Vol. 180, No. 2, 1991 October 31, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 481-488

RESPONSE TO AND EXPRESSION OF AMPHIREGULIN BY OVARIAN CARCINOMA AND NORMAL OVARIAN SURFACE EPITHELIAL CELLS: NUCLEAR LOCALIZATION OF ENDOGENOUS AMPHIREGULIN

Gibbes R. Johnson 1., Toshiaki Saeki 2, Nelly Auersperg 3, Alfred W. Gordon 1, Mohammed Shoyab 4, David S. Salomon 2 and Kurt Stromberg I

1Laboratory of Cell Biology, Division of Cytokine Biology, Food and Drug Administration, Bethesda, MD 20892 2Laboratory of Tumor Immunology and Biology, National Cancer Institute, NIH, Bethesda, MD 20892 3Department of Anatomy, University of British Columbia, Vancouver, B.C., Canada V6T 1W5 4Oncogen, Bristol-Myers Squibb Pharmaceutical Research Institute, Seattle, WA 98121 Received September 6, 1991

Summary; Amphiregulin (AR) is a polypeptide growth regulator which has sequence homology to the epidermal growth factQr-related family of ligands and contains putative nuclear targeting sequences. Human ovarian carcinoma cell lines and their normal counterparts, ovarian surface epithelial cells (OSEs), were assessed for their ability to respond to and express AR. Addition of exogenous AR (8-200 pM) inhibited the growth of 2 of 3 0 S E specimens and 3 of the 6 carcinoma cell lines indicating that AR has the potential to inhibit the growth of normal cells, in addition to carcinoma cells. In contrast, concentrations of AR ranging from 1-5 nM stimulated the growth of all 3 of the OSEs and 4 of the 6 carcinoma cell lines. Immunocytochemical staining of the cells using antipeptide antibodies directed against residues 8-26 of AR indicated that all cells expressed AR and that the staining was localized to the nucleus. The nuclear staining of AR was concentrated in the nucleolus of the carcinoma cells, whereas the staining was diffuse in the nucleus of the OSEs. These results suggest that AR may play a growth regulatory role in the nucleus of cells and this role may be different in normal and malignant epithelial cells. © 199~ Academic Press, Inc.

Amphiregulin (AR) is a newly discovered growth regulatory glycoprotein which was originally purified and characterized from the conditioned medium of the human breast carcinoma cell line, MCF-7, after treatment with 12-O-tetradecanoylphorbol-13-acetate (TPA) (1,2). AR exhibits amino acid sequence homology to the epidermal growth factor (EGF)family of ligands, contains putative nuclear targeting signals and competes with 1251_EGF for

*To whom all correspondence should be addressed. The abbreviations used ar~; AR, amphiregulin; OSEs, ovarian surface epithelial cells; EGF, epidermal growth factor; TPA, 12-O-tetradecanoyl-phorbol-13-acetate; BSA, bovine serum albumin; PBS, phosphate-buffered saline; IgG, immunoglobulin G; TGF-~ transforming growth factor-~ ;bFGF, basic fibroblast growth factor. 0006-291X/91 $1.50 481

Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

Vol. 180, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

binding to the EGF receptor (2). AR stimulates the growth of normal fibroblasts and murine keratinocytes, but potently inhibits the growth of some carcinoma cell lines (1,2). The 84 amino acid mature AR protein is believed to be proteolytically derived from a 252 amino acid transmembrane precursor (3). Many carcinoma cell lines and normal tissues have been shown to express the 1.4 kb mRNA transcript for AR (3). Of normal tissues tested, ovary and placenta were found to express the highest levels of AR mRNA. More recently, human keratinocytes and normal human mammary epithelial cells in culture were shown to express AR mRNA and/or secrete AR protein (4). In the work reported here, we have compared the response to and expression of AR by ovarian carcinoma cell lines and the normal cells from which they are believed to arise, ovarian surface epithelial cells (OSEs). We show for the first time that normal cells can be growth-inhibited by AR and that whether the growth of a cell is inhibited or stimulated may depend upon the concentration of AR to which the cell is exposed. In addition, this study also demonstrates that AR can be targeted to the nucleus of cells that produce it, suggesting that this growth modulatory polypeptide might elicit a direct biological response in the nucleus.

MATERIALS AND METHODS Materials- The IGROV-1, NIH:OVCAR-4 and NIH:OVCAR-8 cell lines were generously provided by the Tumor Repository at the FCRDF (Frederick, MD). The OVCA 420 and 429 cell lines were generously donated by Drs. Robert Bast and Cinda Boyer at Duke Univ. The SK-OV-3 cell line was acquired from the American Type Culture Collection. All' tissue culture reagents were purchased either from GIBCO BRL (Gaithersburg, MD) or Sigma (St. Louis, ME)). 5-[1251]lodo-2'-deoxyuridine was purchased from Amersham (Arlington Heights, IL). Cell (~ulture- The human ovarian surface epithelial (OSEs) specimens from three individuals (referred to as Boo, Gar and Ert) were prepared and grown in culture as previously described (5). Cells within passages 2 through 8 were used for the studies. Ovarian carcinoma cell lines were grown as described by Stromberg eta/. (6). Reaulation of Cell Growth bv Amohireaulin- AR was purified to homogeneity and characterized as described by Shoyab et a/. (1, 2). Growth regulation assays on the carcinoma cell lines were performed in 5% fetal bovine serum as previously described (1,7). All experimental wells were performed in triplicate with the exception of control wells, in which at least 6 wells were used. OSE cells were plated at 3,000 cells per well in a reduced serum medium (8) and the 1251.deoxyuridin e was added 24 h after the addition of AR. The cells were allowed to incorporate label for a period of 24 h and then processed as described (1,7). Generation and Purification of AntipeDtide Antibodies aaainst AmDhireaulinCys-AR8-26-NH2 was synthesized and purified under contract by Multiple Peptide Systems (San Diego, CA). Briefly, the peptide was synthesized by the method of Houghten (9) and purified by preparative reverse phase high performance liquid chromatography. Twenty mg of the peptide was then conjugated through its lone side chain sulfhydryl to 16 mg of keyhole limpet hemocyanin using m-Maleimido-benzoyI-N-hydroxysuccinimide ester and this conjugate was used to immunize rabbits (10). Prior to immunization, the rabbits were bled to generate control preimmune serum. Antisera were screened for a response to the peptide starting 10 days after the first boost, as follows: Forty ng of the peptide in 50 ILl of 0.1 M NaHCO3, pH 9.1 was incubated in each well of an Immulon II ELISA plate (Dynatech Laboratories, Chantilly VA) for 16 h at 4oc. The wells were then blocked with 3% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 2 h at ambient temperature. Fifty p.I of various dilutions of preimmune and post-immune sera in 1% BSNPBS was added to each well and incubated for 2 h at ambient temperature. Bound IgG was detected using biotinylated donkey anti-rabbit IgG, F(ab')2 fragment followed by streptavidin-alkaline phosphatase (Amersham). Color development was performed using the phosphatase substrate system for ELISA

482

Vol. 180, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

(Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD). One rabbit generated postimmune serum which specifically detected the peptide at serum dilutions ranging from 1,000 to 500,000, but showed no cross reactivity with transforming growth factor-~ at all dilutions tested. This antisera was able to specifically detect 5 ng of purified AR and as little as 40 pg of peptide antigen in this ELISA format (data not shown). The antisera was also found to immunoprecipitate AR from the conditioned medium of TPA-treated MCF-7 cells which had been metabolically labeled with [35S]cysteine. This antiserum is designated AR-Abl and is the source of the Immunoglobulin G (IgG) utilized in this study. IgG was purified from control preimmune serum and antiserum using Immunopure Immobilized Protein A/G gel (Pierce, Rockford, IL) as described by the manufacturer. I m m u n o c v t o c h e m i s t r v - Cells were plated into Lab-Tek 4 chamber slides (Nunc, Inc.,

Naperville, IL) at 20,000 cells per chamber and grown to approximately 70% confluence. The cells were washed twice with PBS and fixed with 1.5 % formaldehyde in PBS for 45 min at room temperature. The cells were then washed four times with PBS. At this step, cells could be treated with 0.5 % Triton X-100 in PBS for 5 min at room temperature to slightly enhance the staining of AR in the nucleus of cells. Ten percent goat serum in PBS (0.5 ml) was then added to each chamber and incubated for 45 min at room temperature. This solution was aspirated and 0.5 ml of 10 p.g/ml preimmune (control) IgG or post-immune IgG in 0.1% bovine serum albumin/PBS was added. Post-immune IgG that was preabsorbed with peptide (1 pg IgG per 1 I~g peptide) for 2 h at 37oc was also utilized at this step to confirm specificity. After a 1 h incubation at room temperature the cells were washed twice with PBS and bound IgG was detected using the Vectastain ABC kit for rabbit IgG (Vector Laboratories, Burlingame, CA). The following cell lines, which have been shown to be positive and negative for AR mRNA expression were used as controls: MCF-7 treated with TPA (positive) (3), Geo (positive), LS 174T (positive), JVC (negative) and WiDr (negative) (11). The staining was evaluated and scored as described by Reiner et aL (12).

RESULTS Reaulation of cell orowth bv exoaenous AR- The growth of cells in response to

the exogenous addition of various concentrations of AR was studied as previously described (1,7). AR inhibited the growth of 2 of the 3 normal OSE specimens (Err and Gar), at concentrations ranging from 1.5-200 pM (Fig. 1). Similar concentrations of AR had no significant effect on the growth of the third OSE specimen, referred to as Boo. However, at a concentration of 5 nM AR, the growth of all three OSE specimens was stimulated, and in the case of Boo, by approximately 50%. The effect of AR on the growth of the ovarian carcinoma cell lines was found to be even more complex (Fig. 2). AR had no significant effect on the growth of NIH:OVCAR-8 (panel A), only stimulated OVCA 420 and NIH:OVCAR-4 (panel A) and at all concentrations tested inhibited IGROV-1 (panel B). Two other cell lines, OVCA 429 and SK-OV-3, yielded a biphasic dose-response curve, in which their growth was inhibited at AR concentrations ranging from 8-200 pM and stimulated at concentrations of 1 and 5 nM (panel B). Overall, the following growth-response patterns emerged: (1) doses of AR in the 1.5-40 pM range inhibited or had no significant effect on the growth of the cells tested; (2) doses of 15 nM AR usually led to a stimulation of cell growth and (3) both normal and neoplastic ovarian epithelial cells can yield biphasic dose-response curves, in which their growth is inhibited at low doses of AR but stimulated at higher concentrations. I m m u n o c v t o c h e m i c a l analysis of AR exoression bv OSEs and ovarian carcinoma cell lines- Cells were examined for the expression of AR using antibodies

generated against a synthetic peptide corresponding to residues 8-26 of AR. This region of AR

483

V o l . 180, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

15o

-

125 '

100

75.

50 0

........ , ........ , ........ , .001 .01 .1

........ , 1

........ 10

Amphlregulln (nM) Fi.q, 1. Regulation of Growth of Normal Ovarian Surface Epithelial Cells by A m p h i r e g u l i n . Human ovarian surface epithelial specimens from three individuals, referred to as Err (I), Gar (0) and Boo (X), were assayed for growth modulation by AR as described in Materials and Methods. Data points represent the mean _+SE of triplicate determinations. Asterisks denote data points which were statistically different from control values (P < 0.05).

does not contain the putative nuclear targeting sequences (residues 26-29 and 40-43) and was chosen for the generation of antibodies because it has no significant sequence homology to any known protein (2). As described in the Materials and Methods section, these antibodies recognize purifed native AR as well as the peptide antigen. Optimization and validation of the immunocytochemical procedure w a s performed using a panel of carcinoma cell lines which are

220

A

,,°t

2OO180O O

"5

2

160-

0

140 -

ae

80

120 60.

100~ 80

O

.001

.01

.1

1

10

40

o

.001

.01

.1

1

Amphiregulin (nM)

Amphiregulin (nM)

Fig. 2. Regulation of Growth of Ovarian Carcinoma Cell Lines by Amphiregulin. Growth modulation of human ovarian carcinoma cell lines by amphiregulin was assayed as described in Materials and Methods. A, OVCA 420 (e); NIH:OVCAR-4 (0) and NIH:OVCAR-8 (X). B, IGROV-I(e); SK-OV-3 (0) and OVCA 429 (X). Data points represent the mean _+SE of triplicate determinations. Asterisks denote data points which were statistically different from control values (P < 0.05). 484

lO

Vol. 180, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Table 1. Immunocytochemical Staining of Amphiregulin in Ovarian Carcinoma and Normal Ovarian Surface Epithelial Cellsa Immunocytochemical Stainingb Cells

Intensity

Heterogeneity

Ovarian Surface Epithelial: Gar Ert Boo

1 1 2

2 3 4

Ovarian Carcinoma: OVCA 429 OVCA 420 SK-OV-3 IGROV-1 NIH:OVCAR-4 NIH:OVCAR-8

2 2 2 2 2 2

4 4 3 4 4 4

almmunocytochemistry was performed as described in Materials and Methods using purified IgG directed against residues 8-26 of AR. Carcinoma cell lines that are positive and negative for AR mRNA expression were used as controls as described in Materials and Methods. bStaining was scored as described in Reiner et aL (12) based upon the intensity and heterogeneity of staining. The intensity of the specific immunostaining was scored on a scale of 1 (weak) to 3 (strong). The staining observed with MCF-7 cells treated with TPA represented a score of 3. The heterogeneity of staining was scored based upon the percentage of cells staining positive: a score of 1 represents that G1 transition (13). More recently, the translocation of bFGF to the nucleus has been shown to be a G1 phase cell cycle specific event (14). Intense localized immunostaining of bFGF was seen in the nucleoli of asynchronous growing cells and in Gl-arrested cells undergoing the G1-->S transition after stimulation by bFGF. A weak diffuse nuclear staining was observed in non-proliferative, confluent cells (14). The nucleolar immunostaining of bFGF observed by these workers is strikingly similar to the staining pattern that we have observed for AR which has accumulated in the nucleoli of ovarian carcinoma cells. Taken together, our results suggest that AR plays an important nuclear role in the growth of ovarian cells in vitro. The preferential localization of AR in the nucleoli of the ovarian carcinoma cells as compared to the more diffuse nuclear staining that was observed in the OSEs suggests that the role of AR may relate to either their differential proliferative rate, or perhaps to some more fundamental difference in AR regulation of normal versus malignant cell growth. In this regard, studies are currently planned to examine the expression of AR in ovarian carcinomas in vivo. It is not clear whether the AR which is localized in the nucleus results from internalized material that has been previously secreted by the cells or is the result of the direct targeting of AR to the nucleus after biosynthesis. In this respect, the AR precursor contains both a signal peptide (3) and nuclear targeting signals (2). It has been shown that attachment of a nuclear 487

Vol. 180, No. 2, 1991

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

translocation sequence from yeast histone 2B to a truncated defective heparin-binding growth factor-1 mutant, results in the recovery of its mitogenic activity (19). Recently, Higashiyama et al. (20) discovered a heparin-binding

EGF-like growth factor (HB-EGF), which in overall

structure appears to be more closely related to AR, than to other members of the EGF family. This novel growth factor also contains one of the putative nuclear localization signals that is present in the AR molecule. In summary, it will be critical to determine the biological consequences of AR in the nucleus of both normal and malignant epithelial cells so as to understand the potential differential role that AR plays in their respective growth.

ACKNOWLEDGMENTS The authors wish to thank Sara Maines-Bandiera and Carolyn Jackson for technical assistance and Drs. Raj Puri and Ray Donnelly for the critical review of the manuscript. This work, in part, was supported by a research associateship from the National Cancer Institute of Canada and a grant from the National Research Council of Canada.

1. Shoyab, M., McDonald, V. L., Bradley, J. G., and Todaro, G. J. (1988) Proc. Natl. Acad. Sci. USA 85, 6528-6532. 2. Shoyab, M., Plowman, G. D., McDonald, V. L., Bradley, J. G., and Todaro, G. J. (1989) Science 243, 1074-1076. 3. Plowman, G. D., Green, J. M., McDonald, V. L., Neubauer, M. G., Disteche, C. M., Todaro, G. J., and Shoyab, M, (1990) Mol. Cell. Biol. 10, 1969-1981. 4. Cook, P.W., Mattox, P.A., Keeble, W. W., Pittelkow, M. R., Plowman, G. D., Shoyab, M., Adelman, J. P., and Shipley, G. D. (1991) Mol. Cell. Biol. 11, 2547-2557. 5. Kruk, P. A,, Maines-Bandiera, S. L,, and Auersperg, N. (1990) Lab. Invest. 63, 132-136. 6. Stromberg, K., Collins, T. J., Gordon, A. W., Jackson, C. L., and Johnson, G. R. (1991) submitted for publication. 7. Iwata, K. K., Fryling, C. M., Knott, W. B., and Todaro, G. J. (1985) Cancer Res. 45, 26892694. 8. EIliott, W. M., and Auersperg, N. (1990) J. Cell Biol. 111, 58a. 9. Houghten, R. A. (1985) Proc. Natl. Acad. Sci. USA 82, 5131-5135. 10. Harlow, E., and Lane, D. (1988) In Antibodies, a Laboratory Manual. pp. 53-137. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 11. Ciardiello, F., Kim, N., Saeki, T., Dono, R., Persico, M. G., Plowman, G. D., Garriques, J., Radke, S., Todaro, G. J., and Salomon, D.S. (1991) Proc. Natl. Acad. Sci. USA, in press. 12. Reiner, A., Spona, J., Reiner, G., Schemper, M., Kolb, R., Kwasny, W., Fugger, R., Jakesz, R., and Holzner, J. H. (1986) Am. J. Pathol. 125, 443-449. 13. Bouche, G., Gas, N., Prats, H., Baldin, V., Tauber, J.-P., Teissie, J., and Amalric, F. (1987) Proc. Natl. Acad. Sci. USA 84, 6770-6774. 14. Baldin, V., Roman, A.-M., Bosc-Bierne, I., Amalric, F., and Bouche, G. (1990) EMBO J. 9, 1511-1517. 15. Siemens, C. H., and Auersperg, N. (1988) J. Cell Physiol. 134, 347-356. 16. Defize, L. H. K., Boonstra, J., Meisenhelder, J., Kruijer, W., Tertoolen, L. G. J., Tilly, B. C., Hunter, T., van Bergen en Henegouwen, P. M. P., Moolenaar, W. H., and de Laat, S. W. (1989) J. Cell Biol. 109, 2495-2507. 17. Bellot, F., Moolenaar, W., Kris, R., Mirakhur, B., Verlaan, I., UIIrich, A., Schlessinger, J., and Felder, S. (1990) J. Cell Biol. 110, 491-502. 18. Burwen, S. J., and Jones, A. L. (1987) Trends Biochem. Sci. 12, 159-162. 19. Imamura, T., Engleka, K., Zhan, X., Tokita, Y., Forough, R., Roeder, D., Jackson, A., Maier, J. A. M., Hla, T., and Maciag, T. (1990) Science 249,1567-1570. 20. Higashiyama, S., Abraham, J. A., Miller, J., Fiddles, J. C., and Klagsbrun, M. (1991) Science 251, 936-939. 488