Experimental Cell Research 159 (1985) 47-54

Biosynthesis of EGF Receptor, Transferrin Receptor and Colligin by Cultured Human Keratinocytes

and the Effect of Retinoic Acid

AMANDA TAYLOR, l' * BRIGID L. M. HOGAN t' * and FIONA M. WATT 2'**

1Mammalian Development Laboratory, Imperial Cancer Research Fund, Mill Hill Laboratories, London NW7 lAD, and 2Molecular Cell Biology Laboratory, Kennedy

Institute of Rheumatology, London W6 70W, UK

The biosynthesis of EGF and transferrin receptor by human keratinocytes in culture has been followed using specific monoclonal antibodies. In addition, keratinocytes are shown to synthesise a Mr 47000 protein that binds to gelatin-Sepharose. Peptide mapping confirms the identity of this protein with colligin, a newly described cell surface-associated glycoprotein that also binds to native collagens (Kurkinen et al., J biol chem 259 (1984) 5915) [9]. Vitamin A and its analogues have profound effects on the differentiation, morphology and motility of human keratinocytes in culture. We show here that retinoic acid (RA) has no effect on the growth rate of the cells or the synthesis of EGF receptor and colligin, but stimulates the synthesis of transferrin receptor. © 1985 Academic Press, Inc.

Cultures of human epidermal keratinocytes retain the essential features of the tissue from which they are derived. They grow as stratified colonies in which proliferation is confined to the basal layer and cells that leave it undergo terminal differentiation [1]. In the past, studies on these cells have concentrated on the expression of molecular markers of terminal differentiation, such as keratins [2-5] and involucrin, a precursor of the cornified envelope [6-8]. However, relatively little is known about the expression by cultured keratinocytes of molecules involved in cell proliferation and cell-substrate interactions. Here we report the biosynthesis by keratinocytes of the cell-surface receptors for epider- mal growth factor (EGF) and transferrin. In addition, we show that keratinocytes synthesise the recently described cell surface-associated glycoprotein, colligin [9]. This protein, which has a Mr of 47 000, binds to both gelatin and native collagens, including types I, III, IV and V (Taylor & Hogan, unpublished results). Because of its association with the cell surface it is possible that colligin plays a role in the interaction of cells with a collagenous extracellular matrix.

Vitamin A has profound effects on the pathway of epidermal differentiation both in vivo and in vitro (reviewed in [10]). For example, addition of retinyl

* Present address: Laboratory of Molecular Embryology, National Institute for Medical Research, Mill Hill, London NW7 IAA, UK. ** To whom offprint requests should be sent.

4-858337 Copyright ~ 1985 by Academic Press, Inc. All rights of reproduction in any form reserved

0014-4827/85 $03.00

48 Taylor, Hogan and Watt

ace t a t e to h u m a n e p i d e r m a l cel ls in cu l tu re not on ly in f luences the type o f

kera t ins they s y n t h e s i z e , bu t a lso the m o r p h o l o g y and mot i l i ty o f the co lon ie s [4].

The v i t amin A a n a l o g u e R A has a lso b e e n r e p o r t e d to inc rease the n u m b e r o f

avai lable E G F r e c e p t o r s on the su r face o f m o u s e e p i d e r m a l cel l l ines [11]. W e

have t h e r e f o r e i nves t i ga t ed the R A ef fec t on the b io syn the s i s o f E G F and

t rans fe r r in r e c e p t o r s and col l ig in by cu l t u r ed h u m a n k e r a t i n o c y t e s .

M A T E R I A L S A N D M E T H O D S

Cells and Culture Conditions

Human keratinocytes from newborn foreskin (strain a, fourth to ninth passage) were grown with a feeder layer of mitomycin C-treated 3T3 cells [12, 13]. The culture medium consisted of three parts Dulbecco's modified Eagle medium (DME) plus one part Ham's F 12, supplemented with 1.8 × 10-4 M adenine [14], 0.5 ~tg/ml hydrocortisone [12], 10 -1° M cholera toxin [15], 10 ng/ml EGF [16], 5 ~tg/ml insulin [7], and 10% fetal bovine serum. In some experiments, vitamin A was removed from the fetal bovine serum by solvent extraction (delipidized serum) [4, 17]. A stock solution of RA (all-trans, Sigma) was prepared in alcohol and protected from light. Dilutions were prepared in growth medium immediately before use. Medium was changed every 2 days.

Cultures were seeded at 1.5× 105 keratinocytes per 60 mm diameter Petri dish (Falcon Plastics, Oxnard, Calif.). For determining growth rates, the 3T3 feeder cells were selectively removed from triplicate dishes with EDTA before detaching the keratinocytes with trypsin/EDTA [13]. Human dermal fibroblasts (second to ninth passage) were derived from explants of newborn foreskin. The A431 human epidermoid cell line was obtained from Dr M. Waterfield, and the teratocarcinoma- derived parietal endoderm cell line, PYS, from Dr J. Lehman. All cells were grown in Dulbecco's modified Eagle medium with 10 % fetal bovine serum.

Labelling Conditions

Cultures were washed thoroughly with DME containing 1 ~tg/ml cold methionine and 10 % dialysed fetal bovine serum and then incubated in the same medium containi [35S]methionine (Amersham International) 1000 Ci/mmole.

Antibodies and Gelatin-Sepharose

The monoclonal antibody, R1, specifically recognizing human EGF receptor was kindly provided by Dr M. Waterfield [18]. A mouse monoclonal antibody with the same specificity against human transferrin receptor as OKT9 was kindly provided by Dr C. Schneider [19]. Both antibodies bind to protein A.OX12 is a mouse monoclonal antibody against rat IgG and was kindly provided by Dr A. Williams.

Gelatin (Sigma, type 1 300 bloom) was coupled to CNBr-activated Sepharose 4B according to the manufacturer's instructions [9].

Immunoprecipitation of Cell Extracts Cells were washed in phosphate-buffered saline (PBS) and extracted for 5 min at 4°C in 0.15 M

NaC1, 0.005 M EDTA, 0.05 M Tris pH 8, containing 1% NP40. The lysate was then centrifuged for 5 min at 12 000 g and the supernatant either used immediately or made 50 % v/v with glycerol and stored at -20°C. For immunoprecipitation, aliquots of the lysate were diluted into 1 ml 0.4 M NaCI, 0.005 M EDTA, 0.05 M Tris, pH 8, 1% NP40 and incubated for 1.5 h at room temperature with 5 ~tl of antibody. Immune complexes were isolated using protein-A agarose (Sigma). Details of all the procedures have been given [21]. Samples were analysed by SDS-PAGE under reducing conditions [20] and radioactive proteins visualized by fluorography using Kodak SB5 film [21]. The levels of newly synthesized transferrin receptor were quantitated by scanning the autoradiographs with a Zeineh soft laser densitometer and integrating the areas under the peaks.

Exp Cell Res 159 (1985)

Synthesis of cell-surface proteins by keratinocytes 49

Peptide Mapping Gelatin-binding proteins were separated on a 7.5 % SDS-polyacrylamide gel, which was dried down

and exposed to X-ray film without fixation. Localized Mr 47 000 bands were excised and incubated for 1 h at room temperature with 0.15 M n-chlorosuccinimide in urea as described elsewhere [22]. Under the conditions used the polypeptide is cleaved at tryptophan residues and, less efficiently, as cysteine residues [23].

RESULTS

Synthesis of EGF and Transferrin Receptor

Synthesis of EGF and transferrin receptors was followed using mouse mono- clonal antibodies specific for the two human proteins. Keratinocytes were la- belled for 6 h with [35S]methionine and immunoprecipitates of detergent extracts analysed by SDS-PAGE and autoradiography. Immunoprecipitation with the mouse monoclonal antibody, R1, against human EGF receptor recovered a major methionine-labelled polypeptide of Mr 175 000 (fig. 1, 1). This is the same size as the mature, fully glycosylated EGF receptor recovered with the same antibody from the human epidermoid cell line, A431 (fig. 1, 3, and ref. [24]). After 2 h labelling only the Mr 160000 polypeptide corresponding to the precursor of the intact receptor which lacks terminal sugars [24] is specifically immunoprecipitat- ed from keratinocytes (fig. 1, 7). These cells do not synthesize the polypeptide of Mr 90000 which is prominent in A431 cells labelled for short periods and represents a truncated external domain of the EGF receptor (fig. 1, 9) [24, 25]. At an intermediate labelling time (3.5 h) both the Mr 175 000 and 160 000 forms of the EGF receptor are recovered by R1 from extracts of keratinocytes (fig. 3, 1-3).

Monoclonal anti-transferrin receptor recovered a polypeptide of Mr 90 000 from keratinocytes labelled for 6 h with [35S]methionine (fig. 1, 2). This is the same size as the mature, fully glycosylated transferrin receptor synthesized by A431 cells (fig. 1, 4). Shorter labelling times resulted in the recovery also of a slightly smaller component corresponding to an immature form which lacks terminal sugars [19] (see fig. 3, 4-6).

Synthesis of Colligin

Keratinocytes were labelled with [35S]methionine, and proteins in the deter- gent extract which bind to gelatin-Sepharose were analysed by SDS-PAGE and autoradiography. The major labelled protein recovered by this procedure has a Mr 47 000 (fig. 2A, 4). This is the same molecular weight as the mature, fully glycosylated collagen-binding surface glycoprotein, colligin, recently identified in a variety of cell types, including the murine parietal endoderm cell line, PYS (fig. 2A, 5 and ref. [9]). Minor labelled proteins, including a set of Mr 56000-76000, are also recovered from keratinocytes. These co-migrate on SDS-polyacrylamide gels with previously characterized minor gelatin-binding proteins from PYS cells [9]. However, the relative amounts of these proteins recovered from keratino-

Exp Cell Res 159 (1985)

50 Taylor, Hogan and Watt

HK A431 MrxlC--31 2 3 4 5

175~ ~ I f

90~,- I ~ Q

119

HK

A B 1 2 3 4 5

A431 8 9 Mrx'lO -3

,B!l, < 1 6 0

.~47

kL.

1 2

Fig. t . Biosynthesis of EGF and transferrin receptors by human keratinocytes and A431 cells. Confluent cultures of human keratinocytes and A431 cells were labelled for 6 and 2 h (lanes 6-9) with 100 or 150 ~Ci/ml [35S]methionine respectively. Epidermal growth factor receptor (EGF-R) and transferrin receptor (T-R) were then immunoprecipitated from detergent extracts using monoclonal antibodies. Lane 1, Human keratinocytes, anti-EGF receptor; 2, human keratinocytes, anti-T recep- tor; 3, A431 cells, anti-EGF receptor; 4, A431 cells, anti-T receptor; 6, human keratinocytes, control monoclonal antiserum OXI2; 7, human keratinocytes, anti-EGF receptor; 8, A431 cells, control monoctonal antiserum; 9, A431 cells, anti-EGF receptor; 5, 14C-labelled molecular weight markers (Amersham Intemational). These are myosin heavy chain (Mr 212000) phosphorylase b (Mr 100000 and 92 500), bovine serum albumin (69 000), ovalbumin (43 000), carbonic anhydrase (Mr 30 000) and lysozyme (14300). Fig. 2. Biosynthesis of colligin by keratinocytes and other cell types. (A) Confluent cultures of human keratinocytes (Ker, approx. 5 × 106 cells), human dermal fibroblasts (Fib, 106 cells), and murine 3T3 (3×106 cells) and PYS cells (1.5× 106 cells) were labelled overnight with 66 ~tCi/ml of [3~S]methio- nine. Gelatin-binding proteins were recovered quantitatively from detergent extracts of the total cultures and analyzed on a 7.5 % SDS-polyacrylamide gel which was dried down without fixation and exposed to X-ray film. Lane 1, 14C-labelled Mr markers (see caption to fig. 1); 2, human fibroblasts; 3, murine 3T3 cells; 4, human keratinocytes; 5, murine PYS cells. (B) Partial peptide maps of colligin. The Mr 47000 bands in fig. (A) were excised and the protein cleaved by n-chlorosuccinimide as described [21]. Samples were analyzed on a 10-15% gel under reducing conditions. 1, Human keratinocytes; 2, murine PYS cells; 3, human dermal fibroblasts. 3T3 cells gave a pattern identical with that of PYS cells.

cytes varied somewhat (compare e.g., fig. 2A, 4 with fig. 4, 9) and they will not be considered further here. The identity of the Mr 47 000 protein synthesized by keratinocytes with colligin made by PYS cells was confirmed by one-dimensional peptide mapping of 35S-labelled material (fig. 2B). A Mr 47000 gelatin-binding protein is also synthesized by human dermal fibroblasts and the partial [35S]methionine-labelled peptide map of this protein is very similar to that of PYS cell colligin (fig. 2B, 2,3). Fig. 2 also shows that 3T3 fibroblasts synthesize colligin (fig. 2A, 3). However, it is very unlikely that the small number of feeder 3T3 cells remaining in the confluent human keratinocyte cultures could account for the level of colligin synthesis observed. Moreover control experiments showed colligin synthesis in keratinocytes passaged in the absence of feeder cells (data not shown).

Exp Cell Res 159 (1985)

Synthesis of cell-surface proteins by keratinocytes 51

Mr x 10-3 1

212~

EGF-R¢ I ~

100~.,- 92.5P"

691,-

2 3 4 5 6 M r x10-3., 1 2 3 4 5 6 7 8 -<2 t2 y 10 11 12

EGF-R~ .~ ~ ~

~ : IT -R

T-R~

43~,-

-- 3 ~, ~ to

-~47

Fig. 3. EGF and transferrin receptor synthesis in keratinocytes cultured for 1 week with and without RA. Keratinocytes were cultured in medium containing either 10% delipidized serum alone (DL, lanes 1, 4) or supplemented with RA to a final concentration of 2, 5, 5x 10 -7 M or 3, 6, 1 × 10 -6 M. After 7 days, when the cultures were confluent, the cells were labelled for 3.5 h with 100 lxCi/ml [35S]methionine. Aliquots of detergent extracts containing equal amounts of TCA-precipitable radio- activity were used for the recovery by immunoprecipitation of EGF receptor and transferrin receptor. Samples were analyzed on a 7.5 % gel under reducing conditions. Immunoprecipitation with 1-3, anti- EGF receptor; 4--6, anti-transferrin receptor. Fig. 4. EGF and transferrin receptor, and colligin synthesis in keratinocytes cultured for 2 weeks with and without RA. Keratinocytes were cultured in medium containing either 10 % control fetal bovine serum (C, lanes 1, 5, 9) or 10% delipidized serum alone (DL, 2, 6, 10) or supplemented with RA to a final concentration of 5x 10 -7 M (4, 8,11) or 10 -6 M (3, 7, 12). After 14 days the cultures were labelled with [35S]methionine for 4 h. Aliquots of detergent extracts containing equal amounts of TCA-precipitable radioactivity were used for the recovery by immunoprecipitation of 1-4, EGF receptor; 5--8, transferrin-receptor; 9-12, gelatin-binding proteins. Samples were analyzed either on an 8 % gel (1-8) or on a 5-10% gradient gel (9-12) under reducing conditions. The position of the Mr markers on both gels is shown.

RA Effect on Synthesis of EGF Receptor, Transferrin Receptor and Colligin by Human Keratinocytes

Human keratinocytes were cultured for 7 or 14 days in medium containing 10 % delipidized (vitamin A-free) fetal bovine serum, with or without 5× 10 - 7 o r 10 - 6

M RA. At the end of this time the newly confluent cells were labelled for 3.5 h with [35S]methionine and EGF receptor, transferrin receptor, and gelatin-binding proteins analyzed as described above. In no case was there a significant effect of culture conditions on the amount of newly synthesized EGF receptor and colligin recovered from the keratinocytes (figs 3, 4 and data not shown). However, cells incubated in delipidized serum did show some change in expression of minor gelatin-binding proteins at 14 days (fig. 4, 10).

In contrast to these results, cells supplemented with RA showed a higher level of newly synthesized transferrin receptor than controls. For example, in two

Exp Cell Res 159 (1985)

52 Taylor, Hogan and Watt

A B

30 t t 1-

Z

10 lO /

+i/ +++ + 03 * , ¢

1 2 3 4 5 6 7 El D |0 11 12 13 14 I 2 3 4 5 6 7 8 9 10 11 12 13 14 lg Days Days

Fig. 5. (A, B) Effect of 5× l0 -7 M RA on the population doubling time of human keratinocytes. In two separate experiments, cells were cultured in medium containing either O, i , 10% delipidized serum alone or ©, A, supplemented with RA to a final concentration of 5 × 10 -7 M. At intervals, ceils were labelled for 3.5 h with 150 ~tCi [35S]methionine per dish. Newly synthesized transferrin receptor recovered by immunoprecipitation was quantitated by autoradiography and densitometery. -~, Ratio of transferfin receptor levels in RA-treated (TR RA) to control cultures (TRCONTROL).

separate experiments after 14 days, the ratio of transferrin receptor synthesized by cells treated with 5x 10 -7 M RA compared with controls was 2.4:1 and 2.9:1 (fig. 4 and data not shown). This stimulation could be detected as early as 6 days and was not due to an effect of RA on keratinocyte proliferation (fig. 5 A, B).

DISCUSSION

We have shown here that human keratinocytes in culture synthesize receptors for both EGF and transferrin. The receptors appear to be processed by normal keratinocytes in the same way as the intact receptors of A431 cells and there is no evidence for synthesis of the truncated external domain of the EGF receptor which is made in large amounts by these transformed cells [24]. In mature rat skin in vivo, EGF receptor expression is predominantly on cells of the basal layer, with little or no expression in the fully differentiated cells [25-28]. Transferrin receptor, on the other hand, is expressed by all viable cells throughout the basal and stratified layers [29]. There is no data at present on receptor distribution in cultured keratinocytes.

We have also shown here that keratinocytes synthesize colligin, a Mr 47 000 glycoprotein that binds to gelatin (denatured type I and III collagen) and native type IV collagen [9]. It also binds to native type I and III collagen, and to pepsin-

Exp Cell Res 159 (1985)

Synthesis of cell-surface proteins by keratinocytes 53

extracted type V collagen (Taylor & Hogan, unpublished results). By virtue of its association with the cell surface and its ability to bind collagen, colligin may play a role in the attachment of keratinocytes to a collagenous substratum, both in vivo and in vitro. However, in the absence of specific antibodies to colligin it has not been possible to test this hypothesis directly, and the idea remains specula- tive [9].

One of the aims of this study was to see if vitamin A or its analogue, RA, had any effect on the synthesis of EGF and transferrin receptor and colligin by human keratinocytes. Although the retinoids have no effect on the growth rate of keratinocytes in culture ([4, 30] and fig. 5) they do profoundly affect both the differentiation and motility of the cells [4, 31]. Our finding that RA does not affect synthesis of the EGF receptor is consistent with the idea that vitamin A does not change the proportion of cells in the basal proliferative compartment, but acts on cells already committed to undergo terminal differentiation.

In the presence of retinoids, keratinocyte motility is increased, so that the cultures form whorls at confluence. In contrast, cultures in delipidized serum form blisters and show increased cell-cell adhesion. However, since RA has no effect on colligin synthesis, it is unlikely that this protein mediates the change in cell morphology and behaviour.

In the case of transferrin receptor, exposure to RA does significantly increase its rate of synthesis by keratinocyte cultures. In other systems, increased recep- tor density has been correlated with rapid proliferation (reviewed in [32]), but, as discussed above, the increase we have observed cannot be due to stimulation of proliferation. Instead, the results point to an enhanced requirement for transfer- rin in vitamin A-treated keratinocytes, e.g., for iron-dependent metabolic activi- ties unrelated to proliferation.

We thank Ian R. Kill for skilled technical assistance and Dr K. vonder Mark for generously supplying native collagens. This work was funded in part by the Arthritis and Rheumatism Council.

REFERENCES

1. Green, H, Harvey lect 74 (1980) 101. 2. Sun, T-T & Green, H, J biol chem 253 (1978) 2053. 3. Fuchs, E & Green, H, Cell 15 (1978) 887. 4. - - Ibid 25 (1981) 617. 5. Kubilus, J, McDonald, M J & Baden, H P, Biochim biophys acta 578 (1978) 484. 6. Rice, R H & Green, H, Cell 18 (1979) 681. 7. Watt, F M & Green, H, J cell biol 90 (1981) 738. 8. - - Nature 295 (1982) 434. 9. Kurkinen, M, Taylor, A, Garrels, J I & Hogan, B L M, J biol chem 259 (1984) 5915.

10. Lotan, R, Biochim biophys acta 605 (1980) 33. 11. Jetten, A M, Nature 284 (1980) 626. 12. Rheinwald, J G & Green, H, Cell 6 (1975) 331. 13. Rheinwald, J G, Methods cell biol 21A (1980) 229. 14. Wu, Y-J, Parker, L M, Binder, N E, Beckett, M A, Sinard, J H, Griffiths, C T & Rheinwald, J G,

Cell 31 (1982) 693.

Exp Cell Res 159 (1985)

54 Taylor, H o g a n a n d W a t t

15. Green, H, Cell 15 (1978) 801. 16. Rheinwald, J G & Green, H, Nature 265 (1977) 421. 17. Rothblat, G H, Arbogast, L Y & Howard, B V, In vitro 12 (1976) 554. 18. Waterfield, M D, Mayes, E L V, Stroobant, P, Bennett, P L P, Young, S, Goodfellow, P N,

Banting, G S & Ozanne, B, J cell biochem 20 (1982) 149. 19. Schneider, C, Sutherland, R, Newman, R & Greaves, J, J biol chem 257 (1982) 8516. 20. Laemmli, U K, Nature 227 (1970) 680. 21. Cooper, A R, Kurkinen, M, Taylor, A & Hogan, B L M, Eurj biochem 119 (1981) 189. 22. Lischwe, M A & Ochs, D, Anal biochem 127 (1982) 453. 23. Lane, E B, Hogan, B L M, Kurkinen, M & Garrels, J I, Nature 303 (1983) 701. 24. Mayes, E L V & Waterfield, M D, EMBOj 3 (1984) 531. 25. Downward, J, Yarden, Y, Mayes, E, Scrace, G, Totty, N, StockweU, P, Ullrich, A, Schlessinger,

J & Waterfield, M D, Nature 307 (1984) 521. 26. Green, M R, Basketter, D A, Couchman, J R & Rees, D A, Dev biol 100 (1983) 506. 27. Gusterson, B, Cowley, G, Smith, J A & Ozanne, B, Cell biol int rep 8 (1984) 649. 28. Nanney, L B, McKanna, J A, Stoscheck, C M, Carpenter, G & King, L E, J invest dermatol 82

(1984) 165. 29. Green, M R & Couchman, J R, J embryol exp morphol 82 suppl. (1984) 39. 30. Gilfix, B M & Green, H, J cell physiol 119 (1984) 172. 31. Green, H & Watt, F M, Mol cell biol 2 (1982) 1115. 32. Newman, R, Schneider, C, Sutherland, R, Vodinelich, L & Greaves, M, Trends in biochem sci 7

(1982) 397.

Received October 24, 1984 Revised version received February 20, 1985

Exp Cell Res 159 (1985) Printed in Sweden