(CANCER RESEARCH 48, 2724-2729, May 15, 1988]

Keratin Subtypes of Small Cell Lung CancerAnthony D. Elias,1 Bruce F. Cohen, and Samuel D. Bernal2

Division of Medicine, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115

ABSTRACT

The molecular forms of keratin in small cell lung cancer (SCLC) celllines and tumors were examined with antikeratin monoclonal antibodies.Immunostaining of SCLC by antikeratin antibody and examination byfluorescence microscopy indicates population heterogeneity in keratincontent. Intensity of immunostaining is often weak. However, polyacryl-

amide gel electrophoresis and immunoblotting reproducibly demonstratethe presence of keratin and allow analysis of the keratin subtypes. Thefinding of keratin subtypes closely associated with the development ofkeratinizing epithelium (the 68 kDa basic keratin) in SCLC was unexpected in a tumor that is regarded as poorly differentiated. The cytoskel-

etal composition of SCLC suggests the presence of a heterogeneouspopulation with a significant proportion of cells expressing highly differentiated epithelial properties.

INTRODUCTION

The keratins are a complex set of at least 19 related proteinswhich form the intermediate filaments of epithelial cells. Although their functions within the cell are not well understood,these proteins seem to be expressed in specific combinationsrelated to the cell type and its state of differentiation (1-7).Different isoelectric points and gene structures characterize thetwo families of keratins, acidic and basic. All keratins containa constant central «-helical rod leading to their filamentousstructure and variable regions at both ends which presumablyallow specific functions within the cell. Coordinate expressionof pairs of keratins has been observed; within each keratin pairas defined by frequent coexpression, the basic keratin is usuallyabout 8 kDa larger than its paired acidic counterpart (4).

In the past few years, it has been recognized that in mostcases carcinoma cells contain similar patterns of keratin subtypes as their normal cellular counterparts. This suggests thatanalysis of keratin content may help delineate the relationshipsbetween the lung carcinomas and the highly complex lowerrespiratory epithelium (8, 9). We have previously reported thepresence of keratins in SCLC,3 analyzed by two-dimensional

gel electrophoresis (10). In this report, we describe a characteristic set of keratins in SCLC cell lines and tumors by immunostaining and immunoblotting with antikeratin antibodies. Manyof the SCLC keratins were also found in NBE cells and inNSCLC, suggesting a close relationship between these differentcell types.

MATERIALS AND METHODS

Antikeratin Antibodies. The murine anti-human keratin monoclonalantibodies AIM, AE-2, and AE-3 have been previously characterized(11,12) and were kindly donated by Dr. Tung-Tien Sun. AE-1 has beenshown to react with all of the acidic family of keratins (keratins 10-19)[Moll classification (1)], except 45/46 kDa (keratin 17/18). AE-2 reacts

Received 6/17/87; revised 12/29/87; accepted 1/15/88.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1American Cancer Society Physician's Training Research Fellow.2Supported by Grant ÇA4SS28 from the National Cancer Institute.'The abbreviations used are: SCLC, small cell lung cancer; NBE, normal

bronchial epithelial; NSCLC, non-small cell lung cancers; PBS, phosphate buttered saline; SDS, sodium dodecyl sulfate; 2-ME, 2-mercaptoethanol; TBS, 50mM Tris (pH 7.4)-0.15 M NaCl buffer.

with the 56.5 kDa (keratin 10) and 65-67 kDa (keratins 1-2) keratinsthat may be regarded as markers of skin-type differentiation (keratini-zation). AE-3 reacts with the basic family of keratins (keratins 1-8).The immunoreactivity of keratin 64 kDa (keratin 9) with AE-1, AE-2,and AE-3 is unknown. A mouse monoclonal IgG antibody (287) un-reactive with carcinoma cells was used as a negative control.

Cell Lines and Tumor Specimens. All cell lines were grown in RPMI1640 with 2 HIMglutam ine and 10% calf serum. The variant cell line(low dopa-decarboxylade, few neurosecretory granules, loose suspension growth) SW2-S was established from metastatic tumor in the bonemarrow of a patient with SCLC. The attached subline, SW2-A, wasisolated and subcloned from tissue cultures of SW2-S after 2 years inculture. The classic SCLC line (high dopa-decarboxylase, tight colonymorphology, and many neurosecretory granules) OH-1 has been previously characterized (13). The lung adenocarcinoma cell lines A549and CAÕ,U-3,and the lung squamous cell carcinoma CALU-1 werepurchased from the American Type Culture Collection. The classicSCLC cell line NCI-H69 was obtained from Dr. A. Gazdar, NationalCancer Institute, Bethesda, MD (14). The undifferentiated lung carcinoma LX-1, derived from a tumor with both small cell and squamouscell carcinoma characteristics (15), was provided by Dr. A. E. Bogden,Mason Research Institute, Worcester, MA. The lung squamous cellcarcinomas, HOTZ and U1752, were provided by Dr. P. Groscurth,University of Zurich, Switzerland, and Dr. J. Bergli, Uppsala, Sweden,respectively (16, 17). Other cell lines were supplied by investigators atthe Dana-Farber Cancer Institute: squamous cell carcinomas of thehead and neck SCC-15, SCC-25, SCC-68, SCC-78 (Dr. J. Rheinwald);breast carcinoma MCF-7 (Dr. D. Kufe); melanoma G361 (Dr. M.Wick); and lymphoblastic leukemia CEM (Dr. H. Lazarus). A2182, anattached cell line with SCLC characteristics, was kindly donated byCentocor, Inc. (Dr. S. Mai). Other cell lines were developed in thislaboratory.

Tumor tissues were either extracted within l h of collection or wereflash-frozen in liquid nitrogen and stored until used. Four surgicalspecimens were obtained from Dr. C. O'Hara at New England Deacon

ess Hospital. The SCLC tissue shown in Fig. 2 and Table 1 (SCC) wasa lymph node metastasis from a patient with oat cell histology. SW2-Scells were passaged in nude mice and extracts were made of the solidtumor nodules (SW2-M2). Normal bronchial epithelial cells weregrown from expiants by Dr. K. Weinberg (18).

Indirect Immunofluorescence Staining. Suspension cells or cells grownon glass coverslips were washed with phosphate-buffered saline, pH7.4, twice and then fixed with methanol at -20°Cfor 3 min. The fixed

cells were washed with distilled water, then PBS, and overlaid withprimary antibody for 45 min at 37"C (AE-1, 1:10; AE-2, 1:5; and AE-3, 1:10). After washing with PBS, flunrest-ciliated goat anti-mouse IgG

1:50 (Cappel Laboratories) was incubated for 45 min. After washing inPBS and distilled water, the cells were mounted in 10% PBS, 90%glycerol, 100 mg/dl /vphenylenediamine, pH 8.0 (19), and photographed under a Zeiss epifluorescence microscope with the use of TRI-X ASA 400 Kodak film.

Keratin Extraction. Using a procedure modified from Wu et al. (20),cells were washed with PBS and resuspended in 5 ml of high salt buffer(20 mM Tris-HCl, pH 7.4; 0.6 M KC1; 1% Triton X-100; 1 mMphenylmethylsulfonyl fluoride; 1 mM ethyleneglycol tetraacetate;1 mM EDTA) at 4°C.After homogenization in a Wheaton homog-

enizer, the insoluble fraction was pelleted by centrifugal ion at 10,000x g for 15 min, then incubated with nuclease solution [DNAse I (Sigma Chemical Co.) 100 units/ml-RNAse A (Sigma) 0.05 mg/ml inPBS] for 30-60 min at 4°C.The pellet was then homogenized in

high salt buffer and centrifuged. Finally the pellet was washed in PBSand stored at —70°C.In some experiments, cells were metabolicallylabeled with 100 uC\/ml ["SJmethionine for 3 h at 37°Cprior to

extraction with high salt buffer.2724

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

KERATIN SUBTYPES OF SCLC

Gel Electrophoresis and Western Blotting. For one-dimensional gelelectrophoresis, each lane contained high salt extract from 2x10"

cells. These extracts were dissolved in one-dimensional loading buffer(20 HIMTris-HCl, pH 7.4; 5% SDS; 5% 2-ME) at 100'C for 5 min and

clarified by centrifugal ion at 10,000 x g for 5 min. The proteins wereseparated electrophoretically on a SDS-polyacrylamide (12.9%) slabgel by the Laemmli technique (21). High molecular weight proteinstandards were obtained from Bethesda Research Laboratories, Be-thesda, MD. The gels were equilibrated in transblot buffer (20 HIMTris-67 m,\i glycine in 20% methanol) for 30 min, then sandwiched withnitrocellulose paper and run on a transphor electrophoresis unit at 70V for 3 h at 4"C. The molecular weight lanes were removed and stained

with amido black. After incubation with TBS and 1% Triton X-100 toremove SDS, nonspecific binding sites on the paper were blocked byincubation for 2 h with 5% bovine serum albumin and 3% normal goatserum in TBS. The nitrocellulose strips were incubated with primaryantibody in 5% bovine serum albumin and 3% normal goat serum (AE-1 or AE-3, 1:200; AE-2, 1:100) for 16 h at 4°C.After washing with

TBS buffer, peroxidase-conjugated rabbit anti-mouse immunoglobulin(RAMpox; Cappel) was added and incubated for 2 h at 20°C.Afterrewashing, filtered 3,3'-diaminobenzidine hydrochloride (Sigma) 1 mg/

ml in TBS with 0.05% H2O2was added to develop the brown peroxidasereaction for 3-5 min before quenching with water. Standard molecularweight protein markers in adjacent lanes were used to determinemolecular weights of individual bands. Immunoblots were also performed with the use of the anti-40 kDa keratin antibody described andkindly supplied by Wu and Rheinwald (22).

Two-Dimensional Gel Electrophoresis. Isoelectric focusing was performed according to the method of O'Farrell (23). A high salt extractprepared from [35S]methionine metabolically labeled cells was dissolvedin two-dimensional lysis buffer [9.5 M urea; 2% 2-ME (Sigma); 25 HIMTris, pH 7.4; and 2% LKB ampholytes (pH 4-6.5:pH 5-8:pH 3.5-9.5;1:1:0.33)] and clarified by centrifugation at 10,000 x g for 10 min. Thehigh salt extract from 5 x IO5cells was loaded per tube gel and run at400 V for 16 h, then boosted to 1000 V for 2-3 h. After a 20-minincubation with SDS equilibration buffer [125 mM Tris (pH 6.8)-1 %SDS-10% glycerol-5% 2-ME], the second dimension was run overnightat 8 mA/gel. The gels were fixed in methanol:acetic acid:lM)(7.5:7.5:85), placed in Enhance (New England Nuclear) for 30 min anddried. Autoradiograms were exposed for 1-3 weeks.

RESULTS

Immunofluorescence. Immunofluorescence staining of SCLCcells with antikeratin monoclonal antibody was generally weakand diffuse. AE-2 did not reveal significant immunofluores-cence staining of SCLC cells compared with a control mousemonoclonal antibody. Heterogeneous staining was observed inthe different SCLC cell lines (including classical, variant, andattached sublines), using AE-1 and AE-3 antikeratin monoclonal antibodies as well as rabbit antiserum raised against totalkeratin (courtesy of T-T. Sun). The staining of SCLC cells byAE-1 (Fig. 1, left and middle), AE-3, and antikeratin antiserumwas weak and diffuse in contrast to the strong filamentouscytoplasmic staining generally observed with NSCLC cells byusing these antibodies. Unlike the suspension SCLC cells, theattached subline SW2-A had greater amounts of cytoplasm,and showed more intense filamentous keratin immunofluores-cence with AE-1 (Fig. 1, right).

Gel Electrophoresis and Western Blotting. Unlike immuno-fluorescence staining of intact cells, the presence of keratin wasmore easily and reliably demonstrated by immunoblots of highsalt extracts. SCLC cells and tumors expressed a pattern ofkeratin subtypes that was characteristic of this histology. Themajor keratin subtypes observed by immunoblotting of smallcell lung cancer extracts were 40 kDa, 52.5 kDa, and 68 kDacorresponding to keratins 19, 8, and 1. Minor bands withmolecular weights of 48 kDa, 50 kDa, 54 kDa, 59 kDa, 63

kDa, and 65.5 kDa (keratins 2-4 and 13-16) were also observed. These findings are summarized in Table 1 and shownin Fig. 2. The 68 kDa antigen reactive with AE-2 and AE-3 isexpressed in most SCLC tumors by immunoblot (Fig. 2), variably in NBE cultures, and in trace amounts in squamous cellcarcinomas, but not seen in other lung histologies (data notshown). The classic SCLC cell lines OH-1 and NCI-H69 contain relatively large amounts of this keratin, whereas only asmall amount is detected in variant cell line SW2-S. The 59kDa keratin was present in OH-1 and in SCLC tissue specimensbut not in the other two cell lines. The 54 kDa keratin was notdetectable in SW2-S cultivated in vitro, even though it wasfound in low quantities in SW2-S grown in vivo. The 40 kDaAE-1-reactive keratin has been found to react with the anti-40kDa monoclonal antibody of Wu and Rheinwald (22) (data notshown). SCLC tissue specimens, whether fresh tumor fromlymph node biopsy or xenografts in nude mice, revealed similarkeratin patterns compared with SCLC cell lines. SW2-A contains all of the keratin species found in SW2-S. However, SW2-A appears to contain higher amounts of keratin per mg proteincompared to SW2-S. Similarly, the attached subline 69A contains the same subtypes as its parent cell line NCI-H69, butalso shows increased keratins. The attached sublines, SW2-Aand 69A, contain a 56.5 kDa keratin, not detected by immunoblot in their parent cell lines, but observed in NBE cells. Themajor antikeratin-reactive proteins of 69A, 46 kDa and 49 kDa(AE-3), are also detected in the majority of adenocarcinomas(data not shown). All the keratin species found in SCLC arealso detected in NBE cells. NBE cells express the major antigens 38 kDa [previously described in esophageal epithelium(24)], 40 kDa, 54 kDa, 58 kDa, and 59 kDa, with minor bandsat 48 kDa, 50 kDa, 52.5 kDa, 56.5 kDa, 63 kDa, and 65.6 kDa(Fig. 2; Table 1). Immunoblot bands outside the usual molecularweights for keratins are observed, but are quite variable andmay represent protein aggregates or proteolytic products.

Two-Dimensional Gel Electrophoresis. The keratin subtypesin SCLC cells were also analyzed by two-dimensional gel electrophoresis of [35S]methionine-labeled extracts (Fig. 3, Table2). The classic SCLC cell line OH-1 contained keratins 65.5/68 kDa (keratin 1/2), 59 and 54 kDa (keratins 4 and 13), 58and 50/50' kDa (keratins 5 and 14/15), and 52.5 and 40 kDa

(keratins 8 and 19) by two-dimensional gels. Trace amounts of63 kDa (keratin 3) and 64 kDa (keratin 9) are observed withmetabolic labeling, but could not be distinguished by one-dimensional immunoblot. The 58 kDa keratin is detected bytwo-dimensional gel but appears as a broad band not readilyresolved from the 59 kDa band on one-dimensional blots.Comparison of the results of the two-dimensional gels withone-dimensional immunoblots and the known antibody reactivities of AE-1, AE-2, and AE-3 (Table 2) allowed correlationwith the keratin classification system of Moll et al.(\) and Sunet al. (4). Two-dimensional immunoblots using AE-1 correlatewell with these assignments (3 bands are observed with NCI-H69: 43 kDa, pi 5.4, 3+; 51 kDa, pi 5.1, 1-2+; and 56 kDa,pi 5.3, 1+; data not shown).

DISCUSSION

Small cell lung cancer cells contain a characteristic set ofkeratins. The 40/52.5 kDa pair, the 50, 54/58, 59 kDa pairs,and the 65.5, 68 kDa proteins constitute the major keratinsexpressed in SCLC. This keratin pattern has been detected inSCLC cells and tissues and in classic and variant SCLC histologies, but not in NSCLC or non-lung tumors, although some

2725

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

KERATIN SUBTYPES OF SCLC

Fig. 1. AE-1 immunofluorescence; Left, SW2-S, a variant SCLC cell line; middle, OH-1, a classic SCLC cell line; right, SW2-10A, an attached subline of SW2-S.Immunofluorescence with AE-3 is identical and immunostaining with AE-2 is negative.

Table 1 Keratin analysis of SCLC and normal bronchial epithelium by immunoblot

Molecularwt (kDa)

Classic

OH-1 NC1-H69VariantSW2-S

Attached Tissue

SW2-A 69-A SCC SW2-M2NBEcells

AERxn°

Mollclass*

lia\ic keratins

6865.563595852.54946

+var 2,32,32, (3)2,33

3"3"

Acidic keratins

56.554+++50+ ++(+)48+ ++40++ ++++38+++

+ +++1,2+(+) (+) ++++1(+)

(+) (+) ++1+(+) ++1+++++ -t- (+) +++1++

1"101314/1516194

" AE Rxn, reactivity with each antikeratin monoclonal antibody, 1* Moll class, keratin number by Moll's classification (1).' (+), (3) weak.* Previously unreported reactivities.

AE-1; 2 = AE-2; 3 = AE-3.

individual keratins are shared. In particular, SCLC expressesthe high molecular weight keratins not detected in large cell oradenocarcinomas of the lung. Conversely, in SCLC, we did notobserve the 56 kDa keratin detected in squamous or large cellcarcinoma, nor the 38 kDa keratin detected in adenocarcinomasof the lung."

Some investigators were not able to detect keratin by immunostaining of SCLC (25, 26), but others have detectedkeratin in a few SCLC specimens (10, 27-35). Our results showthat even when immunostaining is negative with an antikeratinantibody such as AE-2, keratins can be reliably identified bybiochemical techniques. The scanty cytoplasm in SCLC addsto the difficulty in observing cytoplasmic components by immunostaining. When staining is observed, it is often diffuse anddoes not show discrete filaments. This may be related to therounded morphology of SCLC cells. By electron microscopy,the low amount of cytokeratin in SCLC tends to be poorlyorganized into tonofilaments (29). Similar difficulty in demonstrating microtubular networks was encountered in some

4 Manuscript in preparation.

transformed cells because of their rounded shape and poorlyspread cytoplasm (36).

The keratin subtypes we have identified in SCLC by immu-noblotting include those associated with simple, stratified, andeven terminally differentiated squamous epithelia. Certain patterns of keratin expression have been associated with differentstates of epithelial differentiation (1, 37). The three lowestmolecular weight acidic (40, 45, 46 kDa) and the two lowestmolecular weight basic keratins (52.5 and 54 kDa) were identified in simple epithelia and the basal layer of skin (7, 20, 35,37). The 48 and 56 kDa keratins were found in hyperprolifer-ative keratinocytes (38). Nonkeratinized stratified epitheliumand corneal epithelium contained the pairs 54/59 kDa, and 55/63 kDa, respectively (39, 40). The highest molecular weightpairs, 56.5/65.5 and 68 kDa, are found in terminally differentiated stratified squamous epithelium (4, 11, 39, 41).

The finding of keratin subtypes related to keratinizing epithelium (the 65.5-68 kDa basic keratins reactive with AE-2and AE-3) in SCLC was unexpected in a carcinoma which is

regarded as so poorly differentiated. These keratins have beenclosely associated with the process of cornification and terminal

2726

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

OH-1 SW-2 SW-SA H-69

KERATIN SUBTYPES OF SCLC

SCO NBE

68

65. »-64

86.6

84.080.04B.O

40.O

52.5 ^-54

T50' 50

40

6 5

OH-l 8W-2 8W-2A H-89 SCO HBB

68.0-65.5-63.0-

86.8-

OH-1 SW-8 SW-2A H-69 SCO NBE

68.0-08.6-63.0-89.0-

82.8-

Fig. 2. A, AE-2 immunoblot of SCLC and NBE; B, AE-2 immunoblot ofSCLC and NBE; C, AE-3 immunoblot of SCLC and NBE.

Fig. 3. Two-dimensional gel electrophoresis of OH-l ["Sjmethionine-labeledhigh salt extract.

Table 2 Two-dimensional gel electrophoresis of OH-l: correlation with keratinclassification

Molecularwt(kDa)6865.5635958565452.56456.556555450'5048464540pirange7.8-7.07.8-7.47.5-7.27.3-6.47.4-6.67.8-7.06.0-5.86.1-5.95.4-5.25.3-5.15.3-5.14.9-4.55.1-4.85.3-5.14.95.1-4.95.1-4.95.7-5.55.2-4.9AERx"2,32,32,32,3333311,21,2NoneNone1Mollclass*12345678910111213141516171819OH-l

molecular wt

(kDa)6865.563595852.5645450'5040pi7.87.77.26.17.66.15.45.35.34.95.3Amount3+3+<+)'1+2+<+)2+2+2+3+(+)

" AE Rx, reactivity with each antikeratin antibody, I = AE-1; 2 = AE-2; 3 =AE-3.

0 Moll class, keratin number by Moll's classification (1).<(+),weak.

differentiation of stratified squamous cells (4, 11, 39, 41),including the development of a granular and a cornified layerin skin and with "pearl formation" in well differentiated squa

mous cell carcinomas. However, several other investigatorshave found similar keratin content in other respiratory epithelium by a variety of techniques and reagents. Emura et al. (42)grew fetal trachéalcells in culture and found markedly positiveimmunofluorescence with antibodies raised against human stratum corneum and with antimerokeratin antibody reactive withsuperficial epidermal keratins between 60 and 70 kDa withoutobserving squamous differentiation. The basal, ciliated, andintermediate cells in normal bronchial epithelium have reactedwith specific antisera to stratum corneum keratins including 68kDa by immunohistochemical staining (25, 33). In these studies, 2/16 SCLC specimens were also reactive. The detection ofthe high molecular weight keratins in SCLC described in thispaper is unlikely to represent artifacts produced by the potentially ubiquitous stratum corneum of experimenters' hands.

2727

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

KERATIN SUBTYPES OF SCLC

Control cell lines such as G361, a melanoma cell line, or CEM,a I cull lymphoblastic leukemia line, did not demonstrate significant reactivity with antikeratin antibodies (neither did immunoblots with irrevelant mouse anti-human IgG as primary

antibody). These negative controls are important for keratinstudies since cell extracts may be contaminated with stratumcorneum present in laboratory equipment (6).

The observation that SCLC expressed a subset of the keratinsfound in normal bronchial epithelial cells is consistent with apulmonary epithelial lineage for SCLC. Normal bronchial epithelial cells contain all the keratins expressed by SCLC, but inaddition exhibit keratins that are observed in NSCLC histologies. The 68 kDa keratin can also be detected in NBE cells,particularly after differentiation is induced by high calcium invitro* However, NBE cells contain other keratins not found in

SCLC such as the 38 kDa band. These lower molecular weightkeratins were also detected in NSCLC cell lines. Blobel et al.(8) have also detected the 40,45/52.5 kDa, the 46/54 kDa, andthe 50-58 kDa pairs by using 2 dimensional nonequilibriumpH gel electrophoresis in NBE from surgical specimens. Thepresence of keratins associated with simple epithelia, stratifiedepithelia, and differentiated epithelia in NBE implies an abilityof the progenitor cell(s) of bronchial epithelium to differentiatealong multiple pathways, expressing different pairs of keratinsduring the process of maturation. This conclusion is supportedby the heterogeneity in keratin content detected by immunoflu-orescence.

The presence of keratin, a fundamental marker of epithelialdifferentiation, casts serious doubt on the interpretation thatSCLC has a monocytic origin simply due to the presence ofsurface antigens shared with monocytes (9). Many monoclonalantibodies, including those recognizing natural killer cells andmonocyte "differentiation" antigens cross-react not only with

SCLC, but also with NSCLC and a variety of tissues such asneural or hematopoietic histology (43). Analysis of intermediatefilament composition may be more useful in determining tissuederivation since it had been shown that cancer cells generallyexpress the same distinct intermediate filament type as theirnormal tissue counterparts. Thus, normal and neoplastic epithelial cells express keratin. Neuroendocrine tissues and tumorscontain neurofilament proteins. Cells of hematopoietic originexpress vimentin, but not keratin. We have previously analyzedthe proteins associated with the cytoskeleton of human hematopoietic cells (44). In addition to vimentin, 10 other proteinswere identified in monocytes which were also present in thecytoskeleton of other hematopoietic cells, including peripheralblood granulocytes and myeloid leukemic cells. None of theseproteins were shared with SCLC or other lung cancers. RecentlyZauli et al. (45) have observed immunofluorescent staining ofkeratin and vimentin in the K562 cell line which was originallyderived from a patient with chronic myelogenous leukemia inblast crisis. About 30% of the cells were diffusely stained byAE-1 but all were negative for AE-3. It would be important toperform immunoblots of keratin extracts to rule out nonspecificstaining, especially since it is unlikely for cells to contain onlyacidic keratins (recognized by AE-1) but not basic keratins(recognized by AE-3).

Our results with antikeratin immunoblots and immunofluo-rescence suggest that SCLC exists as a heterogeneous population with the major proportion expressing highly differentiatedepithelial antigens. A minor population may include a relativelyundifferentiated and multipotent stem cell which can progressalong various pathways of differentiation during continuoustissue culture or during in vivo selection. The ability to express

a certain keratin subtype may indicate a commitment toward acertain program of differentiation. The potential differentiationpathways of stem cells in SCLC tumors may include squamousand adenocarcinomatous differentiation in addition to the neu-roendocrine SCLC. After long term tissue culture (more than2 years), SCLC cell lines develop variants that resemble non-small cell histology (46). Further evidence for this model stemsfrom the biological behavior of SCLC in vivo. Mixed histologies, small cell and non-small cell, are recognized relativelyfrequently, either accompanying the initial presentation, detected in residual disease after appropriate therapy for smallcell, or seen at autopsy (29, 46-51). The most common histo-logical mixture contains large cell or undifferentiated carcinoma with SCLC, but adenocarcinoma or squamous cell mixtures also occur. Further differentiation of the tumor cells ofSCLC to NSCLC elements has been observed to occur in vivo(52). These observations have led to proposals that small cellcarcinoma is derived from an endoderma! stem cell that iscapable of differentiating along multiple pathways (51 ).

Identification of the molecular forms of keratins may behelpful to analyze the changes that occur during SCLC differentiation. Development of molecular probes specific for individual keratins may identify unique populations within a SCLCtumor and population shifts caused by alteration of growthconditions or addition of differentiation agents such as dibu-tyryl-cAMP or a-interferon. Correlation with other potentialmarkers of SCLC differentiation (neurosecretory granules, surface membrane antigens, morphology, and decarboxylase enzymes) may allow definition of the differentiation pathways ofSCLC and its relationship to the normal bronchial epithelium.

ACKNOWLEDGMENTS

Many thanks to Dr. Tung-Tien Sun for his generous gift of antikeratin antibodies and his invaluable criticism and careful reading of themanuscript. I wish to thank Jane Mayotte for her technical help.

REFERENCES

1. Moll, R., Franke, W. W., Schiller, D. L., Geiger, B., and Krepier, R. Thecatalog of human cytokeratins: patterns of expression in normal epithelia,tumors, and cultured cells. Cell, .*/: 11-24, 1982.

2. Franke, W. W., Schmid, E., and Moll, R. The intermediate filament cytoskeleton in tissues and in cultured cells: differentiation specificity of expression of cell architectural elements. In: C. C. Harris and H. N. Am rup (eds.),Human Carcinogenesis, Chap. 1, pp. 3-34. New York: Academic Press, Inc.,1983.

3. Osborn, M.. and Weber, K. Tumor diagnosis by intermediate filament typing:a novel tool for surgical pathology. Lab. Invest., 48: 372-394, 1983.

4. Sun, T-T., Eichner, R., Schermer, A., Cooper, D., Nelson, W. G., and Weiss,R. A. Classification, expression, and possible mechanisms of evolution ofmammalian epithelial keratins: a unifying model. Cancer Cells, 1: 169-176,1984.

5. Bernal, S. D., and Stahel, R. S. Cytoskeleton-associated proteins: their roleas cellular integrators in the neoplastic process. CRC Crii. Rev. Oncol.Hematol., 3: 191-204, 1984.

6. Cooper, D., Schermer, A., and Sun, T-T. Classification of human epitheliaand their neoplasms using monoclonal antibodies to keratins: strategies,applications, and limitations. Lab. Invest., 52:243-2S6, 198S.

7. Franke, W. W., Schiller, D. L., Moll, R., et al. Diversity of cytokeratins:differentiation specific expression of cytokeratin polypeptides in epithelialcells and tissues. J. Mol. Biol., 153:933-959, 1981.

8. Blobel, G. A., Moll, R., Franke, W. W., and Vogt-Moykopf, I. Cytokeratinsin normal lung and lung carcinomas. 1. Adenocarcinomas, squamous cellcarcinomas and cultured cell lines. Virchows Arch. B Cell Pathol., 45: 407-429, 1984.

9. Ruff, M. R., and Pert, C. B. Small cell carcinoma of the lung: macrophage-specific antigens suggest hemopoietic stem cell origin. Science (Wash. DC),225: 1034-1036, 1984.

10. Bernai, S. D., Baylin, S. B., Shaper, J. H„Gazdar, A. F., and Chen, L. B.Cytoskeletal-associated proteins of human lung cancer cells. Cancer Res., 43:1798-1808,1983.

11. Woodcock-Mitchell, J., Eichner, R., Nelson, W. G., and Sun, T-T. Immu-

2728

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

KERATIN SUBTYPES OF SCLC

nolocalization of keratin polypeptides in human epidermis using monoclonalantibodies. J. Cell Biol., 95: 580-588, 1982.

12. Eichner, R., Bonitz, P., and Sun, T-T. Classification of epidermal keratinsaccording to their immunoreactivity, isoelectric point, and mode of expression. J. Cell Biol., 98:1388-1396, 1984.

13. Francis, J., Thompson, R. A., Bernal, S. D., et al. Effect of dibutyryl cyclicadenosine 3':5'-monophosphate on the growth of cultured human small celllung carcinoma and specific cellular activity of L-Dopa decarboxylase. CancerRes., 43: 639-645, 1983.

14. Gadzar, A. F., Carney, D. N., Rüssel,E. K., et al. Establishment of continuous, clonable cultures of small-cell carcinoma of the lung which have amineprecursor uptake and decarboxylation cell properties. Cancer Res., 40:3502-3507, 1980.15. Giovanella, B. ( '., Stehlin, J. S., Williams, L. J., et al. Heterotransplantationof human cancers into nude mice. Cancer (I'llila.), 42: 2269-2281, 1978.

16. Peter, H. J., and Kistler, G. S. Isolation und in vitro Kult ivat ¡onvonTumorzellen aus Menschlichen in der nu/nu Maus transplantierten Broii-chuskarzinomen. Schweiz. Med. Wochenschr., 109:830-832, 1979.

17. Bergh, J., Nilsson, K., Zech, L., and Giovanella, B. Establishment andcharacterization of a continuous lung squamous cell carcinoma cell line (U-1752). Anticancer Res., /: 317-322, 1981.

18. Bemal, S., Weinberg, K., Kakefuda, M., Stahel, R., O'Hara, C, and Wong,

Y. C. Membrane antigens of normal bronchial epithelial cells identified bymonoclonal antibodies. In Vitro Cell. Dev. Biol., in press, 1988.

19. Johnson, G. IX. Araujo, G. M. de C. N. A simple method of reducing thefading of immunofluoresccnce during microscopy. J. 1nun uno I. Methods, 43:349-350, 1981.

20. Wu, Y-J., Parker, L. M., Binder, N. E., et al. The mesothelial keratins: anew family of cytoskeletal proteins identified in cultured mesothelial cellsand nonkeratinizing epithelia. Cell, 31:693-703, 1982.

21. Laemmli, U. K. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature (Lond.), 277.-680-685, 1970.

22. Wu, Y-J., and Rheinwald, J. G. A new small (40 kD) keratin filament proteinmade by some cultured human squamous cell carcinomas. Cell, 25:627-635,1981.

23. O'Farrell, P. H. High resolution two-dimensional electrophoresis of proteins.J. Biol. Chem., 250:4007-4021, 1975.

24. Banks-Schlegel, S., and Harris, C. C. Tissue specific expression of keratinproteins in human esophageal and epidermal epithelium and their culturedkeratinocytes. Exp. Cell Res., 146: 271-280, 1983.

25. Bejui-Thivolet, F., Viac, J., Thivolet, J., and Paure, M. Intracellular keratinsin normal and pathological bronchial mucosa. Virchows Arch. Pathol. Anat.¿95:87-98, 1982.

26. Lehto, V-P., Stenman, S., Miettinen, M., et al. Expression of a neural typeof intermediate filament as a distinguishing feature between oat cell carcinoma and other lung cancers. Am. J. Pathol., 110: 113-118, 1983.

27. Broers, J. L. V., Carney, D. N., De Ley, L., Vooijs, G. P., and Ramaekers,F. C. S. Differential expression of intermediate filament proteins distinguishes classic from variant small-cell lung cancer cell lines. Proc. Nati. Acad.Sci. USA, 82: 4409-4413, 1985.

28. Sappine, A-P., Ellison, M. L., and Gusterson, B. A. Immunohistochemicallocalisation of keratin in small cell carcinoma of the lung: correlation withresponse to combination chemotherapy. Eur. J. Cancer Clin. Oncol., 19:1365-1370,1983.

29. McDowell, E. M., McLaughlin, J. S., Merenyl, D. K., et al. The respiratoryepithelium. V. Histogenesis of lung carcinomas in the human. J. Nati. CancerInst., 61: 587-606, 1978.

30. Van Muijen, G. N. P., Ruiter, D. J., Van Leeuwen, C., et al. Cytokeratinand neurofilament in lung carcinomas. Am. J. Pathol., 116: 363-369, 1984.

31. Said, J. W., Nash, G., Tepper, G., and Banks-Schlegel, S. Keratin proteinsand carcinoembryonic antigen in lung carcinoma: an immunoperoxidasestudy of fifty-four cases, with ultrastructural correlations. Hum. Pathol., 14:70-76, 1982.

32. Said, J. W., Nash, G., Banks-Schlegel, S., et al. Keratin in human lung

tumors: patterns of localization of different-molecular-weight keratin proteins. Am. J. Pathol., 113: 27-32, 1983.

33. Gusterson, B., Mitchell, D., Warburton, M., and Sloane, J. Immunohistochemical localisation of keratin in human lung tumours. Virchows Arch.Pathol. Anat., 394: 269-277, 1982.

34. Banks-Schlegel, S. P., Gazdar, A. F., and Harris, C. C. Intermediate filamentand cross-linked envelope expression in human lung tumor cell lines. CancerRes.,«: 1187-1197, 1985.

35. Banks-Schlegel, S. P., McDowell, E. M., Wilson, T. S., Trump, B. F., andHarris, C. C. Keratin proteins in human lung carcinomas: combined use ofmorphology, keratin immunocytochemistry, and keratin immunoprecipita-tion. Am. J. Pathol., 114: 273-286, 1984.

36. Osborn, M., and Weber. K. The display of microtubules in transformed cells.Cell, /2.-561, 1975.

37. Debus, E., Weber, K., and Osborn, M. Monoclonal cytokeratin antibodiesthat distinguish simple from stratified squamous epithelia: characterizationon human tissues. EMBO J., ;.- 1641-1647, 1982.

38. Weiss, R. A., Eichner, R., and Sun, T-T. Monoclonal antibody analysis ofkeratin expression in epidermal diseases: a 48- and 56-kdalton keratin asmolecular markers for hyperproliferative keratinocytes. J. Cell Biol., 98:1397-1406, 1984.

39. Tseng, S. C. G., Jarvinen, M. J., Nelson, W. G., et al. Correlation of specifickeratins with different types of epithelial differentiation: monoclonal antibody studies. Cell, 30: 361-372, 1982.

40. Nelson, W. G., Battifora, H., Santana, H., and Sun, T-T. Specific keratinsas molecular markers for neoplasms with a stratified epithelial origin. CancerRes., 44:1600-1603, 1984.

41. Sun, T-T., and Green, H. Keratin filaments of cultured human epidermalcells: formation of intermolecular disulfide bonds during terminal differentiation. J. Biol. Chem., 253: 2053-2060, 1978.

42. Emura, M., Mohr, U., Kakunaga, T., and Hilfrích,J. Growth inhibition andtransformation of a human fetal trachéalepithelial cell line by long-termexposure to diethylnitrosamine. Carcinogenesis (Lond.), 6:1079-1085,1985.

43. International Workshop on Small Cell Lung Cancer Antigens: London, 1987.Lung Cancer, 4, in press, 1988.

44. Bernal, S.D., and Chen, L. B. Induction of cytoskeleton-associated proteinsduring differentiation of human myeloid leukemic cell lines. Cancer Res., 42:5106-5116, 1982.

45. Zauli, D., Gobbi, M., Crespi, C., et al. Vimentin and keratin intermediatefilaments expression by K562 leukemic celi line. I euk Res., 10:29-33,1986.

46. Gazdar, A. F., Carney, D. N., Guccion, J. G., and Baylin, S. B. Small cellcarcinoma of the lung: cellular origin and relationship to other pulmonarytumors. In: F. A. Greco, R. K. Oldham, and P. A. Bunn (eds.), Small CellLung Cancer, pp. 145-175. New York: Gruñe& Stratton, Inc., 1981.

47. Gazdar, A. F., Cohen, M. H., Idhe, D. C., et al. Bronchial epithelial changesin association with small cell carcinoma of the lung. In: F. Muggia and M.Rozencweig (eds.). Lung Cancer: Progress in Therapeutic Research, pp. 167-174. New York: Raven Press, 1979.

48. Abeloff, M. D., Eggleston, J. C., Mendelsohn, G., et al. Changes in morphologic and biochemical characteristics of small cell carcinoma of the lung: aclinicopathologic study. Am. J. Med., 66: 757-764, 1979.

49. Brereton, H. D., Mathews, M. M., Costa, J., et al. Mixed anaplastic small-cell and squamous-cell carcinoma of the lung. Ann. Int. Med., 88:805-806,1978.

50. Radice, P. A., Matthews, M. J., Ihde, D. C. et al. The clinical behavior of"mixed" small cell/large cell bronchogenic carcinoma compared to "pure"small cell subtypes. Cancer (Phila.), 50:2894-2902, 1982.

51. Matthews, M. J. Effects of therapy on the morphology and behavior of smallcell carcinoma of the lung—a clinicopathologic study. In: F. Muggia and M.Rozencweig (eds.), Lung Cancer: Progress in Therapeutic Research, pp. 155-165. New York: Raven Press, 1979.

52. Bepler, G., Jaques, G., 1Invernanti. K., ci al. Characterization of two celllines with distinct phenotypes established from a patient with small cell lungcancer. Cancer Res., 47:1883-1891, 1987.

2729

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

1988;48:2724-2729. Cancer Res   Anthony D. Elias, Bruce F. Cohen and Samuel D. Bernal  Keratin Subtypes of Small Cell Lung Cancer

  Updated version

  http://cancerres.aacrjournals.org/content/48/10/2724

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/48/10/2724To request permission to re-use all or part of this article, use this link

on March 11, 2021. © 1988 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from