The Keratin Defect and Hair-cycle of a New Mutant (Matted) in the House-mouse

Keratins probably vary widely in composition, but attempts to study their chemistry have been only partially successful (Bach, 1952; Neurath & Bailey, 1954; Rothman, 1954). All keratins are rich in cystine, and it has been shown that oxidation of thiol SH groups to cystine S-S occurs during keratinization (Giroud & Bulliard, 1930; Hardy, 1952). The strength of keratin depends largely upon cystine linkages (Goddard & Michaelis, 1934), hydrogen bonds (Alexander, 1951), and polar (salt) linkages (Speakman & Townsend, 1938); these unite the long polypeptide fibre molecules. The amino-acid composition is also important in determining the properties of keratin, and Block & Vickery (1931) considered the relative proportions of basic amino-acids to be significant. Auber (1952) believed that lateral compression of the polypeptide fibrils by the turgid inner root sheath was an important factor in keratinization.

In the present study the keratin defect in the matted mutant was examined by ordinary staining methods and by fluorescence microscopy. These hairs were compared with normal mouse-hairs in which the keratin cross linkages had been artificially broken. The molecular structure of the hair has been examined by X-ray diffraction. Thiol and cystine sulphur have been studied, and the aminoacid content of the hairs estimated. The keratin defect is discussed in the light of these results.

Sections of skin containing long-growing hair-follicles were treated by a modification of Auber’s (1952) method. The tissue was bulked stained in Mayer’s haemalum; sections were cut 6 μ thick and stained with 1 per cent, basic fuchsin and 0·1 per cent, picro-indigocarmine. The structure of the skin and hair-follicles in matted animals appeared normal, and we were unable, by this method, to detect any consistent abnormality in the keratinization of growing hairs within the follicle.

Hairs of normal, matted, and Naked (N / + ) mice, were treated with fluorochromes and examined microscopically by filtered ultraviolet rays. The most useful fluorochromes for this study were found to be acridine orange, thioflavine T, and rhodamine B. The hairs were plucked carefully in order to avoid trauma as far as possible. They were degreased in ether, dried, and then placed in 70 per cent, alcohol for 3 hours. After rinsing in distilled water, the hairs were immersed in a 0T per cent, aqueous solution of fluorochrome for 3 minutes. They were then rapidly rinsed in distilled water, dehydrated, cleared, and mounted in a nonfluorescing medium—DePeX. The apparatus employed was that described by Jarrett et al. (1956).

Hairs from normal mice were examined by this technique with acridine orange. These showed a green fluorescence; the contained melanin was dark and non-fluorescent. Very occasionally brilliant red areas were seen in the medulla. Patches of red fluorescence were found in a much higher proportion of, and were more extensive in, hairs of matted mice than in those of normal animals. The majority of matted hairs had one or more patches of red fluorescence (Plate, fig. A). By this means we were able to detect hairs from mutant mice as early as 9 days old before any breakage had occurred or the mice were abnormal to look at.

Normal hairs were cut, and then treated in the manner already described. It was then seen that the medulla at the cut end of the hair fluoresced red (Plate, fig. B). This fluorescence was indistinguishable from that occurring in matted hairs, and in a few of the normal hairs previously examined. This indicated that the normal fluorescence of the medulla when treated with acridine orange was red, but normal hair cuticle prevented the entrance of the fluorochrome. In the matted hair, and in cut normal hair, the fluorochrome penetrated into the medulla and caused it to fluoresce red. The cut tips of matted hairs also fluoresced in the same manner as normal hairs.

These findings led us to the conclusion that in matted hair the cuticle was defective. There was no evidence of any abnormality of the medulla, as the medullary fluorescence with acridine orange was the same in matted as in normal hairs. The abnormality occurred in patches in the cuticle, and these areas were shown by the fluorochrome entering the adjacent medulla and causing it to fluoresce.

The fluorochrome technique was repeated with rhodamine B and thioflavine T. It was seen that when cut normal hairs were treated with rhodamine B they fluoresced for a greater distance from the cut end than with acridine orange (Plate, fig. D). The colour of this fluorescence was salmon pink. It appeared that there was a better diffusion of this fluorochrome along the medulla. This was confirmed by treating matted hairs with rhodamine B; the patches of medullary fluorescence now extended up and down the hair shaft, and these often coalesced, causing the whole length of the medulla to fluoresce (Plate, fig. C). This increased penetration could be seen on macroscopical examination by ordinary light. Matted hairs treated with rhodamine B were stained a magenta colour, whereas normal uncut hairs retained their usual coloration. Thioflavine T had a similar degree of penetration to acridine orange; with this fluorochrome the medulla of matted hairs underlying the areas of defective cuticle fluoresced yellow. The cut ends of normal hairs and of cut matted hairs gave the same yellow fluorescence in the medulla.

These findings confirmed that in normal hair the cuticle prevents the medullary fluorescence from occurring unless it is artificially damaged as by cutting the hairs. The defective cuticle in matted hairs was also demonstrated by staining for 3 minutes with 01 aqueous solution of methylene blue and mounting in glycerine. The dye penetrated the cuticle of matted hairs but failed to do so in normal hairs.

The hairs of Naked (N / + ) mice were examined with the same fluorochromes. In these hairs, in contrast to matted, there were areas of obvious anatomical defects. When treated with acridine orange these affected patches showed a red fluorescence of the cortex. The hair-shaft was often enlarged at these points, and the melanin granules were broken up (Plate, fig. E). Thus these hairs could easily be distinguished from matted hairs. The fluorescence with rhodamine B was salmon pink in the defective places, but the diffusion of this fluorochrome was not so good as in matted, and therefore the area of fluorescence were more limited. No alteration of the medullary fluorescence along the hair-shaft was detected by this method. The fluorescence obtained with thioflavine T confirmed the findings with acridine orange. The defective septa fluoresced yellow and green; the normal medullary fluorescence with this fluorochrome is yellow.

The results obtained after treatment of plucked hairs with various chemicals are given in Table 1. Calcium thioglycollate which reduces S-S linkages (Goddard & Michaelis, 1934) increases normal cuticular permeability, and subsequent fluorescence with fluorochromes is similar to that of matted hairs. Hydrogen peroxide gives a quite different fluorescence from matted hairs, although treated fibres are similarly brittle and inflexible. Other reagents alter the normal hair fluorescence but have to be used in such strong concentration that fibre structure is grossly altered. It is possible that a deficiency of S-S linkages exists in the cuticle of affected areas of matted hairs.

Sinus hairs of matted mice did not show an abnormal fluorescence as did the body-hairs. They gave the same green fluorescence as normal sinus-hairs. It was found, however, that when both normal and matted sinus-hairs were cut, the medulla did not fluoresce red when treated with acridine orange. If therefore there was a defect in the cuticle of these hairs the penetration of the fluorochrome into the medulla would not be shown by this method.

Skin from the back of normal and matted mice was fixed in 70 per cent, alcohol. Sections were cut and double-stained with acridine orange and primulin by the method described by Jarrett et al. (1956). There was no difference between the epidermal keratin of normal mouse skin and that of matted skin; in both the colour fluorescence was orange.

No differences were noted between normal and matted hairs when these were examined by phase contrast microscopy and polariscopy. This suggests that the molecular arrangement in matted fibres is normal. Professor W. T. Astbury kindly examined the X-ray diffraction patterns of matted and normal mousehair. He found both molecular pictures to be of the a keratin type (Astbury, 1950), and there was no detectable alteration of the keratin in matted hairs. The chain molecular orientation in both was better than in most wool or human hair; this may lead to an increased tendency to longitudinal splitting in mouse hairs.

Hairs were examined for SH groups by the ‘nitroprusside test’ (Glick, 1949), but no free SH groups were demonstrated in either matted or normal fibres.

Quantitative estimations of total cystine were made in shaved samples of matted and normal mouse-hair by the colorimetric method developed by Lugg (1932 a, b) and by Shinohara (1937). The hair was degreased, washed, and dried: weighed amounts were hydrolysed by boiling 5 N hydrochloric acid. The cystine S-S was then reduced by sodium bisulphite to thiol SH, and these groups were estimated by the blue colour produced with active phosphotungstic acid reagent. (Some of these analyses were kindly performed by Dr. S. Blackburn.) Normal mouse-hair from the back contains about 2-4 per cent, cystine, but no appreciable thiol S; matted hair does not show any significant difference from the normal in composition.

Although the total cystine content is similar in matted and normal hairs, this does not exclude the possibility of an altered cystine distribution. A relative deficiency in the cuticle, which may render this structure more permeable, could be compensated by an increased amount in the medulla or in other areas of cuticle and cortex. The cystine content in wool-fibres is known to vary along then-length, and the medulla is normally low in cystine (Speakman, 1955). An abnormal distribution, with a cystine deficiency in the cuticle in matted hairs, could on mechanical grounds explain their brittleness.

The amino-acid composition of matted and normal mouse-hair was examined by Dr. S. Blackburn. One-dimensional phenol-ammonia and two-dimensional collidine-phenol-ammonia chromatograms of keratin hydrolysates were prepared. No difference was found between the keratin composition of normal and matted hairs. The N-terminal amino-acids were not examined.

The effects of concentrated sodium hydroxide and hydrochloric acid on normal and matted hairs was examined. Both were less resistant to sodium hydroxide than to hydrochloric acid, but there was no appreciable difference between these hairs in their resistance to sodium hydroxide.

No glucose was detected in the urine of matted mice when tested with Benedict’s reagent. Paper chromatograms of the urinary amino-acids of agouti matted and normal mice were prepared by the method of Harris & Searle (1953); in both, taurine was the only amino-acid detected. There was no evidence of cystinuria.

Thyroids, adrenals, and livers from normal and matted mice were examined histologically after staining with haematoxylin and eosin. No abnormalities were seen.

The hair-cycle on the back was studied in normal and matted mice after lightly clipping the coat. There was no significant difference in the lengths of the first and second hair generations in these two groups; the third generation, however, was significantly shorter in matted mice (Table 2). This difference is possibly due to environmental factors rather than genetically determined. It is known that the mouse-hair-cycle is influenced by mechanical irritation (Borum, 1954). Owing to increased hair-loss in matted mice, the skin is less well protected against friction, and consequently this stimulation would result in a more rapid hair regeneration. This would explain the normal growth for the first two generations before the hair-loss is very great, while the third and subsequent regenerations were more rapid than in normal animals.

1. The keratin abnormalities of the hair in matted and Naked (N / +) mice have been studied by fluorescence microscopy. Numerous areas of red fluorescence were noted in the medulla of intact matted hairs treated with acridine orange. This appearance differed from that of normal mouse-hairs, which showed only very occasional patches of red fluorescence except at their cut ends. The defect of matted hairs is thought to be in the cuticle; where this structure is permeable it allows the fluorochrome to penetrate into the medulla. Confirmatory results were obtained with other fluorochromes and methylene blue.

2. The permeability of normal mouse-hair-cuticle was increased greatly when the keratin was reduced with calcium thioglycollate.

3. Naked hairs unlike matted hairs showed gross anatomical defects of the cross septa, and with acridine orange these fluoresced red in some areas and green in others.

4. Other histological methods failed to reveal any difference between matted and normal hair and skin.

5. Chemical and physical analyses of matted hairs showed no significant difference from normal; neither was there any evidence of abnormal aminoaciduria. It is suggested that an abnormal distribution of S-S linkages might upset the mechanical properties of matted hairs; the total composition remaining normal.

6. The hair-cycle of matted mice is described; this was found to be much shorter in the later hair generations. The cause of this more rapid regeneration is discussed.

We are grateful to Dr. W. N. Goldsmith and Professor H. Grüneberg, F.R.S., for the interest they have shown in this work. We are greatly indebted to Professor W. T. Astbury, F.R.S., of the University of Leeds, for the X-ray analyses, and to Drs. L. Auber and S. Blackburn of the Wool Industries Research Association for suggestions on histology and for certain chemical analyses. We are grateful to Professor L. S. Penrose, F.R.S., for advice on paper chromatography of urinary amino-acids, and to Dr. C. A. B. Smith for help with statistics.

The Naked mice were kindly supplied by Dr. T. C. Carter of the Medical Research Council unit, Harwell.

Mrs. J. A. Hardy prepared the specimens for fluorescence microscopy, and Mr. A. Bligh helped in the production of the photo-fluoromicrographs.

This work was supported by a grant from the Rockefeller Foundation.