Control of collagen fibril diameters in tissues It is proposed that radial growth of collaoen fibrils, which takes place in all connective tissues to varyin 9 extents, accordin 9 to the tensile stresses exerted on them, proceeds mainly by a99regation of protofibrils ( ~ 10 nm) and existing fibrils. In youn 9 tissues,fibrils are preventedfrom making frequent intimate contacts which would lead to aggregation by abundant interfibrillar proteoolycan, that keeps the fibrils apart. Collagen fibrils are probably unable to fuse except when the molecules within them are packed in the same sense, i.e. fusin 9 fibrils are parallel. The roughly equal numbers of parallel and antiparallel fibrils seen in several tissues must limit radial fibril growth in older tissues, where proteoglycan is usually less abundant. Possible origins of the balance o f fibril polarities, which must be conserved after fibril nucleation on cell or non-cell templates, are analysed. The two controlling factors, ambient proteoglycan and fibril polarity, working against the tendency of fibrils to fuse, account for many features of the observed distributions of collagen fibril diameters in diverse tissues and at different ages.

collagen fibrils diameters is quantized 8 and that tendon fibrils disaggregate under very mild conditions into regular subfibrils (ref. 6 and papers cited therein). Stress brings about disaggregation of echinoderm collagenous ligaments into subfibrils 9. Collagen fibrils fraying into bundles of subfibrils were seen in pathological tissues a°. Furthermore, the presence of regularly ordered proteoglycan molecules within cartilage fibrils argues strongly for a modular structure 6. As a corollary, covalent crosslinking between collagen molecules would occur chiefly within the smallest subunits (protofibrils 6). Thus fibrils may be built up and recycled as protofibrils (since aggregation-disaggregation is reversible) rather than via metabolism of individual molecules, thus conserving energy and limiting tissue destruction 6. The discrete distribution of sizes of disaggregated tendon subfibrils (10 12 and 25 nm) suggests that units of this size are stable. It may be relevant that the fibrils in embryonic tendon also average 25 nm in diameter in dehydrated plastic embedded sections 3'a ~ as are those in corneas of mammals and most other species 12. Smaller (10 and 17 nm) fibrils are seen in healing wounds ~3 and in corneas from bony fish respectively a2.

Keywords: Proteoglycans;fibril polarities; fibril fusion

Fibril polarity

The insoluble fibrils (mainly collagen) of connective tissue transmit tensile forces, whereas the interfibrillar soluble polymers (proteoglycans, hyaluronan) elastically absorb compressive stresses 1. Collagen fibril diameters may increase during development to >400 nm postmaturity 2'3. Moreover, those tissues subjected to the highest tensile loads contain the largest fibrils. In cornea, for example, fibril diameters are much smaller than those seen in tendon 4. The mechanisms whereby the increases occur and are controlled during development, ageing, and in pathological conditions have been the subject of controversy for several decades. One possibility not widely considered until recently, is that fibrils grow laterally by fusion 5,6. This mechanism would be strongly affected by the polarities of neighbouring fibrils 5, which must be determined at the earliest stages of fibrillogenesis. The presence of proteoglycans and hyaluronan in large amounts during early development is a second controlling factor, inhibiting lateral fusion of fibrils 3. Coupled with the 'environmental' stimuli of mechanical stress cycles, these predisposing factors lead to distributions of collagen fibril diameters of the form seen in diverse connective tissues. The most important aspects of our proposed scheme are discussed below.

Subflbrillar structure Although collagen fibrils were commonly regarded as homogeneous quasicrystalline aggregates of quarterstaggered collagen molecules 7 many data now suggest that fibrils are modular structures. In support are the observations that the distribution of connective tissue 014l -8130/92/050292-02 © 1992 Butterworth-HeinemannLimited 292

Int. J. Biol. Macromol., 1992, Vol. 14, October

Collagen fibrils are directional because their collagen molecules are oriented parallel to one another. This is shown by staining with heavy metals, which are taken up in an asymmetric pattern labelled a - e that is attributed to the lining-up of groups of hydrophilic or hydroplaobic amino acids in adjacent collagen molecules 14. It was observed 2'5'15'16 that a-e banding on neighbouring collagen fibrils could be parallel or antiparallel. Indeed, the fibrils took one or other direction with about equal frequencies in rat tail tendon 2'5, horse flexor and extensor tendons 17, and in human skin 15. Rabbit and human corneas contain antiparallel fibrils 18. Experimentally induced fibrotic tissue in rat liver also contains equal numbers of parallel/anti-parallel fibrils 19. The relevance of age, fibril diameter and tissue type needs to be further investigated, in this context. It is highly probable that fibril polarity is established at the nucleation of fibrillogenesis, and that it must be conserved from then on. However, the question of how polarity is initially determined has not been considered. A current model of fibrillogenesis 2° suggests that cells excrete fine fibrillar segments that aggregate axially to form long fibrils of uniform diameter. If so, the polarity template must be cellular, implying that (a) a cell has templates of opposite polarities in about equal numbers, or (b) cells have different polarities in about equal numbers, or (c) monopolar cells turn through 180° in laying down neighbouring fibrils. Alternatively, collagen fibril polarity could be established extracellularly either (d) by statistical chance, or (e) via an extracellular template. Such a template may be hyaluronan, which has two identical antiparallel faces (i.e. it is an ambidexteran) 2~. Hyaluronan meshworks therefore show parity in orientating molecules that interact with them.

It is relevant that hyaluronan is present in the extracellular space in large amounts in developing connective tissues 3. Alternatives (d) and (e) require fewer assumptions than (a)-(c), but there is currently little on which to base a prefer+nee for any particular model.

Fusion of fibrils as a major mechanism of fibril diameter growth Gieseking 5 and others proposed that fibrils aggregated to form larger fibrils. Electron micrographs showed probable fusions in process 3, and the very irregular shapes of aged tendon fibrils 22 are easily explained on this basis 6. The quantized distribution of fibril diameters s suggests that this process is active from early fibril growth, using protofibrils ( ,-~ 10 nm diameter) to give a highly ordered modular structure. A major problem with this scheme is that there is no reason for it to stop 6. Gieseking's suggestion that antiparallel fibrils probably could not fuse was criticized 15, since many contiguous parallel fibrils had not fused. We propose to develop Gieseking's suggestion, since later work 14 established that antiparallel arrangements of collagen molecules (an inescapable consequence of the aggregation of antiparallel fibrils or protofibrils) showed fewer stabilizing interactions than parallel arrays. Additionally, antiparallel arrays have bipolar symmetry, i.e. a symmetric banding incompatible with the a - e pattern 14. We have examined many micrographs from skin, tendon and sclera, failing to find any 'hybrid' fibrils, with antiparallel stretches ofa e banding within the same fibril. Thus, the presence of antiparallel groupings of fibrils must limit the frequency with which they could grow by fusion. It follows that the size of fibrils depends on the balance of polarities. If most neighbouring fibrils were antiparallel then fibrils would fuse less often. Conversely, where neighbouring fibrils are parallel, the chances of fusion are much higher. Further, if the polarity of the fibrils is cell-determined, different regions of the tissue could contain thicker or thinner fibrils, depending on the balance of cell polarities. Finally, the rate of growth and perhaps the final distribution of diameters must depend on the frequency with which fibrils meet prior to fusion. We suggest that tissues with rapid and strongly marked cycles of tension and relaxation will have thicker fibrils earlier in development, as in tendon 2'3 and 'active' fish skins 22, as a consequence of the more forceful contacts made between fibrils in these tissues.

Role of proteoglycans in preventing lateral fusion of fibrils Tendon fibrils are uniformly thin ( ~ 2 5 n m ) during development, until a stimulus, which could be programmed and/or be due to mechanical tension from attached muscle, causes a rapid increase in diameter ca. Concomitantly there is a marked decrease in tissue glycosaminoglycan3. On the basis of electron histochemistry, glycosaminoglycan was suggested to keep the fibrils apart 3. Below a critical ratio of glycosaminoglycan: fibril area, lateral fusion took place 2'3 with the displacement of fibril surface-associated proteoglycans. The detailed physicochemical mechanisms behind these interactions are not known. It seems unlikely that specific

glycosaminoglycans are involved to the exclusion of others, since collagen fibril diameters in a range of corneal stromas are constant at ~ 25 nm, in spite of dramatic differences in the ratios of keratan sulphate:chondroitin sulphate:dermatan sulphate 24. Axial extension of the fibrils takes place readily when proteoglycan concentrations are high. In summary we propose that collagen fibrils grow by fusion from protofibrils of about 10 nm diameter and that this process is controlled (1) by interfibrillar proteoglycan, and (2) by the antiparallel grouping of fibrils, each preventing fusion. The proteoglycan effect is dominant in young tissues, and the polarity mechanism assumes importance when, through cycles of stress and relaxation, frequent and forceful contacts between fibrils occur. Fibril polarity, with implications for later growth, must be established at the earliest stage of fibrillogenesis, and is conserved from then on.

Acknowledgement We are grateful to Dr D. Maurice for a useful discussion.

J. E. Scott

and David A. D. Parry

Chemical Morphology, Chemistry Building, Manchester University, Oxford Road, Manchester M13 9PL, UK

Physics and Biophysics, Massey University, Palmerston North, New Zealand

(received 20 February 1992; revised 20 May 1992)

References 1 2 3 4 5 6 7

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Scott, J. E. Phil. Trans. Roy. Soc. Lond. B 1975, 271,235 Parry, D. A. D. and Craig, A. S. Biopolymers 1978, 17, 843 Scott, J. E., Orford, C. R. and Hughes, E. W. Biochem. J. 1981, 195, 573 Schwarz, W. in 'Connective Tissue', (Eds R. E. Tunbridge, M. Keech, J. F. Delafresnaye and G. C. Wood), Blackwell, Oxford, 1957 Gieseking, R. Z. Zellforsch. 1962, 58, 160 Scott, J. E. J. Anat. 1990, 169, 23 Miller, A., Bradshaw, J., Jones, E. Y, Fraser, R. D. B., MacRae, T. P. and Suzuki, E. in 'Fibrosis', (Eds D. Evered and J. Whelan), Pitman, London Ciba Foundation Symp. 1985, 114, p. 65 Parry, D.A.D. andCraig, A.S. Nature(London) 1979,282,213 Erlanger, R. Welsch, U. and Scott, J. E. Unpublished data Holbrook, K. A. and Byers, P. H. in 'The Biochemistry of disease', (Eds J. Uitto and A. J. Perejda), Marcel Dekker, New York, vol. 12, p. 101 Craig, A. S. and Parry, D. A. D. Proc. Roy. Soc. Lond. B 1981, 212, 85 Craig, A.S. andParry, D . A . D . J . Ultrastruct. Res. 1981,74,232 Cooper, M., Ferguson, M. W. and Scott, J. E. Unpublished data Doyle, B. B., Hukins, D. W. L., Hulmes, D. J. S., Miller, A. and Woodhead-Galloway, J. J. Mol. Biol. 1975, 91, 79 Braun-Falco, O. and Rupee, M. J. Invest. Dermatol. 1964, 42,15 Torp, S., Baer, E. and Friedman, B. in 'Structure of Fibrous Polymers', (Eds E. D. T. Atkins and A. Keller), Butterworths, London, 1975, p. 223 Parry, D. A. D., Craig, A. S. and Barnes, G. R. G. Proc. Roy. Soc. Lond. B 1978, 203, 293 Scott, J. E. Unpublished data Scott, J. E., Bosworth, T. R., Cribb, A. M. and Gressner, A. Unpublished data Trelstad, R. L. and Birk, D. E. in 'Fibrosis', (Eds D. Evered and J. Whelan ), Pitman, London, Ciba Foundation Symp. 1985, 114, 4 Scott, J. E. FASEB J. 1992, 6, 2639 Craig, A. S., Eikenberry, E. F. and Parry, D. A. D. Connective Tiss. Res. 1987, 16, 213 Scott, J. E., Haigh, M., Geok-Eng, N. and Gibson. S. Clin. Sci. 1987, 72, 359 Scott, J. E. and Bosworth, T. R. Biochem. J. 1990, 270, 490

Int. J. Biol. Macromol., 1992, Vol. 14, October

293

Control of collagen fibril diameters in tissues.

It is proposed that radial growth of collagen fibrils, which takes place in all connective tissues to varying extents, according to the tensile stress...
262KB Sizes 0 Downloads 0 Views