THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc

Vol. 262, No , 5, Issue of February 15, pp. 2244-2249,1987 Printed in U. S. A.

Identification of Multiple Tropoelastins Secreted by Bovine Cells*

(Received for publication, August 11, 1986)

David S. WrennS, William C. Parks$, Loren A. Whitehouse$, Edmond C. Crouch$, Umberto Kucichll, Joel Rosenbloomn, and Robert P. Mecham$II From the $Respiratory Disease and Critical Care Division, Department of Medicine, Jewish Hospital at Washington University Medical Center, the $Department of Pathology, Jewish Hospital at Washington University Medical Center, St. Louis, Missouri 631 10, and the Wenter for Oral Health Research, School of Dental Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104

High resolution gradient sodium dodecyl sulfate-pol- yacrylamide gel electrophoresis, cell-free translation, and elastin-specific antibodies were used to identify three tropoelastin isoforms secreted by bovine tissue and cells. Tropoelastin isolated from nuchal ligament and from conditioned culture medium or cell-matrix extracts of ligament fibroblasts and auricular chondro- cytes resolved as three distinct bands on sodium dode- cy1 sulfate-polyacrylamide gel electrophoresis with molecular weights of approximately 67,500 (tropoe- lastin I), 65,000 (tropoelastin 11), and 62,000 (tropoe- lastin 111). Three tropoelastin polypeptides with molec- ular mass 2-3 kDa higher than their corresponding tissue forms were also evident in cell-free translation products of ligamentum nuchae RNA, suggesting that each tropoelastin species is encoded by a unique mRNA. The presence of cysteine in all three tropoelas- tin isoforms was demonstrated by the incorporation of [35S]cysteine into newly synthesized tropoelastin poly- peptides and by immunoreactivity with an antibody raised against a synthetic peptide that defines the cys- teine-containing carboxyl-terminal region of tropoe- lastin. Immunological co-localization of the carboxyl- terminal antibody with insoluble elastin in lung vas- culature and parenchyma suggests that intact tropoe- lastin and not a processed form is incorporated into the elastin fiber.

Unlike the collagens, which have been characterized as a multigene family (Fleischmajer et al., 1985), elastin’s unique physical properties have precluded a thorough description of the complexities of the elastin phenotype. Despite accumulat- ing evidence for multiple forms of tropoelastin, questions still remain as to their precise number, the structural relationships between them, and their cellular origin.

Elastin is a polymeric protein composed of individual tro- poelastin molecules cross-linked one with another to form a functional elastomer. Most of what is known about elastin structure is based on biochemical and physiochemical char- acterization of uncross-linked tropoelastin. First isolated from elastic tissues of copper-deficient or lathyritic animals, the elastin precursor was identified as a single, nonglycosylated protein with a molecular weight of approximately 74,000

* This work was supported by National Institutes of Health Grants HL-26499, HL-29594, AM-20553, and Training Grant HL-07317. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

I( To whom correspondence should be addressed Jewish Hospital- Pulmonary Research, 216 S. Kingshighway, St. Louis, MO 63110.

(Sandberg et al., 1969; Smith et al., 1972). It wasn’t until Foster et al. (1980, 1981) described two putative elastin prod- ucts (tropoelastin a and b) in cell-free translation of chick aorta and lung RNA that the existence of multiple tropoelas- tin gene products was suggested. Since then, the presence of at least two soluble elastin forms has been substantiated in in vitro systems (Rich and Foster, 1984; Davidson et al., 1982; Mecham et al., 1985; Chipman et al., 1985; Saunders and Grant, 1985).

TO investigate more thoroughly the genetic complexity of elastin at the protein level, we utilized monoclonal and poly- clonal antibodies to elastin combined with high resolution gel electrophoresis to characterize tropoelastins secreted by bo- vine cells and tissues. Our results indicate that bovine cells synthesize three tropoelastins that are products of unique mRNAs. All three tropoelastins contain cysteine and can be identified in tissue extracts, cell culture medium, and in cell- free translation products of bovine ligamentum nuchae RNA.

EXPERIMENTAL PROCEDURES’

RESULTS

Tropoelastin from Tissue Minces-Fig. 1 shows the elution profile obtained from c18 reverse phase HPLC’ of tropoelas- tin extracted from ligament tissue minces. Elastin-specific enzyme-linked immunosorbent assay and amino acid analysis (Table I) established that the protein material eluting at 12.2 min was tropoelastin. Electrophoresis of the purified product on standard Laemmli gels (Laemmli, 1970) (10% cross-link) produced a single broad band at a molecular weight of ap- proximately 67,000 (Fig. 2, lane B). On ultrathin, high reso- lution gradient gels, however, tropoelastin was resolved into three bands (Fig. 2, lane C ) with approximate molecular weights (relative to globular molecular weight standards) of 67,500 (tropoelastin I), 65,000 (tropoelastin 11), and 62,000 (tropoelastin 111). The relative migration of the three tropoe- lastins was unaffected by reducing agents and the identifica- tion of tropoelastin I was highly dependent upon the resolving properties of the gel system. In many instances, tropoelastin I co-migrated with tropoelastin I1 or was resolved only as a

’ “Experimental Procedures” are presented in miniprint a t the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 86M-2771, cite the authors, and include a check or money order for $1.20 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

The abbreviations used are: HPLC, high performance liquid chro- matography; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline (Miniprint); BSA, bovine serum albumin (Miniprint).

2244

Tropoelastin Multimers 2245

faint band slightly above but not completely separated from tropoelastin 11. Best resolution was obtained when gels were used within 24 h of preparation.

The immunoblots in Fig. 2, lunes D and E, demonstrate that all three tropoelastins are reactive with polyclonal- and monoclonal-derived anti-elastin IgG.

Tropoelastin from Cell Culture-To determine whether the three tropoelastins extracted from ligament tissue are secre- tory forms of elastin, synthetic products of ligamentum nu- chae fibroblasts and ear cartilage chondrocytes where radio-

D

n rr n

FIG. 1. Analysis of bovine ligament tropoelastin by C18 reverse phase HPLC. Tropoelastin extracted from ligament minces was dissolved in water, 0.1% (v/v) trifluoroacetic acid (startingbuffer) and resolved on a Ultrasphere-ODS column (5 pm, 25 cm X 4.6-mm inner diameter) with a RP-18 MPLC NewGuard Column (Pierce Chemical Co.) using a linear gradient of from 0 to 60% acetonitrile over 20 min. At 26 min the acetonitrile concentration was raised to 80% over a 2-min period and at 30 min the column was re-equilibrated with starting buffer. Buffer flow rate was 1 ml/min and the column effluent was monitored at 228 nm with a detector range of 0.5.

TABLE I Amino acid composition (residues/lOOO) of purified ligamentum

nuchae tropoelastin and insoluble elastin Amino acid Trouoelastin Elastin"

Hydroxyproline Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Cystine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Isodesmosine Desmosine Histidine Hydroxylysine Lysine Arginine

I* 12 11 18 21

131 301 212

4 121

0 25 62

7 26 0 0 0 0

44 3

8 5 8 9

16 108 327 220

0 154

0 24 65 8

30 2 2

0 8 4

Trace

a From Whiting, et al. (1974). Low hydroxyproline values reflect the absence of ascorbate during

the incubation period.

A B C D E F G IF ' W

W- 205 KDa

205 KDa-

97.4 6 8 K D a - i KDa - - ' ,,'1 # , - 97.4 KDa , .

E P = . . ] -

'+ 911-68 KDa

: 119- 45 KDa 45 KDa -

29 K D a - C ' , [ i

- 29 KDa

FIG. 2. SDS-polyacrylamide gel electrophoresis of purified tropoelastin. Bovine tropoelastin purified from ligamentum nuchae was analyzed by electrophoresis using a Laemmli SDS-polyacrylam- ide gel (10% cross-link) (lune B ) or an ultrathin, 7-12.5% gradient gel (lune C ) . Both gels were stained with silver. Lanes D and E show immunoblots of different preparations of ligament tropoelastin de- veloped with anti-elastin monoclonal IgG; lune F is with preimmune serum. Gels were run under reducing conditions. Molecular mass standards (lunes A and G ) include myosin (205 kDa), phosphorylase b (97.4 kDa), bovine serum albumin (68 kDa), egg albumin (45 kDa), and carbonic anhydrase (29 kDa).

" Medium Cell Layer

A B C D E F G

Lp ! I ! 1 ,., . - - 205 KDa

; L ' I

, 11 L I - - 92 .5KDa

1.11 , - 1 b - C KDa . . " .

- 1

2- 46 KDa

,. * -. - 18.4 KDa FIG. 3. Tropoelastins synthesized by cultured cells. Fluoro-

graphs of [3H]leucine-labeled tropoelastin immunoprecipitated with a monoclonal antibody from culture medium of ligament fibroblasts (lune A ) or auricular chondrocytes (lune B ) and chromatographed on gradient SDS-PAGE gels as described in the legend to Fig. 2. Immu- noprecipitates from the cell layer of cultured chondrocytes at 4 days (lune D) and 4 weeks (lane E ) post-confluency. Lanes C and F are immunoprecipitates from medium and cell layer extracts, respec- tively, using normal ascites IgG. "C-Labeled molecular mass stand- ards (lane G ) include myosin (205 kDa), phosphorylase b (92.5 kDa), bovine serum albumin (68 kDa), egg albumin (46 kDa), and lacto- globulin A (18.4 kDa).

labeled with [3H]leucine and tropoelastin was precipitated from culture medium with anti-elastin monoclonal IgG. Fig. 3 shows that all three tropoelastins were present in the medium of both cell types.

T o evaluate whether multiple tropoelastins are also asso- ciated with the extracellular matrix of cultured cells, ear chondrocytes were cultured overnight with [3H]leucine in the presence of P-aminopropionitrile and soluble components in the cell layer were extracted with acetic acid. Chondrocytes were chosen for this experiment because they retain a large percentage of secreted tropoelastin in the extracellular matrix (Mecham et al., 1981; Starcher and Mecham, 1981). SDS- PAGE of immunoprecipitates showed that no tropoelastin was associated with the cell layer at 4 days post-confluency but after 4 weeks of culture all three tropoelastins were present in a ratio similar to that seen for the soluble proteins

2246 Tropoelastin Multimers

in culture medium (Fig. 3, lane E ) . Quantification by scanning densitometry of numerous gel

fluorographs showed that the ratio of the three tropoelastins secreted by cultured cells was variable, even when isolated from the same cell type. Tropoelastins I1 and I11 were the major elastin products with tropoelastin I appearing as a minor product in most instances. As an average, the relative proportion of tropoelastin I, 11, and I11 secreted into culture medium by first passage cells was approximately 1:3:2, re- spectively (Fig. 4).

Tropoelastin in Cell-free Translation of Elastin RNA-To examine if the multiple tropoelastins are encoded by unique RNAs, proteins in reticulocyte lysate translations of ligament tissue RNA were analyzed by high resolution SDS-PAGE. Using guanidine thiocyanate extraction and cesium chloride density centrifugation, approximately 16.8 pg of RNA were isolated per gram (wet weight) of ligament tissue. Cell-free translation of ligament RNA gave an approximately 16-fold stimulation over background (no RNA) in [3H]leucine and [35S]cysteine incorporation into trichloroacetic acid-precipi- table protein. Electrophoresis of translation products on high resolution SDS-PAGE showed three major proteins with ap- parent molecular weights of 71,000, 67,500, and 65,000 (Fig. 5). Specific precipitation with anti-elastin monoclonal IgG confirmed that the three bands were tropoelastin. The molec- ular weights of the cell-free translation products are 2000- 3000 larger than their corresponding tissue forms (see Fig. 4).

Evidence for a Cysteine-containing Region Near the Carboxyl Terminus-The presence of cysteine in newly synthesized tropoelastin polypeptides is consistent with elastin gene struc- tural analysis that predicts a unique, cysteine-containing amino acid sequence at the carboxyl-terminal end of the molecule (Cicila et al., 1985). The presence of such a sequence

:ell Free Translation

92.5 Kd 6BKd Ph0s.B BSA

45 Kd OA

I I 1

k Culture MediumIP

Culture Medium I P FCL-270

65.000-

PROTEIN MIGRATION +

FIG. 4. Relative proportion of tropoelastin isoforms in cell- free translation or immunoprecipitation. Densitometry scans of representative SDS gel fluorographs of [3H]leucine-labeled tropoelas- tin immunoprecipitates from cell-free translation of ligamentum nu- chae RNA (top) and from ligament fibroblast culture medium a t first (middk) and third (bottom) passages.

A B C a

FIG. 5 (left). SDS-polyacrylamide analysis of ligament RNA translation products. SDS-PAGE analysis (5-10% gradient) of translation products from ligamentum nuchae RNA labeled with [93] cysteine ( l a n e A ) or [3H]leucine ( l a n e B) . Included for comparison ( l a n e C ) is [3H]leucine-labeled tropoelastin from ligament fibroblast culture medium.

FIG. 6 (right). Reactivity with carboxyl-terminal antibody.

raised against a synthetic septapeptide (GFPGGACLGKSCGRKRK) Immunoblot of purified tropoelastin developed with an antiserum

that corresponds to the last 17 amino acids at the carboxyl terminus of bovine tropoelastin ( l a n e A ) and a monoclonal antibody specific for a repeating, internal hexameric sequence in elastin ( l a n e B) . Each lane contains an equivalent loading of tropoelastin.

on tropoelastin was confirmed using a monospecific antiserum generated to a synthetic peptide with an amino acid sequence determined from nucleotide sequencing of this region of the gene? As shown in Fig. 6, all three tropoelastin bands were immunoreactive with the carboxyl-terminal antibody. Inter- estingly, this antibody reacted only with intact tropoelastin and did not recognize lower molecular weight degradation products of tropoelastin that are otherwise immunoreactive with a monoclonal antibody having specificity for an internal repeating hexapeptide region of the molecule.

To investigate whether the carboxyl-terminal peptide is associated with insoluble elastin, sections of bovine lung were stained with the carboxyl-terminal antibody and immune complexes were visualized using immunoperoxidase. Fig. 7 shows that the carboxyl-terminal antibody co-localized with elastin fibers in lung parenchyma and vasculature.

DISCUSSION

With the identification of multiple tropoelastins, much interest has centered on identifying the exact nature of these proteins. Using severe denaturing conditions to extract and purify tropoelastin from chicken aorta, Rich and Foster (1984) showed that tropoelastin a had more polar amino acids, 5-8 times the number of cysteine residues, and a higher percentage of hydroxyproline than tropoelastin b. Differences in molec- ular weight, amino acid composition, isotope incorporation, antigenic determinants, and peptide maps implies significant differences between the two tropoelastins (Karr and Foster, 1983). Other studies, however, have identified tropoelastin forms that do not differ as extensively. In chick embryo artery cells, two distinct tropoelastin polypeptides were observed only when unhydroxylated proteins were extracted from cells incubated with a,a-bipyridine or 3,4-dehydroproline (Saun- ders and Grant, 1985). It was suggested in these studies that microheterogeneity induced by partial and random hydrox-

Rosenbloom, J., Weinbaum, G., Ornstein-Goldstein, N., Indik, Z., and Kucich, U. (1987) Collagen Relat. Res., in press.

Tropoelustin Multimers 2247

FIG. 7. Co-distribution of carboxyl-terminal epitope with lung elastin. Sections of neonatal bovine lung were reacted with non-immune rabbit IgG (panel A ) , anti-elastin monoclonal IgG (panel B) , or anticarboxyl- terminal IgG (panel C ) . Bound antibodies were visualized by immunoperoxidase staining as described under “Experimental Procedures.” Stained elastin fibers in pulmonary artery adventitia (wide arrow) and alveolar septum (thin arrow) are identified in panels B and C. Magnification of original figure was 125X.

ylation of the proteins in normal cultures was sufficient to obscure differences in their electrophoretic mobilities. Peptide mapping of the underhydroxylated tropoelastins indicated close identity between the two molecules with minor differ- ences best explained by the presence of a peptide sequence in tropoelastin a that is absent in tropoelastin b.

These studies with chicken tissues suggest that tropoelastin a and tropoelastin b represent distinct gene products. In rat aorta, however, multiple tropoelastins may result from pro- teolytic processing of a single secreted tropoelastin precursor. Chipman et al. (1985) identified a single 77,000-dalton tro- poelastin-like molecule in cultures of rat aortic smooth muscle cells that appeared to be processed extracellularly to a 71,000- dalton form. Such a precursor-product relationship between the two proteins raises the possibility that, even in other systems, tropoelastin b is a processed form of tropoelastin a.

Using high resolution SDS-PAGE and antibodies specific for elastin we have identified three tropoelastin isoforms with molecular weights of 67,500, 65,000, and 62,000 in tissue extracts of ligamentum nuchae as well as in culture medium and cell-matrix extracts of ligament fibroblasts and chondro- cytes from auricular cartilage. On standard Laemmli gels, the three isoforms co-migrated as a single broad band at M, = 66,000. The ratio of tropoelastin I, 11, and I11 (approximately 1:3:2, respectively) was similar in first passage cultures of each cell type but fluctuated unpredictably as the cells aged in culture or underwent repeated cell division.

Evidence that the tropoelastins represent unique secreted proteins is supported by the following observations. First, tropoelastin I, 11, and I11 were identified as cell-free transla- tion products of ligament tissue RNA. All three proteins were

approximately 2-3 kDa larger than the tissue forms, which is consistent with the presence of a “signal” peptide on the amino terminus of each molecule, as has been established for chicken (Foster et al., 1981) and sheep (Davidson et al., 1982) tropoelastins. Thus, each protein appears to have been en- coded by a unique RNA. Second, the absence of hydroxylation and glycosylation reactions in cell-free systems argues against these modifications as explanations for the observed differ- ences in electrophoretic mobilities. Third, each of the three tropoelastins reacts with antiserum having specificity for a sequence at the carboxyl terminus of the molecule. If molec- ular weight differences between the three proteins were to arise by proteolytic removal of this carboxyl-terminal se- quence, tropoelastin I1 and I11 would not react with the antibody.

Available physical and chemical information do not yet explain the observed variation in size or chromatographic behavior of the three proteins. Because the molecules have an intact amino and carboxyl terminus, differences between the proteins must be internal to these sites. When subjected to isoelectric focusing, all three tropoelastins migrated as a single band at the basic end of the pH gradient (not shown), con- firming that structural differences between the isoforms do not include alterations which change the overall net charge of any one of the molecules. Furthermore, the co-migration of all three tropoelastins as a single peak on reverse phase HPLC suggests that the proteins have similar physical prop- erties.

In immunoprecipitation and Western blot studies with monoclonal antibodies, it was observed that the characteristic degradation products of tropoelastin (Mecham and Foster,

2248 Tropoelastin Multimers

1977) often migrate on SDS-PAGE as trimers with molecular weight differences identical to those observed between the parent proteins. This fortuitous form of peptide mapping supports the existence of common protease-susceptible sites and large segments of sequence homology between the three isoforms.

The interaction of all three tropoelastins with anticarboxyl- terminal antibody establishes that at least a portion of this novel sequence, predicted from nucleotide sequence analysis of the elastin gene, and extending from the carboxyl terminus to residue -17 (Cicila et al., 1985), is present on the molecule even though a corresponding sequence has not been found in tryptic peptides from tropoelastin (Sandberg et al., 1985). The distinctive features of this sequence include the presence of cysteine residues and a clustering of highly basic amino acids. While the basic amino acids may facilitate the interaction between tropoelastin and acidic components of elastic fiber microfibrils, the migration of tropoelastin as a trimer on SDS- PAGE under nonreducing conditions and the ease with which tropoelastin can be extracted from ligament tissue in the absence of reducing agents suggest that the 2 cysteine residues form an intramolecular bond or exist as free sulfhydryls. However, we cannot exclude the possibility that the formation of intermolecular disulfide bonds between tropoelastin mon- omers or between tropoelastin and other components of the extracellular matrix is a time-dependent process such that extractable tropoelastin represents a pool that has not yet undergone disulfide linkage.

Conclusions about the role of the carboxyl-terminal seg- ment of tropoelastin must remain speculative but since the sequence remains associated with insoluble elastin, one can speculate that the carboxyl-terminal region might be impor- tant for some phase of fibril formation. It is interesting that only intact tropoelastin I, 11, and I11 and not the characteristic lower molecular weight degradation fragments react with the carboxyl-terminal antibody (Fig. 6). These findings, together with the demonstration of intact tropoelastin associated with the extracellular matrix of cultured cells and tissues, suggest that only intact tropoelastin and not a form that is processed at the carboxyl terminus becomes incorporated into the elas- tin fiber.

It has yet to be established whether the three tropoelastin arise from the same or different genes. In situ hybridization experiments using a human cDNA clone and human meta- phase chromosomes have identified only one elastin gene locus in the human genome localized to the q 3 1 q t e r region of chromosome 2 (Emanuel et al., 1985). If the tropoelastin variants are encoded by a single elastin gene, then differences must arise from alternative splicing of the primary transcript. Preliminary S-1 mapping experiments have suggested the occurrence of alternative splicing events in elastin mRNA isolated from bovine ligamentum nuchae (Yoon et al., 1984). Clearly, further studies are required to establish the genetic origins of the three bovine tropoelastins.

Acknowledgments-We thank Gertrude Crump, Dale Pollo, and Michele Bergfeld for excellent technical assistance.

REFERENCES Chipman, S. D., Faris, B., Barone, L. M., Pratt, C. A., and Franzblau,

C. (1985) J. Biol. Chem. 2 6 0 , 12780-12785 Chirgwin, J. M., Przybyla, A. E., McDonald, R. J., and Rutter, W. J.

(1979) Biochemistry 18,5294-5299 Cicila, G., May, M., Ornstein-Goldstein, N., Indik, Z., Morrow, S.,

Yeh, H. S., Rosenbloom, J., Boyd, C., Rosenbloom, J., and Yoon, K. (1985) Biochemistry 24,3075-3080

Davidson, J. M., Leslie, B., Wait, T., Crystal, R. G., and Sandberg, L. B. (1982) Arch. Biochem. Biophys. 218,31-37

Emanuel, B. S., Cannizzaro, L., Ornstein-Goldstein, N., Indik, Z. K., Yoon, K., May, M., Oliver, L., Boyd, C., and Rosenbloom, J. (1985) Am. J. Hum. Genet. 37,873-882

Fleischmajer, R., Olsen, B. R., and Kuhn, K. (1985) Biology, Chem- istry, and Pathology of Colhgen, New York Academy of Sciences, New York

Foster, J. A., Rich, C. B., Fletcher, S., Karr, S. R., and Przybyla, A. (1980) Biochemistry 19,857-864

Foster, J. A., Rich, C. B., Fletcher, S., Karr, S. R., DeSa, M. D., Oliver, T., and Przybyla, A. (1981) Biochemistry 20,3528-3535

Johnson, D. A., Gautsch, J. W., Sportsman, J. R., and Elder, J. H. (1984) Gene (Amst.) Anal. Tech. 1, 3-8

Kagan, A., and Glick, M. (1979) in Methods of Human Radwimmu- noassay (Taffe, B. M., and German, H. A., eds) pp. 328-329, Academic Press, Orlando, FL

Karr, S. R., and Foster, J. A. (1983) Collagen Relat. Res. 3,459-467 Laemmli, U. K. (1970) Nature 227,680-685 Matsudaira, P. T., and Burgess, D. R. (1978) Anal. Biochem. 87,386-

Mecham, R. P., and Foster, J. A. (1977) Biochemistry 16,3825-3831 Mecham, R. P., Lange, G., Maderas, J., and Starcher, B. C. (1981) J.

Mecham, R. P., Madaras, J., McDonald, J. A., and Ryan, U. (1983)

Mecham, R. P., Morris, S. L., Levy, B. D., and Wrenn, D. S. (1985)

OFarrell, P. Z., Goodman, H. M., and OFarrell, P. H. (1977) Cell

Quintarelli, G., Starcher, B. C., Vocaturo, A., Di-GianFilippo, F., Gotte, L., and Mecham, R. P. (1979) Connect. Tissue Res. 7 , l - 9

Rich, C. B., and Foster, J. A. (1984) Biochem. J. 217,581-584 Sandberg, L. B., and Wolt, T. B. (1982) Methods Enzymol. 82,657-

Sandberg, L. B., Weissman, N., and Smith D. W. (1969) Biochemistry

Sandberg, L. B., Leslie, J. G., Leach, C. T., Alvarez, V. L., Torres, A.

Saunders, N. A., and Grant, M. E. (1985) Biochem. J. 230 , 217-225 Smith, D. W., Brown, D. M., and Carnes, W. H. (1972) J. Bwl. Chem.

Starcher, B. C., and Mecham, R. P. (1981) Connect. Tissue Res. 8,

Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc Natl. Acad. Sci. U. S. A. 76,4350-4354

Wrenn, D. S., Griffin, G. L., Senior, R. M., and Mecham, R. P. (1986) Biochemistry 25,5172-5176

Yoon, K., May, M., Goldstein, N., Indik, Z. K., Oliver, L., Boyd, C., and Rosenbloom, J. (1984) Biochem. Biophys. Res. Commun. 118 ,

Whiting, A. H., Sykes, B. C., and Partridge, S. M. (1974) Biochem. J.

396

Cell Biol. 90,332-338

J. Cell. Physiol. 116,282-288

J. Biol. Chem. 259 , 12414-12418

12,1133-1142

665

8,2940-2945

R., and Smith, D. W. (1985) Puthol. Biol. 33 , 266-274

247,2427-2432

255-258

261-269

141,573-575

Tr Supplementary Material to

ldentificatlon of bhdl#ple Tropoelastlns Secreted by Bovine Cells

Davld S. Wrenn, William C. Parks. Loren A. Whaehouse. Edmond C. Cmuch Umberto Kucich. Joel Rosenbbom. and Roben P. Mecham

EXPEAIMENTN PROCEDURES

!&JJQ@: Fibmblasts from &amenturn nuchae were gmwn from explants of fetal Wvlne tlsue as descnbed (Mecham a al.. 1961) Chondmcyfes from b w n e ear cartilage were isolated essenttallv as descnbed bv Qumarelh et ai (1979). Cubes-were mantdned In Dulbecm's modlBed Eagle's medium (DMEM) supplemented with antibiotics, nonessential amlno adds. and 10% (vlv) fetal bovlne serum.

Incubated owmight in leucine- or cystsinedeficient medlum containing 10 -: Cell culture^ were

pCllml L-j4,5-3Hlleuc1ne (Amersham. Chlcego. IL , 120 Cdmmol) or t pCi/ml L-j35SIcysteine (New England Nuclear. Bosion, MA, 7M) Olmmol). respec.

tlvely. and 10% fetal bovine serum lmmunoprec~pitmion of tropoelastin from mndltloned medium and the cell layer was as descnbed (Mecham et al.. 1963) uslng either a polyclonal antlserum agalnst bovlne llgament alpha-elastln or a monoclonai ambody Wtth speafiuty tor one 01 the repeatmg sequenm rsgions of the eiastln miecuie (Wrenn et al.. l966). The predpbtated samples were resolved on 0 75mm thick. 10% sodturn d o d q l sulfate (SDS) polyacrylamlde gels amrding 10 Laemmb (1970) or on 0 45 mm thick. 7.5%-12% SDS polyacrylamlde gradlent gels as descnbed by Matsudatra and Burgess (1976) For tluorography. gels were lmprsgnated With ENHANCE (New England Nuciear. Eo51on, MA) and exposed to Kodak XAR-5 film at -70°C

-The septadeea psptids (Giy-Phe-Pro.Gly-Gly-Ala-Cys-Leu-Gly Lys-Ser-Cys-Gly.Ag-Lys-Arg-Lys) correspondmg to the cartaxy-termtnal sequence of bovlne tropoelastin was syntheszed by Penlnsula Laboratones, Inc , Belmont. CA. and was shown to have the mneu ComPoSitlon bv amino acrid analysis. The synthellc peptide was coupled to keyhole limpet-hemo- cyantn (KLH) usmg giuleraldehyde as described by Kagan and G l i a ((979)

couplad or uncoupled pepllde by mdl-site subcutaneous inpunon. The first Five 10 elght pound New Zealand Whm rabbts were immunmd wth ellher the

lnleclton consisted of immunogen emuis~fied tn an equd volume of complete Frevnds adwant and twe subsepuent iniedcons substtluted \nmmplete

equivalem of 300 01 peptide mlecled every two weeks for three months Freund's adjuvant to minimme ik&l sbn necmsrs Each rabbt recelved the

Anmals were bled by ear vem pumure and serum collected by centrifugalton The IgG fractm was prepared by (NH&SO, precipitation foilowed by DEAE Chmmatography

a Mliwam tissue chopper The minced t l w e was incubated overntght at 37OC In DMEM suF9lemented wlh 100 -Iml fl-amlnopmpionnrile (BAPN), 50 U m l peniallamine, 40 w/ml gentamyun. 1nM dexamethasone. 30 mM N-2-hydmxy- ethylp~peraLin~-N-2-eth~es~lto~ic add (HEPES) and W. faal bovlne 5eNm Tmwlest ln was enraued tmm t ~ s u a alter washing w8h t c 8 m l d water to

contaning 25 W/ml pepstatm and by ovemlght stimng at 4% Insoluble remove serum mmpanems by thorouph homogeniratlon I" 0 5N acetic acid

maenal was removed by cemrilugat~on at 16.WO g lor 30 min at 40C and the Supernatant was adluaed to neutral pH by the dropwise addition of 10 N NaOH Solube proteins In the supernatant were precipitated Overnight at 40C With 45% ammonium sulfate and were pelleted by centrilfugat~on a 16.000 g tor

mnta!ning 5 mM EDTA, 5 mM E-Bm~nOeBprotc acid (EACA), 5mM benr- 30 min. 4% me pellet was d i r r o b d in 5mlot 0.25 M sodium acetate. pH 5.

amidine. 2mM phenyimethykunonyl tluollde (PMSF) and dlalyred agamst 4

dlalyss was removed by centnfugalon as above and lropoe1aslin was snraued lhters (3 changes) of the Same buner. Insoluble material that developed during

Imm the supernatant by the dmpwsa addinton 01 1.5 volumes pmpanoi and 2.5 volumes butanol as descnbed by Sandbeg and Won (1962) As a final punficatlon step. the pmpanol-butanol enrad was s u b w e d lo revem phase chromatography On an Ultrasphere-ODS mlumn (5p. 25cm x 4 6mm I D.) usmg a Bedrman Model 334 HPLC system

. Tmpoelastin was transferred

el. at. (1979) Transfer was at tOOV fof l h a 4OC. NarwIIuiose mntamng -the transfer procedure 01 Towt~n

transferred protein was washed for 30 min at room temperature in Tns-bunered sallne VBS. tOmM Tris. 150 mM NaCI. pH 7.4) mntaining 1 mgml nonfat dried mllk (BLOTTO) (Johnson el ai.. 1984) to block unbound bmding sms. The nltmceilulose vms then washed W8ce with TBS (15 mmlwash) on an orbital

*opoelastin Multimers shaker and ~ncubaled for t h at mom temperature with 5 mp monoclonal IQG or with a t :to0 ddution 01 polycbnal antiserum. The bb l was then washed three times as above with TBWELOTTO followed by a one-hour incuballon at mom temperature with protein A-mupled peroxidase (5 Fglml in TESBLOTTO) After washing three times w lh TBS. treshly made 4chloro-i-naplhol solution (70 mi methanol, t ml of t M po1Bssium citrate. pH 6. 38 ml Waterand 40 p1 of 30% hydmgen peroxide) was added. The mlor reaclion was stopped by nnsmg wlth waer and drying the blot on finer paper

-: Radiolabeled tmpoelestin enracted tmm Ihgamenl ttssue mmms or lmmunoprecipitated from mnddoned culture medium vms funher charaaenred by non-equil#brium iSoeieUnc focusrng In amphollnes pH3-11 (Elo-Rad. Rohmond. CA ) at ZWV tor 4h .s desctibed by OFamll et a1 (1977).

fmm near-term bovine fehrses using a modificatm of the p r d u r e of Chtgwm

pulwenred. and than homogenized with a Bnnkman PoMmn in 4 volumes (wh) GSCN buner (4 M guanidine thiqanate. 2% sahosyi. 25 mM sodium cltrate. pH 7.0, 0.1 M B-mercaptoethanol, O.t% antifoam agent). The enrau was centnfuged at t0,oOOg at 40C tor 20 min The pellet was rehomogenized in t volume (w&) GSCN tuner. and the enraci mnlotuged as before. Supernatants were mmblned. adjusted to 0.5 gml CsCI, layered over 9 ml CsCl pads (5.7 M CsCl in to0 mM EDTA. pH 7) and cerntifuged 81 114.000 gat 20% tor 16h In a Beckman SW26 rotor. The supmalane were removed by asplratlon,and the tubes were thomuphly dried with steriie gauze. RNA pellets were susppended In I ml pertube TES ( I O mM Ttis-HCI. pH 7.5.5 mM EDTA. 1% SDS), combned and enracted once wllh an equal volume of chbm1orm.t- butanol ( 4 1 ) . The organlc phase was back-enmed with a minimal volume of TES. The aqueous phases were mmblned. and the RNA was predpitaed with 1 l t O volume 3 M sodium m e , pH 5.2. and 2.2 Wiumes 95% ethanol at -22% tor 2 h. RNA w86 mllecled by cenlritllgwon a1 10.004 9 tor 30 mm at 4OC. dried under vacuum, d i s s o M in water and repreclpnated with sthanol as described. absorb an^^ ratios at 260 and 280 nm were determined and samples were suspended in water at 2 W m l and stored a1 -7OOC

reaculocyte lysate (Amersham. Chicago. IL) wlth either L-[3H]leucine or p: Purified RNA was translated uslng rabbit

L-[35S]cptsine Each reanion was inarbald for 1 h a1 370C and mntsned 38 PI reticulocyte lysaae. 6 MI 01 label!& amino add (6 pCi L-[sH]leucine or 81.6 pCi L-pSlcyneine) and 4 U of RNA wubstrate or water. Reaniom were terminated by puning on ics. For detem~net~on of inmrpDmed L-[3H]leuane. 2 p1 allqyots were bioned orno Whaman 541 paper, air dried and boiled tor 10 min m 10% t6chbmacsic scd (TCA). The filters were then washed two times wnth lcecold 5% TCA then tW% ethanol, dried all WOC for 5 min and munted for bound RldlMc(ivity using Eudoet S o h (Resaanh Produus Inter- national. MI. Pm-n. IL).

RNA was punfied from tresh hganentum nuchae

et al. (1979). Fresh tiSSuB (about 35 9) Was frozen directly m liquid nitrogen.

For Samples labeled with L-[3~S]cystelne. 2 PI aliquots were added lo 70 PI ot t OM urea. 0.1 M 8-mercaptmthanol and heated at 1OoOC for 2 mm. Then 20 p1 of t M iodoBcBhc aod was added. and the samples were funher mcubated at 37% for 30 mln followed by the addmn of 2 p1 B-mercapto- ethanol. Aiiquots (10 &I) were transferred to clean tubes. and proteins were prec~pitated with t rnl 10% TCA at tOOoC for 10 mi" PrBcip!lates were mllecled on GFlC filter 6164 and washed with ice-mld 5% TCA and 100% ethanol. Dned filters were munted as descnbed. Tropoelast~n in the remsnmg translation sample was ammunopreapltated using antielasfin antibadms

Sections wers deparaffinized and rehydrated. Endogenous peroxidase was buffered tormalh and embedded In paramn. For immmohistOChem~stry. 5&

blocked by treatmem vlth 0 8% (vhl Hz02 In PES for 10 min at mom iemparatum. Seeions were also incubated with 0.1% (w&) trypsm (Sagma.

antlgenlc determinants Non-spedflc immunoglobulin btnding sites were Type 11. porcine pancreas) In PBS, pH 7.4. for t o mm at 220 to unmask

blocked wtlh normal goat serum Sealons were then incubated fn a motst chamber for 2h at 37OC with mouse antibovine a-elastin IgG (25 Wlml). or

mnslsted 01 parallel SmIOnS incubated wolh normal mouse or rabbit IgG as rabbit antlbovlne cartaxoxy-termlnal peptnde IgG (100 wglml). Negative controls

Washed SeRions were Incubated for 20 mm wnh aWmty-puritted btolin- required. The remainlng incubations were performed at room temperature

conjugated goat antimouse (ERL: 1.1600 in PES mmaindng 1% BSA) or goat antirabbt IgG (ERL: 1:400). washed. and lncubated lor 20 min wlth streptavldln-

vlruallzed by inwbat~on for 20 minutes with 3.3'diaminobenridme (Sigma, horseradish pemxldase (ERL. t:400. HRP:SA 01 10'1). Complexes were

munterstaned with Hams hematoxyiln. dehydrated, mounted I" Permount, and 0.5 mglml In 50 mM TriS-HCI, pH 7.4) and HzOz. Secl~ons were washed,

examlned by hght mlcmsmpy.

-: Bovine tissues were fixed I" 10% neutrel-

22 149