REACTIVITY OF MONOCLONAL ANTIBODIES DIRECTED AGAINST LUNG CANCER ANTIGENS WITH HUMAN LUNG, BREAST AND COLON CANCER CELL LINES

A panel of monoclonal antibodies (n=72 including controls) directed against lung cancer antigens was screened immunohistochemically against a panel of seven human lung cancer cell lines (including small cell carcinoma, squamous cell carcinoma, adenocarcinoma and mesothelioma), six human breast cancer cell lines and one human colon cancer cell line, The majority of the antibodies (n=42) reacted also with antigens present on breast and colon cancer cell lines, This cross reactivity especially between lung and breast cancer cell lines is not altogether unexpected since antigens common to breast and lung tissue including their neoplasms such as MUC I antigen have been described, Our results indicate that epitopes shared by lung and breast cancers are probably more common than previously thought. The relevance for prognosis and therapy of these shared antigens, especially as disease markers in breast cancer, has to be investigated,


INTRODUCTION
Some antigens common to lung and breast epithelia can be classified as oncodevelopmental.One of these, the MUCI antigen, detected during development in lung, mammary gland and other tissues of epithelial origin (Braga et al., 1992), is conserved during evolution (Welsch et al., 1990, Spicer et al., 1991, Pemberton et al., 1992) and the antigen can be recovered from the human broncho-alveolar-Iavage (Schumacher et at., 1989).
Since mucin antigens which are expressed in lung and mammary gland have been described, the monoclonal antibody panel (submitted to the panel of the third international association for the study of lung cancer [IASLC] workshop on lung tumour and differentiation antigens) was tested for cross reactivity with human breast cancer cell lines.In addition serving as a further control, a human colon cancer cell line was included in the test panel of cell lines.

MATERIALS AND METHODS
Human breast cancer cell lines (MCF7, BT-S49, T47D, HBLIOO, MDA-MB-1S7, HSS78T) and the human colon cancer cell line HT29 were obtained from the European Cell Culture Collection (Porton Down, Salisbury, Wiltshire, UK) and maintained under the standard culture conditions supplied with the data sheets for each cell line.The human lung cancer cell lines N-417 (derived from small cell carcinoma, variant, kindly provided by Dr. D. N. Carney, Mater Misericordiae Hospital, Dublin, Ireland), SW2 (classic, kindly provided by Dr. S. Bernal, Boston, USA), ZLS and ZL34 (from mesothelioma, generated in Zi.irich), U 17S2 (from squamous cell carcinoma, kindly provided by Dr. J. Berh, Uppsala, Sweden) and Al2S (from adenocarcinoma, obtained from American Type Culture Collection, Rockville, Maryland) were cultured in RPMI-1640 medium supplemented with 2 mMoll1 L-glutamine and 10% fetal calf serum.The immortalised human bronchial epithelial-derived cell line (BEAS-2B, kindly provided by Dr. C. Harris, Bethesda, Maryland) was cultured in Hams FI2 based medium supplemented with Smg/L insulin, Smg/L transferrin , 70ug/L hydrocortisone, 0.1 mg/L vitamin A, 6S0ug/L triiodo-L-thyronine, 2mg/L epinephrine, 30mg/L bovine pituitary extract, SOmg/L bovine serum albumin, Sug/L purified mouse epidermal growth factor, SuM/L ethanolamine and SOmg/L gentamycin.
For the immunofluorescence assay the breast and colon cancer cell lines were grown to confluence on coverslips and briefly fixed in cold methanol.The cells were incubated with the antibody panel (Table 1) including the positive and negative control reagents and the anti mouse or rat FITC-labelled antibodies were used for visualisation.The fluorescence intensity was evaluated semi-quantitatively: -= no fluorescence, (+) = very weak fluorescence, + = weak fluorescence, ++ = modest fluorescence, +++ = intense fluorescence, ++++ = very intense fluorescence .The intensity measurements of the lung cancer cell lines were obtained by FACS; the results of the FACS analyses were transformed into the above scale.

RESULTS
The details of the results of this study are summarised in Table 2 and can be classified into four groups: 1) those antibodies which did not react with any of the cell lines (n = 14, including the negative controls), 2) those which reacted with breast and colon cancer cell lines only (n = 11), 3) those which reacted with lung cancer cell lines only (n = 17) and 4) those which reacted with breast, colon and lung cancer cell lines (n = 30).Amongst groups two to four the numbers of reactive antibodies and the intensity of fluorescence varied considerably.In some cases antibodies reacted only with one cell line (e.g. the antibody number 16 reacted with the cell line T47D only), while other antibodies showed a broader reactivity (antibody number SO reacted with all the cell lines except two breast cancer cell lines MDA-MB-1S7 and HBL 100).Differences in fluorescence intensity were detected not only between the different cell lines but also between cells of the same cell line (e.g.only 40% of the cells from the cell line BT S49 showed strong reactivity with antibodies number 3S & 36).In other cases some antibodies reacted with all cells of a particular cell line (Fig. 1) while other antibodies reacted only with some cells of that cell line (Fig. 2) ., 4,  703-712(1991), Mol. cell. BioI., 13, 1507-1515(1993)      -

DISCUSSION
The present study has shown that monoclonal antibodies designed for detecting oncodevelopmental antigens expressed in lung cancer can also share epitopes of breast and colon cancer cell lines.Furthermore a heterogeneity of the antibody binding towards the different breast cancer cell lines has been observed indicating a phenotypical diversity of these cell lines.Some of the many cross reactivities of the antibodies with the breast and colon cancer cell lines seem to be of functional interest which might have implications concerning the prognosis of both breast and lung cancer: I) The antibodies no 24-26 directed against the neuronal cell adhesion molecule (NCAM) immunoreacted only with the two cell lines derived from small cell carcinoma of the lung, while other antibodies with NCAM specificity showed a broader reactivity: antibody 4 reacted with the mesothelioma derived cell line ZL 34 and with the breast cancer cell line BT 549, antibody no 14 with the breast cancer cell line MCF7 and antibody no 19 with the breast cancer cell line BT549.At the moment it is not clear whether these are indeed different epitopes of the NCAM molecule recognised by the different antibodies, or whether they are cross reactivities with epitopes other than NCAM.Since neuroendocrine markers can be expressed by breast cancer (for review see Nesland et al., 1988) NCAM expression on these breast cancer deri ved cells is likely.
2) The antibody no 72 directed against a mucin antigen cross reacted with all lung cancer cell lines except the mesothelioma derived cell line ZL5 and with the breast cancer derived cell lines BT549, T47D and MCF7.Common expression of mucin antigens in normal and pathological conditions of lung and breast tissue is well known (Welsch et aI., 1990, Spicer et aI., 1991, Pemberton et aI., 1992) and this reac tivity is therefore not unexpected.
3) The mesothelioma specific antibody no 41 showed its strongest reactivity with the mesothelioma derived cell lines ZL5 and ZL34; in addition it reacted with the immortalised human bronchial-derived cell line BEAS-2B and the breast cancer cell line BT549, the nature of this binding pattern being obscure at present.
4) The distribution of cell surface receptors which react with the extracellular matrix differs.In addition to reactivity with various lung cancer derived cell lines, laminin receptor immunoreactivity was de.tected on the breast derived cell lines T47D (antibody no 20) and BT549 and HBLlOO (antibody no 59).The immunoreactivity of this receptor seems to be more widespread than that for beta-l-integrin antibodies detected in HS578T and HT29 (antibody 21).Alpha-6-integrin was detected on lung derived cell lines only (antibody no 22).The biological implications of these findings are not clear at present, but the interaction of these receptors with their Iigand(s) in the extracellular matrix seems to playa crucial role within the metastatic cascade (Hart and Saini, 1992).
Our findings therefore indicate that several antigens thought to be lung cancer specific can also be expressed on breast and colon cancer cell lines.Any claims towards specificity of many of those antibodies therefore have to be treated with great care.
Fig. la

Figure I .
Figure I. Photomicrographs showing FITC labelled cultured (a) HT29 colon cancer cell line and (b) MCF7 breast cancer cell line after reactivity with the antibody number 5 (KM 195).X650.The primary as well as the secondary anti-rat and anti-mouse antibodies were used at I :50 dilutions and incubated with the cell preparations for I hour, both at room temperature. Fig.2a Fig.2b

Figure 2 .
Figure 2. Photomicrographs of cultured (a) HT29 colon cancer cell line and (c) MC7 breast cancer cell line showing some FlTC labelled cells after reactivity with the antibody number 23 (MLuC-I) and the antibody number 54 (1.291) respectively.b) is a phase contrast photomicrograph of the same labelled cells in Fig. (a).X415.

Table I .
List of antibodies used in the present study

Table 2 .
Details of immunofluorescence reactivity.