Observations on the shrink temperature of collagen and its variations with age and disease.

The shrinkage temperature of collagen is related to its stability. When shrinkage takes place, it is thought that the hydrothermal agitation at this temperature is sufficient to overcome the forces, chiefly hydrogen bonds, which hold together the polypeptide chains. When stability is reduced by removal of hydrogen bonds with hydrogen bond breakers or by chemical damage, less energy is required to force the chains apart and so the shrink temperature is reduced. This is also paralleled by reduced resistance to the action of trypsin-like enzymes. Conversely, the introduction of new cross links into collagen, as in tanning, will increase enzyme resistance and the shrink temperature. The different forms of collagen, soluble and insoluble, which occur together in native collagen, vary somewhat in their stability. Banga, Bal6, and Szab6 (1956) have suggested that the soluble has a stabilizing effect on the insoluble. In addition, polysaccharides and non-collagen proteins are concerned in the biogenesis of collagen fibres and their development into fully mature collagen (Bowes and Kenten, 1950; Partridge, 1948; Jackson, 1953). These constituents, always present in intimate association with native collagen, are probably important in determining the final stability and low turnover rate characteristic of fully mature collagen. It was therefore thought of interest to examine the shrink temperatures of human tissues of different ages to ascertain whether there were any age trends, and, if so, to correlate them with those changes in the various constituents which are known to take place with development and ageing. Further, since collagen in the uterus is rapidly deposited during pregnancy and rapidly absorbed after parturition (Harkness and Harkness, 1954), a feature which distinguishes this from collagen of other tissues and organs, the shrink temperature of uterine collagen from the uterus of various ages and in pregnancy was therefore of special interest. In rheumatoid arthritis there is evidence, derived from electron microscopy, x-ray diffraction (Kellgren, Ball, Astbury, Reed, and Beighton, 1951; Kellgren, 1952), and histological studies (Glynn and Reading, 1956), that collagen fibres have suffered damage. We therefore thought it desirable to examine this and tissue from other "collagen disease" for any alteration in shrink temperature. Much of the data on shrink temperature has been derived from species other than man. Further, many experiments have been carried out on such materials as hide powder which have probably become so altered as a result of preparation that the tissue finally examined could hardly be regarded as normal. Small but significant changes in shrink temperatures can occur after relatively mild treatments. Consequently, as far as untreated tissues were concerned, we considered it important to measure their shrink temperatures as soon as possible after collection, avoiding any pre-treatments such as washing, solvent extraction or prolonged storage. Also, since comparatively few observations have been made on human tissue, and since some of the data on non-human tissue are unreliable for reasons already mentioned and sometimes conflicting, a number of problems, such as the effect of hydrogen bond breakers were re-examined using fresh human tissue. For a few experiments, animal tissue was also used. We have chosen normal saline in which to measure shrink temperatures since this appeared to be appropriate physiologically. It was found that the shrink temperature in saline was lower than that in water and was identical for different fibres in any given specimen. It was necessary to carry out shrink temperature measurements in a microapparatus when dealing with biopsied tissue available in only small amounts, and this method was used in most of the experiments described here.