immunoreactivity for hepatocyte growth factor/scatter factor and its receptor, met, in human lung...

JOURNAL OF PATHOLOGY, VOL,. 180 389-394 ( 1 996)

IMMUNOREACTIVITY FOR HEPATOCYTE GROWTH FACTORKCATTER FACTOR AND ITS RECEPTOR, met, IN HUMAN LUNG CARCINOMAS AND MALIGNANT

MESOTHELIOMAS

P. HARVEY", A. WARN", P. NEWMAN", L. J . PERRY?, R. Y. BALL? AND R. M. WARN"

*School of' Biology, University of' East Anglia, Norwich NR4 7TJ, U. K.; ?Department of' Histoputholog~.lCytopatholog~., Norfolk and Nor~.~'icIi Health Care N H S Trust, Norwich NRl 3SR, U. K.

SUMMARY

Paraffin sections from 29 lung carcinomas (28 primary and 1 metastatic) and 9 pleural malignant mesotheliomas were immunostained with antisera to human hepatocyte growth factorlscatter factor (HGFISF) and its receptor, met. For HGFISF, immunoreactivity was demonstrated in all 9 mesotheliomas, 9 of 12 adenocarcinomas, and 7 of 10 squamous cell carcinomas. None of seven cases of small cell anaplastic carcinoma was positive. The adenocarcinomas frequently showed enhanced luminal staining, suggesting possible secretion of HGFISF, and this pattern of staining was also seen occasionally in bronchial epithelium adjacent to the tumour. Stromal fibroblasts also showed immunoreactivity for HGFISF in 6/8 cases of mesothelioma but in only 3/12 adenocarcinomas, 1/10 squamous cell carcinomas, and 114 small cell anaplastic carcinomas. All tumours stained for met, usually strongly. The staining was mainly cytoplasmic in nature, but some plasma membrane staining was usually evident. Adenocarcinomas showed strong luminal membrane staining, as did adjacent, histologically normal bronchial epithelium. This study demonstrates the presence of HGFISF and met in most of the tumour types described, particularly rnesotheliomas, and suggests that the HGFfSFlmet signalling system may play a role in the development of these tumours, either by autocrine or by paracrine mechanisms.

KEY WORDS-HGFISF; met; lung carcinoma; mesothelioma; immunohistochemistry

INTRODUCTION

Hepatocyte growth factor/scatter factor (HGF/SF) is a multifunctional growth factor with major roles in embryogenesis, body tissue maintenance, and wound healing.'.2 These effects are mediated via met, a trans- membrane tyrosine kinase receptor encoded by the c-met proto-~ncogene.'.~ HGF/SF is a potent mitogen for hepa to~y tes ,~ .~ endothelial cells, and a great variety of epithelia1 cell types.' It also has the property of dissociating-scattering-epithelial cell colonies in cul- ture, leading to the appearance of single highly motile cells.' HGF/SF also induces or promotes invasion of scattered cells into collagen gel matrices.* Furthermore, HGF/SF has been shown to be a potent angiogenic factor, both in vitro and in vivo.'.'O

A number of reports have linked HGF/SF and met to tumour growth and metastasis. Various tumour cell lines are scattered in vitro and show enhanced mitosis in response to HGF/SF.'.* There is evidence that HGF/SF is expressed in human malignant breast epithelium;'' the level present is a strong indicator of recurrence of carcinoma and predictor of survival.12 I t is also expressed in human bladder' and p a n ~ r e a t i c ' ~ tuinours. Furthermore, many tumour types have been found to contain amplified met levels compared with normal tissues.I5 The highest levels, over 100 times higher than in normal thyroid tissue,lh were found in human thyroid cancers.

Addressee for correspondence: Dr R. M. Warn, School of Biology, University of East Anglia, Norwich N R 4 7TJ, U.K. [email protected]

$'j 1996 by John Wiley & Sons, Ltd. CCC 0022-34 1719611 20389-06

There is evidence linking HGF/SF to lung carcinomas. Over 90 per cent of a large sample of pleural effusions obtained from cancer patients, including patients with primary lung cancers and mesotheliomas, contained significant amounts of biologically active HGF/SF.I9J0 This suggests that HGF/SF is secreted within the lung or pleura in association with tumour growth and spread, but the origin of the HGF/SF within the lungs of such patients has yet to be determined. This paper reports an immunohistochemical study of bronchial carcinomas and mesotheliomas to determine the presence and distribution of HGFlSF and met.

MATERIALS AND METHODS

Clinical details Thirty-eight patients were studied (Table I). Sections

were cut from archival blocks of tissue obtained at diagnostic biopsy or definitive lung resection. Apart from one trucut biopsy of a metastatic deposit of small cell anaplastic lung cancer, the tumours were primary within the lung or pleura.

Antibodies

Affinity purified polyclonal antibodies against human HGFiSF and met (antibody C-28) were obtained from R & D Systems Europe (Abingdon, U.K.) and Santa- Cruz Biotechnology (NBS Biologicals, Hatfield, U.K.), respectively.

Received 15 November 1995 Accepted 6 June 1996

390 P. HARVEY ET A L

Table I-Types and numbers of tumours studied

No. of Age range No. of No. of Tumour type patients (median), years M:F resections biopsies

Adenocarcinoma 12 49-73 (62) 7:5 12 0 Squamous cell carcinoma 10 54-73 (69) 9: I 10 0 Small cell anaplastic carcinoma 7 49-78 (65) 3:4 1 6% Malignant mesothelioma 9 48-70 (63) 8: 1 0 9

*One of these cases was a supraclavicular nodal metastasis.

Immunohistochemistvy

HGFiSF and met immunoreactivity were determined by the indirect immunoperoxidase procedure. After fixation in 10 per cent formal saline, all specimens were processed by standard methods to paraffin wax blocks. Serial 4 pm sections were applied to polylysine-coated slides (BDH, Poole, U.K.) and dried at 50°C for at least 2 h. The sections were deparaffinized in xylene, rehy- drated through graded alcohols, and incubated in 2 per cent (vo1:vol) hydrogen peroxide for 30 min to block endogenous peroxidase. They were then washed in 70 per cent alcohol for 5 min and treated with trypsin (1 mg/ml; Sigma) at 37°C for 30 min. The sections were then washed in phosphate-buffered saline (PBS), blocked with 10 per cent goat or swine serum (Dakopatts, Copenhagen, Denmark) for 1 h at 37°C and washed for 5 min in PBS. The primary antibodies were diluted in PBS containing 1 per cent bovine serum albumin and the sections incubated for 30 niin at 37°C. Immunostaining was carried out using the Dako StreptABComplex with 30 min incubation at 37°C for both the biotinylated secondary antibodies and the streptavidin-peroxidase step, both followed by three washes in PBS. The colour reaction was developed using diaminobenzidine substrate tablets (Sigma) for 5 min. Sections were counterstained with Mayer's haematoxylin, dehydrated, dried, and embedded in DePeX mountant (BDH, Poole, U.K.).

For both met and HGF/SF negative controls, non- immune immunoglobulins were substituted for the primary antibodies for comparison with each tumour examined. For met, the specificity of the antibodies was further confirmed by pre-absorption of the primary antibody with the peptide used for immunization. Simi- lar experiments were carried out for the HGF/SF antibody with purified recombinant HGF/SF produced in Sf21 insect cells using the baculovirus expression system (Newman, Warn & Warn, unpublished).

RESULTS

All the sections studied were obtained from archival material. The full range of diagnostic stains, including appropriate histochemical and immunohistochemical

preparations, was available for review. This helped in particular with making the distinction between stromal cells and tumour cells in malignant mesothelioma. In general, however, a complete assessment could be made using haematoxylin and eosin-stained sections and the immunohistochemical preparations for HGF/SF and c-met. Staining for HGF/SF and met protein was reli- able and without background. Many sections of lung tumours also included histologically normal lung tissue, allowing a comparative evaluation. Endothelial cell immunoreactivity for HGF/SF provided an internal positive control (35/38 samples). Furthermore, staining of term placenta was identical to previously published observations.21 Internal positive controls were not required for the met protein studies. Specificity was confirmed by the lack of staining of non-immune antibody controls in each case and a lack of staining among non-cellular elements except in malignant mesothelioma (see below). Pre-absorption of the antibodies with appropriate antigens abolished immunostaining for met and reduced it considerably for the HCF/SF. In the latter case, there was probably insufficient excess of antigen for complete blocking. The specificity for the HGF/SF antibody has also previously been demonstrated in metanephric kidneyz2 and that used for met staining in pancreatic tumour material.I4

HGFISF imrnunostaining HGF/SF immunoreactivity was present in some

samples of both normal mesothelium (4/8) and normal bronchial and bronchiolar epithelium ( 1 1/17), where these elements were present, and also in endothelium within the lung. Endothelial and mesothelial staining formed a thin line, but it was impossible to assess further HGF/SF distribution within these cell types. Normal bronchial and bronchiolar epithelium showed variable staining. There was distinct polarity of reactivity on occasion, with strong apical epithelial staining (Fig. 1 a). The cytoplasm of immunoreactive bronchial cells also generally showed weak, diffuse staining. In benign squamous metaplasia of bronchi, the flat surface squamous cells showed occasional staining for HGFISF, the under- lying less mature epithelium being negative (not shown).

All types of tumour except small cell anaplastic carcinoma showed at least some reactivity for HGF/SF

Fig. 1- Immunostaining for HGF/SF of (a) bronchial epithelium, (b) malignant mesothelioma, (c) lung adenocarcinoma, (d) squamous cell carcinoma. and ( e ) small cell amplastic carcinoma. Immunostaining for met of (f) bronchial epitltelium taken from a region adjacent to a tumour. (g) mesothelioma. (h ) adenocarcinoma, ( i ) squamous cell carcinoma, and 6) small cell unaplastic carcinoma. ( a ~ c, c-h. j ) x 400; (d, i) x 200

HGFlSF AND met IN LUNG TUMOURS 39 1


Table 11-Reactivity of tumour cells and stromal fibroblasts for HGFISF in primary pulmonary adenocarcinoma (12 cases), squamous cell carcinoma (10 cases), and small cell anaplastic carcinoma (6 cases), and in malignant mesothelioma of the pleura (9 cases). Two small cell anaplastic carcinomas and one malignant mesothelioma had insufficient stroma to assess. A seventh small cell anaplastic carcinoma, metastatic to a lymph node, was also negative

Tumour type

Staining of tumour Staining of stromal cells fibroblasts

Strong Weaklfocal None Strong Weaklfocal None

Malignant mesothelioma 8 1 0 6 0 2 Adenocarcinoma 5 4 3 0 3 9 Squamous cell carcinoma 3 4 3 0 1 9 Small cell anaplastic carcinoma 0 0 6 0 1 3

Table Ill-Two-by-two table showing the relationship bethecn tuniour cells and stromal fibroblasts for HGF/SF reactivity

but none in tumour cells (Table Ill). In all other cases, reactivity of stromal fibroblasts was associated with reactivity in the neoplastic cells.

Tumour cells Tumour cells positive negative

Stronial cells posithe 10 1 Stromal cells negative 14 9

(Table 11). All malignant mesotheliomas were reactive (Fig. Ib). The staining in nearly all was intense and affected most cells, but in one case only a minority of cells stained rather weakly. Staining of one sarcomatoid malignant mesothelioma was similar to that of tumours with an epithelial or a mixed phenotype. Staining was cytoplasmic and mainly diffuse in all tumours, but some cells showed granular cytoplasmic reactivity. There was some degree of reactivity for HGF/SF in about three- quarters of adcnocarcinomas and squamous cell carcinomas, but staining was variable and strong staining was present in only a minority (Table TI). Adenocarci- nonias showrd diffuse, occasionally granular, cytoplasmic reactivity and often accentuation on their luminal surfaces (Fig. Ic). Squamous cell carcinomas often showed focal staining, the large polygonal cells being reactive and smaller and less mature cells being negative (Fig. Id). None of the six primary or one metastatic small cell anaplastic carcinomas was reactive for HGFiSF (Fig. le).

The tumour stroma of all malignant mesotheliomas except two showed strong staining of its fibroblast population for HGF/SF. One of these was the same case as that which showed only weak reactivity for the tumour cells. Where strong fibroblast staining occurred, the stromal matrix also appeared stained (compare Fig. Ib with Fig. Id). Only a minority of adenocarcinomas, squamous cell carcinomas, and small cell anaplastic carcinomas showed stromal fibroblastic reactivity for HGF/SF, and then only weakly (Table 11). Of 34 cases where the tumours contained sufficient stroma to assess, only one, a small cell anaplastic carcinoma, showed immunoreactivity for HGF/SF in stromal fibroblasts,

Immunostnining for met Both normal and neoplastic elements showed reactiv-

ity for met. Seemingly normal bronchial epithelium taken from regions adjacent to the tumour showed strong, polarized staining for met (Fig. I f ) . The apical plasma membranes, including the cilia, were strongly positive, whereas the lateral and basal surfaces showed little or no enhanced staining. The cytoplasm of these cells also showed diffuse staining; In addition, some nuclei demonstrated apparent reactivity for met (not shown). Basal cells of the bronchial epithelium showed only weak diffuse cytoplasmic staining (Fig. 1 f). Meso- thelial cells also showed polarized staining, with a strong line along their apical plasma membranes (not shown). The basal plasma membranes also demonstrated some staining and there was diffuse cytoplasmic reactivity for met.

Malignant mesothelioma cells usually demonstrated strong met staining (Fig. Ig). Much of this was diffuse and cytoplasmic and, in many tumours, granular cytoplasmic staining was also evident. On occasion, plasma membrane staining was also obvious. Tumour stromal cells also occasionally reacted positively for met in mesotheliomas and also in all other tumours studied. Both cytoplasmic and plasma membrane staining was evident.

All types of lung tumour showed strong to moderate met staining. For adenocarcinomas, the staining was frequently intense, although in other regions weaker staining occurred (Fig. Ih). Much of the staining was cytoplasmic and, focally, granular. Like normal bronchial epithelium, adenocarcinomas frequently showed linear accentuation of staining at their luminal surfaces.

All squamous cell carcinomas showed at least some staining for met (Fig. l i) . The pattern was variable, from strong to faint. The whorls of larger cells sometimes stained more strongly, sometimes less, than surrounding cells. In general, the staining was largely cytoplasmic, but sometimes plasma membrane staining was also apparent.

HGFlSF AND met IN LUNG TUMOURS 393

All small cell anaplastic carcinomas showed met staining (Fig. lj). The pattern of staining and its intensity were variable. In general, there was diffuse cytoplasmic staining but sometimes tumour foci showed distinct granular met reactivity. Staining of the plasma membrane was sometimes evident.

DISCUSSION

This study suggests that for some types of primary lung carcinoma and also for malignant mesothelioma, the tumour itself, as well as stromal fibroblasts, may be a source of HGF/SF. Because HGF/SF has been considered to be largely a paracrine effector of met

it was something of a surprise to find all the sections of malignant mesotheliomas and also the majority of those from adenocarcinomas and squainous cell carcinomas displaying at least some staining of tumour cells of HGF/SF, together with usually strong staining for met. However, breast,’ I bladder,l3 and panc rea t i~ ’~ carcinomas may co-express met and HGF/SF, suggesting possible autocrine effects.

Some cell lines derived from lung adenocarcinomas and squamous cell carcinomas are characterized by both constitutively phosphorylated met and secretion of HGF/SF.24 Some of these lines show HGF/SF antibody- mediated inhibition of proliferation and met tyrosine phosphorylation, demonstrating that HGF/SF can exert autocrine effects. In contrast, nearly all cell lines derived from lung small cell anaplastic carcinomas did not show expression of HGF/SF, but did express met tran- scripts.25 These in vitro data correlate well with our observations for tumours in sitti. Other studies demon- strate the likely complexities of the system. At least one cell line derived from a human lung squamous cell carcinoma secretes HGF/SF but does not express met.26 Furthermore, the lung small cell carcinoma line SBC-5 secretes the two-kringle truncated form of scatter factor, which in this cell line acts in autocrine fashion to stimulate motility.27 Whether the antibody used in this study can distinguish between HGF/SF and the two-kringle variant is not known.

So far, there has been only one other immunohistochemical study of the distribution of HGFlSF in human lung tumours.28 These authors reported strong staining of the basement membranes of tumours, bronchial epithelium and small blood vessels, and of the stroma. The significance of the differences from our study is not clear, but may relate to different methods of fixation and/or the use of different antibodies.

Each tumour type showed a particular pattern of staining for HGF/SF. All malignant mesotheliomas showed HGF/SF immunoreactivity, mostly strong and fairly uniform cytoplasmic staining. The adenocarcinomas exhibited mainly luminal surface staining, suggesting possible polarized secretion in the case of well-differentiated cells. Immunostaining, however, only demonstrates the location of a protein, not its synthesis. For squamous cell carcinoma, the strongest immunostaining was frequently of the large, well-differentiated cells, raising the possibility that the differentiation of the

tumour cells is associated with the switching on of HGFlSF synthesis. The absence of immunoreactivity for HGFlSF in small cell anaplastic tumours suggests that HGFlSF may be less significant in the development of this tumour type, which is characteristically associated with the secretion of a variety of neuroendocrine peptides, including bombesin, which likely functions as an autocrine growth factor.2y The growth regulators of this type of tumour may well be different from those of the other classes.

The finding that stromal fibroblasts frequently stained for HGF/SF was not unexpected, as human lung fibroblasts secrete HGF/SF in vitr0.’3~ As for the tumour cells, stromal fibroblasts stained most strongly and most frequently in malignant mesothelioma. In contrast, the primary lung tumours, including small cell anaplastic carcinoma, showed only sporadic, weak staining of the stromal cells for HGF/SF. A striking finding was that in all cases but one, when stromal cells stained for HGF/SF, so did the tumour cells. In contrast, where no stromal staining at all was seen, the tumour cells showed at least some staining in approximately 60 per cent of cases. Most of these double positives (6/10) came from malignant mesotheliomas and here the extracellular matrix was also frequently stained. Thus, in general, mesotheliomas showed the strongest association of HGF/SF staining of tumour cells and stromal fibroblasts.

Bronchial epithelium and mesothelium adjacent to the tumour, and also endothelium, stained for HGF/SF on occasion. Again, it was unexpected to find both epithelial and endothelial cell staining in sections, in view of in vztvo observations suggesting a paracrine mechanism of HGFlSF a ~ t i o n . ~ . ~ ’ However, immunostaining of normal bronchial epithelium and mesothelium with HGF/SF has been previously rep~rted.~~-’O Partially hepatectomized rats showed raised HGF/SF mRNA levels in the lungs,31 where in situ hybridization revealed a marked increase in endothelial cell HGFiSF mRNA levels, suggesting that this class of cell may be activated to secrete HGFISF. This study also presented evidence that pulmonary macrophages synthesize HGF/SF mRNA.31 We, too, found that alveolar macrophages appeared to stain for HGF/SF, but the controls also did so, albeit less strongly, making interpretation difficult. We suspect that alveolar macrophages may be another source of HGF/SF. Further study is now required to determine in what situations these cell types may synthesize and secrete HGF/SF in human lungs.

All the cell types described in this paper stained for met. Adenocarcinomas were particularly interesting in

that they frequently showed strong apical plasma membrane staining for met, a feature which they shared with histologically normal bronchial epithelial cells from regions surrounding the tumour. The situation in the lung resembles that in the breast32 and pancreas, l 4 where normal epithelial cells and adjacent adenocarcinomas both show strong luminal met staining. In contrast, the basal bronchial cells, which are thought to be pro- genitors for the bronchial epithelium, showed no such polarized staining. There is evidence that for kidney and breast tissue and for cell lines, HGF/SF induces


tubulogenesis.2 HGF/SF may be a factor inducing bronchiolar differentiation during development and in response to injury. The other tumour types also showed met membrane staining, but not in this polarized fashion.

In conclusion, we report complex distribution pat- terns for HGF/SF and its receptor, met, in primary lung carcinomas and pleural mesotheliomas, in associated stromal fibroblasts, and in the bronchial epithelium and mesothelium from areas adjacent to the tumours. For mesothelioma in particular, and also for a number of cases of adenocarcinoma and squamous cell carcinoma, there was co-expression of HGF/SF and met. It is not clear how far paracrine versus autocrine effects may occur. Quite possibly both mechanisms may be in play. Furthermore, many tumours showed strong met immunostaining, suggesting overexpression. Constitu- tive overexpression of met has been found to cause its activation due to autophosphorylation in the absence of HGF/SF;33 this represents an alternative mechanism for the inappropriate regulation of growth, motility, and differentiation pathways. Lastly, the secreted HGF/SF may also have some role in the enhanced vascularization necessary for tumourigenesis, because it has previously been found to be a powerful stimulator of angiogenesis.'

ACKNOWLEDGEMENTS

We thank Miss Julia Burton and colleagues for tech- nical assistance, AICR and Big C for financial support, and Mrs Jill Gorton for putting the manuscript on disk. For part of the research. Dr Harvey was a Fellow of the EU Human Capital & Mobility Programme. (CHRX- CT92-0043)

REFERENCES I.

7 -.

3.

4.

5.

6.

7.

8.

9.

Rosen E. Nigam S. Goldberg 1. Scatter factor and the met receptor: a paradigm for niesencliyniallepitlielial interaction. J CeN Bid 1994; 127: 1783 1787. Warn RM. A scattered jigsaw. Nature 1995; 376 723. Bottaro DP, Rubin JS, Faletto DL, ef ul. Identification of the hepatocyte growth factor receptor as the nicl proto-oncogene product. Science 1991;

Naldini L, Vigna E. Narsimhan RP, er u/. Hepatocyte growth factor (HGF) stimulates the tyrosinc kinase activity of the receptor encoded by the pi-oto-oncogene I J ~ C I . Onmgriw 1991; 6 501 -504. Matsumoto K. Nakamura 7'. Roles of HGF as a pleiotropic in organ regeneration. In: Goldberg 1, Roseii E, eds. Hepatocyte Growth Factor- Scatter Factor and the nwf Receptor. Basel: Birchiuser, 1993; 225-250. Tsubouchi H, Gohda E, Strain A, Daikuhara Y. The role of HGFlSF in animal and human hepatic physiology and pathology. In: Goldberg I , Rosen E. eds. Hepatocyte Growth Factor-Scatter Factor and the mef Receptor. Basel: Birchiuser. 1993; 251 274. Stoker M, Gherardi E. Perryrnan M, Gray J. Scatter factor is a fibroblast- derived modulator of epitheli;rl cell mobility. Nuture 1987; 327: 939-942. Weidncr KM, Bchrens J. Vandekerckhove J, Birchmeier W. Scatter factor: molecular characteristics and effect on the invasiveness of epithelial cells.

Bussolino F. Di Renzo MF. Ziche M, e f ul. Hepatotype growth factor is a potent angiogcnic factor mhich stimulates endothelial cell motility and grouth. J Cell Bid 1992: 119: 629-641

251: 802-804.

.I c(,// Bifli I 990; I I 1: 2097 2 I ox.

10.

11

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

21.

28.

29.

30.

31.

32.

33.

Grant DS, Klcinman HK, Goldbcrg ID. e f ul. Scatter Factor induces blood vessel formation in vivo. Proc Null Acud Scr LISA 1993; 90: 1937 1941 Wang Y, Selden AC, Morgan N. rt ul. Hepatocyte growth factorlscatter factor. Expression in human mammary epithelium Am J Puthol 1994: 1 4 4 615-682. Yamashita J , Ogawa M, Yamashitd S. ef u/. Immunoreactive hepatocyte growth factor is a strong and independent predictor of recurrence and survival in human breast cancer. Cuncer Res 1994; 54: 1630-1633. Joseph A, Weiss G, Jin L, e f d. Expression of scatter factor in human bladder carcinoma. J Nut/ Cuncer [II.Y/ 1995; 87: 372 373. Eberg M, Yokoyama M, Friess H, Buchler MW, Korc M. Coexpression of the met protooncogene and hepatocyte growth factor in human pancreatic cancer. Cuncer Res 1994: 5 4 5775-5778. Prat M, Narsimhan RP, Crepaldi T. Nicotra MK. Natali PG, Comoglio PM. The receptor encoded by the human r n c f oncogene is expressed in hepatocytes, epithelial cells and solid tumors. lur J Cunwr 1991; 49: 323-328. Di Renzo MF, Narsimhan RP, Olivero M. er ul. Expression of the MetlHGF receptor in normal and neoplastic human tissues. Oncogcw 199 I ; 6 1997-2003. Natali PG, Nicotra MR, Di Renzo MF, e r d. Expression of mrrlHGF receptor in human melanocytic neoplasms: demonstration of the relationship to malignant melanoma tumour progression. Br J Cuncw 1993; 6 8 746-750. Di Renro MF, Poulsom K, Olivero M, rf a/. Expression of the meil hepatocyte growth Factor receptor in human pancreatic cancer. Cuncrr Rrs 1995; 5 5 1129-1138. Kenworthy P, Dowrick P, Baillie-Johnson H, G z r ul. The presence of scatter factor in patients with metastatic spread to the pleura. Br J Cuncer 1992; 66: 243-247. Eagles G, Warn A, Ball RY. et ul. Hepatocyte growth factorlscatter factor is present in a high proportion of pleural effusion fluids from cancer patients. Br J Cuncer 1996; 73, 377-38 I . Saito S, Sakakura S, Enomoto M, lchijo M, Matsumoto K. Nakamura T. Hepatocyte growth Factor promotes the growth of cytotrophoblasts by the paracrine mechanism. J Bioclzern 1995; 117: 67 I ~ 676. Woolf AS, Kolatsi-Joannou M, Hardman P, el ol. Roles of hepatocytc growth factor/scatter factor and the met receptor in the early development of the metanephros. J Cell B id 1995; 128: 171 -184. Sonnenberg E, Meyer D, Weidner KM, Birchmeier C. Scatter Factor/ hepatocyte growth factor and its receptor, the mrt tyrosine kinase, can mediate a signal exchange between mesenchyiiic and epithelia during mouse development. J Cell Bid 1993; 123 223-235. Tsao M, Zhu H, Giaid A, Viallet J, Nakamura T. Park M. Hepatocyte growth factorlscatter factor is an autocrine factor for human normal bronchial epithelial and lung carcinoma cells. Cell G~YJNY/I D$/w 1993; 4: 571-579. Rygaard K, Nakamura T, Sprang-Thomsen M. Expression of the proto- oncogenes nlef and c-kit and their ligands. hcpatocyte growth factorlscatter Factor and stem cell factor, in SCLC cell lines and xenografts. Br .I Cunciv 1993; 67: 37-46. Yoshinaga Y, Fujita S, Gotoh M, Nakamura T. Kikuchi M, Hirohashi S. Human lung cancer cell line producing hepatocytc growth Factorlscatter factor. Jpn J Cuncer 1992; 8 3 1257-1261. ltakura Y, Yamanioto T, Matsumoto K, Nakamura T. Autocrine stimu- lation of motility in SBC-5 human lung carcinoma cells by a two-kringle variant of HGF. Cuncer Lett 1994; 8 3 235-243. Yoshinaga Y, Matsuno Y, Fujita S, e f ul. lmmunohistochemical detection of hepatocyte growth factorlscatter factor in human cancerous and inflainmatory lesions of various organs. .//m .I Cuncer Re.! 1993; 84: 1150-1158. Gardar A, McDowell E. Pathobiology of lung cancers. I n : Rosen S, Mulshine J, Cuttitta F, Abrams P, eds. Biology of Lung Cancer: Develop- ment and Treatment. New York: Marcel Dekker, 1988; 1-42, Wolf HK, Zarnegar R, Michalopoulos GK. Localization of hepatocyte growth factor in human and rat tissues: an immunohistochcmical study. Hepatology 1991; 1 4 488494. Yanagita K, Nagaike M. Ishibashi H, Niho Y, Matsumoto K , Nakamura T. Lung may have an endocrine function producing hepatocyte growth factor in response to injury of distal organs. B ~ I J ~ ~ C I W Biop/rj\ Reh Cbnzrnzm 1992; 182: 802-809. Tsarfaty I, Resau J, Rulong S, Keydar 1. Faletto D, Vande Woude G. The met protooiicogeiie receptor and lumen formation S&wc 1992; 257: 1258 1261. Ponzetto C, Giordano S, Peverali F. ef r r l . Mei I S amplified but not mutated in a cell line with activated met tyrosine kinase. Ontogme 1991; 6 553 559.

immunoreactivity for hepatocyte growth factor/scatter factor and its receptor, met, in human lung...

Documents