JOURNAL OF PATHOLOGY, VOL.
168 275-280 (1992)
PRODUCTION A N D CHARACTERIZATION OF A POLYCLONAL ANTIBODY TO THE c-erbB-3 PROTEIN: EXAMINATION OF c-erbB-3 PROTEIN EXPRESSION IN ADENOCARCINOMAS D. N. POLLER*, I. SPENDLOVET, C. BAKER?, R. CHURCH*, I. 0. ELLIS$, G . D. PLOWMAN5 AND R. J. MAYER?
University Departments of Histopathology*and Biochemistry?, Queen's Medical Centre, Nottingham NG7 2UH, U.K.; $Department of Histopathology, City Hospital, Nottingham, NG5 IPB, U.K.;GBristol-Myers Squibb Pharmaceutical Research Institute, 3005 First Avenue, Seattle, Washington 98121, U.S.A. Received30 December 1991 Accepted 6 July 1992
SUMMARY A polyclonal rabbit antibody was raised to the c-erbB-3 protein using a synthetic peptide corresponding to amino acids 1229-1241 of the predicted protein sequence of c-erbB-3. In Western blot analysis this antibody detects a single band at 165 kD in a c-erbB-3 transfected (293/HER-3) human cell line. c-erbB-3 protein expression was then examined in a variety of adenocarcinomas. Expression of c-erbB-3 protein was indicated by membrane and/or cytoplasmic tumour cell immunoreactivity in formalin-fixed, paraffin-embedded tissue sections. c-erbB-3 protein was detected in a series of 13 out of 14 primary breast carcinomas, 3 of 5 gastric adenocarcinomas, 8 of 9 colonic adenocarcinomas, 2 of 9 prostatic adenocarcinomas, 0 of 6 renal cell carcinomas, 1 of 4 primary lung adenocarcinomas, and 5 of 7 endometrial adenocarcinomas. Immunohistochemical expression of the c-erbB-3 protein appears to be a relatively common event in adenocarcinomas, and further studies are now warranted to establish the role of the c-erbB-3 protein in neoplasia.
-
KEY wo~~s+-erbB-2,EGFR, c-erbB-3, adenocarcinoma,
INTRODUCTION The c-erbB-3 gene' also known as HER3,' is a recently identified gene that codes for a member of a family of growth factor receptors showing great similarity in structure to the epidermal growth factor receptor (EGFR) and the c-erbB-2 proteins."* The c-erbB-3 gene has been mapped to chromosome 12ql3.' The c-erbB-3 protein is a membrane-bound, receptor-like protein with presumed tyrosine kinase activity, which shows considerable structural homology to both EGFR and the c-erbB-2 protein^.^ The c-erbB-3 protein at its predicted tyrosine kinase domain shows 64 per cent amino acid sequence homology to the tyrosine kinase domain of EGFR Addressee for correspondence: Dr D. N. Poller, Department of Histopathology, City Hospital, Nottingham NG5 IPB, U.K.
0022-3417/92/110275-06 $08.00 0 1992 by John Wiley & Sons, Ltd.
immunohistochemistry.
and 67 per cent amino acid sequence homology to the tyrosine kinase domain of the c-erbB-2 protein.' Expression of EGFR has been shown to be of prognostic importance in predicting the biological ~.~ behaviour of adenocarcinomas of the b r e a ~ tand colon,6 and also in squamous cell c a r c i n ~ m a . ' ~ ~ Overexpression of the c-erbB-2 protein in human malignancies, a feature usually associated with c-erbB-2 gene amplifi~ation,~ has been shown to be an indicator of adverse pro nosis in breast9-" and ovarian adenocarcinomas," as well as being a feature of mammary ductal carcinoma in situ of comedo type." A polyclonal rabbit antibody to a synthetic peptide corresponding to amino acids 1229-1241 of the predicted amino acid sequence of the c-erbB-3 protein has been produced.'** The expression of c-erbB-3 protein was then examined in a variety of adenocarcinomas.
276
D. N. POLLER ETAL.
presence of sodium dodecyl sulphate, the proteins were transferred to Hybond-C Super (Amersham International PLC, Slough, Berks, U.K.). A Preparation of antibody A peptide corresponding to amino acids 1229-1241 standard western development was followed. The of the predicted sequence of the c-erbB-3 protein transfer was blocked for 1 h in Tris-buffered saline (sequence residues IMPTAGTTPDED) was syn- (200mM NaCl, 5 0 m ~Tris-HC1, pH 7.4), 0.1 thesized on a model 431A peptide synthesizer per cent Tween 20, 5 per cent Marvel, and then (Applied Biosystems, Warrington, U.K.) Using incubated for 2 h with the c-erbB-3 protein primary Fmoc chemistry. The peptide was synthesized with antibody in Tris-buffered saline (TBS) at 1:200 an additional C-terminal cysteine for linkage to dilution, 0.1 per cent Tween 20, 2 per cent Marvel keyhole limpet haemocyanin, using the heterobi- (TBSTM). Following a 1 h wash in TBST, the blot functional reagent m-maleimidobenzoyl-N-hydroxy- was incubated with a ljlO000 dilution of horsesuccinimide ester, and then injected intramuscularly radish peroxidase conjugated swine anti-rabbit and subcutaneously into two rabbits. The resulting secondary antibody (Dako, U.K. Ltd., High polyclonal antiserum was then used without affinity Wycombe, Bucks, U.K.) in TBSTM. Following a purification. High ELISA titres were obtained final20minwashinTBS, theblotwasdevelopedwith Amersham ECL (enhanced chemiluminescence) (1 :500 000) against the synthetic peptide. detection reagent. Duplicate samples were also incubated in the absence of primary antibody Western blotting c-erbB-3 antibody as negative controls. The anti-c-erbB-3 protein antibody was characterized using EGFR, c-erbB-2 and c-erbB-3 transfected Immunohistochemistry cell lines. The EGFR tran~fectant'~ and the c-erbB-2I4 Four pm thick sections of formalin-fixed, paraffintransfectants were cell lines derived from the Fall embedded tumour tissue were cut, dewaxed in Army worm Spodopterafrugiperda (Sf 9), which can be infected by the naturally occurring baculovirus xylene, and then rinsed in alcohol and graded Autographa calfornica nuclear polyhedrosis virus alcohol-water mixtures. A standard avidin-biotin (AcNPV), a non-mammalian expression vector that peroxidase technique was used. Incubation with the uses the strong AcNPV polyhedrin promoter to c-erbB-3 primary antibody was performed for 30 drive production of large amounts of a foreign, min at a dilution of 1:800. Following incubation eukaryotic protein-in this case, EGFR13 or with the primary antibody, a secondary polyclonal c-erbB-2 protein.I4 The Sf 9 cells were cultured as a biotinylated swine anti-rabbit antibody was used monolayer at 27°C in IPL4-I media supplemented (Dako U.K. Ltd.). This was followed by an avidin with 10 per cent FCS, 1 per cent Fungizone, and 0.1 biotinylated peroxidase complex, with 3,3-diaminoper cent gentamycin. A proximately 4 x lo6 cells benzidene as a chromagen. Each case was also were plated into a 25 cm flask and then the EGFR incubated in the absence of the anti-c-erbB-3 or c-erbB-2 containing AcNPV was then added to primary antibody as a negative control. Absorption the flask. The baculovirus concentrations had pre- specificity controls using the preimmune peptide viously been titred to give optimal expressed protein were also performed on some of the positively levels. A non-transfected negative control Sf 9 cell staining tumours, with inhibition of c-erbB-3 line was also used. The cells were harvested 48 h staining. The sections were examined by light after infection, centrifuged, washed in PBS, and microscopy. Formalin-fixed cytospins of a human lysed with Laemmeli sample buffer (A). The c-erbB-3 breast carcinoma cell line MDA MB415, known to transfectant was a human embryonic cell line, 293 express high levels of c-erbB-3 mRNA,' were used as (American Type Culture Collection), which after a positive control for immunohistochemistry. stable c-erbB-3 transfection expresses abundant amounts of recombinant c-erbB-3 protein (personal RESULTS communication, Dr G. D. Plowman). Approximately 100,ug samples of protein were loaded onto a 7 per cent polyacrylamide gel along Western blotting with prestained molecular weight markers (Sigma The results of Western blots using the anti-cChemical Company Ltd., Fancy Road, Poole, erbB-3 antibody are shown in Fig. 1. In Fig. 1, Dorset, U.K.). Following electrophoresis in the lane 1 shows a single discrete band at 165 kD, MATERIALS AND METHODS
P
N
PRODUCTION AND CHARACTERISTICS OF POLYCLONAL ANTIBODY TO THE c-erbB-3 PROTEIN
corresponding to the subunit of c-erbB-3, with the c-erbB-3 transfected 293 cell line. Lane 2 is the control non-transfected 293 cell line which shows absence of the c-erbB-3 protein subunit.
277
Immunohistochemistry The antibody appeared to work well on formalinfixed, paraffin-embedded tissue, with little or no background staining. Smooth muscle showed weak granular cytoplasmic staining, present in the breast, muscularis mucosae and propria of the stomach, small intestine, colon, and myometrium, and within the walls of arteries and muscular veins. The results of the immunohistochemical examination of c-erbB-3 protein expression in adenocarcinomas are shown in Table I. Breast carcinoma The 14 primary breast carcinomas examined comprised 13 invasive ductal carcinomas of no special type, and one invasive lobular carcinoma. Thirteen tumours showed membrane or cytoplasmic c-erbB-3 immunoreactivity and of these, four tumours showed some focal membrane and cytoplasmic immunoreactivity of tumour cells. Nine tumours showed only cytoplasmic staining of tumour cells. Cytoplasmic c-erbB-3 immunoreactivity was also seen in the in situ components of invasive ductal carcinomas, where there was also expression of c-erbB-3 in the invasive component (Figs 2 and 3). Normal mammary duct also showed cytoplasmic staining of epithelial cells with some staining of myoepithelial cells, and areas of apocrine metaplasia within foci of fibrocystic disease. Normal breast acini did not show c-erbB-3 protein expression in the limited number of cases examined.
Gastric adenocarcinoma Five primary gastric carcinomas were examined comprising four of intestinal type and one of diffuse type.I5 Three intestinal type adenocarcinomas showed heterogeneous membrane and cytoplasmic c-erbB-3 immunoreactivity, and one was c-erbB-3 negative, with no evidence of c-erbB-3 expression in the single case of diffuse type gastric carcinoma Fig. I-Lane 1: A single discrete band at 165 kD correspond- examined. Expression of the c-erbB-3 protein was ing to the subunit of c-erbB-3 is seen with the c-erbB-3 transfected seen in the cytoplasm of normal foveolar epithelial 293 cell line. Lane 2: The control (non-transfected) 293 cell line cells of the gastric crypts, predominantly within the showing absence of the c-erbB-3 protein subunit supranuclear Golgi zone. The parietal cells lining the gastric glands also showed strong granular Western blots performed with wild type (non- c-erbB-3 positivity, although there was no staining baculovirus transfected) Sf 9 cells, EGFR transfected of chief cells or mucus neck cells. Sf 9 cells, and the c-erbB-2 transfected Sf 9 cell lines showed absence of cross-reactivity with the Carcinoma of the colon All nine cases examined were primary invasive anti-c-erbB-3 antibody, no band being detected on adenocarcinomas of large intestinal type, one of Western blotting (not illustrated).
-
D. N. POLLER ETAL.
278
Table I.--erbB-3 protein expression and tumour primary site as assessed in formalin-fixed,paraffin-embeddedtissue
Primary site Breast Stomach Colon Prostate Kidney Lung Endometrium
Fig. 2-An
Cytoplasmic
Pattern of immunoreactivity Membrane and No staining Total cytoplasmic
9 0 2 2 0 0 1
4 3 6 0 1 1 4
1 2 1
7 6 3 2
14 5 9 9 7 4 7
invasive ductal carcinoma of the breast showing
strong cytoplasmictumour cell c-erbB-3immunoreactivityin the in situ component (arrow-heads)
Fig. 3-An invasive ductal carcinoma of the breast showing strong cytoplasmic c-erbB-3immunoreactivity
which was classified as a mucinous (colloid) carcinoma.16 Eight out of the nine cases showed expression of the c-erbB-3 protein. Six tumours showed membrane and cytoplasmic immunoreactivity, and two cases showed only cytoplasmic c-erbB-3 staining, including the single case of mucinous carcinoma included in the series. Membrane expression of c-erbB-3 was more intense at invasive tumour margins (Fig. 4). Granular cytoplasmic staining in surrounding normal absorptive and goblet cells of the large intestinal mucosa was also identified.
The pattern of staining was cytoplasmic and granular without evidence of membrane staining. Normal prostatic glandular acini showed no evidence of c-erbB-3 expression, although granular c-erbB-3 staining was present in normal prostatic ducts in both epithelial and myoepithelial cells. Normal transitional mucosa from the bladder neck also showed membrane and cytoplasmic c-erhB-3 immunoreactivity .
Adenocarcinoma of the prostate c-erbB-3 protein expression was present in two of nine invasive prostatic adenocarcinomas examined.
Renal tumours Of the six primary renalcell carcinomas examined, none showed evidence of c-erbB-3 protein expression. A single additional case of collecting duct
PRODUCTION AND CHARACTERISTICS OF POLYCLONAL ANTIBODY TO THE c-er6B-3 PROTEIN
279
membrane and cytoplasmic immunoreactivity in areas of squamous differentiation. The other two endometrioid adenocarcinomas were c-erbB-3 negative. DISCUSSION We have successfully produced a polyclonal antibody to the c-erbB-3 protein via the synthetic peptide route and have demonstrated expression of c-erbB-3 gene protein in a variety of human adenocarcinomas. The antibody works extremely well for immunohistochemistry on formalin-fixed, paraffinembedded tissue, much enhancing its potential utility in studies of c-erbB-3 protein expression in routine pathological specimens. It detects a single subunit of 165 kD in a c-erbB-3 transfected cell line. This is less than the predicted molecular weight of c-erbB-3 protein, but as expression vectors of EGFR probably show deficient or absent receptor protein glycosylation, this is not necessarily an unexpected finding.l3*l4 We have shown that the c-erbB-3 protein is also widely expressed in adenocarcinomas. The presence of cytoplasmic and membrane c-erbB-3 immunoreactivity in tumour cells supports the hypothesis that the c-erbB-3 protein is a membrane-bound receptor protein which is also present in the cell cytoplasm, and therefore shows similar immunohistochemical expression to EGFR and the c-erbB-2 pr~teins.l~'~J~ Our data are broadly in agreement with the results of preliminary immunoblotting studies of small numbers of breast2' and colonic adenocarcinomas*' which also have shown expression of the c-erbB-3 protein or c-erbB-3 mRNA in a proportion of the tumours examined. The presence of c-erbB-3 protein in the supranuclear Golgi zone of enterocytes and also in a more diffuse granular pattern in the cytoplasm of enterocytes is similar to the pattern of staining seen with EGFR in normal small intestine22 and in of EGFR and Barrett's o e s o p h a g ~ sLocalization .~~ c-erbB-3 proteins in the Golgi zone of enterocytes would be consistent with post-translational processing of the c-erbB-3 protein prior to incorporation into the cell membrane, as is known to be the case with EGFR.24 The role of c-erbB-3 in human neoplasia and the relationship of expression of c-erbB-3 protein to that of EGFR and c-erbB-2 remain to be discovered. The relatively high frequency of c-erbB-3 N
Fig. &The invasive edge of a colonic adenocarcinoma showing membrane (arrows) and cytoplasmic immunoreactivity (arrowheads)
carcinoma" showed strong cytoplasmic staining of tumour cells but no evidence of membrane immunoreactivity. The proximal and distal convoluted tubules showed strong granular c-erbB-3 immunoreactivity, but the cells of the glomeruli and the loop of Henle showed no evidence of c-erbB-3 protein expression. Carcinoma of the lung Four primary carcinomas of the lung were examined for c-erbB-3 protein expression, comprising three adenocarcinomas and one bronchioalveolar cell carcinoma. Heterogeneous membrane and cytoplasmic c-erbB-3 immunoreactivity was seen in one primary adenocarcinoma. Normal alveoli and bronchioles showed no evidence of c-erhB-3 expression. Carcinoma qf the endometrium Seven primary endometrial carcinomas were examined. Six were classified as pure endometrioid and one as an endometrioid carcinoma with squamous differentiation (adenoacanthorna).'* Membrane and cytoplasmic c-erbB-3 expression was seen in three endometrioid adenocarcinomas, with one endometrioid carcinoma showing granular cytoplasmic immunoreactivity alone. The single adenoacanthoma examined showed both
280
D. N. POLLER ET AL.
protein expression in this preliminary study of a range of adenocarcinomas suggests that expression of c-erbB-3 protein may prove to be of importance in neoplasia. Further larger studies to examine the extent of c-erbB-3 protein expression in both normal and a wider range of neoplastic tissues are now imperative.
ACKNOWLEDGEMENTS
We would like to thank Dr George Panayotou and Professor M. D. Waterfield for provision of EGFR and c-erbB-2 transfected cell lines, Dr Matthias Kraus for provision of the cell line MDA MB415, Dr Adrian Robins for providing MDA MB415 cell cultures, Mr Ken Morrell for technical assistance, Mr Bill Brackenbury for photography, and Dr Jim Lowe for critical advice.
REFERENCES 1 . Kraus MH, Issing W, Miki T, Popescu NC, Aaronson SA. Isolation
2. 3. 4.
5.
and characterization of ERBB3, a third member of the ERBB/ epidermal growth factor receptor family: evidence for overexpression in a subset ofhuman mammary tumors. Proc NatlAcudSci USA 1989; 86: 9193-9197. Plowman GD, Whitney GS, Neubauer MG, et ul. Molecular cloning and expression of an additional epidermal growth factor receptor related gene. Proc Nut1 Acad Sci USA 1990; 87:49054909. Yarden Y, Ullrich A. Growth factor receptor tyrosine kinases. Annu Rev Biochem 1988; 57: 443478. Sainsbury JRC, Farndon JR, Needham GK, Malcolm AJ, Harris AI. Epidermal growth factor receptor status as predictor of early recurrence of and death from breast cancer. Lancet 1987; ii: 139&1402. Grimaux M, Romain S, Remvikos Y, Martin PM, Magdelenat H. Prognostic value of epidermal growth factor receptor in node-positive breast cancer. Breast Cancer Res Treat 1989; 14 77-90.
6. Steele RJC, Kelly P, Ellul B, Eremin 0. Epidermal growth factor receptor expression in colorectal cancer. Br J Surg 1990; n: 1352-1 354. 7. Yano H, Shiozaki H, Kobayashi K, et al. Immunohistologic detection of the epidermal growth factor receptor in human esophageal squamous cell carcinoma. Cancer 1991;67: 91-98. 8. Di Marco E, Pierce JH, Aaronson SA, Di Fiore PP. Mechanisms by which EGF receptor and TGF-alpha contribute to malignant transformation. Nut Immun Cell Growth Regul1990;9 209-221. 9. Perren TJ. c-erbB-2 oncogene as a prognostic marker in breast cancer. Br JCancer 1991; 63: 328-332. 10. Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 1989; 244: 707-7 12. 11. Lovekin C, Ellis 10, Locker A, el a/. c-erbB-2oncoprotein expression in primary and advanced breast cancer. Br JCuncer 1991;6 3 439443. 12. Van de Vijver MJ, Peterse JL, Mooi WJ, et al. Neu protein overexpression in breast cancer: association with comedo-type ductal carcinoma in sim and limited prognostic value in stage 11 breast cancer. NEnglJMed1988; 319 1239-1245. 13. Greenfield C, Patel G, Clark S, Jones N, Waterfield MD. Expression of the human EGF receptor with ligand stimulatable kinase activity in insect cells using a baculovirus vector. EMBO J 1988;‘7: 139-146. 14. Wilson L. Ph.D. Thesis. Universityof London, 1990;217-246. 15. Lauren P. The two histological main types of gastric carcinoma: diffuse and so-called intestinal type carcinoma. Acta Pathol Microbiol Scand 1965; 64:3149. 16. Sasaki 0, Atkin WS, Jass JR. Mucinous carcinoma of the rectum. Histopathology 1987; 11: 259-272. 17. Fleming S, Lewi HJE. Collecting duct carcinoma of the kidney. Histopathology 1986; 10: 1131-1 141. 18. Clement PB. Pathology of the uterine corpus. Hum Pathol 1991; 2 2 776-791. 19. Gullick WJ, Hughes CM, Mellon K, Neal DE, Lemoine NR. Immunohistochemical detection of the epidermal growth factor receptor in paraffin-embedded human tissues. J Pathol 1991; 164: 28S289. 20. LeJeune S, Plowman G, Horak E, et al. Amphiregulin and c-erbB-3 expression in primary breast cancer: independent expression from EGF receptor and erhB-2. Proc Am Assoc Cancer Res 199; 3 2 51. 21. Ciardiello F, Kim N, Saeki T, et al. Differential expression of epidermal growth factor related proteins in human colorectal tumors. Proc Natl Acad Sci USA I99 I ;8% 7792-7796. 22. Lemoine NR, Jain S, Silvestre F, et a/. Amplification and overexpression of the EGF receptor and c-erbB-2 proto-oncogenes in human stomach cancer. B r J Cancer 1991;&1: 7943. 23. Poller DN, Steele RJC, Morrell K. Expression of epidermal growth factor receptor in Barrett’s esophagus. Arch Puthol Lah Med (in press). 24. Carpenter G. Receptors for epidermal growth factor and other polypeptide mitogens. Annu Rev Biochem 1987; 5 6 881-914.
168 275-280 (1992)
PRODUCTION A N D CHARACTERIZATION OF A POLYCLONAL ANTIBODY TO THE c-erbB-3 PROTEIN: EXAMINATION OF c-erbB-3 PROTEIN EXPRESSION IN ADENOCARCINOMAS D. N. POLLER*, I. SPENDLOVET, C. BAKER?, R. CHURCH*, I. 0. ELLIS$, G . D. PLOWMAN5 AND R. J. MAYER?
University Departments of Histopathology*and Biochemistry?, Queen's Medical Centre, Nottingham NG7 2UH, U.K.; $Department of Histopathology, City Hospital, Nottingham, NG5 IPB, U.K.;GBristol-Myers Squibb Pharmaceutical Research Institute, 3005 First Avenue, Seattle, Washington 98121, U.S.A. Received30 December 1991 Accepted 6 July 1992
SUMMARY A polyclonal rabbit antibody was raised to the c-erbB-3 protein using a synthetic peptide corresponding to amino acids 1229-1241 of the predicted protein sequence of c-erbB-3. In Western blot analysis this antibody detects a single band at 165 kD in a c-erbB-3 transfected (293/HER-3) human cell line. c-erbB-3 protein expression was then examined in a variety of adenocarcinomas. Expression of c-erbB-3 protein was indicated by membrane and/or cytoplasmic tumour cell immunoreactivity in formalin-fixed, paraffin-embedded tissue sections. c-erbB-3 protein was detected in a series of 13 out of 14 primary breast carcinomas, 3 of 5 gastric adenocarcinomas, 8 of 9 colonic adenocarcinomas, 2 of 9 prostatic adenocarcinomas, 0 of 6 renal cell carcinomas, 1 of 4 primary lung adenocarcinomas, and 5 of 7 endometrial adenocarcinomas. Immunohistochemical expression of the c-erbB-3 protein appears to be a relatively common event in adenocarcinomas, and further studies are now warranted to establish the role of the c-erbB-3 protein in neoplasia.
-
KEY wo~~s+-erbB-2,EGFR, c-erbB-3, adenocarcinoma,
INTRODUCTION The c-erbB-3 gene' also known as HER3,' is a recently identified gene that codes for a member of a family of growth factor receptors showing great similarity in structure to the epidermal growth factor receptor (EGFR) and the c-erbB-2 proteins."* The c-erbB-3 gene has been mapped to chromosome 12ql3.' The c-erbB-3 protein is a membrane-bound, receptor-like protein with presumed tyrosine kinase activity, which shows considerable structural homology to both EGFR and the c-erbB-2 protein^.^ The c-erbB-3 protein at its predicted tyrosine kinase domain shows 64 per cent amino acid sequence homology to the tyrosine kinase domain of EGFR Addressee for correspondence: Dr D. N. Poller, Department of Histopathology, City Hospital, Nottingham NG5 IPB, U.K.
0022-3417/92/110275-06 $08.00 0 1992 by John Wiley & Sons, Ltd.
immunohistochemistry.
and 67 per cent amino acid sequence homology to the tyrosine kinase domain of the c-erbB-2 protein.' Expression of EGFR has been shown to be of prognostic importance in predicting the biological ~.~ behaviour of adenocarcinomas of the b r e a ~ tand colon,6 and also in squamous cell c a r c i n ~ m a . ' ~ ~ Overexpression of the c-erbB-2 protein in human malignancies, a feature usually associated with c-erbB-2 gene amplifi~ation,~ has been shown to be an indicator of adverse pro nosis in breast9-" and ovarian adenocarcinomas," as well as being a feature of mammary ductal carcinoma in situ of comedo type." A polyclonal rabbit antibody to a synthetic peptide corresponding to amino acids 1229-1241 of the predicted amino acid sequence of the c-erbB-3 protein has been produced.'** The expression of c-erbB-3 protein was then examined in a variety of adenocarcinomas.
276
D. N. POLLER ETAL.
presence of sodium dodecyl sulphate, the proteins were transferred to Hybond-C Super (Amersham International PLC, Slough, Berks, U.K.). A Preparation of antibody A peptide corresponding to amino acids 1229-1241 standard western development was followed. The of the predicted sequence of the c-erbB-3 protein transfer was blocked for 1 h in Tris-buffered saline (sequence residues IMPTAGTTPDED) was syn- (200mM NaCl, 5 0 m ~Tris-HC1, pH 7.4), 0.1 thesized on a model 431A peptide synthesizer per cent Tween 20, 5 per cent Marvel, and then (Applied Biosystems, Warrington, U.K.) Using incubated for 2 h with the c-erbB-3 protein primary Fmoc chemistry. The peptide was synthesized with antibody in Tris-buffered saline (TBS) at 1:200 an additional C-terminal cysteine for linkage to dilution, 0.1 per cent Tween 20, 2 per cent Marvel keyhole limpet haemocyanin, using the heterobi- (TBSTM). Following a 1 h wash in TBST, the blot functional reagent m-maleimidobenzoyl-N-hydroxy- was incubated with a ljlO000 dilution of horsesuccinimide ester, and then injected intramuscularly radish peroxidase conjugated swine anti-rabbit and subcutaneously into two rabbits. The resulting secondary antibody (Dako, U.K. Ltd., High polyclonal antiserum was then used without affinity Wycombe, Bucks, U.K.) in TBSTM. Following a purification. High ELISA titres were obtained final20minwashinTBS, theblotwasdevelopedwith Amersham ECL (enhanced chemiluminescence) (1 :500 000) against the synthetic peptide. detection reagent. Duplicate samples were also incubated in the absence of primary antibody Western blotting c-erbB-3 antibody as negative controls. The anti-c-erbB-3 protein antibody was characterized using EGFR, c-erbB-2 and c-erbB-3 transfected Immunohistochemistry cell lines. The EGFR tran~fectant'~ and the c-erbB-2I4 Four pm thick sections of formalin-fixed, paraffintransfectants were cell lines derived from the Fall embedded tumour tissue were cut, dewaxed in Army worm Spodopterafrugiperda (Sf 9), which can be infected by the naturally occurring baculovirus xylene, and then rinsed in alcohol and graded Autographa calfornica nuclear polyhedrosis virus alcohol-water mixtures. A standard avidin-biotin (AcNPV), a non-mammalian expression vector that peroxidase technique was used. Incubation with the uses the strong AcNPV polyhedrin promoter to c-erbB-3 primary antibody was performed for 30 drive production of large amounts of a foreign, min at a dilution of 1:800. Following incubation eukaryotic protein-in this case, EGFR13 or with the primary antibody, a secondary polyclonal c-erbB-2 protein.I4 The Sf 9 cells were cultured as a biotinylated swine anti-rabbit antibody was used monolayer at 27°C in IPL4-I media supplemented (Dako U.K. Ltd.). This was followed by an avidin with 10 per cent FCS, 1 per cent Fungizone, and 0.1 biotinylated peroxidase complex, with 3,3-diaminoper cent gentamycin. A proximately 4 x lo6 cells benzidene as a chromagen. Each case was also were plated into a 25 cm flask and then the EGFR incubated in the absence of the anti-c-erbB-3 or c-erbB-2 containing AcNPV was then added to primary antibody as a negative control. Absorption the flask. The baculovirus concentrations had pre- specificity controls using the preimmune peptide viously been titred to give optimal expressed protein were also performed on some of the positively levels. A non-transfected negative control Sf 9 cell staining tumours, with inhibition of c-erbB-3 line was also used. The cells were harvested 48 h staining. The sections were examined by light after infection, centrifuged, washed in PBS, and microscopy. Formalin-fixed cytospins of a human lysed with Laemmeli sample buffer (A). The c-erbB-3 breast carcinoma cell line MDA MB415, known to transfectant was a human embryonic cell line, 293 express high levels of c-erbB-3 mRNA,' were used as (American Type Culture Collection), which after a positive control for immunohistochemistry. stable c-erbB-3 transfection expresses abundant amounts of recombinant c-erbB-3 protein (personal RESULTS communication, Dr G. D. Plowman). Approximately 100,ug samples of protein were loaded onto a 7 per cent polyacrylamide gel along Western blotting with prestained molecular weight markers (Sigma The results of Western blots using the anti-cChemical Company Ltd., Fancy Road, Poole, erbB-3 antibody are shown in Fig. 1. In Fig. 1, Dorset, U.K.). Following electrophoresis in the lane 1 shows a single discrete band at 165 kD, MATERIALS AND METHODS
P
N
PRODUCTION AND CHARACTERISTICS OF POLYCLONAL ANTIBODY TO THE c-erbB-3 PROTEIN
corresponding to the subunit of c-erbB-3, with the c-erbB-3 transfected 293 cell line. Lane 2 is the control non-transfected 293 cell line which shows absence of the c-erbB-3 protein subunit.
277
Immunohistochemistry The antibody appeared to work well on formalinfixed, paraffin-embedded tissue, with little or no background staining. Smooth muscle showed weak granular cytoplasmic staining, present in the breast, muscularis mucosae and propria of the stomach, small intestine, colon, and myometrium, and within the walls of arteries and muscular veins. The results of the immunohistochemical examination of c-erbB-3 protein expression in adenocarcinomas are shown in Table I. Breast carcinoma The 14 primary breast carcinomas examined comprised 13 invasive ductal carcinomas of no special type, and one invasive lobular carcinoma. Thirteen tumours showed membrane or cytoplasmic c-erbB-3 immunoreactivity and of these, four tumours showed some focal membrane and cytoplasmic immunoreactivity of tumour cells. Nine tumours showed only cytoplasmic staining of tumour cells. Cytoplasmic c-erbB-3 immunoreactivity was also seen in the in situ components of invasive ductal carcinomas, where there was also expression of c-erbB-3 in the invasive component (Figs 2 and 3). Normal mammary duct also showed cytoplasmic staining of epithelial cells with some staining of myoepithelial cells, and areas of apocrine metaplasia within foci of fibrocystic disease. Normal breast acini did not show c-erbB-3 protein expression in the limited number of cases examined.
Gastric adenocarcinoma Five primary gastric carcinomas were examined comprising four of intestinal type and one of diffuse type.I5 Three intestinal type adenocarcinomas showed heterogeneous membrane and cytoplasmic c-erbB-3 immunoreactivity, and one was c-erbB-3 negative, with no evidence of c-erbB-3 expression in the single case of diffuse type gastric carcinoma Fig. I-Lane 1: A single discrete band at 165 kD correspond- examined. Expression of the c-erbB-3 protein was ing to the subunit of c-erbB-3 is seen with the c-erbB-3 transfected seen in the cytoplasm of normal foveolar epithelial 293 cell line. Lane 2: The control (non-transfected) 293 cell line cells of the gastric crypts, predominantly within the showing absence of the c-erbB-3 protein subunit supranuclear Golgi zone. The parietal cells lining the gastric glands also showed strong granular Western blots performed with wild type (non- c-erbB-3 positivity, although there was no staining baculovirus transfected) Sf 9 cells, EGFR transfected of chief cells or mucus neck cells. Sf 9 cells, and the c-erbB-2 transfected Sf 9 cell lines showed absence of cross-reactivity with the Carcinoma of the colon All nine cases examined were primary invasive anti-c-erbB-3 antibody, no band being detected on adenocarcinomas of large intestinal type, one of Western blotting (not illustrated).
-
D. N. POLLER ETAL.
278
Table I.--erbB-3 protein expression and tumour primary site as assessed in formalin-fixed,paraffin-embeddedtissue
Primary site Breast Stomach Colon Prostate Kidney Lung Endometrium
Fig. 2-An
Cytoplasmic
Pattern of immunoreactivity Membrane and No staining Total cytoplasmic
9 0 2 2 0 0 1
4 3 6 0 1 1 4
1 2 1
7 6 3 2
14 5 9 9 7 4 7
invasive ductal carcinoma of the breast showing
strong cytoplasmictumour cell c-erbB-3immunoreactivityin the in situ component (arrow-heads)
Fig. 3-An invasive ductal carcinoma of the breast showing strong cytoplasmic c-erbB-3immunoreactivity
which was classified as a mucinous (colloid) carcinoma.16 Eight out of the nine cases showed expression of the c-erbB-3 protein. Six tumours showed membrane and cytoplasmic immunoreactivity, and two cases showed only cytoplasmic c-erbB-3 staining, including the single case of mucinous carcinoma included in the series. Membrane expression of c-erbB-3 was more intense at invasive tumour margins (Fig. 4). Granular cytoplasmic staining in surrounding normal absorptive and goblet cells of the large intestinal mucosa was also identified.
The pattern of staining was cytoplasmic and granular without evidence of membrane staining. Normal prostatic glandular acini showed no evidence of c-erbB-3 expression, although granular c-erbB-3 staining was present in normal prostatic ducts in both epithelial and myoepithelial cells. Normal transitional mucosa from the bladder neck also showed membrane and cytoplasmic c-erhB-3 immunoreactivity .
Adenocarcinoma of the prostate c-erbB-3 protein expression was present in two of nine invasive prostatic adenocarcinomas examined.
Renal tumours Of the six primary renalcell carcinomas examined, none showed evidence of c-erbB-3 protein expression. A single additional case of collecting duct
PRODUCTION AND CHARACTERISTICS OF POLYCLONAL ANTIBODY TO THE c-er6B-3 PROTEIN
279
membrane and cytoplasmic immunoreactivity in areas of squamous differentiation. The other two endometrioid adenocarcinomas were c-erbB-3 negative. DISCUSSION We have successfully produced a polyclonal antibody to the c-erbB-3 protein via the synthetic peptide route and have demonstrated expression of c-erbB-3 gene protein in a variety of human adenocarcinomas. The antibody works extremely well for immunohistochemistry on formalin-fixed, paraffinembedded tissue, much enhancing its potential utility in studies of c-erbB-3 protein expression in routine pathological specimens. It detects a single subunit of 165 kD in a c-erbB-3 transfected cell line. This is less than the predicted molecular weight of c-erbB-3 protein, but as expression vectors of EGFR probably show deficient or absent receptor protein glycosylation, this is not necessarily an unexpected finding.l3*l4 We have shown that the c-erbB-3 protein is also widely expressed in adenocarcinomas. The presence of cytoplasmic and membrane c-erbB-3 immunoreactivity in tumour cells supports the hypothesis that the c-erbB-3 protein is a membrane-bound receptor protein which is also present in the cell cytoplasm, and therefore shows similar immunohistochemical expression to EGFR and the c-erbB-2 pr~teins.l~'~J~ Our data are broadly in agreement with the results of preliminary immunoblotting studies of small numbers of breast2' and colonic adenocarcinomas*' which also have shown expression of the c-erbB-3 protein or c-erbB-3 mRNA in a proportion of the tumours examined. The presence of c-erbB-3 protein in the supranuclear Golgi zone of enterocytes and also in a more diffuse granular pattern in the cytoplasm of enterocytes is similar to the pattern of staining seen with EGFR in normal small intestine22 and in of EGFR and Barrett's o e s o p h a g ~ sLocalization .~~ c-erbB-3 proteins in the Golgi zone of enterocytes would be consistent with post-translational processing of the c-erbB-3 protein prior to incorporation into the cell membrane, as is known to be the case with EGFR.24 The role of c-erbB-3 in human neoplasia and the relationship of expression of c-erbB-3 protein to that of EGFR and c-erbB-2 remain to be discovered. The relatively high frequency of c-erbB-3 N
Fig. &The invasive edge of a colonic adenocarcinoma showing membrane (arrows) and cytoplasmic immunoreactivity (arrowheads)
carcinoma" showed strong cytoplasmic staining of tumour cells but no evidence of membrane immunoreactivity. The proximal and distal convoluted tubules showed strong granular c-erbB-3 immunoreactivity, but the cells of the glomeruli and the loop of Henle showed no evidence of c-erbB-3 protein expression. Carcinoma of the lung Four primary carcinomas of the lung were examined for c-erbB-3 protein expression, comprising three adenocarcinomas and one bronchioalveolar cell carcinoma. Heterogeneous membrane and cytoplasmic c-erbB-3 immunoreactivity was seen in one primary adenocarcinoma. Normal alveoli and bronchioles showed no evidence of c-erhB-3 expression. Carcinoma qf the endometrium Seven primary endometrial carcinomas were examined. Six were classified as pure endometrioid and one as an endometrioid carcinoma with squamous differentiation (adenoacanthorna).'* Membrane and cytoplasmic c-erbB-3 expression was seen in three endometrioid adenocarcinomas, with one endometrioid carcinoma showing granular cytoplasmic immunoreactivity alone. The single adenoacanthoma examined showed both
280
D. N. POLLER ET AL.
protein expression in this preliminary study of a range of adenocarcinomas suggests that expression of c-erbB-3 protein may prove to be of importance in neoplasia. Further larger studies to examine the extent of c-erbB-3 protein expression in both normal and a wider range of neoplastic tissues are now imperative.
ACKNOWLEDGEMENTS
We would like to thank Dr George Panayotou and Professor M. D. Waterfield for provision of EGFR and c-erbB-2 transfected cell lines, Dr Matthias Kraus for provision of the cell line MDA MB415, Dr Adrian Robins for providing MDA MB415 cell cultures, Mr Ken Morrell for technical assistance, Mr Bill Brackenbury for photography, and Dr Jim Lowe for critical advice.
REFERENCES 1 . Kraus MH, Issing W, Miki T, Popescu NC, Aaronson SA. Isolation
2. 3. 4.
5.
and characterization of ERBB3, a third member of the ERBB/ epidermal growth factor receptor family: evidence for overexpression in a subset ofhuman mammary tumors. Proc NatlAcudSci USA 1989; 86: 9193-9197. Plowman GD, Whitney GS, Neubauer MG, et ul. Molecular cloning and expression of an additional epidermal growth factor receptor related gene. Proc Nut1 Acad Sci USA 1990; 87:49054909. Yarden Y, Ullrich A. Growth factor receptor tyrosine kinases. Annu Rev Biochem 1988; 57: 443478. Sainsbury JRC, Farndon JR, Needham GK, Malcolm AJ, Harris AI. Epidermal growth factor receptor status as predictor of early recurrence of and death from breast cancer. Lancet 1987; ii: 139&1402. Grimaux M, Romain S, Remvikos Y, Martin PM, Magdelenat H. Prognostic value of epidermal growth factor receptor in node-positive breast cancer. Breast Cancer Res Treat 1989; 14 77-90.
6. Steele RJC, Kelly P, Ellul B, Eremin 0. Epidermal growth factor receptor expression in colorectal cancer. Br J Surg 1990; n: 1352-1 354. 7. Yano H, Shiozaki H, Kobayashi K, et al. Immunohistologic detection of the epidermal growth factor receptor in human esophageal squamous cell carcinoma. Cancer 1991;67: 91-98. 8. Di Marco E, Pierce JH, Aaronson SA, Di Fiore PP. Mechanisms by which EGF receptor and TGF-alpha contribute to malignant transformation. Nut Immun Cell Growth Regul1990;9 209-221. 9. Perren TJ. c-erbB-2 oncogene as a prognostic marker in breast cancer. Br JCancer 1991; 63: 328-332. 10. Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 1989; 244: 707-7 12. 11. Lovekin C, Ellis 10, Locker A, el a/. c-erbB-2oncoprotein expression in primary and advanced breast cancer. Br JCuncer 1991;6 3 439443. 12. Van de Vijver MJ, Peterse JL, Mooi WJ, et al. Neu protein overexpression in breast cancer: association with comedo-type ductal carcinoma in sim and limited prognostic value in stage 11 breast cancer. NEnglJMed1988; 319 1239-1245. 13. Greenfield C, Patel G, Clark S, Jones N, Waterfield MD. Expression of the human EGF receptor with ligand stimulatable kinase activity in insect cells using a baculovirus vector. EMBO J 1988;‘7: 139-146. 14. Wilson L. Ph.D. Thesis. Universityof London, 1990;217-246. 15. Lauren P. The two histological main types of gastric carcinoma: diffuse and so-called intestinal type carcinoma. Acta Pathol Microbiol Scand 1965; 64:3149. 16. Sasaki 0, Atkin WS, Jass JR. Mucinous carcinoma of the rectum. Histopathology 1987; 11: 259-272. 17. Fleming S, Lewi HJE. Collecting duct carcinoma of the kidney. Histopathology 1986; 10: 1131-1 141. 18. Clement PB. Pathology of the uterine corpus. Hum Pathol 1991; 2 2 776-791. 19. Gullick WJ, Hughes CM, Mellon K, Neal DE, Lemoine NR. Immunohistochemical detection of the epidermal growth factor receptor in paraffin-embedded human tissues. J Pathol 1991; 164: 28S289. 20. LeJeune S, Plowman G, Horak E, et al. Amphiregulin and c-erbB-3 expression in primary breast cancer: independent expression from EGF receptor and erhB-2. Proc Am Assoc Cancer Res 199; 3 2 51. 21. Ciardiello F, Kim N, Saeki T, et al. Differential expression of epidermal growth factor related proteins in human colorectal tumors. Proc Natl Acad Sci USA I99 I ;8% 7792-7796. 22. Lemoine NR, Jain S, Silvestre F, et a/. Amplification and overexpression of the EGF receptor and c-erbB-2 proto-oncogenes in human stomach cancer. B r J Cancer 1991;&1: 7943. 23. Poller DN, Steele RJC, Morrell K. Expression of epidermal growth factor receptor in Barrett’s esophagus. Arch Puthol Lah Med (in press). 24. Carpenter G. Receptors for epidermal growth factor and other polypeptide mitogens. Annu Rev Biochem 1987; 5 6 881-914.
