Keratin Biosynthesis in Normal Mouse Epithelia and in Squamous Cell Carcinomas mRNA-DEPENDENT ALTERATIONS OF THE PRIMARY STRUCTURE OF DISTINCT KERATIN SUBUNITS IN TUMORS*

(Received for publication, December 14, 1982)

Jiirgen Schweizer and Hermelita Winter From the German Cancer Research Center, Institute of Experimental Pathology, 6900 Heidelberg, Federal Republic of Germany

The keratin polypeptide patterns of two murine transplantable squamous cell carcinomas-originally induced by chemical means in the back skin and in the forestomach epithelium-are deficient in high molec- ular weight keratin subunits (>60 kDa) invariably present in the corresponding normal tissues. In addi- tion, the keratin polypeptide composition within the low molecular weight range showed further alterations with regard to the corresponding keratin subset of normal tissues in that both tumors expressed a 40-kDa protein, and a 56-kDa protein was selectively found in the forestomach tumor.

A comparison of the charge properties of normal and tumor keratin polypeptides revealed that the two up- permost tumor proteins at 60 and 58 kDa were basic in nature whereas their normal molecular weight coun- terparts belonged to the acidic subset of the pattern. These tumor proteins also showed mutually identical peptide maps which, however, were considerably dif- ferent from those of the normal proteins. The remain- ing tumor keratin subunits at 52, 50, 48, and 45 kDa, common also to the normal tissues, had retained their normal charge properties. In vitro translation of mRNA, isolated from both

normal and tumor tissue, revealed that every tumor keratin polypeptide is encoded by its own mRNA. In contrast to normal keratinizing tissues, there is there- fore no indication of post-translational protein proc- essing in tumors. The in vitro translation products of tumor RNAs had all properties in common with the in vivo tumor proteins, thus indicating that every devia- tion of the tumor keratin spectrum from the normal state is determined at the mRNA level.

Keratins, the constituent proteins of intermediate filaments (tonofilaments), represent by far the most abundant differ- entiation products of mammalian epidermis. Electrophoret- ically they can be resolved into a protein family of 40-70 kDa. The complexity of the protein patterns varies within species and in relation to anatomical regions (1-6). Changes in ker- atin polypeptide synthesis have been observed during the process of terminal differentiation of adult epidermis (2, 7, 8 ) and in maturing embryonic epidermis (9, 10). Likewise, ex- ogenous stimuli affecting the growth of the epidermis ( i .e . , induction of hyperplasia) drastically interfere with the syn-

~~ ~~

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore he hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

thesis of keratin polypeptides (5, 11, 12). Two cases of dis- turbed terminal differentiation, namely culturing of epidermal cells and malignant transformation of epidermis, give rise to very similar changes in the keratin patterns which are char- acterized by the lack of high molecular weight keratin sub- units. As a rule, only subunits below a species-independent level of about 60 kDa are synthesized by cultured cells (13- 17) and malignant tumors (3, 18-22).

In terminally differentiating epidermis of different species most of the keratin subunits are encoded by their own mRNAs (2, 23, 24), although some keratin members are undoubtedly derived from post-translational processing of mRNA encoded protein species (2, 24). In contrast, all constituent polypep- tides of the keratin patterns of cultured cells are exclusively generated from mRNAs (16, 2 5 ) . At present, little is known about the regulation of keratin synthesis in epidermal tumors. In this investigation we have therefore analyzed the keratin polypeptide synthesis at both the protein and mRNA levels in two normal keratinizing epithelia of the mouse (skin and forestomach) as well as in two squamous cell carcinomas derived from these tissues. We show that as in cultured epidermal cells, all tumor keratin subunits are translated from their own mRNAs and that in comparison to their molecular weight analogues of normal tissues, distinct keratin subunits of the tumors exhibit a considerably altered primary structure.

MATERIALS AND METHODS

adult mouse (strain NMRI, bred at the German Cancer Research Preparation of Epidermis, Epithelia, and Tumors-Epithelia from

Center) forestomach and esophagus as well as epidermis from new- born mouse back skin were separated from the connective tissue after a short incubation in hot (58 “C) isotonic saline (23). All samples were lyophilized and stored at -70 “C. The two kinds of tumors investigated originated from two-stage carcinogenesis experiments with 7,12-dimethylbenz(a)anthracene as initiator and 12-0-tetradec- anoylphorbol-13-acetate as promotor and developed either on the hack skin (26) or in the forestomach (27) of mice. Histologically, both primary tumors were classified as well differentiating SCC.’ Both tumors are kept as subcutaneously passaged transplantation tumors, their histological appearance being routinely checked after every second passage. After 12-14 days of subcutaneous growth, both tumor types reached a size of about 0.5-0.8 cm in diameter. They were excised under nonsterile conditions and the surrounding connective tissue sheet was completely removed. The nodules were immediately frozen in liquid nitrogen, lyophilized, and stored at -70 “C. Histolog- ical investigation at this stage of development revealed the absence of any necrotic material.

Extraction of Keratin Polypeptides-The extraction of keratins from epidermal and epithelial samples and from pooled total tumors was performed by removal of the soluble proteins by repeated homog-

~ ..

I The abbreviations used are: SCC, squamous cell carcinomas; SDS, sodium dodecyl sulfate.

13268

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

Keratin Synthesis in Normal Epithelia and Tumors

enization in a high salt buffer (1.5 M KCI, 10 mM NaCI, 2 mM dithioerythreitol, 05% Triton X-100, 10 mM Tris-HCI, pH 8.0) as described previously (18). The resulting pellets were dissolved in 10 mM phosphate buffer, pH 8.0, containing 5% SDS and 5% 0-mercap- toethanol by heating for 30 min at 60 "C. The protein content of the individual keratin samples was determined according to the method of Schaffner and Weissmann (28).

Isolation of 1'ol.vaden.ylated RNA and in Vitro Translation-Total cellular RNA of newborn mouse epidermis and of both types of SSC were isolated by a one-step procedure using 4 M guanidinium thiocy- anate homogenization and CsCl centrifugation as described previ- ously (23). The polyadenylated RNA was separated by oligo(dT)- cellulose chromatography (type 7, Collaborative Research lnc., Wal- tham, MA) and translated in uitro in the presence of [:'"SSJmethionine (specific activity > 500 Ci/mmol) in a mRNA-dependent reticulocyte lysate (New England Nuclear, Dreieich, Federal Republic of Ger- many).

One- and Two-dimensional Gel Electrophoresis and Peptide Map- ping-SDS-polyacrylamide gel electrophoresis was carried out ac- cording to Laemmli (29) using both homogeneous and gradient sep- aration gels. Analysis of keratin polypeptides by two-dimensional gel electrophoresis was performed as described by O'Farrell (30) and Garrels and Gibson (31) with ampholites of pH 5-8 (LKB, Bromma, Sweden). Gels were stained with 0.1% Coomassie brilliant blue R250 in 25% isopropyl alcohol and 10% acetic acid and destained in 7% acetic acid at 60 "C. Fluorography of gels was carried out according to Ronner and Laskey (32) or by impregnation in ENHANCE solution (New England Nuclear). The dried gels were exposed on Kodak X- Omat films at -70 "C. One-dimensional peptide mapping using Staphylococcus Aureus V8 protease (Miles, Frankfurt, Federal Repub- lic of Germany) was performed following the method of Cleveland et

D E

+7-9+- - 10 40kDaz

- - - - "

FIG. 1. Keratin polypeptide patterns of epithelia of new- born and adult mouse and from mouse SCC. A, newborn mouse back epidermis; R, forestomach epithelium; C , esophagus epithelium; I), back skin SCC; E, forestomach SCC. The estimated molecular weights of the keratin polypeptides are (in kilodaltons, see also Refs. 5 and 9): 2, 67; 3, 64; 4, 62; .5, 60; 6, 58; 7, 52; 8, 50; 9, 47; IO, 45. Subunit 8 of normal tissues is only visible in strongly overloaded gels (5 , 9). The arrowhead in slot I1 indicates the position of actin (43 kDa). The 56-kDa protein of the forestomach SCC is marked by a dot (7-15s gradient gel).

- 6

-9

6 " I -e

1-9

-9

3

I - -e

13269

@ -.

b w c acldrc .<#dm

FIG. 2. Two-dimensional gel electrophoresis of keratin polypeptides systhesized in vivo or translated in vitro. A, Coomassie-stained keratin polypeptides isolated from: a, newborn mouse hack epidermis; b, back skin SCC, c, forestomach SCC. H, autoradiolluorographs of %-labeled keratin polypeptides, synthe- sized in tlitro from polyadenylated RNA (1 pl ofpoly(A+)/2S-pl assay; 30 pooled and high salt-extracted assays) of: a ' , newborn mouse hack epidermis; b', back skin SCC; c', forestomach SCC. The arrowhead in b indicates the position of actin. All probes were resolved in pH 5- 8 isoelectric focusing gels in the first dimension (IEF) and 7.-5% SDS- polyacrylamide gels in the second dimension (SDS).

al. ( 3 3 ) . Gels were stained with silver nitrate (34) or were processed !or autoradiography as described above.

RESULTS

One- and Two-dimensional Analysis of Keratin Polypep- tides-Fig. 1 shows the keratin polypeptide patterns of normal mouse back epidermis (slot A ) and normal forestomach epi- thelium (slot H ) , i.e. of those regions from which the tumors investigated originally developed. (Slot A stems from newborn mouse epidermis which is qualitatively identical to that of adult mouse back epidermis, although in the latter the high molecular weight subunits 2-4 are rather weakly expressed (5). The fact that the nomencalture of the patterns starts with protein 2 is due to the finding that the keratin family of some epidermal tissues of the adult mouse contains a 70-kDa pro- tein which was assigned the number 1 (5)). With the exception of the occurrence of a low molecular weight subunit 10 in the forestomach epithelium, both tissues show a remarkably sim- ilar subunit pattern. In contrast, like other internal epithelia of the mouse (i.e. cheek and tongue, results not shown), esophagus epithelium lacks the large subunits 2-5 (slot C ) . Characteristic for the keratin patterns of both SSC is the complete absence of keratin proteins 2-4 compared to the normal counterpart tissues (slots D and E ) . In addition, both tumors contain a 40-kDa protein (35) and a 56-kDa polypep- tide is uniquely found in the forestomach carcinoma which, in turn, no longer expresses subunit 10 from normal foresto- mach epithelium (slots H and E ) . These patterns are not subject to regional variations within the tumors. Mechanical fractionation of a big tumor nodule ( r = 0.5 cm) and subse- quent extraction of small tissue pieces invariably leads to the protein patterns shown in slots D and E.

When analyzed two-dimensionally by isoelectric focusing, the low molecular weight keratin subunits (up to protein 5) of normal epidermis and forestomach epithelium (3) are con- sistently focused in an acidic pH range of about 5.1-5.6, while

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

13270 Keratin Synthesis in Normal Epithelia and Tumors ~- . . "

A B C D E F G H I J 0. "

.1 . *- - I

I ** " -

c +

FIG. 8. Peptide mapping of normal and tumor in vivo syn- thesized keratin polypeptides 5 and 6. Coomassie-stained poly- peptides 5 and 6 were cut nut from 7.5% gels and digested with 100 ng of V8 protease (during a 30-min digestion period, SO, 100, and 200 ng of V 8 protease/slot of the same polypeptide, where found to generate identical fragmentation patterns). A , protein 5 (epidermis); H, protein 5 (forestomach): C , protein 5 (back skin SCC); I), protein 5 (forestomach SCC); E , protein 6 (epidermis); F, protein 6 (foresto- mach); G , protein 6 (esophagus); H, protein 6 (back skin SCC); I, protein 6 (forestomach SCC); J, V8 protease (15% gel; silver-stain- ing). Arrowheads in slot D indicate protein fragments absent from the normal protein digests. Asterisks in slot E indicate protein frag- ments absent from the tumor protein digests. The arrows in slots A and I.: indicate a faintly stained fragment which is strongly labeled in digests of the in oitro synthesized normal proteins in Fig. 4.

the larger components 4-2 are increasingly basic in nature with protein 2 being outside the focusing range or pH 5-8 (3, 5,9). As an example for such two-dimensional patterns, that of neonatal mouse epidermis is shown in a of Fig. 2A. As a rule, of the uppermost protein doublet 5 and 6 from SSC (b and c ) only a minor portion of protein 6 is retained at the normal isoelectric position while most of the molecules be- come basic and do not enter the focusing gel in the given pH range. Conversely, the keratin polypeptides 7-9, more exten- sively expressed in carcinomas than in normal tissues, show a normal position in two-dimensional gels. The 40-kDa pro- tein is resolved into 3 to 4 isoelectric variants (35). Besides the presence of the 56-kDa protein in the forestomach SCC which shows up as 3-4 isoelectric variants in a rather alkaline pH range of 6.2-6.9 (and which is therefore not a vimentin contamination (4)), the two-dimensional polypeptide patterns of the two different SCC appear almost identical.

In Vitro Translation of Polyadenylated RNA from Epidermis and Tum0r.s-Fig. 2R shows the two-dimensionally resolved in vitro translation products of mRNA preparations from mouse epidermis ( a ' ) and from the two SCC (b' and c'). In this analysis, use was made of the solubility properties of

- e - A ! S

i I

, " -

C D

Y * 'I

t

I

1

E F

FIG. 4. Peptide mapping of normal and tumor in vitro syn- thesized keratin polypeptides 5 and 6. 30 pg of newborn mouse keratin was added as carrier to high salt-insoluble pellets of 30 pooled translation assays of epidermal and SCC mRNA. After solubilization in SDS/B-mercaptoethanol proteins 5 and 6 were cut out from Coo- massie-stained 7.55 gels and digested with 100 ng of V8 protease. A , protein 5 (epidermis); R, protein 6 (epidermis); C, protein 5 (back skin SCC); 11, protein 5 (forestomach SCC); E, protein 6 (back skin SCC); D, protein 6 (forestomach SCC). Autoradiofluorographs; 15% gel. Asterisks in slot R indicate protein fragments absent from the tumor protein digests. Arrowheads in slot D indicate protein frag- ments absent from the normal protein digests. The arrow in slot R indicates a strongly labeled fragment which is only faintly stained in the digests of the in oioo synthesized normal proteins in Fig. 3.

keratin polypeptides in that prior to electrophoresis the sol- uble labeled and unlabeled proteins were removed from the translation assays by exhaustive extraction with the high salt buffer (5, 23). It is readily apparent that with regard to proteins 5-9, the in vitro synthesized products of epidermal mRNA are exact duplicates of the in vivo situation (compare a and a ' ) . In contrast to earlier observations (23), the typical statum corneum proteins 3 and 4 (9) could not be identified among the translation products, and there is growing evidence that these proteins are derived from polypeptide 2 by post- translational processing (5, g).'

The same close resemblance between in vivo and in vitro synthesized keratin patterns holds true for both SCC, each subunit of which, including the 40- and the 56-kDa protein, is identically positioned in fairly comparable intensities (b' and c' ) .

Peptide Mapping of Keratin Proteins from Epidermis, Epi- thelia, and Tumors-V8 protease-mediated digestion of ker- atin polypeptides 5 and 6 from various epidermal regions of the adult and newborn mouse (i.e. tail, footpad, ear) has been found to generate identical fragmentation patterns for both

P. Rowden, personal communication.

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

Keratin Synthesis in Normal Epithelia and Tumors 13271

proteins (results not shown) which, however, are distinctly different from those of other keratin subunits (9). These identical digestion profiles of proteins 5 and 6 from newborn mouse epidermis are shown in slots A and E of Fig. 3. The patterns are characterized by a subset of mainly two high molecular weight protein fragments and a more complex low molecular weight subset of 6 prominent cleavage products. It is apparent from slot B that protein 5 from forestomach epithelium splits like its epidermal counterpart, whereas poly- peptide 6 from forestomach exhibits a completely different amino acid sequence (slot F ) which in turn is characteristic for other internal epithelia, i.e. esophagus (slot G), tongue and cheek (results not shown). Surprisingly, the digestion prod- ucts of protein 6 from both tumors (slot H and I ) are distinctly different from those of any of the normal proteins 6 but are mutually identical and virtually indistinguishable from those of tumor polypeptides 5 (slots C and D ) .

The peptide maps of keratin polypeptides 5 and 6 translated in vitro from mRNA of mouse epidermis and the two SSC are shown in Fig. 4. Without exception, the patterns of the labeled fragments of both normal and tumor proteins are absolutely comparable to those of the corresponding silver-stained frag- ments of Fig. 3. This suggests that the modified primary structure of tumor proteins 5 and 6 is determined by mRNAs which are different from those coding for proteins 5 and 6 from the normal epithelia.

DISCUSSION

Numerous investigations in a variety of species have shown that the epidermis undergoes a defined program of keratin polypeptide synthesis, leading to the sequential and ordered appearance of distinct subunits as cells progress from the proliferative basal layer to the dead outermost layers (2, 7- 10). It may generally be said that subunits below the 60-kDa level are generated from basal cells and the larger keratin proteins are made from differentiating suprabasal cells (2, 7- 10). The close resemblance of the keratin patterns of epider- mal tumors with regard to size distribution of the subunits to that assumed to be characteristic for basal cells is in line with the severely disturbed terminal differentiation of malignant epidermal lesions. Furthermore, tumors frequently express keratin proteins which are not found in the corresponding normal tissues or cells (17, 35, 36). Among these proteins a 40-kDa protein has been described in several human malig- nant cell lines (35) which is also present in the two mouse SCC and its charge properties correspond fairly well to those found in species other than the mouse (35). Our investigation has shown that the altered keratin expression in tumors is regulated at the transcriptional level, since every subunit is encoded by its own mRNA. There is no evidence for post- translational protein processing in tumors, and this phenom- enon is obviously restricted to terminally differentiating tis- sues.

The complexity of normal epidermal keratin expression and the far reaching consequences of malignant transforma- tion are especially apparent with regard to the 58- and 60- kDa protein doublet 5 and 6. These mRNA encoded proteins, invariably present in normal mouse epidermis of any iocaliza- tion (5), belong to the class of acidic keratin polypeptides (5, 9) and have identical peptide maps (9). Although such an acidic protein doublet is also encountered in the orthokeratin- izing forestomach epithelium (3), only the 60-kDa protein 5 is identical with its epidermal counterpart, whereas the 58- kDa protein 6 possesses a completely different primary struc- ture typical for the corresponding protein of the esophagus and other internal epithelia. Moreover, forestomach and esophagus epithelium share the low molecular weight subunits

7-10. This indicates a mixed program of keratin polypeptide synthesis in forestomach epithelium which partially correlates with integumental epidermis and partially with internal epi- thelia.

This structural heterogeneity within the protein subset 5 and 6 of normal tissues is faced with a striking uniformity of the corresponding tumor proteins 5 and 6, whose identical primary structure is distinctly different from that of any of the normal proteins. The uniform digestion patterns of kera- tin subunits 5 and 6 from both tumors have most of their protein fragments in common with those of their integumental counterparts. There are, however, numerous fragments which do not occur in the digestion patterns of the normal proteins. This finding of an altered primary structure is especially striking when considering the fragmentation products of nor- mal and tumor polypeptide 6 from the forestomach epithe- lium. Preliminary results with polyclonal antibodies against proteins 5 and 6 from normal and tumor tissue indicate that the differences in the primary structures are important enough to discriminate between the proteins in Western blot experiments (results not shown). In this context it should, however, be emphasized that the primary structure of the tumor proteins certainly does not account for the altered charge properties of these proteins compared with those of the normal proteins. Rather, their position in two-dimen- sional gels reflects the general finding that independent of the molecular weight range of a distinct keratin polypeptide family the large keratin subunits are invariably neutral to slightly alkaline in nature (22). This polarity within the protein family may be functionally important in the filament assembly of the subunits (37).

At present it can only be speculated whether alterations in the gene structure (or the associated regulatory elements) or a modified processing of the primary gene transcripts may be responsible for the observed discrepancies in keratin gene expression and the properties of the subunits of normal and malignantly transformed tissues. Changes in the environment and growth conditions have recently been shown to induce dramatic alterations in keratin gene expression in that under a variety of culture conditions epidermal cells could be trig- gered to synthesize a set of mRNA-encoded keratin proteins, none of which had an analogue in normal epidermis in vivo (16). What remains, however, is the amazingly unifying effect of the transformation event on keratin patterns and protein properties in comparison to a remarkable heterogeneity of at least the latter of these features in two rather remotely located mouse epithelia. Considering the identical experimental pro- tocol for the induction of both tumors, it can not be excluded that neoplastic epithelial growth proceeded via the selection of a cellular subpopulation most probably located in the interfollicular basal layer (38-42). Evidence for cell selection in malignancy has recently been provided in a comparative study on keratin polypeptide patterns of a human basal cell carcinoma and the epithelium of the pilosebaceous tract (36).

Subfractionation of basal epidermal cells of normal and carcinogen-treated epidermis and the construction of cDNA clones for normal and tumor keratin mRNAs, currently under way in our laboratory, may help to elucidate whether the keratin phenotype in epidermal tumors is the result of a selection process or due to alterations in the gene expression.

Acknowledgments-We thank Dr. A. Balmain for helpful discus- sion and H. Loehrke for technical assistance. Dr. A. Marston provided stylistic help in preparing the manuscript.

REFERENCES 1. Lee, L. D., Kubilus, J., and Baden, H. P. (1979) Biochern. J . 117,

2 . Fuchs, E., and Green, H. (1980) Cell 19, 1033-1042 187-196

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

13272 Keratin Synthesis in Normal Epithelia and Tumors

3. Winter, H., and Schweizer, J. (1981) Carcinogenesis 2, 613-621 4. Franke, W. W., Schiller, D. L., Moll, R., Winter, S., Schmid, E.,

Engelbrecht, I., Denk, H., Krepler, R., and Platzer, B. (1981)

5. Schweizer, J., and Winter, H. (1982) Cancer Res. 42, 1517-1529 6. Milstone, L. M. (1981) J . Cell Bid. 88, 317-322 7. Banks-Schlegel, S. P., Schlegel, R., and Pinkus, G. S. (1981) Exp.

8. Skerrow, D., and Skerrow, C. J. (1983) Exp. Cell Res. 143, 27-

9. Schweizer, J., and Winter, H. (1982) Differentiation 22, 19-24 10. Banks-Schlegel, S. P. (1982) J. Cell Biol. 93, 551-559 11. Baden, H. P., Kubilus, J., and Argyris, T. S. (1980) J. Inuest.

Derrnatol. 75,383-387 12. Laskin, J. D., Mufson, R. A,, Piccinini, L., Engelhardt, D. L., and

Weinstein, I. B. (1981) Cell 25, 441-449 13. Fuchs, E., and Green, H. (1978) Cell 15, 887-897 14. Sun, T.-T., and Green, H. (1978) J. Biol. Chem. 253, 2053-2060 15. Doran, T. I., Vidrich, A,, and Sun, T.-T. (1980) Cell 22, 17-25 16. Roop, D. R., Hawley-Nelson, P., Cheng, C. K., and Yuspa, S. H.

17. Banks-Schlegel, S. P., and Howley, P. M. (1983) J. Cell Biol. 96,

18. Winter, H., Schweizer, J., and Goerttler, K. (1980) Carcinogenesis

19. Winter, H., Schweizer, J., and Goerttler, K. (1983) Arch. Der-

20. Kubilus, J., Baden, H. P., and MacGilvray, N. (1980) J. Natl.

21. Breitkreutz, D., Tilgen, W., Boukamp, P., and Fusenig, N. E.

22. Moll, R., Franke, W. W., Schiller, D., Geiger, B., and Krepler, R.

J. Mol. Biol. 153, 933-959

Cell Res. 136,465-469

35

(1983) Proc. Natl. Acad. Sci. U. S. A. 80, 716-720

330-337

1, 391-398

matol. Res. 275, 27-34

Cancer Inst. 65,869-873

(1981) Anticancer Res. 1, 323-328

(1982) Cell 31, 11-24

23. Schweizer, J., and Goerttler, K. (1980) Eur. J . Biochem. 112,

24. Bladon, P. T., Bowden, P. E., Cunliffe, W. J., and Wood, E. J.

25. Fuchs, E., and Green, H. (1979) Cell 17, 573-582 26. Goerttler, K., and Loehrke, H. (1976) Virchows Arch. A. Pathol.

27. Goerttler, K., Loehrke, H., Schweizer, J., and Hesse, B. (1979)

28. Schaffner, W., and Weissmann, C. (1973) Anal. Biochem. 56,

29. Laemmli, U. K. (1970) Nature (Lond.) 227,680-685 30. O'Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021 31. Garrels, J. I., and Gibson, W. (1976) Cell 9, 793-805 32. Bonner, W. M., and Laskey, R. A. (1974) Eur. J . Biochem. 46,

33. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli,

34. Oakley, B. R., Kirsch, D. R., and Morris, N. R. (1980) Anal.

35. Wu, Y. J., and Rheinwald, J. G. (1981) Cell 25, 627-635 36. Moll, R., Franke, W. W., Volc-Platzer, B., and Krepler, R. (1982)

37. Steinert, P. M., Idler, W. W., and Zimmermann, S. B. (1976) J.

38. Lavker, R. M., and Sun, T.-T. (1982) Science (Wash. D. C.) 215,

39. Raick, A. N. (1973) Caneer Res. 33,269-286 40. Klein-Szanto, A. J. P. (1977) J. Cutaneous Pathol. 4, 275-280 41. Klein-Szanto, A. J. P., Topping, D. C., Heckman, C. A., and

42. Klein-Szanto, A. J. P., Major, S. K., and Slaga, T. J. (1980)

243-249

(1982) Biochern. J . 208, 179-187

Anat. Histol. 372, 29-38

Cancer Res. 39, 1293-1297

502-514

83-88

U. K. (1977) J . Biol. Clzern. 252, 1102-1106

Biochern. 105, 361-363

J. Cell Biol. 95, 285-295

Mol. Bid. 108, 547-567

1240-1241

Nettesheim, P. (1980) A m . J . Pathol. 98, 83-100

Carcinogenesis 1, 399-406

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

J Schweizer and H Winterin tumors.

mRNA-dependent alterations of the primary structure of distinct keratin subunits Keratin biosynthesis in normal mouse epithelia and in squamous cell carcinomas.

1983, 258:13268-13272.J. Biol. Chem. 

  http://www.jbc.org/content/258/21/13268Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/21/13268.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from