CHARACTERIZATION OF TYPE MICHAELA I AND TYPE III COLLAGENS IN BODE HUMAN TISSUES Department of Clinical Chemistry

OULU 2000

MICHAELA BODE

CHARACTERIZATION OF TYPE I AND TYPE III COLLAGENS IN HUMAN TISSUES

Academic Dissertation to be presented with the assent of the Faculty of Medicine, University of Oulu, for public discussion in Auditorium 9 of the University Hospital of Oulu, on April 7th, 2000, at 12 noon.

OULUN YLIOPISTO, OULU 2000 Copyright © 2000 Oulu University Library, 2000

Manuscript received 16 February 2000 Accepted 18 February 2000

Communicated by Docent Kari Majamaa Docent Onni Niemelä

ISBN 951-42-5553-4

ALSO AVAILABLE IN PRINTED FORMAT

ISBN 951-42-5552-6 ISSN 0355-3221 (URL: http://herkules.oulu.fi/issn03553221/)

OULU UNIVERSITY LIBRARY OULU 2000 Bode, Michaela, Characterization of type I and type III collagens in human tissues. Department of Clinical Chemistry, University of Oulu, FIN-90014 University of Oulu, Finland 2000 Oulu, Finland (Manuscript received 14 February 2000)

Abstract

Fibrillar type I and III collagens are the major constituents of the extracellular matrix, providing the tissue with tensile strength and influencing cell attachment and migration. The amount of type III collagen and the extent of its processing and cross-link maturation were studied in human atherosclerotic plaques, abdominal aortic aneurysms, colon and ovarian cancer, and finally, colon diverticulosis, using a novel radioimmunoassay for the cross-linked aminoterminal telopeptide of type III collagen. In addition, immunoassays for different structural domains of type I and type III collagens, together with immunohistochemical methods, were applied. In atherosclerotic plaques, the fully cross-linked type III collagen was the major collagen type. Type III collagen was completely processed, since the amount of type III pN-collagen was negligible. The amounts of free type I and III procollagen propeptides in the soluble fraction were small, indicating a low rate of collagen turnover. The proportion of type III collagen was lower in abdominal aortic aneurysms than in atherosclerotic aortic control samples. Furthermore, the amount of type III pN-collagen was significantly increased in aneurysms. Type I and III collagens were also maturely cross-linked in colon diverticulosis, the only difference from normal colon tissue being the increased amount of the aminoterminal propeptide of type III procollagen in the soluble tissue extract, indicating a slightly increased metabolic activity of type III collagen. In malignant ovarian tumors, the cross-linking of type I and III collagens was defective. A similar trend was also seen for type I collagen in colon cancer. Even though procollagen synthesis was increased in these malignancies, the total collagen content and the amounts of cross-linked collagens were decreased. The amount of type III pN-collagen was increased in malignant ovarian tumors, whereas no such tendency was seen in colon cancer. These findings suggest a wide variety of changes in the metabolism of type I and III collagens in diseases. Defective processing and cross-link maturation of these collagen types might result in impaired fibril formation or increased susceptibility of collagens to proteolytic attack – both of them processes with a potential role in the pathogenesis of diseases.

Key words: aortic aneurysm, atherosclerosis, diverticulosis, neoplasms Acknowledgements

This study was initiated at the Department of Medical Biochemistry in 1994, and continued at the Department of Clinical Chemistry. Professor Taina Pihlajaniemi, Head of the Department of Medical Biochemistry, University of Oulu, and Professor Arto Pakarinen, Chief Physician of the laboratory of the Oulu University Hospital, are acknowledged for the opportunity to work at these facilities. I am deeply indebted to my supervisor, Professor Juha Risteli, for his guidance during these long years and his confidence in my project. I also wish to express my gratitude to my co-supervisor, Docent Leila Risteli, Director of Learning and Research Services, University of Oulu, whose expert knowledge of collagen and help with the manuscripts have been of great value. Docent Onni Niemelä and Docent Kari Majamaa are acknowledged for their thorough reviewing and valuable comments on the manuscript of this thesis. I thank my collaborators, Professor Tatu Juvonen, Professor Frej Stenbäck, Docent Tuomo Karttunen, Docent Jyrki Mäkelä, Docent Ylermi Soini, Jukka Melkko, M.D., Ph.D., and Martti Mosorin, M.D., for their contribution to this research. I am very grateful to my co-author, Jari Satta, M.D., Ph.D., for his never-ending optimism, encouragement and valuable contribution to this research. Special thanks go to Saila Kauppila, M.D., Ph.D., from whom I have learned so much and who has always been a role model for me, like a big sister in the scientific world. Aimo Heinämäki, Ph.D., Päivi Annala, Kristiina Apajalahti, Tiina Holappa, Ulla Pohjoisaho and Sanna Sääskilahti are acknowledged for their expert technical assistance. I would also like to thank Anne Ollila for secretarial help. The photography laboratory and the library of the Medical Faculty are greatly appreciated. I thank Sirkka-Liisa Leinonen, Lic. Phil., for the careful revision of the language of this thesis. I wish to thank the whole staff of our laboratory. I am especially grateful to my colleague Heidi Eriksen for her friendship and support during these years. Without her, the long and occasionally difficult days at work would have been unbearable. Finally, I owe my deepest gratitude to my fiancé Joni Mäki, whose love, encouragement and confidence in me have been the thriving forces in my life since the day we first met at the freeze-dryer at the Department of Medical Biochemistry. I am also indebted to my mother Kaarina, and my grandparents Laila and Pentti, for their endless love, support and care during my whole life. I hope that when I have children of my own, I will be able to provide them as good a home as I have had myself. This work was financially supported by the Oulu University Hospital, Suomen Laboratoriolääketieteen Edistämissäätiö and my mother.

Oulu, February 2000 Michaela Bode Abbreviations

AAA abdominal aortic aneurysm ACP aldol condensation product AOD aortoiliac occlusive disease C- carboxy- CI confidence interval (95 %) CNBr cyanogen bromide Col 1 the most aminoterminal globular domain of aminoterminal propeptide of a procollagen Col 2 carboxyterminal end of aminoterminal propeptide Col 3 collagenous helix in the middle of aminoterminal propeptide CTGF connective tissue growth factor DEAE diethylaminoethyl deH-DHLNL dehydrodihydroxylysinonorleucine deH-HLNL dehydrohydroxylysinonorleucine deH-LHNL dehydrolysinohydroxynorleucine deH-LNL dehydrolysinonorleucine ECM extracellular matrix EDS Ehlers-Danlos Syndrome EDTA ethylenediamine tetra-acetate HHL histidinohydroxylysinonorleucine HHMD histidinohydroxymerodesmosine HIS histidine HL-Pyridinoline hydroxylysylpyridinoline HL-Pyrrole hydroxylysylpyrrole HPLC high performance liquid chromatography HSP47 heat shock protein 47 (a collagen specific chaperone) HYL hydroxylysine HYLALD hydroxylysylaldehyde ICTP cross-linked carboxyterminal telopeptide of type I collagen IFN-γ interferon gamma IIICTP carboxyterminal telopeptide of type III collagen IIINTP cross-linked aminoterminal telopeptide of type III collagen IL-1 interleukin-1 LDL low density lipoprotein LO lysyl oxidase LOL lysyl oxidase-like protein LOR lysyl oxidase-related protein L-Pyridinoline lysylpyridinoline LYS lysine LYSALD lysylaldehyde MMP matrix metalloproteinase MT-MMP membrane type matrix metalloproteinase N- amino- NTx aminoterminal telopeptide of type I collagen PAGE polyacrylamide gel electrophoresis PBS phosphate-buffered saline PDGF platelet derived growth factor PDI protein disulfide isomerase PICP carboxyterminal propeptide of type I procollagen PIIICP carboxyterminal propeptide of type III procollagen PIIINP aminoterminal propeptide of type III procollagen PINP aminoterminal propeptide of type I procollagen pN-collagen collagen molecule with retained aminoterminal propeptide RER rough endoplasmic reticulum SDS sodium dodecyl sulfate SMC smooth muscle cell SP4 linear synthetic peptide (SAGFDFSFLPQPPQEKY, amino acid residues nos. 1193-1208 + tyrosine residue) corresponding to the carboxyterminal telopeptide of the α1-chain of type I collagen SP6 linear synthetic peptide (DVKSGVAVGGLAGY, amino acid residues nos. 159-162) corresponding to the aminoterminal telopeptide of the α1-chain of type III collagen TGF-β transforming growth factor beta TIMP tissue inhibitor of metalloproteinase TPCK N-tosyl-L-phenylalanine chloromethyl ketone (enzyme inactivator) List of original articles

This thesis is based on the following articles, which are referred to in the text by their Roman numerals:

I Bode MK, Mosorin M, Satta J, Risteli L, Juvonen T & Risteli J (1999) Complete processing of type III collagen in atherosclerotic plaques. Arterioscler Thromb Vasc Biol 19: 1506-1511.

II Bode MK, Mosorin M, Satta J, Risteli L, Juvonen T & Risteli J (2000) Increased amount of type III pN-collagen in abdominal aortic aneurysms. Submitted as a revised version.

III Bode MK, Karttunen TJ, Mäkelä J, Risteli L & Risteli J (2000) Type I and III collagens in human colon cancer and diverticulosis. Scand J Gastroenterol. In press.

IV Kauppila S, Bode MK, Stenbäck F, Risteli L & Risteli J (1999) Cross-linked telopeptides of type I and III collagens in malignant ovarian tumours in vivo. Br J Cancer 81: 654-661.

In addition to papers I, II, III and IV, some new results on abdominal aortic aneurysms are included.

Contents

Abstract Acknowledgements Abbreviations List of original articles Contents 1. Introduction...... 13 2. Review of the literature...... 14 2.1. Structure and biosynthesis of collagens ...... 14 2.2. Fibril-forming collagens ...... 17 2.2.1. Type III collagen ...... 17 2.2.1.1. General features and assembly of type III collagen...... 17 2.2.1.2. Aminoterminal propeptide of type III procollagen, PIIINP ...... 18 2.2.1.3. Aminoterminal telopeptide of type III collagen, IIINTP...... 20 2.2.1.4. Type III pN-collagen...... 21 2.2.1.5. Carboxyterminal telopeptide (IIICTP) and propeptide (PIIICP) of type III collagen ...... 22 2.2.1.6. Distribution of type III collagen in tissues ...... 23 2.2.1.7. Type III collagen mutations ...... 23 2.2.2. Type I collagen and its metabolites...... 24 2.2.3. Enzymatic cross-linking of type I and III collagens...... 25 2.2.3.1. Lysyl oxidase...... 25 2.2.3.2. Cross-linking pathways ...... 26 2.2.4. Degradation of type I and III collagens...... 28 2.3. Type I and III collagens in atherosclerosis and abdominal aortic aneurysms ...... 30 2.3.1. General features of atherosclerosis...... 30 2.3.2. Type I and III collagens in normal and atherosclerotic aorta ...... 30 2.3.3. Synthesis and degradation of type I and III collagens in atherosclerotic plaques ...... 32 2.3.4. Other collagen-related events in atherosclerotic plaques ...... 33 2.3.5. Type I and III collagens in abdominal aortic aneurysms (AAA)...... 34 2.4. Type I and III collagens in cancer of the ovary and colon ...... 35 2.5. Type I and III collagens in colon diverticulosis...... 36 3. Outlines of the present study...... 38 4. Materials and methods ...... 39 4.1. Tissue samples ...... 39 4.2. Purification of the aminoterminal telopeptide of type III collagen ...... 40 4.3. Production of antiserum...... 40 4.4. Radioimmunoassay for IIINTP...... 41 4.5. Other immunoassays for type I and III collagen domains...... 41 4.6. Immunohistochemical staining ...... 41 4.7. Chromatographic analyses of type I and III collagen antigens in tissue samples...... 42 4.8. Hydroxyproline measurements ...... 42 4.9. Investigation of the proteolytic fragments of PIIINP...... 43 4.10. Statistics...... 43 4.11. Ethical considerations ...... 43 5. Results...... 44 5.1. Purification of the human cross-linked aminoterminal telopeptide of type III collagen (IIINTP)...... 44 5.2. New radioimmunoassays for type III collagen...... 44 5.3. Type I and III collagens in atherosclerotic plaques (I)...... 45 5.4. Type I and type III collagens in human abdominal aortic aneurysms (II)...... 46 5.5. Type I and III collagens in colon diverticulosis and colon cancer (III) ...... 48 5.6. Type I and III collagens in ovarian tumors (IV) ...... 49 6. Discussion...... 51 6.1. Insoluble vessel wall collagens – methodological and structural aspects ...... 51 6.2. Type I and III collagens – soil for atherosclerosis ...... 52 6.3. Abdominal aortic aneurysms and atherosclerosis ...... 54 6.4. Type I and III collagens in AAA and diverticular disease of colon ...... 56 6.5. Maturation and processing of type I and III collagens in ovarian and colon cancer...... 57 7. Conclusions...... 59 8. References...... 61 1. Introduction

Insoluble fibers, microfibrils, soluble proteins and glycoproteins are the components of the extracellular matrix (ECM). The ECM has an important role not only in providing tissues with their individual mechanical and physiological properties, but also in influencing cell attachment and cell migration. Collagens are the most abundant proteins of the ECM. Each of the collagen types shows a characteristic tissue distribution and thus has different functions and properties. Type I collagen is the major collagen type present in most tissues. Type III collagen is the second most abundant collagen in human tissues and occurs particularly in tissues exhibiting elastic properties, such as skin, blood vessels and various internal organs. These collagen types are synthesized as large precursor molecules containing propeptides at both ends. During extracellular processing, these propeptides are cleaved off and the collagen molecule participates in the intermolecular cross-linking responsible for tensile strength. Occasionally, the aminoterminal propeptides of type I and III collagens are retained, resulting in pN-collagen, which can only be found on the surface of collagen fibrils. The present study examines the characteristics of type I and especially type III collagens in the pathogenesis of several human diseases, in which these collagen types have been suggested to play a role. Type III collagen plays an important structural role in hollow organs, such as the aorta and bowel. Furthermore, type III collagen, together with type I collagen, takes part in the invasion of malignant tumors. For these reasons, atherosclerosis, abdominal aortic aneurysms, diverticulosis, and colon and ovarian cancer were studied in tissue samples by using specific immunoassays for several type I and III collagen metabolites and corresponding purified antibodies for immunohistochemical studies. In addition, new methods for quantifying type III collagen in tissue digests were developed. The immunomethods used here detect newly synthesized procollagens on the one hand and cross-linked, fibrillar type I and III collagens on the other, thus giving a good view into the metabolic state and matrix organization of tissues under investigation. 2. Review of the literature

2.1. Structure and biosynthesis of collagens

Collagens are one of the most important structural proteins in vertebrates. The characteristic features of collagens are the presence of at least one triple-helical domain consisting of polypeptide chains with a repeated Gly-X-Y sequence and the ability to form supramolecular aggregates with a structural function in the extracellular matrix (ECM) (see Kielty et al. 1993, Prockop & Kivirikko 1995, Bateman et al. 1996 for reviews). In addition to collagens, several other proteins, such as the complement component C1q, acetylcholine esterase and macrophage scavenger receptors, contain collagenous sequences, but since these proteins do not have any structural functions in the ECM, they are not classified as collagens (see Kielty et al. 1993). Collagens consist of three identical or different polypeptide chains, called α-chains, each of which is coiled into a left-handed helix, the three together being twisted around one another to form a right-handed super-helix. Glysine in every third position, frequently hydroxylated proline and lysine residues in the X and Y positions, and exact registration of the three α-chains are required for the zipper-like folding of the protein and its consequent triple-helical conformation (for review, see Engel & Prockop 1991). Collagens have traditionally been divided into two subgroups, fibril-forming and non- fibril-forming collagens, according to their structural features. The latter group may be further divided into subfamilies: network-forming collagens, a beaded filament-forming collagen, a collagen which forms anchoring fibrils, FACIT collagens (fibril-associated collagens with interrupted triple helices), collagens with a transmembrane domain, and MULTIPLEXINs (proteins with multiple triple helix domains and interruptions). To date, nineteen genetically distinct collagen types with specialized functions and characteristic tissue distributions have been identified (Table 1) (see Chu & Prockop 1993, Kielty et al. 1993, Prockop & Kivirikko 1995, Bateman et al. 1996, Adachi et al. 1997 for reviews). After transcription of the procollagen genes and processing of the pre-mRNAs, the α- chains are synthesized on the ribosomes of the endoplasmic reticulum. The signal peptides at the aminoterminal ends of the chains are removed by a signal peptidase after translocation across the membrane of the rough endoplasmic reticulum (RER). A large number of post-translational modifications are involved in collagen biosynthesis, for 15 which several specific enzymes are required. Proline and lysine residues in the Y-position are hydroxylated to 4-hydroxyproline and hydroxylysine, respectively, and some of the prolines in the X-position are hydroxylated to 3-hydroxyproline. Galactosyl moieties can be attached to some hydroxylysine residues by hydroxylysyl galactosyltransferase, and glucose can be attached to some of the galactosyl hydroxylysine residues by galactocyl hydroxylysyl glucosyltransferase. In addition to the lysines in the triple-helical region, the lysines in the short telopeptides can also be glycosylated (Cannon & Davison 1978). Furthermore, the carboxyterminal propeptides of at least type I and III procollagens contain asparagine-linked high-mannose type oligosaccharide side chains. Intra- and interchain disulfide bonds are formed between cysteine residues by protein disulfide isomerase (PDI) (Koivu & Myllylä 1987), which is a subunit of prolyl-4-hydroxylase. Procollagen folding and association into a triple helix requires the involvement of molecular chaperones, such as HSP47 (Nakai et al. 1992, Clarke et al. 1993). HSP47 is needed in collagen assembly and also during the transport of collagen from the endoplasmic reticulum to Golgi (see Nagata 1998). Increased HSP47 in the endoplasmic reticulum has been demonstrated in osteogenesis imperfecta fibroblasts, where the structure of type I collagen can be abnormal (Kojima et al. 1998). In addition to HSP47, PDI has been shown to have chaperone functions as well (Wilson et al. 1998). After the folding of the three polypeptide chains, the procollagen molecule is transported to the Golgi complex, where phosphorylation of some serine residues and sulfation of tyrosine residues in the propeptides of type I and III collagens take place, respectively. Finally, the procollagens are secreted out of the cell, where the extracellular processing takes place. Propeptides of the fibrillar collagens are cleaved off by specific N- and C- proteinases, the collagen is assembled into fibrils, and cross-links are formed between certain lysine and hydroxylysine residues after their conversion into aldehyde derivatives by lysyl oxidase. For further reading and reviews on the biosynthesis of collagen, see Kielty et al. (1993), Prockop & Kivirikko (1995) and Bateman et al. (1996).

Table 1. Collagen types, their genes, molecular forms and distributions in human tissues. Type Subgroup Gene Molecular forms Distribution

I Fibrillar COL1A1 [α1(I)]2α2(I) Most tissues COL1A2 [α1(I)]3

II Fibrillar COL2A1 [α1(II)]3 Cartilage, cornea, vitreous humor, intervertebral disc

III Fibrillar COL3A1 [α1(III)]3 Soft tissues, with type I collagen

IV Network-forming COL4A1 [α1(IV)]2α2(IV) Basement membranes COL4A2 [α3(IV)]2α4(IV) COL4A3 other forms COL4A4 COL4A5 COL4A6 16 Table 1. continued.

Type Subgroup Gene Molecular forms Distribution

V Fibrillar COL5A1 [α1(V)]3 Minor amounts in most COL5A2 α1(V)α2(V)α3(V) tissues with type I collagen COL5A3 other forms VI Beaded filament- COL6A1 α1(VI)α2(VI)α3(VI) Minor amounts in most forming COL6A2 tissues COL6A3

VII Anchoring COL7A1 [α1(VII)]3 Skin, cervix, oral mucosa fibril-forming

VIII Network-forming COL8A1 [α1(VIII)]2α2(VIII) Many tissues COL8A2 IX FACIT COL9A1 α1(IX)α2(IX)α3(IX) With type II collagen, COL9A2 e.g. cartilage COL9A3

X Network-forming COL10A1 [α1(X)]3 Hypertrophic cartilage XI Fibrillar COL11A1 α1(XI)α2(XI)α1(II) With type II collagen COL11A2 other forms e.g. cartilage COL2A1

XII FACIT COL12A1 [α1(XII)]3 Many tissues with type I collagen XIII Transmembrane COL13A1 Unknown Minor amounts in many domain tissues

XIV FACIT COL14A1 [α1(XIV)]3 Many tissues with type I collagen XV MULTIPLEXINs COL15A1 Unknown Many tissues

XVI FACIT COL16A1 [α1(XVI)]3 Many tissues

XVII Transmembrane COL17A1 [α1(XVII)]3 Hemidesmosomes of domain stratified squamous epithelia XVIII MULTIPLEXINs COL18A1 Unknown Liver, kidney, placenta, etc. XIX FACIT COL19A1 Unknown Several tissues See text for references. FACIT = fibril-associated collagens with interrupted triple helices MULTIPLEXINs = proteins with multiple triple helix domains and interruptions 17 2.2. Fibril-forming collagens

The type I, II, III, V and XI collagens make up the group of fibrillar collagens. By forming highly organized fibers, these collagens provide the structural support for the body in the skeleton, skin, blood vessels, nerves, intestines, and the fibrous capsules of the organs. The characteristic features of fibrillar collagens include triple-helical domains consisting of about 1000 amino acids per chain, propeptides at both ends of the molecular precursor, and finally, highly complex fibrillogenesis. Collagen fibril formation is intrinsically a self-assembly process, but considerable cellular control is also involved. For references and further reading, see Prockop & Kivirikko (1995) and Kadler et al. (1996). Type I collagen is the most abundant collagen type, being present in most tissues, but mainly in bone, tendon and skin. Type III collagen is often found in association with type I collagen in all soft tissues, but not in mineralized bone. In addition, hybrid fibrils of type I and III collagens are found in at least skin, tendon and amnion (Keene et al. 1987). Type II collagen is often found in cartilage and vitreous humor in association with type XI collagen (Sandberg et al. 1993), whereas type V usually occurs as a minor component with type I collagen fibers (Birk et al. 1988). In mammals, type I and III collagens account for approximately 70 % and 5 to 20 % of total collagen, respectively (see Adachi et al. 1997).

2.2.1. Type III collagen

2.2.1.1. General features and assembly of type III collagen

Type III collagen is a homotrimer of three α1(III) chains encoded by a single gene found on human chromosome 2 at location q24.3-q31 (Emanuel et al. 1985). This gene, which is called the COL3A1 gene, is about 44 kb in length and contains 52 exons (see Chu & Prockop 1993). The isolation of collagen molecules with the chain composition of [α1 (III)]3 was first reported almost 30 years ago (Miller et al. 1971). One α1(III) chain consists of an uninterrupted triple-helical domain about 1000 amino acids and 300 nm in length. Similarly to the other fibrillar collagens, type III collagen is also synthesized as a large precursor molecule containing propeptides at both ends. These propeptides are then enzymatically cleaved off during the extracellular processing of type III collagen. Between these propeptides and the main triple helix, there are short, non-triple-helical telopeptides containing the sites essential for intermolecular cross-linking (see Yamauchi & Mechanic 1988, Kielty et al. 1993, Prockop & Kivirikko 1995 for reviews). The folding of collagen requires association of the α-chains through the globular C- terminal domains, formation of a nucleus of triple-helical conformation at the C- terminus, and finally, propagation of the nucleus from the C-terminus to the N-terminal domain with a zipper-like mechanism (see Engel & Prockop 1991, Baum & Brodsky 1999 for reviews). Intra- and interchain disulfide bonds are formed within C-propeptide 18 and C-telopeptide, and when the triple helix is forming, disulfide bonds are also formed within the N-propeptide (Bruckner et al. 1978). There are a number of peptide bonds, to which proline contributes the nitrogen, in the collagen polypeptide chains (see Engel & Prockop 1991). Since the peptide bonds must be in the trans configuration in the native triple helix, the cis-trans isomerization limits the rate of collagen folding. Type III collagen folds in two phases. In the fast phase, about 25 tripeptide units fold between the disulfide bonds at the C-terminal region and the first cis peptide bond. During the second, much slower phase, which is determined by the slow rate of cis-trans isomerization of the cis peptide bonds, the triple helix then propagates in a zipper-like fashion (Bächinger et al. 1978, Bächinger et al. 1980, Bächinger 1987, see Engel & Prockop 1991). The currently available data indicates that the native type III collagen molecule is unusually susceptible to proteolysis by trypsin when compared to type I collagen, which is highly resistant to this proteolytic activity. This phenomenon has been attributed to the lack of helicity in the trypsin-susceptible region of the type III collagen molecule (Miller et al. 1976). Furthermore, human and various animal type III collagens are differentially susceptible to hydrolysis by human collagenase 1 (MMP-1), probably due to variation in helix stability (Welgus et al. 1985).

2.2.1.2. Aminoterminal propeptide of type III procollagen, PIIINP

The aminoterminal propeptide of type III procollagen, PIIINP, is cleaved off by a specific N-proteinase during the extracellular processing of type III procollagen molecules (Halila & Peltonen 1986). The PIIINP molecule has a molecular weight of approximately 42 000 and consists of a cysteine-rich globular domain (Col 1), a triple-helical region (Col 3), and another non-collagenous domain (Col 2) ending in the N-telopeptide. The Col 1 domain contains 5 intrachain disulfide bridges along with 79 amino acids. The Col 2 domain is an area 12 amino acids long and located in the carboxyterminal region of the propeptide containing three interchain disulfide bonds. Finally, the Col 3 domain with 39 amino acids form a triple-helical structure between the Col 1 and Col 2 domains (Bruckner et al. 1978) (Fig. 1). The aminoterminus, known as Col 1, is the most immunogenic region in the molecule (Rohde et al. 1983). The Col 1 region has a substrate site for transglutaminase (Bowness et al. 1987) (Fig. 1), which appears to play an important role in, for example, wound healing (Bowness et al. 1988) and the binding of other molecules, such as fibrinogen and LDL, to procollagen (Bowness et al. 1989). The PIIINP molecule is stabilized by three interchain and five intrachain disulfide bridges. The melting temperature of the collagenous domain is 53°C, and refolding of the helix from denatured peptides takes place extremely fast, probably due to the disulfide bonds that already keep the three chains together (Bruckner et al. 1978). Tyrosine sulfation is a widespread post-translational modification (Huttner 1982). The presence of acidic amino acids, most frequently an aspartic or glutamic acid, is the key feature of tyrosine sulfation sites (Lee & Huttner 1983). Sulfation is a general characteristic of type III procollagen, into which the anion is incorporated in the form of tyrosine-O-sulfate (Jukkola et al. 1986) (Fig. 1). Interestingly, inhibition of tyrosine 19 sulfation does not seem to affect the secretion or processing of type III procollagen in in vitro experiments (Jukkola et al. 1990). Since it has been suggested that charged and hydrophobic amino acids could contribute to the prevention of premature fibril formation and the regulation of collagen fiber diameter (see Fleischmajer & Perlish 1986), one possible function of tyrosine sulfation lies in collagen fibrillogenesis. Furthermore, these sulfate residues, due to their negative charge, could influence the interactions of type III procollagen with other extracellular matrix components. Intact circulating PIIINP is cleared by scavenger receptors located on the endothelial cells of the liver (Smedsrød et al. 1988, Melkko et al. 1994). The highly acidic nature of the PIIINP molecule due to sulfation is believed to play a role in this recognition. The most aminoterminal part, Col 1, is present as monomers, which are eventually removed via the kidneys. Because the metabolism of type III collagen is altered in various clinical situations, e.g. fibrosis and malignant processes, the measurement of PIIINP in human serum by radioimmunoassay has established its status in clinical practice (Niemelä et al. 1985, Risteli et al. 1988a, Risteli & Risteli 1995). Elevated levels of circulating PIIINP have been observed in hormonally induced growth (Trivedi et al. 1989) and in diseases presenting with fibrosis (see Risteli & Risteli 1990 for a review). Decreased PIIINP levels have been observed in Ehlers-Danlos syndrome type IV (Steinmann et al. 1989). Furthermore, PIIINP assay may also be useful in monitoring dilated cardiomyopathy (Klappacher et al. 1995) or vascular sclerosis in hypertension (see Weber 1998). In addition, measurement of PIIINP concentrations in cerebrospinal fluid may be of value in the clinical diagnosis of meningeal fibrosis (Sajanti & Majamaa 1999).

Fig. 1. Schematic illustration of the aminoterminal propeptide of type III procollagen. The arrow represents the N-proteinase cleavage site. Possible tyrosine sulfation sites are indicated by asterisks. The transglutamination site is marked by a circle. For clarity, these modifications are only indicated in one of the three chains. The bars indicate the location of the alleged disulfide bonds between cysteine residues (modified from Niemelä 1985, Risteli & Risteli 1990). 20 2.2.1.3. Aminoterminal telopeptide of type III collagen, IIINTP

A non-helical segment of 14 amino acid residues in type III collagen, the amino-terminal telopeptide, has been recognized as a fundamental element for proper functioning of the protein (Becker et al. 1976, Glanville & Fietzek 1976), since it contains intra- and intermolecular cross-linking sites. In addition, the tyrosine residues in IIINTP could be sulfated, as described in the previous chapter concerning PIIINP. After the cleavage of the propeptides, the collagen molecules self-assemble into fibrils, which are then stabilized by cross-links. In order to form these bonds, the collagen molecules are aligned in a staggered manner, so that the molecules are laterally displaced by about one-quarter of their length relative to the previous molecule along the axis of the fibril. Thus, the individual 300-nm collagen molecules overlap each other by a distance of 67 nm, leaving a gap of 40 nm between the ends of the non-overlapping molecules. For further reading and references, see Kielty et al. (1993), Prockop & Kivirikko (1995) and Kadler et al. (1996). Four homologous loci of cross-linking are present in the molecules of type III collagen. Two are aldehyde sites, one in each telopeptide region. The other two sites are hydroxylysines located in the helical region at about 90 residues from each end of the molecule (see Eyre et al. 1984 for a review). During extracellular processing, a trivalent pyridinoline cross-link structure is formed between the lysine/hydroxylysine residues in two aminoterminal telopeptides from the same collagen molecule and one in the carboxyterminal triple-helical region of the neighboring collagen molecule (Fig. 2). There is also data to support the existence of divalent cross-links joining only one N-telopeptide to the carboxyterminal helical region (Henkel et al. 1979).

Fig. 2. Structure of the cross-linked aminoterminal telopeptide of type III collagen, IIINTP. The actual non-triple-helical telopeptide domain is underlined. The arrows represent the N- proteinase and trypsin cleavage sites. The asterisks indicate tyrosine residues, which can possibly be sulfated. The pyridinoline cross-link connects two aminoterminal telopeptides of one molecule with the carboxyterminal helical region of the adjacent collagen molecule. 21 2.2.1.4. Type III pN-collagen

The aminoterminal propeptides of both type I and III collagens have been suggested to have a regulatory role in collagen fibril formation (see Fleischmajer & Perlish 1986). The aminopropeptide of type III collagen is usually released more slowly than the carboxyterminal propeptide (Fessler et al. 1981). Occasionally, the type III collagen molecule retains its aminoterminal propeptide domain (PIIINP), resulting in type III pN- collagen. Fleischmajer et al. (1981) found type III pN-collagen along thin collagen fibrils in human skin. Hedman et al. (1982) cultured amniotic epithelial cells, which only synthesize type III collagen, and demonstrated the incorporation of pN-type III collagen in early fibrillogenesis. Studies on human adult skin indicate that when the fibril reaches the diameter of 35 – 55 nm, the aminoterminal propeptide of type III procollagen is retained at the surface of the fibril, possibly resulting in the prevention of further fibril growth (Fleischmajer et al. 1985) (Fig. 3). This differs from type I collagen, where the aminoterminal propeptide is cleaved when the fibrils reach a diameter of 35-45 nm, thus suggesting that the aminopropeptide has only a transient role in fibrillogenesis (Fleischmajer et al. 1985). The same study also suggested that type I pN-collagen was only present on very thin collagen fibrils, which are typically seen in fetal skin, whereas type III pN-collagen was evident in both adult and fetal skin on fibrils ranging from 25 to 55 nm in diameter. This observation has also been previously reported in studies on chicken embryo skin (Fleischmajer et al. 1983).

Fig. 3. Metabolism of type III collagen fibrils. When the procollagen molecule has been secreted out of the fibroblast, the carboxyterminal propeptide (PIIICP) is cleaved. The remaining type III pN-collagen, which still retains the aminoterminal propeptide domain, is attached to the surface of the fibril. Depending on the thickness of the existing fibril, the PIIINP molecule is cleaved off and released into the circulation. However, if the whole pN- collagen is degraded, the collagenous domain of PIIINP can also be further cleaved, so that the most aminoterminal region of the molecule, known as Col 1, is released (modified from Jensen & Høst, 1997). 22 Evidence is also available to indicate that type III collagen remains associated with the type I-containing collagen fibrils after removal of the amino- and carboxyterminal propeptides, and that several tissues contain copolymers of at least type I and III collagens (Keene et al. 1987, Fleischmajer et al. 1990). Furthermore, there are several observations to support the hypothesis that type III pN-collagen also regulates the diameter of type I collagen fibrils by coating the surface of the fibrils and thereby allowing tip growth but no lateral growth of the fibrils, leaving them thinner (Fleischmajer et al. 1990, Romanic et al. 1991). Type III pN-collagen is suggested to be an important component of reticular fibers in lymph nodes, both under normal conditions and in Hodgkin’s disease (Karttunen et al. 1988). The distribution of type III pN-collagen in the reticular fibers of the spleen is different in developing and adult tissues, suggesting a possible role of type III collagen in the development and function of normal spleen (Liakka et al. 1995). Interestingly, type III pN-collagen is reduced in severely photodamaged skin, possibly resulting in altered organization of fibrillar collagen and thus contributing to the wrinkled appearance of injured skin (Talwar et al. 1995).

2.2.1.5. Carboxyterminal telopeptide (IIICTP) and propeptide (PIIICP) of type III collagen

The C-propeptide (amino acid residues 1206-1466 of the pro-α1(III)-chain) plays a role in the early stages of procollagen assembly (see McLaughlin & Bulleid 1998). The C- propeptide has a recognition signal, which directs the type-specific assembly of individual procollagen chains (Lees et al. 1997). By exchanging this sequence between different procollagen chains, it is possible to direct chain association and, potentially, assemble molecules with defined chain compositions (Lees et al. 1997). Bulleid et al. (1997) showed that the triple helix nucleation took place even when the C-propeptide was substituted with a transmembrane domain. Thus, it was concluded that the C-propeptide is required only to ensure association of the monomeric chains. Furthermore, the same study demonstrated that triple helix formation is directed by the most carboxyterminal Gly-X-Y triplet of the triple helix and not by the propeptide. The C-propeptide of the α1(III) chain contains 8 cysteine residues, which are essential for the formation of intra- and interchain disulfide bridges. Disulfide bonds are thought to form between the C-propeptide chains prior to the triple helix formation (Lukens 1976). However, mutant chains lacking the ability to form interchain disulfide bonds are still able to assemble, indicating that interchain disulfide bonding within the C-propeptide or C-telopeptide is not required for triple helix formation (Bulleid et al. 1996). The carboxyterminal propeptides of the secreted type III procollagen molecules are rapidly cleaved off by the same C-proteinase that is also known to cleave type I and II collagens (Kadler & Watson 1995). The C-telopeptide of the α1(III) chain contains a conserved partial sequence Glu-Lys-Ala involved in enzymatic cross-linking. This sequence has a polar region where nuclear magnetic resonance studies have suggested existence of a β-turn (Liu et al. 1993, Henkel 1996). Apart from the interchain disulfide 23 bridges within the C-telopeptide, there is also evidence of disulfide cross-links between adjacent collagen molecules (Cheung et al. 1983, Henkel 1996).

2.2.1.6. Distribution of type III collagen in tissues

Type III collagen is the second most abundant collagen in human tissues and occurs particularly in tissues exhibiting elastic properties, such as skin, blood vessels and various internal organs. Several studies concerning the relative amounts of type I and III collagen have been published, all giving more or less different results, depending mainly on the tissue extraction methods used to solubilize the insoluble collagen (see Rauterberg et al. 1993). For example, the initial work on arterial collagens suggested a predominance of type III collagen over type I, but later results have suggested that type I collagen accounts for 55 – 88 % of the total amount of these two (see Barnes 1985 for a review). Hanson & Bentley (1983) reported in one study that the vena cava contained 21 % of type III collagen, whereas achilles tendon exhibited only type I collagen. The type III collagen content of bone in normal subjects is about 4 % (Bailey et al. 1993). Type III collagen accounts for 11 % and 20 % of the total collagen protein in normal myocardium and valve leaflets, respectively (see Weber 1989). Dilated cardiomyopathy has recently been found to be associated with significant changes in the collagen type I/III ratio (Pauschinger et al. 1999). Liver has been reported to have as much as 45 % and human lung 21 % of type III collagen (Kelley et al. 1984, van Kuppevelt et al. 1995). The proportion of type III collagen out of the total collagen content is about 20 % and 50 % in adult human skin and embryonic dermis, respectively (Epstein & Munderloch 1975, Cheung et al. 1990). There are also age-related changes in the relative amounts of type I and III collagens. In studies on rats, type III collagen represented about one-third of the sum of types I and III in all tissues at 2 weeks of age. After this, the proportion of type III collagen increased for up to one year in both heart (54 %) and lung (48%), whereas it decreased in skin (19 %) (Mays et al. 1988). Similar results on human skin collagen had also been obtained previously (Chan & Cole 1984). Tagasako et al. (1992) reported the total quantitity of type III collagen to decrease upon ageing in all tissues.

2.2.1.7. Type III collagen mutations

The whole nucleotide and amino acid sequences of the human type III procollagen chain were presented in 1989 (Ala-Kokko et al. 1989). Since then, several disease-causing mutations in the COL3A1 have been found (see Kuivaniemi et al. 1997 for a review, Schwartze et al. 1997, Bateman et al. 1998, Gilchrist et al. 1999). Most of the mutations have been seen in patients with Ehlers-Danlos syndrome IV, the most severe form of EDS, which may cause sudden death due to a rupture of large arteries and hollow organs, in addition to causing skin and joint changes. The mutations include various single-base substitutions that mostly convert a codon of glycine to a codon of some other amino acid, 24 in combination with deletions and mRNA splicing defects (see Kuivaniemi et al. 1997). EDS III is characterized by general laxity of joints and extensibility of skin, and a glycine-to-serine mutation in the α1(III) chain has also been reported in this disease (Narcisi et al. 1994). The different mutations in the COL3A1 gene have effects on the secretion, folding, fibrillogenesis, and thermal stability of collagen as well as the tissue architecture (Anderson et al. 1997, Smith et al. 1997). In order to study the role of COL3A1 in the body, the murine COL3A1 gene was inactivated in embryonic stem cells (Liu et al. 1997). About 10 % of the homozygous mutant animals survived, but had a shortened lifespan. The major cause of death in the mutant knock-out mice was rupture of the major blood vessels. Furthermore, it was shown that type III collagen is essential for normal fibrillogenesis of type I collagen in the cardiovascular system (Liu et al. 1997). Mutations in the type III collagen gene are also associated with the pathogenesis of abdominal aortic aneurysms without other features of EDS IV (Kontusaari et al. 1990, Anderson et al. 1996). Spontaneous cervical arterial dissections have also been studied for mutations, but even though some of the patients have a low ratio of type III to type I collagen, mutations in the type III collagen gene are not considered responsible for cervical arterial dissections (van den Berg et al. 1998). The structural unstability of type III procollagen observed in patients with cerebral aneurysms (Majamaa et al. 1992) might suggest mutations in the COL3A1 gene. Such mutations are, however, an infrequent cause of these or any other familial aneurysms without other signs of connective tissue manifestations (see Kuivaniemi et al. 1997).

2.2.2. Type I collagen and its metabolites

Type I collagen is the major collagen type in most tissues. For example, most of the organic matrix of bone consists of type I collagen. The main cells synthesizing type I collagen in soft tissues are fibroblasts, whereas bone collagen is produced by osteoblasts. The most abundant molecular form of type I collagen is a heterotrimer of two α1(I) chains and one α2(I) chain. In addition to this classical type I collagen, there is evidence of a type I α1-homotrimer collagen consisting of three α1(I) chains (Jimenez et al. 1977, Wohllebe & Carmichael 1978). A third variant, the laminin-binding collagen found in colon and breast cancer tissue, has been suggested to consist of one α1(I) chain, one α1(III) chain, and a unique acidic chain (Pucci-Minafra et al. 1993). Similarly to other fibrillar collagens, type I collagen is synthesized as large precursor molecules containing extra propeptide domains at both ends. The aminoterminal propeptide (PINP) and the carboxyterminal propeptide (PICP) of type I collagen are cleaved off enzymatically by specific endoproteinases, after the procollagen molecule has entered the extracellular space (see Risteli & Risteli 1999). Cleavage of the carboxyterminal propeptide is required for the initiation of fibril formation (Kadler et al. 1987). The molecules can retain their aminoterminal propeptide for a while, resulting in type I pN-collagen, which has been suggested to regulate the lateral growth of type I collagen fibers (Fleischmajer et al. 1985). The PINP molecule is similar to PIIINP and consists of three structural subdomains: Col 1 (the globular part in the most 25 aminoterminal region containing intrachain disulfide bridges, which is also phosphorylated), Col 2 (the short non-helical part) and Col 3 (the central collagenous domain) (see Risteli & Risteli 1999 for a review). The molecular mass of the PINP molecule is 35 000, and it is cleared from the blood via scavenger receptors of liver endothelial cells (Melkko et al. 1994). The PICP molecule, in turn, is cleared via the mannose receptor of the same liver endothelial cells (Smedsrød et al. 1990) and has a molecular mass of about 100 000. The PICP molecule is a glycoprotein containing high- mannose type oligosaccharides (Clark 1979). The concentrations of PINP and PICP antigens in serum have been shown to reflect the synthesis of type I collagen in both physiological and pathological situations (see Risteli & Risteli 1999). Radioimmunossays for these two metabolites have been developed for clinical use (Melkko et al. 1990, Melkko et al. 1996). Especially PINP has proved useful in detecting changes in the rate of bone collagen metabolism, such as those seen during estrogen treatment in postmenopausal women (Suvanto-Luukkonen et al. 1997), Paget’s disease of bone (Sharp et al. 1996), and in patients with aggressive breast cancer (Jukkola et al. 1997). Between the propeptides and the helical collagen molecule, there are short, non-triple- helical telopeptide domains containing sites for intra- and intermolecular cross-linking. The lysine or hydroxylysine residue of the aminoterminal telopeptide of type I collagen can be oxidized into aldehyde, which can further mature into trivalent pyridinoline or pyrrole cross-links (see next chapter). The carboxyterminal telopeptide forms a divalent cross-link by binding to a helical region of another collagen molecule, which, in turn, can mature into pyridinoline or other mature structures (see next chapter) (see Bailey et al. 1998, Risteli & Risteli 1999). An interesting feature in the C-terminal telopeptide is the β-isomerization of the aspartic acid as a result of the normal ageing process (Fledelius et al. 1997), the biological effects of which are yet to be elucidated. As for the propeptides of type I collagen, namely PINP and PICP, specific assays have also been developed for the telopeptide regions in order to monitor the degradation of type I collagen in serum and urine. The NTx assay detects the trivalently cross-linked aminoterminal telopeptide structure of type I collagen (Hanson et al. 1992). The original CrossLaps assay was developed for a synthetic peptide, the amino acid sequence (EKAHDGGR) of which was derived from the carboxyterminal telopeptide (Bonde et al. 1994). The ICTP assay, which was also used in the current studies, detects degradation products of mature type I collagen with a trivalent cross-link (Risteli et al. 1993, Sassi et al. 2000).

2.2.3. Enzymatic cross-linking of type I and III collagens

2.2.3.1. Lysyl oxidase

Lysyl oxidase is a copper-dependent extracellular enzyme, which is responsible for the oxidation of certain lysine and hydroxylysine residues in collagen and lysine residues in elastin to form α-aminoadipic-δ-semialdehyde, which, in turn, can react with the 26 neighboring molecules to form covalent cross-links (see Smith-Mungo & Kagan 1998 for a review). The cross-linking process goes on continuously on the surface of the growing fibril and therefore occurs at a very early stage of fibril formation (Cronlund et al. 1985). Up to date, lysyl oxidase, LO (Hämäläinen et al. 1991), has two identified isoforms: lysyl oxidase-like protein, LOL (Kenyon et al. 1993, Kim et al. 1995) and lysyl oxidase- related protein, LOR (Saito et al. 1997). The existence of these highly related genes implies the presence of a lysyl oxidase gene family, additional members of which may yet remain to be identified (see Smith-Mungo & Kagan 1998). The expression of lysyl oxidase responds to many physiological conditions, such as growth, development, ageing and wound healing. A number of genetic diseases affecting copper metabolism and the function of lysyl oxidase, such as cutis laxa and Menkes syndrome, have been identified (see Smith-Mungo & Kagan 1998). LO expression is increased in fibrotic conditions (see Kagan 1994). An opposite situation prevails in malignant cell line cultures, where LO expression is down-regulated at the transcriptional level (Hämäläinen et al. 1995). In addition, there is evidence of a lack of LO in the stroma accompanying invading tumors, consistent with the deficiency of properly cross- linked fibers (Peyrol et al. 1997, see Smith-Mungo & Kagan 1998). Lysyl oxidase is also a candidate gene for a tumor suppressor (Contente et al. 1990).

2.2.3.2. Cross-linking pathways

Collagen cross-linking is essential for the tensile strength of tissue, since it increases the resistance of collagen fibers against proteolysis (Vater et al. 1979). Two pathways of enzymatic cross-linking have been described for fibrillar collagens: one based on lysine aldehydes, and another based on hydroxylysine aldehydes. The cross-linking of type III collagen has not been so widely studied as that of type I collagen, but the cross-linking pathways have been thought to be similar in these two collagen types. The specific hydroxylation of lysine residues in the telopeptides is responsible for the nature of the cross-link and thus for the biomechanical properties of the tissue. The lysine aldehyde pathway predominates at least in adult skin, cornea, sclera and rat tail tendon, whereas the hydroxylysine aldehyde pathway predominates in bone, cartilage, ligaments, tendons and most internal organs, being also present in embryonic skin (see Eyre 1987, Yamauchi & Mechanic 1988, Bailey et al. 1998 for reviews). There is recent evidence of a bone- specific telopeptide lysyl hydroxylase, which could be partially responsible for the tissue- specific collagen cross-linking (Bank et al. 1999, Uzawa et al. 1999). Cross-linking does not occur only between the chains of one collagen type: type I collagen, for example, may be co-polymerized with type III collagen (Henkel & Glanville 1982). Enzymatic cross-linking of collagen is illustrated in Figure 4. The telopeptide lysyl and hydroxylysyl aldehydes are capable to react with the unmodified lysine or hydroxylysine residues of another collagen molecule to form immature reducible divalent cross-links. In addition, two lysyl aldehydes from the same molecule can react with each other to form an intramolecular aldol condensation product (ACP). In the hydroxylysyl aldehyde pathway, the divalent cross-links spontaneously undergo an Amadori rearrangement, resulting in acid and heat-stable keto-imine cross-links, which then 27 mature further into trivalent pyridinoline and pyrrole cross-links. Divalent cross-links derived from the lysyl aldehyde pathway can react with hydroxylysine and histidine residues resulting in trivalent histidinohydroxylysinonorleucine (HHL) and tetravalent histidinohydroxymerodesmosine (HHMD) cross-links. However, the in vivo existence of the latter cross-link along with the pyrroles is controversial. Of the trifunctional cross- links, histidinohydroxylysinonorleucine (HHL) is found at least in skin and cornea, whereas hydroxylysylpyridinoline (HL-Pyridinoline) is very typical of type II collagen in cartilage, and lysylpyridinoline (L-Pyridinoline) in calcified tissue. HL-Pyridinoline has been reported to be located equally at both the N- and the C-terminus, whereas L- Pyridinoline and the pyrroles are preferentially located at the N- terminus of the molecule (Hanson & Eyre 1996). The other mechanism of intermolecular cross-linking of collagen is the non-enzymatic reaction with glucose. For references and more thorough reviews on the cross-linking of collagen, see Eyre et al. (1984), Eyre (1987), Yamauchi & Mechanic (1988) and Bailey et al. (1998).

Fig. 4. Schematic representation of enzymatic cross-link formation and maturation in type I and III collagens (modified from Yamauchi & Mechanic 1988, Bailey et al. 1998). The possible maturation product of deH-LNL is currently unknown (marked by a question mark). For abbreviations, see the list of abbreviations at the beginning of this book.

The cross-link profiles are, to some extent, tissue-specific and may change with age. For example, the content of reducible cross-links in bone collagen decreases markedly between birth and 25 years (Eyre et al. 1988), followed by a suggested maturation mainly 28 into pyrrolic trivalent structures, whereas the amount of pyridinolines may remain almost the same (Knott et al. 1995). The age-related changes causing a functional deficiency of collagenous tissues are possibly due to increased intermolecular cross-linking (Bailey et al. 1998). In addition to normal ageing, changes in collagen cross-linking may be involved in the pathogenesis of many diseases, including liver fibrosis (Ricard-Blum et al. 1996), dilated cardiomyopathy (Gunja-Smith et al. 1996) and diabetes (Reiser 1991). Cannon & Davison (1978) reported that type III collagen isolated from bovine amnion tissue contained large amounts of immature lysine-derived cross-links, particularly dihydroxylysinonorleucine. In addition, almost 80 % of these cross-links were glycosylated. Divalent hydroxylysinonorleucine cross-links containing only one N- telopeptide plus the carboxyterminal helical region and trivalent mature cross-links containing two N-telopeptides in addition to the helical region have been found in human uterine leiomyoma (Henkel et al. 1979). The cross-link analysis of the C-terminus of type III collagen from calf aorta revealed covalent cross-linking between the C-terminal telopeptide of type III molecule and the N-terminal helical cross-linking region of another. This structure contained mainly dihydroxylysinonorleucine with a small amount of hydroxylysinonorleucine, indicating that the lysine residues involved in cross-linking are both hydroxylated (Henkel 1996). In addition to the intermolecular aldehyde-derived cross-links, type III collagen also has disulfide cross-links (Cheung et al. 1983). For example, Henkel (1996) suggested a presence of intermolecular disulfide cross-links between the cysteine residues of adjacent collagen molecules in the C-terminus.

2.2.4. Degradation of type I and III collagens

Degradation of collagens can occur via intracellular or extracellular routes. Matrix metalloproteinases (MMPs), which are zinc-dependent endopeptidases, play a crucial role in the extracellular degradation of collagens. Breakdown of the ECM is needed in many physiological situations, such as development, tissue repair and angiogenesis. Several pathological conditions involve changes in the synthesis and activity of the matrix metalloproteinases. MMPs are grouped into collagenases, gelatinases, stromelysins and stromelysin-like MMPs, membrane type-MMPs (MT-MMPs), and novel MMPs, depending on their substrate specificity and primary structure. MMP activity is regulated by gene transcription, activation of the inactive proenzyme forms and the action of specific inhibitors, the major ones being the group of tissue inhibitors of MMPs (TIMPs). For references and further reading, see Kähäri & Saarialho-Kere (1997) and Kähäri & Saarialho-Kere (1999). Collagenase-1 (fibroblast collagenase, MMP-1), collagenase-2 (neutrophil collagenase, MMP-8) and collagenase-3 (MMP-13) (Freije et al. 1994) can cleave fibrillar collagens at about three fourths of the way from the aminoterminal end, resulting in two separate fragments susceptible to denaturation and further proteolysis (see Kähäri & Saarialho-Kere 1997). MMP-8 preferentially degrades type I collagen, whereas MMP- 1 cleaves type III and MMP-13 type II collagen (Knäuper et al. 1996a, see Kähäri & Saarialho-Kere 1997). In addition, human MMP-13 cleaves type I collagen in the aminoterminal non-helical telopeptide (Krane et al. 1996). 29 Gelatinases are very important in the final degradation of collagens after cleavage by collagenases. At least 72-kDa gelatinase (gelatinase A, MMP-2) has been shown to cleave native type I collagen into fragments identical to those generated by collagenases (Aimes & Quigley 1995). MMP-2 has also been shown to activate at least MMP-13 (Knäuper et al. 1996b). Type I collagen, similarly to the types II and V, is digested in the N-terminal non-helical telopeptide by 92-kDa gelatinase (gelatinase B, MMP-9) (Okada et al. 1995). The stromelysins and stromelysin-like MMPs are able to degrade a wide range of substrates in the extracellular matrix, including some of the collagens, but not type I and III collagens in their native form. Of the membrane-type MMPs, MT1-MMP (MMP-14) degrades the native collagens I, II and III (D’Ortho et al. 1997, Ohuchi et al. 1997) and activates gelatinases and MMP-13 (see Murphy et al. 1999). MT3-MMP has been shown to cleave type III collagen (Matsumoto et al. 1997). The substrate specificities of the newer MMPs are yet to be established. In addition to extracellular collagen degradation by specific endoproteinases, such as MMPs, fragments of collagen fibers can be endocytosed and degraded within the lysosomes. Furthermore, part of the newly synthesized collagen can be degraded before it has even been secreted out of the cell (see Everts 1996 et al. for a review). Intracellular degradation plays an important role particularly when the synthesized collagen is abnormal due, for example, to insufficient post-translational modifications or mutations (Berg et al. 1980, Bateman et al. 1984). In addition to the degradation of type IV collagen in basement membranes, destruction of the stromal type I and III collagens is an essential step in the invasive and metastatic behavior of tumor cells (see Kähäri & Saarialho-Kere 1999 for a review). TIMPs play a key role in malignant tissue, not only by controlling the activity of MMPs, but also by affecting cell proliferation and survival (see Blavier et al. 1999). In addition to malignant processes, MMPs or TIMPs have been associated with at least atherosclerosis (Sukhova et al. 1999, Lindstedt et al. 1999), abdominal aortic aneurysms (see Grange et al. 1997), congestive heart failure (Spinale et al. 1999), dilated cardiomyopathy (Thomas et al. 1998), rheumatoid arthritis (Maeda et al. 1995) and fibrotic conditions, such as scleroderma of the skin (see Kähäri & Saarialho-Kere 1997). Unrevealing the detailed structures of MMPs and TIMPs and their role in the pathogenesis of diseases have provided a basis for the development of drugs aimed at modulation of MMP activity (see Skotnicki et al. 1999). Measurement of plasma or serum MMP and TIMP levels may provide important data for selecting and following patients considered for treatment with these drugs (see Zucker et al. 1999 for a review). In addition to MMPs, there are lysosomal cysteine proteinases, such as cathepsins, along with different serine and aspartic proteinases, all capable of degrading proteins of the extracellular matrix. Cathepsin K is abundant in at least human osteoclasts (Drake et al. 1996) and macrophages (Sukhova et al. 1998). Cathepsin K cleaves native type I collagen at the N-terminal end of the triple helix and at multiple sites within the helical region (Garnero et al. 1998, Kafienah et al. 1998). Furthermore, this enzyme cleaves the trivalently cross-linked C-terminal fragment of type I collagen (Sassi et al. 2000). As for the other cathepsins, cathepsin B has been reported to raise the activity of the other MMPs and to stimulate angiogenesis in malignancies, which together may promote tumor invasion (Kostoulas et al. 1999). 30 2.3. Type I and III collagens in atherosclerosis and abdominal aortic aneurysms

2.3.1. General features of atherosclerosis

Atherosclerosis is a response to injury to the endothelium, characterized by a fibroproliferative inflammatory reaction, degenerative changes and accumulation of extracellular matrix and lipids. One of the parameters associated with the endothelial cell dysfunction that results from exposure to oxidized LDL is increased adherence of macrophages and other inflammatory cells. Macrophages are known to localize subendothelially, where they scavenge lipids and turn into large foam cells. Proliferation and migration of vascular smooth muscle cells (SMCs) together with their transition from the “contractile” type into the “synthetic” phenotype are key features in the development of atherosclerosis. SMCs, T-cells and foam cells together form a fatty streak, which in turn may develop into an intermediate, fibrofatty lesion and, finally, into a fibrous plaque with calcification. The whole process is extremely complex, involving growth factors, cytokines, hormones, hemostatic agents and changes in collagen synthesis and smooth muscle cells (see Ross 1993, Rekhter 1999 for reviews).

2.3.2. Type I and III collagens in normal and atherosclerotic aorta

As in most connective tissues, the quantitatively most abundant group of arterial collagens are the fibril-forming type I and III collagens (see Barnes 1985 for a review). These collagen types together are responsible for tensile strength, but also contribute, together with elastin, somewhat to the extensile properties of the aorta (Dobrin et al. 1984, Dobrin & Mrkwicka 1994). The quantitative ratios of type I and III collagens in normal and atherosclerotic aortas have usually been determined after extraction of the collagens or collagen fragments from tissue specimens and quantification, after separation, into different types (see Barnes 1985, Rauterberg et al. 1993 for reviews). The increased proportion of type I collagen is thought to be characteristic of the atherosclerotic intimal plaque. In spite of extensive research, our knowledge of the quantitative ratios of type I and III collagens in normal and atherosclerotic aortic wall is incomplete and controversial (see Barnes 1985, Rauterberg et al. 1993). The currently existing data on type I and type III collagen ratios in atherosclerosis are shown summarized in Table 2. Data from immunohistological studies indicate only small differences in the normal aorta in the distribution of type I and III collagens (see Rauterberg et al. 1993). Both these collagen types are present in the normal fibrillar structures around smooth muscle cells and elastic membranes (McCullagh et al. 1980, Shekhonin et al. 1985, Katsuda et al. 1992). Furthermore, they co-localize with fibronectin (Shekhonin et al. 1987) and stain more intensively in the adventitia than in the media (Katsuda et al. 1992). The type I and III collagens appear together in the thickened intimas of atherosclerotic lesions, 31 where their staining intensity is also greater than in healthy aorta (Katsuda et al. 1992, see Rauterberg et al. 1993).

Table 2. Summary of data on type I and III collagen contents in normal and atherosclerotic human aorta. Method Finding Reference Pepsin digestion, 65 % type I in atherosclerosis McCullagh & SDS-PAGE 30 % type I in normal aorta Balian 1975

Pepsin digestion, 66 % type I in atherosclerosis Ooshima 1981 SDS-PAGE

Pepsin / 76 % type I in atherosclerosis Morton & CNBr cleavage, No major shift when compared to Barnes 1982 SDS-PAGE healthy aorta

CNBr cleavage, 58 % type I in normal aorta Hanson & CMC chromatography, 90 % type I in atherosclerosis Bentley 1983 SDS-PAGE

Pepsin, 38 % type I in normal aorta Hill & Harper DEAE-cellulose 1984 chromatography, CNBr, SDS-PAGE

CNBr, SDS-PAGE 70 % type I in normal aorta Halme et al. 1986

Pepsin, 70 % type I in atherosclerosis Leuschner & CMC chromatography No major shift when compared to Haust 1986 HPLC, CNBr, normal aorta SDS-PAGE

Pepsin, CNBr, 70 % type I collagen, type III ↓ Murata et al. SDS-PAGE with advancing atherosclerosis 1986

Pepsin, CNBr, 62 % type I in normal thoracic aorta Miller et al. 1991 HPLC, UV absorbance 47 % type I in normal abdominal aorta

CNBr, HPLC, 59 % type I in normal thoracic aorta Deyl et al. 1997 SDS-PAGE, capillary 52 % type I in normal abdominal aorta electrophoresis CMC = carboxymethyl-cellulose; CNBr = cyanogen bromide; DEAE = diethylaminoethyl; HPLC = high-performance liquid chromatography; SDS-PAGE = sodium dodecyl sulfate polyacrylamide gel electrophoresis 32 2.3.3. Synthesis and degradation of type I and III collagens in atherosclerotic plaques

The increased amount of type I collagen-producing cells has been established in all types of human atherosclerotic lesions (Andreeva et al. 1997). In animal models, the collagen content in atherosclerotic lesions has been shown to increase consistently over time (Kratky et al. 1999). The plaque collagen is mainly produced by smooth muscle cells (SMCs), but endothelial cells are also able to synthesize collagen (Canfield et al. 1992). Furthermore, cells different from typical SMCs produce type I collagen in human atherosclerotic lesions (Rekhter et al. 1996, Andreeva et al. 1997, Tintut et al. 1998). SMC phenotype, proliferation, migration and collagen production play an important role in the pathophysiology of atherosclerosis (see Rekhter 1999). There is evidence to indicate that ageing is associated with dysregulation of the proliferative responses of SMCs from old aortas, possibly predisposing to atherosclerotic changes (McCaffrey et al. 1988). In addition, DNA alterations have been found in SMCs of human atherosclerotic lesions, suggesting that mutational events may be part of the atherosclerotic process (De Flora et al. 1997). At least type I collagen synthesis has been clearly shown to be upregulated in atherosclerotic plaques compared with the media and normal arteries (Rekhter et al. 1993). Recently published data suggest that genes consistent with increased proliferation and collagen production are upregulated in SMCs during atherogenesis (Laury-Kleintop et al. 1999). A number of cytokines and growth factors produced and secreted by macrophages and other inflammatory cells have been described as being involved in the process of atherosclerotic plaque formation, in part by altering the rate of collagen synthesis (see Rekhter 1999 for a review). Transforming growth factor beta (TGF-β) is the most potent and consistent stimulator of collagen synthesis in vascular SMCs (Amento et al. 1991). There is evidence of genomic instability of the type II TGF-β1 receptor gene in human atherosclerotic and restenotic vascular cells, probably causing resistance to the antiproliferative effects mediated by this particular receptor type (McGaffrey et al. 1997). Platelet-derived growth factor (PDGF) (Amento et al. 1991), interleukin-1 (IL-1) (Amento et al. 1991), angiotensin-II (Kato et al. 1991), homocysteine (Majors et al. 1997), endothelin-1 (Rizvi et al. 1996) and oxidized LDL (Bachem et al. 1999) have also been reported to stimulate collagen synthesis in SMC culture. Human connective tissue growth factor (CTGF) is undetectable in normal blood vessels, but overexpressed in atherosclerotic lesions, suggesting that CTGF plays a role in atherogenesis (see Oemar & Lüscher 1997). Furthermore, mechanical stress has been shown to have a stimulatory effect on collagen synthesis (Kolpakov et al. 1995a). In contrast, fibroblast growth factor (Majors & Ehrhart 1993, Pickering et al. 1997), nitric oxide (Kolpakov et al. 1995b, Myers & Tanner 1998), interferon gamma (IFN-γ) (Amento et al. 1991) and prostaglandins (Fitzsimmons et al. 1999) all inhibit collagen production (see Rekhter 1999). In addition, SMC may exhibit different biosynthetic activity depending on the matrix environment, and the extracellular matrix itself is thus able to control collagen synthesis in SMC culture (Thie et al. 1991). Plaque rupture is responsible for the acute manifestations of atherosclerosis. Rupture prone plaques are characterized by a large lipid pool, a thin fibrous cap, an increased amount of inflammatory cells and a reduced collagen and SMC content (see Lee & Libby 33 1997, van der Wal & Becker 1999 for thorough reviews). Previous studies (Henney et al. 1991, Galis et al. 1994, Brown et al. 1995, Nikkari et al. 1995, Shah et al. 1995, Li et al. 1996) have implicated matrix metalloproteinases, mostly MMP-1, gelatinases and stromelysins, together with certain tissue inhibitors of metalloproteinases, in the pathogenesis, destabilization and rupture of atherosclerotic lesions. Recently, Sukhova et al. (1999) demonstrated increased levels of MMP-1 and MMP-13 in atheromatous versus fibrous carotid atherosclerotic plaques. A large portion of matrix-degrading enzymes is produced by inflammatory cells, especially macrophages. The numbers of mast cells, T lymphocytes and MMP-9-containing macrophages increase as the atherosclerotic lesions become more severe (Kaartinen et al. 1998). In addition to the MMPs, macrophages in human atheroma have been reported to contain abundant amounts of cathepsins K and S (Sukhova et al. 1998). Certain infections, such as Chlamydia pneumoniae (see Movahed 1999 for a review), have been suggested to play a role in the pathogenesis of atherosclerotic lesions, possibly by inducing the inflammatory process and thus the production of MMPs.

2.3.4. Other collagen-related events in atherosclerotic plaques

It has been suggested that oxidized low-density lipoprotein (LDL) is a key component in the endothelial injury in atherosclerotic lesions. The type I collagen content in the aortas of cholesterol-fed birds has been shown to be increased compared to controls (Jarrold et al. 1999). LDL may directly injure the endothelium or play a role in the adherence and migration of inflammatory cells (see Ross 1993). In addition, several collagen types are known to bind oxidized lipoproteins, and the accumulation of collagen in the intima may thus promote lipoprotein accumulation (Greilberger et al. 1997). Furthermore, oxidized LDL-containing immunocomplexes have been shown to play an important role in macrophage activation, which, in turn, may result in the secretion of cytokines capable of inducing collagen synthesis (Virella et al. 1995). The aminoterminal propeptide of type III procollagen, PIIINP, has been found to be elevated in the blood of patients with coronary artery disease (Bonnet et al. 1988). In addition, thrombolytic treatment of patients with acute myocardial infarction results in collagen breakdown (Peuhkurinen et al. 1991, Peuhkurinen et al. 1996). Furthermore, thrombin, a very important peptide in blood coagulation and platelet activation, stimulates procollagen production in cultured SMCs and thus in the arterial wall (Dabbagh et al. 1998). One of the most common features of advanced atherosclerosis is vascular calcification. Type I collagen has been shown to promote calcification of vascular cells in in vitro conditions (Watson et al. 1998). Type I collagen may also be important in plaque neoangiogenesis (Jackson & Jenkins 1991) and the organization of thrombus (Rekhter et al. 1996). Furthermore, human coronary restenosis involves rapid accumulation of collagen fibers (Pickering et al. 1996). Collagen accumulation has also been found to correlate with the severity of restenosis after angioplasty in an animal model (Lafont et al. 1999). Not only synthesis but also degradation of collagen has an important role in collagen turnover after balloon angioplasty (see Rekhter 1999). 34 2.3.5. Type I and III collagens in abdominal aortic aneurysms (AAA)

Atherosclerotic plaques are present in aneurysmal walls, and atherosclerosis is thought to be, at minimum, a permissive factor in aneurysm development (see Grange et al. 1997 for a review). There is evidence to suggest that the ratio of type III to type I collagen is not altered in the aneurysmal aorta (Rizzo et al. 1989). However, occasional lower amounts of type III collagen at the tissue level have been described in patients with a family history of aneurysms (Menashi et al. 1987, Powell & Greenhalg 1989). Kuga et al. (1998) suggested that the amount of type III collagen is significantly decreased in atherosclerotic AAAs compared with atherosclerotic aortas without aneurysmal changes. In addition, he found abnormal fragments of type III collagen in AAA tissue. Deak et al. (1992) and van Keulen et al. (1999) reported decreased type III procollagen secretion in fibroblasts in a small group of patients with familial AAA. The expression of type I and III collagens is increased in human abdominal aortic aneurysms (McGee et al. 1991, Mesh et al. 1992, Minion et al. 1993). The increased expression of collagen is located in adventitial fibroblasts, medial smooth muscle cells near the inflammatory infiltrates, and myofibroblasts inside the plaque (Hunter et al. 1996, see Grange et al. 1997). The existing data on the amount of collagen are controversial: there is evidence of increased (Menashi et al. 1987, Rizzo et al. 1989, Baxter et al. 1994, He & Roach 1994), unchanged (McGee et al. 1991) or even decreased (Sumner et al. 1970) proportion of total collagen in AAAs. Aneurysm formation is a complex remodeling process that involves both synthesis and degradation of matrix proteins (see Ghorpade & Baxter 1996 for a review). The identification of at least the MMPs 1, 2, 3, 9 and, recently, 13 in AAA tissue has been reported by several investigators (Irizarry et al. 1993, Newman et al. 1994, McMillan et al. 1995, Thompson et al. 1995, Sakalihasan et al. 1996, Tamarina et al. 1997, Mao et al. 1999). Freestone et al. (1995) reported more MMP-2 activity in small aneurysms, whereas MMP-9 was significantly increased in large aneurysms. On the other hand, MMP-9 expression has also been reported to be significantly higher in medium-sized AAAs than in either small or large aneurysms (McMillan et al. 1997). In addition, the activity of several cathepsins has been found to be increased in aortic aneurysms (Gacko & Chyczewski 1997). Clinical methods of MMP inhibition are under development. For example, preoperative treatment with doxycycline has been associated with reduction in the aortic wall expression of MMP-2 and MMP-9 (Thompson & Baxter 1999), and marimastat (an MMP inhibitor) has been shown to significantly inhibit matrix degradation and active MMP-2 production in a porcine model of aneurysm disease (Treharne et al. 1999). MMP-9, which is increased in the plasma of AAA patients, could be a potential diagnostic marker for clinical purposes (McMillan & Pearce 1999). Other diagnostic approaches have been studied as well: the aminoterminal propeptide of type III collagen, PIIINP, has previously been found to be clinically useful in the follow-up of patients with abdominal aortic aneurysms (Satta et al. 1997). Ehlers-Danlos syndrome IV (EDS IV), which is characterized by multiple aneurysms and other connective tissue manifestations, is caused by mutations in the gene encoding the α1-chain of type III collagen (see Kuivaniemi et al. 1997). In addition, there are many studies reporting a strong familial tendency to aneurysms not associated with EDS IV (see Verloes et al. 1996 for a review). In these studies, the proportion of AAA probands 35 with a positive family history has ranged from 12 to 20 %, and the relative risk for siblings has been 10 - 28. Van Keulen et al. (1999) recently reported that 29 % of 56 consecutive AAA patients had a positive family history of aneurysms. The characteristic features of familial aneurysms are an early age of diagnosis and a higher risk of rupture. Furthermore, even though both sporadic and familial AAAs are more common in men, the patients with familial AAA are more likely to be women (see Verloes et al. 1996). Type III collagen gene mutations have been suggested to contribute to the development of AAAs (Kontusaari et al. 1990, Anderson et al. 1996), and other genes, such as those for collagenases and elastin, have also been mapped for abdominal aneurysms (see Verloes et al. 1996). Some of the patients with familial aneurysms show decreased type III collagen production in cultured skin fibroblasts (Deak et al. 1992, van Keulen et al. 1999), or even lack type III collagen in skin and aortic wall (Hamano et al. 1998).

2.4. Type I and III collagens in cancer of the ovary and colon

Altered synthesis and increased breakdown of extracellular proteins have been shown to play very important roles in the invasion and growth of malignant ovarian tumors (Autio- Harmainen et al. 1993, Moser et al. 1994, Afzal et al. 1996, Wilson et al. 1996, Santala et al. 1998). Probably due to the increased collagenolytic activity, the type I and III collagen fibers are poorly organized and disintegrated in ovarian tumor tissue (Zhu et al. 1993a, Zhu et al. 1995). The synthesis of type I and III procollagens is normally very slow in ovarian tissue, and an increase of the production of these collagen types is related to the degree of malignancy (Zhu et al. 1993b, Kauppila et al. 1996). Some of the malignant cells even seem to have a capacity to produce type I collagen themselves (Zhu et al. 1995). Type I collagen is crucial for the invasion and metastasis of ovarian cancer, because it promotes the migration of carcinoma cells through an integrin-mediated mechanism and induces the activation of at least MMP-2 (Fishman et al. 1998, Boyd & Balkwill 1999). Insertion polymorphism in the promoter region of the MMP-1 gene resulting in greater transcriptional activity of the gene has been suggested to be associated with at least ovarian cancer (Kanamori et al. 1999). The increased synthesis and degradation of type I and III collagens in ovarian malignancy can be monitored by measuring the type I and III collagen metabolites in serum, cyst fluid or ascitic fluid (Santala et al. 1998). Measurement of the circulating aminoterminal propeptide of type III procollagen, PIIINP, has turned out valuable in the assessment of the prognosis and clinical course of the disease (Risteli et al. 1988b, Kauppila et al. 1989, Tomás et al. 1990). Elevated serum concentrations of the cross- linked carboxyterminal telopeptide of type I collagen, ICTP, reflecting increased degradation of type I collagen, are strongly related to impaired prognosis in patients with ovarian cancer (Santala et al. 1995, Santala et al. 1999). Altered synthesis and degradation of the extracellular matrix are also seen in human gastrointestinal malignancies, where the increased collagen synthesis takes place in stromal fibroblasts (Ohtani et al. 1992). Even though the expression of type I and III procollagens is increased in gastrointestinal carcinoma, the overall collagen content is decreased (Turnay et al. 1989). Retained secretion of collagen molecules together with 36 altered type I to type III procollagen ratios have been described in tumor-associated fibroblast-like cells, which are responsible for the production of collagen in colon adenocarcinoma (Turnay et al. 1989, Turnay et al. 1991). Collagenolytic activity is increased in colorectal malignant tissue and correlates with tumor differentiation (van der Stappen et al. 1990). In various gastrointestinal cancers, the gene expression of MMP-1, which degrades particularly type III collagen, is very abundant in stromal cells, especially at the margin of tumor invasion (Otani et al. 1994). On the other hand, the invasive areas of human colon cancer display signs of active fibrosis and tissue contraction, which has been suggested to be responsible for obstructing type colonic carcinomas (Miura et al. 1993). There is also evidence of a variant of type I and III collagens in both colon and breast carcinoma tissue. This onco-fetal/laminin-binding collagen is a heterotrimer of α1(III)- chain, α1(I)-chain and a unique acidic chain, being completely absent in normal tissue, which makes it a possible marker of malignancy (Pucci-Minafra et al. 1993). Furthermore, α2(I) chains have been suggested to be either overmodified or completely absent in colon cancer tissue (Pucci-Minafra et al. 1998). The occurrence of these matrix irregularities in malignant tissue may promote tumor invasion by affecting the supramolecular assembly of the collagenous matrix (Pucci-Minafra et al. 1998).

2.5. Type I and III collagens in colon diverticulosis

Diverticulosis of the colon is characterized by thickening of the muscle layers. The bowel lumen is encroached by thickened bars of circular muscle, which divide it into a series of small chambers (for reviews, see Smith 1986, Ming 1998). Surprisingly, the muscle cells in diverticular disease appear to be normal, with no evidence of hyperplasia or hypertrophy (Whiteway & Morson 1985). There is, however, evidence of a threefold increase in the elastin content of the taeniae coli when compared with the controls (Whiteway & Morson 1985). Diverticular disease is also associated with increased intraluminal pressure of the colon (Painter et al. 1965) and a deficiency of dietary fiber consumption (Painter et al. 1971). Genetic factors are also involved, since diverticulosis has been reported in Marfan’s syndrome and Ehlers-Danlos syndrome IV (Beighton et al. 1969, Suster et al. 1984), which involve mutations in fibrillin and type III collagen genes, respectively. Type I and III collagens are the most abundant collagen types in colon tissue, being interspersed in the muscle layer and thus providing the tissue with tensile strength. There is a link between diverticular disease, the ageing process, and the decreased tensile strength of the colonic wall. There is also evidence of different kinds of diverticular disease, one being mainly associated with muscle abnormalities, whereas another, a non- muscle form, is possibly caused by a failure of collagen support (see Smith 1986, Ming 1998). Collagen content has been suggested to decrease progressively or to be altered upon ageing either by rearrangement of the matrix or by increasing weakness of the individual fibers (see Smith 1986). Collagen fibrils in the left colon have been shown to become smaller and more tightly packed over increasing age (Thomson et al. 1987). Furthermore, the cross-linking of colonic collagen appears to increase with age and thus 37 in diverticulosis (Wess et al. 1995). A high-fiber diet has been found to be associated with a decrease in collagen cross-linking in an animal model (Wess et al. 1996a). Furthermore, the offspring of rats on a high-fiber diet presented no colonic diverticulosis, suggesting that the maternal diet plays an important role in the development of diverticular disease (Wess et al. 1996b). 3. Outlines of the present study

Fibrillar collagens are the major proteins of the extracellular matrix (ECM) and provide the tissue with tensile strength. Furthermore, collagens play an important role in pathological remodeling of tissues, such as invasion of malignant tumors. At the time the present work was initiated, a lot of information was already available on the metabolism of type I and III collagens in various diseases. This information had mostly been gained by measuring collagen metabolites in patient serum or other biological fluids, such as ascites, pleural effusion and synovial fluid. However, the situation at tissue level had usually only been studied with immunohistochemical techniques or hydroxyproline analyses, which were not able to shed light on the absolute amounts of the different collagen types or the state of maturation of their cross-links. The purpose of this study was to develop new immunological methods that would allow us to investigate the cross- linked type I and III collagens in tissues and, eventually, to apply these methods to study clinical conditions in which type III collagen, in particular, is known to play a role. During the course of the work, atherosclerosis, abdominal aortic aneurysms, colon and ovarian malignancies, and colon diverticulosis were chosen for further analysis. The present results not only represent the changes in collagen metabolism and processing in selected clinical situations, but also increase our understanding of the complex nature of collagen metabolism in pathological conditions in general.

The specific aims of this study were as follows:

1. to characterize the cross-linking of type I and III collagens and the processing of type III pN-collagen in human atherosclerotic plaques 2. to examine the possible role of fibrillar collagens, particularly type III collagen and pN-collagen, in the pathogenesis of human abdominal aneurysms 3. to study the cross-linking and processing of type I and III collagens in malignant processes, such as cancer of the ovary and colon 4. to study cross-linking of type I and III collagens in diverticulosis of the human colon 4. Materials and methods

4.1. Tissue samples

The human leiomyomas for the purification of type III collagen telopeptide were obtained from the Department of Obstetrics and Gynecology, Oulu University Hospital. Altogether 17 atherosclerotic plaques from the carotid (n = 15) and femoral (n = 2) arteries and 8 pieces of atherosclerotic aorta (I) were obtained from routine surgery at the Department of Surgery, Oulu University Hospital. One piece of healthy aorta, obtained from the Department of Pathology, Oulu University Hospital, was added to the study to validate the method. All the samples were stored at -20°C until processed. Nine pieces of abdominal aortic aneurysms (II) were collected during operations at the Department of Surgery, Oulu University Hospital. The control material for this study consisted of ten pieces of extremely atherosclerotic abdominal aorta from autopsies at the Department of Forensic Medicine, Oulu University Hospital. All the samples were stored at -20°C until processed. Altogether 24 pieces of colonic diverticulosis, six tissue samples of cancer of the colon and nine controls (III) were obtained from routine operations at the Department of Surgery, Oulu University Hospital. The histology of the samples was confirmed by a pathologist (TJK, see paper III) at the Department of Pathology, Oulu University Hospital. All the samples were stored at -20°C until processed. The ovarian cancer material (IV) obtained from the Department of Obstetrics and Gynecology, Oulu University Hospital, consisted of 18 samples of benign serous cystadenomas and malignant adenocarcinomas. The diagnoses were confirmed by routine histopathological analysis at the Department of Pathology, and the samples were stored at -20°C. Parts of the tissue were used for immunohistochemical analysis and therefore fixed in 10 % buffered formalin, embedded in paraffin and cut into 5 µm sections. 40 4.2. Purification of the aminoterminal telopeptide of type III collagen

The aminoterminal telopeptide of type III collagen, IIINTP, was originally purified from human uterine leiomyoma, which was first cut into small pieces, homogenized, and treated with 6 mol/L urea in 0.05 mol/L Tris/HCl buffer (pH 7.4) at +4°C for 24 hours. The residue was then washed with distilled water and reduced with NaBH4 in 0.1 mol/L sodium phosphate buffer, pH 7.4 (see Eyre 1987). The fat was extracted with acetone/methanol, and the residue washed briefly with distilled water and 0.2 mol/L ammonium bicarbonate (NH4HCO3), and freeze-dried. The insoluble pellet was suspended in 500 ml of 0.2 mol/L NH4HCO3, denatured at +70°C for 1 hour, and treated with 1 mg of trypsin (Worthington; TPCK-treated) per 100 mg of tissue at +37°C for 4 hours. After this, the residue was re-denatured, homogenized with an Ultra-Turrax homogenizer and digested with the same amount of trypsin overnight. The residual trypsin activity was destroyed at +70°C for 30 minutes, and the material was centrifuged at 10 000 g for 30 minutes. The supernatant was collected and freeze-dried. IIINTP was purified using Sephacryl S-100 (Pharmacia) gel filtration (equilibrated in 0.2 mol/L NH4HCO3 at room temperature), chromatography on a reverse-phase C8 column (Vydac) (0.4 % ammonium acetate + 0 → 60 % acetonitrile, pH 7.4), and finally, Sephacryl S-300 (Pharmacia) gel filtration (equilibrated in 0.2 mol/L NH4HCO3 at room temperature). Polyacrylamide gel electrophoresis was performed on 18 % gels with sodium dodecyl sulfate (SDS) to determine the purity and size of the purified peptide. The exact concentration was determined by amino acid analysis after acid hydrolysis. IIINTP was deblocked with pyroglutamate aminopeptidase (Sigma) (Podell & Abraham 1978, Hörlein et al 1979) and sequenced.

4.3. Production of antiserum

Antibodies against the aminoterminal telopeptide region of type III collagen were raised in New Zealand White rabbits by giving them purified or synthetic peptides conjugated with bovine thyroglobulin and mixed with equal amounts of 0.9 % NaCl and Freund’s incomplete adjuvant (Sigma) as intradermal injections at three- to four-week intervals. The conjugate was prepared by dissolving 2.5 mg of antigen and 10 mg of bovine thyroglobulin to 500 µl of 10 mM sodium phosphate buffer, to which 100 mg carbodi- imide (Sigma) diluted to 1 ml of sodium phosphate buffer was added. The mixture was allowed to stay on ice in a mixer for two hours and dialyzed overnight against PBS at pH 7.2. Finally, 100 to 200 µl of this conjugate-antigen mixture per one animal was used for immunization. 41 4.4. Radioimmunoassay for IIINTP

The purified IIINTP was labelled with 125I by the Chloramine-T method, and free iodine was separated from the iodinated antigen by disposable Sep pak C18 cartridges (Waters) (see Risteli & Risteli 1987). 100-µl aliquots of a properly diluted sample were incubated with 200 µl of the antiserum dilution and 200 µl of 125I-IIINTP solution for 2 h at 37°C. After that, 0.5 ml of the second antibody in 10 % polyethylene glycol (M.W. 6 000) was added, and the tubes were incubated at 4°C for 30 min. The samples were centrifuged at 2000 x g for 30 min at 4°C and the radioactivity in the precipitates was counted.

4.5. Other immunoassays for type I and III collagen domains

Radioimmunoassays for the aminoterminal propeptide of type I procollagen (PINP) (Melkko et al. 1996), the carboxyterminal propeptide of type I procollagen (PICP) (Melkko et al. 1990), the carboxyterminal telopeptide of type I collagen (ICTP) (Risteli et al. 1993), and the aminoterminal propeptide of type III collagen (PIIINP) (Risteli et al. 1988a) were used to determine their concentrations, when required. In addition, two in- house methods were used for analyses: an immunoassay for the synthetic peptide SP 6 having the amino acid sequence DVKSGVAVGGLAGY (Neosystem Laboratories) from the aminoterminal telopeptide of the α1-chain of type III collagen (amino acid residues nos. 159 – 162) and an immunoassay for the synthetic peptide SP 4 having the amino acid sequence SAGFDFSFLPQPPQEKY (Neosystem), derived from the carboxyterminal telopeptide of the α1-chain of type I collagen (amino acid residues nos. 1193 – 1208 + tyrosine residue). Samples in duplicate were diluted appropriately, and the analyses were performed as described above.

4.6. Immunohistochemical staining

The anti-aminoterminal propeptide of type I procollagen (anti-PINP), the anti- carboxyterminal telopeptide of type I collagen (anti-ICTP), and the anti-aminoterminal propeptide of type III procollagen (anti-PIIINP) were used for immunohistochemical analysis. These antibodies were purified by cross-adsorption with several extracellular matrix proteins (7S domain of human type IV collagen, P1 fragment of human laminin, different type I and III collagen domains) and with the specific antigen in question coupled to CNBr-activated Sepharose 4B. The immunohistochemical stainings for type I and III collagens were carried out using the avidin-biotin-immunoperoxidase technique (Zhu et al. 1993a, Zhu et al. 1995). Deparaffinized, rehydrated and PBS-washed sections were treated with 0.4 % pepsin (Sigma) in 0.01 mol/L HCl, pH 2, at 37°C for 30 minutes to 2 hours, washed with PBS and incubated in 0.3 % H2O2 in 100 % methanol for 30 minutes. Afterwards, the sections were washed with distilled water and incubated with 2 % goat serum for 30 minutes and, finally, with primary antibody diluted 1:25 (anti-ICTP) and 1:100 (anti-PINP and anti- 42 PIIINP) in PBS in a moist chamber for 2 hours at room temperature. Then the specimens were incubated for 30 minutes with biotinylated anti-rabbit immunoglobulin diluted in 1:400 in PBS and for 30 minutes with an avidin-peroxidase complex, washed with PBS and treated with 3,3-diaminobenzidine tetrahydrochloride and H2O2 in Tris buffer, pH 7.4. The samples were then washed with distilled water and stained with hematoxylin.

4.7. Chromatographic analyses of type I and III collagen antigens in tissue samples

Sephacryl S-300 gel filtration column was used to analyze the soluble tissue extracts. 2 ml of sample (100 mg/ml) was put to a column equilibrated in PBS-Tween at room temperature, and fractions of 2 ml were collected at 20-minute intervals. These fractions were used for PIIINP, PINP and PICP analyses after suitable dilutions. The molecular sizes of telopeptides of insoluble cross-linked collagens, together with the aminoterminal propeptide antigens of type III pN-collagen, were characterized on a Sephacryl S-100 gel filtration column equilibrated in 0.2 mol/L NH4HCO3 at room temperature. 1 ml of sample (10 mg/ml) was analyzed, and fractions of 2 ml were collected at 20-minute interval. ICTP, IIINTP, PIIINP, SP4 and SP6 were assayed in properly diluted fractions (see chapters 4.4. and 4.5.).

4.8. Hydroxyproline measurements

The content of hydroxyproline was measured as described previously (Kivirikko et al. 1967). The samples were hydrolyzed overnight in 6 mol/L HCl at 120°C. The hydrolytes were then evaporated to dryness and diluted in 2 ml of distilled water. pH was adjusted to neutral by using 0.005 mol/L KOH. 1.5 g of KCl was used to saturate the samples, and 250 µl of 10 % L-alanine and 500 µl of potassiumborate buffer (1 mol/L H3BO3, 3 mol/L KCl, pH 8.7) were added. The samples were stirred and incubated at room temperature for 20-30 minutes. Then, 500 µl of 0.2 mol/L chloramin-T (Riedel-deHaën) (2.82 g/ 50 ml 1,3-propandiol) was added and the solution was allowed to incubate for 25 minutes. The reaction was stopped by adding 1.5 ml of 3.6 mol/L Na2S2O3. After this, 2.5 ml of toluene was added, the samples were stirred and centrifuged at 2000 x g, and the upper phase was removed. The samples were then incubated in boiling water for 30 minutes, after which toluene was added and the samples were re-centrifuged as described above. After centrifugation, 2 ml of the upper phase was pipetted and mixed with 0.8 ml of Ehrlich´s reagent (Sigma) (120 g of p-dimethylaminobentsaldehyde, 400 ml of ethanol and 27.4 ml of H2SO4), and incubated for 30 minutes. Absorbances were measured at 560 nm, and the amount of hydroxyproline was estimated from the standard curve (standards 1 µg, 2.5 µg, 5 µg and 10 µg of L-hydroxyproline (S.A.F. Hoffmann – La Roche & Co. LTD.) in 2 ml of distilled water (standards prepared as the samples except for hydrolysis). 43 4.9. Investigation of the proteolytic fragments of PIIINP

PIIINP was purified from human ascitic fluid as described earlier (Niemelä et al. 1985, Risteli et al. 1988a) and digested with trypsin at 37°C after heat denaturation. A smaller trypsin-generated fragment of PIIINP, probably corresponding to the most aminoterminal part of the monomeric PIIINP, was purified from the human uterine leiomyoma with Sephacryl S-100 gel filtration, chromatographies on a reverse-phase C8 column (0.4 % ammonium acetate + 0 → 75 % acetonitrile, pH 7.4), DEAE (50 mmol/L ammonium acetate + 0 → 0.3 mol/L NaCl, pH 5), and finally, a reverse-phase C18 column (0.1 % TFA + 0 → 70 % 2-propanol). The effect of trypsin treatment on the antigenic activity of PIIINP was studied further in radioimmunoinhibition assay. The molecular sizes of these trypsin-generated peptides were characterized by polyacrylamide gel electrophoresis and Sephacryl S-100 gel filtration.

4.10. Statistics

Statistical analyses were carried out using the SPSS analysis tool and CIA confidence interval analyses. Spearman’s rank correlation was used for correlation analysis. Means and 95 % confidence intervals were calculated. Non-parametrical Mann-Whitney U test and one-way ANOVA with corrected LSD (Bonferroni) test were used for estimating statistical significances when appropriate.

4.11. Ethical considerations

Approvals of the local ethics committee for the collection of tissue samples were obtained by our clinical collaborators for the studies I, II and IV and by our own research group for study III. The tissue samples were obtained during routine surgical operations, and nothing else was removed from the patients. In the case of cadavers, only the pieces of tissue removed for diagnostic purposes were used. 5. Results

5.1. Purification of the human cross-linked aminoterminal telopeptide of type III collagen (IIINTP)

The cross-linked aminoterminal telopeptide of type III collagen, IIINTP, was purified from human uterine leiomyoma after extraction with urea, reduction with NaBH4, and digestion with TPCK-treated trypsin, followed by a series of Sephacryl S-100 and S-300 gel permeations and C8 reverse-phase chromatography on HPLC. The purity and size of the peptide were determined by SDS-polyacrylamide gel electrophoresis. Aminoterminal amino acid sequencing yielded a sequence equivalent to the aminoterminal telopeptide of human type III collagen (QYDSYDV). Without deblocking with pyroglutamic aminopeptidase, the obtained amino acid sequence GAAGI corresponded to the carboxyterminal triple-helical area of α1(III) containing a lysine residue involved in intermolecular cross-linking. Cross-link analysis of the IIINTP antigen indicated that at least 80 % of the cross-links were hydroxylysylpyridinoline (unpublished data, courtesy of Jason Mansell, Bristol University). With the purified IIINTP antigen, a radioimmunoassay for the cross-linked aminoterminal telopeptide of type III collagen was established.

5.2. New radioimmunoassays for type III collagen

A radioimmunoassay for IIINTP from human uterine leiomyoma was developed based on polyclonal antibodies and iodine-labeled IIINTP antigen. A 1:400 dilution of the antibodies was chosen, giving 53 % binding. In paper IV, another antiserum with a different dilution (1:200) but similar immunological properties was used. Serum samples did not give inhibition in the assay, indicating that the physiological degradation product of type III collagen likely to exist in the circulation differs from the trypsin-generated telopeptide antigen. The intra- and interassay coefficients of variation were tested by using 100 µl trypsin-digested human colon tissue diluted 1:1000. The intra- and interassay coefficients of variance were 6 % and 10 %, respectively. 45 An immunoassay for the synthetic peptide SP 6, having the sequence DVKSGVAVGGLAGY from the aminoterminal telopeptide of the α1-chain of type III collagen, was also established in the same way as the IIINTP assay. A 1:2000 dilution of the antibodies was chosen, giving 45 % binding in the assay. In this assay, the serum samples gave linear inhibition, and 1:2 dilution of serum (about 50 % inhibition) was used to determine the intra- and interassay variations, which were 4.3 and 11.5 %, respectively.

5.3 Type I and III collagens in atherosclerotic plaques (I)

Trypsin digestion was found to release effectively the cross-linked type I (ICTP) and type III (IIINTP) collagen telopeptide antigens from atherosclerotic plaques into analysis (see Figure 1 in paper I). The ICTP and IIINTP antigens were partially purified from the trypsin-digested plaque material, and both these cross-linked telopeptides were found to be similar to those of the corresponding standard antigens, indicating mature cross- linking of these collagen types (see Figure 2 in paper I). In addition, the samples gave a linear inhibition in the ICTP and IIINTP assays. Furthermore, the IIINTP and ICTP antigens in the atherosclerotic plaques corresponded in size to the trivalent, fully cross- linked peptide standards, when studied by gel filtration (see Figure 5 in paper I). The amount of type III collagen (calculated as a percentage of the sum of types I and III, respectively), was 61 % (CI 58 – 65) in human atherosclerotic plaques (n = 17) and 56 % (CI 44 – 68) in atherosclerotic aortic specimens (n = 8). The one healthy young aorta contained 72 % of type III collagen. For comparison, in human leiomyoma and placenta, the proportions of type III collagen calculated in the same way were 35 % and 33%, respectively. In the calcified fraction of the atherosclerotic plaques, the mean proportion of type III collagen was 54 %. The amount of type III pN-collagen was calculated from the molar amounts of PIIINP and IIINTP. To validate the results obtained, the effect of trypsin digestion on the antigenicity and size of the PIIINP antigen was studied. After trypsin digestion, a truncated form of the monomeric propeptide chain appeared, indicating partial cleavage of the propeptide leading to a 50 % decrease in the immunoreactivity of this peptide (see Figures 3 and 4 in paper I). Thus, the trypsin loss of PIIINP was corrected by multiplying the obtained PIIINP concentrations by two. The amount of type III pN-collagen in the insoluble matrix of the plaques was 0.0081 % (CI 0.0042-0.0121) (see Table 2 in paper I). In the soluble tissue extracts, type III pN-collagen accounted for 42 % of the total type III collagen content. The mean concentrations of type I procollagen propeptides were very low (PINP 0.13 µg/g and PICP 1.18 µg/g) in the soluble tissue extracts, indicating a low rate of type I procollagen synthesis. The mean PIIINP concentration was 0.31 µg/g. Gel filtration analysis of the soluble tissue extracts revealed that the type III collagen was in the form of pN-collagen, whereas PINP was always present as a free propeptide. For PICP, in turn, nearly equal amounts of authentically cleaved propeptide and type I pC- collagen were seen (see Figure 6 in paper I). 46 5.4. Type I and type III collagens in human abdominal aortic aneurysms (II)

The average amounts of type I and type III collagens expressed per total hydroxyproline were 5.3 and 3.4 times higher in the abdominal aortic aneurysms (AAA) than in the atherosclerotic aortic samples (AOD), respectively (see upper panel of Figure 2 in paper II). However, the mean hydroxyproline concentrations were practically the same in both AAAs (7.8 ± 1.4 mg/g) and AODs (7.8 ± 1.2 mg/g). The relative amount of type III collagen was 44 % (CI 38-41) in AAAs (n = 9) and 60% (CI 53-67) in AODs (n = 10) (p < 0.001), indicating a significantly smaller relative amount of cross-linked type III collagen in aneurysm tissue (see lower panel of Figure 2 in paper II). The concentrations of ICTP and IIINTP antigens correlated significantly in both AAAs (r = 0.867; p < 0.002) and AODs (r = 0.648; p < 0.043). In both groups, most of the IIINTP antigen eluted as one peak corresponding to the elution position of the trivalently cross-linked aminoterminal telopeptide antigen of type III collagen (see Figure 3 in paper II). The mean absolute amounts of PIIINP in the insoluble fractions of AAA and AOD samples were 14.9 µg/g and 0.27 µg/g, respectively (see Figure 7A in paper II). The average relative amount of type III pN-collagen in the insoluble collagenous matrix was over 12 times higher in AAAs (0.74 %) than in AODs (0.06%) (see Table 1 in paper II). This difference was also evident in the gel filtration analysis, where the PIIINP antigenicity eluted mostly in one major peak in AAAs, whereas no propeptide derived antigens were seen in AODs (see Figure 3 in paper II). The major peak of PIIINP antigenicity in insoluble AAA tissue shown in Figure 3 in paper II corresponded in size to the partially cleaved PIIINP molecule (see chapter 5.3. and Figure 3 in paper I). Thus, we were able to use the same correction factor 2 as in paper I. However, some PIIINP antigenicity, clearly smaller in size, was found in the gel-filtrated samples (see Figure 3 in paper II). When purified from human uterine leiomyoma, this smaller peptide was found to be a single monomeric chain corresponding to the Col 1 part of the aminoterminal propeptide (see Figure 5 in paper II). This smaller form of digested PIIINP reacted about 30 times less intensely in the radioimmunoinhibition assay for the intact aminoterminal propeptide of type III collagen than did the larger trypsin-generated PIIINP (see Figure 6 in paper II). The mean absolute amounts of PIIINP in the soluble tissue extracts of AAA and AOD were 0.68 µg/g and 1.15 µg/g (see Figure 7B in paper II). As much as 78 % of PIIINP was found to be soluble in AODs, whereas the corresponding percentage in AAAs was only 6 %. In the soluble tissue extracts, the relative amount of type III pN-collagen accounted for 69 % and 41 % in AAAs and AODs, respectively (see Table I in paper II). The solubilized PIIINP antigens were studied on Sephacryl S-300 gel exclusion chromatography; most of the PIIINP antigenicity in AAAs eluted as type III pN-collagen, whereas AODs also appeared to contain free propeptide (see Figure 4 in paper II). Immunohistochemical studies (M.K. Bode, Y. Soini, J. Melkko, J. Satta, L. Risteli, J. Risteli, unpublished data) indicated abundant PIIINP staining in the media of abdominal aortic aneurysms, whereas type I pN-collagen localized in the intima in both AAAs and AODs (Fig. 5). The healthy aorta, in turn, showed no immunoreactivity for either PIIINP or PINP (not shown). 47

Fig. 5. Immunohistochemical type I (A) and type III (B) pN-collagen staining in abdominal aortic aneurysms. Type I (C) and type III (D) pN-collagen staining in atherosclerotic abdominal aorta. Hematoxylin counterstain, original magnification, x 10, i = intima, m = media, a = adventitia. 48 5.5. Type I and III collagens in colon diverticulosis and colon cancer (III)

The total collagen content was measured in trypsin-digested colon tissue by hydroxyproline analysis, which showed about 50 % less collagen in malignant tissue than controls (see Figure 1A in paper III). The amount of fibrillar type III collagen measured by IIINTP presented a similar pattern: the cancer samples showed less cross-linked type III collagen than did the controls (0.14 mg/g and 0.36 mg/g, respectively). For type I collagen, no such significant difference was found (see Table I in paper III). The proportion of type III collagen (as a percentage of the sum of type I and III collagens) was significantly decreased in malignant samples (see Table I in paper III). The correlations were 0.884 between hydroxyproline and ICTP, 0.940 between hydroxyproline and IIINTP, and 0.877 between ICTP and IIINTP. The amount of fibrillar type III pN-collagen was also assessed (calculated from the molar amounts of PIIINP and IIINTP). The mean amounts of type III pN-collagen were 0.48 % (range 0.00-2.97), 1.71 % (range 0.15-5.96) and 0.40 % (range 0.05-1.38) in diverticulosis, cancer samples, and controls, respectively. However, due to the obvious large variation, there were no statistically significant differences between the groups analyzed. The trypsin-digested type I and III collagen antigens were characterized by gel filtration analysis on Sephacryl S-100, to further analyze their molecular sizes and the state of cross-linking. In order to do that, immunoassays for ICTP, IIINTP, SP4 and SP6 were applied. In the samples of malignancy, the ICTP antigen eluted in one peak, corresponding to the trivalently cross-linked mature type I collagen. SP4, however, eluted in two peaks, the latter representing immaturely cross-linked type I collagen. In the diverticulosis and control samples, no such tendency was evident (see Figure 2 in paper III). In all the samples, type III collagen eluted in one peak, representing the maturely cross-linked type III collagen, IIINTP (see Figure 2 in paper III). The aminoterminal propeptide domains of type I and III procollagens, PINP and PIIINP, respectively, were quantified in the soluble fraction, where the molecules are in transit from biosynthesis to collagen fibers. The mean PINP concentrations were significantly higher in the colon cancer samples than in the controls (1.57 µg/g and 0.24 µg/g, respectively). The mean PIIINP values, in turn, were significantly higher in diverticulosis than in the controls (1.73 µg/g and 0.83 µg/g, respectively). The proportion of type III collagen (as a percentage of the sum of type I and III collagen) was significantly decreased in the tumor samples (see Table II in paper III). The sizes of the soluble PINP and PIIINP antigens were assessed by gel filtration analysis on Sephacryl S-300. The major PINP and PIIINP antigenicities in the tumor samples eluted as free propeptides related to the ongoing synthesis. In diverticulosis, PINP eluted as a free propeptide, whereas the profile for PIIINP showed practically identical amounts of type III pN-collagen and free aminoterminal propeptide. In the control samples, most of the PINP and PIIINP antigenicities eluted in their pN-form (see Figure 3 in paper III). 49 5.6. Type I and III collagens in ovarian tumors (IV)

The tissue contents of both ICTP and IIINTP were lower in the malignant than in the benign tissues. For type I collagen, this tendency was even more obvious: the amount of ICTP in the benign cystadenomas was about 8 times higher than that in adenocarcinomas (mean 0.57 mg/g and 0.07 mg/g, respectively, see Table 2 in paper IV). Furthermore, the amount of total collagen measured by hydroxyproline analysis showed a nearly five-fold decrease in the malignant samples compared to the benign ones (see Table 2 in paper IV). There was a strong correlation between hydroxyproline and IIINTP (0.96), hydroxyproline and ICTP (0.98), and IIINTP and ICTP (0.96). The amount of type III collagen (calculated as a percentage of the sum of type I and III collagens) was 35 % (CI 28-42 %) in the benign ovarian tumor tissue. In the insoluble tissue, the PIIINP/IIINTP ratio was used to estimate the amount of type III pN-collagen, which was clearly elevated in the malignant samples (3.20 % in adenocarcinomas and 0.70 % in cystadenomas, see Table 3 in paper IV). There was no statistically significant difference, however, in the total amount of PIIINP in the trypsin-digested ovarian tumors between the two groups (see Table 3 in paper IV). In the soluble extracts, the mean PINP concentration was higher in the carcinoma samples than in the benign tumors (2.45 µg/g and 1.00 µg/g, respectively). No such tendency was evident for PIIINP (2.04 µg/g for malignant and 1.44 µg/g for benign samples). The sizes of the type I and III collagen telopeptide fragments were analyzed on gel exclusion chromatography. In the benign samples, most of the ICTP and IIINTP antigens eluted in one major peak, corresponding to the trivalently cross-linked telopeptide antigens (see Figures 1A-C and 2A in paper IV). In the malignant samples, however, there were also smaller forms present, indicating immaturely cross-linked type I and III collagens (see Figures 1D-F and 2B in paper IV). The soluble PINP and PIINP antigens were also characterized by gel filtration. In both the benign and the malignant samples, the major PINP antigenicity eluted as a free propeptide, whereas PIIINP was present in its pN-form (see Figure 3 in paper IV). In the benign samples, PINP staining showed type I pN-collagen mainly at the epithelial-stromal junction, whereas in poorly differentiated tumors PINP was seen near the malignant cells. Mature cross-linked type I collagen (ICTP staining) was weaker in the malignant samples compared to the benign ones. Type III pN-collagen staining resembled that seen for type I pN-collagen. Immunostainings also revealed disorganized type I and type III collagen fibers in malignant tumors (see Figure 4 in paper III). 50 Table 3. Summary of the quantitative and qualitative findings of type I and III collagens in the human tissues studied.

Type III/ Type III pN/ Immaturely cross-linked (Type III +I) % Total Type III % forms of ______type I type III

Atherosclerotic plaques (n = 17) 61 0.01 -/+ -/+

Aorta AOD (n = 10) 60 0.06 -/+ -/+ AAA (n = 9) 44 0.74 -/+ -/+

Ovarian tumor Benign (n = 8) 35 0.70 -/+ -/+ Malignant (n = 10) 42 3.20 +++ ++

Colon Cancer (n = 6) 53 1.71 +++ -/+ Diverticulosis (n = 24) 62 0.48 -/+ -/+ Control (n = 9) 65 0.40 -/+ -/+

Placenta (n = 1) 33 8.50 ND ND Leiomyoma (n = 1) 35 1.41 ++ ++

ND = not determined -/+ = no or minor amounts ; ++ = moderate, and +++ = large amounts of immature forms, determined by gel exclusion chromatography by antigen size (qualitative analysis) 6. Discussion

6.1. Insoluble vessel wall collagens – methodological and structural aspects

One of the most striking events of atherosclerosis is the accumulation of fibrillar collagens, which is why the quantities of especially type I and III collagens in atherosclerotic lesions have been extensively investigated (see Barnes 1985, Rauterberg et al. 1993 for reviews). However, the evidence has remained controversial. The initial work by McCullagh & Balian (1975) suggested type III collagen to predominate in normal arterial wall. Since then, research results have indicated type I collagen to be the major collagen type in both normal and atherosclerotic arteries (see Barnes 1985). Our study (paper I), in turn, suggested that type III collagen predominates in atherosclerotic lesions. The amounts of collagen types found in arterial wall may depend on the method of extraction, and type III collagen seems to be more easily extractable than type I collagen (Szymanowicz et al. 1982). Pepsin, for example, cleaves collagen molecules from the telopeptide region, which contains the sites for intermolecular cross-linking. Thus, pepsin digestion may result in different degrees of solubilization due to the different extents of intermolecular cross-linking (see Rauterberg et al. 1993). Most previous studies on the quantitative ratios of type I and III collagens have been made by using cyanogen bromide (CNBr), which cleaves after the methionine residue (see Rauterberg et al. 1993). This method has several advantages compared to pepsin digestion, one of the most important being the high yield of soluble collagenous peptides. However, the cleavage of the collagen chains may be incomplete due to methionine oxidation. In addition, the CNBr method is very often based on simple staining of the cleaved peptide fragments on SDS- PAGE, which is not ideal for collagenous sequences. Recently, improved methods based on the quantitation of collagens by capillary electrophoresis after CNBr digestion have been developed (Deyl et al. 1997). Trypsin is known to cleave after the amino acids lysine and arginine, liberating the cross-linked telopeptides into analysis. Furthermore, previous studies have shown that trypsin solubilizes 99 % of cross-linked type I collagen (Kuboki et al. 1981). Due to these reasons, in addition to the easy availability and low price of the enzyme, trypsin was selected for these studies. 52 The presence of differently organized type I and III collagen fibrils could theoretically influence the solubility of collagens and the interpretation of the data. Two populations of fibrils, the smaller ones about 28 nm and the larger ones about 42 nm in diameter, have been described in the aortic wall by Shiraishi et al. (1980). In contrast, Raspanti et al. (1989) detected only one kind of fibril population in the aortic media, but found evidence of two morphologically different fibril types in other tissues, one arranged into unidirectionally packed fibrils and the other with a helical conformation. Raspanti et al. (1989) further suggested that unidirectional packing would be characteristic of skeletal tissues, whereas helical fibrils might be present in, for example, skin and vessel walls. There is evidence to suggest that helically arranged fibrils are present in tissues rich in type III collagen, but there are also exceptions to this rule (see Rauterberg et al. 1993). Henkel (1996) previously investigated the supramolecular assembly and cross-linking of the carboxyterminal telopeptide of type III collagen. He suggested that several dimers containing two quarter-staggered molecules with hydroxylysine or lysine-derived cross- links can be thought as being joined by disulfide bonds, resulting in a tetrameric structure stabilized by lysine derived cross-links connecting the N-telopeptides of different type III collagen molecules (Henkel 1996). The different ways of organization and cross-linking of fibrillar collagens are likely to influence the mechanical properties of tissues and thus their functions. They may also have an effect on the results and their interpretations in studies aiming to quantify collagens in tissue digests. The total collagen contents of tissue samples have been traditionally measured by hydroxyproline analysis (Kivirikko et al. 1967). Many different types of collagens have been described in vessel walls (see Barnes 1985, Rauterberg et al. 1993), and since the amount of one collagen type may increase at the same time as that of another decreases, hydroxyproline content cannot give any easily interpretable information on specific collagen types. In paper II, the total amount of hydroxyproline was not significantly different between AAAs and AODs, similarly to the finding published by Minion et al. (1994). However, type I and III collagen specific analyses showed the amounts of the two fibrillar collagen types to be larger in AAAs than in undilated atherosclerotic aortas. Thickening of the intima due to atherosclerosis may increase the amounts of at least type V, VI (Katsuda et al. 1992) and VIII collagens (Plenz et al. 1999), which may explain the lower amounts of type I and III collagens per hydroxyproline in AODs. Furthermore, hydroxyproline is not collagen specific, since several other proteins, such as elastin, the C1q component of complement, scavenger and some other cell surface receptors, and acetylcholinesterase, contain hydroxyproline (see Kielty et al. 1993).

6.2. Type I and III collagens – soil for atherosclerosis

Fibrillar type I and III collagens constitute a major part of the total plaque protein (see Barnes 1985 for a review) and occupy about 80 % of the section area in samples of human coronary restenotic lesions (Pickering et al. 1996). Thus, these collagen types play an important role in plaque growth and arterial lumen narrowing. In addition, collagen stimulates atherogenesis by serving as a depot for proatherogenic peptides and molecules, such as lipoproteins, growth factors and cytokines, and modulates the responses of 53 inflammatory cells and SMC proliferation (see Rekhter 1999). Furthermore, injury to the endothelium leads to activation of the coagulation system, since the thrombogenic material of the plaque - lipids and collagens in particular - is exposed, resulting in thrombus formation. A typical feature of advanced atherosclerosis is calcification, which shares many histological features with bone and involves many bone-related proteins, such as osteopontin, osteonectin and osteocalcin (Hirota et al. 1993, Bini et al. 1999). These bone matrix proteins, apolipoproteins, fibrinogen, and proteases, are suggested to interact in the formation of calcification and thus the progression of the atherosclerotic lesions (Bini et al. 1999). Furthermore, intraplaque thrombus is suggested to contribute to arterial calcification as a source of osteocalcin (Bini et al. 1999). Osteonectin promotes binding of calcium to collagen (Termine et al. 1981), and vascular calcification is, in turn, promoted by type I collagen and fibronectin in in vitro conditions (Watson et al. 1998). In the current quantitative analyses of type I and III collagens in atherosclerotic plaques, no marked differences were seen between the non-calcified and calcified insoluble type I and III collagens. Thus, despite the reported similarities in bone mineralization and vascular calcification, the presence of type III collagen in mineralized aortic tissue does not suggest the calcification process to be identical with that in bone. The present study demonstrated that the amount of partially processed type III procollagen molecules with a retained aminoterminal propeptide (type III pN-collagen) is negligible in atherosclerotic plaques. In addition, the synthetic activity of these collagen types was found to be very low, as suggested by the small concentrations of free propeptides in the soluble fraction, where collagen molecules are in transit from biosynthesis to collagen fibers. Thus, the present data supports the idea of metabolically inert, silent plaques (Kockx et al. 1998). Type I and III collagens appear to be fully cross- linked in atherosclerotic plaques. This is consistent with the previously reported increased activity of lysyl oxidase, an enzyme responsible for the cross-linking of collagen in the rabbit aorta with diet-induced atherosclerosis (Kagan et al. 1981). Thus, mature cross- linking reflects the fibrotic process seen in atherosclerosis. Several previous studies have implicated matrix metalloproteinases (MMPs), which are enzymes responsible for collagen degradation, in the destabilization and rupture of atherosclerotic lesions (Henney et al. 1991, Galis et al. 1994, Brown et al. 1995, Nikkari et al. 1995, Shah et al. 1995, Li et al. 1996, Sukhova et al. 1999). It can be suggested that the synthesis and degradation of collagens in the plaque are relatively balanced processes, but over time the degradation of collagens may overcome the synthesis, resulting in plaque rupture. On the other hand, if collagen synthesis decreases due, for example, to the reported inhibitory effect of monocyte prostaglandins (Fitzsimmons et al. 1999) and T- lymphocyte-derived IFN-γ (Amento et al. 1991) on collagen production, or apoptosis of SMCs (Bauriedel et al. 1999), an acute lack of plaque-stabilizing collagenous matrix might develop and result in complications (see Lee & Libby 1997, Rekhter 1999). This is consistent with the finding that macrophages, T-lymphocytes and activated monocytes are particularly abundant at the site of plaque rupture, whereas the number of SMCs is decreased (van der Wal et al. 1994, Kovanen et al. 1995). Many reports have suggested fibrous plaques to be clinically more stable than those containing large pools of lipids (see van der Wal & Becker 1999). Feeley et al. (1991) reported that even though asymptomatic carotid plaques contain more fibrous material (88 %) than symptomatic 54 plaques (66 %), no relationship between plaque composition and the number of ischemic events was evident. The development of atherosclerosis and restenosis is a very complex process involving many different but interrelated factors. Therefore, further studies on the regulation of collagen synthesis, processing and degradation are necessary to reveal the underlying mechanisms.

6.3. Abdominal aortic aneurysms and atherosclerosis

Because atherosclerosis is an almost universal finding in the walls of aneurysms, AAAs have been thought to be caused by atherosclerosis. In addition, the risk factors for these two clinical conditions – high blood pressure, high serum cholesterol, and cigarette smoking – are the same (Reed et al. 1992). Proliferation of SMCs is a typical feature in atherosclerosis (see Ross 1993), whereas the density of these cells in aneurysmal wall is low due to apoptosis (López-Candales et al. 1997, Henderson et al. 1999). SMCs together with adventitial fibroblasts are the source of collagens in the aortic wall. Atherosclerosis is characterized by accumulation of the extracellular matrix. Most of the existing data on AAAs suggest increased procollagen expression (McGee et al. 1991, Mesh et al. 1992, Minion et al. 1993). The proportion of collagen has been reported to be increased (Menashi et al. 1987, Rizzo et al. 1989, Baxter et al. 1994, He & Roach 1994), unchanged (McGee et al. 1991) or even decreased (Sumner et al. 1970) in AAAs compared to undilated aortas. In our study, no such comparison was possible due to the different extents of calcification between the AOD and AAA samples. However, since the amounts of aminoterminal propeptide of type III procollagen, PIIINP, in the soluble tissue extracts of AAAs and AODs did not differ, the synthesis of type III procollagen does not seem to be markedly increased in dilated aorta. Type III collagen molecule with a retained aminoterminal propeptide is often found on the surface of collagen fibrils. When the fibril has reached its final diameter, the aminoterminal propeptide is retained, thus preventing the attachment of other collagen molecules to the fibril surface (Fleischmajer et al. 1985). Satta et al. (1995) found higher serum PIIINP concentrations below than above the aneurysmal region, which also provided a rationale for us to investigate type III collagen at tissue level. Indeed, the current study (paper II) indicated that the amount of type III pN-collagen is increased in abdominal aortic aneurysms, thus suggesting the presence of thinner than usual type III collagen fibers. This might result in increased susceptibility of collagen to proteases, which are clearly increased in abdominal aortic aneurysms. There are studies indicating the presence of MMP-1, the main substrate of which is native type III collagen, in AAA tissue (Irizarry et al. 1991, Newman et al. 1994). Most studies have concentrated on gelatinases (MMP-2 and –9), the activities of which have usually been reported to be higher in AAAs compared to AODs, with a few exceptions, where no differences between these two clinical conditions were seen (see Grange et al. 1997 for a review). More recent findings have suggested that MMP-9 is elevated in both AAA and AOD tissue, whereas the MMP-2 mRNA and protein levels are only significantly higher in 55 AAA tissue samples (Davis et al. 1998). At the same time, however, Elmore et al. (1998) reported completely opposite results, suggesting that MMP-9, but not MMP-2, is an important factor in the etiology of AAAs. The role of these MMPs in aneurysm pathogenesis has been mainly attributed to their elastolytic properties. However, native type I and denatured collagen have also been reported as substrates for gelatinases (see Kähäri & Saarialho-Kere 1997). Fragmentation and decreased concentrations of elastin have been described as one the most striking histologic features in aneurysm tissue (Rizzo et al. 1989, Baxter et al. 1992). Changes in elastin content and architecture are not, however, exclusive to AAAs, but have also been described in AOD tissue (Campa et al. 1987). Degradation of elastin leads to aortic dilatation, whereas degradation of collagen results in both dilatation and, above all, rupture (Dobrin & Mrkvicka 1994). Thus, an increased amount of type III pN- collagen might be responsible for at least part of the clinical manifestations of aneurysms, making the collagen molecules more susceptible to degradation. Infiltration of inflammatory cells is a characteristic feature of aortic aneurysms and atherosclerosis. Since some of the matrix-regulating peptides together with proteolytic enzymes are produced by inflammatory cells, this has an important role in both collagen and elastin metabolism (see Ghorpade & Baxter 1996, Grange et al. 1997). Furthermore, both atherosclerosis (see Movahed 1999) and abdominal aortic aneurysms (Juvonen et al. 1997) have also been previously suggested to be associated with Chlamydia pneumoniae infection. Our recent immunohistochemical studies have shown abundant type III pN-collagen staining in the media of aneurysms (see Fig. 5 in Results-section). Type I pN-collagen staining, in contrast, was seen in the intima of both AAA and AOD specimens. Thus, it seems that the atherosclerotic injury in the intima causes fibroproliferative reactions characterized by enhanced synthesis of type I procollagen, whereas the pathogenesis of abdominal aortic aneurysms includes changes in type III collagen metabolism in the media layer. Recent data suggests that the MMPs 1, 2 and 9 are differently expressed in the aortic wall of AAAs and AODs (Palombo et al. 1999). Therefore, type III collagen in the media of aneurysms might be under a more severe proteolytic attack than the wall of a corresponding atherosclerotic aorta without dilatation. In our biochemical studies (paper II), the relative amount of type III collagen was significantly lower in aneurysms than in AODs and atherosclerotic plaques. These observations together with the previously reported observations on decreased type III collagen in AAAs (Kuga et al. 1998, van Keulen et al. 1999) and cerebral aneurysms (Pope et al. 1981, van den Berg et al. 1997) suggest a very important role of type III collagen in aneurysm formation. In addition, Pauschinger et al. (1999) reported a decreased relative amount of type III collagen in dilated cardiomyopathy, a disorder similar to AAAs. The disorder in the processing of type III collagen might predispose aneurysmal tissue to atherosclerosis, because the PIIINP molecule contains a transglutamination site available to LDL binding, for example (Bowness et al. 1989). On the other hand, this was not the case in atherosclerotic plaques in paper I, since the amount of type III pN-collagen was negligible. A clear familial tendency of AAAs has been demonstrated on several occasions (see Verloes et al. 1996). Ehlers-Danlos syndrome type IV is caused by several mutations in the type III procollagen gene (see Kuivaniemi et al. 1997 for a review), and there is also 56 evidence of type III procollagen gene mutations in familial aneurysms (Kontusaari et al. 1990, Anderson et al. 1996). In the present study, all the AAAs were atherosclerotic and without a family history. Yet, the relative proportion of type III collagen in AAA tissue was decreased. Therefore, one might question if there is a genetic factor involved, which only becomes apparent after the damaging effect of atherosclerosis on the aortic wall.

6.4. Type I and III collagens in AAA and diverticular disease of colon

The macroscopic appearance of both colon diverticula and abdominal aortic aneurysms suggests a loss of tensile strength of the tissue. Abnormalities in the architecture of elastin can be seen in both AAAs (Rizzo et al. 1989) and colon diverticulosis (Whiteway & Morson 1985). In addition, both of these clinical conditions may be associated with at least genetic factors and increased intraluminal pressure. The collagen content has been reported to be increased in AAAs (Menashi et al. 1987, Rizzo et al. 1989, Baxter et al. 1994, He & Roach 1994). In colon diverticulosis, the total amount of collagen is not altered, but an increase in the proportion of insoluble collagen has been suggested (Wess et al. 1995). The present data indicates that type I and III collagens are maturely cross-linked in both AAAs and colon diverticulosis. However, the increase in type III pN-collagen was only seen in aneurysm tissue. In contrast to Wess et al. (1995), there was no sign of increased cross-linking of collagen in colon diverticulosis. However, the result of Wess et al. was only based on hydroxyproline analysis and observations on collagen solubility in weak acids, which makes the obtained results susceptible to several errors. Mature collagen cross-linking is thought to be a general feature of fibrotic conditions (Ricard- Blum et al. 1993), which is supported by the abundant ICTP staining in liver fibrosis (Ricard-Blum et al. 1996). Because the amount of maturely cross-linked telopeptides (ICTP and IIINTP) and hydroxyproline was not increased in colon diverticulosis, our results did not indicate a marked fibrotic process in this clinical situation. In AAAs, mature cross-linking of type III collagen was found. There is evidence of a lower total amount of cross-links in ruptured intracranial aneurysms than in unruptured aneurysms or healthy controls (Gaetani et al. 1998). Furthermore, the involvement of proteases in the pathogenesis of aneurysms has been established on several occasions. Thus, it seems that the whole situation may depend on the presence or absence of balance between the synthesis, cross-linking and degradation of collagen. On the other hand, we were unable to find any signs of decreased cross-linking in the two AAA specimens, which had ruptured. The relative amount of type III collagen was decreased in AAA tissue but not in colon diverticulosis, suggesting that type III collagen has a different function in the wall of the colon compared to the aorta. However, the rate of type III procollagen synthesis seemed quite active in colon diverticulosis, where the amount of PIIINP in the soluble tissue extract was significantly increased compared to normal controls. In addition, soluble PIIINP eluted partly in its cleaved form in gel filtration analysis, suggesting ongoing synthesis. In AAA tissue, in turn, the synthesis of type III procollagen did not seem to be particularly increased when analyzed quantitatively, thus 57 suggesting that the delayed maturation of type III collagen might be the reason for increased type III pN-collagen.

6.5. Maturation and processing of type I and III collagens in ovarian and colon cancer

The previously reported increased expression of type I and III procollagens in ovarian (Kauppila et al. 1996) and type I collagen in gastrointestinal carcinomas (Ohtani et al. 1992) was supported by the increased synthesis of type I collagen measured by PINP assay in the soluble tumor extracts. However, the amounts of cross-linked type I and III collagen antigens and hydroxyproline were lower in the enzyme-digested malignant tissues than in the benign or normal controls. In addition, the cross-linking of especially type I collagen was found to be defective in tumor tissue. This is in agreement with the previous findings indicating that the expression of lysyl oxidase, a key enzyme in cross- linking process of fibrillar collagens, is decreased in malignancies (Hämäläinen et al. 1995, Peyrol et al. 1997). Thus, the malignant process seems to be very different from many other situations, e.g. wound healing, where the expression of lysyl oxidase increases together with increasing collagen expression (Fushida-Takemura et al. 1996). This might suggest that lysyl oxidase, in addition to having a role in collagen cross- linking, also has many other important regulatory functions in malignant tissues. Indeed, it has been suggested that lysyl oxidase might have a function as a tumor suppressor (Contente et al. 1990). In paper IV, the immunohistochemical studies on malignant ovarian tumors showed irregular arrangement of collagen fibers, as also described previously in our laboratory (Zhu et al. 1993a, Zhu et al. 1995). This is most probably due to the increased proteolytic activity in malignant tissue. Increased levels of MMP-2 have been reported in both ovarian and gastrointestinal cancer (Levy et al. 1991, Autio-Harmainen et al. 1993, Moser et al. 1994, Afzal et al. 1996). In addition to MMP-2, MMP-1 has also been shown to be associated with ovarian and colorectal cancer (van der Stappen et al. 1990, Kanamori et al. 1999). MMP-1 is able to cleave fibrillar type I and III collagens, whereas MMP-2 has been shown to degrade native type I collagen in addition to type IV collagen (Aimes & Quigley 1995). Indeed, MMP-1 produced by breast tumor cells has been reported to effectively mediate invasion by degrading type I collagen (Benbow et al. 1999). Interestingly, there is new data to suggest that MMP-1 is a target gene for tumor suppressor protein p53 (Sun et al. 1999). Thus, the ECM degradation seen in malignancies due to high levels of MMPs might be related to inactivation of p53. Many other matrix metalloproteinases, e.g. MMP-7 in gastrointestinal cancer (Adachi et al. 1999), have been associated with tumor progression and invasion as well (see Kähäri & Saarialho-Kere 1999), but these matrix metalloproteinases do not necessarily cleave fibrillar type I and III collagens. Even though the type I collagen synthesis detected by PINP was increased in malignant samples, no such tendency was seen in the case of type III collagen measured by PIIINP. The amount of type III collagen with the retained aminoterminal propeptide, called type III pN-collagen, was significantly increased in the insoluble ovarian 58 carcinoma tissue, suggesting enhanced type III collagen metabolism. This was not the case in colon cancer. The increased amounts of type III pN-collagen might contribute to tumor invasion by leaving the fibers thinner (Fleischmajer et al. 1985) and thus more susceptible to proteolysis by MMPs. An interesting observation in both ovarian and colon cancers was that type I collagen was more poorly cross-linked than type III collagen. This phenomenon might suggest that type III collagen undergoes mature cross-linking more rapidly than type I collagen, being therefore more resistant to proteolysis. On the other hand, firmly cross-linked collagen may have decreased endurance for tension (Allain et al. 1978), and hence be mechanically weak and susceptible to degradation. The above differences between type I and III collagen synthesis and processing might be related to a recent observation by Boyd & Balkwill (1999) suggesting that type I collagen induces MMP-2 production in ovarian carcinoma. Thus, type I collagen may be more important in the regulation and function of tumor matrix proteins and malignant cells than type III collagen. 7. Conclusions

The aim of the present study was to biochemically characterize the tissue contents of type I and III collagens in various clinical conditions.

1. In contrast to the current opinion, human atherosclerotic plaques contained a lot of type III collagen. In addition, the type III collagen was found to be completely processed, since the amount of type III pN-collagen was low. Both type I and III collagens were maturely cross-linked. This together with the low levels of type I and III procollagen propeptides in soluble tissue extracts warrants the conclusion that metabolic inertia prevails in atherosclerotic plaques. Thus, the possible thrombogenicity of type III collagen must be based on completely processed and cross-linked collagen.

2. In abdominal aortic aneurysms (AAAs), the relative amount of type III collagen was decreased, whereas the amount of incompletely processed type III collagen with the retained aminoterminal propeptide was increased. Because type III pN-collagen has been reported to regulate the fibril diameter, it is suggested that the collagen fibers in AAAs are thinner, possibly resulting in susceptibility to proteolysis and decreased tensile strength. Furthermore, the retained aminopropeptide might interfere with the cross-linking process.

3. Collagen synthesis was enhanced in ovarian and colon cancer, whereas the total collagen content in tissue was decreased. The cross-linking of especially type I collagen was defective in both these clinical conditions. These findings suggest an increased turnover of collagen in ovarian and colon cancer. The decreased cross- linking might be a general feature in malignant tissue, making collagen fibers more susceptible to proteolysis and thereby promoting tumor invasion and growth.

4. Type I and III collagens were very similar in normal colon tissue and in colon with diverticular disease. The only difference was the increased amount of free aminoterminal propeptide of type III collagen, related to ongoing type III collagen synthesis. There was no evidence of the previously reported increased cross-linking of collagen. 60

Taken together, the current results suggest that disturbances in the processing, cross- linking and fibril formation of type I and III collagens may lead or contribute to numerous pathological conditions. For example, the increased proteolysis in AAAs and malignant processes reported on several occasions might require defective processing of collagen molecules in order to take place in significant amounts. New biochemical markers for follow-up or even diagnostic purposes may arise when we learn about the pathogenesis of these conditions. Knowledge of the pathogenesis of these clinical conditions may also facilitate the development of new therapeutic applications. 8. References

Adachi E, Hopkinson I & Hayashi T (1997) Basement-membrane stromal relationships: Interactions between collagen fibrils and the lamina densa. Int Rev Cytol 173: 73-156. Adachi Y, Yamamoto H, Itoh F, Hinoda Y, Okada Y & Imai K (1999) Contribution of matrilysin (MMP-7) to the metastatic pathway of human colorectal cancers. Gut 45: 252-258. Afzal S, Lalani E-N, Foulkes WD, Boyce B, Tickle S, Cardillo MR, Baker T, Pignatelli M & Stamp GWH (1996) Matrix metalloproteinase-2 and tissue inhibitor of metalloproteinase-2 expression and synthetic matrix metalloproteinase-2 inhibitor binding in ovarian carcinomas and tumor cell lines. Lab Invest 74: 406-421. Aimes RT & Quigley JP (1995) Matrix metalloproteinase-2 is an interstitial collagenase. J Biol Chem 270: 5872-5876. Ala-Kokko L, Kontusaari S, Baldwin CT, Kuivaniemi H & Prockop DJ (1989) Structure of cDNA clones coding for the entire preproα1(III) chain of human type III procollagen. J Biol Chem 260: 509-516. Allain JC, Le Lous M, Bazin S, Bailey AJ & Delaunay A (1978) Isometric tension developed during heating of collagenous tissues. Biochim Biophys Acta 533: 147-155. Amento EP, Ehsani N, Palmer H & Libby P (1991) Cytokines and growth factors positively and negatively regulate interstitial collagen gene expression in human vascular smooth muscle cells. Arterioscler Thromb 11: 1223-1230. Anderson DW, Edwards TK, Ricketts MH, Kuivaniemi H, Tromp G, Stolle CA, Deak SB & Boyd CD (1996) Multiple defects in type III collagen synthesis are associated with the pathogenesis of abdominal aortic aneurysms. Ann NY Acad Sci 800: 216-228. Anderson DW, Thakker-Varia S, Tromp G, Kuivaniemi H & Stolle CA (1997) A glycine (415)-to- serine substitution results in impaired secretion and decreased thermal stability of type III procollagen in a patient with Ehlers-Danlos syndrome type IV. Hum Mutat 9: 62-63. Andreeva ER, Pugach IM & Orekhov AN (1997) Collagen-synthesizing cells in initial and advanced atherosclerotic lesions of human aorta. Atherosclerosis 130: 133-142. Autio-Harmainen H, Karttunen T, Hurskainen T, Höyhtyä M, Kauppila A & Tryggvason K (1993) Expression of 72 kilodalton type IV collagenase (gelatinase A) in benign and malignant ovarian tumors. Lab Invest 69: 312-321. Bachem MG, Wendelin D, Schneiderhan W, Haug C, Zorn U, Gross HJ, Schmid-Kotsas A & Grunert A (1999) Depending on their concentration oxidized low density lipoproteins stimulate extracellular matrix synthesis or induce apoptosis in human coronary artery smooth muscle cells. Clin Chem Lab Med 37: 319-326. Bailey AJ, Paul RG & Knott L (1998) Mechanisms of maturation and ageing of collagen. Mech Ageing Dev 106: 1-56. Bailey AJ, Wotton SF, Sims TJ & Thompson PW (1993) Biochemical changes in the collagen of human osteoporotic bone matrix. Connect Tissue Res 29: 119-132. 62 Bank RA, Robins SP, Wijmenga C, Breslau-Siderius LJ, Bardoel AFJ, Van der Sluijs HA, Pruijs HEH & TeKoppele JM (1999) Defective collagen crosslinking in bone, but not in ligament or cartilage, in Bruck syndrome: indications for a bone-specific telopeptide lysyl hydroxylase on chromosome 17. Proc Natl Acad Sci USA 96:1054-1058. Barnes MJ (1985) Collagens in atherosclerosis. Collagen Rel Res 5: 65-97. Bateman JF, Chiodo AA, Weng YM, Chan D & Haan E (1998) A type III collagen gly559 to arg helix mutation in Ehlers-Danlos syndrome type IV. Hum Mutat Suppl 1: 257-259. Bateman JF, Lamandé SR & Ramshaw JAM (1996) Collagen superfamily. In: Comper WD (ed) Extracellular matrix: Molecular components and interactions. Harwood Academic Publishers, Amsterdam, p 22-67. Bateman JF, Mascara T, Chan D & Cole WG (1984) Abnormal type I collagen metabolism by cultured fibroblasts in lethal perinatal osteogenesis imperfecta. Biochem J 217: 103-115. Baum J & Brodsky B (1999) Folding of peptide models of collagen and misfolding in disease. Curr Opin Struct Biol 9: 122-128. Bauriedel G, Hutter R, Welsch U, Bach R, Sievert H & Luderitz B (1999) Role of smooth muscle cell death in advanced coronary primary lesions: implications for plaque instability. Cardiovasc Res 41: 480-488. Baxter BT, Davis VA, Minion DJ, Wang YP, Lynch TG & McManus BM (1994) Abdominal aortic aneurysms are associated with altered matrix proteins of the nonaneurysmal aortic segments. J Vasc Surg 19: 797-803. Baxter BT, McGee GS, Shively VP, Drummond IAS, Dixit SN, Yamauchi M & Pearce WH (1992). Elastin content, cross-links, and mRNA in normal and aneurysmal human aorta. J Vasc Surg 16: 192-200. Becker U, Fietzek PP, Nowack H & Timpl R (1976) Isolation and structure of the amino-terminal cross-linking region in insoluble type III collagen. Hoppe-Seyler’s Z. Physiol Chem 357: 1409- 1415. Beighton PH, Murdoch JL & Votteler T (1969) Gastrointestinal complications of Ehlers-Danlos syndrome. Gut 10: 1004-1008. Benbow U, Schoenermark MP, Orndorff KA, Givan AL & Brinckerhoff CE (1999) Human breast cancer cells activate procollagenase-1 and invade type I collagen: invasion is inhibited by all- trans retinoic acid. Clin Exp Metastasis 17: 231-238. Berg RA, Schwartz ML & Crystal RG (1980) Regulation of the production of secretory proteins: intracellular degradation of newly synthesized “defective” collagen. Proc Natl Acad Sci USA 77: 4746-4750. Bini A, Mann KG, Kudryk BJ & Schoen FJ (1999) Noncollagenous bone matrix proteins, calcification and thrombosis in carotid artery atherosclerosis. Arterioscler Thromb Vasc Biol 19: 1852-1861. Birk DE, Fitch JM, Babiarz JP & Linsenmayer TF (1988) Collagen type I and type V are present in the same fibril in the avian corneal stroma. J Cell Biol 106: 999-1008. Blavier L, Henriet P, Imren S & Declerck YA (1999) Tissue inhibitors of matrix metalloproteinases in cancer. Ann NY Acad Sci 878: 108-119. Bonde M, Qvist P, Fledelius C, Riis BJ & Christiansen C (1994) Immunoassay for quantifying type I collagen degradation products in urine. Clin Chem 40: 2022-2025. Bonnet J, Garderes PE, Aumailley M, Moreau C, Gouverneur G, Benchimol D, Crockett R, Larrue J & Bricaud H (1988) Serum type III procollagen peptide levels in coronary artery disease (a marker of atherosclerosis). Eur J Clin Invest 18: 18-21. Bowness JM, Folk JE & Timpl R (1987) Identification of a substrate site for liver transglutaminase on the aminopropeptide of type III collagen. J Biol Chem 262: 1022-1024. Bowness JM, Tarr AH & Wiebe RI (1989) Transglutaminase-catalysed cross-linking: a potential mechanism for the interaction of fibrinogen, low density lipoprotein and arterial type III procollagen. Thromb Res 54: 357-367. Bowness JM, Tarr AH & Wong T (1988) Increased transglutaminase activity during skin wound healing in rats. Biochim Biophys Acta 967: 234-240. Boyd RS & Balkwill FR (1999) MMP-2 release and activation in ovarian carcinoma: the role of fibroblasts. Br J Cancer 80: 315-321. 63 Brown DL, Hibbs MS, Kearney M, Loushin C & Isner JM (1995) Identification of 92-kD gelatinase in human coronary atherosclerotic lesions: Association of active enzyme synthesis with unstable angina. Circulation 91: 2125-2131. Bruckner P, Bächinger HP, Timpl R & Engel J (1978) Three conformationally distinct domains in the amino-terminal segment of type III procollagen and its rapid triple helix coil transition. Eur J Biochem 90: 595-603. Bulleid NJ, Dalley JA & Lees JF (1997) The C-propeptide domain of procollagen can be replaced with a transmembrane domain without affecting trimer formation or collagen triple-helix folding during biosynthesis. EMBO J 16: 6694-6701. Bulleid NJ, Wilson R & Lees JF (1996) Type-III procollagen assembly in semi-intact cells: chain association, nucleation and triple-helix folding do not require formation of inter-chain disulphide bonds but triple-helix nucleation does require hydroxylation. Biochem J 317: 195- 202. Bächinger HP (1987) The influence of peptidyl-prolyl cis-trans isomerase on the in vitro folding of type III collagen. J Biol Chem 262: 17144-17148. Bächinger HP, Bruckner P, Timpl R & Engel J (1978) The role of Cis-Trans isomerization of peptide bonds in the coil triple helix conversion of collagen. Eur J Biochem 90: 605-613. Bächinger HP, Bruckner P, Timpl R, Prockop DJ & Engel J (1980) Folding mechanism of the triple helix in type-III collagen and type-III pN-collagen. Eur J Biochem 106: 619-632. Campa JS, Greenhalg RM & Powell JT (1987) Elastin degradation in abdominal aortic aneurysms. Atherosclerosis 65: 13-21. Canfield AE, Wren FE, Schor SL, Grant ME & Schor AM (1992) Aortic endothelial cell heterogeneity in vitro. Lack of association between morphological phenotype and collagen biosynthesis. J Cell Sci 102: 807-814. Cannon DJ & Davison PF (1978) Crosslinking in type III collagen of fetal tissue. Biochem Biophys Res Commun 85: 1373-1378. Chan D & Cole W (1984) Quantitation of type I and III collagens using electrophoresis of alpha chains and cyanogen bromide peptides. Anal Biochem 139: 322-328. Cheung DT, Benya PD, Perelman N, Dicesare PE & Nimni ME (1990) A highly specific and quantitative method for determining type III /I collagen ratios in tissues. Matrix 10: 164-171. Cheung DT, DiCesare P, Benya PD, Libaw E & Nimni ME (1983) The presence of intermolecular disulfide cross-links in type III collagen. J Biol Chem 258: 7774-7778. Chu M-L & Prockop DJ (1993) Collagen: Gene structure. Royce PM & Steinmann B (eds) Connective tissue and its heritable disorders. Molecular, genetic and medical aspects. Wiley- Liss Inc., New York, p 149-165. Clark CC (1979) The distributional initial characterization of oligosaccharide units on the COOH- terminal propeptide extensions of the pro-α2 chains of type I procollagen. J Biol Chem 254: 10798-10802. Clarke EP, Jain N, Brickenden A, Lorimer IA & Sanwal BD (1993) Parallel regulation of procollagen I and colligin, a collagen-binding protein and a member of the serine protease inhibitor family. J Cell Biol 121: 193-199. Contente S, Kenyon K, Rimoldi D & Friedman RM (1990) Expression of gene rrg is associated with reversion of NIH 3T3 transformed by LTR-c-H-ras. Science 249: 796-798. Cronlund AL, Smith BD & Kagan HM (1985) Binding of lysyl oxidase to fibrils of type I collagen. Connect Tissue Res 14: 109-119. D’Ortho MP, Will H, Atkinson S, Butler G, Messent A, Gavrilovic J, Smith B, Timpl R, Zardi L & Murphy G (1997) Membrane-type matrix metalloproteinases 1 and 2 exhibit broad-spectrum proteolytic capacities comparable to many matrix metalloproteinases. Eur J Biochem 250: 751- 757. Dabbagh K, Laurent GJ, McAnulty RJ & Chambers RC (1998) Thrombin stimulates smooth muscle cell procollagen synthesis and mRNA levels via PAR-1 mediated mechanism. Thromb Haemost 79: 405-409. Davis V, Persidskaia R, Baca-Regen L, Itoh Y, Nagase H, Persidsky Y, Ghorpade A & Baxter BT (1998) Matrix metalloproteinase–2 production and its binding to the matrix are increased in abdominal aortic aneurysms. Arterioscler Thromb Vasc Biol 18: 1625-1633. 64 De Flora S, Izzotti A, Walsh D, Degan P, Petrilli GL & Lewtas J (1997) Molecular epidemiology of atherosclerosis. FASEB J 11: 1021-1031. Deak S, Ricotta JJ, Mariani TJ, Deak ST, Zatina MA, MacKenzie JW & Boyd CD (1992) Abnormalities in the biosynthesis of type III procollagen in cultured skin fibroblasts from two patients with multiple aneurysms. Matrix 12: 92-100. Deyl Z, Novotná J, Mikšík I, Jelínková D, Uhrová M & Suchánek M (1997) Quantitation of collagen types I, III and V in tissue slices by capillary electrophoresis after cyanogen bromide solubilization. J Chromatogr B 689: 181-194. Dobrin PB, Baker WH & Gley WC (1984) Elastolytic and collagenolytic studies on arteries. Arch Surg 119: 405-409. Dobrin PB & Mrkvicka R (1994) Failure of elastin or collagen as possible critical connective tissue alterations underlying aneurysmal dilatation. Cardiovasc Surg 2: 484-488. Drake FH, Dodds RA, James IE, Connor JR, Debouk C, Richardson S, Lee-Rykaczewski E, Coleman L, Rieman D, Barthlow R, Hastings G & Gowen M (1996) Cathepsin K, but not cathepsins B, L, or S, is abundantly expressed in human osteoclasts. J Biol Chem 271: 12511- 12516. Elmore JR, Keister BF, Franklin DP, Youkey JR & Carey DJ (1998) Expression of matrix metalloproteinases and TIMPs in human abdominal aneurysms. Ann Vasc Surg 12: 221-228. Emanuel BS, Cannizarro LA, Seyer JM & Myers JC (1985) Human α1(III) and α2(V) procollagen genes are located on the long arm of chromosome 2. Proc Natl Acad Sci USA 82: 3385-3389. Engel J & Prockop DJ (1991) The zipper-like folding of collagen triple helices and the effects of mutations that disrupt the zipper. Annu Rev Biophys Biophys Chem 20: 137-152. Epstein EH & Munderloch NH (1975) Isolation and characterization of CNBr peptides of human [α1(III)]3 collagen and tissue distribution of [α1(I)]2α2 and [α1(III)]3 collagens. J Biol Chem 25: 9304-9312. Everts V, van der Zee E, Creemers L and Beertsen W (1996) Phagocytosis and intracellular digestion of collagen, its role in turnover and remodelling. Histochemical J 28: 229-245. Eyre DR (1987) Collagen cross-linking amino acids. Methods Enzymol 144: 115-139. Eyre DR, Dickson IR & van Ness K (1988) Collagen cross-linking in human bone and articular cartilage. Biochem J 252: 495-500. Eyre DR, Paz MA & Gallop PM (1984) Cross-linking in collagen and elastin. Annu Rev Biochem 53: 717-748. Feeley TM, Leen EJ, Colgan M-P, Moore DJ, Hourihane DO & Shanik GD (1991) Histologic characteristics of carotid artery plaque. J Vasc Surg 13: 719-724. Fessler LI, Timpl R & Fessler JH (1981) Assembly and processing of procollagen type III in chick embryo blood vessels. J Biol Chem 256: 2531-2537. Fishman DA, Kearns A, Chilukuri K, Bafetti LM, O’Toole EA, Georgacopoulos J, Ravosa MJ & Stack MS (1998) Metastatic dissemination of human ovarian epithelial carcinoma is promoted by alpha2beta1-integrin-mediated interaction with type I collagen. Invasion Metastasis 18: 15- 26. Fitzsimmons C, Proudfoot D & Bowyer DE (1999) Monocyte prostaglandins inhibit procollagen secretion by human vascular smooth muscle cells: implications for plaque stability. Atherosclerosis 142: 287-293. Fledelius C, Johnsen AH, Cloos PAC, Bonde M & Qvist P (1997) Characterization of urinary degradation products derived from type I collagen. Identification of a β-isomerized Asp-Gly sequence within the C-terminal telopeptide (α1) region. J Biol Chem 272: 9755-9763. Fleischmajer R, Olsen BR, Timpl R, Perlish JS & Lovelace O (1983) Collagen fibril formation during embryogenesis. Proc Natl Acad Sci USA 80: 3354-3358. Fleischmajer R & Perlish JS (1986) The role of aminopropeptide in collagen fibrillogenesis. Rheumatology 10: 91-103. Fleischmajer R, Perlish JS, Burgeson RE, Shaikh-Bahai F & Timpl R (1990) Type I and III collagen interactions during fibrillogenesis. Ann NY Acad Sci 580: 161-175. Fleischmajer R, Perlish JS & Timpl R (1985) Collagen fibrillogenesis in human skin. Ann NY Acad Sci 460: 246-257. 65 Fleischmajer R, Timpl R, Tuderman L, Raisher L, Wiestner M, Perlish JS & Graves PN (1981) Ultrastructural identification of extension aminopropeptides of type I and III collagens in human skin. Proc Natl Acad Sci USA 78: 7360-7364. Freestone T, Turner RJ, Coady A, Higman DJ, Greenhalg RM & Powell JT (1995) Inflammation and matrix metalloproteinases in the enlarging abdominal aortic aneurysm. Arterioscler Thromb Vasc Biol 15: 1145-1151. Freije JMP, Díetz-Itza I, Balbín M, Sánchez LM & Blasco R (1994) Molecular cloning and expression of collagenase-3, a novel human matrix metalloproteinase produced by breast carcinomas. J Biol Chem 269: 16766-16773. Fushida-Takemura H, Fukuda M, Maekawa N, Chanoki M, Kobayashi H, Yashiro N, Ishii M, Hamada T, Otani S & Ooshima A (1996) Detection of lysyl oxidase gene expression in rat skin during wound healing. Arch Dermatol Res 288: 7-10. Gacko M & Chyczewski L (1997) Activity and localization of cathepsin B, D and G in aortic aneurysm. Int Surg 82: 398-402. Gaetani P, Tartara F, Grazioli V, Tancioni F, Infuso L & Rodriguez y Baena R (1998) Collagen cross-linkage, elastolytic and collagenolytic activities in cerebral aneurysms: a preliminary investigation. Life Sci 63: 285-92. Galis ZS, Sukhova GK, Lark MW & Libby P (1994) Increased expression of matrix metalloproteinases and matrix degrading activity in vulnerable regions of human atherosclerotic plaques. J Clin Invest 94: 2493-2503. Garnero P, Borel O, Byrjalsen I, Ferreras M, Drake FH, McQueney MS, Foged NT, Delmas PD & Delaissé J-M (1998) The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J Biol Chem 273: 32347-32352. Ghorpade A & Baxter BT (1996) Biochemistry and molecular regulation of matrix macromolecules in abdominal aortic aneurysms. Ann NY Acad Sci 800: 138-150. Gilchrist D, Schwartze U, Shields K, MacLaren L, Bridge PJ & Byers PH (1999) Large kindred with Ehlers-Danlos syndrome type IV due to a point mutation (G571S) in the COL3A1 gene of type III procollagen: low risk of pregnancy complications and unexpected longevity in some affected relatives. Am J Med Genet 82: 305-311. Glanville RW & Fietzek PP (1976) Amino acid sequence of the N-terminal non-triple-helical cross- link region of type III collagen. FEBS letters 71: 99-102. Grange JJ, Davis V & Baxter BT (1997) Pathogenesis of abdominal aortic aneurysm: an update and look toward the future. Cardiovasc Surg 5:256-265. Greilberger J, Schmut O & Jürgens G (1997) In vitro interactions of oxidatively modified LDL with type I, II, III, IV, and V collagen, laminin, fibronectin, and poly-D-lysine. Arterioscler Thromb Vasc Biol 17: 2721-2728. Gunja-Smith Z, Morales AR, Romanelli R & Woessner JF Jr (1996) Remodeling of human myocardial collagen in idiopathic dilated cardiomyopathy. Role of metalloproteinases and pyridinoline cross-links. Am J Pathol 148: 1639-1648. Halila R & Peltonen L (1986) Purification of human procollagen type III N-proteinase from placenta and preparation of antiserum. Biochem J 239: 47-52. Halme T, Peltonen J, Sims TJ, Vihersaari T & Penttinen R (1986) Collagen in human aorta. Changes in the type III/I ratio and concentration of the reducible crosslink, dehydrohydroxylysyninorleucine in ascending aorta from healthy subjects of different age and patients with annulo-aortic ectasia. Biochim Biophys Acta 881: 222-228. Hamano K, Kuga T, Takahashi M, Fujioka K, Katoh T, Zempo N, Fujimura Y & Esato K (1998) The lack of type III collagen in a patient with aneurysms and an aortic dissection. J Vasc Surg 28: 1104-1106. Hanson AN & Bentley JP (1983) Quantitation of type I to type III collagen ratios in small samples of human tendon, vessels and atherosclerotic plaques. Anal Biochem 130:32-40. Hanson DA & Eyre DR (1996) Molecular site specificity of pyridinoline and pyrrole cross-linking in type I collagen of human bone. J Biol Chem 271: 26508-26516. Hanson DA, Weis MAE, Bollen A-M, Maslan SL, Singer FR & Eyre DR (1992) A specific immunoassay for monitoring human bone resorption: Quantitation of type I collagen cross- linked N-telopeptides in urine. J Bone Miner Res 7: 1251-1258. 66 He CM & Roach MR (1994) The composition and mechanical properties of abdominal aortic aneurysms. J Vasc Surg 20: 6-13. Hedman K, Alitalo K, Lehtinen S, Timpl R & Vaheri A (1982) Deposition of an intermediate form of procollagen type III (pN-collagen) into fibrils in the matrix of amniotic epithelial cells. EMBO J 1: 47-52. Henderson EL, Geng Y-J, Sukhova GK, Whittemore AD, Knox J & Libby P (1999) Death of smooth muscle cells and expression of mediators of apoptosis by T lymphocytes in human abdominal aortic aneurysms. Circulation 99: 96-104. Henkel W (1996) Cross-link analysis of the C-telopeptide domain from type III collagen. Biochem J 318: 497-503. Henkel W & Glanville RW (1982) Covalent crosslinking between molecules of type I and type III collagen. Eur J Biochem 122: 205-213. Henkel W, Rauterberg J & Glanville RW (1979) Isolation of crosslinked peptides from insoluble human leiomyoma. Eur J Biochem 96: 249-256. Henney AM, Wakeley PR, Davies MJ, Foster K, Hembry R, Murphy G, Humphries S (1991) localization of stromelysin gene expression in atherosclerotic plaques by in situ hybridization. Proc Natl Acad Sci USA 88: 8154-8158. Hill RJ & Harper E (1984) Quantitation of types I and III collagens in human tissue samples and cell culture by cyanogen bromide analysis. Anal Biochem 141: 83-93. Hirota S, Imakita M, Kohri K, Ito A, Morii E, Adachi S, Kim H-M, Kitamura Y, Yutani C & Nomura S (1993) Expression of osteopontin messenger RNA by macrophages in atherosclerotic plaques: a possible association with calcification. Am J Pathol 143: 1003-1008. Hunter GC, Smyth SH, Aguirre ML, Baxter BT, King DD, Wang YP, Hall KA & Putnam CW (1996) Incidence and histologic characteristics of blebs in patients with abdominal aortic aneurysms. J Vasc Surg 24: 93-101. Huttner WB (1982) Sulphation of tyrosine residues is a widespread modification for proteins. Nature 299: 273-276. Hämäläinen E-R, Jones TA, Sheer D, Taskinen K, Pihlajaniemi T & Kivirikko KI (1991) Molecular cloning of human lysyl oxidase and assignment of the gene to chromosome 5Q23.3- 31.2. Genomics 11: 508-516. Hämäläinen E-R, Kemppainen R, Kuivaniemi H, Tromp G, Vaheri A, Pihlajaniemi T & Kivirikko KI (1995) Quantitative polymerase chain reaction of lysyl oxidase mRNA in malignantly transformed human cell lines demonstrates that their low lysyl oxidase activity is due to low quantities of its mRNA and low levels of transcription of the respective gene. J Biol Chem 270: 21590-21593. Hörlein D, Fietzek PP, Wachter E, Lapière CM & Kühn K (1979) Amino acid sequence of the aminoterminal segment of dermatosparactic calf-skin procollagen type I. Eur J Biochem 99: 31- 38. Irizarry E, Newman KM, Gandhi RH, Nackman GB, Halpern V, Wishner S, Scholes JV & Tilson MD (1993) Demonstration of interstitial collagenase in abdominal aortic aneurysm disease. J Surg Res 54: 571-574. Jackson CJ & Jenkins KL (1991) Type I collagen fibrils promote rapid vascular tube formation upon contact with the apical side of cultured endothelium. Exp Cell Res 192: 319-323. Jarrold BB, Bacon WL & Velleman SG (1999) Expression and localization of the proteoglycan decorin during the progression of cholesterol induced atherosclerosis in Japanese quail: implication for interaction with collagen I and lipoproteins. Atherosclerosis 146: 299-308. Jensen LT & Høst NB (1997) Collagen: scaffold for repair or execution. Cardiovasc Res 33: 535- 539. Jimenez SA, Bashey RI, Benditt M & Yankowski R (1977) Identification of collagen α1(I) trimer in embryonic chick tendons and calvaria. Biochem Biophys Res Commun 78: 1354-1361. Jukkola A, Risteli J, Niemelä O & Risteli L (1986) Incorporation of sulphate into type III procollagen by cultured human fibroblasts. Identification of tyrosine O-sulphate. Eur J Biochem 154: 219-224. 67 Jukkola A, Risteli J & Risteli L (1990) Chlorate inhibits tyrosine sulfation of human type III procollagen without affecting its secretion or processing. Biochem Biophys Res Commun 170: 264-269. Jukkola A, Tähtelä R, Thölix E, Vuorinen K, Blanco G, Risteli L & Risteli J (1997) Aggressive breast cancer leads to discrepant serum levels of the type I procollagen propeptides PINP and PICP. Cancer Res 57: 5517-5520. Juvonen J, Juvonen T, Laurila A, Alakärppä H, Lounatmaa K, Surcel H-M, Leinonen M, Kairaluoma MI & Saikku P (1997) Demonstration of Chlamydia pneumoniae in the walls of abdominal aortic aneurysms. J Vasc Surg 25: 499-505. Kaartinen M, van der Wal AC, van der Loos CM, Piek JJ, Koch KT, Becker AE & Kovanen PT (1998) Mast cell infiltration in acute coronary syndromes: implications for plaque rupture. J Am Coll Cardiol 32: 606-12. Kadler KE, Hojima & Prockop DJ (1987) Assembly of collagen fibrils de Novo by cleavage of the type I pC-collagen with procollagen C-proteinase. J Biol Chem 260: 15696-15701. Kadler KE, Holmes DF, Trotter JA & Chapman JA (1996) Collagen fibril formation. Biochem J 316: 1-11. Kadler KE & Watson RB (1995) Procollagen C-peptidase: procollagen C-proteinase. Methods Enzymol 248: 771-781. Kafienah W, Brömme D, Buttle DJ, Croucher LJ & Hollander AP (1998) Human cathepsin K cleaves native type I and II collagens at the N-terminal end of the triple helix. Biochem J 331: 727-732. Kagan HM (1994) Lysyl oxidase: mechanism, regulation and relationship to liver fibrosis. Path Res Pract 190: 910-919. Kagan HM, Raghavan J & Hollander W (1981) Changes in aortic lysyl oxidase activity in diet- induced atherosclerosis in the rabbit. Arteriosclerosis 1: 287-291. Kanamori Y, Matsushima M, Minaguchi T, Kobayashi K, Sagae S, Kudo R, Terakawa N & Nakamura Y (1999) Correlation between expression of the matrix metalloproteinase-1 gene in ovarian cancers and an insertion/deletion polymorphism in its promoter region. Cancer Res 59: 4225-4227. Karttunen T, Alavaikko A, Apaja-Sarkkinen M & Autio-Harmainen H (1988) An immunohistochemical study of laminin, type IV collagen and type-III pN-collagen with relation to reticular fibres in Hodgkin´s disease. Int J Cancer 41: 52-58. Kato H, Suzuki H, Tajima S, Ogata Y, Tominaga T, Sato A & Saruta T (1991) Angiotensin II stimulates collagen synthesis in cultured vascular smooth muscle cells. J Hypertens 9: 17-22. Katsuda S, Okada Y, Minamoto T, Oda Y, Matsui Y & Nakanishi I (1992) Collagens in human atherosclerosis. Immunohistochemical analysis using collagen type-specific antibodies. Arterioscler Thromb Vasc Biol 12: 494-502. Kauppila A, Puistola U, Risteli J & Risteli L (1989) Amino-terminal propeptide of type III procollagen: A new prognosis indicator in human ovarian cancer. Cancer Res 49: 1885-1889. Kauppila S, Saarela J, Stenbäck F, Risteli J, Kauppila A & Risteli L (1996) Expression for mRNAs for type I and III procollagens in serous ovarian cystadenomas and cystadenocarcinomas. Am J Pathol 148: 539-548. Keene DR, Sakai LY, Bächinger HP & Burgeson RE (1987) Type III collagen can be present on banded collagen fibrils regardless of fibril diameter. J Cell Biol 105: 2393-2402. Kelley J, Chrin L & Evans JN (1984) Microquantitation of insoluble tissue collagen (types I and III) by radiodilution assay. Anal Biochem 139: 115-125. Kenyon K, Modi WS, Contente S & Friedman RM (1993) A novel human cDNA with a predicted protein similar to lysyl oxidase maps to chromosome 15q24-q25. J Biol Chem 268: 18435- 18437 Kielty CM, Hopkinson I & Grant ME (1993) Collagen: the collagen family: structure, assembly, and organization in the extracellular matrix. In: Royce PM & Steinmann B (eds) Connective tissue and its heritable disorders. Molecular, genetic and medical aspects. Wiley-Liss Inc., New York, p 103-147. Kim Y, Boyd CD & Csiszar K (1995) A new gene with sequence and structural similarity to the gene encoding human lysyl oxidase. J Biol Chem 270: 7176-7182. 68 Kivirikko KI, Laitinen O, Prockop DJ (1967) Modifications of a specific assay for hydroxyproline in urine. Anal Biochem 19: 249-255. Klappacher G, Franzen P, Haab D, Mehrabi M, Binder M, Plesch K, Pacher R, Grimm M, Pribill I, Eichler HG & Glogar HD (1995) Measuring extracellular matrix turnover in the serum of patients with idiopathic or ischemic dilated cardiomyopathy and impact on diagnosis and prognosis. Am J Cardiol 75: 913-918. Knott L, Whitehead CC, Fleming RH & Bailey AJ (1995) Biochemical changes in the collagenous matrix of osteoporotic avian bone. Biochem J 310: 1045-1051. Knäuper V, López-Otin C, Smith B, Knight G & Murphy G (1996a) Biochemical characterization of human collagenase-3. J Biol Chem 271: 1544-1550. Knäuper V, Will H, López-Otin C, Smith B, Atkinson SJ, Stanton H, Hembry RM & Murphy G (1996b) Cellular mechanisms for human procollagenase-3 (MMP-13) activation. J Biol Chem 271: 17124-17131. Kockx MM, De Meyer GR, Buyssens N, Knaapen MW, Bult H & Herman AG (1998) Cell composition, replication, and apoptosis in atherosclerotic plaques after 6 months of cholesterol withdrawal. Circ Res 83: 378-387. Koivu J & Myllylä R (1987) Interchain disulfide bond formation in types I and II procollagen. J Biol Chem 262: 6159-6164. Kojima T, Miyaishi O, Saga S, Ishiguro N, Tsutsui Y & Iwata H (1998) The retention of abnormal type I procollagen and correlated expression of HSP 47 in fibroblasts from a patient with lethal osteogenesis imperfecta. J Pathol 184: 212-218. Kolpakov V, Gordon D & Kulik TJ (1995b) Nitric oxide-generating compounds inhibit total protein and collagen synthesis in cultured vascular smooth muscle cells. Circ Res 76: 305-309. Kolpakov V, Rekhter MD, Gordon D, Wang WH & Kulik TJ (1995a) Effect of mechanical forces on growth and matrix protein synthesis in the in vitro pulmonary artery. Analysis of the role of individual cell types. Circ Res 77: 823-831. Kontusaari S, Tromp G, Kuivaniemi H, Romanic AM & Prockop DJ (1990) A mutation in the gene for type III procollagen (COL3A1) in a family with aortic aneurysms. J Clin Invest 86: 1465- 1473. Kostoulas G, Lang A, Nagase H & Baici A (1999) Stimulation of angiogenesis through cathepsin B inactivation of the tissue inhibitors of matrix metalloproteinases. FEBS Lett 455: 286-290. Kovanen PT, Kaartinen M & Paavonen T (1995) Infiltrates of activated mast cells at the site of coronary atheromatous erosion or rupture in myocardial infarction. Circulation 92: 1084-1088. Krane SM, Byrne MH, Lemaître V, Henriet P, Jeffrey JJ, Witter JP, Liu X, Wu H, Jaenisch R & Eeckhout Y (1996) Different collagenase gene products have different roles in degradation of type I collagen. J Biol Chem 271: 28509-28515. Kratky RG, Ivey J & Roach MR (1999) Local changes in collagen content in rabbit aortic atherosclerotic lesions with time. Atherosclerosis 143: 7-14. Kuboki Y, Tsuzaki M, Sasaki S, Liu CF & Mechanic GL (1981) Location of the intermolecular cross-links in bovine dentin collagen, solubilization with trypsin and isolation of cross-link peptide containing dihydroxylysinonorleucine and pyridinoline. Biochem Biophys Res Commun 102: 119-126. Kuga T, Esato K, Zempo N, Fujioka K & Nakamura K (1998) Detection of type III collagen fragments in specimens of abdominal aortic aneurysms. Surg Today 28: 385-390. Kuivaniemi H, Tromp G & Prockop DJ (1997) Mutations in fibrillar collagens (types I, II, III, and XI), fibril-associated collagen (type IX), and network-forming collagen (type X) cause a spectrum of diseases of bone, cartilage, and blood vessels. Hum Mutat 9: 300-315. Kähäri V-M & Saarialho-Kere U (1997) Matrix metalloproteinases in skin. Exp Dermatol 6: 199- 213. Kähäri V-M & Saarialho-Kere U (1999) Matrix metalloproteinases and their inhibitors in tumour growth and invasion. Ann Med 31: 34-45. Lafont A, Durand E, Samuel JL, Besse B, Addad F, Lévy BI, Desnos M, Guérot C & Boulanger CM (1999) Endothelial dysfunction and collagen accumulation. Two independent factors for restenosis and constrictive remodeling after experimental angioplasty. Circulation. 100: 1109- 1115. 69 Laury-Kleintop LD, Gleason M & Tulenko TN (1999) Expression of the heterogenous nuclear ribonucleoprotein complex K protein and the prolyl-4-hydroxylase α-subunit in atherosclerotic arterial smooth muscle cells. Biochem Biophys Res Commun 260: 382-389. Lee RT & Libby P (1997) The unstable atheroma. Arterioscler Thromb Vasc Biol 17: 1859-1867. Lee RWH & Huttner WB (1983) Tyrosine-O-sulfated proteins of PC 12 pheochromocytoma cells and their sulfation by tyrosylprotein sulfotransferase. J Biol Chem 258: 11326-11334. Lees JF, Tasab M & Bulleid NJ (1997) Identification of the molecular recognition sequence which determines the type-specific assembly of procollagen. EMBO J 16: 908-916. Leuschner JRA & Haust MD (1986) Interstitial collagens in fibrous atherosclerotic lesions of human aorta. Path Biol 34: 14-18. Levy AT, Cioce V, Sobel ME, Garbisa S, Grigioni WF & Liotta LA (1991) Increased expression of the Mr 72,000 type IV collagenase in human colonic adenocarcinoma. Cancer Res 51: 439-444. Li Z, Li L, Zielke R, Chenng L, Xiao R, Crow MT & Stetler-Stevenson WG (1996) Increased expression of 72-kd type IV collagenase (MMP-2) in human aortic atherosclerotic lesions. Am J Pathol 148: 121-128. Liakka A, Karjalainen H, Virtanen I & Autio-Harmainen H (1995) Immuno-electron-microscopic localization of types III pN-collagen and IV collagen, laminin and tenascin in developing and adult human spleen. Cell Tissue Res 282: 117-127. Lindstedt L, Saarinen J, Kalkkinen N, Welgus H & Kovanen PT (1999) Matrix metalloproteinases- 3, -7, and –12, but not –9, reduce high density lipoprotein-induced cholesterol efflux from human macrophage foam cells by truncation of the carboxyl terminus of apolipoprotein A-I. Parallel losses of pre-beta particles and the high affinity component of efflux. J Biol Chem 274: 22627-22634. Liu X, Otter A, Scott PG, Cann JR & Kotovych G (1993) Conformational analysis of the type II and type III collagen α-1 chain C-telopeptides by 1H NMR and circular dichroism spectroscopy. J Biomol Struct Dyn 11: 541-555. Liu X, Wu H, Byrne M, Krane S & Jaenisch R (1997) Type III collagen is crucial for collagen fibrillogenesis and for normal cardiovascular development. Proc Natl Acad Sci 94: 1852-1856. López-Candales A, Holmes DR, Liao S, Scott MJ, Wickline SA & Thompson RW (1997) Decreased vascular smooth muscle cell density in medial degeneration of human abdominal aortic aneurysms. Am J Pathol 150: 993-1007. Lukens LN (1976) Time of occurrence of disulfide linking between procollagen chains. J Biol Chem 251: 3530-3538. Maeda S, Sawai T, Uzuki M, Takahashi Y, Omoto H, Seki M & Sakurai M (1995) Determination of interstitial collagenase (MMP-1) in patients with rheumatoid arthritis. Ann Rheum Dis 54: 970-975. Majamaa K, Savolainen ER & Myllylä VV (1992) Synthesis of structurally unstable type III procollagen in patients with cerebral artery aneurysm. Biochim Biophys Acta 1138: 191-196. Majors A & Ehrhart LA (1993) Basic fibroblast growth factor in the extracellular matrix suppresses collagen synthesis and type III procollagen mRNA levels in arterial smooth muscle cell cultures. Arterioscler Thromb 13: 680-686. Majors A, Ehrhart A, Pezacka EH (1997) Homocysteine is a risk factor for vascular disease – enhanced collagen production and accumulation by smooth muscle cells. Arterioscler Thromb Vasc Biol 17: 2074-2081 Mao D, Lee JK, VanVickle SJ, & Thompson RW (1999) Expression of collagenase-3 (MMP-13) in human abdominal aortic aneurysms and vascular smooth muscle cells in culture. Biochem Biophys Res Commun 261: 904-910. Matsumoto S, Katoh M, Saito S, Watanabe T & Masuho Y (1997) Identification of soluble type of membrane-type matrix metalloproteinase-3 formed by alternatively spliced mRNA. Biochim Biophys Acta 1354: 159-170. Mays PK, Bishop JE & Laurent GJ (1988) Age-related changes in the proportion of types I and III collagen. Mech Ageing Dev 45: 203-212. McCaffrey TA, Du B, Consigli S, Szabo P, Bray PJ, Hartner L, Weksler BB, Sanborn TA, Bergman G & Bush HL, jr. (1997) Genomic instability in the type II TGF-β1 receptor gene in atherosclerotic and restenotic vascular cells. J Clin Invest 100: 2182-2188. 70 McCaffrey TA, Nicholson AC, Szabo PE, Weksler ME & Weksler BB (1988) Aging and arteriosclerosis. The increased proliferation of arterial smooth muscle cells isolated from old rats associated with increased platelet-derived growth factor-like activity. J Exp Med 167: 163-174. McCullagh KA & Balian G (1975) Collagen characterization and cell transformation in human atherosclerosis. Nature 258: 73-75. McCullagh KG, Duance VC & Bishop KA (1980) The distribution of collagen types I, III and V (AB) in normal and atherosclerotic human aorta. J Pathol 130: 45-52. McGee GS, Baxter BT, Shively VP, Chisholm R, McCarthy WJ, Flinn WR, Yao JST & Pearce WH (1991) Aneurysm or occlusive disease – factors determining the clinical course of atherosclerosis of the infrarenal aorta. Surgery 110: 370-376. McLaughlin SH & Bulleid NJ (1998) Molecular recognition in procollagen chain assembly. Matrix Biol 16: 369-377. McMillan WD, Patterson BK, Keen RR & Pearce WH (1995) In situ localization and quantification of seventy-two-kilodalton type IV collagenase in aneurysmal, occlusive, and normal aorta. J Vasc Surg 22: 295-305. McMillan WD & Pearce WH (1999) Increased plasma levels of metalloproteinase-9 are associated with abdominal aortic aneurysms. J Vasc Surg 29: 122-129. McMillan WD, Tamarina NT, Cipollone M, Johnson DA, Parker MA & Pearce WH (1997) Size matters: the relationship between MMP-9 expression and aortic diameter. Circulation 96: 2228- 2232. Melkko J, Hellevik T, Risteli L, Risteli J & Smedsrød B (1994) Clearance of NH2-terminal propeptides of types I and III procollagen is a physiological function of the scavenger receptor in liver endothelial cells. J Exp Med 179: 405-412. Melkko J, Kauppila S, Risteli L, Niemi S, Haukipuro K, Jukkola A & Risteli J (1996) Immunoassay for the intact amino-terminal propeptide of human type I procollagen. Clin Chem 42: 947-954. Melkko J, Niemi S, Risteli L & Risteli J (1990) Radioimmunoassay of the carboxy-terminal propeptide of human type I procollagen. Clin Chem 36: 1328-1332. Menashi S, Campa JS, Greenhalg RM & Powell JT (1987) Collagen in abdominal aneurysm: Typing, content, and degradation. J Vasc Surg 6: 578-582. Mesh CL, Baxter BT, Pearce WH, Chisholm RL, McGee GS & Yao JST (1992) Collagen and elastin gene expression in aortic aneurysms. Surgery 112: 256-262. Miller EJ, Epstein EH Jr. & Piez KA (1971) Identification of three genetically distinct collagens by cyanogen bromide cleavage of insoluble human skin and cartilage collagen. Biochem Biophys Res Commun 42: 1024-1029. Miller EJ, Finch JE jr., Chung E & Butler WT (1976) Specific cleavage of the native type III collagen molecule with trypsin. Arch Biochem Biophys 173: 631-637. Miller EJ, Furuto DK & Narkates AJ (1991) Quantitation of type I, III and V collagens in human tissue samples by high-performance liquid chromatography of a selected cyanogen bromide peptides. Anal Biochem 196: 54-60. Ming SC (1998) Diverticular diseases of the colon. In: Ming SC & Goldman H (eds) Pathology of the gastrointestinal tract. Williams & Wilkins, Baltimore, Maryland, p 801-818. Minion DJ, Davis VA, Nejezchleb BS, Wang Y, McManus BM & Baxter BT (1994) Elastin is increased in abdominal aortic aneurysms. J Surg Res 57: 443-446. Minion DJ, Wang Y, Lynch TG, Fox IJ, Prorok GD & Baxter BT (1993) Soluble factors modulate changes in collagen gene expression in abdominal aortic aneurysms. Surgery 114:252-257. Miura S, Kodaira S & Hosoda Y (1993) Immunohistologic analysis of the extracellular matrix components of the fibrous stroma of human colon cancer. J Surg Oncol 53: 36-42. Morton LF & Barnes MJ (1982) Collagen polymorphism in the normal and diseased blood vessel wall. Investigation of collagens types I, III and V. Atherosclerosis 42: 41-51. Moser TL, Young TN, Rodriguez GC, Pizzo SV, Bast RTC Jr & Stack MS (1994) Secretion of extracellular matrix-degrading proteinases is increased in epithelial ovarian carcinoma. Int J Cancer 56: 552-559. Movahed MR (1999) Infection with Chlamydia pneumoniae and atherosclerosis: a review. JSC Med Assoc 95: 303-308. 71 Murata K, Motayama T & Kotake C (1986) Collagen types in various layers of the human aorta and their changes with the atherosclerotic process. Atherosclerosis 60: 251-262. Murphy G, Knäuper V, Cowell S, Hembry R, Stanton H, Butler G, Freije J, Pendás AM & López- Otín C (1999) Evaluation of some newer matrix metalloproteinases. Ann NY Acad Sci 878: 25- 39. Myers PR & Tanner MA (1998) Vascular endothelial cell regulation of extracellular matrix collagen: role of nitric oxide. Arterioscler Thromb Vasc Biol 18:717-722. Nagata K (1998) Expression and function of heat shock protein 47: a collagen-specific molecular chaperone in the endoplasmic reticulum. Matrix Biol 16: 379-386. Nakai A, Satoh M, Hirayoshi K & Nagata K (1992) Involvement of the stress protein HSP47 in procollagen processing in the endoplasmic reticulum. J Cell Biol 117: 903-914. Narcisi P, Richards AJ, Ferguson SD & Pope FM (1994) A family with Ehlers-Danlos syndrome type III/articular hypermotility syndrome has a glycine 637-to-serine substitution in type III collagen. Hum Molec Genet 3: 1617-1620. Newman KM, Malon AM, Shin RD, Scholes JV, Ramey WG & Tilson MD (1994) Matrix metalloproteinases in abdominal aortic aneurysm: characterization, purification, and their possible sources. Connect Tissue Res 30: 265-276. Niemelä O (1985) Radioimmunoassays for type III procollagen amino-terminal peptides in humans. Clin Chem 31: 1301-1304. Niemelä O, Risteli L, Parkkinen J & Risteli J (1985) Purification and characterization of the N- terminal propeptide of human type III procollagen. Biochem J 232: 145-150. Nikkari ST, O’Brien KD, Ferguson M, Hatsukami T, Welgus HG, Alpers CE & Clowes AW (1995) Interstitial collagenase (MMP-1) expression in human carotid athesclerosis. Circulation 92: 1393-1398. Oemar BS & Lüscher TF (1997) Connective tissue growth factor. Friend or foe? Arterioscler Thromb Vasc Biol 17: 1483-1489. Ohtani H, Kuroiwa A, Obinata M, Ooshima A & Nagura H (1992) Identification of type I collagen- producing cells in human gastrointestinal carcinomas by non-radioactive in situ hybridization and immunoelectron microscopy. J Histochem Cytochem 40: 1139-1146. Ohuchi E, Imai K, Fujii Y, Sato H, seiki M & Okada Y (1997) Membrane type I matrix metalloproteinase digests interstitial collagens and other extracellular matrix macromolecules. J Biol Chem. 272: 2446-2451. Okada Y, Naka K, Kawamura K, Matsumoto T, Nakanishi I, Fujimoto N, Sato H & Seiki M (1995) Localization of matrix metalloproteinase-9 (92 kDa gelatinase/type IV collagenase=gelatinase B) in osteoclasts; implications for bone resorption. Lab Invest 72: 311-322. Ooshima A (1981) Collagen αB chain: Increased proportion in human atherosclerosis. Science 213: 666-668. Otani Y, Okazaki I, Arai M, Kameyama K, Wada N, Maruyama K, Yoshino K, Kitajima M, Hosoda Y & Tsuchiya M (1994) Gene expression of interstitial collagenase (matrix metalloproteinase 1) in gastrointestinal tract cancers. J Gastroenterol 29: 391-397. Painter NS & Burkitt DP (1971) Diverticular disease of the colon: a deficiency disease of western civilization. Br Med J 2: 450-454. Painter NS, Truelove SC, Ardran GM & Tuckey M (1965) Segmentation and the localization of intraluminal pressures in the human colon, with special reference to the pathogenesis of colonic diverticula. Gastroenterology 49: 169-177. Palombo D, Maione M, Cifiello BI, Udini M, Maggio D & Lupo M (1999) Matrix metalloproteinases. Their role in degenerative chronic diseases of abdominal aorta. J Cardiovasc Surg 40: 257-260. Pauschinger M, Knopf D, Petschauer S, Doerner A, Poller W, Schwimmbeck PL, Kuhl U & Schultheiss HP (1999) Dilated cardiomyopathy is associated with significant changes in collagen type I/III ratio. Circulation 99: 2750-2760. Peuhkurinen K, Risteli L, Jounela A & Risteli J (1996) Changes in interstitial collagen metabolism during acute myocardial infarction treated with streptokinase or tissue plasminogen activator. Am Heart J 131: 7-13. 72 Peuhkurinen KJ, Risteli L, Melkko J, Linnaluoto M, Jounela A & Risteli J (1991) Thrombolytic therapy with streptokinase stimulates collagen breakdown. Circulation 83: 1969-1975. Peyrol S, Raccurt M, Gerard F, Gleyzal C, Grimaud JA & Sommer P (1997) Lysyl oxidase gene expression in the stromal reaction to in situ and invasive ductal breast carcinoma. Am J Pathol 150: 497-507. Pickering JG, Ford CM & Chow LH (1996) Evidence for rapid accumulation and persistently disordered architecture of fibrillar collagen in human coronary restenosis lesions. Am J Cardiol 78: 633-637. Pickering JG, Ford CM, Tang B & Chow LH (1997) Coordinated effects of fibroblast growth factor-2 on expression of fibrillar collagens, matrix metalloproteinases and tissue inhibitors of matrix metalloproteinases by human vascular smooth muscle cells. Evidence for repressed collagen production and activated degradative capacity. Arterioscler Thromb Vasc Biol 17: 475- 482. Plenz G, Dorszewski A, Völker W, Ko YS, Severs NJ, Breithardt G & Robenek H (1999) Cholesterol-induced changes of type VIII collagen expression and distribution in carotid arteries of rabbit. Arterioscler Thromb Vasc Biol 19: 2395-2404. Podell DN & Abraham GN (1978) A technique for the removal of pyroglutamic acid from the amino terminus of proteins using calf liver pyroglutamate amino peptidase. Biochem Biophys Res Commun 81: 176-185. Pope FM, Nicholls AC, Narcisi P, Bartlett J, Neil-Dwyer G & Doshi B (1981) Some patients with cerebral aneurysms are deficient in type III collagen. Lancet 1:973-975. Powell J & Greenhalgh RM (1989) Cellular, enzymatic, and genetic factors in the pathogenesis of abdominal aortic aneurysms. J Vasc Surg 9: 297-304. Prockop DJ & Kivirikko KI (1995) Collagens: Molecular biology, diseases, and potentials for therapy. Annu Rev Biochem 64: 403-434. Pucci-Minafra I, Andriolo M, Basiricó L, Luparello C, Buccellato C, Garbelli R & Minafra S (1998) Absence of regular α2(I) collagen chains in colon carcinoma biopsy fragments. Carcinogenesis 19: 575-584. Pucci-Minafra I, Luparello C, Andriolo M, Basiricó L, Aquino A & Minafra S (1993) A new form of tumor and fetal collagen that binds laminin. Biochemistry 32: 7421-7427. Raspanti M, Ottani V & Ruggeri A (1989) Different architectures of the collagen fibril: morphological aspects and functional implications. Int J Macromol 11: 367-371. Rauterberg J, Jaeger E & Althaus M (1993) Collagens in atherosclerotic vessel wall lesions. In: Roessner A, Vollmer E (eds) Current Topics in Pathology. Springer-Verlag, Berlin, p 163-192. Reed D, Reed C, Stemmermann G & Hayashi T (1992) Are aortic aneurysms caused by atherosclerosis? Circulation 85: 205-211. Reiser KM (1991) Nonenzymatic glycation of collagen in aging and diabetes. Proc Soc Exp Biol Med 196: 17-29. Rekhter MD (1999) Collagen synthesis in atherosclerosis: too much and not enough. Cardiovasc Res 41: 376-384. Rekhter MD, O’Brien E, Shah N, Schwartz SM, Simpson JB & Gordon D (1996) The importance of thrombus organization and stellate cell phenotype in collagen I gene expression in human, coronary atherosclerotic and restenotic lesions. Cardiovasc Res 32: 496-502. Rekhter MD, Zhang K, Narayanan AS, Phan S, Schork MA & Gordon D (1993) Type I collagen gene expression in human atherosclerosis. Localization to specific plaque regions. Am J Pathol 143: 1634-1648. Ricard-Blum S, Bresson-Hadni S, Guerret S, Grenard P, Volle PJ, Risteli L, Grimaud JA & Vutton DA (1996) Mechanism of collagen network stabilization in human irreversible granulomatous liver fibrosis. Gastroenterology 111: 172:182. Ricard-Blum S, Esterre P & Grimaud JA (1993) Collagen cross-linking by pyridinoline occurs in non-reversible skin fibrosis. Cell Mol Biol 39: 723-727. Risteli J, Elomaa I, Niemi S, Novamo A & Risteli L (1993) Radioimmunoassay for the pyridinoline cross-linked carboxy-terminal telopeptide of type I collagen: a new serum marker of bone collagen degradation. Clin Chem 39: 635-640. 73 Risteli J, Niemi S, Trivedi P, Mäentausta O, Mowat AP & Risteli L (1988a) Rapid equilibrium radioimmunoassay for the amino-terminal propeptide of human type III procollagen. Clin Chem 34: 715-718. Risteli J & Risteli L (1995) Analysing connective tissue metabolites in human serum. Biochemical, physiological and methodological aspects. J Hepatol 22(Suppl. 2): 77-81. Risteli J & Risteli L (1999) Products of bone collagen metabolism. In: Dynamics of bone and cartilage metabolism: principles and clinical applications. Seibel MJ, Robins SP, Bilezikian JP (eds) Academic Press, London, p 275-287. Risteli L, Kauppila A, Mäkilä UM & Risteli J (1988b) Aminoterminal propeptide of type III procollagen in serum – an indicator of clinical behavior of advanced ovarian carcinoma? Int J Cancer 41: 409-414. Risteli L & Risteli J (1987) Analysis of extracellular matrix proteins in biological fluids. Methods Enzymol 145: 391-411. Risteli L & Risteli J (1990) Noninvasive methods for detection of organ fibrosis. In: Connective tissue in health and disease. Rojkind M (ed) CRC press, Boca Raton, FL, p 61-98. Rizvi MAD, Katwa L, Spadone DP & Myers PR (1996) The effects of endothelin-1 on collagen type I and type III synthesis in cultured porcine coronary artery vascular smooth muscle cells. J Mol Cell Cardiol 28: 243-252. Rizzo RJ, McCarthy WJ, Dixit SN, Lilly MP, Shively VP, Flinn WR & Yao JST (1989) Collagen types and matrix protein content in human abdominal aneurysms. J Vasc Surg 10: 365-373. Rohde H, Bruckner P & Timpl R (1983) Immunochemical properties of the aminopropeptide of procollagen type III. Eur J Biochem 135: 197-202. Romanic AM, Adachi E, Kadler KE, Hojima Y & Prockop DJ (1991) Copolymerization of pNcollagen III and collagen I. J Biol Chem 266: 12703-12709. Ross R (1993) The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 362: 801- 809. Saito H, Papaconstantinou J, Sato H & Goldstein S (1997) Regulation of a novel gene encoding a lysyl oxidase-related protein in cellular adhesion and senescence. J Biol Chem 272: 8157-8160. Sajanti J & Majamaa K (1999) Detection of meningeal fibrosis after subarachnoid haemorrhage by assaying procollagen propeptides in cerebrospinal fluid. J Neurol Neurosurg Psychiatry 67: 185- 188. Sakalihasan N, Delvenne P, Nusgens BV, Limet R & Lapiere CM (1996) Activated forms of MMP2 and MMP9 in abdominal aortic aneurysms. J Vasc Surg 24: 127-133. Sandberg MM, Hirvonen HE, Elima KJ & Vuorio E (1993) Co-expression of collagens II and XI and alternative splicing of exon 2 of collagen II in several developing human tissues. Biochem J 294: 595-602. Santala M, Risteli J, Risteli L, Puistola U, Kacinski BM, Stanley ER & Kauppila A (1998) Synthesis and breakdown of fibrillar collagens: concomitant phenomena in ovarian cancer. Br J Cancer 77: 1825-1831. Santala M, Risteli L, Puistola U, Risteli J & Kauppila A (1995) Elevated serum ICTP concentrations reflect poor prognosis in patients with ovarian carcinoma. Ann Med 27: 57-61. Santala M, Simojoki M, Risteli J, Risteli L & Kauppila A (1999) Type I and III collagen metabolites as predictors of clinical outcome in epithelial ovarian cancer. Clin Cancer Res 5: 4091-4096. Sassi M-L, Eriksen H, Risteli L, Niemi S, Mansell J, Gowen M & Risteli J (2000) Immunochemical characterization of the assay for the carboxyterminal telopeptide of human type I collagen (ICTP). Loss of antigenicity by treatment with cathepsin K. Bone, in press. Satta J, Haukipuro K, Kairaluoma MI & Juvonen T (1997) Aminoterminal propeptide of type III procollagen in the follow-up of patients with abdominal aortic aneurysms. J Vasc Surg 25: 909- 915. Satta J, Juvonen T, Haukipuro K, Juvonen M & Kairaluoma MI (1995) Increased turnover of collagen in abdominal aortic aneurysms, demonstrated by measuring the concentration of the aminoterminal propeptide of type III procollagen in peripheral and aortal blood samples. J Vasc Surg 22: 155-160. 74 Schwartze U, Goldstein JA & Byers PH (1997) Splicing defects in the COL3A1 gene: marked preference for 5’ (donor) spice-site mutations in patients with exon-skipping mutations and Ehlers-Danlos syndrome type IV. Am J Hum Genet 61:1276-1286. Shah PK, Falk E, Badimon JJ, Fernandez-Ortiz A, Mailhac A, Villareal-Levy G, Fallon JT, Regnstrom J & Fuster V (1995) Human monocyte-derived macrophages induce collagen breakdown in fibrous caps of atherosclerotic plaques. Potential role of matrix-degrading metalloproteinases and implications for plaque rupture. Circulation 92: 1565-1569. Sharp CA, Davie MWJ, Worsfold M, Risteli L & Risteli J (1996) Discrepant blood concentrations of type I procollagen propeptides in active Paget’s disease of bone. Clin Chem 42: 1121-1122. Shekhonin BV, Domogatsky SP, Idelson GL, Koteliansky VE & Rukosuev VS (1987) Relative distribution of fibronectin and type I, III, IV, V collagens in normal and atherosclerotic intima of human arteries. Atherosclerosis 67: 9-16. Shekhonin BV, Domogatsky SP, Muzykantov VR, Idelson GL & Rukosuev VS (1985) Distribution of type I, III, IV and V collagen in normal and atherosclerotic human arterial wall: Immunomorphological characteristics. Collagen Rel Res 5: 355-368. Shiraishi M, Segawa M & Takebayashi S (1980) An electron microscopic study on aging of collagenous fibrils in the large blood vessels of humans. J Electron Microsc 29: 139-43. Skotnicki JS, Zask A, Nelson FC, Albright JD & Levin JI (1999) Design and synthetic considerations of matrix metalloproteinase inhibitors. Ann NY Acad Sci 878: 61-72. Smedsrød B (1988) Aminoterminal propeptide of type III procollagen is cleared from the circulation by receptor-mediated endocytosis in liver endothelial cells. Coll Rel Res 8: 375-388. Smedsrød B, Melkko J, Risteli L & Risteli J (1990) Circulating C-terminal propeptide of type I procollagen is cleared mainly via the mannose receptor in liver endothelial cells. Biochem J 271: 345-350. Smith AN (1986) Colonic muscle in diverticular disease. Clin Gastroenterol 15:917-935. Smith LT, Schwartze U, Goldstein J & Byers PH (1997) Mutations in the COL3A1 gene result in the Ehlers-Danlos syndrome type IV and alterations in the size and distribution of the major collagen fibrils of the dermis. J Invest Dermatol 108: 241-217. Smith-Mungo LI & Kagan HM (1998) Lysyl Oxidase: properties, regulation and multiple functions in biology. Matrix Biol 16: 387-398. Spinale FG, Coker ML, Krombach SR, Mukherjee R, Hallak H, Houck WV, Clair MJ, Kribbs SB, Johnson LL, Peterson JT & Zile MR (1999) Matrix metalloproteinase inhibition during the development of congestive heart failure. Effects on left ventricular dimensions and function. Circulation Res 85: 364-376. Steinmann B, Superti-Furga A, Joller-Jemelka H, Cetta G & Byers PH (1989) Ehlers-Danlos syndrome type IV: A subset of patients distinguished by low serum levels of the amino terminal propeptide of type III procollagen. Am J Med Genet 34: 68:71. Sukhova GK, Schönbeck U, Rabkin E, Schoen FJ, Poole R, Billinghurst RC & Lippy P (1999) Evidence for increased collagenolysis by interstitial collagenases-1 and –3 in vulnerable human atheromatous plaques. Circulation 99: 2503-2509. Sukhova GK, Shi GP, Simon DI, Chapman HA & Libby P (1998) Expression of the elastolytic cathepsins S and K in human atheroma and regulation of their production in smooth muscle cells. J Clin Invest 102: 576-583. Sumner DS, Hokanson DE & Strandness DE Jr. (1970) Stress-strain characteristics and collagen- elastin content of abdominal aortic aneurysms. Surg Gynecol Obstet 130: 459-466. Sun Y, Sun Y, Wenger L, Rutter JL, Brinckerhoff CE & Cheung HS (1999) Human metalloproteinase-1 (collagenase-1) is a tumor suppressor protein p53 target gene. Ann NY Acad Sci 878: 638-641. Suster SM, Ronnen M & Bubis JJ (1984) Diverticulosis coli in association with Marfan´s syndrome. Arch Intern Med 144: 203. Suvanto-Luukkonen E, Risteli L, Sundström H, Penttinen J, Kauppila A & Risteli J (1997) Comparison of three serum assays for bone collagen formation during postmenopausal estrogen-progestin therapy. Clin Chim Acta 266: 105-116. 75 Szymanowicz AG, Bellon G, Laurain-Guillaume G, Delvincourt T, Randoux A, Caulet T & Borel JP (1982) An evaluation by sequential extraction of the proportions of collagen types from medium sized arteries. Artery 10: 250-265. Tagasako T, Nakamura K, Kashiwagi S, Inoue S-H, Oto H & Takeo K (1992) Analysis of collagen type III by uninterrupted sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting: Changes in collagen type III polymorphism in aging rats. Eletrophoresis 13: 373-378. Talwar HS, Griffiths CEM, Fisher GJ, Hamilton TA & Voorhees JJ (1995) Reduced type I and type III procollagens in photodamaged adult human skin. J Invest Dermatol 105: 285-290. Tamarina NA, McMillan WD, Shively VP & Pearce WH (1997) Expression of matrix metalloproteinases and their inhibitors in aneurysms and normal aorta. Surgery 122: 264-271. Termine JD, Kleinman HK, Whitson SW, Conn KM, McGarvey ML & Martin GR (1981) Osteonectin, a bone-specific protein linking mineral to collagen. Cell 26: 99-105. Thie M, Schlumberger W, Semich R, Rauterberg J & Robenek H (1991) Aortic smooth muscle cells in collagen lattice culture: effects on ultrastructure, proliferation and collagen synthesis. Eur J Cell Biol 55: 295-304. Thomas CV, Coker ML, Zellner JL, Handy JR, Crumbley AJ & Spinale FG (1998) Increased matrix metalloproteinase activity and selective upregulation in LV myocardium from patients with end-stage dilated cardiomyopathy. Circulation 97: 1708-1715. Thompson RW & Baxter BT (1999) MMP inhibition in abdominal aortic aneurysms. Rationale for a prospective randomized clinical trial. Ann NY Acad Sci 878: 159-178. Thompson RW, Holmes DR, Mertens RA, Liao S, Botney MD, Mecham RP, Welgus HG & Parks WC (1995) Production and localization of 92-kilodalton gelatinase in abdominal aortic aneurysms. J Clin Invest 96: 318-326. Thomson HJ, Busuttil A, Eastwood MA, Smith AN & Elton RA (1987) Submucosal collagen changes in the normal colon and in diverticular disease. Int J Colorect Dis 2: 208-213. Tintut Y, Parhami F, Boström K, Jackson SM & Demer LL (1998) cAMP stimulates osteoblast-like differentiation of calcifying vascular cells. Potential signaling pathway for vascular calcification. J Biol Chem 273: 7547-7553. Tomás C, Penttinen J, Risteli J, Risteli L & Kauppila A (1990) Simultaneous evaluation of epithelial cell function by CA 125 and stromal cell activity by aminoterminal propeptide of type III procollagen (PIIINP) in ovarian carcinoma. Ann Med 22: 115-121. Treharne GD, Boyle JR, Goodall S, Loftus IM, Bell PR & Thompson MM (1999) Marimastat inhibits elastin degradation and matrix metalloproteinase 2 activity in a model of aneurysm disease. Br J Surg 86: 1053-1058. Trivedi P, Hindmarsh P, Risteli J, Risteli L, Mowat AP & Brook CGD (1989) Growth velocity, growth hormone therapy, and serum concentrations of the amino-terminal propeptide of type III procollagen. J Pediatr 114: 225-230. Turnay J, Olmo N, Gavilanes JG & Lizarbe MA (1989) Collagen metabolism in human colon adenocarcinoma. Connect Tissue Res 23: 251-260. Turnay J, Olmo N, Gavilanes JG & Lizarbe MA (1991) Fibroblastlike primary cells from human colon adenocarcinoma explants: collagen biosynthesis. In Vitro Cell Dev Biol 27A: 447-452. Uzawa K, Grzesik WJ, Nishiura T, Kuznetsov SA, Robey PG, Brenner DA & Yamauchi M (1999) Differential expression of human lysyl hydroxylase genes, lysine hydroxylation, and cross- linking of type I collagen during osteoblastic differentiation in vitro. J Bone Mineral Res 14: 1272-1280. van den Berg JS, Limburg M, Kappelle LJ, Pals G, Arwert F & Westerveld A (1998) The role of type III collagen in spontaneous cervical arterial dissections. Ann Neurol 43: 494-498. van den Berg JS, Limburg M, Pals G, Arwert F, Westerveld A, Hennekam RC, Albrecht KW (1997) Some patients with intracranial aneurysms have a reduced type III/type I collagen ratio. A case-control study. Neurology 49: 1546-1551. van der Stappen WJ, Hendriks T & Wobbes T (1990) Correlation between collagenolytic activity and grade of histological differentiation in colorectal tumors. Int J Cancer 45: 1071-1078. van der Wal AC & Becker AE (1999) Atherosclerotic plaque rupture – pathologic basis of plaque stability and instability. Cardiovasc Res 41: 334-344. 76 van der Wal AC, Becker AE, van der Loos CM & Das PK (1994) Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 89: 36-44. van Keulen CJ, van de Akker E, Pals G & Rauwerda JA (1999) The role of type III collagen in the development of familial abdominal aortic aneurysms. Eur J Endovasc Surg 18: 65-70. van Kuppevelt TH, Veerkamp JH & Timmermans JAH (1995) Immunoquantification of type I, III, IV and V collagen in small samples of human lung parenchyma. Int J Biochem Cell Biol 27: 775-782. Vater CA, Harris ED & Siegel RC (1979) Native cross-links in collagen fibrils induce resistance to human synovial collagenase. Biochem J 181: 639-645. Verloes A, Sakalihasan N, Limet R & Koulischer L (1996) Genetic aspects of abdominal aortic aneurysm. Ann NY Acad Sci 800: 44-55. Virella G, Munoz JF, Galbraith GM, Gissinger C, Chassereau CH & Lopes-Virella MF (1995) Activation of human monocyte-derived macrophages by immune complexes containing low density lipoprotein. Clin Immunol Immunopathol 75: 179-189. Watson KE, Parhami F, Shin V & Demer LL (1998) Fibronectin and collagen I matrices promote calcification of vascular cells in vitro, whereas collagen IV matrix is inhibitory. Arterioscler Thromb Vasc Biol 18: 1964-1971. Weber KT (1989) Cardiac interstitium in health and disease: The fibrillar collagen network. J Am Coll Cardiol 13: 1637-1652. Weber KT (1998) Monitoring vascular sclerosis in hypertension. A new window of opportunity. Circulation 98: 498-500. Welgus HG, Burgeson RE, Wootton JAM, Minor RR, Fliszar C & Jeffrey JJ (1985) Degradation of monomeric and fibrillar type III collagens by human skin collagenase. J Biol Chem 260: 1052- 1059. Wess L, Eastwood M, Busuttil A, Edwards C & Miller A (1996b) An association between maternal diet and colonic diverticulosis in an animal model. Gut 39: 423-427. Wess L, Eastwood MA, Edwards CA Busuttil A, Miller A (1996a) Collagen alteration in an animal model of colonic diverticulosis. Gut 38: 701-706. Wess L, Eastwood MA, Wess TJ, Busuttil A & Miller A (1995) Cross linking of collagen is increased in colonic diverticulosis. Gut 37: 91-94. Whiteway J & Morson BC (1985) Elastosis in diverticular disease of the sigmoid colon. Gut 26: 258-266. Wilson KE, Langdon ASP, Lessells AM & Miller WR (1996) Expression of the extracellular proteins tenascin in malignant and benign ovarian tumours. Br J Cancer 74: 999-1004. Wilson R, Lees JF & Bulleid NJ (1998) Protein disulfide isomerase acts as a molecular chaperone during the assembly of procollagen. J Biol Chem 273: 9637-9643. Wohllebe M & Carmichael DJ (1978) Type-I trimer and type-I collagen in neutral-salt-soluble lathyritic-rat dentine. Eur J Biochem 92: 183-188. Yamauchi M & Mechanic GL (1988) Cross-linking of collagen. In: Nimni ME (ed) Collagen. CRC press, Boca Raton, FL, p 157-172. Zhu GG, Risteli J, Puistola U, Kauppila A & Risteli L (1993b) Progressive ovarian carcinoma induces synthesis of type I and type III procollagens in the tumor tissue and peritoneal cavity. Cancer Res 53: 5028-5032. Zhu GG, Risteli L, Mäkinen M, Risteli J, Kauppila A & Stenbäck F (1995) Immunohistochemical study of type I collagen and type I pN-collagen in benign and malignant ovarian neoplasms. Cancer 75: 1010-1017. Zhu GG, Stenbäck F, Risteli L, Risteli J & Kauppila A (1993a) Organization of type III collagen in benign and malignant ovarian tumors. An immunohistochemical study. Cancer 72: 1679-1684. Zucker S, Hymowitz M, Conner C, Zarrabi HM, Hurewitz AN, Matrisian L, Boyd S, Nicolson G & Montana S (1999) Measurement of matrix metalloproteinases and tissue inhibitors of metalloproteinases in blood and tissues. Clinical and experimental applications. Ann NY Acad Sci 878: 212-227.