0013-7227/92/1301-0335$03.00/0 Endocrinology Copyright 0 1992 by The Endocrine

Recruitment Luteinizing Cycle* GWEN Department

V. CHILDS, of Anatomy

Vol. 130, No. 1 Society

Printed

in U.S.A.

and Maturation of Small Subsets of Hormone Gonadotropes during the Estrous GEDA

UNABIA,

and Neurosciences,

AND JONATHAN University

LLOYD

of Texas Medical

Branch,

Galveston,

Texas 77550

Because the overall percentage of immunoreactive LH cells did not change after diestrus, small LH cells may have enlarged or increased their density to join the medium-sized pool. During proestrus, the proportion of large immunoreactive LH gonadotropes increased from 41 -C 2% to 65 f 2% (by the morning of estrus) as the proportion of small or medium-sized LH cells declined to 17-18 & l%, suggesting further increases in size or density. These data suggest that small or medium-sized gonadotropes are activated during early diestrus to enlarge and produce LHP. They contribute to the increased number of cells in medium-sized and large fractions in proestrous or estrous rats. The predominance of the smaller subtypes during mete&us and diestrus suggests that LH gonadotropes may revert to a smaller or lighter subset to await activation during the next cycle. (Endocrinology 130: 335-344,1992)

ABSTRACT. Small and medium-sized gonadotropes may enlarge and produce more LH in order to contribute to the proestrous surge. To test this hypothesis, dispersed pituitary cells from cycling female rats were separated by centrifugal elutriation into small, medium, and large fractions and labeled for LH@ antigens or mRNA (by in situ hybridization with a biotinylated oligonucleotide probe complementary to sequences encoding amino acids 26-40). The percentage of cells bearing LHfl mRNA in the pituitary cell population increased from 6 f 0.4% in the evening of diestrous day 2 to 16 f 0.7% in the morning of estrus (average f SEM). Over 80% of these labeled cells were large or small subtypes. The proportion of small gonadotropes labeled with LHB mRNA declined from 43 f 3% at metestrus to 29 + 1% on the evening of proestrus as the proportion of mediumsized gonadotropes labeled for LH/3 antigens (15 f 1%) or mRNA (17 + 1%) increased to 25 f 2% or 38 f 2%, respectively.

G

and the small gonadotropes are reserve subsets. However, because previous studies were of separated cells from randomly cycling female rats (3), no correlations between the morphology and a physiological state (i.e. the stage of the cycle) could be drawn. Some gonadotropes in the smallest fractions respond to GnRH in vitro by enlarging (8, 9) and changing from a monohormonal to a bihormonal storage state (1, 2, 9, 10). Small gonadotropes could also be induced to bind GnRH by pretreatment with estradiol (10). An expression of a particular size or storage state does not necessarily mean that small gonadotropes have converted to the larger subset during the 3- to 4-h GnRH stimulation. A number of investigators have shown that the high rate of LH secretion characteristic of large gonadotropes has not been evident in the small gonadotrope populations (11-19). Perhaps key maturational steps are needed to support peak secretory activity. Such a maturational sequence might be seen if different sized subsets of gonadotropes are collected from rats in one stage of the cycle. Their responses could then be compared as the population is being prepared for the LH surge. A study of this type would only be feasible, however, after separation and analyses of cells from individ-

ONADOTROPES change storage patterns, shape, and size during the estrous cycle or as they are stimulated (l-lo). The average area of immunoreactive gonadotropes increases as they approach estrus, after which it is reduced (1, 2). Some cells become pleiomorphic during peak secretory activity, and their granules are found in processes near blood vessels (1). However, other cells remain ovoid or round and may or may not be well granulated. Are smaller gonadotropes in a reserve state? Previous ultrastructural studies (2, 3) demonstrated that the smallest gonadotropes contain few pleiomorphic granules (2, 3), like those from immature (neonatal) rats (2, 4). Gonadotropes from medium-sized cell fractions are packed with intensely labeled granules (3). The largest gonadotropes have small and large granules scattered among vacuoles or vesicular profiles of rough endoplasmic reticulum. Perhaps these are the mature forms Received July 11, 1991. Address all correspondence and requests for reprints to: Gwen V. Childs, Ph.D., Department of Anatomy and Neurosciences, University of Texas Medical Branch, 200 University Boulevard H-43, Galveston, Texas 77550 * This work was supported by a bridging grant from the Sealy-Smith Foundation (Galveston, TX) and NIH Grant ROl-HD-15472. 335

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

336

IN LH GONADOTROPES

ual pituitaries. Developments in counterflow centrifugation separation protocols have now made it possible to study different sized pituitary cells from individual rats (20). This is performed in a chamber designed by Sanderson et al. (21-23) for use with low numbers of cells (10,000 minimum, 10,000,000 maximum). We have used this chamber successfully to separate and enrich corticotropes (20). Therefore, we initiated studies of cell fractions from cycling rats separated by size and/or density in the Sanderson chamber. The objective was to determine if we could detect movement of gonadotropes into a smaller or larger pool as the cells involuted, enlarged, or changed density of hormone stores. This presentation begins by validating this separation protocol, showing that there is no selective loss of a particular cell population at any stage of the cycle. It then presents changes in the distribution of pituitary cells among the small, medium, and large fractions throughout the estrous cycle which suggest that some of the pituitary cells are indeed shifting to a larger subset as the cells approach estrus. The contribution of LH cells to this shift is tested further by labeling the cells for LHP antigens or mRNA and studying their distribution among the three fractions. The hypothesis that was tested is that small LH gonadotropes may synthesize LHB mRNA, enlarge, or change density of LH stores early in the cycle, and thereby shift to a larger cell fraction. A companion study of the secretory responses of these same cell populations is presented in a separate report (24). Materials Collection and dispersion

and Methods of pituitaries

Female Sprague-Dawley (200-250 g) rats were acclimated for 7-10 days, with food and water available ad libitum, in controlled lighting (on at 0600 h; off at 2000 h). Daily vaginal smearswere then begun, and rats were used only if they had completed two successfulcycles. The protocol has been approved by the University of Texas Medical Branch Animal Care and Use Committee (ACUC Protocol 89-10-220, ll/Ol/ 90). The times tested were the following: 0900 h on estrus, metestrus, diestrous day 2 (diestrous II), and proestrus; and 1400 h on proestrus. At least four to six rats per group were collected;cellsfrom eachpituitary were dispersedand elutriated separately (25, 26). In addition, pituitaries from five to seven rats per group were dispersedand plated without elutriation to compare secretion and hormone storage in gonadotropes in unseparatedcultures. An additional time point wasaddedwhen unseparatedcultures were tested (1400 h on diestrous II) in order to determine whether there was an early increment in the percentageof cells with LHP mRNA. After rats were killed by guillotine, the pituitaries were removedrapidly and placedin cold Dulbecco’sModified Eagle’s Medium (DME; Hazleton Biologics, Inc., Lenexa, KS) contain-

Endo. Voll30.

1992 No 1

ing 0.3% BSA (Sigma Chemical Co., St. Louis, MO) and 2.5 g/ 500 ml HEPES (Sigma Chemical). The neurointermediate lobeswere removed, and the anterior lobeswere choppedinto 8-12 pieces,washedin fresh DME, and placed in 0.5% trypsin. The dissociationprotocol was carried out as describedin previous studies (15, 20, 21, 25). Cells were tested for viability by the trypan dye exclusion test (0.25% trypan blue in DME). Normally, the protocol yielded 3-4 million cells/pituitary that were 98% viable. Just before elutriation, the cells were monodispersedwith gentle suction through an 18-gaugeneedle. Centrifugal

elutriation

technique

Pituitary cells from a singlerat were separatedin a Beckman 52-21 centrifuge (Palo Alto, CA) with the JE-6B elutriator rotor at 1920rpm at 6-8 C. The elutriator and tubing had been sterilized and equilibrated with the elutriation buffer, as described in our recent study (20). After loading at 8 ml/min, three major fractions were collected at 15 ml/min (small), 25 ml/min (medium),and 35 ml/min, as the centrifuge wasslowly stopped(large). The contents of the chamberwere pooledwith the largecell fraction. Cellscollected during loadingwerepooled with the small cell fraction. The averageyield from elutriation experiments was 70-80%, and the viability was97-100%. The percentagesof recoveredcells were calculated, and values from the experimental groups (i.e. one stage of the cycle) were comparedto learn if there wasany selective cell losswith the stage of the cycle. To determine whether the Sanderson chamber separatedby size, cells from each fraction were collected at 5 ml/min intervals, and a samplewas diluted in DME plus trypan blue and placed in a hemocytometer. At least 50 living cells/fraction were measuredimmediately with Bioquant System IV imageanalysis equipment (20) (R & M Biometrics, Nashville, TN). The cursor was used to draw around the perimeter of eachcell. Areas and diameters(longestdimension) were then calculated automatically by the computer. Plating

and stimulation

The unseparatedcells or different sized fractions were suspended in DME to which were added5 rg/ml insulin (Sigma Chemical),30 nM sodiumselenite (JohnsonMatthey Chemical, Ltd., New York, NY), 50 pg/ml transferrin (Sigma Chemical), and 4.2 rg/ml fibronectin (Chemicon, Temecula, CA). They were plated on glasscoverslips (A. H. Thomas Scientific, Suredesboro,NJ) in 24-well trays so that the plating density was at least 10,000cells/well. At least 90% of the cells adheredin 1 h at pH 7.6-7.8. As describedin other studies(9, 10, 24, 27), the dissociationprotocol doesnot causelossof binding activity or responsivenessto GnRH. The plating medium, which was alsothe diluent for the secretagogues to be usedin thesestudies, was DME containing 0.3% BSA and 2.5 g/500 ml HEPES buffer. Gentamicin was usedto prevent bacterial growth, and aprotinin (100 kallikrein inhibitor units; SigmaChemical)was usedas a proteaseinhibitor. One hour after plating, the cells were exposedto 0 or 1 nM [n-Lys’]GnRH in DME for 4 h. The cultures were divided into two groups. One group was designatedfor single labeling for LH@ antigens. The second group was prepared for in situ hybridization to detect LH/3 mRNA. Therefore, most of the experiments employed cells that had been separatedfrom the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

IN LH GONADOTROPES

337

in viuo environment for 6-6.5 h. In the groups from rats taken

at 1400h on proestrus,the time period spannedthe time of the expected proestrousLH surge (1). Previous studieshave demonstrated that LH gonadotropesretain+he memory of their in uiuo environment, including priming by steroids (28-31). However, the delay might have affected the percentagesof cells bearingmRNA. Therefore, in later experimentsadditional cells were collectedand fixed 1 h after plating (2-2.5 h after removal from the rat). Labelingprotocolsfor LHfl antigens The fixed cells were first exposedto 5% normal goat serum in phosphate buffer (pH 7.4) containing crystalline human bovine serum (2.5 mg/ml, Sigma Chemical) for at least 15 min. Then, the cells were exposedfor 2 h to antibovine LHP serum, diluted 1:40,000in the above solution (including normal goat serum and human serum albumin). The ABC Elite detection protocol (Vectastain ABC Kit, Vector Laboratories, Burlingame,CA) was then carried out with nickel-intensified diaminobenzidine (7, 26, 32). Controls involved omission of the primary antibody or absorption with LH or FSH (iodination preparations, NIDDK). Only LH absorption abolishedreactivity in the antiserum to bovine LH/3 (7,26). In situ hybridization with photobiotinyluted oligonucleotide probesto mRNA for LH/3 The in situ hybridization protocol to detect complementary biotinylated probesfor LH/3 mRNA wascarried out asdescribed previously (32, 33). Tests of the effect of the delay during the 4-h exposureperiod showed no differences in percentagesof labeled cells. Therefore, the data were pooled with data from cells grown for 4 h. After fixation in 2% glutaraldehyde for 30 min, as described in previous studies(32,33), the cellswere washedfor 1 h in 0.1 M phosphate buffer plus 4.5% sucrose (pH 7.4). A 39-mer oligonucleotide probe complementary to the portion of LHP mRNA encodingamino acids 28-40 wasproduced and purified by the University of Texas Medical Branch Nucleotide Synthesizing Laboratory. Its sequenceis 5’-GCT AGG ACA GTA GCC GGC ACA GAT GCT GGT GGT GAA GGT-3’. The probe wasphotobiotinylated (Photoprobe,Vector Laboratories) as previously described (32, 33). The in situ hybridization protocol was identical to that describedpreviously (33), except RNase was omitted becausethe probe was not a cRNA. The controls were also run aspreviously described(32, 33). Different concentrations of the biotinylated oligonucleotide probeswere tested to determine saturation conditions, and the results were comparedto those of studiesemploying the biotinylated antisensecRNA probes (8, 32, 33). Labeled cells from estrous or proestrous [morning (AM)] rats were analyzed for their area, labeling density, and averagearea of label per cells with imageanalysis protocols identical to those describedrecently (33). As in previous studies, saturation was detected after exposureto concentrations higher than 10 rig/ml. There were no differences in the average density of the label when populations from estrous and proestrous rats were compared. The averageareaof label in proestrousrats was slightly higher, however.

Each pituitary or each fraction produced cells distributed evenly among24-32 coverslips.Half of the coverslipsbore cells stimulated by GnRH. Six to 8 coverslips from each set were labeledfor LHB antigens,and 6-8 werelabeledfor LHj3 mRNA. The remaining 12-16 were labeled for FSH/3 antigens or mRNA. Data from these setsof CS will be describedin a later report (Childs, G. V., G. Unabia, and J. Lloyd, manuscript submitted). Analysis of the percentageof labeledcellswasperformed by randomly choosing fields at x10-20 magnification and then increasingthe magnification to x40-100 and counting the cells in the field in view. Usually 200 cells/coverslip were counted x 6-8 coverslips/set x 5-7 rats. BecauseGnRH did not stimulate changes in the percentagesof labeled cells, the percentages from the GnRH-treated set actually servedasa duplicatevalue. Therefore, up to 11,200 cells were analyzed per experimental group. The percentages of labeled cells from control or GnRHtreated coverslipswere averagedto obtain a value for an individual rat or fraction. However, the final values represented the average of cell populations from different rats. Thus the variance tested was that of five to sevendifferent populations of cells. The n for each experimental group (stageof the cycle; fraction) was the number of different rats (5-7). Analysis of variance followed by Duncan’s multiple range test (at the 5% level) were carried out to determine differencesbetweenexperimental groups.

Results Changes in percentages of LH gonadotropes in the unseparated cultures Counts of immunolabeled LH gonadotropes in unseparated cultures showed increased percentages on diestrous II AM, which remained high until the AM of estrus. At that time, they dropped and remained low until the AM of mete&us (Fig. 1). In contrast, the percentages of gonadotropes bearing LH/3 mRNA rose steadily from the AM of proestrus to e&us, when they reached a peak (Fig. 1) which was identical to the highest percentages of immunolabeled LH cells (proestrous PM). This was followed by a decline that reached a nadir on the AM of diestrous II. Figures 2, 3, and 4 illustrate populations

from rats on diestrous II AM, proestrous AM, and estrus. The strong labeling for mRNA in cells from estrous rats is evident, especially in the largest cells. Validation of the separation methodology

The cell counts in each elutriation experiment produced information about the percent recovery and the overall distribution of the cells among the three fractions. Values for rats in different stages of the estrous cycle were compared to test the reproducibility of the experiments and determine whether there was selective loss in a given fraction at a particular time of the cycle. When

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

338 Mixed

D-II

Stage

AN

D-II

LH GONADOTROPES

Endo. Vol130

l

1992 No 1

cultures -o-

-+ - Antigens

MO1

IN

PM

Pro

mRNA

AM

ot the ertrous

Pro

PM

cycle

FIG. 1. Counts of LH gonadotropes labeled for LHP antigens or mRNA. Percentages of immunolabeled LH cells increase significantly from lo-14% on diestrous AM. The higher percentages (up to 16%) persist until estrus (Est.), at which time they decline and become similar to those of metestrous (Met) cells. In contrast, percentages of LH gonadotropes bearing LHB mRNA are highest at estrus. There is a decline on mete&us and diestrous AM, followed by a steady increase to reach peak estrous values of 16%. t, Significantly different from nadir values. D-II, Die&us II; Pro, proestrus.

overall cell yields were compared, there were no differences from the stage of the cycle. Analyses of the average yield or percent recovery was 78 + 11% (+SEM) cells (99% viable). The recovery did not depend on the number of cells loaded, cells from small pituitaries (1.7-2 million cells) were recovered as readily as those from large pituitaries (3-4 million cells). Furthermore, when the distributions of cells in individual fractions were compared, there were no changes that correlated with the percent recovery (yield) or the number of cells loaded. However, there were significant differences with the stage of the cycle. Data from five or six rats per experimental group are shown in Table 1. Half of the cells from metestrous rats were eluted in the small cell fraction, and 30% were large. However, there was a significant increase in the proportion of small cells to 60% on diestrus II, followed by a decline to 41% of the population by estrus. This decrease was accompanied by an increase in the percentage of medium-sized cells in the AM of proestrus, followed by increases in the percentages of cells in the largest fractions in the PM of proestrus and the AM of estrus. The increase in the percentage of small cells from estrus to diestrus II was accompanied by a decrease in the percentage of large cells. The Sanderson chamber was also tested for its ability to separate cells by size. Previous studies of cells from randomly cycling female rats or male rats showed that the average diameter or area of cells in each fraction increased with the flow rate in either the standard cham-

FIG. 2. Unseparated the field labeled for (arrows). One labeled plane, however. Only LHP mRNA (b; arrow).

cells from rat taken on the AM of die&us II. In LH antigens (a), there are several labeled cells process in the lower left corner is out of the focal one labeled cell is evident in the field labeled for r, Red blood cells. Magnification, ~733.

ber (2,3,8,13) or the Sanderson chamber (20). However, these studies had not been performed with cells from individual female rats. When tests were repeated on cells from rats in three different stages of the cycle, there was a significant increase in the average cell diameter as the flow rate increased by lo-12 ml/min (Fig. 5). Slight differences were seen when different stages of the estrous cycle were compared. Figure 5 shows that the average diameter of cells from diestrous II rats eluted at 15 ml/ min was smaller than comparable fractions from proestrous (AM) or estrous (AM) rats. This agrees with the data presented in Table 1 showing the predominance of small cells in populations from diestrous rats. The first significant increase in the average diameter of fractions from diestrous II rats was seen when the flow rate was increased 12 ml/min (from 8 to 20 ml/min). Figure 5 also shows that cells from proestrous female rats eluted at the highest flow rates (35 ml/min or greater) were, on the average, l-2 pm larger than their counterparts from estrous or diestrous rats. This may

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

IN LH

GONADOTROPES

339 +

EEtru-O-

Diert

+

Proe

0.001

:

:

:

:

:

:

:

,

6

8

15

20

25

30

35

40

ch

Flow rate

(mls/min)

5. Average diameter of cells in each fraction collected at different flow rates. The y-axis is the diameter (longest dimension). Cells from rats taken in the AM of three different stages of the estrous cycle [estrus (Estru), diestrus II (D-II), and proestrus (Pro)] show identical profiles, except that small cells from diestrous II rats tend to be smaller, on the average. Proestrous rats tend to have larger cells in the largest fractions. Values are the average diameter + SEM. t, Average diameters significantly different from other populations. FIG.

FIG. 3. Unseparated pituitary cells taken from a rat on the AM of proestrus. The field is labeled for LH@ antigens, showing strong labeling in medium-sized cells. The inset shows a field labeled for LHP mRNA (arrowhead) in a large cell. r, Red blood cells. Magnification, ~733.

FIG. 4. Unseparated pituitary cells taken from rat on the AM of estrus labeled for LH@ mRNA. Three small cells have patches of label (arrows). Labeled cells are more numerous than in the population from diestrous rats shown in Fig. 2b. The insets show strong labeling in two larger gonadotropes. Magnification, ~733. TABLE 1. Distribution of total cells in small, medium-sized, and large elutriation fractions: changes with the stage of the estrous cycle

Fraction Small Medium Lame

Stage of the cycle Metestrus Diestrus II Proestrus AM Proestrus PM 49 + 5 60 + 3” 54 + 3 54 + 5 21 + 3 19 + 3 28 + 2” 19 * 3 30 f 4 21 f 1 18 f 2 27 f 3”

Estrus 41 + 4” 21 + 3 38 f 4”

Values are a percentage of the total cells recovered + SEM (n = 5-6 rats/group). “Significantly different from value in metestrous rat cell populations.

stem from increases in size or density in the proestrous rat gonadotropes (3). Tests of this hypothesis essentially

formed the basis for this study. The next series of tests determined whether the fractions contained percentages of gonadotropes similar to those in previous studies (3, 13) in which small fractions contained lo-12% LH cells (flow rates, ~15 ml/min), medium-sized fractions (flow rate, ~25 ml/min) contained 20% LH cells, and the largest fractions (flow rate, 30-40 ml/min) contained 40-45% LH cells. The same three fractions were separated in these studies (Table 2). Large cells from metestrous rats were percentages of immunoreactive gonadotropes lower than 40%. There were no changes in immunolabeled LH cells in the medium-sized cell population with the stage of the cycle. The small cells collected contained 7-11% LH cells, except on the AM of proestrus. GnRH did not stimulate increases in the percentages of immunoreactive LH cells (data not shown). Changes in percentages of cells bearing LHP mRNA

Table 2 also shows that when cells labeled for LH/? mRNA were analyzed, the small cells showed higher percentages of labeled cells in metestrus and diestrus than those identified by immunolabel. The highest percentages of small cells bearing LH mRNA were found in populations from metestrous rats. After a reduction at diestrus, the lowest percentages were found on the PM of proestrus. Medium-sized fractions showed 12-B% cells labeled for LHP mRNA. No changes were seen with GnRH stimulation. The percentage of large cells labeled for LHP mRNA

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

340

TABLE 2. Percentage of total cells in different elutriation

IN LH GONADOTROPES

Endo

l

Voll30.

1992 No

1

fractions labeled for LHfl antigens or mRNA Stage of cycle

Fraction Small % LHfl Agb % LHB mRNAd Medium’ % LH,9 Agb % LH mRNA“ Large’ % LH6 Agb % LH mRNAd

Mete&us

Diestrous II

Proestrous AM

Proestrous PM

Rstrus

11 * 1 24 f 3’

7+1 13 f 1

19 zk 2’ 8+1

7fl 8+1

10 f 2 10f 1

17 * 1 18 -c 2

25 zk 4 14 f 1

21+4 17 f 1

24 + 3 16 + 1

21 f 3 13 * 2

29 f 1 25 z!z2’

44 f 4’ 19 f 2

43 f 6’ 16k 1

44 + 7’ 9+1

43 f 6’ 18+ 1

’ Collected at 8-15 ml/min. b Percentage of total cells in fraction labeled for LHj3 antigens + SEM (n = 7). cHighest percentages (by Duncan’s multiple range test). d Percentage of total cells in fraction labeled for LHB mRNA f SEM (n = 7). ’ Collected-at 25 ml/min. f Collected at more than 35 ml/min.

was also highest in cells from rats in metestrus. A significant decrease was seen in proestrus, which persisted until the afternoon. Then, the percentages rose by the AM of estrus to levels that were nearly as high as those in metestrous rats. GnRH stimulation did not significantly increase the percentage of large cells bearing LH@ mRNA. Figure 6 illustrates large gonadotropes from rats in mete&us. Label for LHP mRNA is strong and occurs in patches, linear arrays, or vesicles. Changes in distribution of LH gonndotropes in the small, medium, and large cell pools

Table 1 had shown a flux of cells in and out of the different sized pools that varied with the stage of the cycle. We hypothesized from previous measurements (1, 2) that some of this could be due to changes in the size

FIG. 6. Largest gonadotropes from metestrous rats labeled for LH@ mRNA. Several cells are noted by the art-ours. Label for the mRNA was performed with streptavidin-colloidal gold, intensified by silver. The inset shows labeling in linear or vesicular arrays, which are typical of the rough endoplasmic reticulum (rer) found in this subtype (1, 2). Magnification, ~733.

and density of LH gonadotropes. The overall percentages of LH cells in each fraction (Table 2) may not provide accurate information about such a shift, however, because the cells may be diluted by other cell types that are also changing size or density. Therefore, to test the hypothesis about changes in the smaller subsets, the percentages of LH cells were first used to determine the absolute numbers of gonadotropes in each fraction. Then, their distribution among the three pools was calculated. The proportion of gonadotropes bearing LH/3 mRNA that were small was 43 + 3% of the total gonadotropes in metestrus and 45 f 2% of those on diestrus II (Figs. 7 and 8; values are averages f SEM). At metestrus, the proportion of medium-sized gonadotropes bearing LHP antigens or mRNA was 15 + 1% or 17 + l%, respectively. The proportion of medium-sized gonadotropes then increased to 25 f 2% cells bearing LHP antigens and 38 + 2% cells with LH/3 mRNA by late proestrus (Figs. 9 and 10). This was accompanied by a significant decrease in the proportion of small cells bearing LHP mRNA to 29 f: 1%. The majority of the gonadotropes bearing LH antigens were large by the PM of proestrus (53 zt 2%; Fig. 10) and early estrus (65 +- 3%; Fig. 11). As the proportion of large immunoreactive gonadotropes increased from proestrus to estrus, the proportion of small (17 f 1%) and mediumsized (18 f 1%) immunoreactive gonadotropes decreased. However, the pool of gonadotropes bearing LHP mRNA still contained significant numbers of small cells. By estrus and metestrus, over 80% of gonadotropes bearing mRNA were either in the small or large subset. Discussion The study was designed to test the hypothesis that the small or medium-sized gonadotropes may serve as reserve

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

IN LH GONADOTROPES PROESTRUS

METESTRUS % of total

LH Antigen

341

% of total

gonadotropes

LH mRNA

FIG. ‘7. Pie graph showing the distribution of total LH gonadotropes among the three fractions collected from five metestrous rats. The percentages of gonadotropes labeled for LH@ antigens or mRNA were used to obtain absolute numbers of gonadotropes in each fraction. These numbers were added to yield total numbers of gonadotropes. The absolute numbers of gonadotropes in each fraction were then divided by the total number of gonadotropes to obtain the percentage of gonadotropes in that fraction. In cells from metestrous rats, 50% of the gonadotropes bearing LHfl antigens are large; among cells bearing LH@ mRNA, only 40% are large. In contrast, more small gonadotropes bear LHfi mRNA than LH/3 antigens.

DIESTRUS AM % of total gonadotropes

LH antigens

LH mRNA

FIG. 9. The distribution of gonadotropes in cells from proestrous rate (AM) shows the following changes. There is an increase in the proportion of medium-sized cells bearing LH antigens or mRNA from 15% to 25% or 17% to 28%, respectively, from mete&us to the AM of proestrus. This is coupled with a significant decrease in the proportion of small gonadotropes bearing LHj3 mRNA (from 43-45% during diestrus to 37% by proestrous AM). Thus, the major pool that is changing appears to be the medium-sized gonadotropes. This correlates with the changes shown in Table 1. Because there are no changes in overall percentages of LH during this time (see Fig. l), the working hypothesis is that small gonadotropes synthesize and translate mRNA, enlarge, and become more dense, thus populating the medium-sized pool.

PROESTRUS % of total

LH antigens

AM

gonadotropes

PM

gonadotroper

LH mRNA

FIG. 8. The approach described in Fig. 7 was used bution of gonadotropes from diestrous II rats. basically similar to that in the metestrous group. A percentages of labeled cells in the medium-sized The change is not significant, however.

to study the distriThe distribution is trend toward higher fraction is evident.

cells. In other words, small gonadotropes may be activated early in the cycle, enlarge, and increase LH stores, presumably to support the proestrous surge. The detection of increases in the proportion of small gonadotropes that express LH/3 mRNA early in the cycle followed by increases in medium-sized cells containing the translated product (LH) might support this hypothesis. Considerations in the interpretation of data

Centrifugal elutriation separates by both size and density (21-23). As the LH stores increase, the gonadotropes may become more dense. Therefore, an increase in size, density, or both would cause the gonadotropes to move into a larger sized pool. Such changes might be accom-

LH antigens

LH mRNA

FIG. 10. Pie graphs showing the distribution of total LH gonadotropes from rats taken in the PM (1400 h) of proestrus. The pool of small gonadotropes with LH antigens or mRNA is reduced (compared with that on proestrous AM). Medium-sized fractions contain more cells with LHB mRNA. Finally, there is an increase in the percentage of cells with LH/3 antigens from proestrous AM to proestrous PM.

panied by decreases in the proportion of smaller gonadotropes. Although each fraction has been named by the average size of its cells (small, medium, or large), it should be recognized that a subset may enter a different pool simply because of a change in density, and there may be a size overlap. In previous studies we showed that gonadotropes increase in both size and density as they approach estrus (1, 2). Therefore, a shift based on either characteristic is physiologically significant. These studies validated the separation methodology in a number of ways. First, we found no difference in the overall percentage of cells recovered with the stage of the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

342

IN LH

ESTRUS AM gonadotropcs

% of total

GONADOTROPES

cycle. The range in yields matched that of previous studies (2, 3,8, 13,20). Thus, there did not appear to be a selective loss with a given stage of the cycle. Furthermore, there was no correlation between the percent recovery and the distribution of the cells or the number of cells loaded (as long as the starting numbers remained under maximal levels recommended for the Sanderson chamber). There was a change in the distribution of cells among the fractions with different stages of the cycle. Diestrous rats tended to have more small cells (shown in both Table 1 and Fig. 5). Then, there was an increase in the proportion of medium-sized cells in early proestrus, followed by an increase in the proportion of large cells in late proestrus and early estrus. This suggested that some of the pituitary cells were becoming more dense or larger as they approached proestrus. The remaining experiments tested contributions by the LH cell population to this shift. The three fractions contained percentages of immunolabeled gonadotropes similar to those recovered in previous studies (2, 3, 8, 13). The lower percentages of LH cell in large fractions from metestrous rats may reflect the shift out of this subset after the peak secretory activity. However, a number of other cell types could be changing size or density during the cycle, which could dilute the gonadotropes or concentrate them in a given fraction. The percentages of immunoreactive LH cells in unseparated populations and in the fractions during proestrus were higher than those in previous reports (l-7,13; other reports reviewed in Refs. 1 and 2). In earlier studies we used randomly cycling rats (3, 7, 13) and less sensitive immunoperoxidase detection protocols (l-7, 13). The highest percentages of LH cells agree with previous

l l

1992 No 1

counts of cells labeled with biotinylated [D-Lys’]GnRH (34,35) or dual labeled for LH and FSH (7). These same high percentages are also found when LH@ mRNA is detected during estrus. This may be the concentration of fully activated LH gonadotropes. Shifts in LH gonadotropes

LH mRNA LH antigens FIG. 11. The distribution of LH gonadotropes from estrous rata shows a continued shift in favor of large antigen-bearing LH cells from proestrus to estrus. At the same time, small and medium-sized fractions show fewer cells labeled for LH@ antigens. However, there is a significant increase in the percentage of small or large cells bearing LH@ mRNA at the expense of the medium-sized fractions. The latter group shows a reduction from 38% labeled cells (during proestrous PM) to 18% labeled cells (by estrus). Note that this same distribution of cells bearing LHB mRNA persists in the metestrous cell population (see Fig. 7).

Endo Voll30

with the stage of the cycle

Because there were no changes from diestrus to proestrus PM, the shifts seen in the distribution study were not due to the addition of new gonadotropes. The shifts do correlate with overall changes in the cell population seen in Table 1. A high percentage (43-45%) of gonadotropes bearing LH/3 mRNA from metestrous and diestrous II rats are small. The proportion of small gonadotropes with LH@ mRNA then declines to 29% by proestrous PM. This decline correlates with the decline in overall numbers of small cells from diestrus to proestrous PM. The proportion of medium-sized gonadotropes labeled for LH@ mRNA grows during diestrus to 38% of gonadotropes by the PM of proestrus. This suggests that the subset of small cells may have enlarged or become more dense in order to populate the pool of medium-sized gonadotropes. This change also correlates with the overall increase in numbers of medium-sized cells seen in the AM of proestrus. There is a striking increase in the proportion of immunolabeled large gonadotropes from proestrus AM (41%) to estrus (65%), which is accompanied by a decrease in the proportion of immunolabeled small or medium-sized cells to 17-18%. Again, this could come from enlargement or increases in the density of these smaller cells. This change also correlates with the increase in the overall numbers of large cells from proestrus AM to estrus. The mechanisms behind the enlargement are uncertain. Previous studies have shown that it occurs with GnRH stimulation (2, 8) or by removal of testosterone feedback (castration) in viuo (32,33). It may result from increased surface area as granule membranes are added during exocytosis. It may also be associated with increased translation and packaging. The continued predominance of large gonadotropes until estrus is not unexpected and confirms previous measurements (1). The secretory activity of both medium-sized and large gonadotropes would continue after the LH surge, because most are multihormonal and release FSH in early estrus (1, 2). This estrous activity may continue any enlargement mediated by exocytosis. Furthermore, the predominance of small and medium-sized LH gonadotropes during metestrus and diestrous AM suggests a response to the lower GnRH stimulation during that period. After estrus, surface membranes may be recycled back to the Golgi complex, thereby reducing the average cell area.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

PLASTICITY

IN LH

Loss of LH or FSH stores through secretion may also reduce the density of the cells. This reduction in size or density may cause them to populate a smaller sized cell pool. At this point, they may be in a reserve state, awaiting activation during the next cycle. In previous studies of cells from randomly cycling female rats, small cells were the only population that could be stimulated by GnRH to store more LH (become bihormonal) (9). They could also be stimulated by estradiol to bind GnRH (10). Changes in expression of mRNA by LH gonadotropes

The production of LH P-subunit mRNA is one of the rate-limiting steps in the production and secretion of LH (36-38). Assays have shown that the gonadotrope population expresses more LH/3 mRNA at key times in the cycle (37-39). Shupnik et al. (38) reported increases in rates of transcription of LHP mRNA on the AM of diestrus II, which continued during the PM of proestrus, reaching a peak at 1700 h on proestrus and falling to proestrous AM values on the AM of estrus. Zmeili et al. (37, 39) also reported a peak in LHP mRNA during the day of diestrus (0800-2000 h) and just before the proestrous LH surge (1600-1800 h). The assays of mRNA content do not discriminate whether changes in the rates of transcription, processing to mature mRNA, transport to the cytoplasm, and degradation (40) contributed to the changes. The in situ protocols may determine whether there are changes in the number of reactive cells or in the amount of available mRNA per cell. Furthermore, in situ hybridization can distinguish activity exhibited by a subpopulation of cells. The increased rate of transcription during proestrus (38) and the peak in LH/3 mRNA content (37, 39) correlate with an increment in the overall percentages of gonadotropes that express the mRNA. Furthermore, most of this early increase appears to be in the largest gonadotrope population. The peak in LHP mRNA content assayed on diestrus II (37,39) is not associated with an increase in the overall percentage of gonadotropes bearing the mRNA. It may reflect increases in the amount of available mRNA per cell, which can be detected by image analysis. An analysis of changes in the proportion of LHP mRNA-bearing gonadotropes suggests that the proestrous increases in transcription rates (38) may occur in both small and large gonadotropes. After the overall increase in large gonadotropes bearing mRNA at estrus, there were increases in the proportion of small gonadotropes bearing LH/3 mRNA. The rapid loss of LH cells from the large cell fraction after estrus suggests involution. Perhaps some of the small cells bearing LH@ mRNA during metestrus are involuted large gonadotropes.

GONADOTROPES

343

Summary and conclusion

To summarize, there is a shift in the overall distribution of different sized cells in the population with the stage of the cycle. There appear to be more large cells during proestrus and estrus. The studies in this presentation suggest that LH gonadotropes may contribute to this shift. Small gonadotropes appear to enlarge or increase in density before diestrus II and express mRNA for LHP. This fits with their proposed role as a subset of reserve cells. However, the large gonadotropes also contribute to the increased mRNA activity seen in late proestrus, which indicates that they are not solely a secretory subset. After estrus, the large gonadotropes may have secreted their stores and internalized their membranes, thereby becoming smaller and less dense, so that they repopulate the small cell fractions. The companion report (24) will discuss the contributions of each of the subsets to LH secretion throughout the estrous cycle. A later report will describe the changes in the FSH cell population. Acknowledgments The authors wish to thank Dr. J. G. Pierce for providing the antibovine LH/3 antiserum. They are also grateful to Dr. W. Chin for help and suggestions during the production of the complementary oligonucleotide probes to LHP mRNA. They appreciate the excellent typing skills of Ms. Betty Williams. References 1. Childs GV, Unabia G, Tibolt R, Lloyd JM 1987 Cytological factors that support non-parallel secretion of LH and FSH during the estrous cycle. Endocrinology 121:1546-1558 2. Childs GV 1986 Functional ultrastructure of gonadotropes: a review. In: Pfaff D (cdl Current Tonics in Neuroendocrinoloev. “I Springer-Verlag, New York, vol7D:49-97 3. Childs GV, Hyde C, Naor Z, Catt K 1983 Heterogeneous luteinizing hormone and follicle stimulating hormone storage patterns in subtypes of gonadotropes separated by centrifugal elutriation. Endocrinology 113:2120-2128 4. Childs GV, Ellison DG, Foster L, Ramaley JA 1981 Postnatal maturation of gonadotropes in the male rat pituitary. Endocrinology 109:1683-1693 5. Moriarty GC 1975 Electron microscopic-immunocytochemical studies of rat pituitary gonadotrophs: a sex difference in morphology and cytochemistry of LH cells. Endocrinology 97:1215-1225 6. Childs GV. Ellison DG 1980 An immunocvtochemist’s view of gonadotropin storage in the adult male rat. Cytochemical and morphological heterogeneity in serially sectioned gonadotropes. Am J Anat 158:397-410 I. Childs GV 1985 Shifts in gonadotropin storage in cultured gonadotropes following GnRH stimulation in uitro. Peptides 6:103-107 8. Childs GV, Lloyd JM, Unabia G, Wierman ME, Gharib SD, Chin WW 1988 Differential regulation of LH beta and FSH beta subunit gene expression. In situ hybridization studies of individual gonadotropes. In: Lakoski JM. Perez-Polo JR. Rassin DK (edsl Neural Control of Reproductive Function. Proceedings of the Fifth Galveston Neurosciences Symposium. A. R. Liss, Inc., New York, chapt 25 9. Lloyd JM, Childs GV 1988 Differential storage and release of LH and FSH from individual gonadotropes separated by centrifugal

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.

344

PLASTICITY

IN LH GONADOTROPES

elutriation. Endocrinology 122:950-961 10. Lloyd JM , Rougeau D, Childs GV 1988 Enrichment and differential regulation of gonadotrope subpopulations at specific stages of the estrous cycle. ’71st Annual Meeting of The Endocrine Society, Seattle WA, 1989, p 229 (Abstract) 11. Denef C, Hautekeete E, Dewals R 1978 Monolayer cultures of gonadotrophs separated by velocity sedimentation: heterogeneity in response to luteinizing hormone-releasing hormone. Endocrinology 103:736-747 12 Denef C, Hautekeete E, Dewals R, De Wolf A 1980 Differential control of luteinizing hormone and follicle-stimulating hormone secretion by androgens in rat pituitary cells in culture: functional diversity of subpopulations separated by unit gravity sedimentation. Endocrinology 106:724-729 13. Hyde CL, Childs G, Wahl LM, Naor Z, Catt KJ 1982 Preparation of gonadotroph-enriched cell populations from adult rat anterior pituitary cells by centrifugal elutriation. Endocrinology 111:14211423 14. Limor R, Ayalon D, Capponi A, Childs GV, Naor Z 1987 Cytosolic free calcium levels in cultured pituitary cells separated by centrifugal elutriation: effect of gonadotropin-releasing hormone. Endocrinology 120:2129-2142 15. Smith CE, Wakefield I, King JA, Naor Z, Millar RP, Davidson JS 1987 The initial phase of GnRH-stimulated LH release from pituitary cells is independent of calcium entry through voltage-gated channels. FEBS L&t 225:247-250 16. Chang JP, Stojilkovic SS, Greater JS, Catt KJ 1988 Gonadotropinreleasing hormone stimulates luteinizing hormone secretion by extra-cellular calcium-dependent and -independent mechanisms. Endocrinology 122:87-97 17. Conn PM, Mariam J, McMillian M, Stern J, Rogers D, Hamby M, Penna A, Grant E 1981 Gonadotropin-releasing hormone action in the pituitary: a three step mechanism. Endocr Rev 2174 18. Catt KJ, Loumaye E, Wynn PC, Iwashita M, Hirota K, Morgan RO. Charm JP 1985 GnRH receutors and actions in the control of reproduc&e function. J Steroid*Biochem 23~677 19. Conn PM, McArdle CA, Andrews WV, Huckle WR 1987 The molecular basis of gonadotropin-releasing hormone (GnRH) action in the pituitary gonadotrope. Biol Rep 3617-35 20. Childs GV, Lloyd JM, Unabia G, Rougeau D 1988 Enrichment of corticotropes by counterflow centrifugation. Endocrinology 123:1226-1232 21. Sanderson RJ, Bird KE, Palmer NF, Brenman J 1976 Design principles for a counterflow centrifugation cell separation chamber. Appendix: a derivation of the equation of motion of a particle under combined centrifugal and hydrodynamic fields. Anal Biochem 71:615-622 22. Sanderson RJ, Bird KE 1977 Cells separations by counterflow centrifugation. In: Prescott DM (ed) Methods in Cell Biology. Academic Press, New York, vol 15:1-14 23. Sanderson RJ 1982 Separation of different kinds of nucleated cells from blood by centrifugal elutriation. In: Pretlow II TG, Pretlow TP (eds) Cell Separation. Academic Press, New York, vol 1:153168 24. Childs GV, Unabia G, Lee BL, Rougeau D 1992 Heightened secretion by small and medium-sized luteinizing hormone (LH)-secreting gonadotropes late in the cycle suggests contributions to the LH surge or possible paracrine interactions. Endocrinology 130:345352

Endo. Voll30

l

1992 No 1

25. Wilfinger WW, Larsen WJ, Downs TR, Wilbur DL 1984 An in vitro model for studies of intercellular communication in cultured rat anterior pituitary cells. Tissue Cell l&483-497 26. Childs GV 1987 Use of immunocytochemical techniques to study cell secretion. In: Poisner. Trifaro (eds) In Vitro Methods for Studying Cell Secretion. Elsevier, New York, vol3:235-253 27. Marchetti C, Childs GV, Brown AM 1990 Voltage-dependent calcium currents in rat gonadotropes separated by centrifugal elutriation. Am J Physiol E589-E596 28. Baldwin DM, Downs TR 1981 Release of LH and FSH by anterior pituitary cell suspension from female rats during the estrous cycle and from estrogen-treated ovariectomized rats. Biol Rep 24:581542 29. Evans WS, Boykin BJ, Kaiser DL, Borges JLC, Thorner MO 1983 Biphasic luteinizing hormone secretion in response to gonadotropin-releasing hormone during continuous perifusion of dispersed rat anterior pituitary cells: changes in total release and the phasic components during the estrous cycle. Endocrinology 112:535-542 30. O’Conner JL, Clary AR, Kellom TA 1988 Superfused pituitary cell cultures: comparative responsiveness of cells derived from various stages of the estrous cycle of LHRH stimulation administered. Life Sci 42:61-72 31. Fallest PC, Hiatt ES, Schwartz NB 1989 Effects of gonadectomy on the in vitro and in viva gonadotropin responses to gonadotropinreleasing hormone in male and female rats. Endocrinology 124:1370-1379 32. Childs GV, Lloyd J, Unabia G, Gharib SD, Wierman ME, Chin WW 1987 Detection of LH-beta mRNA in individual gonadotropes after castration: use of a new in situ hybridization method with a photobiotinylated cRNA probe. Mol Endocrinol1:926-932 33. Childs GV, Unabia G, Wierman ME, Gharib SD, Chin WW 1990 Castration induces time-dependent changes in the follicle-stimulating hormone b-subunit messenger ribonucleic acid-containing gonadotrope cell population. Endocrinology 126:2205-2213 34. Childs GV, Naor Z, Hazum E, Tibolt R, Westlund KM, Hancock MB 1983 Localization of biotinylated gonadotropin releasing hormone on pituitary monolayer cells with avidin-biotin peroxidase complexes. J Histochem Cytochem 31:1422-1425 35. Lloyd JM, Childs GV 1988 Changes in the number of GnRHreceptive cells during the rat estrous cycle: biphasic effects of estradiol. Neuroendocrinology 48138-146 36. Gharib S, Wierman ME, Shupnik MA, Chin WW 1989 Pituitary gonadotropin subunit genes: structure and hormonal regulation of expression. In: Lakoski J, Perez-Polo JR, Rassin D (eds) Neural Control of Reproductive Function. Liss. New York, DD 395-409 37. Zmeili SM, Papavasiliou SS, Thorner MO, Evans WS, Marshall JC, Landefeld TD 1986 Alpha and luteinizing hormone beta subunit messenger ribonucleic acids during the rat estrous cycle. Endocrinology 119:1867-1869 38. Shupnik MA, Gharib SD, Chin WW 1989 Divergent effects of estradiol on gonadotropin gene transcription in pituitary fragments. Mol Endocrinol3:474-480 39. Marshall JC, Haisenleder DJ, Ortolano GA, Dalkin AC, Paul SJ, Landefeld TD 1989 Regulation of gonadotropin subunit gene expression. In: Lakoski J, Perez-Polo JR, Rassin D (eds) Neural Control of Reproduction Function. Liss, New York, pp 411-426 40. Levin N, Roberts JL 1991 Positive regulation of proopiomelanocortin gene expression in corticotropes and melanotropes. Front Neuroendocrinol12:1-22

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 11:17 For personal use only. No other uses without permission. . All rights reserved.