The structure of type IX collagen

We present a detailed analysis both of tryptic peptides and amino-terminal sequences of the subunits of two collagenous fragments (HMW and LMW) previously isolated from pepsin extracts of chicken cartilage (Reese, C.A., and Mayne, R. (1981) Biochemistry 20, 5443-5448). This analysis and a comparison with the nucleotide sequence of the cDNApYN1738 (Ninomiya, Y., and Olsen, B.R. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 3014-3018) shows that HMW and LMW are pepsin-resistant fragments of a unique collagen composed of molecules with three different polypeptide chains (alpha-chains). This collagen has been assigned the type number IX, and the alpha-chain encoded by pYN1738 has been given the designation alpha 1 (IX). Type IX collagen contains three triple-helical domains and at least two sets of interchain disulfide bridges. At the amino and carboxyl ends are noncollagenous domains which do not appear to be homologous to amino and carboxyl propeptides of interstitial collagens.     

Elongation of the interchain disulfide bridges of insulin. A synthetic analog.

The sythesis and isolation in purified form of an analog of insulin with the interchain disulfide bridges elongated by a methylene group is described. This analog differs from the parent molecule in that the cystein residues occupying positions A-7 and A-20 and involved in the formation of the two interchain disulfide bridges of insulin have been replaced by homocysteine residues. For the synthesis of this compound the Hcy-7, 20-A chain of sheep insulin was chemically synthesized and isolated in the S-sulfonated form. Conversion of the latter product to the sulfhydryl derivative and combination with the S-sulfonated form of the B chain of sheep insulin yielded the [Hcy-7, 20-A] insulin. Isolation of the analog from the combination mixture was effected by chromatography on a carboxymethylcellulose column with acetate buffer (pH 3.3) and an exponential sodium chloride gradient. This analog, by the mouse convulsion assay methods and in doses at least 40-fold higher than those normally used for insulin assay, was inactive. By the radioimmunoassay method this synthetic analog was found to possess a potency of 2 i.u./mg. It is concluded that the biological activity of insulin depends critically on a particular geometry conferred on the molecule by the proper placement of the A and B chains.     

Dissociation of bovine seminal ribonuclease into catalytically active monomers by selective reduction and alkylation of the intersubunit disulfide bridges.

The hypothesis previously advanced that interchain disulfide bridges link the two identical subunits of bovine seminal ribonuclease BS-1 has been confirmed. The sedimentation rate and the electrophoretic mobility of the protein are not affected by denaturing agents unless thiol reagents are present in the denaturation mixtures. Reduction under controlled conditions results in the immediate cleavage of only 2 disulfide bonds out of 10 percent in the dimeric protein. Under these conditions, and the results do not change when partial reduction is followed by S-alkylation, 30% of the protein dissociates, while the remaining is found to consist of a dimeric species easily dissociable by denaturing agents without addition of thiol reagents. This indicates that the dimeric structure of seminal ribonuclease is maintained not only by disulfide bridges, but also by noncovalent forces. The protein derivative prepared by selective reduction and alkylation has been identified as monomeric bis-S-carboxymethylcysteine-31,32-ribonuclease BS-1. This is on the basis of the characterization of the 14C-labeled S-carboxymethylated peptides isolated from a thermolytic hydrolysate of the derivative prepared with iodo-2-[14C]acetic acid. Monomeric, selectively alkylated ribonuclease BS-1 is stable and catalytically active. The importance of such a derivative is discussed both in the light of the recent studies on the biological actions of seminal ribonuclease and as the fourth component of an experimental system of ribonucleases consisting of two homologous dimers (bovine seminal ribonuclease BS-1 and dimerized bovine pancreatic ribonuclease A) and two homologous monomers (ribonuclease A and the monomeric derivative of ribonuclease BS-1.     

Expression of A Chain and B Chain of β-Bungarotoxin from Taiwan Banded Krait: The Functional Implication of the Interchain Disulfide Bond Between A Chain and B Chain

-Bungarotoxin (-Bgt), the main presynaptic neurotoxin purified from the venom of Bungarus multicinctus, consists of two dissimilar polypeptide chains, the A chain and B chain, cross-linked by an interchain disulfide bond. The A and B chain cDNAs were subcloned into expression vectors pT7-7 and pET20b(+), respectively, and transformed into Escherichia coli strain BL21(DE3). The expressed protein was isolated from the inclusion bodies of E. coli and subjected to refolding into its folded structure. The yields of the refolded A and B chains increased markedly by at least 100-fold after substituting Ser for Cys15 of A chain and Cys55 of B chain, which formed an interchain disulfide bond. Either the A(C15S) chain or B(C55S) chain alone or in combination cannot exhibit the phospholipase A₂ activity or synaptosome binding activity of -Bgt. Nevertheless, the results of competitive enzyme-linked immunoassay, CD spectra, and fluorescence measurement revealed that the A(C15S) chain and B(C55S) chain possessed a native-like structure like the subunits of native -Bgt. Moreover, the interfacial interaction between the A and B chains explored by glutaraldehyde cross-linking revealed the essential aspects of the intact interchain disulfide bond in this interaction. This suggests that the formation of the interchain disulfide bond should not be a crucial step for the formation of folded A and B chains in the venom glands, and that the integrity of the interchain disulfide linkage favors the subunit interaction that consequently fulfills the functional mechanism of -Bgt.     

Location of disulfide bonds within the sequence of human serum cholinesterase

Human serum cholinesterase was digested with pepsin under conditions which left disulfide bonds intact. Peptides were isolated by high pressure liquid chromatography, and those containing disulfide bonds were identified by a color assay. Peptides were characterized by amino acid sequencing and composition analysis. Human serum cholinesterase contains 8 half-cystines in each subunit of 574 amino acids. Six of these form three internal disulfide bridges: between Cys65-Cys92, Cys252-Cys263, and Cys400-Cys519. A disulfide bond with Cys65 rather than Cys66 was inferred by homology with Torpedo acetylcholinesterase. Cys571 forms a disulfide bridge with Cys571 of an identical subunit. This interchain disulfide bridge is four amino acids from the carboxyl terminus. A peptide containing the interchain disulfide is readily cleaved from cholinesterase by trypsin (Lockridge, O., and La Du, B. N. (1982) J. Biol. Chem. 257, 12012-12018), suggesting that the carboxyl terminus is near the surface of the globular tetrameric protein. The disulfide bridges in human cholinesterase have exactly the same location as in Torpedo californica acetylcholinesterase. There is one potential free sulfhydryl in human cholinesterase at Cys66, but this sulfhydryl could not be alkylated. Comparison of human cholinesterase, and Torpedo and Drosophila acetylcholinesterases to the serine proteases suggests that the cholinesterases constitute a separate family of serine esterases, distinct from the trypsin family and from subtilisin.     

Extraordinarily stable disulfide-linked homodimer of human growth hormone

Although a 22-kDa human growth hormone (hGH) is the predicted protein product of the hGH-N gene, a pleiotropic collection of uncharacterized molecular weight and charge isoforms is also produced. Using chromatography and preparative SDS-PAGE under reducing conditions we isolated an unusually stable mercaptoethanol-resistant (MER) 45-kDa hGH. A 5-h incubation at 100 degrees C in the presence of 2-mercaptoethanol was required to convert approximately 90% of MER-45-kDa hGH into a 22-kDa hGH. Other reductants were not as effective in splitting MER-45-kDa hGH. After fracturing MER-45-kDa hGH, the 22-kDa hGH fragments would spontaneously reassociate if the reductant was removed; however, alkylation of cysteine residues prevented their reassociation. Identical amino acid sequences for the first six N-terminal residues were obtained for MER-45-kDa hGH and its 22-kDa hGH cleavage product. Structural identity of MER-45-kDa hGH and 22-kDa hGH was demonstrated by MALDI-TOF mass spectrometry of tryptic digests. MER-45-kDa hGH did not break up upon incubation with EDTA and EGTA. The significance of this work to our understanding of the structure of hGH isoforms is that it demonstrates that MER-45-kDa hGH is not a single chain polypeptide but is instead a homodimer of 22-kDa hGH monomers. The MER-45-kDa hGH dimer is held together by interchain disulfide bonds and not by divalent metal cation bridges. Additionally, MER-45-kDa hGH's interchain disulfide links are exceptionally resistant to reducing agents and thus confer extreme stability to the homodimer.     

Isolation and characterization of pepsin-treated type III collagen from calf skin.

Calf skin collagen was solubilized by incubating acid-extracted calf skin with pepsin at pH 2.0 and 25 degrees C, conditions that did not cause degradation of the triple helical region of collagen. Type III collagen was separated from type I collagen by differential salt precipitation at pH 7.5. The isolated type III collagen contained mainly gamma and higher molecular weight components cross-linked by reducible and/or non-reducible bonds. The isolated alpha1 (III) chains had an amino acid composition characteristic of type III collagen. Denatured but unreduced type III collagen, chromatographed on carboxymethyl-cellulose, eluted in the alpha 2 region, while after reduction and alkylation the alpha1 (III) chains eluted between the positions of alpha1 (I) and alpha2. The mid-point melting temperature temperature (tm) of type III collagen (35.1 degrees C) in a citrate buffer at pH 3.7 was somewhat lower than that of type I collagen (35.9 degrees C). Renaturation experiments at 25 degrees C showed that denatured type III collagen molecules with intact intramolecular disulfide bridges (gamma components) reform the triple helical structure of collagen much faster than reduced and carboxymethylated alpha1 (III) chains.     

The crystallizable human myeloma protein Dob has a hinge-region deletion.

During experiments to prepare heavy-metal derivatives of the crystallizable human IgG1 (k) immunoglobulin Dob, it became apparent that this protein has several unusual features. (1) Instead of the four labile interchain disulfide bridges ordinarily found in IgG1, the Dob protein has only a single interchain disulfide bridge, which connects its two light chains. (2) The Dob heavy chain appears to be slightly smaller than a control gamma1 chain, as judged by polyacrylamide gel electrophoresis in sodium dodecyl sulfate and by gel filtration in guanidine. (3) The Dob heavy chain has three fewer residues of half-cystine than expected in gamma1 chains. (4) The Dob IgG is relatively resistant to digestion with papain and trypsin; however, it is readily digested with pepsin, although at an unusual site. These findings suggest that some or all of the gamma1 hinge region is missing in Dob. To localize the deletion, we prepared an F(ab')2 fragment consisting of two heavy-chain pieces (Fd') noncovalently associated with the light-chain dimer. The Fd' piece was isolated and digested with trypsin. The sequence of the C-terminal tryptic peptide was Val-Ala-Pro-Glu-Leu-Leu-Gly-Gly-Pro-Ser-Val. Positions 2-11 of this peptide are identical with residue positions 231-240 of the gamma1 chain. The N-terminal valine could be either Val-211 or Val-215 of the gamma1 sequence. A tryptic peptide, Val-Asp-Lys-Lys, was also isolated from Dob Fd'; this sequence is not found in the variable region of the Dob heavy chain [Steiner, L. A., Garcia Pardo, A., & Margolies, M. N. (1979) Biochemistry (following paper in this issue)] but corresponds to positions 211-214 of the gamma1 constant region. Therefore, the deletion cannot include these residues and must begin after Val-215; normal gamma1 sequence resumes at Ala-231. The same 15-residue deletion has been found in two other IgG1 proteins, Mcg [Fett, J. W., Deutsch, H. F., & Smithies, O. (1973) Immunochemistry 10, 115] and Lec [Rivat, C., Schiff, C., Rivat, L., Ropartz, C., & Fougereau, M. (1976) Eur. J. Immunol. 6, 545]. Possible explanations for the occurrence of identical hinge-region deletions in three different immunoglobulins are suggested by recent experiments demonstrating that the three constant domains and the hinge region of mouse gamma1 chains are each encoded by separate segments of DNA [Sakano, H., Rogers, J. H., Hüppi, K., Brack, C., Traunecker, A., Maki, R., Wall, R., & Tonegawa, S. (1979) Nature (London) 277, 627].     

Monoclonal antibody against chicken type IX collagen: preparation, characterization, and recognition of the intact form of type IX collagen secreted by chondrocytes

A series of monoclonal antibodies was prepared against the pepsin-resistant fragment of type IX collagen designated HMW. One of these antibodies (called 2C2) was selected for further analysis. Antibody 2C2 showed no cross-reactivity with other collagen types by inhibition enzyme-linked immunosorbent assays. It recognized an epitope present in native HMW, but failed to recognize any of the three chains of HMW fractionated after denaturation followed by reduction and alkylation of interchain disulfide bridges. Electron microscopic observations after rotary shadowing showed that the location of the epitope for antibody 2C2 was close to the carboxy-terminus of HMW. Immunofluorescent staining of sections of embryonic and adult cartilage with antibody 2C2 after removal of proteoglycans by testicular hyaluronidase digestion showed that type IX collagen is distributed throughout the cartilage matrix, and is not present in other connective tissues or skeletal muscle. The intact type IX collagen molecule, which was secreted by a suspension culture of freshly isolated embryonic chick chondrocytes, was recognized by rotary shadowing in the presence of antibody 2C2 after first precipitating the procollagens from the culture medium with ammonium sulfate (30%). Two different collagenous molecules were present in the precipitate: a longer molecule of type II procollagen (average length, 335 nm) with both amino- and carboxy-propeptides still remaining uncleaved, and a shorter molecule (average length, 190 nm) which was identified as type IX collagen. Antibody 2C2 consistently bound to the shorter molecules at a site located 136 nm from a distinctive knob at one end of the molecule, and did not bind to any specific site on the type II procollagen molecules. The structure of the intact type IX collagen molecule with the location of both collagenous and noncollagenous domains was as predicted after converting the nucleotide sequence of a cDNA clone encoding for one of the chains of type IX collagen to an amino acid sequence (Ninomiya, Y., and B. R. Olsen, 1984, Proc. Natl. Acad. Sci. USA, 81:3014-3018).     

Human surfactant polypeptide SP-B. Disulfide bridges, C-terminal end, and peptide analysis of the airway form.

Human hydrophobic surfactant polypeptide, SP-B, purified from lung tissue by exclusion chromatography in organic solvents, has been characterized. The polypeptide is 79 residues long, has a C-terminal methionine, and contains seven Cys residues. Native human SP-B lacks free thiol groups. Three intrachain disulfide bridges were defined, linking Cys8 to Cys77, Cys11 to Cys71 and Cys35 to Cys46. The remaining Cys48 is concluded to link the protein chains into homodimers via an interchain disulfide to its counterpart in a second SP-B polypeptide. These SS bridges are identical to those in the porcine form and confirm a consestant and unique disulfide pattern for SP-B polypeptides in general.     

Chemical properties and amino acid composition of beta1-bungarotoxin from the venom of Bungarus multicinctus (Formosan banded krait)

beta-Bungarotoxin purified from the venom of Bungarus multicinctus (Formosan banded krait) contained no carbohydrate and behaved as a homogeneous protein on polyacrylamide gel electrophoresis at pH 4.1 and SDS-polyacrylamide gel electrophoresis without 2-mercaptoethanol treatment. Its molecular weight and isoelectric point were estimated to be about 21,000 by gel filtration and about 9.5 by isoelectric focusing, respectively. The toxin treated with the reducing agent was split into two polypeptide chains as revealed by SDS-polyacrylamide gel electrophoresis and their molecular weights were calculated to be about 13,000 and 7,000. The two polypeptide chains (the large one named the A chain and the small one the B chain) were isolated by gel filtration after reduction of disulfide bonds in the toxin followed by alkylation. The A chain contained 120 amino acid residues including 13 half-cystines and the B chain 60 residues including 7 half-cystines. The two chains were supposed to link by disulfide bond(s) in the intact toxin which contained no free sulfhydryl groups. The N-terminal residues of the A and B chains were asparagine and arginine and the C-terminal ones were glutamine and proline, respectively, in accordance with the results of the terminal analyses of the intact toxin.     

Cleavage of human fibroblast type I procollagen by mammalian collagenase: demonstration of amino- and carboxy-terminal extension peptides.

Human skin fibroblasts in monolayer culture synthesize and secrete precursor forms of collagen into the culture medium. The type I collagen precursor, the major precursor in the culture medium, was isolated on DEAE cellulose chromatography and subjected to mammalian collagenase cleavage. The amino terminal cleavage fragments had a higher molecular weight than α1^A and α2^A, but did not contain interchain disulfide bonds. The carboxy-terminal cleavage fragments formed high molecular weight aggregates which contained interchain disulfide bonds. These results indicate that human type I procollagen contains noncollagenous amino and carboxy-terminal extension peptides and that all of the interchain disulfide bonds are on the carboxy-terminal portion of the molecule.     

The disulphide bridges of a mouse immunoglobulin G1 protein

[(35)S]Cystine-labelled immunoglobulin MOPC21 (IgG1) was prepared from myeloma cells in tissue culture. Carrier myeloma protein was added and the protein was digested with pepsin. The digest was fractionated on Sephadex G-50 into two fractions, further digested with trypsin and again fractionated on Sephadex. Disulphide-bridge peptides were purified by electrophoresis and chromatography and identified by radioautography. A peptide of 96 residues was isolated, which contains both the heavy-light interchain disulphide bridge and all the inter-heavy-chain disulphide bridges. Other peptides were isolated, accounting for all the intrachain disulphide bridges (which could be placed by homology with proteins of other species), except for the variable section of the light chain. Sequences describing this missing disulphide bridge were obtained from totally reduced and alkylated light chains. Peptides related to the interchain disulphide-bridge peptide were isolated from partially reduced and alkylated myeloma protein and from totally reduced heavy chain. The interchain disulphide-bridge peptide was placed at the C-terminal position of the F(ab')(2) fragment, prepared by digestion of the protein with pepsin at pH4.0. Sequences from the heavy-chain intrachain disulphide bridges of MOPC 21 immunoglobulin are compared with homologous sequences from mouse myeloma proteins of other subclasses and proteins of other species.     

Spontaneous dissociation of an IgG-3 paraprotein through disulfide interchange.

A paraprotein of the gamma3 subclass was observed to dissociate "spontaneously" under various experimental conditions, such as during chromatography on DEAE-cellulose and on Sephadex G-200 or on starch block electrophoresis. This phenomenon was accompanied by the formation of various complexes of higher molecular weight, displaying antigenic properties different from those of the original paraprotein. These changes did not occur in the presence of iodoacetamide, indicating that dissociation of the paraprotein was due to disulfide interchange(s) and not to the absence of interchain disulfide bridges.     

Type IV collagen from chicken muscular tissues. Isolation and characterization of the pepsin-resistant fragments

Type IV collagen has been isolated from adult chicken gizzard after limited pepsin digestion and subsequent differential salt fractionation in acidic and neutral conditions. After denaturation, three fragments (called F1, F2, and F3) were isolated by agarose gel filtration and carboxymethylcellulose chromatography. F1 and F2 possessed apparent molecular weights of 53 000 and 50 000, respectively, and were consistently isolated in a 2:1 proportion. F3 was larger and after reduction of disulfide bonds gave rise to three fragments (called F3A, F3B, and F3C) of apparent molecular weights 68 000, 40 000, and 29 000. No alpha-chain-sized components of Type IV collagen were observed. A native fraction containing F1 and F2, but no F3, was isolated after extraction using less pepsin and an additional salt fractionation in acidic conditions. F1 and F2 in the native form were not separated by carboxymethylcellulose or diethylaminoethylcellulose chromatography performed in nondenaturating conditions or by differential salt precipitation in acidic or neutral conditions; these results suggest that F1 and F2 arise as a single native component of structure (F1)2F2. The fraction containing F1 and F2 also gave rise to a single segment long spacing crystallite pattern and to a circular dichroism spectrum which was typical for a native collagen. F1 and F2 were also isolated from chicken heart, blood vessels, and skeletal muscle, whereas from bovine aorta, using the same isolation procedures, two alpha-chain-sized components were obtained, which appeared to be similar to the two Type IV chains recently described by other groups. The data suggest that (i) pepsin fragmentation of type IV collagen from chicken tissues occurs in a different manner compared to Type IV collagen from mammalian tissues and (ii) for the chicken there must be at least two Type IV chains which are assembled into a single native molecule.     

Basement membrane (type IV) collagen is a heteropolymer

Type IV collagen was isolated in high yield from bovine kidney cortex. The protein revealed Mr = 380,000 and contained, in a 2:1 ratio, two different disulfide-linked polypeptide chains, C-1 and D-1 (Mr = 125,000). Carboxymethyl-cellulose chromatography before and after reduction proved that the two polypeptide chains are arranged in a single triple helical molecule with the chain composition (C-1)2(D-1). The disulfide bridges appear to be located 180 amino acid residues from the NH2 terminus of the chains.     

Characterization of a type IV procollagen synthesized by human amniotic fluid cells in culture.

Fetal epithelioid cells, isolated from human amniotic fluid, synthesize and secrete a type IV-like procollagen characterized by a unique pattern of cyanogen bromide (CNBr)-produced peptides. The procollagen is disulfide-bonded and, after reduction, migrates on sodium dodecyl sulfate-polyacrylamide gel electrophoresis as a doublet between collagen beta components and pro-alpha 1(I) chains. No conversion of the procollagen to collagen or to procollagen intermediates is observed in cell culture. The procollagen was purified by salt fractionation and ion exchange chromatography; its amino acid composition resembles that of collagenous proteins extracted from basement membranes, with a high 3- and 4-hydroxyproline and hydroxylysine content and low levels of alanine and arginine. The major products obtained after limited proteolytic digestion of the protein retain interchain disulfide bonds and, after reduction, migrate on sodium dodecyl sulfate-polyacrylamide gel electrophoresis near intact pro-alpha 1(I) chains. The procollagen is secreted efficiently by amniotic fluid cells despite almost complete inhibition of peptidyl hydroxylation but, unlike type I procollagen, the secreted underhydroxylated chains lack interchain disulfide bonds. Since these cells also secrete fibronectin and elaborate an extensive extracellular matrix, the system should prove useful in the study of cell-matrix interactions.     

Echistatin disulfide bridges: selective reduction and linkage assignment.

Echistatin is the smallest member of the disintegrin family of snake venom proteins, containing four disulfides in a peptide chain of 49 residues. Partial assignment of disulfides has been made previously by NMR and chemical approaches. A full assignment was made by a newly developed chemical approach, using partial reduction with tris-(2-carboxyethyl)-phosphine at acid pH. Reduction proceeded in a stepwise manner at pH 3, and the intermediates were isolated by high performance liquid chromatography. Alkylation of free thiols, followed by sequencer analysis, enabled all four bridges to be identified: (1) at 20 degrees C a single bridge linking Cys 2-Cys 11 was broken, giving a relatively stable intermediate; (2) with further treatment at 41 degrees C the bridges Cys 7-Cys 32 and Cys 8-Cys 37 became accessible to the reagent and were reduced at approx. equal rates; (3) the two bicyclic peptides produced in this manner were less stable and could be reduced at 20 degrees C to a peptide that retains a single bridge linking Cys 20-Cys 39; and (4) the monocyclic peptide can be reduced to the linear molecule at 20 degrees C. Some disulfide exchange occurred during alkylation of the bicyclic intermediates, but results unambiguously show the pattern to be [2-11; 7-32; 8-37; 20-39]. A comparison is made with kistrin, a longer disintegrin whose disulfide structure has been proposed from NMR analysis.     

Partial amino acid sequence of porcine elastase II. Active site and the activation peptide regions

The partial amino acid sequence of porcine elastase II, isolated from crude trypsin Type II, was determined. The enzyme consists of two polypeptide chains, a light chain composed of 11 residues, and a heavy chain (Mr = 23 500) with four intrachain disulfide bridges; the two chains are held together by one interchain disulfide bond. Elastase II was fragmented into several peptides by chemical cleavages with CNBr at the two methionine residues, 99 and 180 (chymotrypsinogen numbering), and with hydroxylamine at the peptide bond following DIP-Ser195. About 50% of the sequence was determined and the positions of 120 amino acids were located, including the light chain residues and most of the active site residues. The partial sequence shows 64% difference between porcine elastase II and elastase I and only 26% difference between porcine elastase II and bovine chymotrypsin B.     

p-HMW-collagen, a minor collagen obtained from chick embryo cartilage without proteolytic treatment of the tissue

The fragments of minor collagens of cartilages, called HMW and LMW, were isolated after pepsin treatment of sternal cartilages of young chickens and were shown to be entirely triple-helical molecules as judged by their circular dichroic spectra. Studies on renaturation kinetics of HMW suggested that the interchain disulfide bonds in HMW reside at one of the ends of the so-called long arm. Polyclonal antibodies against HMW were raised and affinity purified. These antibodies did not cross-react with type II collagen nor with other minor collagens such as LMW and 1 alpha, 2 alpha, 3 alpha collagen in native or denatured structure. The antibodies were used to identify HMW-related molecules which were synthesized by embryonic chick cartilages in vitro. Some of these molecules were secreted into the organ culture medium and could be recovered from it by ammonium sulfate precipitation. Polyacrylamide gel electrophoresis of this precipitate gave one band of high molecular weight which could be reduced to two bands migrating slightly faster than the alpha 1(II) chain when identified by immunoblotting. These bands could also be identified among about six radiolabelled polypeptides present in the ammonium sulfate precipitate of medium proteins when analysed by polyacrylamide gel electrophoresis followed by fluorography. The same polypeptides could be recovered from the medium by immunoprecipitation with anti-HMW antibodies. Their presence in cartilage tissue was shown by immunoblotting of material extracted from cartilage tissue and separated on polyacrylamide gels. We suggest that the protein containing these polypeptide chains represents the parent molecule of the peptic fragment HMW as it is synthesized in vivo and have designated it p-HMW-collagen.