IN VITRO Volume 15, No. 6, 1979 All rights reserved 9

C O L L A G E N S Y N T H E S I S IN N O R M A L B H K C E L L S AND T E M P E R A T U R E - S E N S I T I V E C H E M I C A L L Y T R A N S F O R M E D BHK CELLS B. D. SMITH, D. BILES, W. GONNERMAN, B. FARIS, A. LEVINE, N. CAPPARELL, F. MOOLTEN, ANDC. FRANZBLAU Departments of Biochemistry and Microbiology, Boston University School of Medicine, Boston, Massachusetts O2118 (D. B., IV. G., B. F., A. L., N. C., F. M., C. F.); and Connective Tissue-Aging Research Laboratory, Veterans Administration Outpatient Clinic, 17 Court Street, Boston, Massachusetts 02108 (B.D.S. )

~ReccivedAugust 18, 1978; acceptedNovember27, 1978} SUMMARY Collagen synthesis in normal BHK 21/cl 13 and chemically transformed temperaturesensitive BHK 21/cl 13 cells (MezN4) was assessed by examination of hydroxyproline formation and collagenase-susceptible protein. The Me2N4 cells lost their ability to synthesize collagen at both permissive and nonpermissive temperatures for transformation. These conclusions were confirmed by polyacrylamide-gel electrophoresis and CM-cellulose chromatography. Prolyl hydroxylase activity was present in both normal and transformed cells even when no collagen could be demonstrated. The production of noneollagen protein, although decreased in the transformed cell, did not change as drastically as the collagen synthesis. INTRODUCTION

Cells transformed by either viruses or chemicals have been used to study aspects of cellular growth control. Transformation causes significant changes in cellular morphology, function, transport and synthesis. Certain cell lines exhibit temperature-dependent expression of the transformed phenotype after exposure to chemical carcinogens (1,2) or viral agents (3). A criterion for distinguishing transformed from normal phenotypes [described by MacPherson and Montagnier {4)] has been the ability to grow in soft agar. Di Mayorca et al. (1), using this assay, have reported temperature-sensitive variants of BHK 21/cl 13 cells exposed to the carcinogen dimethylnitrosamine. These transformed cells display the normal phenotype at 32~ C, but express the transformed phenotype when incubated at 38.5 ~ C. The phenomenon is reversible when the temperature is changed. The synthesis and characterization of collagen have been the locus of a variety of studies involving virally transformed and normal cell lines (5-14). In all cases the percentage of collagen synthesized decreases with transformation I5-14). Hata and Peterkofsky (7) have shown that the type of collagen also changes with transformation.

Some studies suggest that conversion of procollagen to collagen is defective (11,12), whereas others indicate that decreased collagen messenger RNA levels (13,14) are responsible for decreased collagen synthesis. Temperature-sensitive viruses have been used to show that the changes in collagen synthesis occur with transformation hut not with virus infection (8, 11). Little information has been published concerning the synthesis of collagen in temperaturesensitive, chemically transformed cells. It is of interest to find out if relative collagen synthesis changes in chemical transformation as well as in viral transformation. If these changes, are temperature-sensitive, the elucidation of the process may lead to greater understanding of collagen synthesis. Toward this end the present communication examines collagen biosynthesis in normal BHK 21/cl 13 cells and their chemically transformed counterpart {Me2N4) at permissive {38.5~ C) and nonpermissive {32~ C} temperatures. MATERIALS AND METHODS

Cell Culture The BHK 21/cl 13 and Me2N4 cells were initially established and described by di Mayorca et

455

456

SMITH ET AL.

al. tl). Low-passage cells were kindly provided by di Mayorca and stored in liquid nitrogen until used for biochemical studies. The cells were tested for temperature-sensitive characteristics, and only Me2N4 cells grown at 38.5 ~ C formed colonies on soft agar as described previously (1). The cells were tested periodically for mycoplasma (15) and were always negative. For the collagen experiments, cells were grown at either 32 ~ C or 38.5 ~ C in 75-em 2 flasks using Dulbecco's medium supplemented with 10% fetal bovine serum. Each flask was gassed with 5% COs and sealed.

Incorporation of Radioactive Proline or Glycine Confluent cells to be incubated with [14C]proline were washed several times with small volumes of medium containing no fetal bovine serum. The cells then were preineubated for 1 hr at either 32 ~ C or 38.5 ~ C with medium supplemented with 50 tag per ml of sodium ascorbate but lacking both proline and fetal bovine serum. After this period, the medium was changed to one containing 1 taCi per ml of [~4C]proline {specific activity 260 taCi per mM) and sodium ascorbate as described above. The cells were incubated for 6 hr alter which time the cultures were assayed for hydroxyproline formation or eollagenase-suseeptible protein as described below. In some of the above experiments/3-amino propionitrile fumarate {BAPN) was added 150 tag per ml) along with the radiolabeled proline. In other experiments [3H]glycine (10 taCi per ml) was used in addition to [3H]prolinc (10 taCi per ml) in order to obtain higher specific activity in the collagen produced.

Collagen Synthesis Sample preparation. Medium from labeled cultures was poured off and centrifuged at approximately 2500 x g for 15 min. The resulting supernate was dialyzed versus distilled I-I20 at 4 ~ C until radioactivity in the diffusate reached background levels. Cells were washed 3 times with phosphate buffered saline IPBS), pH 7.4, scraped off the flasks and resuspended in PBS. Cell suspensions were dialyzed as above. All dialysates were lyophilized separately and the residues were weighed. Hydroxyproline determination. Some lyophilized samples from above were hydrolyzed in 6 N HCI at 108 ~ C for 20 hr in sealed vials. The hydrolysates then were placed on a Technicon

amino acid analyzer equipped with a streamsplitting device. Fractions of 1.3 ml were collected and the radioactivity in 1.0-ml aliquot from each fraction was counted in a Packard liquid scintillation counter. Collagenase assay. The amount of collagen and noncollagen protein was determined using a protease-frec bacterial collagenase as previously described 116,17). Briefly, lyophilized samples from the labeled cultures were suspended individually in 1 ml of 0.15 M NaCI containing 2 mg per ml of bovine serum albumin. After a trichloroacetic acid (TCA)-tannic acid mixture was added, the resulting precipitate was rediseolved and neutralized in H E P E S buffer containing N-ethylmaleimide and calcium. After overnight incubation at 37 ~ C with purified bacterial collagenase, an additional TCA-tannic acid precipitation was carried out. Radioactivity of the resulting superhate was determined. Control assays were carried out without collagenase and all samples were assayed in duplicate.

Prolyl Hydroxylase Peptide substrates were prepared by incubating calvaria from 16-day chick embryos in prolinefree Dulbecco's medium (2 calvaria per ml~ containing 3,4-[3H]proline (New England Nuclear, 5 taCi per ml) and 0.2 mM a-a'-dipyridyl for 24 hr. The incubation medium was dialyzed extensively against distilled water and aliquots were stored frozen. Before the medium was used as substrate, it was heated in a boiling water bath for 10 min. Routinely 20,000 to 22,000 cpm of nondialyzable radioactivity were used for each assay tube. Prolyl hydroxylase activity was estimated by modifications of the tritium release assay of Hutton, Tappel and Udenfriend (18~. Cofactors were added according to the procedure of Kivirrikko and Prockop (19). The B H K 21/cl 13 and Me2N4 cells were harvested at confluency by decanting the incubation medium, rinsing the cells with PBS, and then suspending them in PBS. After the suspension was homogenized in a Polytron, aliquots of the whole homogenate were used for nitrogen determination by the Kjeldahl procedure. The remainder of the homogenates was diluted with buffer to yield a final concentration of 0.1 M glycine, 0.2 M NaCI, 50 mM dithiothreitol, 20 mM tris, 0.1% Triton X100, pH 7.4, and allowed to stand on ice for 2 hr with occasional stirring. After centrilugation the supernate was used for the enzyme assay.

457

COLLAGEN SYNTHESIS BY BHK A N D ME2N4 CELLS

~2 ~0

§

c~ +I 9-

o_~

4~

II

• ~o~

Z

'i

~I~-~ ~m

-I-I

§

m

~

§ x

"~

~ ~

~lV

=~

~"

458

SMITH ET AL.

Each flask of cells was assayed in duplicate. Values obtained from a blank containing the enzyme and substrate without cofactors were subtracted from the activity of each flask. Four flasks were used for each point.

TABLE2 PROLYL HYDROXYLASE ACTIVITY INNORMAL ANDTRANSFORMED CELLS Temp.

Cell

[3H]-Release/Total Cell N2a

38.5 ~ C

Collagen Chain Distribution Cells that had been pulsed with radioactive proline a n d / o r glycine in the presence of BAPN as described above were examined for collagen and procollagen chains. Media from identically treated culture flasks were pooled and centrifuged (2500 x g) to remove debris and cells. The clear supernate from the media was dialyzed against a buffer containing 0.15 M NaCI, 0.05 M tris, 0.02 M E D T A , pH 7.4, including the protease inhibitors p-chloromercuribenzoate ~1 raM) and phenylmethylsulfonylfluoride {10 pM} as previously described (20). The cell layers were extracted overnight in 0.5 N acetic acid with the addition of 20 mg rat skin collagen standard and then centrifuged at 15,000 x g for 15 rain. The supernate then was dialyzed overnight against 0.05 M sodium acetate buffer, pH 4.8, heated to 90 ~ C for 5 min and chromatographed on CM-eclhlose columns by the method of Piez, Eigner and Lewis 121). Aliquots of media and acid extracts were monitored by SDS-polyacrylamide slab gel electrophoresis by the method of Laemmli ~22}. Approximately equal amounts of radioactivity were applied. Fluorograms of the radioactive gels were made by saturating the gels in dimethylsuifoxide and P P O using the method of Bonner and Laskey (23). Radioactive collagen and procollagen from chick calvaria were utilized as standards for gel electrophoresis. RESULTS

Proline incorporation. The average amount of collagen synthesized in 6 hr by triplicate cultures of B H K 21/c113 cells and Me2N4 cells is reported in Table 1. The hydroxyproline assays were repeated in three other sets of experiments and the collagenase assays in two other sets of experiments. The data always showed that transformed cultures incubated at 32 ~ C or 38.5 ~ C show a striking decrease in formation of hydroxyproline or collagenase-susceptible protein when compared to their corresponding control cells. This effect was observed with both the lyophilized nondialyzable material obtained from the medium and the cell layer. The Me2N4 cultures have less colla-

BHK 1366 _+421 b Me2N4 1353 _+ 77 32~ C BHK 1168 _+155 Me2N4 687 _+ 28 a Total cell nitrogen is derived from a Kjeidahl determination on the cell homogenate. Enzyme is extracted from total cell homogenate and assayed as described in text. Data is from four individual flasks (not pooled). b SEM ~standard error of the mean).

genous counts in the medium than in the cell layer. The presence of noncollagenous protein also could be evaluated by the collagenase method or by examination of total ['4C]proline incorporation into protein. On the basis of data presented in Table 1, the noncollagenous protein production does not appear to decrease as significantly as does the coUagenous protein. Prolyl hydroxylase assay. Since there is such a marked decrease in the levels of hydroxyproline in the Me2N4 cells, it was of interest to find out if prolyl hydroxylase activity decreased in a coordinated fashion with collagen synthesis. As shown in Table 2 the difference in prolyl hydroxylase ~000

!~ 0

l/t t 20 40 60 FRACTION NUMBER

80

FIG. 1. Acid-extracted cell layer from cells grown at 32~ C chromatographed by CM-cellulose chromatography. The column was run in 0.05 M Na acetate buffer with a linear gradient from 0 to 0.1 M NaCI over a total volume of 400 mi. Five-ml fractions were collected and 0.5-ml aliquots from every other tube were counted. OH-O, [14C]-cpmof BHK 21/c113 cells ~186,340 cpm); @ - - @ , ['4C]-cpm of Me2N4 cells ~216,370 cpm); --, absorbance at 223 nm of added lathyritic rat skin collagen standard.

COLLAGEN SYNTHESIS BY BHK AND ME2N4 CELLS

459

t4

x sF MI

2

~ la~l" i

o

I

~

32

I

~

i

~o

J

~o

FRACTIONNUMBER FIG. 2. Acid-extracted cell layer from cells grown at

38.5 ~ C chromatographed by CM-cellulose chromatography. Chromatography conditions were exactly the same as in Fig. 1. O---C), [~H]-cpm of BHK 21/cl ]3 cells (2,826,900 cpm); O - - O, [3H]-cpm of Me2N4 cells (3,603,330 cpm); - - , absorbance at 223 nm of added lathyritic rat skin collagen standard. activity between the normal and transformed cells, although decreased at 32 ~ C, was not as significantly decreased as the hydroxyproline content shown in Table 1. Collagen characterization. An attempt was made to demonstrate collagen a-chains in the normal and transformed cultures by CM-cellulose chromatography and gel electrophoresis. The acetic acid extract of the B H K 21/c113 cell layer, but not that of the Me2N4 cell layer, contained a 1 and a2-chains as shown by CM-cellulose chromatography {Figs. 1, 2) and by gel electrophoresis {Fig. 3, gel D). The ratio of a l to a2 was always much greater than the normal type I collagen ratio although the standard carrier collagen ratio was normal {Figs. 1, 2). At 38.5 ~ C the B H K 21/el 13 made essentially no a2-chains (Fig. 2). The figures are representative chromatographs from at least 10 CM-cellulose columns. The transformed cell had no discrete peaks in the a l - or the a2-chain region on CM-cellulose chromatography {Figs. 1, 2} or on gel electrophoresis {Fig. 3, gel E). In some experiments both glycine and proline were used to label the cultures which should increase the specific activity of collagen; however, we still could not observe any discrete collagen achains in the transformed cell cultures at either temperature. The gel fluorogram of unfractionated medium from B H K 21/cl 13 cells (Fig. 3, gel B) contains

FIo. 3. Fluorogram of media and acid-extractedcell layersfrom cellsgrown at 32~ C electrophoresedon 5% polyacrylamideSDS gels.All sampleswere reducedwith I m M dithiothreitol.A, Standard chicken collagenand procollagen (30,000 cpm); B, [3H]-cpm in BHK 21/c113 medium {190,000 cpm); C, [3H]-cpm in Me2N4 medium ~150,000 cpm); D, [3H]-cpm in BHK 21/cl 13 acidextracted cell layer (168,000 cpm); E, [3H]-cpm in Me2N4 acid-extractedcell layer (188,000 cpm).

several protein bands, some of which correspond to collagen and procollagen standards. Since this was a 5 % polyacrylamide gel, proteins less than 40,000 mol wt appear at the bottom of the gel. Strikingly,the M e 2 N 4 medium (Fig. 3, gel C) displays only one high molecular weight protein band. DISCUSSION Previous studies have shown that in general collagen synthesis is reduced in transformed cells. When hydroxyproline measurements were employed as an indication of collagen synthesis,

460

SMITH ET AL.

collagen production varied with transformation depending on culture conditions (10). Peterkofsky and her colleagues 17, 16, 17), using purified bacterial collagenase to measure collagen production, have shown that 3T3 cells normally synthesized 1% to 2% of their total protein as collagen, and that transformed 3T3 cells synthesize a significantly smaller percentage of collagen ~7). Levinson, Bhatnagar and Liu (9), using secondary cultures of chicken embryo fibroblasts, showed that Rous sarcoma virus infection decreased collagen production by 50% after a 24-hr exposure and by 80% to 90% after a 72-hr exposure to virus. Further, Arbogast et al. (11), using similar Rous sarcoma virus-transformed fibroblasts, and Sundarraj and Church (12), using SV40 infected human cells, found procoRRagen secreted into the medium without any processing of procollagen to collagen. Our data indicate that collagen synthesis is dramatically altered in the chemically transformed B H K 21/cl 13 cells at both temperatures studied. The total collagen synthesized in the normal cell at 32 ~ C represented 1.8% of the total protein synthesized and 0.9% at 38.5 ~ C. On the other hand, the transformed cells synthesized approximately 0.2% of their total protein as collagen at either temperature. This decrease in collagen synthesis is much greater than that which Hata and Peterkofsky (7) observed for 3"1"3 cells and slightly greater than that reported for Rous sarcoma virus-transformed chicken fibroblasts (8, 9). In contrast to Arbogast et al. (11) or Sundarraj and Church (12), no procollagen could be demonstrated in the media of Me2N4 cultures, and acmaRRy less collagenous material was found in the media than in the cell layers (Table 1). Possible explanations for the observed reduction in collagen include inhibition of collagen synthesis, inhibition of secretion, impaired hydroxylation of proline and/or an increase in collagen degradation. The level of the enzyme prolyl hydroxylase was not altered significantly; therefore impaired enzyme does not appear to be the cause of the obvious decrease in hydroxyproline, and the prolyl hydroxylase activity is not coordinated with collagen synthesis. Examination of the distribution of collagen in the cell layer and medium revealed that the transformed cells secrete proportionally less collagen into the medium as well as synthesizing less collagen. This reduced secretion is not limited to collagen synthesis since examination of the polyacrylamide gel electrophoretic pattern of the protein molecules secreted by the normal and transformed cells into the me-

dium indicates differences in a variety of other proteins as well as collagen. In fact, the transformed cell secretes only one detectable high molecular weight protein moiety by this method. Some of these findings could be explained by increased breakdown of protein in transformed cells. Protease and collagenase inhibitors were added immediately after removing the medium to reduce protein breakdown to a minimum during sample processing. Preliminary data (unpublished data) indicate that collagenase activity could not be detected in these cell lines. No other protease activities were examined, however, so general protcolysis during the labeling period re. mains a possible explanation for the general reduction in medium proteins by the Me~N4 cells. Our data strongly suggest that the synthetic machinery for collagen molecule production is altered in the chemically transformed cell. Recent evidence (13,14) to support this finding suggests that Rous sarcoma virus-transformed chick embryo fibroblasts contain no translatable messenger RNA for collagen. Since there are genetically distinct collagen molecules, an attempt was made to determine collagen chain composition produced by these cells. As one might expect from the insignificant amounts of collagen shown in Table 1, no discrete collagen chains could be demonstrated in the cell layer from the transformed cells by CM-cellulose chromatography or gel electrophoresis, and no procollagen molecules could be demonstrated by gel electrophoresis. It is difficult, however, with the methods used in this study, to rule out the possibility that small amounts of collagen are being produced by the transformed cells. In contrast to Me2N4 cells, Hata and Peterkofsky (7) were able to demonstrate collagen production in transformed 3T3 cells as their transformed 3T3 cells produced more collagen than the Me2N4 cells. These authors suggest that the collagen types made by the transformed 3T3 cells differ from those of the parent cell line. The normal BALB 3T3 cells produced collagen with an excess of alchains, whereas the transformed cells produced type I I I collagen and type I collagen with no excess of a-chains. The cell layer of the parent B H K 21/cl 13 cell also displays collagen with a high al to a2 ratio. Other laboratories have reported similar findings in a variety of pathological situations or in cell cultures (24-28). Several investigators have suggested the presence of an ~1 (I) trimer responsible for this phenomenon (25-28). Future studies, includiug peptide maps, will be necessary

COLLAGEN SYNTHESIS BY BHK AND ME2N4 CELLS to determine if this collagen is indeed an a l {I) trimer. A key observation in this study is that collagen synthesis in these transformed cells was not temperature-sensitive. In contrast, Kamine and Rubin (8), using a temperature-sensitive Rous sarcoma virus, found that collagen synthesis was reduced at the permissive temperature and reached normal levels at the nonpermissive temperature. Although the Me2N4 ceils grow on soft agar at 38.5 ~ C, hut not at 32 ~ C, the cells produce no collagen. One might conclude from these data that collagen synthesis does not affect the ability of these cells to grow on soft agar. The morphology of the transformed cells is different from that of the parent cell line and these differences are present to some extent even at 32 ~ C. Thus the collagen changes may be more closely correlated with morphological changes in cells rather than with the viability of cells in soft agar. Certainly, the assignment of a "normal" status to these transformed cells at a particular temperature based on the agar plate technique is clearly not sufficient and must be made with caution. Studies on collagen synthesis in temperaturesensitive mutants were performed using a temperature-sensitive virus (8,11). However, in this study a chemical affected one or possibly more cellular genes. If more than one gene is affected in the Me2N4 cell, it is possible that certain characteristics are temperature-sensitive while others are not. Further study of the genetic mutation causing the selectively greater collagen decrease may lead to insights regarding collagen regulation.

REFERENCES 1. Di Mayorca, M., M. Greenblatt, T. Trauthen, A. Soller, and R. Giordano. 1973. Malignant transformation of BHK21 clone 13 ceils in vitro by nitrosamines--a conditional state. Proc. Nat. Acad. Sci. U.S.A. 70: 46-49. 2. Weinstein, I. B., N. Yamaguchi, R. Gerbert, and M. E. Kaign. 1975. Use of epithelial cell cultures for studies on the mechanisms of transformation by chemical carcinogens. In Vitro 11: 130-141. 3. Martin, G. S. 1970. Rolls sarcoma virus: A function required for the maintenance of the transformed state. Nature 227: 1021-1023. 4. MacPherson, I., and L. Montagnier. 1964. Agar suspension culture for the selective assay of cells transformed by polyoma virus. Virology 23: 291-294. 5. Green, H., and B. Goldberg. 1963. Kinetics of collagen synthesis by established mammalian cell lines. Nature 200: 1097-1098.

461

6. Green, H., G. J. Todaro, and B. Goldberg. 1966. Collagen synthesis in fibroblasts transformed by oncogenicviruses. Nature 209: 916-917. 7. Hata, R. I., and B. Peterkofsky. 1977. Specific changes in the collagen phenotype of BALB 3T3 cells as a result of transformation by sarcoma viruses or a chemical carcinogen. Proe. Nat. Acad. Sci. U.S.A. 74: 2933-2937. 8. Kamine, J., and H. Rubin. 1977. Coordinate control of collagen synthesis and cell growth in chick embryo fibroblasts and the effect of viral transformation on collagen synthesis. J. Cell. Physiol. 92: 1-12. 9. Levinson, W., R. S. Bhatnagar, and T.-Z. Liu. 1975. Loss of ability to synthesize collagen in fibroblasts transformed by Rous sarcoma virus. J. Nat. Cancer Inst. 55: 807-810. 10. Temin, H. 1965. The mechanism of carcinogenesis by avian sarcoma viruses. I. Cell multiplication and differentiation. J. Nat. Cancer Inst. 35: 679-693. 11. Arbogast, B. W., M. Yoshimura, N. A. Kefalides, H. Holtzer, and A. Kaji. 1977. Failure of cultured chick embryo fibroblasts to incorporate collagen into their extracellular matrix when transformed by Rous sarcoma virus. J. Biol. Chem. 252: 8863-8868. 12. Sundarraj, N., and R. L. Church. 1978. Alterations of post-translational modifications of proeollagen by SV40-transformed human fibroblasts. FEBS Lett. 85: 47-51. 13. Adams, S. L., M. E. Sobel, B. H. Howard, K. Olden, K. M. Yamada, B. de Crombruggle, and I. Pastan. 1977. Levels of translatable mRNA's for cell surface protein, collagen precursors, and two membrane proteins are altered in Rous sarcoma virus-transformed chick embryo fibroblasts. Cell 74: 3399-3403. 14. Rowe, D. W., R. C. Moen, J. M. Davidson, P. H. Byers, P. Bornstein, and R. D. Palmiter. 1978. Correlation of procollagen mRNA levels in normal and transformed chick embryo fibroblasts with different rates of procollagen synthesis. Biochemistry 17: 1581-1590. 15. Schneider, E. L., E. J. Stanbridge, and C. J. Epstein. 1974. Incorporation of 3H-uridine and 3Huracil into RNA--a simple technique for detection of mycoplasma contamination of cultured cells. Exp. Cell Res. 84: 311-318. 16. Peterkofsky, B. 1972. The effect of ascorbic acid on collagen polypeptide synthesis and proline hydroxylation during the growth of cultured fibroblasts. Arch. Biochem. Biophys. 152: 318-328. 17. Peterkofsky, B., and R. Diegelmaun. 1971. Use of a mixture of proteinase-free collagenases for the specific assay of radioactive collagen in the presence of other proteins. Biochemistry 10: 988-994. 18. Hutton, J. J., Jr., A. L. Tappel, and S. A. Udenfriend. 1966. A rapid assay for collagen proline hydroxylase. Anal. Biochem. 16: 384-394. 19. Kivirrikko, K.I., and D.J. Prockop. 1967. Hydroxylation of proline in synthetic polypeptides with purified protocollagen hydroxylase. J. Biol. Chem. 242: 4007-4012.

462

SMITH ET AL.

20. Smith, B. D., K. H. MeKenney, and T. J. Lustberg. 1977. Characterization of collagen precursors found in rat skin and bone. Biochemistry 16: 2980-2985. 21. Piez, K. A., E. A. Eigner, and M. S. Lewis. 1963. The chromatographic separation and amino acid composition of the subunits of several collagens. Biochemistry 2: 58-66. 22. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685. 23. Bonner, W. M., and R. A. Laskey. 1974. A film detection method for tritium labelled proteins and nucleic acids in polyacrylamide gels. Eur. J. Biochem. 46: 83-88. 24. Little, C . D . , R. L. Church, R . A . Miller, and F. H. Ruddle. 1977. Procollagen and collagen produced by teratocarcinoma-derived cell line, TSD4: Evidence for a new molecular form of collagen. Cell 10: 287-295.

25. Mayne, R., M. S. Vail, P. M. Mayne, and E. J. Miller. 1976. Changes in type of collagen synthesized as clones of chick chondrocytes grow and eventually lose division capacity. Proc. Nat. Acad. Sci. U.S.A. 73: 1674-1678. 26. Mayne, R., M. S. Vail, and E. J. Miller. 1975. Analysis of changes in collagen biosynthesis that occur when chick chondrocytes are grown in 5bromo-2'-deoxyuridine. Proc. Nat. Acad. Sci. U.S.A. 73: 4511-4515. 27. Moro, L., and B. D. Smith. 1977. Identification of collagen crl (I) trimer and normal type I collagen in a polyoma virus-induced mouse tumor. Arch. Biochem. Biophys. 182: 33-41. 28. Narayanan, A.S., and R . C . Page. 1976. Biochemical characterization of collagens synthesized by fibroblasts derived from normal and diseased human gingiva. J. Biol. Chem. 251: 5464-5471.

This paper was supported in part by a grant from the Public Health Service (AG00001), and by the Medical Research Service of the Veterans Administration.