c-erbB-2 oncogene expression in hepatocellular carcinoma and cholangiocarcinoma

Journal of Hepatology, 1992; 14: 377-380 0 1992 Elsevier Science Publishers B.V. All rights reserved.

377

0168~8278/92/$05.00

HEPAT 01088

c-er

JD. Collier’, K. 6u01, J. Mathe+,

r easci

an

F.E.B. Ma?, M.K. Bennett3, I.P. Corbett2, M.F. Bassendinel and A.D. Burt2

Departments of ‘Medicine and ‘Pathology, Universityof Newcastle upon Tyne and 3Departrnent of Pathology, Freeman Hospital, Newcastle upon Tyne, United Kingdom (Received

12 August 1991)

The c-erlrB-2 proto-oncogene encodes a transmembrane protein which is homologous to the epidermal growth factor receptor. This protein can be localized immunohistochemically in formalin-fixed paraffin-embedded material using a monoclonal antibody NCL-CBll; positive membrane staining correlates with gene amplification and protein overexpression in breast cancer. Using this technique we have shown that only 2126 (8%) of hepatocellular carcinomas, 0110 (0%) of cholangiocarcinomas and O/2 (0%) hepatoblastomas overexpressed c-erbB-2 as evidenced by membrane staining. Moreover c-&B-2 mRNA was not detected in seven hepatocellular carcinomas examined by Northern blot analysis. c-e&B-2 overexpression is, therefore, unlikely to be contributing to the malignant phenotype in hepatocellular carcinoma and cholangiocarcinoma.

Proto-oncogenes encode proteins which have diverse normal cellular functions related to differentiation and proliferation; abnormal or altered expression of these genes is thought to be involved in both the initiation and/or maintenance of the malignant phenotype. The c-e&B-2 proto-oncogene, located a9 chromosome 17, encodes a 190 kDa protein homologous to the epidermal growth factor receptor (1). It consists of an external domain linked to an internal domain by a transmembrane region. The ligand for this putative membrane receptor has not yet been identified. Gene amplification and overexpression of the oncogene has been reported in many adenocarcinomas, including breast carcinoma, where it is associated with a poor prognosis (2). Moreover expression of actived c-neu, (rat homologue of c-e&B-2) in transgenic mice induces malignant transformation of mammary epithelium suggesting a direct role in tumorigenesis (3). It has recently been reported that 86% of hepatscellular carcinomas (HCCs) and 73% of cholangiocarcinomas overexpress the c-erbI3-2 protein as demonstrated by immunohistochemistry using a polyclonal antibody (4).

This is much higher than that found in uther tumours where the highest reported incidence of 20-30% is in breast carcinoma (5). If such a high percentage of HCCs overexpress the c-e&B-2 protein it suggests a role for this protein in the initiation and/or maintenance of the malignant phenotype in this tumour, and, therefore, an important future research field. The inhibition in vivo of tumour growth in rats by monoclonal antibodies raised against the external domain of neu (6) illustrates the potential for immunotherapy targetted to the c-e&B-2 protein. To substantiate the previous findings in hepatocellular carcinoma and cholangiocarcinoma, we have studied expression of the c-erbB-2 gene, detecting the protein immunohistochemically with a highly specific monoclonal antibody and the mRNA by Northern blot analysis.

Case material

Archival

Correspondence: Dr. A.D. Burt, Division of Pathology, School of Pathological Sciences, 41-P. U.K.

formalin-fixed

paraffin-embedded

liver bi-

Royal Victoria Infirmary, Newcastleupon Tyne, ~1

J.D. COLLIER et al.

378 a

TABLE 1 Patient details Hepatocellular carcinoma

Cholangiocarcinoma

Hepatoblastoma

Number

26

10

2

Age range

54-13 years

38-76 years

l-7 months

Sex M F

19 7

5 5

1 1

HBsAg positive HBeAg positive

4 0

0 0

0 0

Alpha-fetoprotein > 500 IV/l (normal < 10)

23

0

0

from 26 cases of HCC, ten cholangiocar-

cinemas

two hepatoblastomas,

were

used

1

ehiik

la4-

opsy material and

beds

in this

FEg. 1. Northern blot analysis. Total cellular RNA extracted from breast cancer cell-line SK-BR3 (lane a, 20 gg; lane b, 2 pg), normal human liver (lane c, 20 pg), HCC-derived cell-line Hep3B (lane d, 20 pg) and HCC samples (lanes e-k, 20 ,ug) was hybridized to a cDNA erbB-2 probe. Numbers on the left indicate marker sizes in kilobases.

study. In seven cases of HCC, tumour samples had been

snap frozen at the time of operative resection and were available for nucleic acid extraction. Clinical information was obtained on all patients from the medical records (Table 1). All 38 patients were European Caucasian with the exception of one patient with HCC who was Japanese. Correct classification of the ten cases of cholangiocarcinoma was ensured by simultaneous review of both the clinical details and pathological data. Northern blot analysis Total RNA was extracted from the seven snap-frozen HCC samples, a human liver cancer cell line (Hep3B), and normal human liver (obtained from a kidney donor whose liver was ‘unsuitable for organ transplantation’). The samples were homogenized in guanidine thiocyanate and pelleted through a 5.7 M caesium chloride gradient (8). RNA samples (20 yg per track) were electrophoresed in a 1.2% agarose gel containing 0.25 M formaldehyde and then transferred to a nylon membrane (Hybond N+, Amersham, U.K.) by capillary action using 20 x SSC (1 x SSC is 150 mM sodium chloride, 15 mM sodium citrate, pH 7.0). After transfer, the RNA was covalently crosslinked by ultraviolet (UV) irradiation. Twenty- and 2qg samples of RNA from a human breast carcinoma cell line (SK-BR3) known to overexpress c-erbB-2 mRNA (7) were included as positive controls. The blot was prehybridized for 2 h at 42°C in 50% formamide, 10 X Denhardt’s solution (1 x Denhardt’s = 0.02% bovine serum albumin, 0.02% Ficoll, 0.02% polyvinylpyrrolidine), 6 X SSPE (1 x SSPE = 150 mM sodium chloride, 10 mM sodium dihydrogen orthophosphate, 1 mM

EDTA, pH 7.4), 5% dextran sulphate, 0.5% sodium dodecyl sulphate (SDS), and 100 pg salmon sperm DNA. It was then hybridized with a 32P-labelled c-erbB-2 probe (EcoRl fragment of the c-erbB-2 proto-oncogene spanning the central portion of the cDNA; a gift from T. Yamamoto, Tokyo), synthesized’ with a random prime DNA labelling system (Boehringer Mannheim) in the above buffer at 42°C for 16 h. After hybridization the blot was washed in 2 x SSPE, 0.1% SDS buffer at 42°C in two changes (10 min each), then 1 x SSPE, 0.1% SDS at 68°C for 30 min and 0.1 x SSPE, 0.1% SDS at 68°C for 10 min. The blot was then exposed to film at -70°C for 3 days. The size of the RNA transcript was confirmed by comparison with a denatured 0.24-9.5 kb RNA ladder (BRL) stained with ethidium bromide. Immunochemistry Three-pm sections were cut from the iissue blocks. Endogenous peroxidase activity was blocked with 0.5% hydrogen peroxide in methanol for 10 min. The primary monoclonal antibody NCL-CBll was raised against a synthetic peptide corresponding to the C-terminus of the internal domain of the c-erbB-2 protein (5). Sections were incubated overnight at 4°C with a 1:40 dilution of NCL-CBll. Immunostaining was performed using an indirect immunoperoxidase method: 3,3’-diaminobenzidine (Sigma) was used as the chromagen. A positive control (a case of breast carcinoma known to overexpress c-erbB-2) was included in all experiments.

c-crbB-2 IN MALIGNANT

LIVER

TUMOURS

379

Fig. 2. (a) Membrane staining with NCL-CBll in positive control section of breast carcinoma. (b) Cytoplasmic immunoreactivity NCL-CBll in moderately well differentiated hepatocellular carcinoma. (Indirect immunoperoxidase; haematoxylin counterstain).

esullts Expression

iscussion of c-erbB-2 &WA

The c-erbB-2 probe hybridized strongly to a transcript

of approx. 4.4 kb in RNA extracted from the breast caucer cell-line SK-BR3 (Fig. 1). In contrast c-e&B-2 transcripts were undetectable in all seven HCCs. However, after prolonged exposure of the autoradiograph very low levels of the c-e&B-2 transcript were detected in RNA extracted from the HCC-derived cell-line Hep3B. Expression

with

of c-erbB-2 protein

The presence of cell membrane labelling was considered an essential criterion for identification of c-e&B-2 overexpression; this pattern was demonstrated by the positive breast control (Fig. 2a). Only two out of 26 HCCs had membrane staining and in both cases the immunoreactivity was focal. One of these was a clear cell variant of HCC; the intensity of membrane staining was much weaker than in the breast control. Weak cytoplasmic staining was seen in 5126 HCCs (Fib. 2b). Membrane immunoreactivity for c-e&B-2 was absent in both hepatoblastomas and all ten cholangiocarcinomas. However, weak cytoplasmic staining was seen in 1110cholangiocarcinemas and l/2 hepatoblastomas.

The c-e&B-2 oncoprotein is a transmembrane receptor and positive immunohistochemical staining in the form of membrane immunoreactivity has been shown to correlate with overexpression of the protein and gene amplification in breast cancer. In this study only two out of 26 HCCs and none of ten cholangiocarcinomas or two hepatoblastomas showed such positive membrane staining using a recently developed monoclonal antibody to the c-e&B-2 protein NCL-CBll. These results contrast with those of Voravud et al. who found that 40/63 (73%) cholangiocarcinomas and 12/14 (86%) hepatocellular carcinomas stained positively with a polyclonal antibody, 21N to the c-e&B-2 oncoprotein (4). However, they did not distinguish between cytoplasmic and membrane staining when reporting their findings. This is important as the significance of cytoplasmic staining remains one of conjecture; cytoplasmic staining in the absence of membrane staining does not correlate with overexpression of c-erbB-2.

Cytoplasmic staining has been reported with both the polyclonal antibody 21N and the monoclonal antibody NCL-CBll. We detected cytoplasmic staining in 5/26 (19%) of hepatocellular carcinomas, 1110 (10%) of cholangiocarcinomas and l/2 hepatoblastomas. Possible ex-

J.D. COLLIER et al.

planations for cytoplasmic staining include crossreactivity with an unknown protein, as both NCL-CBll and 21N react j,vith a 150 kDa band on Western blotting (5,9), abnormal processing of the RNA or protein, or internalization of the c-&B-2 receptor. Internalization of the epidermal growth factor receptor (EGFR), which has homology to the c-e&B-,2 protein, has been demonstrated in epidermoid carcinoma (A431) cells (10). In that study the EGFR was demonstrated on both the cell membrane and endosomal compartments in a transfer pathway of EGF/EGFR complex to the lysosomes. c-erbB-2 has also been demonstrated immunohistochemically, in endoplasmic reticulum using electron microscopy (10). In this study c-erbB-2 RNA was not detected in two HCCs which showed cytoplasmic immunoreactivity confirming that overexpression of c-erbB-2 oncogene, comparable to that found in breast adenocarcinoma was not occurring. It is unlikely that overexpression of c-erbB-2 was occurring in our series, while not being detected immunohistochemically, as RNA was not detected in the other References 1 Yamamoto T, Ikawa S. Akiyama T, et al. Similarity of protein encoded by the human c-&B-2 gene to epidermal growth factor receptor. Nature 1986; 319: 230-4. 2 Lovekin C, Ellis IO, Locker A, et al. c-er&B-2 oncogene expression in breast cancer: relationships and prognostic signihcance. J Path01 1989; 49: 2087-91. 3 Muller WJ, Sinn E, Pattengate PK. Wallace R, Leder P. Single step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 1988; 54: 105-15. 4 Voravud N, Foster FC, Gilbertson JA, Silkora K, Waxman J. Oncogene expression in cholangiocarcinoma and in normal hepatic development. Hum Path01 1990; 20: 1163-8. 5 Corbett IP, Henry JA, Angus B. NCL-CBll, a new monoclonal antibody recognising the internal domain of the c-erbB-2 oncogene protein effective for use on formalin fixed paraffin embedded tissue. J Path01 1990; 161: 15-25.

five HCCs which were negative immunohistochemically for c-erbB-2 protein. We did not look for DNA amplification as has been done in breast carcinoma as, in the absence of overexpression of RNA, DNA amplification is unlikely to be of any biological significance. Our results indicate that overexpression of the c-e&B-2 proto-oncogene is uncommon in hepatocellular carcinoma and cholangiocarcinoma and is unlikely, therefore, to be important in hepatocarcinogenesis. Determining the presence of the protein product cannot distinguish between neoplastic and non-neoplastic tissue or be of value in determining prognosis. Moreover, immunotherapy using the c-e&B-2 transmembrane receptor cannot be considered as potential therapy in these tumours. Acknowledgements Collier was supported by Newcastle Health Authority Research Committee. K. Guo was supported by the Wellcome Trust. F.E.B. May thanks the Royal Society for a University Research Fellowship. J.D.

6 Drebin JA, Link VC, Greene MI. Monoclonal antibodies specific for the neu oncogene product directly mediate anti-tumour effects in vivo. Oncogene 1988; 2: 387-94. 7 Kraus MH, Popescu NC, Amsbaugh C, King CR. Overexpression of the EGF receptor-related proteo-oncogene erbB-2 in human mammary tumour cell lines by different molecular mechanisms. EMBO J 1987; 6: 605-10. 8 Chirgwin JM, Przybla AE, MacDonald RJ, Rutter WJ. Isolation of biologically active ribonucleic acid from source enriched in ribonuclease. Biochemistry 1979; 18: 5294-9. 9 Gullick WJ, Berger MS, Bennett PLP, Rothbard JB, Watertield MD. Expression of c-erbB-2 protein in normal and transformed cells. J Cancer 1987; 40: 246-54. 10 Miller K, Beardmore J, Kanety H, Schlessinger J, Hopkins CR. Localisation of the epidermal growth factor (EGF) receptor within the endosome of EGF-stimulated epidermoid carcinoma (A431) cells. J Cell Biol 1986; 102: 500-9.