Amino and carboxyl propeptides in bone collagen fibrils during embryogenesis

Tendon Extracellular Matrix Assembly, Maintenance and Dysregulation Throughout Life

In his Lissner Award medal lecture in 2000, Stephen Cowin asked the question: "How is a tissue built?" It is not a new question, but it remains as relevant today as it did when it was asked 20 years ago. In fact, research on the organization and development of tissue structure has been a primary focus of tendon and ligament research for over two centuries. The tendon extracellular matrix (ECM) is critical to overall tissue function; it gives the tissue its unique mechanical properties, exhibiting complex non-linear responses, viscoelasticity and flow mechanisms, excellent energy storage and fatigue resistance. This matrix also creates a unique microenvironment for resident cells, allowing cells to maintain their phenotype and translate mechanical and chemical signals into biological responses. Importantly, this architecture is constantly remodeled by local cell populations in response to changing biochemical (systemic and local disease or injury) and mechanical (exercise, disuse, and overuse) stimuli. Here, we review the current understanding of matrix remodeling throughout life, focusing on formation and assembly during the postnatal period, maintenance and homeostasis during adulthood, and changes to homeostasis in natural aging. We also discuss advances in model systems and novel tools for studying collagen and non-collagenous matrix remodeling throughout life, and finally conclude by identifying key questions that have yet to be answered.

Diverse biological functions of extracellular collagen processing enzymes

Collagens are abundant proteins in higher organisms, and are formed by a complex biosynthetic pathway involving intracellular and extracellular post-translational modifications. Starting from simple soluble precursors, this interesting pathway produces insoluble functional fibrillar and non-fibrillar elements of the extracellular matrix. The present review highlights recent progress and new insights into biological regulation of extracellular procollagen processing, and some novel functions of byproducts of these extracellular enzymatic transformations. These findings underscore the notion that released propeptides and other proteolytic products of extracellular matrix proteins have important biological functions, and that structural proteins are multifunctional. An emerging concept is that a dynamic interplay exists between extracellular products and byproducts with cells that helps to maintain normal cellular phenotypes and tissue integrity.

Collagen self-assembly and the development of tendon mechanical properties

The development of the musculoskeleton and the ability to locomote requires controlled cell division as well as spatial control over deposition of extracellular matrix. Self-assembly of procollagen and its final processing into collagen fibrils occurs extracellularly. The formation of crosslinked collagen fibers results in the conversion of weak liquid-like embryonic tissues to tough elastic solids that can store energy and do work. Collagen fibers in the form of fascicles are the major structural units found in tendon. The purpose of this paper is to review the literature on collagen self-assembly and tendon development and to relate this information to the development of elastic energy storage in non-mineralizing and mineralizing tendons. Of particular interest is the mechanism by which energy is stored in tendons during locomotion. In vivo, collagen self-assembly occurs by the deposition of thin fibrils in recesses within the cell membrane. These thin fibrils later grow in length and width by lateral fusion of intermediates. In vitro, collagen self-assembly occurs by both linear and lateral growth steps with parallel events seen in vivo; however, in the absence of cellular control and enzymatic cleavage of the propeptides, the growth mechanism is altered, and the fibrils are irregular in cross section. Results of mechanical studies suggest that prior to locomotion the mechanical response of tendon to loading is dominated by the viscous sliding of collagen fibrils. In contrast, after birth when locomotion begins, the mechanical response is dominated by elastic stretching of crosslinked collagen molecules.

Collagen fibril biosynthesis in tendon: a review and recent insights

The development and evolution of multicellular animals relies on the ability of certain cell types to synthesise an extracellular matrix (ECM) comprising very long collagen fibrils that are arranged in very ordered 3-dimensional scaffolds. Tendon is a good example of a highly ordered ECM, in which tens of millions of collagen fibrils, each hundreds of microns long, are synthesised parallel to the tendon long axis. This review highlights recent discoveries showing that the assembly of collagen fibrils in tendon is hierarchical, and involves the formation of fairly short "collagen early fibrils" that are the fusion precursors of the very long fibrils that occur in mature tendon.

pN collagen type III within tendon grafts used for anterior cruciate ligament reconstruction

This study measured the amount of immature collagen type III present in tendon rafts obtained from anterior cruciate ligament (ACL) reconstructions. These values were compared with those obtained from control grafts typically used for reconstruction--Achilles, patellar, and fascia lata--and also to the normal ACL. Analyses were performed using a commercially available radioimmunoassay (RIA). The RIA made use of a rabbit polyclonal antibody specific to the amino terminus of procollagen type III. The specificity of the Ab was confirmed by a western blot. Fibril diameter of each of the above samples was measured by transmission electron microscopy (TEM). We thus were able to determine if there was a relationship between pN collagen III content and fibril diameter. The mean amount of pN collagen type III in the normal tendon control group was 0.8 +/- 0.3 ng/microg total protein (range 0.0-2.5 ng/microg). There was significantly greater pN collagen III (16 +/- 3.7 ng/microg total protein) in the grafts containing an average fibril diameter <55 nm than in the normal tendons or ACL (P < 0.05). Grafts with an average fibril diameter >55 nm had similar levels of pN collagen III (1.0 +/- 0.79 ng/microg) as the controls. There was also significantly less pN-collagen III within the functional grafts (5.3 +/- 1.9 ng/microg) as compared to failed grafts, (21.6 +/- 5.1 ng/microg, P < 0.05). These results indicate that incomplete processing of procollagen III may be responsible for some of the ultrastructural alterations seen in tendon grafts. Since ultrastructural organization is believed to influence mechanical properties of these tissues. pN collagen III levels may be a possible indicator of ligament or tendon weakness.

Localization of pN-type IIA procollagen on adult bovine vitreous collagen fibrils

Type II procollagen is synthesized in long (type IIA) and short (type IIB) forms because of alternative splicing of mRNA; the long form containing an additional cysteine-rich domain in the amino-propeptide. An antiserum (IIA) that recognizes this domain was used for immunolocalization studies on adult bovine vitreous at light and electron microscopic levels and for Western blot analyses. The immunolocalization studies revealed labelling by the IIA antiserum of the vitreous collagen fibrils. This labelling was removed by prior extraction of the fibrils with 6 M guanidine hydrochloride (GuHCl) and the extract was shown to contain pN-type IIA procollagen. Adult vitreous collagen fibrils are coated with pN-type IIA procollagen, a finding with potential implications for vitreous collagen fibril structure and function.

Surface located procollagen N-propeptides on dermatosparactic collagen fibrils are not cleaved by procollagen N-proteinase and do not inhibit binding of decorin to the fibril surface

Dermatosparaxis is a recessive disorder of animals (including man) which is caused by mutations in the gene for the enzyme procollagen N-proteinase and is characterised by extreme skin fragility. Partial loss of enzyme activity results in accumulation of pNcollagen (collagen with N-propeptides) and abnormal collagen fibrils in the fragile skin. How the N-propeptides persist in the tissue and how abnormal fibril morphology results in fragile skin is poorly understood. Using biochemical and quantitative mass mapping electron microscopy we showed that the collagen fibrils in the skin of a dermatosparactic calf contained 57% type I pNcollagen and 43% type I collagen and the fibrils were irregularly arranged in bundles and hieroglyphic in cross-section. Image analysis of the fibril cross-sections suggested that the deviation from circularity of dermatosparactic fibrils was caused by N-propeptides of pNcollagen being located at the fibril surface. Comparison of experimental and theoretical axial mass distributions of the fibrils showed that the N-propeptides were located to the overlap zone of the fibril D-period (where D=67 nm, the characteristic axial periodicity of collagen fibrils). Treatment of the dermatosparactic fibrils with N-proteinase did not remove the N-propeptides from the fibrils, although the N-propeptides were efficiently removed by trypsin and chymotrypsin. However, the N-propeptides were efficiently cleaved by the N-proteinase when the pNcollagen molecules were extracted from the fibrils. These results are consistent with close packing of N-propeptides at the fibril surface which prevented cleavage by the N-proteinase. Long-range axial mass determination along the fibril length showed gross non-uniformity with multiple mass bulges. Of note is the skin fragility in dermatosparaxis, and also the appearance of mass bulges along the fibril long axis symptomatic of the fragile skin of mice which lack decorin. Western blot analysis showed that the dermatosparactic fibrils bound elevated levels of the proteoglycan, compared with normal skin fibrils. The results showed that N-propeptides can distort the morphology of fibrils, that they do not inhibit binding of gap-associated macromolecules (such as decorin) and that the normal mechanical properties of skin are strongly dependent on the close association of near-cylindrical fibrils, thereby enabling maximal fibril-fibril interactions.

Collagen fibril formation

Collagen is most abundant in animal tissues as very long fibrils with a characteristic axial periodic structure. The fibrils provide the major biomechanical scaffold for cell attachment and anchorage of macromolecules, allowing the shape and form of tissues to be defined and maintained. How the fibrils are formed from their monomeric precursors is the primary concern of this review. Collagen fibril formation is basically a self-assembly process (i.e. one which is to a large extent determined by the intrinsic properties of the collagen molecules themselves) but it is also sensitive to cell-mediated regulation, particularly in young or healing tissues. Recent attention has been focused on "early fibrils' or "fibril segments' of approximately 10 microns in length which appear to be intermediates in the formation of mature fibrils that can grow to be hundreds of micrometers in length. Data from several laboratories indicate that these early fibrils can be unipolar (with all molecules pointing in the same direction) or bipolar (in which the orientation of collagen molecules reverses at a single location along the fibril). The occurrence of such early fibrils has major implications for tissue morphogenesis and repair. In this article we review the current understanding of the origin of unipolar and bipolar fibrils, and how mature fibrils are assembled from early fibrils. We include preliminary evidence from invertebrates which suggests that the principles for bipolar fibril assembly were established at least 500 million years ago.

Effects of ascorbic acid on collagen matrix formation and osteoblast differentiation in murine MC3T3-E1 cells

Treatment of mouse MC3T3-E1 cells with ascorbic acid initiates the formation of a collagenous extracellular matrix and synthesis of several osteoblast-related proteins. We recently showed that ascorbic acid dramatically increases alkaline phosphatase and osteocalcin mRNAs and that this induction is blocked by inhibitors of collagen triple-helix formation (Franceschi and Iyer, J Bone Miner Res 7:235). In the present study, the relationship between collagen matrix formation and osteoblast-specific gene expression is explored in greater detail. Kinetic studies revealed that ascorbic acid increased proline hydroxylation in the intracellular procollagen pool within 1 h and stimulated the cleavage of type I collagen propeptides beginning at 2.5 h. Mature alpha 1(I) and alpha 2(I) collagen components were first detected at 10 h and continued to increase in both cell layer and culture medium for up to 72 h. Ascorbic acid also increased the rate of procollagen secretion from cell layers to culture medium. The secretion of another matrix protein, fibronectin, was only slightly affected. Alkaline phosphatase or its mRNA was first detected 2-3 days after ascorbic acid addition, but osteocalcin mRNA was not seen until day 6. Two inhibitors of collagen triple-helix formation, ethyl-3,4-dihydroxybenzoate and 3,4-dehydroproline, inhibited procollagen hydroxylation and alkaline phosphatase induction. 3,4-Dehydroproline also inhibited the induction of alkaline phosphatase and osteocalcin mRNAs. Surprisingly, induction was not blocked if cells were exposed to ascorbic acid before inhibitor addition. Alkaline phosphatase was also partially inhibited if cells were grown in the presence of purified bacterial collagenase. These results indicate that the induction of osteoblast markers by ascorbic acid does not require the continuous hydroxylation and processing of procollagens and suggest that a stable, possibly matrix-associated signal is generated at early times after ascorbic acid addition that allows subsequent induction of osteoblast-related genes.

Interactions of the neural cell adhesion molecule and the myelin-associated glycoprotein with collagen type I: involvement in fibrillogenesis

To gain insights into the functional role of the molecular association between neural adhesion molecules and extracellular matrix constituents, soluble forms of the myelin-associated glycoprotein (MAG) and the neural cell adhesion molecule (N-CAM), representing most of the extracellular domains of the molecules, were investigated in their ability to modify fibrillogenesis of collagen type I. MAG and N-CAM retarded the rate of fibril formation, as measured by changes in turbidity, and increased the diameter of the fibrils formed, but did not change the banding pattern when compared to collagen type I in the absence of adhesion molecules. Scatchard plot analysis of the binding of MAG and N-CAM to the fibril-forming collagen types I, II, III, and V suggest one binding site for N-CAM and two binding sites for MAG. Binding of MAG, but not of N-CAM, to collagen type I was decreased during fibril formation, probably due to a reduced accessibility of one binding site for MAG during fibrillogenesis. These results indicate that the neural adhesion molecules can influence the configuration of extracellular matrix constituents, thus, implicating them in the modulation of cell-substrate interactions.

Morphology of sheet-like assemblies of pN-collagen, pC-collagen and procollagen studied by scanning transmission electron microscopy mass measurements

At high concentrations, type I pN-collagen, pC-collagen and procollagen (the first 2 generated from procollagen by enzymic cleavage of C-propeptides and N-propeptides, respectively) can all be made to assemble in vitro into thin D-periodic sheets or tapes. Scanning transmission electron microscopy mass measurements show that the pN-collagen sheets and procollagen tapes have a mass per unit area corresponding to that of approximately 6.8 monolayers of close-packed molecules. pN-collagen sheets are extensive and remarkably uniform in mass thickness (fractional S.D. 0.035); procollagen tapes are neither as extensive nor as uniform in thickness. The mean thickness of pC-collagen tapes is less and the variability is greater. In pN-collagen sheets, the overlap: gap mass contrast in a D-period is increased from 5:4 (the ratio in a native collagen fibril) to 6:4, showing that the N-propeptides do not project into the gap but are folded back over the overlap zone. Assuming all N-propeptides to be constrained to the two surfaces of a sheet, their surface density can be found from the mass thickness of the sheet. In a lateral direction (i.e. normal to the axial direction where the spacing is D-periodic), the N-propeptide domains are calculated to be spaced, centre to centre, by 2.23 (+/- 0.1) nm on both surfaces. This value (approx. 1.5 x the triple-helix diameter) implies close-packing laterally with adjacent domains in contact. Sheet formation and the "surface-seeking" behaviour of propeptides can be understood in terms of the dual character of the molecules, evident from solubility data, with propeptides possessing interaction properties very different from those displayed by the rest of the molecule. The form and stability of sheets (and of first-formed fibrils assembling in vivo) could, it is suggested, depend on the partially fluid-like nature of lateral contacts between collagen molecules.

Developmental changes in the type I procollagen processing pathway in chick-embryo cornea

Type I procollagen processing in chick-embryo corneas was studied at days 12, 14 and 17 of development. Pulse-chase experiments and electrophoretic analysis of salt-soluble extracts showed developmental changes in the processing pathway. A kinetic model was fitted to the data to determine rate constants for processing of both N- and C-propeptides. Data for pro alpha 1(I)-chain processing and pro alpha 2(I)-chain processing were fitted separately (where pro means procollagen). Between days 12 and 17 the relative flux through the pC-collagen (procollagen chain lacking the N-propeptide) and pN-collagen (procollagen chain lacking the C-propeptide) pathways increased approx. 4-fold. Pro alpha 1(I) chains and pro alpha 2(I) chains were processed by slightly different routes. Variations in the rate constants were compared with electron-microscopic measurements of collagen fibril diameters at each stage of development. Diameters increased by less than 10% over the period from 12 to 17 days. It was concluded that fibril diameters are relatively insensitive to the pathway of procollagen processing in the salt-soluble pool.

Dermal collagen fibrils are hybrids of type I and type III collagen molecules

It has been suggested that dermal collagen fibrils with 67-nm periodicity consist of hybrids of type I and type III collagens. This is based on the assumption that all these banded fibrils are coated with type III collagen regardless of their diameter. However, conclusive evidence for this form of hybridization is lacking. In order to clarify this problem dermal collagen fibrils were disrupted into microfibrils using 8 M urea. Single and double indirect immunoelectron microscopy showed type III collagen at the periphery of intact collagen fibrils but no labeling with type I collagen antibodies, suggesting that the epitopes for this collagen were masked. Disrupted collagen fibrils revealed type I collagen throughout the fibril except for the periphery which was coated with type III collagen. Almost no type III collagen was noted in the interior of the collagen fibrils. Since type III collagen is present only at the periphery it suggests that this collagen has a different role than type I collagen and may have a regulatory function in fibrillogenesis.

Immunoelectron microscopy of osteonectin and type I collagen in osteoblasts and bone matrix

The pathway of production, secretion, and extracellular deposition of type I collagen and osteonectin was studied by immunoelectron microscopy using respective polyclonal antibodies. Protein A gold and immunogold methods yielded to similar results in human callus tissue used as a model of bone formation. The intracellular distribution of osteonectin in active osteoblasts is found as a faint immunolabeling of vesicular Golgi fields and some lamellae of rough endoplasmic reticulum. A more intensive labeling occurs in opaque cytoplasmic vesicles pointing to the process of secretion as some of the vesicles are connected with the basal cell membrane. Our type I collagen antibody did not label the respective intracellular compartments. Extracellularly, the type I collagen antibody showed a continuous labeling from the immature subcellular osteoid to the mineralized bone. Osteonectin antibodies were bound first to the deeper layer of osteoid maturation with intensity increasing below the mineralization front. Osteonectin is thought to be associated with mineralization of bone matrix.

D-periodic assemblies of type I procollagen

The solubility limit of purified chick type I procollagen, incubated at 37 degrees C in phosphate-buffered saline, was found to be in the range 1 to 1.5 mg/ml. At higher concentrations large aggregates formed. These comprised: (1) D-periodic assemblies; (2) narrow filaments with no apparent periodicity; and (3) segment-long-spacing-like aggregates. The D-periodic assemblies, which predominated at high concentrations, were separated from the other types of aggregate and found to be ribbon-like. Ribbons were uniform in thickness (approximately 8 nm) and up to 1 micron wide. Staining patterns showed features similar to those in native-type collagen fibrils. Immunolabelling indicated that the carboxyl-terminal propeptide domains were close to the carboxyl-terminal gap-overlap junction, and that the amino-terminal propeptide domains were folded over into the amino-terminal side of the overlap zone. Both propeptide domains appeared to be located on the surface of the assemblies. These observations show that intact propeptide domains hinder, but do not prevent, the formation of D-periodic assemblies. The presence of the propeptide domains on the surface of a growing assembly could restrict its lateral growth and limit its final thickness.

The regulation of size and form in the assembly of collagen fibrils in vivo

A possible mechanism for regulating the lateral growth of collagen fibrils in vivo is considered. A growth inhibitor associated with a particular part of the long semiflexible collagen molecule restricts that part of the molecule to the surface of the growing assembly. Lateral accretion ceases when these inhibitors form a complete circumferential layer around the fibril surface. Cell-mediated removal of the inhibitors allows lateral growth to proceed to a second limiting layer, and so on to subsequent limiting layers. In this way, cycles of inhibitor removal and limited lateral accretion permit growth to be synchronized over large populations of fibrils. Observed diameter distributions in bundles of embryonic and neonatal fibrils are those expected from a mechanism of this kind. The mechanism depends on the existence of axial order (D-periodicity) in fibrils, but not on any specific lateral packing of molecules. Rather, contacts between newly assembled molecules are presumed to be partly fluid-like in lateral directions (except where covalent cross-links have formed). Some initial fluidity in lateral packing prior to cross-linking does not preclude the subsequent emergence of quasi-crystalline packing as cross-links form. The cylindrical shape of fibrils in vivo may also be attributable in part to fluidity of intermolecular contacts at the growing surface.

Characterization of multiple forms of small collagenous apatite-binding proteins in bone

A number of small (Mrs 25-28 kDa) collagenous apatite-binding (SCAB) proteins that stain blue with 'Stains-All' have been isolated from fetal porcine bone by sequential extractions with 4M GuHCl (G1), followed by 0.5M EDTA (E), and again with 4M GuHCl (G2). Following purification under dissociative conditions, two types of SCAB proteins both with approximately one-third of their structure being collagenous, were identified in the EDTA extract. One type, which appears to be a novel protein, was revealed in two forms (SCABs 1 and 2, Mrs 25 and 28 kDa) that were recognized by a monoclonal antibody (MBP-322). The second type, SCAB 3, was also present in two forms; one form (SCAB 3a) having a lower affinity for hydroxyapatite than the other (SCAB 3b). These proteins were resistant to CNBr and displayed the chemical and immunochemical properties of the alpha 1 pN-propeptide of type I collagen. A third form of the propeptide (G2-28K) was a prominent component of the second 4M GuHCl extract. The chromatographic properties of serum alpha 1 (I) pN-propeptide were similar to SCAB 3a, indicating that SCAB 3b and G2-28K are post-translationally modified forms of the propeptide produced by bone cells. These propeptides may provide a link between the hydroxyapatite and collagen fibrils, and also have the potential to suppress collagen synthesis during bone resorption.

Retention of carboxypropeptides in type-II collagen fibrils in chick embryo chondrocyte cultures

An antibody reacting with the C-propeptide of chick type-II procollagen was used in an attempt to localize this terminal extension of the procollagen molecule (by immunogold labelling) during early collagen fibrillogenesis in chondrocyte cultures. After 2 days in culture the chondrocytes were surrounded by pericellular type-II collagen, as demonstrated by an indirect immunofluorescence labelling technique. An electron microscopy study of these cultures showed that the collagen fibrils were thin (approximately 15 nm diameter), with a poorly visible cross striation, sometimes enhanced by slight thickenings. The antibody against the C-propeptide of type-II procollagen labelled most of the collagen fibrils, according to a very regular pattern constituting a 60 nm periodicity. After 3 days the label was still present on the pericellular collagen fibrils but disappeared from the collagen fibrils of the extracellular matrix. Our results indicate that the C-propeptide of type-II procollagen is retained in the newly formed fibrils.

Identification of small collagenous proteins with properties of procollagen alpha 1 (I) pN-propeptide in fetal porcine calvarial bone

Several small collagenous apatite binding (SCAB) proteins have been extracted from the mineralized matrix of fetal porcine calvarial bone. One protein (SCAB 3), released on demineralization of bone with 0.5 M EDTA, appears to represent the alpha 1 pN-propeptide that is normally released during proteolytic processing of type I procollagen. The 28 Kd protein, which stains blue with "Stains-all", is reduced to a 19 Kd fragment by bacterial collagenase digestion, but is not susceptible to cyanogen bromide. The amino acid composition, blocked amino-terminus and immunological properties are all consistent with properties of alpha 1 (I) pN-propeptide. Fractionation on hydroxylapatite in the presence of urea has revealed a nonbinding (SCAB 3a) and a binding (SCAB 3b) form. Extraction of the demineralized matrix of bone with 4 M GuHCl revealed a third form (G2-28) which was similar to SCAB 3a on hydroxylapatite chromatography but showed differences on FPLC "Mono Q" resin. The occurrence of these different forms of pN-propeptide in bone may be of significance in collagen fibril-associated hydroxylapatite formation and in the regulation of osteoblastic function during bone resorption.

The carboxypropeptide trimer of type II collagen is a prominent component of immature cartilages and intervertebral-disc tissue

Immature bovine cartilages and intervertebral-disc tissue all revealed a prominent protein, not present in the adult tissues, in non-denaturing extracts made with chondroitin ABC lyase (EC 4.2.2.4), Streptomyces hyaluronidase (EC 4.2.2.1) or 1 M NaCl. The protein ran on SDS-polyacrylamide electrophoresis, before disulphide reduction, as a close doublet of bands of apparent molecular weight 110,000 and 105,000. After reduction, they dissociated respectively into two protein bands at 37,000 and 35,000, indicating that the initial molecules were disulphide-bonded trimers. Amino-terminal sequence analysis established the identity of both proteins (Mr 110,000 and Mr 105,000) as forms of the carboxypropeptide of type II collagen. The larger molecule appeared to be the trimer of intact alpha 1(II) carboxypropeptides and the smaller, a version composed of chains that were ten residues shorter at their amino-terminal ends. The material appears to be identical to chondrocalcin, a protein previously found to be enriched in fetal growth plate and named on the basis that it may play a role in cartilage calcification. The present findings, however, indicate that the protein is equally abundant in all type II collagen-synthesizing young cartilages, including nucleus pulposus of the intervertebral disc and other cartilages that never calcify.

The carboxylpropeptide of type I procollagen in skin fibrillogenesis

Previous studies suggested that the aminopropeptide of type I procollagen may initiate fibril formation. The purpose of this investigation was to study the location of the carboxylpropeptide of type I procollagen during collagen fibrillogenesis. Chick embryonic and posthatching skin specimens were studied by immunofluorescence and immunoelectron microscopy and by immunoblotting with antibodies against the amino and carboxylpropeptide of type I procollagen. The carboxylpropeptide was demonstrated at the surface of collagen fibrils, 20-40 nm in diameter (10-day embryos) and in fibrils, 40-65 nm (21-day embryos). In addition, the carboxylpropeptide was found at the cell surface and free in the ground substance. The aminopropeptide was only seen in fibrils, 20-30 nm in diameter, as previously reported. Ratios of pN-collagen/pC-collagen increased from 16 days embryonic to 3 and 9 days postembryonic skins. This study suggests that both pN-collagen (aminopropeptide plus collagen) and pC-collagen (carboxylpropeptide plus collagen) participate in fibrillogenesis.