Hereditary hypertrophic neuropathy in the Trembler mouse

Journal of the Neurological Sciences, 1976, 30:343-368 © Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

HEREDITARY MOUSE

HYPERTROPHIC

NEUROPATHY

IN

THE

343

TREMBLER

Part 2. Histopathological Studies: Electron Microscopy

P. A. LOW Department of Medicine, University of Sydney, Sydney, N.S.W., 2006 (Australia) (Received 28 April, 1976)

SUMMARY The Trembler mouse suffers from a dominantly inherited hypertrophic neuropathy. Electron microscopy, including a quantitative analysis of myelination was performed on the nerves of Trembler mice from birth to senility and compared with the findings in control mice. Axons in adult Trembler nerves were thinly myelinated and were surrounded by very few myelin lamellae which in turn were often uncompact circumferentially and longitudinally. Schwann cell cytoplasm was copious and had a normal content of organelles. Well-developed "onion-bulb" fbrmations which consisted of thinly myelinated axons surrounded by empty membrane configurations were frequently seen. The initiation of myelination was studied. The diameter distribution of promyelin fibres of control and Trembler sciatic nerve at ages day 2, 4, and 7 was calculated Myelination in Trembler nerves commenced on axons of larger diameters than controis. The effectiveness of myelination was studied by relating the number of turns of myelin to the axon area of control and Trembler sciatic nerves from age 2 days to adult mice. At all ages Trembler axons were less well myelinated than controls and the difference was more pronounced with age. Schwann cell activity was examined by relating the area of the Schwann cell cytoplasm to the area of the axon it invests. The relative amount of Schwann cell cytoplasm decreased progressively in control axons with age and as the axon became better myelinated. By contrast, that of Tremblers did not undergo a similar reduction as the animal matured and the relative amount of Schwann cell cytoplasm was markedly increased in adult Tremblers when compared with controls. This work was supported by grants from The National Health and Medical Research Council of Australia, and from the Postgraduate Medical Foundation, University of Sydney.

344 The periodicity of control and Trembler compact myelin was compared. The myelin period of Trembler mouse was significantly greater than that of controls. The defect in Trembler peripheral nerves was considered to be that of dysmyelinogenesis. The Schwann cell was active but ineffective in the synthesis, compaction and maintenance of myelin.

INTRODUCTION The Trembler mouse suffers from a dominantly inherited hypertrophic neuropathy. The pathological features of thinly-myelinated axons surrounded by empty membrane configurations (Ayers and Anderson 1973; Low and McLeod 1975)and slowing of conduction to le~s than 10 m/sec (Low and McLeod 1975) are features which are also seen in Dejerine-Sottas neuropathy. An electron microscopic study of the pathological changes from birth to adulthood, including a quantitative analysis of myelination has been performed in order to gain insight into the evolution of the neuropathy in the Trembler mouse and similar neuropathies in man. MATERIALS AND METHODS

Animals Twenty-four control mice from birth to 15 months, and 24 Trembler mice from birth to 27 months were used for the histopathological studies. Immature Trembler mice were identified by the use of the Trember-Rex linkage (Low 1976a).

Histological techniques Fixation, dehydration and embedding were performed as described (Low 1976). Ultra-thin sections of about 50 nm were cut with a diamond knife, stained on copper grids with uranyl acetate (Watson 1958) and lead citrate (Reynolds 1963), and examined in a Philips EM 200 or 201 electron microscope.

Quantitative studies (1) Promyelin fibres These were measured on control and Trembler mice of ages, 2, 4 and 7 days. A promyelin fibre is an unmyelinated fibre which has intimate contact over the whole of its circumference with one Schwann cell, the cytoplasm of which contains abundant organelles (Friede and Samorajski 1968). These axons are about to become myelinated (Friede and Samorajski 1968; Fraher 1972). The fibre was not accepted as a promyelin fibre unless the Schwann cell contained only one axon. Contiguous nonoverlapping sections were photographed and for each nerve a minimum of 0.011 mm z was assessed. The fibres were counted and their diameters measured at a final magnification of x 7350, with a Zeiss T G Z z Particle Size Analyser, set in the linear mode. Fibres of diameter 0.2 #m or less were grouped together and larger fibres subdivided into 0.2-#m groups.

345

(2) Myelin lamellae: axon area (M/A ratio) In normal mammalian nerve fibres there is a linear relationship between the size of the axon and the number of turns of myelin (Friede and Samorajski 1968; Dyck, Lambert, Sanders and O'Brien 1971; Sharma and Thomas 1974). The M/A ratio is reduced in primary disorders of Schwann cell (Samorajski, Friede and Reimer 1970; Dyck et al. 1971), and increased in axonal degeneration (Friede and Martinez 1970; Dyck et al. 1971). This ratio was studied from day 2 to adulthood in control and Trembler mice in order to provide a quantitative analysis of myelin formation. The axon with its surrounding Schwann cell was photographed at x 8700 and printed at a final magnification of x 26,100. The choice of axons to be photographed was random but care was taken to ensure that the fibre was technically suitable for analysis (Fraher 1972) and that the axon was round or oval. Elongated fibres or markedly irregular fibres were excluded. The axon was measured by a planimeter and the number of major dense lines was counted under magnification with a hand lens (Friede and Samorajski 1968; Dyck, Lais and Offord 1974). In fibres where myelin was not compact circumferentially, the lamellae were counted only if they were compact over at least 75 ~ of the axon circumference. Myelin lamellae with a lesser degree of compaction were not counted. Only axons containing more than 3 Schwann cell spirals were included in the analysis. The M/A ratio of control and Trembler sciatic nerves was calculated at days 2, 4, 7, 14 and adulthood and in 1 Trembler nerve at 27 months. At least 100 fibres were measured for each age.

(3) Cytoplasm + axoplasm: axoplasm (C/A) ratio During the period of most active myelinogenesis a consistent feature is the increase in Schwann cell cytoplasm so that the C/A ratio is high. The ratio decreases progressively as myelination proceeds towards completion and provides a good index of Schwann cell activity (Friede and Samorajski 1967, 1968). The ratio was determined in control and Trembler mice from infancy to adulthood in order to examine quantitatively the activity of Schwann cells in Trembler mice relative to those in controls. The C/A ratio was determined, using the same photographic prints that were used for the M/A ratio, in control and Trembler sciatic nerves at ages day 4, 7, 14 and adulthood and in 1 Trembler nerve at 27 months. At least 100 fibres were measured for each age. The cross-sectional area of Schwann cell cytoplasm was measured with a planimeter. In developing fibres, where cytoplasm was abundant, it was divided into 2--4 approximately equal segments by means of a marking pen. These areas were then measured, using a planimeter, and the values were totalled. Where a nucleus was present, it was also measured and its area subtracted from the total area. In adult control nerves, cytoplasm was scanty and often formed a discontinuous rim around the myelin. These islands of cytoplasm were measured individually with a planimeter and totalled. The ratio was expressed as a percentage.

(4) Measurement of myelin period Measurement of myelin period was performed on the sciatic nerves of 1 adult

346 control and 1 age-matched Trembler mouse. Shrinkage occurs in fixed nerve, and as this has been shown to occur mainly in the dehydration stage (Finean, Sj6strand and Steinmann 1953), both nerves were processed together. The grids were photographed at the same session and filament voltage and current, condenser lens current, apertures, and magnification were held constant. Care was taken to measure the periodicity of the myelin lamellae in regions of the sheath which were free from compression artefact due to cutting (Kaarlson 1966). The grids were photographed at × 20,000 and printed at a final magnification of × 60,000. A diffraction grating was photographed and submitted to the same developing, enlarging and printing processes. Trembler axons were not infrequently surrounded by Schwann cell spirals. These fibres were not photographed so that only Trembler fibres with compact myelin were included in the analysis. The distance between the inner and outer major dense lines was measured and the number of myelin periods for each axon was counted under magnification with the aid of a hand lens and dividers. The mean myelin period was then calculated by dividing the total number of myelin periods by the distance. Twenty control and 23 Trembler myelinated fibres were measured in this way. The mean and standard deviation was calculated and the two groups were compared statistically using the Student t-test.

(5) Unmyelinated fibre density and distribution Unmyelinated fibres were counted in 5 control and 5 Trembler adult nerves and their diameters were measured with a Zeiss TGZ3 Particle Size Analyser set in the linear mode on prints at a final magnification of x 7350. Contiguous non-overlapping sections were photographed and fibres were measured and counted in an area of at least 0.011 mm 2 for each nerve. Over 360 fibres were counted for each nerve. Unmyelinated fibres were identified by the usually round or oval contour, the presence of a mesaxon and the prominence of neurotubules, and the increased density of axolemma when compared with Schwann cell membrane. Fibres of diameter 0.2/~m or less were grouped together and larger fibres subdivided in 0.2 #m groups (Ochoa and Mair 1969). The exact final enlargement was determined by photographing a ruled diffraction grating and subjecting the plate to the same processes of developing and printing.

(6) Schwann cell and fibroblast nuclei The numbers of Schwann cell and fibroblast nuclei were counted and their densities calculated on the same prints as those used for measurement of unmyelinated fibres. A cell was counted if at least half its nucleus appeared to be within the print under assessment. Schwann cells were clearly identifiable by virtue of their relationship to myelinated and unmyelinated axons, and by the presence of a basement membrane (Thomas 1963; Gamble and Eames 1964) and 10 nm filaments within their cytoplasm (Elfvin 1961). Fibroblasts, by contrast, do not contain basement membrane, but contain prominent endoplasmic reticulum and possess elongated processes (Thomas 1963).

347 RESULTS

Myelinogenesis in control mice Fourteen immature control mice were studied. Their ages ranged from 0 to 21 days. Myelinogenesis has been well described previously and the present findings confirm the observations of other workers (see reviews, Webster 1975; Bischoff and Thomas 1975). At day 2 there were large clusters of small calibre unmyelinated fibres not individually enclosed in Schwann cell cytoplasm. Occasional axons at the periphery were enlarged and enveloped in a tongue of Schwann cell cytoplasm. Numerous promyelin fibres were seen. These were larger unmyelinated axons which were surrounded in abundant Schwann cell cytoplasm, which in turn was active, containing numerous polyribosomes, rough endoplasmic reticulum and mitochondria. One or two spirals of Schwann cell processes were often seen. A few axons contained 3 spirals and compaction usually commenced on these so that axons encircled by Schwann cell spirals of 4-5 layers were rarely seen. The myelin seen at this age was compact but less so than that of adult mice, so that the myelin period was longer and the interperiod line not discernible. Axons contained few microtubules and neurofilaments. The changes in day 4, 7 and 14 controls were those of progressive maturation. At day 7, myelination was well advanced. There were fewer fetal fibres and also they were less numerous than in younger mice. By day 14 there were few fetal fibres and promyelin fibres, and myelinating axons were of larger diameter, and had up to 70-80 lamellae in the largest axons. By this stage, the myelin period did not differ from that of adult mice. Schwann cell cytoplasm was abundant in myelinating fibres. The cytoplasm here tended to be dense, contained numerous mitochondria, polyribosomes, rough endoplasmic reticulum, which was often dilated, and in addition contained Golgi complexes. By 2 weeks the relative volume of cytoplasm became less, tending towards adult proportions. Myelinogenesis in Trembler mice Fifteen Trembler-Rex mice of ages 0-23 days were used. Ten of these were 7 days or younger and were identified by the curliness of their whiskers. The 5 older mice were clinically recognizable as Trembler mice. Using this method there is a 12-15 ~ error when a mouse with Rex characteristics might not be a Trembler. It was observed that the sciatic nerve of Trembler mice was able to be identified from the earliest age by the presence of large uncompact spirals. There were other characteristics such as the lesser degree of myelination and larger promyelin fibres but these features were less helpful in a qualitative examination. One of the 10 Trembler-Rex mice which did not have large uncompact spirals was assumed to be a Rex but not a Rex-Trembler, and was excluded from the study. Schwann cell proliferation and its segregation of fetal fibres into promyelin fibres appeared normal in Trembler sciatic nerve. Promyelin fibres appeared to be of larger diameter than those in control mice (Fig. 1). At all ages axons were poorly myelinated and 3-5 ~ of fibres were undergoing myelin degeneration between days 7

348

Fig. 1. Sciatic nerve of a day 7 Trembler mouse. Myelinating fibres (Mf) have reduced myelin thickness. Uncompact spirals (arrows), large promyelin fibres (P) and myelin debris (d) are seen, Scale 1 pro.

and 4 weeks (Fig. 1). Fibroblasts were present in normal numbers and collagen fibrils did not appear to be present in increased numbers. Adult mice

The findings of thinly myelinated axons surrounded by empty membrane configurations and the presence of Schwann cell spirals have been described in previous publications (Ayers and Anderson 1973; Low and McLeod 1975). The genesis of "onion bulb" configurations has been described elsewhere (Low 1976b). Only additional findings will be described here. Compact myelin in Trembler mice was usually less compact than myelin in control nerves and the minor dense line was usually not seen (Figs. 2 and 3). Schwann cell cytoplasm was abundant and contained copious free and bound polyribosomes, Golgi complexes and mitochondria, which were often hypertrophied (Fig. 3). From the age of 3 to 6 months there was a

349

Fig. 2. Three months old Trembler sciatic nerve. Myelin is non-compact over about 30 ~ of the circumference of the axon. Arrow indicates an island of microfibrils. Scale 0.25 #m.

progressive increase in the density of neurofilaments (Fig. 4) and axons assumed bizarre contours (Fig. 5). Collagen fibrils varied in calibre between 15 and 80 nm and there were frequent islands of microfibrils of approximately 10 nm diameter (Fig. 4). On longitudinal section, areas of uncompact myelin were frequently seen (Fig. 6). Long lengths of axon were denuded of myelin. Indeed, normal internodes were never seen. Nodes of Ranvier were present but no completely normal nodes were seen. The axon component of the node appeared normal. Well marked paranodal narrowing was present and the characteristic electron-dense segment of nodal membrane was seen (Fig. 7). However, the myelin sheath usually contained few lamellae and one or other internode bordering the node of Ranvier was usually denuded (Fig. 7). N o t infrequently, Schwann cell nuclei were situated within the node of Ranvier (Fig. 8).

350

Fig. 3. Three months old Trembler sciatic nerve. Schwann cell cytoplasm is copious; numerous free and bound polyribosomes and Golgi complexes (g) are seen. Mitochondria (m) appear hypertrophied. Arrow indicates a pi body. Scale 0.5/~m.

Quantitative studies (1) The initiation of myelination Promyelin fibres are those which are about to myelinate (Friede and Samorajski 1968; Fraher 1972). The densities and diameter distribution are shown in Fig. 9 and Table 1. The density of promyelin fibres was increased in Trembler mice at all ages. The diameter distribution of control promyelin fibres is unimodal with a peak of 0.8-1.0/~m. By contrast, the peak, in the case of Trembler mice, tended to be more dispersed with a suggestion of bimodality (Fig. 9). These findings confirm the qualitative observation that there was an increased number of very large promyelin fibres.

(2) Myelin lamellae: axon area (M/A) ratio The ratio was assessed from the onset of myelination into the aged state in order to gain insight into the mechanism of both synthesis and maintenance of myelin. At day 2, the number of myelin lamellae ranged from 0 to 20 (mean 7.1; SD, 4.2) for control mice, and 0 to 12 (mean 3.9; SD, 2.5) for Trembler mice (Table 2). The difference was highly significant (P < 0.001, Student's t-test). When the number of

351

Fig. 4. Twenty-sevenmonths old Trembler sciatic nerve. There is a marked increase in neurofilaments. Neurotubules are reduced and clumped. Small arrows indicate empty basement membranes and large arrow indicates microfibrils. Scale 0.5/~m. myelin lamellae was related to axon area, no significant regression line was obtained for either Tremblers or control fibres (Table 3). At days 4, 7, 14, adult (non aged) and aged (27 months) control and Trembler mice (Table 2), the axons of Trembler mice had fewer turns of myelin than control mice at each age and the difference became more pronounced with age. The slope of the regression line was flatter in Trembler mice than controls and the difference was

352

Fig. 5. Twenty-sevenmonths old Trembler sciatic nerve. The axon is denuded and has bizarre configurations, x 7400. accentuated with age (Fig. 10, Table 3). There was a negative regression coefficient for aged Trembler axons, which indicates that the larger axons have fewer myelin lamellae than smaller axons.

(3) Cytoplasm: cytoplasm + axoplasm (C/A) ratio Day 4 (Fig. 11). The C/A ratio ranged from 29 to 9 2 ~ (mean, 65.6; SD, 13.6) for control mice, and from 31 to 9 2 ~ (mean, 67.4; SD, 14.2) for Trembler mice. The difference was not significant (P < 0.05). Day 7 (Fig. 11). The C/A ratio ranged from 23 to 8 8 ~ (mean, 49.3; SD, 16.8) for control mice, and from 16 to 95 ~ (mean, 67.2; SD, 18.0) for Trembler mice. The difference was highly significant (P < 0.001, Student's t-test). Day 14 (Fig. ll). The C/A ratio ranged from 12 to 8 6 ~ (mean, 40.6; SD, 16.6) for control mice, and from 31 to 93 ~ (mean, 63.1 ; SD, 16.6) for Trembler mice. The difference was highly significant (P < 0.001, Student's t-test). Adult mice (Fig. 11). The C/A ratio ranged from 2 to 65 ~ (mean, 15.9; SD, 11.6) for control mice. By contrast, that of Trembler mice ranged from 5 to 84 ~ (mean, 47.9; SD, 18.0). The difference was highly significant (P < 0.001, Student's t-test). Comparison of C/A ratio of adult and aged Trembler (,Fig. 11). The C/A ratio of adult Trembler sciatic nerve ranged from 5 to 84 ~ (mean, 47.9; SD, 18.0) and that of an aged Trembler sciatic nerve ranged from 5 to 7 6 ~ (mean, 29.9; SD, 16.2). The difference was highly significant (P < 0.001, Student's t-test).

353

Fig. 6. Longitudinal section of a 3 months old Trembler sciatic nerve. The lower photomicrograph is an enlargement of the area indicated above. Compact myelin (My) merges abruptly with uncompact myelin which contains Schwann cell cytoplasm (cyt.). Ax. represents axon. Scale 5/~m and 1/~m for upper and lower photomicrographs respectively.

(4) The axon area The axon area of c o n t r o l sciatic nerve was c o m p a r e d with that of T r e m b l e r mice for a n i m a l s f r o m day 2 to a d u l t h o o d (Fig. 12).

354

Fig. 7. Longitudinal section of a 12 months old Trembler sciatic nerve. Nodal narrowing, terminal loops (TL) mitochondria (m) and specialised axolemmal dense segments (arrow) are seen, but to the right of that (arrow head), the axon has been denuded of myelin, × 19,600.

The axon area at day 2 was significantly greater (P < 0.001) in Trembler mice (mean 2.1/~m2; SD, 1.2) than control mice (mean, 1.2 #m2; SD, 0.7). At day 4 the axon area was again significantly greater (P < 0.001) in Trembler sciatic nerve (mean, 2.1 /tmZ; SD, 0.9) when compared with that in control mice (mean, 1.2 #m2; SD, 0.4). By day 7 there was no significant difference between control (mean, 2.0,/~m2; SD, 1.2) and Trembler (mean, 2.2/~mZ; SD, 1.4) axon areas. The axon area at day 14 was significantly greater (P < 0.001) for control mice (mean, 3.5/~m2; SD, 2.7) than for Trembler mice (mean, 2.6/~m2; SD, 1.8). The axon area of adult control sciatic nerves (mean, 9.0 #m2; SD, 10.2) was greater than that of Trembler mice (mean, 5.1 # m 2 SD, 5.5). The difference is significant, P < 0.001. Neither control nor Trembler mice were older than 12 months of age.

(5) The myelin period (Fig. 13) The myelin period was measured in the sciatic nerve of control and of an agematched Trembler mouse aged 3 months (see Methods). Twenty control and 23 Trembler fibres were measured. The myelin period in Trembler fibres (mean, 19.4 nm; SD, 49.6) was significantly greater (P < 0.001) than that in control fibres (mean, 13.7 nm; SD, 5.4).

(6) Unmyelinatedfibre density and distribution (Fig. 14, Table 4) Unmyelinated fibres were identified, measured and counted using the criteria described previously (see Methods).

355

Fig. 8. Supernumerary Schwann cell nucleus (N) situated in the node of Ranvier. The bottom photomicrograph is an enlargement of the area indicated above. Ax, axon; arrow, terminal loop. Upper photomicrograph, x 6000; lower photomicrograph, x 21,300.

356 clay 2

day 7

day 4

"IREMBLER

a. 2o

DIAMETER

Qum)

Fig. 9. Percent distribution of diameters of control and Trembler promyelin fibres at different ages.

The unmyelinated fibre density in control sciatic nerve ranged from 34.1 × 103 to 52.3 × 103 fibres/mm 2 (mean, 41.5 × 103; SD, 7.2 × 103); the density in Trembler sciatic nerve ranged from 34.8 × 103 to 69.4 × 103 fibres/mm 2 (mean, 51.5 × 103; SD, 12.9 × 103). The difference was not significantly different (P ~ 0.05). In no case did the density in Trembler nerve fall below the control range. The unmyelinated fibre diameter distribution in Trembler nerves did not differ from that of control nerves. In all fibres there was a unimodal distribution from 0.4 to 0.8/~m (Fig. 14).

(7) Schwann cell density (Table 4) Schwann cells were identified according to the criteria described (see Methods). The density (cells/ram 2) in sciatic nerves of control nerves ranged from 818 to 1727 (mean, 1109; SD, 487). By contrast, the density in adult Trembler sciatic nerves ranged from 8778 to 10,454 (mean, 9238; SD, 727), and in all cases was well above the control range. The mean cell density in sciatic nerves of adult Trembler mice was × 8.7 that of adult control nerves and was of a similar order to that in control and Trembler sciatic nerves during the most active period of myelinogenesis (Table 1). (8) Fibroblast density (Table 4) Criteria for the identification of fibroblasts have been described (see Methods). The density (cells/mm 2) in adult control nerves ranged from 90 to 450 (mean, 234; SD, 164). By contrast, the density in adult Trembler sciatic nerves ranged from 722 to 1636 cells/mm 2 (mean, 1108; SD, 448) and in all cases it was well above the control range. The mean fibroblast density in the sciatic nerves of Trembler mice was × 4.7 that of adult control nerves. The densities (cells/mm 2) of fibroblasts in control sciatic nerves of ages 2, 4 and 7 days were similar to those in age-matched Trembler mice (Table 1). These densities in immature control and Trembler mice were in turn similar to these of adult controls

C52 C54 C52 T63 T57 T37

2 4 7 2 4 7

Animal Age (days)

8

2 6 10

0.6

31 57 45 7 5 23

0.8

45 33 26 24 35 22

1.0

Fibre diameter (pm)

14 3 17 15 36 12

1.2

5 1 2 22 18 18

1.4

22 6 11

3

1.6

5

9

> 1.8

0.011 0.011 0.011 0.011 0.011 0.011

Area (mm e)

290 254 89 368 318 207

No. of promyelin fibres

27338 23944 8390 34691 29977 18818

promyelin fibres

15273 13544 9090 11363 11544 14998

Schwann cell

Density (no./mm 2)

PROMYELIN FIBRE DENSITY A N D P E R C E N T A G E DISTRIBUTION IN CONTROL A N D T R E M B L E R MICE

TABLE 1

364 454 181 181 364 273

fibroblast

"....I

358 TABLE 2 COMPARISON OF MYELIN LAMELLAE: AXON AREA (M/A RATIO) IN CONTROL AND TREMBLER MICE Age

Day 2 Day 4 Day 7 Day 14 Adult Aged

Number of measurements Range

Mean ± standard deviation

control

Trembler

control

Trembler

control

140 127 119 169 330

100 102 102 186 335 103

0- 20 0- 30 0- 50 9- 73 10-110

0-12 0-21 0-25 0-32 0~5 0-45

7.1 ± 14.8 ± 24.8 ± 35.6 ± 50.9 ±

Trembler 4.2 7.5 11.7 15.3 25.4

3.9 ± 2.5 8.1 ± 5.1 9.6 ± 6.9 9.0 ~ 9.2 5.0 ± 7.1 5.3 ~:~8.7

and confirm the qualitative finding that fibroblast proliferation was not a feature in the sciatic nerves of immature Trembler mice. DISCUSSION The fine structural findings in the present study of reduced myelin thickness, Schwann cell spirals, uncompact myelin both circumferentially and longitudinally, and the presence of concentric arrays of Schwann cell processes or basement membranes confirm the prior observations of Ayers and Anderson (1973). Ayers and Anderson (1973, 1975) described the histopathological abnormalities in the peripheral nerves of immature offspring of heterozygous Trembler females which had been mated with control mice. They considered that the segregation and maturation of axons to the promyelin fibre stage was delayed. This observation was not confirmed in the present study. The density of Schwann cell nuclei in the first week of life did not differ from that of control mice and the density of promyelin fibres was not reduced (Fig. 1). Ayers and Anderson's (1975) conclusion is also inconsistent with their observation that there were no fine structural differences between the sciatic nerves of their control mice of days 1 and 2, and those of 13 age-matched offspring of a Trembler female. This is the period in mice when the mitosis of Schwann cells and segregation of axons is maximal (Asbury 1967; Friede and Samorajski 1968). Quantitative studies were not performed. Duncan's (1934) original concept of a critical diameter for myelination to commence does not seem tenable since myelination has been shown to commence on axons which range in diameter from 0.9 to 3.2 ~m (Ochoa 1971). However, when diameter distribution histograms of promyelin fibres were constructed, a unimodal distribution was consistently found in the present study with a peak of approximately 0.8-1 # m ; this finding confirms the observations of other workers (Friede and Samorajski 1968; Fraher 1972). Whether it is axon size per se, or whether it is some other associated signal that initiates myelination, myelination commences in the majority of control sciatic axons when their diameter is 0.8-1 #m. By contrast, in the case of Trembler mice, most axons do not commence myelination until their fibres

l~ay2 Day4 Day 7 D a y 14 Adult Aged

Age

140 127 119 169 330

control

100 102 102 186 335 103

Trembler

Y = 7.15 + 0 . 7 1 X Y= 7.36+6.47X Y = 15.78 + 4.47 X Y = 21.29 + 4.12 X Y = 33.50 + 1.95 X

control

Y = 3.49 + 0 . 2 1 X Y=5.43+l.29X Y -- 6.80 + 1.26 X Y = 4.01 + 1.91 X Y = 4.06 + 0.19 X Y = 6.71 - - 0 . 2 0 X

Trembler

0.12 0.38 0.46 0.74 0.78

control

Correlation coefficient

0.10 0.23 0.25 0.38 0.15 0.15

MICE

NS P<0.001 P < 0.001 P < 0.001 P < 0.001

control

NS P<0.05 P < 0.05 P < 0.001 P < 0.001 P < 0.05

Trembler

Significance o f slope

AND TREMBLER

Trembler

AXON AREA (M/A RATIO) IN CONTROL

Regression e q u a t i o n

OF MYELIN LAMELLAE:

Number of measurements

COMPARISON

TABLE 3

P P P P

< < < <

0.001 0.001 0.001 0.001

control

Trembler

Comparison of c f slopes

4 3~ 4 6 3

14 12 5 4 2

C 13 C 16 C 14 C43 C44

T 15 T 17 T 19 T 22 T 26

Animal Age (months)

2 10 2 3 2

14 3 4

5

0.2

41 29 50 21 44

34 32 47 54 51

0.4

25 35 27 25 42

21 40 20 30 32

0.6

Fibre diameter (#m)

21 27 14 22 10

24 22 10 9 12

0.8

6 9 6 13 2

9 4 5 2 2

1.0

2

6

8 1

1

2

1.4

2 1

2 1

3

1.2

3

2

2

1.6

1

1.8

0.011 0.018 0.013 0.011 0.011

0.011 0.011 0.011 0.011 0.021

Area (mm 2)

476 977 920 568 369

555 477 362 395 827

Mean

Mean

No. of unmyelinated fibres

1545 727 1727 727 818 1109 10272 8778 9090 9544 10454 9238

44900 55300 69400 53500 34800 51500

Schwann cell

52300 45000 34100 37200 39000 41520

unmyelinated fibres

Density (no./mm 2)

727 722 1545 909 1636 1108

90 180 450 90 360 234

fibroblast

SCIATIC NERVE U N M Y E L I N A T E D FIBRE DENSITY AND PERCENT DISTRIBUTION IN A D U L T CONTROL (C) AND T R E M B L E R (T) MICE

TABLE 4

o

361 100 •

I

d

u

l

t

..m :£

<

60 •

Z .i u/ ).

1

4

:E

20

I "14

/.'.-'"

~ ~. i i .i .~ .4

......... adult _ .- Idllt

i

I

lO

20

AXON

I

I

30 AREA

40

o

(pro2)

Fig. 10. The number of myelin lamellae related to the axon area of control (continuous lines) and Trembler (discontinuous lines) at different ages. Numbers indicate age in days.

6G

2C <

X <

~.+ O I->. u

60

2G DAY 4

DAY 7

DAY 14

ADULT

AGED

Fig. 11. The mean cytoplasm: cytoplasm -F axoplasm (%) of control and Trembler (stippled) sciatic nerves. Bars represent 2 SD. have attained considerably larger diameters. These findings may be interpreted as indicating a defect of the Schwann cell so that its response to the signal to myelinate is delayed. An alternative interpretation is that the signal tor the Schwann cell to commence myelination is defective, while the signal for axonal enlargement is normal. The number of myelin lamellae is related to axon circumference or area, and has provided a more accurate assessment of myelin :axon ratio than has been possible using light microscopy (Friede and Samorajski 1967, 1968; Friede and Martinez 1970; Dyck, Ellefson, Lais, Smith, Taylor and Van Dyke 1970; Dyck et al. 1971;

362

E < .-

_,

z

0

"day 2

day 4

day 7

day 14

+ adult

Fig. 12. The mean a×on area of control and Trembler (stippled) sciatic nerves at different ages. Bars represent 2 SD.

o<

200-'r-

100' Z

Fig. 13. The myelin period of control and Trembler (stippled) sciatic nerve. Bars represent 2 SD.

"'~ ~°'~ .~ ~- 60 20t u.

~

OL

a

"

601

TREMBLER

,.J

z

2O

O,AM.ER

(pro)

Fig. 14. Mean percentage diameter distribution of unmyelinated fibres in control and Trembler sciatic nerves. Bars represent 2 SD.

363 Fraher 1972; Dyck et al. 1974; Sharma and Thomas 1974). The close correlation between the number of myelin lamellae and axon area in nerves of adult mice and rats (Friede and Samorajski 1967, 1968; Fraher 1972) indicates that, irrespective of whether it is axon size per se, or whether it is some associated signal, that directs the Schwann cell to lay down the appropriate number of myelin lamellae, the M/A ratio provides a useful index of the effectiveness of myelination. Friede and Samoraj ski (1967) reported a linear relationship between the number of myelin lamellae and axon circumference in the sciatic and vagus nerves of control mice. They observed a similar relationship in the sciatic nerves of rats aged from 1 day to 16 weeks. Fraher (1972) reported similar findings in adult rat sciatic nerves but found that correlation between the number of myelin lamellae and axon circumference was poorer in younger rats and that the regression coefficient was not significant in animals under the age of 4 days. By comparing the ratio of the number of myelin lamellae to axon circumference in the Quaking mouse with that of controls, Samorajski et al. (1970) were able to demonstrate the presence of hypomyelination which had not been noted previously. The use of this parameter has allowed Dyck and coworkers (1971, 1974) to separate hypomyelination (1971) in Dejerine-Sottas neuropathy from secondary demyelination in uraemic neuropathy (1971) and hypertrophic Charcot-Marie-Tooth disease (1974). In the present study M/A ratio in control mice followed a similar trend to that demonstrated in the sciatic nerve of rats (Friede and Samorajski 1968; Fraher 1972) and in mice (Friede and Samorajski 1967). With an increase in age, the slope relating the number of turns of myelin to axon size became flatter, which indicates that, at the earlier stage, myelination was proceeding rapidly relative to axonal growth. This finding confirms the observations of Friede and Samorajski (1968) that myelin lamellae were laid down 4-5 times more ~'apidly at the onset of myelination than at 30-40 days. While there was in control mice a progressive increase in the number of myelin lamellae with age, increasing to 110 lamellae in adulthood, the Trembler mice follow a different trend. At all ages the mean number of myelin lamellae was reduced when compared with controls. The mean number oflamellae increased from 3.9 at day 1, to 8.1 at day 4, 9.6 at day 7, and then became reduced to 9 at day 14. By adulthood the mean number was only 5.0. These findings confirm the qualitative observation that myelin breakdown was prominent between day 7 and 1 month. When the regression lines relating the number of myelin lamellae to axon size are examined it may be seen in Trembler mice that the slope at all ages is significantly flatter than the slopes in control mice and the difference is most marked in adult nerves. This reduced slope provides clear quantitative evidence that hypomyelination was present from birth and represented the main abnormality in the nerves although demyelination contributed at a later stage to the relative reduction in number of myelin lamellae. The regression lines are much more abnormal than those seen in the Quaking mouse (Samorajski et al. 1970) and resemble those in Dejerine-Sottas disease (Dyck et al. 1971). The hypomyelination persisted into senility. The M/A ratio in the

364 aged mouse was less in larger axons than smaller ones, which indicates that, in the aged mouse, the axon-Schwann cell relationship has become so disturbed that even the markedly reduced number of myelin lamellae is no longer maintained. During the initiation of myelinogenesis Schwann cell replication is very active pari passu with the formation of promyelin fibres (Abercrombie and Johnson 1946; Friede and Samorajski 1968). Friede and Samorajski (1967, 1968) found that during the period of most active myelinogenesis Schwann cell hypertrophy occurs at the same time as myelin lamellae are added and axoplasm is increased. These authors reported a C/A ratio of 40-80 % in myelinating fibres whereas the C/A ratio in adult rat and mice was only 20 %. Similar values were obtained in control mice in the present study. The mean C/A ratio in control sciatic myelinating fibres was 65.6 % (SD, 13.6) at day 4, and was only 15.9% (SD, 11.6) by adulthood. The mean C/A ratio in day 4 Trembler mice of 67.4% (SD, 14.2) did not differ significantly from control values. However, in adult Tremblers the mean C/A ratio was still 47.9 % (SD, 18.0). This finding of a persistently high C/A ratio in the adult Trembler indicates that the Schwann cells were very active although they were ineffective in forming myelin. The C/A ratio is reduced and the slope relating the C/A ratio to the axon area was not significantly different in the aged Trembler mouse when compared with that of adult non-aged Tremblers. This reduced Schwann cell activity in response to increasing axon size in the aged Trembler was associated with a corresponding reduced Schwann cell effectiveness in forming myelin which was noted in considering the M/A ratio, when it was found that the slope of the regression line relating the number of turns of myelin to axon area was much flatter than in adult non-aged Tremblers. Indeed, the ~lope had a negative coefficient, which indicated that in those larger diameter fibres which had proportionately less active Schwann cells than smaller diameter fibres, even the small number of myelin lamellae was not maintained. The Schwann cells of Trembler mice persist in a hypertrophied stage throughout life. It does not appear necessary to invoke the postulate of Ayers and Anderson (1975) that the stimulus for Schwann cell hypertrophy and hyperplasia was the ingestion of myelin debris. Schwann cell activity was increased before myelin breakdown occurred, and at day 4 was not different in Tremblers and control mice. The presence of a developing and markedly hypomyelinated axon would appear to be an adequate stimulus for Schwann cell hypertrophy. Schwann cell hyperplasia, however, was present and persisted into adult life. Proliferation of Schwann cells is a consistent feature in any neuropathy where demyelination is extensive. The stimulus to this multiplication is not known (Ballin and Thomas 1969). Remyelination is, in many ways, a recapitulation of early myelinogenesis (Ochoa and Mair 1969) when the stimulus is axonal and Schwann cell replication is most vigorous before myelin appears (Asbury 1967; Friede and Samorajski 1969). More likely possibilities than stimulation by Schwann cell debris are the presence of a bare axon or the loss of contact inhibition discussed by Ballin and Thomas (1969). It appears likely, on present evidence, that the stimulus to Schwann cell division is axonal, but the precise nature of the stimulus is not understood.

365 The myelin period in Trembler mice was significantly greater (P < 0.001) than that of control mice. Minor dense lines were not evident and many areas of lack of compaction were seen both longitudinally and circumferentially. The myelin was morphologically abnormal. Its synthesis was abnormal and it was inadequately compacted. It is likely that demyelination occurs because myelin is morphologically and, presumably, biochemically abnormal. It appears that hypomyelination is the result of dysmyelinogenesis and is reflected in the reduced M/A ratio. It is of critical importance to determine if the primary or predominant abnormality in the Trembler peripheral nerves is that of the Schwann cell or of the neurone. It is possible that the primary defect is neuronal so that the signal(s) to initiate myelination, and to synthesize, compact and maintain the appropiiate number of myelin lamellae for the area of the axis cylinder is defective; however, the bulk of evidence favours a Schwann cell abnormality .The cell bodies (dorsal root ganglia and spinal motor neufones) are present in normal appearance and numbers. The fine structural details of axons appear normal well into adulthood whereas hypomyelination and demyelination commence at an early stage of development. There is a relatively abrupt transition from severely hypomyelinated anterior and posterior roots to well-myelinated central fibres when the same axons pass from Schwann cell domain into oligodendrocyte domain. Finally, the fine structural characteristics of supernumerary Schwann cells provide more direct evidence of a Schwann cell defect. While the central Schwann cell (which has captured the denuded axon) has copious cytoplasm replete with organelles, there are prominent degenerative changes present in the cytoplasm of supernumerary Schwann cells (Low 1976b). Necrotic changes are seen and the cytoplasm recedes to leave empty membrane configurations. The supernumerary Schwann cells, which are not in contact with the axon, perhaps reflect the state of the Schwann cell when it is independent of the influence of the axons. The pathological changes in Trembler peripheral nerve resemble those of Dejerine-Sottas neuropathy. Extrapolation of the findings in Trembler mice to DejerineSottas neuropathy suggests that there may be a primary Schwann cell defect in the latter condition. The findings of Brimijoin, Capek and Dyck (1973), of an abnormality in the axonal transport of dopamine-fl-hydroxylase in adrenergic unmyelinated axons in vitro in Dejerine-Sottas neuropathy, may indicate a concomitant or coincidental defect in the neurone. It is even possible that it reflects a severe Schwann cell abnormality, since similar abnormalities in axoplasmic flow have been reported in the peripheral nerves of dystrophic mice (.Bradley and Jaros 1973; Komiya and Austin 1974; Jablecki and Brimijoin 1974) which has islands of amyelinated axons (Bradley and Jenkison 1973) and in which a Schwann cell abnormality has been reported (Bray and Aguayo 1975). Axons deprived of normal Schwann cell support in Trembler mice undergo marked irregularities in contour from 12 months and there is a progressive increase in neurofilaments. Of further interest is the effect of hypomyelination trom birth on the development of axon size. Dyck et al. (1971) produced quantitative evidence that the circumference of the axis cylinder was considerably reduced in patients with DejerineSottas neuropathy when compared with control subjects. The leduced size of the axis

366 cylinders in Dzjerine-Sottas neuropathy and hypertrophic Charcot-Marie-Tooth disease was thought to be due either to the loss of trophic influence of the Schwann cell as a result of demyelination or to a coincidental metabolic abnormality of the axon (Dyck et al. 1970). In the present study, measurements of axon area were made on large numbers of control and Trembler nerves from day 2 to 27 months, and permit an examination of the mechanism of the reduced axis cylinder size in the Trembler mouse. The findings may be extrapolated to explain the mechanism of hypomyelinated neuropathies like Dejerine-Sottas neuropathy. In the sciatic nerves of Trembler mice at the ages of days 2 and 4 the axon areas were greater than those of controls (P < 0.001, Fig. 12). The difference in axon area was not significant at day 7. Howevec, in older animals (days 14, adult, and aged) the axon area of Trembler sciatic nerves was significantly reduced (P < 0.001) when compared to control mice. The findings that the area of the developing axon was larger in Trembler sciatic nerve than in controls in the first 4 days of myelinogenesis and that after day 7 its increase in axon area lagged behind that of control mice, are best explained by assuming that the increase in axon area during early myelinogenesis was relatively independent of the Schwann cell and that after day 7 there was a greater interplay of influences. There is support for this explanation from the study of M/A ratio in the present study and that of Fraher (1972). In both studies the slope of the regression line relating the number of myelin lamellae to axon size was not statistically significant until the age of 4 days. After this age the slope reached statistical significance and the level of significance increased with age in both studies. There is, then, statistical evidence that before day 4 the increase in axon area and the laying down of myelin behaved relatively independently. There is experimental evidence that it is the axon that instructs the Schwann cell to produce myelin (Simpson and Young 1945; Hillarp and Olivecrona 1946; Weinberg, Spencer and Raine 1975). At a later stage the relationship between the number of turns of myelin lamellae and the axon area was closer and this suggests that the Schwann cell in normal fibres then responded to an increase in axon size by laying down the appropriate number of myelin lamellae. There is also evidence that the axon depends on the Schwann cell for its integrity (Singer and Salpeter 1966). Axon size may diminish as a result of demyelination (Lubifiska 1958). In an axon that has been deprived of the trophic support of the Schwann cell from the earliest days, it is likely that the axon area would be developmentally reduced. The lack of development of axon size is seen only when the axons are severely hypomyelinated over a great length, as in the Trembler mouse and in Dejerine-Sottas disease. When the length of hypomyelinated axons is short, as in the dystrophic mouse, the axis cylinders are of normal calibre (Bradley and Jenkison 1973; Weinberg et al. 1975), or when the hypomyelination is less, as in the Quaking mouse, the axons are of normal calibre and have a normal content of organelles (Samorajski et al. 1970). ACKNOWLEDGEMENTS The author was in receipt of the Roche Research Fellowship of the Royal Australasian College of Physicians. The helpful advice and criticism of Professor J. G. McLeod in the preparation of the paper was much appreciated.

367 REFERENCES Abercrombie, M. and M. L. Johnson (1946) Quantitative histology of Wallerian degeneration, J. Anat. (Lond.), 80" 37-50. Asbury, A. K. (1967) Schwann cell proliferation in developing mouse sciatic nerve - - A radioautographic study, J. Cell BioL, 34: 735-743. Ayers, M. M. and R. McD. Anderson (1973) A model of hypertrophic interstitial neuropathy (Dejerine-Sottas) in man, Acta neuropath. (Berl.), 25: 54-70. Ayers, M. M. and R. McD. Anderson (1975) Development of onion bulb neuropathy in the Trembler mouse - - Comparison with normal nerve maturation, Acta neuropath. (BerL), 32: 43-59. Ballin, R. H. M. and P. K. Thomas (1969) Electron microscope observations on demyelination and remyelination in experimental allergic neuritis, Part 2 (Remyelination), J. neurol. Sci., 8 : 225-237. Bischoff, A. and P. K. Thomas (1975) Microscopic anatomy of myelinated nerve fibers. In: P. J. Dyck, P. K. Thomas, and E. H. Lambert (Eds.), Peripheral Neuropathy, Saunders, Philadelphia, Pa., pp. 104-130. Bradley, W. G. and E. Jaros (1973) Axoplasmic flow impaired in axonal neuropathies, Part 2 (Axoplasmic flow in mice with motor neurone disease and muscular dystrophy), Brain, 96: 247-258. Bradley, W. G. and M. Jenkison, (1973) Abnormalities of peripheral nerves in murine muscular dystrophy, J. neuroL Sci., 18: 227-247. Bray, G. M. and A. J. Aguayo, Neonatal development of spinal roots in dystrophic mice - - Electronmicroscopic and radioautographic observations. Delivered at the Meeting of the Peripheral Nerve Club, 18-21 Sept., 1975, Rochester, Minn. Brimijoin, S., P. Capek and P. J. Dyck (1973) Axonal transport of dopamine-fl-hydroxylase by human sural nerves in vitro, Science, 180: 1295-1297. Duncan, D (1934) A relation between axon diameter and myelination determined by measurements of myelinated spinal root fibers, J. comp. Neurol., 60: 437-472. Dyck, P. J. and A. C. Lais (1971) Evidence for segmental demyelination secondary to axonal degeneration in Friedreich's ataxia. In: 2rid International Congress on Muscle Diseases, 22-26 November, Perth, Australia, pp. 1-19. Dyck, P. J., A. C. Lais and K. P. Offord (1974) The nature of myelinated nerve fiber degeneration in dominantly inherited hypertrophic neuropathy, Proc. Mayo Clin., 49: 34-39. Dyck, P. J., E. H. Lambert, K. Sanders, and P. C. O'Brien (1971) Severe hypomyelination and marked abnormality of conduction in Dejerine-Sottas hypertrophic neuropathy - - Myelin thickness and compound action potential of sural nerve in vitro, Proc. Mayo Clin., 46: 432-436. Dyck, P. J., R. D. Ellefson, A. C. Lais, R. C. Smith, W. F. Taylor and R. A. Van Dyke (1970) Histologic and lipid studies of sural nerves in inherited hypertrophic neuropathy - - Preliminary report of a lipid abnormality in nerve and liver in Dejerine-Sottas disease, Proc. Mayo Clin., 45 : 286-327. Elfvin, L. G. (1961) The ultrastructure of the nodes of Ranvier in cat sympathetic nerve fibres, J. Ultrastruct. Res., 5: 374-387. Finean, J. B., F. S. SjSstrand and E. Steinmann (1953) Submicroscopic organisation of some layered lipoprotein structures (nerve myelin, retinal rods, and chloroplasts), Exp. Cell Res., 5: 557-559. Fraher, J. P. (1972) A quantitative study of anterior root fibres during early myelination, J. Anat. (Lond.), 112: 99-124. Friede, R. L. and A. J. Martinez (1970) Analysis of axon sheath relations during early Wallerian degeneration, Brain Res., 19 : 199-212. Friede, R. L. and T. Samorajski (1967) Relation between the number of myelin lamellae and axon circumference in fibers of vagus and sciatic nerves of mice, J. comp. Neurol., 130: 223-232. Friede, R. L. and T. Samorajski (1968) Myelin formation in the sciatic nerve of the rat, J. Neuropath. exp. Neurol., 27: 546-570. Gamble, H. J. and R. A. Eames (1964) An electron microscope study of the connective tissues of human peripheral nerve. J. Anat. (Lond.), 98: 633-655. Hillarp, N. A. and H. Olivecrona (1946) The role played by the axon and Schwann cells in the degree of myelination of the peripheral nerve fibre, Acta Anat. (Basel), 2: 17-32. Jablecki, C. and S. Brimijoin (1974) Reduced axoplasmic transport of choline acetyltransferase activity in dystrophic mice, Nature (Lond.), 250: 151-154. Kaarlson, U. (1966) Comparison of the myelin period of peripheral and central origin by electron microscopy, J. Ultrastruct. Res., 15: 451-468.

368 Komiya, Y. and L. Austin (1974) Axoplasmic flow of protein in the sciatic nerve of normal and dystrophic mice, Exp. Neurol., 43: 1-12. Low, P. A. (1976a) Hereditary hypertrophic neuropathy in the Trembler mouse, Part 1 (Histopathological studies: light microscopy), J. neurol. Sci., 30: 327-341. Low, P. A. (1976b) Hereditary hypertrophic neuropathy in the Trembler mouse - - The evolution of "onion bulbs", Submitted for publication. Low, P. A. and J. G. McLeod (1975) Hereditary demyelinating neuropathy in the Trembler mouse, J. neuroL Sci., 26: 565-574. Lubiflska, L. (1958) Short internodes "intercalated" in nerve fibres, Acta Biol. exp. (Warszawa) 18: 117-136. Ochoa, J. (1971) The sural nerve of the human foetus - - Electron microscope observations and counts of axons, J. Anat. (Lond.), 108: 231-245. Ochoa, J. and W. G. P. Mair (1969) The normal sural nerve in man, Part 1 (Ultrastructure and numbers of fibres and cells), Acta neuropath. (Berl.), 13: 197-216. Reynolds, E. S. (1963) The use of lead citrate at high pH as an electron-opaque stain in electron microscopy, J. Cell Biol., 17: 208-213. Samorajski, T., R. L. Friede and P. R. Reimer (1970) Hypomyelination in the quaking mouse - - A model for the analysis of disturbed myelin formation, J. Neuropath. exp. Neurol., 29: 507-523. Simpson, S. A. and J. Z. Young (1945) Regeneration of fibre diameter after cross-unions of visceral and somatic nerves, J. Anat. (Lond.), 79: 48-65. Singer, M. and M. M. Salpeter (1966) Transport of tritium labelled L-histidine through the Schwann and myelin sheaths into the axon of peripheral nerves, Nature (Lond.), 210: 1225-1227. Sharma, A. K. and P. K. Thomas (1974) Peripheral nerve structure and function in experimental diabetes, J. neural. Sci., 23:1-15. Thomas, P. K. (1963) The connective tissue of peripheral nerve - - An electron microscope study, J. Anat. (Lond.), 97: 35-44. Watson, M. L. (1958) Staining for tissues for electron microscopy with heavy metals, J. biophys. biochem. CytoL, 41 : 475-478. Webster, H. de F. (1975) Development of peripheral myelinated and unmyelinated nerve fibers. In: P. J. Dyck, P. K. Thomas and E. H. Lambert (Eds.), Peripheral Neuropathy, Saunders, Philadelphia, Pa., pp. 37-61. Weinberg, H. J., P. S. Spencer and C. S. Raine (1975) Aberrant PNS development in dystrophic mice, Brain Res., 88: 432-537.