Differential screening of ovarian cDNA libraries detected the expression of the porcine collagenase inhibitor gene in functional corpora lutea

Molecular and Cellular Endocrinology, 83 (1992) 65-11 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00

MOLCEL

65

02673

Differential screening of ovarian cDNA libraries detected the expression of the porcine collagenase inhibitor gene in functional corpora lutea Toshiaki

Tanaka

‘, Naoki Andoh

b, Tatsuo

Takeya

a and Eimei Sato b

‘I Institute for Chemical Research, Kyoto lJnil,ersity, Uji, Kyoto 61 I, Japan, and h Department of Animal Science, Faculty of Agriculture, Kyoto Uni~w-sity,Sakyo-ku, Kyoto 601, Japan

Key words: cDNA

library;

Differential

(Received

24 July 1991; accepted

screening;

Ovary;

Collagenase

24 September

inhibitor;

1991)

(Porcine)

Summary cDNA libraries were constructed from porcine granulosa cells of antral follicles as well as functional corpus luteum, and clones encoding stage-specific genes have been isolated by differential screening. A clone specific to the functional stage of corpus luteum was found to encode the porcine collagenase inhibitor gene and the stage-specific expression in luteinizing tissue was confirmed by Northern blot analysis. The complete open reading frame of the porcine collagenase inhibitor was deduced from the nucleotide sequence, and the localization of the product was examined by immunohistochemical staining as well in pig ovary; the inhibitor was detected in the intercellular space of luteal cells and in the connective tissue around blood vessels in the functional corpus luteum.

Introduction In mammalian follicles, ovarian granulosa cells are known to play essential roles in oocyte development and maturation (Masui and Clarke, 1979; Sato and Koide, 1982), and several lines of evidence suggest that biochemical transmitters produced by granulosa cells are directly or indirectly involved in these processes (Charming et al., 1982). After ovulation of an oocyte, granulosa cells differentiate into luteal cells and proliferate to form the corpus luteum (Charming et al., 1980). Steroidal and nonsteroidal regulators are synthesized in luteal cells and secreted to prepare the uterus for possible ovum implantation and to

Correspondence to: Tatsuo Takeya, Institute for Chemical Research, Kyoto University, Uji, Kyoto 611, Japan. Tel. 077432-8336; Fax 0774-33-1247.

maintain the corpus luteum as well (Charming et al., 1980). Taken together, the co-ordinated expression of various kinds of genes seems to be deeply involved in each stage throughout the oocyte development and maturation, granulosa cell differentiation, luteinization of the follicles and maintenance of the corpus luteum. Therefore, identifying genes that express in a stage- or spatial-specific manner may be crucial to understand the oocyte development and maturation at the molecular level. We constructed cDNA libraries from porcine granulosa cells as well as corpus luteum, and screened cDNA clones by differential hybridization between these two libraries. Deduced nucleotide sequences and predicted amino acid sequences were compared with those of the reported genes, and the expression of mRNA corresponding to each cDNA clone was examined by

06

Northern blot analysis. As a result, a clone among those specific to the functional corpus luteum turned out to encode for the porcine collagenase inhibitor gene; the complete open reading frame was deduced and the localization of the product was examined in pig ovary by immunohistochemical staining.

Materials

and methods

Pig olsary; preparation of granulosa cells and isolation of corpus luteurn Ovaries were obtained from 4-12-month-old pigs at a local slaughterhouse. Follicular contents were sucked out by syringes from medium-sized follicles (3-5 mm in diameter) (Charming et al., 1975) and used directly as a source of granulosa cells. Corpora lutea were classified into three stages; early, functional and regressing stages based on the morphological characteristics of the surface. The tissue was fixed in Bouin solution, stained with the Masson-Goldner method, and histological features of corpora lutea in these stages were examined. After removing the capsule from the corpus luteum, the tissues were kept frozen in liquid nitrogen.

RNA extraction and construction of cDNA libraries Total RNA was extracted by using guanidium thiocyanate (Sambrook et al., 1989) from granulosa cells and corpora lutea. Poly(A) RNA was selected for the construction of cDNA libraries by passing through oligo(dT)-cellulose column. Both cDNA libraries from granulosa cells and corpora lutea were constructed by using hgtll as a cloning vector (Sambrook et al., 1989).

Screening of cDNA clones by difyerential hybridization cDNA clones specific to either granulosa cells or corpora lutea were isolated by differential hybridization between the above two cDNA libraries. Two kinds of cDNA probes were prepared from poly(A)-selected RNA of granulosa cells and of a mixture of three stages of corpora lutea, respectively, using reverse transcriptase (Takara Shuzo, Japan) and [ cu-“‘P]dCTP.

DNA sequence analysis Inserts of phage cDNA clones thus identified were subcloned into plasmid vector pUC18 and sequenced by the chain terminator method (Sanger et al., 1977). Northern blot analysis 10 pug of total RNA samples from granulosa cells and corpora lutea at each stage (early, functional and regressing stages) were denatured with glyoxal, electrophoresed in 1.2% agarose gels, and transferred onto nylon membranes (Immobilone, Millipore) using standard techniques (Sambrook et al., 198Y). An insert of plasmid cDNA clone was labeled using [ a--i2 Pldeoxycytidine triphosphate with random hexamers as primers and was used as a clone-specific probe. A cDNA clone encoding porcine inhibin subunit-A gene (T. Tanaka, unpublished) was used as a probe to examine the mRNA extracted from granulosa cells (Ying, 1988). Immunological staining of 0r:aries 5 pm-thick frozen sections of porcine ovaries were fixed with cold acetone for 10 min at 4°C washed with phosphate-buffered saline (PBS), and treated for 30 min with absolute methanol containing 0.3% hydrogen peroxide. After washing with PBS, tissue sections were incubated with buffer solution (0.1 M phosphate buffer, pH 6.5, 0.005% thimerosal, 0.1% bovine serum albumin) containing horseradish peroxydase-labeled antibovine collagenase-inhibitor antibody (monoclonal antibody 7-6~1, lot No. F-24, Fuji Chemicals, Japan; Kishi and Hayakawa, 1984) overnight at 4°C. 3,3’-Diaminobenzidine-tetrahydrochloride (DAB-4HCl) (Nakane and Pierce, 1967) was used to raise the signal. The sections were immersed in hematoxylin solution for 10 s, washed with running water for 20 min, and dehydrated. Results Isolation of cDNA clones encoding stage-specific genes For screening of corpus luteum-specific cDNA clones, 5 x 10’ independent plaques of the corpus luteum-derived cDNA library were replicaplated. A pair of plates were hybridized with the

was drastically decreased at the regressing stage of corpus luteum (Fig. 1B). In contrast, a significant expression of the inhibin subunit-A gene was detected in granulosa cells (Fig. 1C) (Ying, 19SS>, indicating that the absence of the 0.8 kb band with the clone TS.543 in granulosa cells was not simply due to the quality of the RNA preparation.

respective cDNA probe, and plaques which hybridized preferentially with the corpus luteum probe were selected as candidates for cDNA clones specific to corpus luteum, followed by several purification steps. cDNA clones specific to granulosa cells were obtained vice versa. Among clones thus obtained, we will focus on a corpus luteum-specific clone, TS.543. The insert of A phage clone, 0.8 kb, was subcloned into a plasmid vector, pUC18, and was used for further analysis.

Complete nucleotide sequence of porcine collagenase inhibitor gene The 0.8 kb insert was sequenced and the predicted amino acid sequence was deduced; computer search revealed that the clone TS543 probably encodes the porcine counterpart of human collagenase inhibitor (or a tissue inhibitor of metalloproteinase inhibitor: TIMP) (Carmichael et al., 1986). The insert of the clone TS543 is expected to cover almost the entire region of mRNA based on its size (Fig. lB), and the entire nucleotide sequence is shown in Fig. 2A. Since TIMP including collagenase inhibitor is believed

Northern blot analysis The stage specificity of the clone TS543 was :xamined by Northern blot analysis; total RNA was extracted from granulosa cells and three stages of corpus luteum, run on formaldehyde gel and transferred to nylon membrane (Fig. 1A). The clone TS543 was hybridized to RNA at a band of 0.8 kb and the classification of this clone into corpus luteum specific was further confirmed; the expression was prominent in the early through functional stages of corpus luteum, and

A

C

B 1

2

3

I

4

2

3

4

1

2

3

4

-28s

-18s

-28s

-28s

.ias

-ias

Fig. 1. Northern blot analysis. Total RNA was extracted from granulosa cells and corpus luteum (early, functional and regression stages), run on a formaldehyde gel and transferred to nylon membrane filters. Lanes 1: granulosa cells; lanes 2: early stage; lanes 3: functional stage; and lanes 4: regressing stage. (A) A filter was stained to examine the quality and the equivalence of RNA loads. Probes: (B) clone TS543, and (C) porcine inhibin subunit-A gene. Lengths of the bands were estimated (kb) on the basis of the mobilities of the ribosomal RNAs.

6X

to play important physiological roles in follicular luteinization (LeMaire et al., 1987; Lipner, 1988), we extended the characterization of the gene structure and its product. To begin with, the predicted amino acid sequence was compared with those of human, bovine, rabbit and mouse counterparts (Carmichael et al., 1986; Edwards et al., 1986; Horowitz et al., 1989; Wagner et al., 1990) (Fig. 2B). Collagenase inhibitor is known to be a secretory protein and, hence, the presence of the leader sequence and processing to the mature molecule are expected. Therefore, the predicted porcine gene product was divided into the leader sequence and the mature molecule, and the similarities were compared at both nucleotide and amino acid levels with the corresponding regions of other mammalian inhibitor genes (Table 1). The highest similarity was observed between porcine and bovine genes in the mature inhibitor as well as the leader peptide. The 12 cysteine residues are completely conserved among these genes, as discussed by Wagner et al. (1990). Expression and localization of collagenase inhibitor in pig ouaiy Northern blot analysis clearly showed that the expression of collagenase inhibitor was elevated at specific stages during the follicular luteinization, especially in the functional stage of corpus luteum. To confirm this finding, and further to localize the inhibitor molecules in pig ovary, we examined the expression of collagenase inhibitor by immunohistochemical staining. Sections of corpora lutea at the functional and the regressing stages were stained with horserad-

TABLE

Fig. 2. Porcine collagenase inhibitor. (A) Deduced nucleotide sequence and predicted amino acid sequence of the cDNA clone TS543. Nucleotides are numbered above the line and amino acids are identified on the left. The underlined sequence (AATAAA) indicates the putative poly(Al additional signal. (B) Comparison of the deduced amino acid sequence of TS543 with those of human, bovine and mouse collagenase inhibitors (Carmichael et al., 1986; Edwards et al., 1986; Horowitz et al., 1989; Wagner et al., 1990). Identities are represented by hyphens. The arrow indicates the potential cleavage site of the leader peptide (Carmichael et al., 1986) Gaps are introduced into the sequences (amino acid position at 176 of rabbit and at 180 of mouse) for better alignment.

1

COMPARISON HIBITOR

Human Bovine Rabbit Mouse ” Numbers

OF

THE

AMINO

Amino

acid sequence

Leader

peptide

87.0 a 91.3 73.9 73.9 indicate

ACID

AND

Nucleotide Mature

inhibitor

83.2 88.0 80.2 68.1 percent

NUCLEOTIDE

similarities

Leader

SEQUENCES

the porcine

MAMMALIAN

sequence

peptide

84.1 91.3 84.1 72.5 against

OF

collagenase

COLLAGENASE

Reference Mature 88.3 91.5 84.3 71.4

inhibitor.

inhibitor Carmichael et al. (1986) Wagner et al. (1990) Horowitz et al. (1989) Edwards et al. (1986)

IN-

ish peroxidase-labeled anti-collagenase inhibitor antibodies. In accordance with the results obtained by Northern blot analysis, the strong signal was detected with sections prepared from the functional stage (Fig. 3A); inhibitor molecules were found to be localized in the connective tissue around blood vessels as well as in the capsule of corpus luteum. In contrast, only faint and diffused staining was observed with sections of regressing corpora lutea in the correspondig region (Fig. 3B). Discussion

By applying differential screening to granulosa cell- and corpus luteum-derived cDNA libraries, we obtained cDNA clones encoding genes expressing predominantly in either cell type. The stages where they are expressed during the follicular development and luteinization were confirmed by Northern blot analysis on RNAs ex-

10

Pig

20

tracted from granulosa cells of antral follicles and the defined stage of corpus luteum. A cDNA clone encoding porcine collagenase inhibitor was identified as one of corpus luteumspecific clones and characterized in this paper. The expression was most prominent at the functional stage by Northern blot analysis and this was further confirmed by immunohistochemical analysis on pig ovary; the product was rather specifically localized in the connective tissue and in the luteal cell junction. It seems likely that collagenase inhibitor plays a role in the regulation of the collagenase activity (Brannstrom et al., 19881 and the disappearance of the inhibitor at the regressing stage was observed, suggesting that the rapid regression of corpus luteum and the progression to corpus albicans reflect the level of inhibitor activity. Curry et al. (1990) demonstrated the presence of TIMP in human follicular fluid and further indicated that the granulosa cells are an impor-

30

40

50

Human

MSPFRPLASGILLLLnTAPSRACTCVPPHPQTAFCSSDLVIRAKFVGAPEFNQTASYQR _A-_S____________~__________________N_________--T--V___TL___

Bovine

_A-_--~-____________________________N__V---_____TA_V_B__L___

Rabbit

___L_A___SML_____V____-__-__________N_______---____V_H_TL___

MOUSGS

~_A___S___------S-~-S-K--S-A---------N_________N_S--~___TL___

10

80

90

100

110

pig

YEII(MTKMFKGFNALGDAPDIRFIYTP~SVCGYFHRSQNRSQEFLIAGQL~GHLHIT

Human

________Y___Q_____A____V______---------H---E--____K_QD_L____

Bovine

-_----______~__~______________---------____~________~-------

Rabbit

-___T_______D___H_T____V______-----S-K--___~________~_-L----

Mouse

-~---___~___~_~_~_~___YA--_V___L---A_K__~~~~~~_~_____~____~-~-~--~_-_~

Pig

TCSFVAPWNSLSSAQRQGFTEIYRAGCEECTVFPCTSIPCKLQSDTHCLWTDQLLTGSDK

Human

____________~___~___~~_~~__________~___-----~-------___Q__~_

Bovine

_---------~-----~---~~-------------~--______________________

130

140

Rabbit Mouse

190

200

Pig

GFQSRHLACMPREPGMCTWQSLRPRVA

Human

---------L-----L-------SQI-

Bovine

_____-___L_____L_______AQ”_

Rabbit

_________L_Q___L_A_E_____KD

Mouse

~----~__~--~-_~--_~__~~

150

160

170

60

120

180

sense TIMP mRNA could be detected at any stage, and neither immature nor antral follicles contained detectable levels of TIMP transcripts, and the density of hybridization grains in oocytes was similar with both positive- and negative-strand RNA probes (Nomura et al., 1989). Our results have demonstrated that the TIMP gene product was identified in the connective tissue around blood vessels in the corpus luteum and in the capsule of corpus luteum. Taken together, TIMP is likely to be synthesized in the luteal cells and accumulated in the connective tissue of the corpus luteum and the capsule. The antibody used in this study was originally raised against bovine collagenase inhibitor. According to the developer, this antibody also cross-reacted against its human counterpart (J&hi and Hayakawa, 1984). While the cross-reactivity has not been checked between bovine and porcine proteins, the strong similarity of the primary sequences among human, bovine and porcine products (Table 1) strongly suggests that the immunostaining observed in this study was due to the porcine coltagenase inhibitor molecule. Fig. 3. Immunohistochemical stainings of corpora lutea at the functional (A) and regressing (!3) stage. Sliced sections were incubated with horseradish peroxidase-labeled anti-collagenase inhibitor antibody, followed by the reaction with DAB-4HCl as described in Materials and methods. Nuclei of the preparation were stained with hematoxylin solution. X 100.

tant source of inhibitory

activity. However, they removed granulosa cells from human ovaries injected with either human menopausal gonadotropin alone or clomiphene citrate alone, or in combination. Moreover, human chorionic gonadotropin was administered to stimulate the finaI stages of folli~ular maturation. Under these hormonal environments, granulosa cells might be already committed to luteinization and, thereby, TIMP mRNA was likely to be induced. On the other hand, high expression of TIMP mRNA has been detected in the corpus luteum of the mouse ovary using single-stranded antisense RNA as a probe (Nomura et al., 1989). Sections of ovaries from both adult cycling and pregnant females showed similar patterns of TIMP expression. No significant hybridization to thecal cells with anti-

References Brannstrom. M.. Woessner, J.F., Koos, R.D., Sear, C.H.J. and LeMaire, W.J. (198X) Endocrinology 122, 1715-1721. Carmichael, D.F.. Sommer, A.. Thompson. R.C., Anderson. D.C., Smith, C.G., Welgus, H.G. and Stricklin. G.P. (1986) Proc. Nat]. Acad. Sci. USA 83, 2407-2411. Channing, C.P. and Ledwitz-Rigby, F. (1975) in Methods in Enzymology (Hardmann, J.G. and O’Malley, B.W., eds.), Vol. 39. pp. 183-230, Academic Press, New York. Channing, C.P.. Schaerf. F.W., Anderson, L.D. and Tsafriri, A. (1980) Int. Rev. Physiol. 22, 117-201. Channing, C.P., Anderson, L.D.. Hoover, D.J., Kolena, J., Osteen, K., Pomerantz, S.H. and Tanabe, K. (1982) Recent Prog. Horm. Res. 38, 33 I-408. Curry, Jr., T.E., Mann, J.S.. Estes, R.S. and Jones. P.B. (1990) Endocrinology 127, 63-68. Edwards, D.R., Waterhouse, P., Holman, M.L. and Dcnhardt, D.T. (1986) Nucleic Acids Res. 14, 8863-8878. Horowitz, S.. Dafni, N., Shapiro, D.L., Helm, B.A., Notter, R.H. and Quible, D.J. (1989) J. Biol. Chem. 264, 70927095. Kishi, J. and Hayakawa, T. (1984) J. Biochem. 96, 395-404. LeMaire, W.J., Curry. T.E., Morioka, N.. Brannstrom. M., Clark. M.R., Woessner, J.F. and Koos, R.D. (1987) in The Primate Ovary (Stouffer, R.L., ed.). pp. 91-105. Plenum Publishing Corp., New York.

71 Lipner, H. (1988) in The Physiology of Reproduction (Knobil, E. and Neill, J., eds.), pp. 447-461, Raven Press, New York. Masui, Y. and Clarke, H.J. (1979) Int. Rev. Cytol. 57,185-282. Meyer, G.T. and Bruce, N.W. (1989) in Development in Ultrastructure of Reproduction (Motta, P.M., ed.), pp. 131-151, Alan R. Liss, New York. Nakane, P.K. and Pierce, Jr., G.B. (1967) J. Cell Biol. 33, 307. Nomura, S., Hogan, B.L., Wills, A.J., Heath, J.K. and Edwards, D.R. (1989) Development 105, 575-583.

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual (2nd edn.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. Sato, E. and Koide, S.S. (1982) Int. Rev. Cytol. 106, l-33. Wagner, J.F., Luck, R.M., Einspanjer, R. and Scheit, K.H. (1990) Biochem. Biophys. Res. Commun. 171, 250-256. Ying, S.-Y. (1988) Endocr. Rev. 9, 267-293.