J. Cell Sci. 86, 83-93 (1986) 83Printed in Great Britain © The Company of Biologists Limited 1986

CHARACTERIZATION OF A SPONTANEOUSLY

TRANSFORMED PULMONARY EMBRYONIC RAT

(PER) EPITHELIAL CELL LINE

M. PAYE1, CH. ETIEVANT1, F. MICHIELS2, D. PIERARD1,B. NUSGENS1 AND CH. M. LAPIERE1'*Laboratoire de Dermatologie Experimental, Universite de Liege, CHU du Sart-Tilman,

B23, B-4000 Sart-Tilman, par Liege 1, Belgium and 2Institut de RechercltesScientifiques sur le Cancer, CNRS, Villejuif, France

SUMMARY

A spontaneously transformed pulmonary embryonic rat epithelial cell line (PER) is described interms of growth, tumorigenicity, growth factor responsiveness and biosynthetic capacity. At low-passage subcultures, PER cells grew as a monolayer and did not form colonies in soft agar. Afterlong-term subcultivation, they lost contact inhibition, became anchorage-independent and formedtumours in nude mice. Low concentrations of foetal calf serum permit the maximum growth rate.The multiplication and metabolic activity, assessed by 2-deoxy-D-glucose uptake, was significantlystimulated by growth factors. PER cells synthesized collagen types I, III, IV and V, laminin andfibronectin, and organized a pericellular matrix made up of only basement membrane components(type IV collagen and laminin) and fibronectin. These data enabled us to define PER cells as atransformed epithelial cell line evolving towards malignancy with long-term subcultivation. Thesecells appeared to be a valuable tool in studies of tumour cell-matrix interactions and regulation ofgrowth factor receptors in tumorigenesis.

INTRODUCTION

The regulation and/or the maintenance of cell differentiation operated by thesupporting extracellular matrix or the basement membrane might be significant inthe preservation of physiological characteristics. This interaction might depend oninformation received by the cells through membrane receptors recognizing extra-cellular components such as fibronectin (Grinnell & Feld, 1979; Takashima &Grinnel, 1984), collagen type I (Klebe, 1974; Burke et al. 1983) and laminin(Terranova et al. 1980; Couchman et al. 1983).

Experimental models in vitro and in vivo provide support for the importance ofthis interaction in the process of tumour development, invasion and metastasis thatmight be significant in the expression of a transformed phenotype (Gospodarowicz,1983). For example, Michiels et al. (1981) failed to observe malignant transform-ation developing in embryonic lung explants exposed to carcinogens while thedissociation of the cells from the explants resulted in the establishment of a trans-formed tumorigenic cell line. A similar cell line was obtained without treatment with

•Author for correspondence.

Key words: epithelial lung cells, transformation, culture.

84 M. Paye and others

a carcinogen just by dissociation of lung explants (Michiels et al. 1981). These cellsproved to be most interesting in terms of cell-matrix relationship (Paye & Lapiere,1986) as well as in the alteration of the turnover of epidermal growth factor (EGF)receptors (Nishiyama et al. unpublished data).

We describe here the growth properties of such cells in vitm, their biosyntheticcapacity in terms of collagen, basement membrane components and attachmentglycoproteins, as well as their metabolic and mitotic response upon stimulation withdefined growth factors.

MATERIALS AND METHODS

MaterialsReagents were obtained from the following sources: 3,5-diamino-benzoic acid (DABA), fibro-

blast growth factor (FGF), epidermal growth factor (EGF) from Sigma; [methvl-3Vi]thymidme(6-7Cimmor ' ) , 2-deoxy-D-[l-3H]glucose (8-3 Cimmol"1), [5-3H]proline (lOCimmol"1) fromNew England Nuclear. Culture medium was purchased from Gibco and foetal calf serum fromGibco and Flow Laboratories. Tissue culture plates were obtained from Nunc. Anti-rabbitfluorescein-conjugated goat immunoglobulin G (IgG) and anti-guinea-pig rhodamine-conjugatedgoat IgG were purchased from Dako.

Cell culture

Pulmonary embryonic rat (PER) epithelial cells were obtained by dissociation of lung ex-plants from 15-day-old Wistar rat embryos after organ culture for 16 days according to thetechnique described by Michiels et al. (1981). These cells are known to undergo transformationspontaneously upon prolonged subculture.

The cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM) with 10%(v/v) foetal calf serum (FCS), glutamine (292mgr ' ) , ascorbic acid (50|/gml~'), penicillin-streptomycin (100 units ml~') at 37°C in a controlled atmosphere containing 5 % CO2. They werepropagated by brief treatment with 0-1 % trypsin and 0-02% EDTA solution at pH 7-4 by a 1:3split ratio every 5 days. Unless stated otherwise, the cells were used between passages 27 and 45.Normal bovine skin fibroblasts used for comparison were cultivated and propagated using the sameconditions.

DA'A measurement

The DNA content was measured by the fluorimetnc technique described by Johnson-Wint &Hollis (1982) using recrystallized DABA reagent. Briefly, cell layers were fixed for 24h at 4°Cin 10 % formaldehyde in 0' 1 M-borate buffer, pH 8, washed with 15 % isopropanol solution, driedfor 30min at 50°C. DABA reagent was added and allowed to react for 45 min at 60°C. Afterstabilizing the DABA-DNA reaction by adding 6M-HC1, the fluorescence of the reaction mediumwas measured in a spectrofluorimeter at an excitation wavelength of 420 nm and an emissionwavelength of 510nm.

Measurement of the mitogenic responseThe cells (7-5X 104) were plated in 30 mm dishes in DMEM containing 10% FCS. Immediately

after attachment, they were made quiescent by incubation in DMEM containing 0-05% FCS.After 24 h, the quiescence medium was replaced with DMEM containing FCS at increasingconcentrations or purified growth factors. Twenty-four hours later, [3H]thymidine (7-5 /iCi, 1 /AM)was added for 18 h. After washing three times with cold phosphate-buffered saline (PBS), the cellswere lysed in lml of 0-5M-NaOH for 4h at 37°C. DNA was precipitated by addition of cold

Transformed lung epithelial cell line 85

trichloracetic acid (30 % final concentration) and harvested by filtration under reduced pressureon Whatman GFA filters and repeatedly washed. The radioactivity retained on the filters wasmeasured by liquid scintillation.

Determination of 2-deoxy-D-glucose (2-DOG) transportCells were plated on 24-well tissue culture plates (Nunc) and allowed to reach confluence in

DMEM with 10% FCS. One day after confluence the cells were maintained overnight in aquiescence medium (DMEM containing 0 1 % FCS) before being preincubated for 5h in thequiescence medium containing increasing concentrations of dialysed FCS or purified growthfactors as indicated. The cells were washed with PBS buffer at 37°C and incubated for 15 min in1 ml of PBS containing the same concentration of growth factors or dialysed FCS as above. Themeasurement of 2-DOG transport was initiated by adding 1-5 jiCi of [ H]2-DOG to each well.After 20 min of incubation, cells were washed three times with 1 ml of cold PBS and lysed at roomtemperature in 0'4ml of 0'2% sodium dodecyl sulphate (SDS) solution in 0-2M-NaOH. Radio-activity of the lysate was measured by liquid scintillation.

Anchorage-independent growth assayAnchorage-independent growth assay of the cultured cells at passages 38 and 75 was performed

in soft agar according to the technique described by Courtenay (1976).

Characterization of the biosynthetic productsFour days after reaching confluence, cultures of PER cells were labelled in the maintenance

medium freshly supplemented with ascorbic acid (50j/gml~') and containing beta-aminopropio-nitrile (50/igmP1) and [5-3H]proline (10/iCiml"1). After a 24-h period of labelling, the mediumwas removed and the cells were washed with physiological saline, scraped in 0-05 M-Tris- HC1,pH7-2, 0-15M-NaCl, and briefly sonicated. Culture medium and cell homogenate were sup-plemented with a mixture of protease inhibitors (final concentration 20mM-EDTA, 0-5 mM-A'-ethylmaleimide and 0-1 mM-phenylmethane sulphonyl fluoride) and precipitated by addingammonium sulphate to 40% saturation. After 18 h at 4°C, the precipitated material was collectedby centrifugation and solubilized in O'lM-acetic acid. Samples of the labelled material weredigested with pepsin for 18 h at 4CC. Native and pepsin-digested material were submitted topolyacrylamide slab gel electrophoresis in 6-25 % separation gel under non-reducing and reducingconditions, according to the technique of Laemmli (1970) slightly modified by the addition of4M-urea in the separation gel. The labelled bands were revealed by fluorography (Bonner &Laskey, 1974) and exposure to Royal SO-mat Kodak film.

Immunofluorescence studiesCells were seeded on glass coverslips in bacteriological dishes and allowed to reach confluence.

After washing with PBS, they were air-dried and fixed in acetone at —20°C for 10 min. Indirectimmunofluorescence studies were performed according to the technique described by Foidart &Reddi (1980). Antibodies, raised in rabbit or guinea-pig, were directed against type I and pN-IIIcollagen purified from bovine foetal skin (Pierard et al. 1984), type IV collagen extracted fromhuman placenta (Van Cauwenberge et al. 1983), fibronectin isolated from human plasma andlaminin from EHS tumour (Nusgens et al. 1986). The specificity of the antibodies was tested byELISA as described by Rennard et al. (1980) and by immunoblotting using the technique ofTowbin et al. (1979). The cytokeratin filaments were revealed by a monoclonal antibody fromSanbio (Utrecht, Holland). Control preparations were performed using non-immune serum fromrabbit, guinea-pig or mouse. The preparations revealed by goat anti-rabbit fluorescein- or anti-guinea-pig rhodamin-conjugated IgG were observed in a Leitz ultraviolet microscope with epi-illumination and photographed.

86 M. Paye and others

RESULTS

Morphology and transformed phenotype

At low cell density (Fig. 1A), PER cells displayed a roundish and irregular shapeand a large amount of cytoplasm containing inclusions that did not stain withPapanicolaou coloration. A small number of cells (7 %) had an elongated fibroblasticshape. At high cell density, the cells had a hexagonal aspect with a small amount ofcytoplasm (Fig. IB). Until passage 50, cells grew as a monolayer without over-lapping. At higher passages, contact inhibition at confluence was lost and multilayersformed (Fig. 1C). The cytoplasm was very small and no inclusion was visible.

At passage 38, the PER cells do not grow in soft agar (Fig. 2A) whereas afterlong-term subcultivation (75 passages) cells were clearly anchorage-independent andformed visible colonies in soft agar 9 days after cloning (Fig. 2B).

Responsiveness to FCS and growth factors

At a concentration of FCS of 10% in DMEM the doubling time of PER cells is23 h. The incorporation of thymidine by sparse culture of PER cells for increasingconcentrations of FCS is illustrated in Fig. 3. The stimulation of cellular multipli-cation was dose-dependent up to 1 % FCS. At this concentration, the uptakeof thymidine was maximal and levelled off at higher FCS concentrations. Thenutritional requirements of normal skin fibroblasts taken for comparative purposewere significantly higher since the plateau of maximum uptake was not reached evenat 5 % FCS. EGF significantly stimulated the DNA synthesis of the PER cells whilealmost no response was observed upon addition of FGF up to lOOngml"1 (Fig. 4) inconfluent culture. The response of fibroblasts to these growth factors was the reverseof that observed in the PER cells, strong activation by FGF and mild stimulation byEGF.

The responsiveness to these growth factors in terms of metabolic activity wasassessed by the uptake of 2-DOG (Table 1). 2-DOG transport was stimulated,with a dose-dependent response, by FCS. Both EGF and FGF also induced astimulation, with the maximum response at lOngml"1.

Characterization of the biosynthetic products of the PER cells

The polypeptides synthesized in vitro after radioactive labelling with [ H]prolinewere characterized by SDS-polyacrylamide slab gel electrophoresis (Fig. 5). Theradiolabelled collagen polypeptides were found mainly in the form of soluble pre-cursors in the culture medium. After pepsin digestion (Fig. 5, lanes b and d) theyproved to be collagen type I (75 %), type III (25 %) and type V (traces). A labelledband of 220x10 MT co-migrating with authentic fibronectin and susceptible topepsin was also observed in the culture medium (Fig. 5, lane c).

By indirect immunofluorescence performed on the cell layer, a well-developedextracellular network of collagen type IV (Fig. 6A), laminin (Fig. 6B) and fibro-nectin (Fig. 6C) was observed. No type I or type III collagen could be detected in

Transformed lung epithelial cell line 87

Fig. 1. Phase-contrast micrographs of PER cells: A, at low density (x340); B, atconfluency and passage 27 (X170); and C, growing in multilayers at passage 75 (X 170).

Hf M. Paye and others

the matrix although fine granular intracytoplasmic deposits of pN-I and pN-IIIcollagen were visible (data not shown). The cytoplasm of all cells contained a well-developed network of cytokeratin filaments (Fig. 6D).

Fig. 2. Phase-contrast micrographs of PER cells seeded in soft agar (104 cells) andmaintained in culture for 9 days. A. PER cells at passage 38; and B, PER cells at passage75 (X100).

FCS (%)

Fig. 3. [3H]thymidine incorporation (ctsmin"1 per dish) into DNA of PER cells (O)and normal skin fibroblasts (A) seeded at low density (7-5X104 cells per 3cm dish)and made quiescent in 0-05% FCS for 24 h. After stimulation by increasing concen-tration of FCS for 24 h, [we</iy/-3H]thymidine was added for 18 h and the radioactivityincorporated into DNA was measured as described in Materials and Methods.

Transformed lung epithelial cell line 89

Growth factor (ngml"

Fig. 4. [3H]thymidine incorporation (ctsmin ', X 104 per dish) into DNA of confluentPER cells supplemented with increasing concentrations (1—100 ngml"1) of EGF (•) andFGF (D).

Table 1. Effect ofFCS, EGF and FGF on 2-DOG transport by PER cells

Control (0-1 %FCS)+ FCS0-5%

1%2%5%

10%+ E G F 2 n g m r '

lOngml"1

+ F G F 2 n g m r 'lOngml"1

Disintsmin ',Xl0-3/l05 cells

6-78-28-6

10-112-014-822-526-916-519-4

% of control

100122128151179221245292179211

Confluent PER cells made quiescent by incubation overnight in culture medium containing0-1 % FCS were stimulated by increasing concentrations of FCS or purified EGF and FGF. Theuptake of 2-DOG was performed as described in Materials and Methods. Results are expressedin disintsmin" , Xl0~3/l05 cells, and are the average of duplicate determinations. The averagevariation is less than 10%.

DISCUSSION

This study was performed to define the morphological characteristics, growthproperties, nutritional requirements, biosynthetic activity and metabolic response togrowth factors of an established pulmonary embryonic rat (PER) epithelial cell line.These characteristics will be discussed in relation to some specific properties of thesecells that might be related to their transformed phenotype and described in detail inthe accompanying paper (Paye & Lapiere, 1986) and elsewhere (Nishiyama et al.unpublished data).

As often reported for transformed cells (Sanford, 1965; San et al. 1979; Exilieet al. 1980; Mukherjie/ al. 1984), the PER cells display an irregular, roundish shape

M. Paye and others

-220

Fig. 5. Fluorogram of 3H-labelled polypeptides precipitated from the culture medium ofPER cells by ammonium sulphate at 40 % saturation. The electrophoresis was performedin a 6-25 % acrylamide gel under non-reducing (lanes a and b) and reducing (lanes cand d) conditions, before (lanes a and c) and after (lanes b and d) pepsin digestion.

at low density. Contact inhibition of growth is observed at low passage and lost athigh passage where they form foci of multilayered cells. The malignant potential issupported by the development of anchorage-independent growth and the formationof tumours after inoculation in syngeneic rats (Michiels et al. 1981, and unpublisheddata).

We have demonstrated the ability of PER cells to live in a synthetic mediumcontaining a low concentration of foetal calf serum while maximum growth is reachedwith a foetal calf serum concentration as low as 1 %. Multiplication is stimulated bylow concentrations of EGF and less significantly by FGF. The metabolic response ofPER cells, estimated by 2-DOG transport, is triggered by foetal calf serum, EGFand FGF. As observed in several transformed cell lines (Oshiro & Dipaolo, 1974;San et al. 1979; Siddigi & Lype, 1975), the basal level of sugar transport is aboutthree times higher than that expressed by normal skin fibroblasts and C3H10T 1/2(unpublished observations).

PER cells display in vitro biosynthetic activities compatible with their epithelialorigin as suggested by their content of cytokeratin filaments. A well-developedextracellular network of type IV collagen, laminin and fibronectin supports theircapacity to synthesize and to assemble basement membrane components. Labelling

Transformed lung epithelial cell line H

experiments show that they also synthesize interstitial collagen type I and type IIIand traces of collagen type V as observed in several other lines of epithelial cells(Keski-Oja et al. 1982; Alitalo et al. 1982). The PER cells were, however, unable topolymerize these collagens in the extracellular matrix, perhaps due to a lack ofprocollagen peptidase activity and/or to a lack of appropriate receptors on the cellsurface.

All together these data strongly suggest that PER cells are transformed epithelialcells evolving towards malignant transformation with long-term subcultivation. Theabsence of Papanicolaou positive granules in the cytoplasm, a feature of type IIalveolar cells (Kikkawa & Yoneda, 1974) that is, however, rapidly lost in culture,prevents us from confirming that PER cells are a line of pneumocytes II.

Fig. 6. Indirect immunofluorescent staining of the extracellular matrix of confluent PERcells using antiserum against: A, type IV collagen (X4S0); B, lamiriin (X450); andC, fibronectin (X450). Subconfluent PER cells were stained with a monoclonal anti-keratin antibody (D) and with non-immune mouse serum (E); X1200.

92 M. Paye and others

This cell line represents an experimental model that is useful in the study of theregulatory activity of growth factors in tumorigenesis since, unlike A431, theirmultiplication is stimulated by EGF. Moreover, the down-regulation of the EGFreceptors of the PER cells is unusual as 50% of the receptors remain at the cellsurface after contact with EGF (Nishiyama et al. unpublished data). A similarfeature has been recently described in normal rat kidney (NRK) fibroblasts exposedto TGF-beta (Assoian & Sporn, 1984).

PER cells also demonstrate defective attachment properties to interstitial collagensthat can be corrected by fibronectin modified by a transglutaminase (activatedFXIII) of blood coagulation. This provides an interesting in vitro assay for in-vestigating the requirements of tumour cells for adhesion to an extracellular matrix(Paye & Lapiere, accompanying paper).

We acknowledge the excellent technical assistance of Mrs Y. Goebels and the typographical helpof Mrs I. Leclercq. This work has been supported in part by a grant no. 3.4512.85 from the FRSMof Belgium and a grant from the Fonds CanceYologique de la Caisse Ge'ne'rale d'Epargne et deRetraite.

REFERENCES

ALITALO, K., KESKI-OJA, J., HEDMAN, K. & VAHERI, A. (1982). Loss of different matrixcomponents of rat cells transformed with a T-class ts mutant of rous sarcoma virus. Virology 119,347-357.

ASSOIAN, R. K. & SPORN, M. B. (1984). Cell surface EGF receptors are regulated by transforminggrowth factor beta. Cell. Biochem. Suppl. 8A, 256.

BONNER, W. M. &LASKEY, A. B. (1974). A film detection method for tritium labelled proteins andnucleic acid in polyacrylamide gels. Eur.J. Biochem. 46, 83-88.

BURKE, D., BROWN, M. J. & JACOBSON, B. S. (1983). HeLa cell adhesion to microcarriers in highshear conditions: evidence for membrane receptors for collagen but not laminin or fibronectin.Tissue® Cell 15, 181-191.

COUCHMAN, J. R., HOOK, M., REES, D. A. & TIMPL, R. (1983). Adhesion, growth, and matrixproduction by fibroblasts on laminin substrates. J. CellBiol. 96, 177-183.

COURTENAY, V. D. (1976). A soft agar colony assay for Lewis lung tumor and B16 melanoma takendirectly from the mouse. Br.J. Cancer 34, 39-45.

EXILIE, M. F., TURC, C , CHESSEBEUF, M. & PADIEU, P. (1980). Non-random chromosomechanges in nine rat liver epithelial cell lines. Cell Biol. Int. Rep. 4, 459-469.

FoiDART, J. M. & REDDI, A. M. (1980). Immunofluorescent localization of type IV collagen andlaminin during endochondral bone differentiation and regulation by pituitary growth hormone.Devi Biol. 75, 130-136.

GOSPODAROWICZ, D. (1983). The control of mammalian cell proliferation by growth factors,basement lamina, and lipoproteins. J. invest. Derm. 81, 40s-50s.

GRINNELL, F. & FELD, M. K. (1979). Initial adhesion of human fibroblasts in serum-free medium:possible role of secreted fibronectin. Cell 17, 117-129.

JOHNSON-WINT, B. & HOLLIS, S. (1982). A rapid in situ deoxyribonucleic acid assay fordetermining cell number in culture and tissue. Analyt. Biochem. 122, 338-344.

KESKI-OJA, J., GAHMBERG, C. G. & ALITALO, K. (1982). Pericellular matrix and cell surfaceglycoproteins of virus-transformed mouse epithelial cells. Cancer Res. 42, 1147-1153.

KIKKAWA, Y. & YONEDA, K. (1974). The type II epithelial cell of the lung. I. Method of isolation.Lab. Invest. 30, 117-129.

KLEBE, R. J. (1974). Isolation of a collagen-dependent cell attachment factor. Nature, Land. 250,248-251.

LAEMMU, U. K. (1970). Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature, Land. 227, 680-685.

Transformed lung epithelial cell line 93

MICHIELS, F., DUVERGER, A. & CHOUROULINKOV, I. (1981). Correlation between preneoplasticlesions in rat embryo lung treated with B(a)P or CSC in organ culture and tumor productionin vivo. Carcinogenesis 2, 885-896.

MUKHERJI, B., MACALISTER, T. J., GUHA, A., GILLIES, C. G., JEFFERS, D. C. & SLOCUM, S. K.

(1984). Spontaneous in vitro transformation of human fibroblasts. J. natn. Cancer Inst. 73,583-593.

NUSGENS, B., DELAIN, D., SENECHAL, H., WINAND, R., LAPIERE, CH. M. & WAHRMANN, J. P.

(1986). Metabolic changes in the extracellular matrix during differentiation of myoblasts of theL6 line and of a Myo-non-fusing mutant. Expl Cell Res. 162, 51-62.

OSHIRO, Y. & DIPAOLO, J. A. (1974). Changes in the uptake of 2-deoxy-D-glucosc in BALB/3T3cells chemically transformed in culture.X cell. Physiol. 83, 193-202.

PAYE, M. & LAPIERE, CH. M. (1986). Induction of attachment of transformed embryonic lungepithelial cells to collagen I by fibronectin and FXIIIa.J . Cell Sci. 86, 95-107.

PIERARD, D., NUSGENS, B. & LAPIERE, CH. M. (1984). Radioimmunoassay for the amino-terminalsequences of type III procollagen in human body fluids measuring fragmented precursorsequences. Analyt. Biochem. 141, 127-136.

RENNARD, S. I., BERG, R., MARTIN, G. R., FOIDART, J. M. & GHERON-ROBEY, P. (1980).

Enzyme-linked immunoassay (ELISA) for connective tissue components. Analyt. Biochem. 104,205-214.

SAN, R. H. C , LASPIA, M. F., SOIFFER, A. I., MASLANSKY, C. J., RICE, J. M. & WILLIAMS, G. M.

(1979). A survey of growth in soft agar and cell surface properties as markers for transforming inadult rat liver epithelial-like cell cultures. Cancer Res. 39, 1026-1034.

SANFORD, K. K. (1965). Malignant transformation of cells in vitro. Int. Rev. Cytol. 18, 249-307.SIDDIGI, M. & LVPE, P. T. (1975). Studies on the uptake'of 2-deoxy-D-glucose in normal and

malignant rat epithelial liver cells in culture. Int.jf. Cancer 15, 773-780.TAKASHIMA, A. & GRINNELL, F. (1984). Human keratinocyte adhesion and phagocytosis promoted

by fibronectin. J . invest. Derm. 83, 352-358.TERRANOVA, V. P., ROHRBACH, D. H. & MARTIN, G. R. (1980). Role of laminin in the attachment

of PAM212 (epithelial) cells to basement membrane collagen. Cell 22, 719-726.TOWBIN, H., STAEHELIN, T. & GORDON, J. (1979). Electrophoretic transfer of proteins from

polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. natn. Acad.Sci. U.SA. 76, 4350-4354.

VAN CAUWENBERGE, D., PIERARD, G. E., FOIDART, J. M. & LAPIERE, CH. M. (1983). Immuno-

histochemical localization of laminin, type IV and type V collagen in basal cell carcinoma. Br.jf.Derm. 108, 163-170.

(Received 22 April 1986 -Accepted 26 June 1986)