J Cancer Res Clin Oncol (1990) 116:29 37

C~i/cer ~esearch Clinical 9 @ Springer-Verlag1990

Different pattern of expression of cellular oncogenes in human non-small-cell lung cancer cell lines P.E. Kiefer, B. Wegmann, M. Bacher, C. Erbil, H. Heidtmann, and K. Havemann Philipps-UniversityMarburg/Department of Internal Medicine, Divisionof Haematology/Oncology,D-3550 Marburg, Federal Republic of Germany

Summary. Altered and deregulated cellular oncogenes were found in many human solid tumors. Except for a few types of tumors that consistently exhibited specific altered proto-oncogenes, the majority of tumors are associated with a number of transcriptionally activated cellular oncogenes. In the heterologous group of non-small-cell lung cancer (NSCLC), nothing about a specific pattern of proto-oncogene expression is known. Therefore, we investigated the expression of a panel of cellular oncogenes in NSCLC cell lines. D N A and RNA from 11 established NSCLC cell lines (4 adenocarcinoma cell lines, 3 squamous cell carcinoma cell lines, 3 large-cell carcinoma cell lines and 1 mesothelioma cell line) were isolated and analysed using the Southern, dot blot and Northern hybridization technique, c-myc R N A expression was found in all NSCLC cell lines, L-myc expression only in 1 adenocarcinoma cell line, N-myc and c-myb expression in none of the t 1 cell lines examined. No c-myc amplification could be detected in the DNAs. v-sis-related m R N A was observed in 5/11 cell lines without association to a specific NSCLC subtype, v-src-related mRNA, found in all tested cells, exhibited increased levels in 1 adenocarcinoma cell line (A-549) compared to the other cell lines. Binding sites for epidermal growth factor (EGF) had been described previously in NSCL, therefore we found erbB homologue transcripts coding for the E G F receptor in all N S C L C cell lines. Also, c-rafl-, N-ras-, Ki-ras-, and H-ras-related R N A expression was observed in all lines. We conclude that L-myc, N-myc, and c-myb expression does occur less frequently in NSCLC than in SCLC. Also Abbreviations used." kb, kilobase(s); SSC, saline sodium citrate (1 x SSC is 0.15 M sodium chloride: 0.015M sodium citrate); oligo(dT), oligodeoxythymidylate; SDS, sodium dodecylsulfate; SCLC: small cell lung cancer; NSCLC: non-small-celllung cancer; EGF, epidermalgrowth factor; PDGF, platelet-derivedgrowth factor Offprint requeststo: K. Havemann

amplification does not appear to be an important mechanism by which the c-myc proto-oncogene is activated in NSCLC. A specific pattern of oncogene expression could not be detected in N S C L C cells; each cell line examined showed its own pattern. However, transcriptional activation of a proto-oncogene like erbB, ras, raf, src, and c-myc, which are all involved in the progression pathway of EGF, may be a common feature of NSCLC. Key words: Cellular oncogenes - Non-small-cell lung cancer

Introduction Lung cancer is currently divided in two major groups: the small-cell lung cancer (SCLC) group, and the heterogeneous non-small-cell lung cancer (NSCLC) group. SCLC has typical clinical and biological features: rapid tumor proliferation, frequent paraneoplastic syndromes, amine precursor uptake and decarboxylation properties, high metastasizing rate, responsiveness to chemotherapy (Carney et al. 1985; Gazdar et al. 1985). However, NSCLC, comprising approximately 75% 85% of all lung cancers, are characterized by few, if any, c o m m o n biological parameters. In this study we took the opportunity of finding common features in N S C L C on the basis o f p r o t o - o n cogene expression. A growing amount of evidence strongly suggests that abnormal activation of cellular oncogenes is involved in neoplastic transformation. Point mutation, deletion, translocation and gene amplification may be mechanisms that give quiescent proto-oncogenes oncogenic potential. It is generally assumed that the deregulated function of cellular oncogenes causes a loss of growth control in malignant cells (Bishop 1983;

30

Duesberg 1985; Heisterkamp et al. 1983; Collins et al. 1983; Alitalo et al. 1983; Pelicci et al. 1984; Dalla Favera et al. 1982a; Schwab et al. 1983; Santos et al. 1984; Orkin et al. 1984). Activated cellular ras genes including the Ha-ras, N-ras, and Ki-ras genes, found in some human NSCLC cell lines coding for altered p21 proteins, are capable of transforming phenotypically normal cells into tumorigenic cells (Santos et al. 1984; Pulciani et al. 1982; Yuasa et al. 1983, 1984; Capon et al. 1983b). Recently Rodenhuis et al. (1988) reported on Ki-ras activated by a point in codon 12 in fresh human lung adenocarcinoma. In an extended study, the authors found Ki-ras point mutations in 14/45 examined adenocarcinomas (Rodenhuis et al. 1987, 1988). With the exception of a few types of human tumors in which specifically altered oncogenes are consistently expressed, such as in Burkitt's lymphoma, SCLC, breast cancer, and chronic lymphatic leukemia, most tumors are associated with a number of transcriptionally activated proto-oncogenes (Klein 1983; Groffen et al. 1984; Dalla-Favera et al. 1982; Slamon et al. 1984; Kolata 1987; Slamon et al. 1987). It seems reasonable to suppose that multiple transcriptionally but not necessarily structurally activated proto-oncogenes are involved in growth regulation of human tumors. A list of preferably expressed cellular oncogenes may be typical of certain lung cancer types and may provide an approach for a better understanding of the biology in various types of lung cancer. Such studies cannot elucidate the mechanisms of tumor formation, but may give a better picture of the actual functional state in the tumor cells. We previously reported the expression of proto-oncogenes in human SCLC cell lines and found a high level of c-myc expression and DNA amplification in some cell lines. All cell lines expressed c-myb, N-ras, Ki-ras, Ha-ras, and c-rafl, 4/12 lines had elevated RNA levels of N-myc and 3 of these lines showed N-myc amplification and 1 had a simultaneous amplification of N-myc and cmyb (Kiefer et al. 1987). To compare the findings with the proto-oncogene expression in NSCLC we screened the transcription of 16 proto-oncogenes in 11 established NSCLC cell lines including 3 squamous cell carcinoma cell lines, 4 adenocarcinoma cell lines, 3 large-cell carcinoma cell lines, and 1 mesothelioma cell line. Materials and methods Cell lines. The cell lines used in this study were established in our laboratory (the SCLC cell line: SCLC-21 H, the large-cell carcinoma cell lines: LCLC-103H, LCLC-97TM, the adenocarcinoma cell line: ADLC-5M2, the squamous cell carcinoma cell lines: EPLC-65H, EPLC-32M, and the mesothelioma cell line: MSTO-211H) or were donated by Drs. Carney and Gazdar, NCI, Bethesda, Maryland (the

adenocarcinoma cell lines: NCI-H23, NCI-H125, NCI-549, and the epidermoid carcinoma cell line: A-431) and Dr. Bergh, Department of Pathology, University Hospitals, Uppsala (the large-cell carcinoma cell line: U-1752, and the squamous carcinoma cell line: U1810). The characteristics of these cell lines have been described elsewhere (Giard et al. 1973; Bergh et al. 1981, 1985; Bepler et al. 1988). The SCLC cell line grew as floating cell aggregates whereas all other cell lines grew substrate-adherent. They were kept in RPMI 1640 medium supplemented with 10% fetal bovine serum in a well humidified atmosphere of 5% CO2 at 37 ~ C. In certain NSCLC cell lines included in this paper, expression of some oncogenes had been tested by other authors: myc-related expression in U-1810, A549, H125 (Nau et al. 1985, 1986; Saksela et al. 1985), myb expression in H23, H125, A549, and U-1752 (Griffin and Baylin 1985) and the plateletderived-growth-factor-related c-sis expression and gene product in U-1810 (Betsholtz et aI. 1989). Isolation and blotting analysis o f genomic DNA. For isolation of high-molecular-mass cellular DNA, 108 cells obtained from a logarhithmically growing culture were lysed in 10 ml lysis buffer (0.5% SDS, 0.1 M NaC1, 0.02 M TRIS/HC1 pH 7.4, 0.001 M EDTA and proteinase K (100 ~tg/ml). This solution was uncubated overnight at 37 ~ C, extracted once with phenol and twice with chloroform/isoamyl alcohol (24/1). D N A was precipitated with 0.15 M sodium acetate (pH 5.) and two volumes of cold 100% ethanol, pooled, dried under vacuum and redissolved in water. For D N A dot blotting, DNA was denaturated in 0.5 M N a O H at 60 ~ C for 30 min, neutralised with 2 M a m monium acetate (NH4OAc), diluted serially and spotted on to nitrocellulose membrane equilibrated with I M N H 4 0 A c . For Southern blot analysis, restriction-endonuclease-digested DNA was electrophoresed in 1% agarose gel and transfered to nitrocellulose using the Smith and Summers method (Smith and Summers 1980). Filters were prehybridized (50% formamide, 5 x SSC, 5 z Denhard's solution, 50 mM sodium phosphate buffer pH 6.5,250 j.tg/ml denaturated salmon DNA) and hybridized (50% formamide, 5 x SSC, 1 x Denhard's solution, 20 m M sodium phosphate buffer pH 6.5, 10% dextran sulfate, 100 ~tg/ml denatured salmon DNA) to nick-translated probes. Blots were washed for 4 z 5 rain in 3 x SSC, 0.1% SDS at room temperature, then for 3 x 20 rain in 0.1 x SSC, 0.1% SDS at 50 ~ C and exposed to Kodak XAR-5 film for various lengths of time. Isolation and blotting analysis of RNA. Total R N A was isolated from I0 s logarithmically growing cells by the thiocyanate/guanidium procedure and poly(A)/RNA was further purified by oligo(dT)cellulose chromatography (Chirgwin et al. 1979; Aviv und Leder 1972). The RNA preparations were tested for integrity by gel electrophoresis. Gel electrophoresis, R N A dot blotting and Northern blot analysis were done essentially as described previously; briefly, for RNA dot blots, the RNA was denatured in 6 • SSC, 0.22 M formaldehyde for 15 rain at 60 ~ C, diluted serially and spotted on nitrocellulose filters; for Northern blots RNA was electrophoresed in glyoxal gel and transferred with 20 x SSC to nitrocellulose (Kiefer et al. 1987). The blots were washed as noted above for D N A filter hybridisation. Recombinant DNA probes. The nick-translated DNA probes used in hybridisation were: a chicken fl-actin probe (P7000, Inotech); human c-myc, a 1.3-kb ClaI-EcoRI fragment, and a 1.7-kb Sstl-SstI fragment (Nau et al. 1985; Little et al. 1983); human Haras, a 6.6-kb BamHI fragment (Capon et al. 1983a), human Ki-ras a 1.1-kb PstI fragment of the Ki-ras2 (McCoy et al. 1984), pSW11-1 clone; human N-ras, a 1.5-kb EcoRI fragment (Murray et al. 1983); human c-mos, a 2.7-kb EcoRI fragment (Watson et al. 1987); human N-myc 1.0-kb EcoRI-BamHI fragment (Nau et al. 1986); human Lmyc, a 1.8-kb SmaI-EcoRI fragment (Nau et al. 1985); chicken cmyb, a 1.3-kb HindIII fragment of the pCM 1.3 clone (Gonda and Bishop 1983); human c-rafl, a 1.8-kb EcoR1 fragment of pilE1 clone (from U. Rapp); human c-fos, a 9.0-kb EcoRI fragment ofpc-fos(hu-

31 man)-1 clone (Miller et al. 1984); v-erbA, 0.8-kb PvulI-SstI fragment ofp-erb/t clone (Vennstr6m et al. 1980); v-fes, 0.5-kb PstI fragment (Sherr et al. 1980); v-fms, a 1.5-kb PstI fragment of pSM3 clone (Donner et al. 1982); v-src, a 0.8-kb PvuII fragment (DeLorbe et al. 1980), and V-sis, a 1.2-kb PstI fragment oflambda-SSVI clone subcloned in pBR322 (Robins et al. 19831).

Results

Quantitative evaluation of proto-oncogene expression in N S C L C cell lines Total cellular R N A was isolated f r o m 11 exponentially growing N S C L C cell lines and the same R N A batches were analyzed by the d o t blot techniques with 32P-labelled oncogene probes using a/~actin p r o b e and a glyceraldehyde p h o s p h a t e d e h y d r o genase probe (not shown) as internal control. In order to exclude the possibility that lack o f hybridization was due to n o n h o m o l o g y between the viral oncogene

A

9

I Absorbonce

I 0

0

I .15

X

9

I .3

XX

9

I .6

XXX

', 1.4

XXXX

-related probes and their cellular counterparts, we dotted h u m a n D N A as positive controls on each blot. The a u t o r a d i o g r a m s were quantified by densitometer tracing. To simplify the presentation o f the oncogene expression data we used a scale o f increasing intensities o f hybridisation (Fig. 1A and B). D a t a on the expression related to nine oncogenes are summarized in Table 1. Transcripts o f e r b A , f e s , f m s , N-myc, and myb were n o t detected in any cell line. L-myc transcripts were f o u n d only in the R N A f r o m the adenocarcin o m a cell line N C I - H 2 3 . C-rafl, erbB, v-src, N-ras, Ki-ras, and Ha-ras-related R N A expression was observed in all cell lines. The raf and ras expression exhibits approximately equivalent levels, whereas the src R N A level war m a r k e d l y increased in the adenocarcin o m a cell line A-549. erbB m R N A also exhibits higher levels in U-1752 and A-549 than in the other cell lines. As a positive control (Table 1), R N A f r o m the cell lines A-431 containing an amplified erbB gene was blotted together with the R N A f r o m N S C L C cells. The R N A levels approximately reflected the E G F binding sites expressed by N S C L C and A-431 cells (Haeder et al. 1988; Sherwin et al. 1981). c-myc transcripts were expressed in various a m o u n t s by all cell lines examined. The expression data were confirmed in two separate experiments and the R N A levels in d o t blot assays were generally consistent with the results o f N o r t h e r n blots (Figs. 1A and 2A).

Fig. 1. A Scale of intensity of hybridization signals. The scale ranged from 0, corresponding to no expression or RNA digested with RNase A and T1 as negative control, to + + + + for high expression. Numbers indicate intensities as determined by densitometer tracings using a Beckman spectrophotometer DU 6. B Dot blot hybridization of c-myc, v-src and fl-actin. Total RNA was serially I : 1 diluted (from 15 gg to 0.95 gg), spotted onto nitrocellulose paper and hybridized with the 32p-labelled oncogene probes and with the fl-actin probe used as internal control

32 A correlation of the proto-oncogene expression observed only in some cell lines (src, L-myc, sis) with the morphological cell type of N S C L C could not be revealed. The expression of c-myc was more abundant in N S C L C cell lines with short population doubling times than in cell lines with long doubling times, but it was not possible to trace this to a specific N S C L C cell-type.

Analysis of proto-oncogene-related transcripts The sizes of proto-oncogene-related transcripts were determined by the Northern blot technique. One cmyc transcript of 2.3 kb was found in all cell-lines tested with the exon III probe (Clal/EeoRI) (Fig. 2A). In cells with high levels of c-myc m R N A , we detected an accessory smaller band of 2.0 kb, which may represent the transcript derived from promoter P2, 150 base pairs (Prehn et al. 1984) downstream from promoter Pl. Two src-related 3.5-kb and 5.0-kb transcripts were demonstrable in the R N A from all cell lines examined, bands with differing proportions to each other. An Lmyc transcript of 3.8-kb was found exclusively in the adenocarcinoma cell line NCI-H23. A l t e t al. (1986) demonstrated that the 3.8-kb transcript produced by SCLC cell lines represents all three L-myc exons whereas the smaller R N A species (2.0) found more frequently in SCLC cells contains only exon 1 and exon 2. A single 3.5-kb c-rafl m R N A was detected in N S C L C cells. The size of sis-related transcript was determined as 4.2 kb as described elsewhere (Betsholtz et al. 1989) (Fig. 2B).

c-myc is not amplified or rearranged in N S C L C cell lines In order to determine whether the high level of c-myc expression could be caused by gene amplification, as in SCLC where the c-myc gene is frequently amplified or by other genomic alterations such as rearrangements found in Burkitt's lymphoma, we analyzed the genomic D N A by restriction endonuclease analysis

Fig. 2. A Northern blot analysis shows the typical 2.3-kb c-myc transcript in all non-small-celllunger center (NSCLC) cell lines examined. In cells with high levels of c-myc expression, a smaller band of 2.0 kb was detected, which may correspond to the P2 promoter transcript. B Transcripts related to c-sis, L-myc, c-src, and c-raJ7detected in various human NSCLC cell lines.

33 T a b l e 1. Synopsis o f the relative R N A expression in non-small-cell lung c a n c e r cell lines a

c-myc

Cell line Squamous EPLC-32 M EPLC-65H U-1752

L-myc

x

0 0 0

0

x x x x x x

v-src

v-sis

v-erbB

c-rafl

Ha-ras

Ki-ras

N-ras

x

0

x

x

x

x

x

x

x

x x

x

x

x x

x x

x

x x

x x

x

x x

x x

x

Adeno ADLC-5U2

x x x

NCI-H23

x x

x

0

x

x x

x

x

x x

0

x

x x

x

x

NCI-H125 A-549

x x

0 0

x x x x x x x

x x 0

x x x

x x x

x x

x x x

x x x x

L a r g e cell LCLC-97TM LCLC-103H U-1810

x x x x

0 0 0

x x x x x x x

0 x x x

x x x

x x x

x x x

x x x x

x x x x x x

Meso MSTO-2M-H

x x

0

x x x

0

x

x

x

x

x x

SCLC SCLC-21H

x x x x

0

0

0

0

x x

x x

x x x

x x x

n.d.

n.d.

n.d.

x x x

n.d.

n.d.

Epidermoid A431

x x x

n.d.

0, no expression; + to + + + + , low to high expression. N-myc-, c-myb-, c-fi}s-, v:/~s-, a n d a n y line. Meso, m e s o t h e l i o m a ; S C L C , small-cell lung cancer, A d e n o , a d e n o c a r c i n o m a

C-MYC

EXONI

H

Xh PROBES:

EXON2

EXON3

SS

S

Xb

3g

r

1

S

$

C

I C ,lkb

I R

v-fms-related

x x x

n.d.

x x

n.d.

t r a n s c r i p t s were not o b s e r v e d in

Fig. 3A, B. Analysis o f c-myc in N S C L C cell lines. D N A f r o m the t u m o r cells was digested with: A SstI, a n a l y z e d b y S o u t h e r n blotting a n d h y b r i d i z e d with the e x o n II SstI/SstI c-myc f r a g m e n t as i n d i c a t e d in the figure. Xb, XbaI; Xh, XhoI; R, EcoRI;; S, SstI; Bg, Bg/lI; H, HindIII; C, ClaI, B Digestion with EcoRI a n d then with either XbaI, Xhol or Bg/II. T h e s e blots were h y b r i d i z e d with the c-myc e x o n I I I ClaI/EcoRI p r o b e (see A). Restriction e n d o n u c l e a s e analysis revealed only characteristic, n o r m a l - s i z e d c-myc f r a g m e n t s in all N S C L C cell lines

34

EPLC - 32 M E PLC- 65 M U -1752

ADLC- 5 N 2 o z

NCI -H 23

(:3