Isolation and partial characterization of the entire human pro alpha 1(II) collagen gene

Exclusion of the COL2A1 gene as the mutation site in diastrophic dysplasia

The involvement of the cartilage specific type II collagen gene (COL2A1) was studied in nine patients with diastrophic dysplasia in the Finnish population, where the prevalence of this chondrodystrophy clearly exceeds that reported for other populations. COL2A1 was chosen as the candidate gene based on previous morphological and chemical studies which suggested abnormal structure of type II collagen in diastrophic dysplasia. Southern analysis of the patients' DNA showed no disease related differences in any of the restriction fragments covering the 30 kb COL2A1 gene. As a second approach, the nine patients and their 74 relatives were studied for the inheritance of the type II collagen gene. Three of the patients with diastrophic dysplasia were not homozygous for the intragenic RFLP markers, which suggests that the disease is not linked to the type II collagen gene. Multipoint linkage analysis gave a lod score of -2.95, which conclusively excluded the COL2A1 gene as the mutation site in diastrophic dysplasia in these families.

Predisposition to familial osteoarthrosis linked to type II collagen gene

The genetic background of two families, in whom a predisposition to primary osteoarthrosis is inherited as a dominant trait, was investigated. Use of restriction fragment length polymorphisms within and around the type II collagen gene on chromosome 12 revealed a linkage between this cartilage-specific gene and primary osteoarthrosis.

Organization of the exons coding for pro alpha 1(II) collagen N-propeptide confirms a distinct evolutionary history of this domain of the fibrillar collagen genes

The organization of the exons coding for the N-terminal portion of human type II procollagen has been determined. Aside from inferring the previously unknown primary structure of type II N-propeptide, this study has revealed that this coding domain of the gene exhibits an organization uniquely distinct from those of type I and type III collagens. This finding substantiates the notion that the N-propeptide coding domains of the fibrillar collagen genes evolved under less stringent selection than those encoding the C-propeptide and triple helical regions.

The 5' splice site: phylogenetic evolution and variable geometry of association with U1RNA

The 5' splice site sequences of 3294 introns from various organisms (1-672) were analyzed in order to determine the rules governing evolution of this sequence, which may shed light on the mechanism of cleavage at the exon-intron junction. The data indicate that, currently, in all organisms, a common sequence 1GUAAG6U and its derivatives are used as well as an additional sequence and its derivatives, which differ in metazoa (G/1GUgAG6U), lower eucaryotes (1GUAxG6U) and higher plants (AG/1GU3A). They all partly resemble the prototype sequence AG/1GUAAG6U whose 8 contigous nucleotides are complementary to the nucleotides 4-11 of U1RNA, which are perfectly conserved in the course of phylogenetic evolution. Detailed examination of the data shows that U1RNA can recognize different parts of 5' splice sites. As a rule, either prototype nucleotides at position -2 and -1 or at positions 4, 5 or 6 or at positions 3-4 are dispensable provided that the stability of the U1RNA-5' splice site hybrid is conserved. On the basis of frequency of sequences, the optimal size of the hybridizable region is 5-7 nucleotides. Thus, the cleavage at the exon-intron junction seems to imply, first, that the 5' splice site is recognized by U1RNA according to a "variable geometry" program; second, that the precise cleavage site is determined by the conserved sequence of U1RNA since it occurs exactly opposite to the junction between nucleotides C9 and C10 of U1RNA. The variable geometry of the U1RNA-5' splice site association provides flexibility to the system and allows diversification in the course of phylogenetic evolution.

Further evidence for the correlation between the primary structure and the stain exclusion banding pattern of the segment-long-spacing crystallites of collagen

Recently, we have noted the direct correlation between the primary structure of type I collagen and the electron microscopical banding pattern of the negatively stained segment-long-spacing (SLS) crystallites (K. Kobayashi, T. Ito, and T. Hoshino (1986), J. Electr. Microsc. 35, 272-275). In this paper, we examined the correlation in the other types of collagen. Unstained light bands (stain excluding bands) of the negatively stained SLSs of type II and type III collagens were located at the clusters of large hydrophobic amino acid residues along the respective molecules. Photographic averaging of the pattern improved the visual comparison of the correlation. We also noted a few occasions of discrepancy from the above-mentioned correlation. Preliminary computer simulation experiments revealed that, among amino acid parameters so far reported, only the hydrophobicity values of G. D. Rose and S. Roy (1980, Proc. Natl. Acad. Sci. USA 77, 4643-4647) explained the ability of amino acids for the negative staining (stain exclusion) of the collagen SLSs.

Does Sc1-70 modulate collagen production in systemic sclerosis?

Sc1-70, an autoantigen in systemic sclerosis, may accelerate collagen gene transcription by virtue of its activity as a topoisomerase I (topo I), a DNA template-modifying enzyme. A survey of sequences corresponding to all or part of the known topo I binding sequence AGAACTTAGAGAAAATTTAAA in four fibrillar collagen genes (three of them dermal) and sixteen non-collagen genes showed a striking preponderance of the tetramer 5'-CTTA-3', comprising the core of this binding sequence, at the exon-intron junctions of the fibrillar collagen genes (59% compared with 16% in the control group). In addition, a non-random clustering of three potential topo I binding sites was seen within 350 base-pairs of 5' flanking DNA in the dermal collagen gene alpha 2(I), and a fourth site occurred in the promoter region of the alpha 1(III) gene. The findings suggest that a selective vulnerability to the action of Sc1-70/topo I is built into the structure of the dermal collagen genes.

The Stickler syndrome: evidence for close linkage to the structural gene for type II collagen

The Stickler syndrome is an autosomal dominant hereditary disorder of connective tissue with pleiotropic features including premature osteoarthropathy, mild spondyloepiphyseal dysplasia, vitreoretinal degeneration, and the Pierre-Robin sequence. Genetic linkage studies in two families with the Stickler syndrome have been performed using restriction fragment length polymorphisms associated with the structural gene for type II collagen, COL2A1. No recombinants between the Stickler phenotype and COL2A1 were observed. The total LOD score for linkage of the Stickler syndrome and COL2A1 at a recombination fraction (theta) of zero is 3.59. These findings suggest that, at least in some families, the mutation causing Stickler syndrome affects the structural locus for type II collagen.

Determination of the single polyadenylation site of the human pro alpha 1(II) collagen gene

Several cDNA clones for the human type II procollagen mRNA were isolated from a cartilage cDNA library. Six of the clones containing the longest inserts were subjected to restriction site mapping for alignment. All these six clones extended to the poly A tail. The longest clone, containing a 1470 bp insert, was named pHCAR3. Sequencing of pHCAR3 made it clear that neither of the two canonical AATAAA sequences of the human type II collagen gene is used as the polyadenylation signal. Two 60 bp stretches of high interspecies homology terminating in a hexanucleotide ATTAAA, located 23 nucleotides upstream of the poly A tail, apparently have an important role in determining the single polyadenylation signal for this gene. S1 protection experiments confirmed these observations.

Characterization of three constituent chains of collagen type VI by peptide sequences and cDNA clones

Pepsin-solubilized collagen VI was prepared from human placenta and used to separate three constituent chains for determining partial amino acid sequences. Antibodies raised against the chains assisted in the identification and purification of several cDNA clones from three expression lambda gt11 libraries. Most of the clones hybridized to either a 3.5-kb or 4.2-kb mRNA species which by matching peptide and nucleotide sequences could be identified as coding for the alpha 2(VI) or alpha 1(VI) chain, respectively. Other clones hybridized to either an 8.5-kb mRNA which very likely encoded the alpha 3(VI) chain or to an unknown 2.0-kb mRNA. Northern blots revealed a considerable variation in the mRNA levels for each collagen VI chain in both skin and cornea fibroblasts and in several tumor cell lines. Limited sequence data generated from peptides and cDNA clones demonstrated a characteristic cysteine pattern at the junction between N-terminal globular domain and triple helix in all three chains. In addition, the data showed occasional interruptions of triplet sequences within the triple-helical domain and the presence of two Arg-Gly-Asp sequences which are potential cell-binding structures.

Disulfide bonding as a determinant of the molecular composition of types I, II and III procollagen

Procollagen molecules have amino-terminal and carboxy-terminal propeptides at the respective ends of the collagenous triple helix. The carboxy-terminal propeptides enhance and direct the association of pro alpha-chains into procollagen molecules, but the mechanism of this registration function is still obscure. A hypothesis concerning the function of disulfide bonding in the assembly of types I, II and III procollagen is put forward here.

Partial characterization of a low molecular weight human collagen that undergoes alternative splicing

A cDNA library prepared from RNA isolated from a cultured human tumor cell line, HT-1080, was screened with a mouse cDNA clone coding for part of the -Gly-Xaa-Yaa- domain of the alpha 2(IV) collagen chain. Four overlapping cDNA clones were characterized that coded for a low molecular weight human collagen. The cDNA clones did not, however, code for the short-chain collagens, types IX and X. The amino acid sequences derived from the clones resembled type IV collagen in that there were short interruptions in the repeating -Gly-Xaa-Yaa- sequence. The noncollagenous, carboxyl-terminal domain was, however, much shorter and contained only 18 amino acid residues. Interestingly, one of the cDNA clones contained an additional 36 nucleotides not found in an overlapping clone. The 36 nucleotides encoded four -Gly-Xaa-Yaa- repeats without changing the reading frame. Nuclease S1 mapping demonstrated that the difference between the clones was due to existence of two different mRNAs. A synthetic 24-residue peptide corresponding to the last two -Gly-Xaa-Yaa- triplets and the entire carboxyl-terminal domain was used to generate polyclonal antibodies. Electrophoretic transfer blot analysis of HT-1080 cells and normal human skin fibroblasts identified two polypeptides, Mr 67,000 and Mr 62,000, that were sensitive to bacterial collagenase.

The pro alpha 2(V) collagen gene is evolutionarily related to the major fibrillar-forming collagens

A number of overlapping cDNA clones, covering 5.2 kb of sequences which code for the human pro alpha 2(V) collagen chain, have been isolated. Analysis of the structural data have indicated a close evolutionary kinship between the pro alpha 2(V) chain and the major fibrillar collagen types. Isolation and analysis of an 8 kb genomic fragment has further supported this notion by revealing a homologous arrangement of nine triple-helical domain exons. These studies have therefore provided conclusive evidence which categorizes the Type V collagen as a member of the Group 1 molecules, or fibrillar-forming collagens.

COOH-terminal propeptides of the major human procollagens. Structural, functional and genetic comparisons

The sequences of the carboxy-terminal extensions (COOH-propeptides) of at least one chain of all of the major human procollagens have only recently been deduced, and include those of the interstitial (alpha 1(I), alpha 2(I), alpha 1(II), alpha 1(III)), basement membrane (alpha 1(IV)) and pericellular (alpha 2(V)) procollagens. Comparisons of DNA and protein sequences, corresponding to these COOH-propeptides domains, established the early divergence of the basement membrane alpha 1(IV) COOH-propeptide from the corresponding sequences of the interstitial and pericellular procollagens. The latter are relatively highly conserved and share 58% primary peptide sequence similarities, whereas sequence similarities relative to alpha 1(IV) are limited. Hydropathy profiles and secondary structure potentials further emphasize the clustering of conserved and variable regions among the interstitial and pericellular COOH-propeptides, and provided further evidence for significant structural differences between these sequences and the alpha 1(IV) COOH-propeptide. The most highly conserved sequences of the alpha 1(I), alpha 2(I), alpha 1(II), alpha 1(III) and alpha 2(V) COOH-propeptides include regions surrounding the carbohydrate attachment site, cysteine-containing regions and the COOH-terminal sequences. Cysteinyl, tyrosyl and tryptophanyl residues were found to be highly conserved as were most charged residues. Localization of variable regions, in general, occurs within hydrophilic sequences with high beta-turn potentials that are proximal to intron/exon splice junctions. The most variable sequences are associated with the telopeptides and adjoining NH2-terminal portions of the COOH-propeptides as demonstrated by predictive secondary structure analyses. These results, combined with similar analyses of abnormal alpha 2(I) COOH-propeptide (osteogenesis imperfecta) permitted the identification of subsequences that are likely to be a prerequisite for COOH-propeptide functions, namely procollagen chain recognition and nucleation sites for triple helix formation. These functions are also common to the alpha 1(IV) COOH-propeptide; however, the lack of cleavage of this region and its additional postulated structural role in extracellular matrix interactions likely account for its divergent primary and secondary structure.

Genetic disorders of collagen

Osteogenesis imperfecta, Ehlers-Danlos syndrome, and Marfan syndrome form a group of genetic disorders of connective tissue. These disorders exhibit remarkable clinical heterogeneity which reflects their underlying biochemical and molecular differences. Defects in collagen types I and III have been found in all three syndromes.

Nonexpression of cartilage type II collagen in a case of Langer-Saldino achondrogenesis

A lethal short-limbed dwarfism was diagnosed at autopsy as the Langer-Saldino variant of achondrogenesis by radiological, histological, and gross pathological criteria. Cartilage was obtained for biochemical and ultrastructural analyses from the ends of long bones, from ribs and from a scapula of the newborn infant. At all sites, it had an abnormal gelatinous texture and translucent appearance. Biochemical analyses of the cartilages to identify pepsin-solubilized collagen alpha-chains and collagen-specific CNBr-peptides failed to detect type II collagen at any site where it would normally be the main constituent. Instead, type I was the predominant collagen present. However, three cartilage-specific minor collagen chains identified as 1 alpha, 2 alpha, and 3 alpha chains by their electrophoretic mobility were present at about 10% of the total collagen. Cartilage-specific proteoglycans also appeared to be abundant in the tissue judging by its high hexosamine content and high ratio of galactosamine to glucosamine. The findings indicate that a chondrocyte phenotype had differentiated but without the expression of type II collagen. In addition to the skeletal abnormalities, the severe pulmonary hypoplasia was also felt to be directly related to the underlying pathology in collagen expression. The term chondrogenesis imperfecta rather than achondrogenesis should be considered a more accurate description of this and related conditions.

Partial purification and characterization of a procollagen C-proteinase from the culture medium of mouse fibroblasts

A procollagen C-proteinase was purified about 100-fold from the medium of cultured mouse fibroblasts by a combination of ammonium sulfate precipitation, gel-filtration, and affinity chromatography on a column of Sepharose coupled to the carboxyl propeptide of type I procollagen. The purified enzyme did not exhibit other proteolytic activities, and it cleaved type I, II and III procollagens to produce the corresponding pN alpha chains and carboxyl propeptides as the only products. Amino acid sequencing of the first 14-18 residues at the N-terminus of the carboxyl propeptides generated by the enzyme from human pro alpha 1(I), pro alpha 2(I) and pro alpha 1(III) chains showed that the cleavage occurred at the physiological site, i.e. at the specific Ala-Asp bond in the pro alpha 1(I) and pro alpha 2(I) chains, and at the specific Gly-Asp bond in the pro alpha 1(III) chain. The pH optimum of the enzyme is 8.5 and its molecular weight as estimated by gel-filtration is about 125,000 daltons. The enzyme is inhibited by metal-chelators, various amines, dithiothreitol, N-ethylmaleimide and serum, but it is insensitive to pepstatin, leupeptin and serine proteases inhibitors. The enzyme differs from the C-proteinase described by Njieha et al. (Biochemistry 21:757-764, 1982), and the catheptic activities reported by Davidson et al. (Eur. J. Biochem 100:551-558, 1979) and Helseth and Veis (Proc. Natl. Acad. Sci. USA 81:3302-3306, 1984). The specificity of the enzyme is offered as evidence for a unique, C-proteinase, and its recovery from culture medium supports an extracellular location for procollagen processing.

Discrimination among multiple AATAAA sequences correlates with interspecies conservation of select 3' untranslated nucleotides

The DNA sequence corresponding to the 1.3 kb 3' untranslated region of the 6.5 kb human procollagen alpha 1(IV) mRNA was determined and compared with the mouse sequence obtained from 3' cDNA and genomic clones overlapping the reported 5' half (Oberbaumer et al., 1985, Eur. J. Biochem. 147:217). Although four AAUAAA hexanucleotides are found in the human and seven in the mouse RNAs, Northern blot hybridization showed almost exclusive utilization of the most 3' sequence, in contrast to the pattern seen when using alpha 1(I), alpha 2(I), alpha 1(III) and alpha 2(V) procollagen probes. Moreover, the ninety nucleotides 5' to the poly A tail in the major alpha 1(IV) mRNAs exhibit a much greater degree of interspecies homology than those encompassing the other three shared AAUAAA recognition signals. Further examination of this highly conserved area revealed the presence of two "consensus sequences" found in the 3' noncoding region of a number of RNA polymerase II transcribed genes (Mattaj and Zeller, 1983, Embo J. 2:1883) and, unexpectedly, some similarity with the nucleotides 5' to the poly A attachment signals in other procollagen mRNAs.

Exon/intron organization of the chicken type II procollagen gene: intron size distribution suggests a minimal intron size

Overlapping genomic clones have been isolated that contain the alpha chain and COOH-terminal propeptide coding regions of the chicken type II procollagen gene. All type II procollagen exon sequences present in these clones have been identified and mapped by DNA sequencing. These include 43 exons coding for the alpha-chain triple helix, 1 exon coding for the junction between the COOH-terminal propeptide and the alpha-chain region, and 3 exons coding for the COOH-terminal propeptide and 3' noncoding sequences. With the exception of one additional intron between 2 exons coding for amino acids 568-585 and 586-603, exon-intron boundaries have been conserved when compared with genes for all other characterized genes for fibrillar collagens. The chicken type II procollagen gene differs from most other collagen genes in having introns of considerably smaller average size. The size distribution of the introns suggests that approximately equal to 80 base pairs may be a minimal functional size for introns in this gene. This size of intron may be necessary in a gene with a very large number of small exons to prevent aberrant splicing from removing exon sequence together with intron sequence.

Further evidence for the dispersion of the human fibrillar collagen genes

Recombinant DNA probes specific for the human pro alpha 1(II) and pro alpha 1(III) collagen chains have been used for the chromosomal localization of the two genes. Restriction endonuclease analysis of DNA from human-rodent hybrid cell lines in conjunction with in situ hybridization of human metaphasic chromosomes have shown that the gene coding for the pro alpha 1 chain of type II collagen (COL2A1) is located on chromosome 12 in the segment 12q131----12q132. Likewise, the gene coding for the pro alpha 1 chain of type III collagen (COL3A1) was assigned to the segment 2q31----2q323 of chromosome 2.

Promoter region of the human pro-alpha 1(II)-collagen gene

We have isolated a genomic clone containing the 5'-terminal portion of the human pro-alpha 1(II)-collagen gene. This clone, HC2C, contains 10 kb of the gene and 6 kb of 5'-flanking sequences. It was selected by cross hybridization using a [32P]DNA probe containing the promoter region of the rat alpha 1(II)-collagen gene. We have determined the nucleotide sequence of the first exon and of 400 bp upstream. There is considerable homology between this human sequence and the corresponding rat sequence. As in the rat, the first exon contains a 155-bp untranslated segment and a 85-bp sequence coding for the signal peptide and a part of the N-propeptide of type-II procollagen. The segment preceding the transcription initiation site contains a conserved 'ATATAA' element, and several similar G + C-rich stretches. As in the rat gene, no 'CAT' element is evident between -70 and -120. We also find the sequence 5'-GTGGTCAGA-3' reported as an enhancer element in both viral and cellular genes, located around -290 bp. A high degree of homology exists between the atypical rat and human promoter structures; however, such homology is absent among the alpha 1(I), alpha 2(I) and alpha 1(III) promoters, which suggests that the unique alpha 1(II) sequences may be related to the specific expression of the alpha 1(II)-collagen gene.

Isolation and characterization of the human fibrillar collagen genes

In order to elucidate some of the mechanisms leading to the pathological expression of the human fibrillar collagens, as well as to understand the evolution of these loci, specific cDNA and genomic clones have been isolated. The primary structure of the COOH-terminal propeptide of the four collagen chains and either part or the entire exon/intron arrangement of the genes have been determined. Interspecies and pairwise comparison revealed that the four loci have evolved at slightly different rates, maintaining, however, remarkably similar exon/intron arrangement. The fibrillar genes, albeit sharing the same elaborate structure, exhibit different sizes that correlate with the average length of their intron sequences, possibly because of their different chromosomal origin.

Collagen genes and proteins in osteogenesis imperfecta

Type I collagen is a heteropolymer of alpha 1(I) and alpha 2(I) chains, each of which is a separate product of genes localised to chromosomes 17 and 7 respectively. Molecular defects of type I collagen produce a group of inherited disorders of connective tissue primarily affecting bones, which are easily broken and collagen depleted (osteogenesis imperfecta). Sillence classifies these diseases into four groups, two of which are autosomal dominant and relatively mild, the others being either genetic lethals or responsible for very severe progressive disease. Here we describe two specific molecular abnormalities of type I collagen. One, a cysteine substitution in alpha 1(I) collagen, causes a mild Sillence type I disease, the other, a four base deletion in the C terminal extension of alpha 2(I) collagen, causes progressive Sillence type III disease in the homozygously affected patient and mild premature osteoporosis in his clinically symptomless parents. We have briefly reviewed a variety of other similar mutations causing various OI syndromes, which are tabulated, including various helical and non-helical deletions and a variety of structural protein changes. Several restriction fragment length polymorphisms for alpha 2(I) and alpha 1(II) collagens have also been described, and 5' EcoRI and 3' MspI polymorphisms for alpha 2(I) collagen segregate with Sillence type IV OI.

A highly polymorphic region 3' to the human type II collagen gene

We have characterised a highly polymorphic region 1.3kb downstream of the human Type II collagen gene. It consists of a highly AT-rich tandem repetitive region (minisatellite) approximately 650bp long. Two alleles had been observed previously, differing in size by approximately 300bp. When this region was cloned from four unrelated individuals carrying the larger allele, DNA sequence data identified three alleles, suggesting far higher polymorphism than was originally supposed. This minisatellite was shown to be present in a single copy in the human genome, and to have arisen after the divergence of Old and New World monkeys.