuropeptide Y and Agouti Gene-Related Protein Messenger Ribonucleic Acid Expression in the Arcuate Nucleus Endocrinology, January 1, 2003; 144(1): 75 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Pillai, J. M. Jones, and R. D. Koos Treatment of Rats with 17{beta}-Estradiol or Relaxin Rapidly Inhibits Uterine Estrogen Receptor {beta}1 and {beta}2 Messenger Ribonucleic Acid Levels Biol Reprod, December 1, 2002; 67(6): 1919 - 1926. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
