Peter P. Fietzek

Information contained in the amino acid sequence of theα1(I)-chain of collagen and its consequences upon the formation of the triple helix, of fibrils and crosslinks

SummaryThe molecule of type I collagen from skin consists of twoα1(I)-chains and oneα2-chain. The sequence of the entireα1-chain comprising 1052 residues is summarily presented and discussed. Apart from the 279 residues ofα1(I)-CB8 whose sequence has been established for rat skin collagen, all sequences have been determined for calf skin collagen. In order to facilitate sequence analysis, theα1-chain was cleaved into defined fragments by cyanogen bromide or hydroxylamine or limited collagenase digestion. Most of the sequence was established by automated stepwise Edman degradation.Theα1-chain contains two basically different types of sequences: the triple helical region of 1011 amino acid residues in which every third position is occupied by glycine and the N- and C-terminal regions not displaying this type of regularity. Both of these non-triple helical regions carry oxidizable lysine or hydroxylysine residues as functional sites for the intermolecular crosslink formation. Implications of the amino acid sequence for the stability of the triple helix and the fibril as well as for formation of cross-links are discussed.Evaluation of the sequence in connection with electron microscopical investigations yielded the parameters of the axial arrangement of the molecules within the fibrils. Axial stagger of the molecules by a distance D = 670 Å = 233 amino acid residues results in maximal interaction of polar sequence regions of adjacent molecules and similarly of regions of hydrophobic residues. Ordered aggregation of molecules into fibrils is, therefore, regulated by electrostatic and electrophobic forces. Possible loci of intermolecular crosslinks between theα1-chains of adjacent molecules may be deduced from the dimensions of the axial aggregation of molecules.

The amino acid sequence of the carboxyterminal nonhelical cross link region of the α 1 chain of calf skin collagen

The collagen molecule terminates at both ends in regions endowed with particular properties. These regions cannot participate in the triple helical conformation since glycine does not occupy every third position as in the central areas of the molecule. The terminal regions contain lysine residues usually converted into cw-amino-adipic-semialdehyde by lysine oxidase [ 11 . Allysine is the essential participant in the crosslinking reaction [2,3]. Structure and function of the N-terminal region have been determined some time ago [4-91. In addition, Stark et al. [lo] reported the isolation and characterization of the nonhelical peptide of the C-terminal end of the al chain of calf skin collagen. The amino acid sequence of this peptide is reported in the present paper. It was isolated from a chymotryptic digest of the C-terminal cyanogen bromide peptide al-CB6. The sequence was established applying automated Edman degradation to the whole peptide and manual Edman-dansyl-technique to fragments obtained by proteolytic cleavage.

The covalent structure of collagen: Amino acid sequence of α1‐CB3 from calf skin collagen

The peptide ol-CB3 is derived from the ol-chain of collagen by cyanogen bromide (CNBr) cleavage and consists of 149 amino acids [ 11. It occupies positions 421 through 570 of the 1055 residues of the alchain [2]. In a previous communication we have described the isolation and characterization of defined fragments of al-CB3 from calf skin collagen by cleavage with trypsin, chymotrypsin, and thermolysin [3]. The present publication describes the elucidation of the complete amino acid sequence of o1 l-CB3, obtained exclusively by automated stepwise degradation according to Edman and Begg [4].

Isolation and characterization of the cyanogen bromide peptides from the α2 chain of calf skin collagen

Cleavage at the methionyl residues with CNBr has been proved to be a useful method for studies on the primary structure of the collagen molecule. CNBr (CB)-peptides of aland cuZchains of different species and tissues have been described [l-9] . Reuterberg and Kuhn [3, lo] reported the isolation, characterization and ordering of the CB-peptides of the al -chains of calf skin collagen. This paper describes the results of a similar investigation on the aZchains of calf skin collagen. The methods applied were very similar to those used by Fietzek and Piez [6] to isolate the CB-peptides of the oZchain of rat skin collagen. The peptides have been named on the basis of homology to six CB-peptides of the cuZchain of rat skin collagen.

Pro—Gln: The procollagen peptidase cleavage site in the α1 (I) chain of dermatosparactic calf skin procollagen

Collagen i~ synthesized as a precursor form called procollagen which contains both aminoand carboxyterminal extensions, the so-called procollagen peptides. During conversion of procollagen to collagen the extensions are removed by procollagen peptidases which are specific either for the Nor C-terminal extensions [ 1 ]. Investigation of the structure of the N-terminal extension has been greatly facilitated by the genetic disease, dermatosparaxis, found in cattle [2] and sheep [3], in which the enzyme responsible for removing the N-terminal procollagen peptide is inactive [4,5]. This results in the accumulation of procollagen in the skin which contains pal (I) chains. Here we report the isolation of the N-terminal extension of the pal(I) chain from a cyanogen bromide digest of dermatosparactic calf skin. The extension was fragmented using bacterial collagenase and peptides separated. The C-terminal collagenase fragment was sequenced and the cleavage site of the procollagen peptidase determined as the peptide bond between proline and glutamine.

Evidence for two antigenic determinants in the C-terminal region of rat skin collagen

The ability of the antigenic determinants of collagen to react with antibodies does not depend on the intact triplehelical structure. This property qualifies collagen for a comprehensive immunochemical investigation. Recent success in the chemical characterization of peptides obtained by cleavage with cyanogen bromide [l ,2] or with collagenase [3] has favoured such studies. Thus, Pontz et al. [4] were able to localize two different antigenic determinants on the C-terminal and on a middle region, respectively, of the calf skin collagen molecule. Michaeli et al. [S] assigned an antigenic determinant to the N-terminal region of guinea-pig collagen. In these studies, rabbit antibodies have been employed and it may be predicted that both calf skin and guinea-pig collagen should differ in their sequence from rabbit collagen at least at the antibody-recognition sites. A consistent application of immunological methods should highly facilitate the localization and elucidation of structural variations among different collagens and might also lead to a better understanding of the phenomenon of immunological discrimination. For these reasons some of the antigenic determinants of rat skin collagen were characterized by the same methods as already applied to calf collagen [4].

The covalent structure of collagen. Amino acid sequence of the N-terminal region of α2-CB4 from calf and rat skin collagen

Elucidation of the primary structure of collagen has rapidly progressed in recent years (reviewed by Traub and Piez [l] . Thus, approx. 65% of the ammo acid sequence of the cul-chain have been established [l-9]. The distribution of methionyl residues within the a2-chain is much less favourable, cyanogen bromide cleavage giving rise to three very large and three very small peptides. The large peptides are of equal length of approx. 330 amino acid residues, and the three short peptides comprise three, fourteen and thirty residues, respectively [lo]. So far, only the sequences of the small peptides have been determined [ 1, 11,121. The automated degradation procedure introduced by Edman and Begg [ 131 now rendered the large peptides likewise amenable to sequence work. In the present communication we report the sequence of the first 42 N-terminal ammo acid residues of (u2-CB4 from calf and rat skin collagen [ 10, 141. The entire peptide (112~CB4 comprises 330 residues and extends over the N-terminal third of the cr2-chain [ 14, 151. Comparison of these sequences of rat and calf collagen revealed four interspecies substitutions, all of them involving apolar residues. We also compared the first 60 N-terminal residues of the aland a2-chains of rat skin collagen.

The covalent structure of collagen: Amino acid sequence of the N-terminal region of α2-CB5 from rat skin collagen

The triple helical molecule of skin collagen contains two different peptide chains: two (~1(1)and one (u2chain [cf 11. However, there are collagen molecules present in other tissues such as cartilage which are composed of three identical chains, designated ozl(II)chains [2]. Collagen from evolutionary primitive species such as sea anemone likewise consists of three identical polypeptide chains [3]. Artificial molecules of composition [al( which may be obtained upon renaturation of type I al-chains display properties similar to those of natural skin collagen [4]. Thus, their renaturation temperature is similar to that of [al(I)]2 and they may aggregate to form fibrils of the native type. In contrast, much less stability is displayed by molecules consisting entirely of cu2-chains [4]. Therefore, the question may be raised, what biological advantage is associated with collagen molecules consisting of two different chains such as the collagens present in skin, bones and tendons. An answer may be expected from comparison of the sequence data of cw 1. and cu2-chains. The sequence of the al(I)-chain has been completely elucidated by investigations of calf and rat skin collagen [5-161. In contrast, investigations of the a2-chain are still in the early stages. So far, only the sequences of the three small CNBr-derived peptides (w2-CBl , (~2~CBO, and o12-CB2 have been reported. (r2-CBl is the N-terminal peptide of the aZchain and comprises of 14 amino acid residues. (r2-CBO is the adjacent tripeptide, and o12-CB2 is derived from a central region of the chain and consists of 30 residues [17-19,221. The sequence of the 42 N-terminal

Amino acid sequence of the amino-terminal region of calf skin collagen.

Since the N-terminus of the α2 chain is blocked by pyroglutamic acid, the chain cannot be subjected to stepwise Edman degradation. Limited pepsin digestion of the native molecule liberated at the N-terminus the peptide pyroglutamyl-phenylalanine. Automated sequencing of the remaining α2P for 17 steps resulted in elucidation of the amino acid sequences of α2-CB1, α2-CB0 and of the first residues of α2-CB4. The data were confirmed by investigation of the isolated tryptic and chymotryptic peptides of α2-CB1.

Contribution of the α2 chain to the molecular stability of collagen

Collagen, the main protein of connective tissue, has long stiff rod-like molecules consisting of three polypeptide chains wound about a common axis in a triple helical structure [ 1 ] . Each chain has a little over one thousand amino acid residues, including glycine in every third position, except near the ends of the chains, and unusually large amounts of the imino acids proline and hydroxyproline which help to stabilise the molecular conformation. In many types of collagen all three chains have identical sequences, but the predominant type of collagen in most vertebrate tissues has two identical chains, designated al, and a third, a2, of somewhat different sequence. All three chains extend over the full length of the molecule and have their N-terminal ends on the same side. The left-handed helical symmetry of the collagen molecule relates the structurally equivalent Gly. X.Y. tripeptides by a translation of 2.9 a and a rotation of some 108” (fig.1). Because every third residue lies near the central axis of the molecule there is room in this position for only the smallest amino acid, glycine. Residues in both the X and Y positions lie on the surface of the molecule, but differ in their backbone conformations and intramolecular stereochemical environments. The three chains are joined by systematic hydrogen bonding between NH groups of glycine residues and backbone CO groups of residues in X positions. With the determination of amino acid sequences of ol chains from several vertebrate collagens [2] it became clear that some amino acid residues were non-randomly distributed between X and Y positions