THE JOURNAL 0 1989 by The American Society for Biochemistry

OF BIOLOGICAL CHEMISTRY and Molecular Biology, Inc

Vol. 264, No. Issue of October 5, pp. 16941-16947,1989 Printed in U.S.A.

Estrogen Regulation of Tissue-specific Expression of Complement C3* (Received for publication, June 2, 1989)

Susan A. Sundstrom, Barry S. KommS, Helga Ponce-de-Leon, Zheng Yi, Cory Teuscher, and C. Richard Lyttlej From the Dfzpartment of Obstetrics and Gynecology, University of Pennsylvania, School of Medicine, Philadelphia, Pennsylvania 19104 and the $Department of Biochemistry, University of Arizona, Tucson, Arizona 95724

The administration of estradiol to immature rats results in the increased synthesis and secretion of a 180-kDa protein, composed of 115- and 65-kDa sub- units, by the uterine luminal epithelial cells. A mono- clonal antibody against the 180-kDa protein was uti- lized to isolate the cmorresponding cDNA (LE-1) from a rat uterine luminal epithelial cell cDNA X g t l 1 expres- sion library. This LE-1 cDNA was sequenced and shown to be homologous to complement component C3. The sequence was a.pproximately 81 and 90% homol- ogous to human and mouse C3, respectively. The LE-1 cDNA sequence was homologous with the 3’ portion of the C3 mRNA containing the (Y subunit (115 kDa). Uterine mRNA isol.ated from immature rats treated with 1 pg of estradiol for 24 h demonstrated a 25-fold increase in the concentration of a 6.0-kilobase mRNA by Northern hybridization with either LE- 1 or authen- tic human C3 cDNA probes. To further examine the possibility that the estradiol-regulated secretory pro- tein was C3, an aliquot of radiolabeled media protein from control and estradiol-stimulated rat uteri was incubated with goat. anti-rat C3 antibody. The immu- noprecipitated radiolabeled protein from estradiol- treated animals was: increased significantly (p c 0.01) compared to media from control animals. Analysis of the immunoprecipitated proteins on nonreducing so- dium dodecyl sulfalte-polyacrylamide gel electropho- resis revealed a single protein of 180 kDa from estra- diol-stimulated uterine media, whereas no detectable proteins were immunoprecipitated from media ob- tained from control uteri. Also, when the immunopre- cipitated protein was reduced (20 mM dithiothreitol) it dissociated into two subunits of 115 and 65 kDa. Im- munohistochemical studies demonstrated the presence of C3 only in the epithelial cells of estrogen-stimulated rat uteri. In addition, the estradiol-stimulated mRNA was only detectable in uterine epithelial cell RNA. Analysis of liver RNA demonstrated a 6.0-kilobase mRNA, as in the uterus, when hybridized with LE-1. However, unlike the uterus, its concentration was not influenced by estrogen administration with up to three daily injections of 100 pg of diethylstilbestrol. Based on biophysical, DNA sequence, and antibody data we

HD 20025 (to C. R. L.), :HD 21926 (to C. T.), and HD 06274 (to C. * This work was supported by National Institutes of Health Grants

R. L. and C. T.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Preliminary reports containing this work was presented at the 70th and 71st Annual Meetings of the Endocrine Society in June 1988 and 1989, Abstracts 1302 and 725.

§ To whom correspondence should be addressed Dept. of Obstet- rics and Gynecology; University of Pennsylvania, School of Medicine, Room 306, John Morgan Bldg., Philadelphia, PA 19104-6080.

conclude that rat uterine epithelial cells produce C3 in response to estradiol whereas the expression in the liver was not modulated by estrogens.

The rat uterus is a well-studied target tissue for sex steroid hormone action. Estrogen stimulates uterine growth and in- creases expression of mRNAs coding for proteins of known function as well as others whose functions have not yet been identified (1-3). Additionally, our laboratory and others have demonstrated that estrogen stimulates the synthesis and se- cretion of several uterine proteins (4-6). We have reported that estrogens stimulate the synthesis of two uterine proteins having molecular weights of 115,000 and 65,000 (7). The synthesis and secretion of these proteins is also regulated by interactions of progestins and estrogens (8). Therefore, these proteins may serve as good models or marker proteins for the examination of the mechanisms of steroid regulation of gene expression in a mammalian reproductive system.

It is currently hypothesized that the control of uterine gene expression by estrogen requires the estrogen receptor, located in the cell nucleus, which regulates transcription of specific genes by binding to specific DNA sequences or estrogen responsive elements (9-12). While this has been shown to be the case for a few genes in breast cancer cell lines or in other tissues, it has not been demonstrated to occur in the uterus. Additionally, little attention has been given to the possible differential regulation of the three major uterine cell types, namely the luminal and glandular epithelial cells, stromal cells, and myometrial cells. Recent data from our laboratory indicate that the 115,000- and 65,000-dalton proteins in the rat uterus are, in fact, disulfide-linked subunits of a single 180,000-dalton protein synthesized almost exclusively by the epithelial cells (13). Therefore, not only could this protein serve as a model for the examination of steroid-regulated gene expression in the reproductive tract, but it may be a specific marker for epithelial cells.

In this paper we report the cloning of a cDNA probe for this estrogen-regulated secretory protein and identify the protein as complement component C3. This identification was based on the biochemical nature of the proteins, the sequence of a cDNA clone obtained from a Xgtll cDNA library synthesized from mRNA isolated from estradiol-stimulated uterine luminal epithelial cells, and immunoprecipitation data using goat anti-rat C3 antibody. Furthermore, we demonstrate that the expression of this gene in the liver is not modulated by estrogens.

MATERIALS AND METHODS

Animals-Immature female (21 days of age) Sprague-Dawley rats were obtained from Charles River Breeding Laboratories (Farming- ton, CT). All animals were given free access to food and water and

16941

16942 Estrogen-regulated Expression of C3 were maintained on a 14-h light and 10-h dark-light cycle. Steroids were injected subcutaneously in sa1ine:ethanol (8020) and controls were injected with vehicle alone.

In Vitro Incubations and Protein Analysis-Uteri were removed, split longitudinally, rinsed three times in minimal essential medium (Sigma) minus methionine, and incubated for 6 h at 37 "C, 5% CO, in minimal essential medium minus methionine containing penicillin (100 pg/ml), streptomycin (1 pg/ml), and 50 pC/ml [%]methionine (1200 C/mmol, Amersham Corp.). The incubation media were col- lected and cells and tissue debris removed by centrifugation. The amount of radioactivity incorporated into protein was determined by trichloroacetic acid precipitation and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)' as previ- ously described (7).

Immunoprecipitation-A volume of incubation media containing 50,000 trichloroacetic acid precipitable counts/min was incubated with 10 pg of goat anti-rat C3 antibody (Cappel Laboratories Inc., Organon Teknika Corp.) for 18 h at 4 "C, following which 50 p1 of washed Pansorbin (Behring Diagnostics, La Jolla, CA) was added and incubated at room temperature for 1 h. All incubations were performed on a rotary platform. The pellet obtained after a 5-min centrifugation at 10,000 X g was washed by resuspension followed by centrifugation. This washing was repeated three times with Tris- buffered saline containing 0.5% Triton, three times with Tris-buff- ered saline-Triton containing 0.5% SDS, and once with Tris-buffered saline. The washed pellet was resuspended in 100 p1 of 2 X final sample buffer without reducing agents, heated at 90 "C for 15 min, centrifuged, and 10-pl aliquots of the supernatant counted in a scintillation counter. The identity of the immunoprecipitated protein was confirmed by SDS-PAGE.

Immunohistochemical Localization-Cryostat sections (6-pm) were cut, air dried for 2-4 h, fixed in acetone for 10 min a t 4 "C, and washed in phosphate-buffered saline. Endogenous peroxidase activity was destroyed by reacting the tissue sections with 0.02 M sodium azide containing 0.1% Hz02 in phosphate-buffered saline for 10 min at room temperature. The sections were then incubated with peroxi- dase-conjugated goat anti-rat C3 antibody (1:lOOO) (Cappel, Organon Teknika Corp.) for 30 min. Peroxidase activity was indicated with 3- amino-9-ethylcarbazole. Tissue sections were counterstained with Delafield's hematoxylin (14).

RNA Isolation-RNA was extracted from either whole tissues or isolated cells as previously described (3). RNAs were separated by electrophoresis in 1% agarose gels containing 2.2 M formaldehyde, transferred electrophoretically to Nytran (Schleicher & Schuell) (151, and hybridized to nick-translated cDNA probes (16).

cDNA Library Construction and Screening-Uterine luminal epi- thelial cells were isolated using the enzymatic digestion methods developed by Sherman et al. (17) as modified by McCormack and Glasser (18). A cDNA library was constructed using 2 pg of mRNA isolated from rat uterine luminal epithelial cells (19, 20).

Approximately 500,000 plaques were screened using monoclonal antibody 1C prepared as described previously (21, 22). Hybridoma media diluted 125 was used as the primary antibody. First antibody binding was detected using a 'T-labeled rabbit anti-mouse IgM (kindly supplied by Dr. M. Cancro, Dept. of Pathology, University of Pennsylvania). Positive plaques were selected and purified by several rounds of selection until all plaques were positive for antibody bind- ing.

DNA Sequencing-The cDNA clone isolated from the Agtll library designated LE-1 was subcloned into pGEM3 and pGEM4 and se- quenced using the GemSeq K/Rt system (Promega, Madison, WI) as previously described (23).

RESULTS

Previous results from our laboratory demonstrated that the 180,000 protein is produced almost exclusively by the luminal epithelial cells (8). In addition, we have developed a mono- clonal antibody to this protein which can be used on "Western blots" (8). Based on these findings, mRNA was extracted from a preparation of isolated luminal epithelial cells from estro- gen-stimulated rat uteri. This mRNA was used as a template to produce a cDNA library in Xgtll. The library was screened

The abbreviations used are: SDS-PAGE, sodium dodecyl sulfate- polyacrylamide gel electrophoresis; kb, kilobase.

using our monoclonal antibody and an lZ5-labeled rabbit anti- mouse IgM. Initial attempts to isolate the insert cDNA from the phage using the restriction enzyme EcoRI were not suc- cessful, therefore it was necessary to restrict the X DNA using a double digest (KpnI and EcoRI) to isolate the insert DNA attached to the X DNA. This digestion resulted in the isolation of a 1.9-kb insert containing 1.08 kb of X DNA. Restriction mapping of the insert revealed a PstI site within approxi- mately 50 bases of the original EcoRI insertion site. Therefore, insert DNA was isolated using PstI and EcoRI and subcloned into pGEM3 and pGEM4 plasmid vectors. Analysis on a 1% agarose DNA gel indicated that the isolated insert was ap- proximately 845 base pairs in length. This insert, termed LE- 1 (Luminal Epithelial), was used in subsequent experiments to examine rat uterine mRNA and to obtain sequence data.

Messenger RNA isolated from either control or estradiol- stimulated rat uteri was examined by RNA blot analysis. The results presented in Fig. lA show that the LE-1 cDNA rec- ognized a 6.0 f 0.2-kb mRNA which was detected only in mRNA isolated from uteri of estradiol-treated rats. The re- sults presented in Fig. 1B were obtained by scanning Fig. lA using a video densitometer and more clearly demonstrate the effect of estradiol treatment on the LE-1 mRNA concentra- tions. Fig. 1 also shows the analysis of the blot which was stripped and rehybridized using probe 1A. We previously showed that probe 1A is not hormonally regulated in the rat uterus (23). Therefore, it was employed to demonstrate the relative concentrations of mRNA. This figure clearly dem- onstrates that estradiol treatment caused approximately a 25- fold increase in the concentration of a 6.0-kb mRNA. This size mRNA is sufficient to code for a protein having a molec- ular weight of 180,000.

The LE-1 cDNA was sequenced in order to identify this estradiol-induced secretory 180,000 protein. The sequence data shown in Fig. 2 was 81% homologous to human comple- ment C3 (24) and 90% homologous to mouse C3 (25). Thus, from the sequence of this portion of the mRNA it appeared that the secreted protein could be complement component C3. Analysis of rat uterine RNA was also performed using two authentic human C3 cDNA probes obtained from the America Tissue Culture Collection (Rockville, MD). The probe designated pHC3.11 contains most of the C3 mRNA sequence, but lacks the 5' end of the mRNA, and this sequence is contained in pHC3.59 (24). Rat uterine mRNA analysis using probe pHC3.59 showed a band at 6 kb only in mRNA obtained from estradiol-stimulated uteri (Fig. 3A). The same results were obtained usingpHC3.11 (Fig. 3B). These findings were essentially identical to those presented in Fig. lA. South- ern blot analysis of these probes demonstrated that the uter- ine probe LE-1 hybridized only to a 3' section of pHC3.11 (Fig. 4, lane 4 ) but did not hybridize to a 5' section of pHC3.11 (lune 3) or to pHC3.59 (lane 5 ) .

In order to examine the possibility that estrogen induced C3 synthesis, radiolabeled media proteins from both control and estrogen-stimulated rat uteri were immunoprecipitated with goat anti-rat C3 antibody. The counts/min immunopre- cipitated from 50,000 trichloroacetic acid-precipitable counts/ min are shown in Table I. The administration of estradiol for 1 day resulted in approximately a 4-fold increase in secreted C3 whereas 3 days of treatment resulted in a 14-fold increase over nonspecific binding when compared with vehicle-treated controls. Nonspecific binding was determined using goat anti- rat IgG and was 961 +. 449 cpm or 1.9 f 0.9% of the radiola- beled media proteins. Immunoprecipitated radiolabeled pro- teins obtained from animals treated for 1 day were examined by SDS-PAGE. The fluorograph in Fig. 5 demonstrates that

Estrogen-regulated Expression of C3 16943

A C E "

9.5 = 7.5 = - 4.4 - 2.4 - 1.4 -

A B

ID.

C E "

LE- 1 l A

c E C E FIG. 1. Effect of estradiol on LE-1 mRNA. A, Northern blot

analysis of 2 pg of uterine mRNA (poly(A')) from control (C) and estradiol ( E ) (1 pg for 1 day)-treated rats. The left side of A shows the blot hybridized with LE-1 cDNA. The right side of A is the same blot that was stripped and re-hybridized with lA, a nonregulated uterine cDNA. B, relative densities of the autoradiograms of blots shown in 1A. The density of the bands seen in A was determined using a video scanning densitometer.

the immunoprecipitated radiolabeled proteins migrated as a 180,000-dalton protein under nonreducing conditions and as two peptides of 115,000 and 65,000 daltons in the presence of dithiothreitol, while no bands were seen in the immunopre- cipitates obtained from control animals.

We previously demonstrated that the 180,000-dalton pro-

tein is synthesized and secreted by isolated luminal epithelial cells of estradiol-stimulated uteri (13). The results presented in Fig. 6 clearly demonstrate, using peroxidase-conjugated goat anti-rat C3 antibody, that staining was seen only in the luminal and glandular epithelial cells of 1 or 3 day estradiol- stimulated uteri (Fig. 6, B and C), while no staining was seen in uteri from control rats (Fig. 6A). The extent of stimulation is in agreement with the data presented in Table I illustrating the effect of estrogen on the synthesis of C3.

The immunohistochemical and immunoprecipitation data indicated that C3 was present and most likely synthesized in the uterine epithelial cells. In order to further examine this possibility mRNA isolated from luminal epithelial cells and stromal/myometrial cells from estradiol-stimulated rat uteri was analyzed. The results presented in Fig. 7 clearly demon- strate that only the mRNA isolated from the luminal epithe- lial cells contained LE-1 mRNA.

The main site of synthesis of complement C3 is believed to be the liver. We have previously shown that while a 180,000- dalton doublet is clearly visible in the media obtained from incubations of liver tissue, no regulation by estradiol is seen at this dose (13). This observation was confirmed in this study by immunoprecipitation which did not demonstrate any sig- nificant difference in the rate of synthesis and secretion of liver C3 in response to estradiol or to diethylstilbestrol (100 pg/day for 3 days). SDS-PAGE of these immunoprecipitated proteins was identical to those obtained from the uterus (data not shown). Moreover, in the same experiment, liver RNA obtained from both control and diethylstilbestrol-stimulated rats was examined using the LE-1 cDNA probe and estrogen had no effect on mRNA concentrations suggesting a tissue- specific regulation (Fig. 8). Similar results were obtained using pHC3.59, an authentic C3 cDNA.

DISCUSSION

The administration of estradiol to either immature or ma- ture ovariectomized rats was previously shown to result in an increased synthesis and secretion of two proteins (4,7). How- ever, when analyzed in the absence of reducing agents a doublet having a molecular weight of 180,000 is induced while neither of the two proteins having molecular weights of 115,000 or 65,000 are observed (13). These results suggest that the lower molecular mass peptides are subunits of the 180,000-dalton protein.

Our previous results suggest that the uterine estradiol- stimulated protein synthesis and secretion is dependent on continued transcription and translation (4). This is based on the use of the inhibitors of transcription and translation, actinomycin D and emetine, respectively. This conclusion is supported by the data presented here which indicates that estradiol treatment resulted in a 25-fold increase in the con- centration of mRNA coding for the secreted protein (26). The concentrations seen in the control animals were either ex- tremely low or in most cases nondetectable. Therefore it was difficult to accurately determine the fold increase. However, in comparison to other rat uterine mRNAs, such as the collagens which are increased 2- to &fold following estradiol treatment, the stimulation of the LE-1 mRNA was much greater (3,23). The LE-1 mRNA was approximately 6.0 kb in size and was seen as a single band. This size is adequate to code for a protein of a t least 180,000 daltons.

The sequence analysis of the 845-nucleotide LE-1 cDNA was 90% homologous to the mouse and 81% homologous to the human DNA sequences of complement C3 (24, 25). The LE-1 cDNA was homologous to the 3' section of the C3 gene which codes for the carboxyl terminus of the (Y chain. Whether

16944

First at.

+lM

+lH

+lR

+IOlM

+101H

+101R

+201M

+ZOlH

+ZOlR

+301M

+UIlH

+301R

4 0 1 M

+401H

40111

+501M

+%IH +SO111

+601M

+601H +601 R

+701M

+701H

+701R

+ W M

+8OlH

+801R

Estrogen-regulated Expression of C3

FIG. 2. Comparison of human, mouse, and rat sequence data. The mouse sequence (M) represents nucleotides 1932-2776. The human (H) sequence data represents nucleotides 4002-4846 and the rat sequence (R) is the sequence of the LE-1 cDNA.

the extent of homology continues through the entire molecule is not known and is presently under investigation. However, the results obtained by Northern blot analysis using probe pHC3.59, a sequence distinct from LE-1 (Fig. 4), would sug- gest that there is also excellent homology in the 5' end.

Immunoprecipitation of the radiolabeled uterine secretory protein with a commercially available antibody to rat serum C3 strongly supports the conclusion that the secretory protein is C3. This is in agreement with the findings of Takeda et al. (6) who demonstrated that the uterine secretory proteins identified by us are absorbed by an anti-rat serum antibody affinity column. However, our data do not demonstrate that the 115,000- and 65,000-dalton proteins are of serum origin, but rather indicate that the uterine proteins synthesized in uitro, while of uterine origin, may be the same as serum complement produced by the liver. Moreover, while estrogen treatment resulted in increased uterine synthesis of C3, as demonstrated by immunoprecipitation and mRNA blots, it did not affect the synthesis in the normal rat liver. Since the liver has the ability to metabolize estradiol to inactive prod- ucts, we additionally examined the regulation of C3 stimula-

tion using a high dose of the synthetic estrogen diethylstil- bestrol (Fig. 8). The results indicated that estrogens, even in very high doses, did not regulate the expression of C3 in the liver. These findings suggest a tissue-specific stimulation, similar to that reported for the effect of estradiol on the transferrin gene in chick oviduct and liver (27).

Taken together, several lines of evidence support our con- clusion that the 180,000-dalton rat uterine secretory protein was complement C3. First, the molecular weight of 180,000 was in good agreement with complement component C3, which in humans and mice has a molecular weight of 190,000 (28) and 180,000 for rat serum C3 (29). Second, both our secreted protein and complement C3 are composed of two disulfide-linked nonidentical subunits (115,000 and 65,000) for rat serum C3 (29). Third, there was a high percentage of sequence homology between the LE-1 cDNA and that of human and mouse C3. Fourth, the radiolabeled proteins pro- duced by the uterus following estradiol stimulation were im- munoprecipitated by goat antibodies directed against rat serum C3. Last, the results of the immunohistochemical lo- calization studies and RNA blots of epithelial versus stromal/

9.5 - 7.5 - 4.4 - 2.4 - 1.4 -

0.24 -

C E

"

m

..

A

Estrogen-regulated Expression of C3 16945

C E C E

" "

TABLE I Effects of estradiol on uterine C3 synthesis

Radiolabeled media proteins from control and estradiol-stimulated rat uteri (1 pg/day for 1 and 3 days) were immunoprecipitated with a goat anti-rat C3 antibody. Results are expressed as a mean of 4 uteri/ treatment erouD t the standard deviation.

Treatment Uterinefiody c3 c3 PJUPS weight immunoppt" immunopptb

mRIR cpm % Control 0.5 2 0.1 1,958 f 319 3.9 f 0.7 Estradiol (1 day) 1.1 2 0.2' 5,315 t 1,582' 10.6 t 3.2' Estradiol (3 days) 1.6 2 0.2' 14,858 f 4,930' 29.8 t 9.9' a Total counts/min immunoprecipitated by goat anti-rat C3 anti-

body from media containing 50,000 cpm of total trichloroacetic acid- precipitable proteins.

*Percent of total protein synthesized and secreted into media which is immunoprecipitated by C3 antibody.

' p < 0.01 compared to control.

biii &B B C

FIG. 3. RNA blot analysis of uterine poly(A+)RNA (2 pg) from control (C) and estradiol @)-treated immature rats using authentic human C 3 cDNA probes, pHC3.59 and pHC3.11. The left pair of lanes in A shows the blot probed with pHC3.59. The middle pair ( B ) shows the blot re-hybridized with pHC3.11. The right pair ( C ) were hybridized with 1A cDNA.

1 2 3 4 5

"-0

2.1 -

0.8 -

b .,- ..

- 4

D T T A B

FIG. 5. Fluorogram of immunoprecipitated protein using goat anti-rat C 3 antibody. Radiolabeled proteins immunoprecipi- tated with goat anti-rat C3 antibody were analyzed by SDS-PAGE under ( A ) nonreducing and (R) reducing (20 mM dithiothreitol) conditions. Arrows indicate 180-kDa protein ( A ) , and 116- and 65- kDa subunits under reducing conditions ( B ) .

myometrial RNA indicate that the C3 was of epithelial origin. Thus these data support the concept that complement com- ponent C3 was produced by the luminal epithelial cells of the estrogen-stimulated rat uterus and that the expression in the

FIG. 4. Southern blot analysis of LE-1 cDNA and authentic liver was not modulated by estrogens. C 3 cDNAs. Plasmid DNA containing pHC3.11 and pHC3.59 was Even though the liver is the major site of C3 synthesis, digested with Sal1 and ClaI in order to isolate the insert DNAs. Since other extrahepatic tissues and cells such as monocytes and pHC3.11 contained an additional Sal1 site, it was digested into the 5' macrophages have been reported to produce and secrete this and 3' ends. Each fragment was isolated and analyzed using nick-

DNA (1 &; lane 3, 5' end of pJ-JC3.11 (1 pg); lane 4, 3' end of reported that the synthesis of c3 by epithelial Cells Of the pHC3.11 (1 pg); and lane 5, pHC3.59 (1 pg). liver could be enhanced by the monokine, interleukin 1.

translated LE-1 cDNA. Lane 1, LE-1 DNA (75 ng); lane 2, plasmid complement component (30). The work Of Guif?let et (31)

16946 Estrogen-regulated Expression of C3

Additionally, the results of Strunk et d (32) demonstrate C3 synthesis by pulmonary alveolar type I1 epithelial cells. Re- cent work from the laboratory of Fey and Gauldie (33) sug- gests that C3 synthesis, as well as other acute phase proteins, appears to be regulated by interleukin 6. Several proteins including C4, C5, sex limited protein, and az-macroglobulin share some sequence homology and common evolutionary origins with complement C3. However, a search of the human, mouse, and rat libraries in GeneBank with our LE-1 cDNA sequence only detected significant homology with human and mouse C3, rat C3 sequence data has not yet been reported. It is also of interest to note that one of these, a2-macroglobulii, a 180,000-dalton glycoprotein, is present in the uterus and its expression is regulated during pregnancy (34, 35). However,

FIG. 7. Effect of estradiol on C3 RNA in the uterine cell types. Northern blot analysis of total RNA (15 pg) isolated from luminal epithelial ( A ) or stromal and myometrial ( B ) cells from estradiol-stimulated rat uteri. The resulting blot was hybridized with nick-translated LE-1 cDNA.

9.5- 7.5-

4.4-

2.4-

1.4-

FIG. 8. Effect of diethylstilbestrol (100 rcglaay for 3 days) on C3 mRNA in the liver. Equal concentrations (10 pg) of total RNA isolated from either control (C) or diethylstilbestrol (D)-treated animals were applied to each well and hybridized with nick-translated LE-1 cDNA (A) . The blot was stripped and re-hybridized with 1A cDNA (B) .

in the immature rat uterus estradiol exerted no regulation of az-macroglobulin mRNA concentrations? Furthermore, the regulation of az-macroglobulin in the uterus appears to be distinct from that seen in the liver. The results presented .

a C. R. Lyttle and G. Fey, unpublished data.

Estrogen-regulated Expression of C3 16947

herein, regarding C3 expression, are in good agreement with their findings.

While it seems clear that the luminal epithelial cells syn- thesize and secrete C3 in response to estradiol, the biological function of this protein in the uterus remains unclear. Several biological functions nolrmally ascribed to C3 in the comple- ment system can be postulated to occur in the uterine lumen. These would include, for example, a role in the targeting and lysis of foreign cells or b'acteria. The activation of C3 by either the classical or alternative activation pathways results in the cleavage of C3 into C3a and C3b. Component C3b is able to bind to cells which may be lysed by the further action of the complement pathway (28, 36). The low molecular weight peptide C3a has been shown to possess leukocytic chemotactic activity (28, 36) and perhaps may play a role in estradiol- stimulated eosinophil influx in the rat uterus (14).

Whether a similar regulation occurs in the human is not known. However, it is of interest to note that recent data from our laboratory demonstrate the synthesis of C3 by the glandular epithelial cells found in endometriotic implants (37). It is unclear if this secreted C3 is related to the many immunologic phenomena observed in patients with endome- triosis. Thus, while many roles can be postulated for C3 in the control of the reproductive process, it has not been estab- lished whether any of t:hese possibilities occur.

Acknowledgments-We wish to thank Matthew Galman for his commitment and technical expertise. We are also grateful to Dr. George Fey, Research Institute of Scripps Clinic, for his comments and helpful discussion.

REFERENCES 1. Komm, B. S., and Lyttle, C. R. (1984) J. Steroid Biochem. 21 ,

2. Murphy, L. J., and Friesen, H. G. (1988) Endocrinology 122,

3. Komm, B. S., Frankel, F. R., Myers, J. C., and Lyttle, C. R.

4. Komm, B. S., Keeping, B. S., Sabogal, G., and Lyttle, C. R. (1985)

5. Kuivanen, P. C., and DeSombre, E. R. (1985) J. Steroid Biochem.

6. Takeda, A., Takahashi, N., and Shimizu, S. (1988) Endocrinology

7. Komm, B. S., Rusling, D. J., and Lyttle, C. R. (1986) Endocri-

8. Wheeler, C., Komm, B. S., and Lyttle, C. R. (1987) Endocrinology

9. King, W. J., and Greene, G. L. (1984) Nature 307,745-747

571-577

325-332

(1987) Endocrino1og.y 120, 1403-1410

Biol. Reprod. 32,443-450

22,439-451

122, 105-113

nology 118 , 2411-2416

120,919-923

10. Welschons, W. V., L,ieberman, M. E., and Gorski, J. (1984) Nature 307, 747-74-9

11. Dean, D. C., Knoll, B. J., Riser, M. E., and O'Malley, B. W. (1983) Nature 305 , 551-554

12. Green, S., Kumar, V.. Krust, A,, Walter, P., and Chambon, P. (1986) in Cold Spring Harbor Symp. Quant. Biol. 5 1 , 751-758

13. Lyttle, C. R., Wheeler, C., and Komm, B. S. (1987) in Cell and Molecular Biology of the Uterus (Leavitt, W. W., ed) PP. 119- 136, Plenum Publishing Corp., New York

14. Zheng, Y., Zhou, Z., Lyttle, C. R., and Teuscher, C. (1988) J. Leukocyte Biol. 44,27-32

15. Davis, L. G., Dibner, M. D., and Battey, J. G. (1986) Basic Methods in Molecular Biology, pp. 143-146, Elsevier/North- Holland, New York

16. Rigby, P. W. J., Dieckmann, M., Rhodes, D., and Berg, P. (1977) J. Mol. Biol. 113 , 237-261

17. Sherman, M. (1978) in Methods in Mammalian Reproduction (Daniel, J. C., Jr., ed) pp. 247-257, Academic Press, New York

18. McCormack, S. A., and Glasser, S. R. (1980) Endocrinology 106,

19. Gubler, U., and Hoffman, B. J. (1983) Gene (Amst.) 25,263-269 20. Huynh, T., Young, R., and Davis, R. (1984) in DNA Cloning, A

Practical Approach (Glover, D., ed) Vol. 1, pp. 49-78, IRL Press, Oxford

21. Keeping, H. S., and Lyttle, C. R. (1984) Biochim. Biophys. Acta

22. Kennet, R. H., McKear, T. J., and Bechtol, K. T. (1980) Mono- clonal Antibodies, Plenum Publishing Corp., New York

23. Hsu, C. J., Komm, B. S., Lyttle, C. R., and Frankel, F. (1988) Endocrinology 122,631-639

24. DeBruijnard, M. H. L., and Fey, G. H. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 708-712

25. Lundwall, A., Wetsel, R. A., Domdey, H., Tack, B. F., and Fey, G. H. (1984) J . Biol. Chem. 259 , 13851-13856

26. Sundstrom, S. A., Komm, B. S., Ponce-de-Leon, H., and Lyttle, C. R. (1988) 70th Annual Meeting of the Endocrine Society, New Orleans, LA, Abstr. 1302, The Endocrine Society, Be- thesda, MD

27. Lee, D. C., McKnight, G. S., and Palmiter, R. D. (1977) J. Biol. Chem. 253,3494-3503

28. Muller-Eberhard, H. J. (1988) in Inflammation: Basic Principles and Clinical Correlates (Gallin, J. I., Goldstein, I. M., and Snyderman, R., eds) pp. 21-53, Raven Press, New York

29. Oda, K., Miki, K., Hirose, S., Takami, N., Misumi, Y., and Ikehara, Y. (1986) J. Biochem. (Tokyo) 100 , 1669-1675

30. Cole, F. S., and Colten, H. R. (1988) in The Complement System (Till, 0. G., and Rother, K. O., eds) pp. 44-70, Springer-Verlag, Heidelberg

31. Guiguet, M., Frigere, M. E., Dethieux, M., Bidan, Y., and Mack, G. (1987) In Vitro Cell. Deu. Biol. 23,821-829

32. Strunk, R. C., Eidlen, D. M., and Mason, R. J. (1988) J. Clin. Invest. 81, 1419-1426

33. Fey, G., and Gauldie, J. (1989) in Progress in Liver Disease (Popper, H., and Schaffner, F., eds) 9th Ed., W. B. Saunders Publishers, Philadelphia, PA, in press

34. Northemann, W., Shiels, B. R., Braciak, T. A., Hanson, R. W.,

9203 Heinrich, P. W., and Fey, G. H. (1988) Biochemistry 27,9194-

35. Hayashida, K., Tsuchiya, Y., Kurokawa, S., Hattori, M., Ishi- bashi, H., Okubo, H., and Sakaki, Y. (1986) J. Biochem. (Tokyo)

1634-1649

802,309-406

100,989-993 36. Lamhris, J. D. (1988) Zmmunol. Today 9,387-393 37. Isaacson, K. B., Coutifaris, C., Garcia, C., and Lyttle, C. R. (1989)

36th Annual Meeting of the Society for Gynecologic Znvestiga- tion, San Diego, CA, Abstr. 119, The Society for Gynecologic Investigation, Washington, D.C.