THE JOURNAL OF COMPARATIVE NEUROLOGY 312~599-609 (1991)

Lysosomal Activity in Developing Cat Alpha-Motor Axons Under Normal Conditions and During Retrograde Axonal Transport of Horseradish Peroxidase ~~

~

KLIMENT P. GATZINSKY, CLAES-HENRIC BERTHOLD, AND CHRISTIAN FABRICIUS Section of Neuroanatomy, Department of Anatomy, University of Goteborg, S-400 33 Goteborg, Sweden

ABSTRACT The occurrence of acid phosphatase (AcPase)-positive bodies, i.e., lysosomes, in lumbosacral alpha-motor axons of kittens, 0-16 weeks of age, was analyzed by light and electron cytochemical methods under normal conditions and after intramuscular injection of horseradish peroxidase (HRP). Axonal lysosomes were rare early postnatally. In 3-week-old animals, a few AcPase-positive bodies appeared in the axoplasm at some nodes of Ranvier in the peripheral nervous system (PNS) and internodally in the intrafunicular motor axon parts within the central nervous system (CNS). From 6 weeks postnatally, a nodal concentration of AcPasepositive bodies was also noted in the CNS. The number of AcPase-positive bodies continued to increase gradually in the course of neuronal maturation. In 16-week-old animals, axonal AcPase activity was still at considerably lower levels than at adult stages. At all ages, acid hydrolase-containing organelles were most commonly found at ventral root nodes. After injection of HRP in the medial gastrocnemius muscle, accumulations of AcPase-positive bodies were seen in the axoplasm at some PNS nodes of the HRP-injected sides of kittens aged 8, 12, and 16 weeks. Incubation for demonstration of both HRP and AcPase activity showed that some organelles at HRP-transporting nodes contained both types of reaction product. The nodal AcPase activity in the intrafunicular, CNS parts of alpha-motor axons of the HRPexposed sides did not differ from that of the contralateral, uninjected sides. In view of our previous observations in alpha-motor neurons of adult cats in which a lysosome-mediated degradation of axonally transported materials may take place at PNS nodes of Ranvier, the present study illuminates possible differences in the ability to interfere with axonal transport between developing and mature neurons. The infrequent presence of lysosomes in developing alpha-motor axons and the implied disability of their nodal regions to interfere with axonally transported constituents in a way similar to that seen in adult animals may be of significance in that trophic and chemical signals can pass unhindered between the periphery and perikaryon. However, this could also have negative consequences for the vulnerable immature neuron in that various materials retrieved at the axon terminals outside the CNS are permitted a more-or-less free access to the perikaryon. Key words: acid phosphatase, motoneuron, axon development, Ranvier's node

Recent studies on acid hydrolase activity in various neuron types have indicated that lysosome-mediated degradation of endogenous and exogenous materials is more extensive in the axon than previously assumed (Broadwell and Cataldo, '84; LaVail and Margolis, '87; Schmied and Holtzman, '87; Gatzinsky et al., '88; Gatzinsky and Berthold, '90). In the adult cat, axonal lysosomes observed in large myelinated hind limb and spinal root nerve fibres are normally concentrated at the nodes of Ranvier (Gatzinsky O

1991 WILEY-LISS, INC.

et al., '88; cf. Holtzman and Novikoff, '65). During retrograde axonal transport in alpha-motor neurons of intramuscularly administered horseradish peroxidase (HRP), a tracer protein, which en route to the cell body accumulates at Accepted June26,1991. Address reprint requests to Dr. K.P. Gatzinsky, Section of Neuroanatomy, Dept. of Anatomy, University of Goteborg, Box 33031, S-400 33 Goteborg, Sweden.

600

K.P. GATZINSKY ET AL.

nodes of Ranvier in the peripheral nervous system (PNS) (Berthold and Mellstrom, '82, '86; Berthold et al., '86), there is a lysosomal rearrangement and apparently an increase in number of lysosomes at several nodes of the HRP-transporting axons (Gatzinsky and Berthold, '90). Many lysosomes at these nodes contain HRP reaction product. The findings have been taken to indicate that nodal lysosomal activity varies with variations in axoplasmic transport and that a degradation of substances transported toward the neuronal perikaryon may be initiated locally at some nodes of Ranvier (cf. Berthold, '82; Janetzko et al., '89). Such an ability could possibly protect the motor neuron perikarya by restricting intraneuronal transport of exogenous, blood-borne materials, which, if being imbibed by the axon terminals outside the CNS, will circumvent the blood-brain barrier (cf. Broadwell and Brightman, '76). Since neither nodal accumulations of transported HRP nor a change in lysosomal activity is observed at nodes along the intrafunicular, central nervous system (CNS) parts of alpha-motor axons, the ability to interact with axonally transported constituents seems to differ between nodes along PNS and CNS parts of the same axon (Berthold et al., '88; Gatzinsky and Berthold, '90). Preliminary observations in our laboratory have shown that retrogradely transported HRP accumulates at nodes along the PNS portions of developing feline alpha-motor axons in a way similar to that seen in the adult cat first at the end of the second postnatal month (Fabricius et al., '87). These observations suggest that immature PNS nodes lack the adult ability to interfere with axonally transported substances. Although the lysosomal activity in association with the ensheathing Schwann and CNS glial cells has been investigated thoroughly in nodal and internodal parts of developing feline myelinated nerve fibres (Hildebrand and Skoglund, '71; Berthold, '73a), special attention seems neither to have been given to the lysosomal activity in the axon nor to the issue of whether there is an engagement of axonal lysosomes in degradation of axonally transported materials during the first months of life. The present study investigates axonal AcPase activity in the PNS and CNS parts of developing normal and HRPtransporting cat alpha-motor neurons by using light and electron cytochemical methods. The results are evaluated against the normal structural differentiation and maturation of the nodal regions in future large myelinated axons (Berthold, '68b; Hildebrand, '71b; Berthold, '74; Berthold and Mellstrom, '81). An abstract reporting preliminary findings has been presented (Gatzinsky and Berthold, '88).

MATERIALS AND METHODS Thirty-eight kittens, 0-16 weeks old, were used (see Table 1).In each age group, all animals except one, which was used as a control, were under pentobarbital anaesthesia and sterile conditions injected in the left medial gastrocnemius muscle (MGM) with a solution of 25 mg HRP (Type VI, Sigma Chemical Co., St Louis) dissolved in 125 ~1saline per kg body wt. Kittens up to 6 weeks of age were returned to the mother, elder kittens were left in the postoperative room free to feed and move. Survival times ranged from 4 to 70 hours postinjection (Table 1). Both HRP-injected kittens and animals not exposed to HRP were anaesthetized with pentobarbital (40 mgikg b. wt.) and fixed by vascular perfusion with cold (4°C) 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4). In all animals, the nerve to the

TABLE 1. Survey of Investigated Animals Survival time (h) after HRP-injection Age

Normal

4

8

24

Newborn 10 days 3 weeks 4 weeks 6 weeks 8 weeks 12 weeks 16weeks

1 1 1

1

1 1

1 2 1 2 2 1 1 2

1

1 1 1 1

2

42

48

50

60

70

2 1 1 1 1

6

I 1

1 1 1

Totalno.

1

1 1

3 4 5 4 4 5

medial gastrocnemius muscle (MGM nerve) and the spinal cord segments L7 and S1 together with their ventral roots were dissected out and postfixed for 90 minutes. After rinsing in buffer, the MGM nerves were teased into fascicles as thin as possible (1-5 nerve fibres per fascicle). Spinal cord segments were cut sagittally on an Oxford Vibratome in order to obtain longitudinal sections (20-25 Frn in thickness) through the intrafunicular, CNS parts of the ventral root axons. Consecutive longitudinal serial sections were cut from 1.5-2 mm long pieces of the middle portions of the ventral roots. In all specimens the proximodistal orientation with reference to the perikaryon was noted. Specimens from the uninjected animals, as well as from the left (experiment) and the right (control) sides of HRP-injected ones, were incubated for demonstration of AcPase activity in the Gomori medium (beta-glycerophosphate as substrate, Gomori, '52) for 30 or 60 minutes. After incubation, the specimens were rinsed several times and exposed to 1% ammonium sulphide for 2 minutes for conversion of lead phosphate to light microscopically detectable lead sulphide. In four HRP-injected animals aged 6 , 8 , 12, and 16 weeks, with survival times of 4 2 4 8 hours, some specimens from both the left and the right side were double-incubated; first in H,O, and 3',3'-diaminobenzidine (Graham and Karnovsky, '66) during 20 minutes for the demonstration of HRP and then in the Gomori medium for 30 minutes. For double incubation, survival times of 42-48 hours were chosen with reference to our previous investigation of adult cats, which showed that there is a rearrangement and increase in AcPase activity in some HRPtransporting nodal regions 2-3 days after intramuscular injection of the protein (Gatzinsky and Berthold, '90). Specimens incubated in substrate-free media served as histochemical controls. All specimens were osmicated, dehydrated, embedded in a thin layer of Vestopal W, and submitted to polymerization (for details see Berthold and Mellstrom, '82). After polymerization, all specimens were coded and systematically searched in the light microscope for axonal AcPase/HRP activity. In each age group, the occurrence of AcPase-positive bodies in paranode-node-paranode (PNP) segments (for terminology see Gatzinsky et al., '88) of the largest ventral root axons of uninjected animals and from the right sides of HRP-injected ones was analyzed quantitatively and compared with that in axons of the HRP-injected sides. PNP segments, which showed various light microscopical characteristics with regard to the occurrence and distribution of AcPase and/or HRP activity, were selected for ultrastructural analysis from each of the three investigated axon levels. Series of consecutive ultrathin (80-100 nm) longitudinal sections were cut through these PNP segments on an LKB 4800 Ultrotome, collected on Formvar coated one-hole copper grids, and examined in the electron

601

ACID PHOSPHATASE ACTIVITY AT NODES OF RANVIER TABLE 2. Light MicroscopicalEvaluation of Occurrence and Distribution of AcPase Activity at Different Ages in PNP Segments of Large Cat Ventral Root Axons'

Age Newborn 10 days 3 weeks 4 weeks 6 weeks 8 weeks 12 Weeks 16 weeks 1year'

PNP distribution of AcPase-Pos. bodies ('%I

AcPase-pos. PNP segments (%I

No. of AcPase-pos. bodies per AcPase-pos. PNP segment (mean S.E.M.)

CON-area

l(0-2) 6 (4-10) 15 (4-22) 18 (8-27) 34 (25-39) 30 (29-31) 44 (33-54) 58 (50-64) 88 (84-93)

1.167 i 0.204 1.278 t- 0.115 1.467 _t 0.153 1.472 f 0.106 1.735 i 0.103 1.522 f 0.076 1.788 i 0.083 2.162 f 0.112 5.012 k 0.206

100 61 27 13 13 8 5 2 1

*

CON+ASN-area

ASN-area

-

-

6 4 9 7 3 1 2 4

33 69 78 80 89 94 96 95

'Three animals from each age-group were included iu the quantitative analysis (in kittens, specimens were obtained from the animal not exposed to HRP and from the right, uninjeeted sides of 2 randomly selected, HRP-injected animals), From each animal 200 PNP segments of Vibratome sections incubated in the Gomori medium for 30 min were examined as they appeared during the systematic search of the specimens. All values are mean values, except for those in parentheses, which indicate animal ranges. PNP-paranode-node-paranode; CON-constricted axon segment. 'From Gatzinsky et al., '88.

microscope without further staining. In addition to examination of single sections, the ultrastructural distribution of AcPase activity was reconstructed in 20 strongly AcPasepositive PNP segments using a previously described superimposing technique (Gatzinsky et al., '88).

RESULTS The types of intraaxonal nonenzymatic lead deposits accounted for by Gatzinsky et al. ('88) were occasionally observed and are not considered further.

AcPase activity in axons of the MGM nerve and the L7 and S1 ventral roots Light microscopy Nerve fibres not directly exposed to HRP. The AcPase activity in specimens from animals not injected with HRP was similar to that in specimens from the right sides of animals exposed to HRP in their left hind limb. As in adult cats, axonal AcPase-positive bodies were more common in the ventral root than in the MGM nerve at all investigated ages (cf. Gatzinsky et al., '88). Axonal AcPase activity was rare in kittens younger than 3 weeks. At this age a few solitary AcPase-positive bodies appeared in some PNP segments. The bodies were situated either in the constricted (CON) axon segment and the paranodal axoplasm immediately adjacent to it, or just inside the bend where the myelin sheath turns to end on the constricted axon segment. In adult animals this latter region of the paranodal axon is often occupied by complexes of intenvoven axoplasm and adaxonal Schwann cell cytoplasm denoted axon-Schwann cell networks (ASNs) (Berthold, '68a; Spencer and Thomas, '74; Gatzinsky et al., '91). In accordance with Gatzinsky and Berthold ('901, the two parts of the PNP segment mentioned above are hereafter referred to as the CON-area and the ASN-area, respectively (see Fig. 2). The AcPase-positive bodies situated in the CON-area were generally smaller than those in the ASN-area, the latter often > 0.4 pm in size. Both CON-area- and ASN-areaassociated AcPase-positive bodies mainly appeared distal to the nodal midlevel. The quantitative analysis of AcPase activity in ventral root PNP segments of kittens not exposed to HRP and from the right sides of HRP-injected animals showed that there was an increase both in number of AcPase-positive nodes and in AcPase-positive bodies per node with increasing age (Table 2). Only 1%of the investigated PNP segments were

TABLE 3. Occurrence of CON-Area-AssociatedAccumulations of AcPase-positive Bodies in L7 and S1 Ventral Root Axons From HRP-Injected Animals'

No.AcPase-pos. bodies HRP-injectedside Control side

8 weeks 24 h

____

12 weeks 48 h

8 weeks 48 h

___

___

34

>4

3 4

>4

3-4

>4

25

2

14

5

30

5

7

-

3

-

1

1

-

16 weeks

48 h

___ 34 >4 29

8

-

1

'In each L7 and S1 ventral root the investigation was based on examination of all Vibratome sections from a ventral root piece, which was longitudinally sectioned through its whole extent. Half the sections were incubated in the Gomori medium for 30 min, the rest for 60 min. In each Vibratome section, all PNP segments along a distance of 1 mm were examined. Only animals, in which accumulations of > 4 AcPase-positive bodies were found are presented. The table shows, for example, that 5 of the investigated L7 and S1 ventral root PNP segments from the HRP-injected side of the 12 weeks old kitten, 48 h of survival, contained CON-area-associated accumulations of > 4 AcPase-positive bodies. Accumulations of 3 4 bodies were detected in 30 PNP segments. The corresponding numbers in specimens from the control side of this animal were 0 and 11, respectively. CON-constricted axon segment.

AcPase-positive in the newborn kittens. This frequency had increased to 58%in the 16-week-oldanimals, which should be compared with the 88%of adult animals (Table 2). The AcPase-positive bodies showed a higher predilection for the ASN-area than the CON-area with increasing age (Table 2). At all ages AcPase-positive bodies were very rare internodally. Nerve fibres directly exposed to HRP. Investigation of the intraaxonal peroxidase activity in the MGM nerves and L7 and S1 ventral roots of the HRP-injected animals used in this study showed that, in all animals, the tracer protein was restricted to nerve fibres from the injected sides (Fabricius et al., in preparation). The examination under code of specimens incubated in the Gomori medium showed that the occurrence and distribution of AcPase-positive bodies in PNP segments from the HRP-injected sides of kittens younger than 8 weeks did not differ from that observed in PNP segments from the right, uninjected sides. However, in contrast to the uninjected sides, the CON-areas of some of the examined ventral root PNP segments in the 8-week-old kittens with survival times of 24 hours and 48 hours contained accumulations of > 4 light microscopically detectable AcPasepositive bodies situated preferentially distal to the nodal midlevel (Table 3). Since the size of many of these bodies was just above the resolution limit of the light microscope (0.2-0.3 pm), their detection required a thorough examination of the axoplasm at different focal depths. Similar but

602

D

K.P. GATZINSKY ET AL.

-

P

Figures 1 and 2

ACID PHOSPHATASE ACTIVITY AT NODES OF RANVIER more distinct accumulations were also seen in some PNP segments of the left ventral roots in the 12- and 16-week-old animals with survival times of 48 hours (Table 3). Examination of all Vibratome sections from a longitudinally sectioned piece from each L7 and S l ventral root in the above mentioned animals showed that accumulations of this type were mainly concentrated to some sections of the S1 roots. These sections did also contain higher numbers of CONareas with 3-4 AcPase-positive bodies, as compared to sections of similarly investigated ventral root specimens from the corresponding uninjected sides (Table 3). CONarea-associated accumulations of AcPase-positive bodies were present in a few PNP segments of the left MGM nerve of the 16-week-oldkitten, survival time 24 hours, but were otherwise rare in the nerve. In double-incubated specimens from the 6-, 8-, 12-, and 16-week-old kittens, 42-48 hours of survival, the redbrownish HRP-positive bodies were either aggregated in the CON-area distal to the nodal midlevel, i.e., in a pattern similar to that of HRP-transporting so-called type A nodes of adult animals, or more commonly scattered throughout the whole CON-area, in a pattern resembling that of so-called type C nodes (cf. Berthold and Mellstrom, '86; Fabricius et al., '87). No HRP-distribution reminiscent to that seen at so-called type B nodes, which resemble type A nodes with the addition of a proximal disc-like accumulation of HRP activity, was observed in any of these animals. The HRP-positive bodies obscured detection of AcPase activity in the CON-area, whereas the comparatively large and black ASN-area-associated AcPase-positive bodies could be distinguished from the HRP reaction product, the latter rarely being found in this area (see Gatzinsky and Berthold, '90). Quantification of AcPase-positive bodies in the ASNareas of HRP-positive PNP segments showed that their number was similar to that at PNP segments of the uninjected sides in all 4 animals. Electron microscopy Nerve fibres not directly exposed to HRP. In the 0-4week-old kittens, the few AcPase-positive bodies present in

Fig. 1. All electron micrographs show aspects of longitudinally sectioned, uncontrasted PNP segments. D, distal to the nodal midlevel; P, proximal to the nodal midlevel (= direction given with reference to the perikaryon). Serial section reconstructions illustrate the distribution of AcPase-positive organelles in an approximately 2 - ~ mthick longitudinal median layer of the paranode-node-paranode (PNP) segments shown in the corresponding micrographs. Three-week-old kitten injected with HRP, survival time 24 hours. Medial gastrocnemius muscle (MGM)nerve PNP segment from the right, control side. Gomori medium 30 minutes. (a)An AcPase-positive body (arrow) is situated in the axoplasm distal to the nodal midlevel (indicated by vertical pair of arrowheads). x 12,500. (b) Serial section reconstruction of AcPase activity. Arrow points to the body shown in (a). Fig. 2. Six-week-old kitten not exposed to HRP. Ventral root PNP segment. Gomori medium 30 minutes. (a) Twq axon-Schwann cell network (ASN)complexes (asterisks) occupy the space (i.e., the ASNarea) just inside the bend where the proximal myelin sheath turns inward to end on the constricted axon segment. Single bars on the horizontal line (indicating the long axis of the axon) delimit the constricted axon segment. Double bars delimit the CON-area, which in addition to the constricted (CON) axon segment also includes the adjacent distal and proximal paranodal axoplasm, excluding the ASNarea. AcPase-positive bodies are scattered throughout the CON-area. X 12,500.(b)Serial section reconstruction ofAcPase activity. A distribution of AcPase-positive bodies distal to the nodal midlevel predominates.

603

the CON-area showed no definite proximodistal preponderance in relation to the nodal midlevel (Fig. 1).From 6 weeks of age a segregation was noted in the distribution of AcPase-positive bodies, in that bodies > 0.2 pm showed a predilection for the axoplasm distal to the nodal midlevel (Fig. 2). These bodies, which were mainly of a multivesicular or lamellar type, were intermingled both with smaller AcPase-positive vesicles and tubes, 50-150 nm in diameter, and with AcPase-negative organelles of the same morphological appearance as those containing AcPase (Fig. 3; see also Fig. 4). Paranodal ASNs were not seen before 4 weeks of age and were comparatively rare in animals younger than 8 weeks (Fig. 2). Up to this age, ASN-area-associated AcPasepositive bodies were mainly situated in the main axoplasmic compartment. From 8 weeks of age most AcPase-positive bodies >0.2 pm appeared in association with ASN complexes (Fig. 3). Nerve fibres directly exposed to HRP. The CON-areas of PNP segments from the HRP-injected sides, which contained light microscopically detectable accumulations of AcPase-positive bodies, also accommodated high numbers of small AcPase-positive vesicles and blunt-ended tubes, as well as many AcPase-negative bodies (Fig. 4).Sometimes, most acid hydrolase-containing vesiculotubular profiles were concentrated in the proximal half of the constricted axon segment, the tubes occupying the most proximal aspects (Fig. 4; cf. Gatzinsky and Berthold, '90). In double-incubated specimens, the AcPase and HRP reaction products could be distinguished from one another, the former being of a higher electron density (Figs. 5,6; see Broadwell and Brightman, '79; LaVail and Margolis, '87; Gatzinsky and Berthold, '90). The AcPase activity in the CON-area differed widely between PNP segments that showed about the same distribution and/or amount of HRP-positive bodies. Most of the examined HRP-positive PNP segments contained no or only a few AcPase-positive bodies. Some PNP segments, however, contained higher numbers of AcPase-positive bodies intermingled with the HRP-containing ones (Fig. 5 ) . Several organelles in the axoplasm of these PNP segments contained both HRP and AcPase reaction product (Figs. 5 , 6). In none of the doubleincubated, HRP-positive PNP segments that were investigated ultrastructurally did we observe an AcPase activity as intense as that in the strongest AcPase-positive, singleincubated PNP segments.

AcPase activity in the intrafunicular (CNS) parts of ventral root axons Because of the absence of clear constrictions, nodes of Ranvier in the CNS were recognized with some difficulty in newborn and 10-day-old kittens (cf. Hildebrand, '71b). The nodal regions of the largest fibres became more distinctly demarcated from the third postnatal week and further on. Intraaxonal AcPase-positive bodies were rare in the CNS before 3 weeks of age. In the 3-week-old kittens, a few AcPase-positive bodies were seen mainly internodally in some axons. From 6 weeks of age, a nodal concentration of AcPase activity was noted also in the CNS axons. The occurrence and distribution of AcPase-positive bodies in PNP segments from the HRP-injected sides were similar to that seen in specimens from uninjected kittens and from the right, control sides of HRP-exposed animals. The AcPase-positive bodies were situated exclusively in the CON-area, mainly distal to the nodal midlevel (Fig. 7). No accumulations of AcPase-positive vesiculotubular profiles,

K.P. GATZLNSKY ET AL.

604

P

Figures 3 and 4

ACID PHOSPHATASE ACTIVITY AT NODES OF RANVIER

605

rials of various origin is most likely to take place in the ventral roots, where the highest contents of axonal lysosomes are found. Still at 16 weeks of age the nodal lysosomal activity is at a considerably lower level than in motor axons of adult cats (cf. Gatzinsky et al., '88). The acquirement of a more manifest nodal lysosomal activity coincides with the time when PNS nodes of feline alpha-motor neurons start to delay intramuscularly administered, retrogradely transported HRP in a way similar to that seen at adult stages (Fabricius et al., '87). The characteristics and age-related temporal course of the nodal interference in HRP transport are subject for extended studies at present (Fabricius et al., in preparation). Although an HRP degradation, as illustrated by the dual presence of HRP and AcPase reaction product in the same organelle, was observed to take place in the axoplasm of some double-incubated, HRP-transporting PNP segments, DISCUSSION we found no apparent HRP-associated changes in nodal General interpretation of the results AcPase activity as compared to specimens from the uninThe present study shows that lysosomes are rare in cat jected sides. However, in the animals 8-16 weeks of age, lumbosacral alpha-motor axons early postnatally. Thus where CON-area-associated accumulations of lysosomes retrogradely transported substances, imbibed by the axon were present in some PNS axons incubated for demonstraterminals in the PNS, should have to reach the neuronal tion only of AcPase activity, a relation between this type of perikarya before degradation during this period. The post- accumulations and HRP transport may exist for two reanatal increase in intraaxonal lysosomal activity is gradual, sons: (1)the accumulations appeared preferentially in PNP and alpha-motor neurons seem not to have the prerequi- segments of the HRP-injected sides, and (2) in ventral root sites to initiate a more extensive lysosome-mediated degra- pieces, which were sectioned longitudinally through their dation within their axons until 2-3 months after birth. This whole extent, PNP segments containing such accumuladegradative ability develops during a period when many tions were often grouped toogether and concentrated to feline lumbo-sacral alpha-motor neuron populations just some of the investigated Vibratome sections, probably have attained perikaryal sizes close or equal to adult being of those parts of the roots through which axons of the dimensions (Tatton and Theriault, '88; Ulfhake and Cull- MGM nerve pass. We have previously noted that HRPheim, '88; Cameron and Fang, '89; Horcholle-Bossavit et transporting axons belonging to alpha-motor neurons of al., '901, but is several months ahead of the time at which the MGM motor nucleus (about 260 alpha-motor neurons, adult internodal mean lengths and axon diameters are see Burke et al., '77) show a similar distribution during reached (Nilsson and Berthold, '88). As in adult animals, their passage through the ventral root. Furthermore, as in axonal lysosomes are concentrated at nodes of Ranvier, adult cats, CON-area-associated lysosomal accumulations which due to the considerable reduction in axon cross- were most prominent in the ventral roots around 2 days after HRP injection (cf. Gatzinsky and Berthold, '90). Since section area are suitable sites for temporary retardation of organelle transport and for chemical interactions between the HRP reaction product obscured detection of AcPasepositive bodies in the CON-area at the light microscopical intraaxonal membrane constituents, e.g., fusion of lysolevel and due to the variation in AcPase activity between somes and retrogradely transported endosomes (Berthold, individual PNP segments, detection of similar accumula'82; Gatzinsky et al., '88; Janetzko et al., '89; Gatzinsky and tions of hydrolase-containing constituents in doubleBerthold, '90). A putative intraaxonal degradation of mate- incubated specimens would probably require ultrastructural examination of a large number of randomly selected, light microscopically HRP-positive PNP segments. Moreover, in double-incubated specimens of adult cats, CON-areaFig. 3. Sixteen-week-old kitten not exposed to HRP. Ventral root PNP segment. Gomori medium 60 min. (a)Two of the AcPase-positive associated lysosomal accumulations mainly appeared at HRP-transporting nodes of the so-called B type, which bodies situated in the CON-area are indicated by arrows. Larger AcPase-positive bodies occupy an ASN-area-associated protrusion of show the highest concentrations of HRP-positive bodies adaxonal Schwann cell cytoplasm (arrowhead) proximal to the con- (Gatzinsky and Berthold, '90). No HRP-positive nodes of stricted axon segment. This latter compartment, which is shown at a this type were observed in the double-incubated specimens higher magnification in (b),contains both AcPase-positive and -nega- investigated in this study. (In the work by Gatzinsky and tive membraneous constituents. (a) ~9,700.(b) ~69,200. Berthold, '90, problems concerning the interpretation of changes in nodal AcPase activity in both Gomori- and Fig. 4. Sixteen-week-old kitten injected with HRP, survival time 48 double-incubated axons, and the possible connection to hours. Ventral root PNP segment from the left, injected side. Gomori medium 30 minutes. This PNP segment contained six light microscopi- HRP transport, have been discussed.) Possible age-dependent differences between immature cally detectable CON-area-associated AcPase-positivebodies. (a)AcPasepositive bodies of different morphological characteristics are situated in and mature alpha-motor axons in local turnover of subthe CON-area both distal and proximal to the nodal midlevel. The stances in the nodal region are also illustrated by the axoplasm also contains many AcPase-negative organelles. x 9,700. (b) postnatal scarcity of paranodal ASN complexes, which are Serial section reconstruction of AcPase activity. Note concentration of larger (> 0.2 pm) AcPase-positive bodies immediately distal to the entitites suggested to participate in the removal of degenerconstricted axon segment and the accumulation of AcPase-positive ate axonal materials, e.g. associated with secondary lysosvesiculotubular profiles in the proximal part of the constricted axon omes (Spencer and Thomas, '74; Ellisman et al., '84; segment, where the tubes occupy the most proximal aspects. Gatzinsky et al., '88; Gatzinsky and Berthold, '90). Taken reminiscent of those present in the proximal half of CONareas in the PNS, were seen. The nodal axoplasm generally showed lower contents of organelles than at the PNS nodes. Compartments corresponding morphologically to the paranodal ASNs were not seen. AcPase-positive bodies were abundant in the motor neuron perikarya at all investigated ages. Examination of CNS specimens incubated for demonstration of HRP activity showed that the tracer protein was restricted to neuron cell bodies of the injected sides (cf. Fabricius et al., '87). Both in these specimens and in the double-incubated specimens used in the present study, the PNP segments along the intrafunicular part of ventral root axons from the HRP-injected sides were virtually free from light microscopically detectable peroxidase activity.

606

Fig. 5. Twelve-week-oldkitten injected with HRP, survival time 48 hours. Double-incubated ventral root PNP segment from the left, injected side, which appeared strongly HRP-positive in the light microscope. Vertical pair of arrowheads denotes the nodal midlevel. Several HRP-positive bodies, some of which are indicated by small arrowheads, are distributed in the CON-area distal to the nodal midlevel. The axoplasm in this section also contains one AcPasepositive body (arrow), the electron density of which is considerably higher than the density indicative of HRP reaction product. I t is not possible to tell whether the AcPase-positive body contains any HRP. x 20,500. Inset shows a CON-area-associated HRP-positive body with contents of AcPase (arrows) observed in another longitudinal section of this PNP segment. ~46,400. Fig. 6. Micrographs showing two HRP-positive bodies (one contains AcPase reaction product-arrow, the other is without such contents-

K.P. GATZINSKY ET AL.

arrowhead) in a double-incubated,HRP-transporting ventral root PNP segment. The same bodies are shown in (a) and (b). (a1 represents a normal exposure, emphasizing the labeling with HRP reaction product. (bj is an underexposed image, where the electron density of the HRP reaction product is reduced, thereby enhancing the visibility of the black lead precipitate indicative of AcPase reaction product. x 58,700.

Fig. 7. Sixteen-week-old kitten not exposed to HRP. PNP segment from the intrafunicular (CNSj part of the ventral root. Gomori medium 30 minutes. (a)Some AcPase-positive bodies (arrows) are situated in the axoplasm mainly distal to the nodal midlevel, which is denoted by vertical pair of arrowheads. m, mitochondria. X9,700. (b) Serial section reconstruction of AcPase activity. This type of distribution in AcPase activity was also the one found in CNS specimens from the HRPinjected sides.

ACID PHOSPHATASE ACTIVITY AT NODES OF RANVIER together, all findings mentioned above illuminate what we consider as important differences between PNP segments of developing and mature feline alpha-motor neurons with reference to their ability to interfere with, and probably also take part in degradation of, materials undergoing axonal transport. The new findings concerning the AcPase activity expressed within axons of developing cat alpha-motor neurons stand in contrast to the postnatal AcPase activity in the ensheathing Schwann cells, which is most prominent during the first 2 weeks of life (Berthold, '73a). The high Schwann cell-associated hydrolytic activity early postnatally is assumed to be involved in a local breakdown of the myelin sheath, leading to elimination of internodes preferably in large myelinated nerve fibres. This demyelination is part of a normally occurring, structural remodeling of the PNP-segments denoted t h e "nodalzation process" (Berthold, '68b, '73b, '74; Berthold and Mellstrom, '811, which is finished during the 4th to 6th postnatal week, when large nerve fibres attain a diameter of 4-5 pm. In addition, unpublished observations in our laboratory (Andersson, personal communication) have indicated that the degree of nodal axon constriction in relation to internodal axon diameter in developing feline alpha-motor neurons approaches adult values (cf. Rydmark, '81) during the 4th to 6th postnatal week. This raises the possibility that the gradually elevated nodal concentration of AcPasereactive bodies in kittens up to 6 weeks of age mainly reflects an increased trapping of axonally transported lysosomes at the PNS nodes. The number of nodal lysosomes continues, however, to increase also after the "mature" constriction values have been achieved. Since lysosomes are normally abundant in developing motor neuron perikarya (Decker, '74; Maslinska and Thomas, '75; Clarke and Martin, '85; this study), and since the main properties of fast axoplasmic transport appear to be uniform from the first weeks of life and through adulthood (Ochs, '84), the infrequent presence of axonal lysosomes early postnatally could imply that no or only small numbers of acid hydrolasecontaining constituents are directed from the perikaryon into the axon in very young animals. Such age-dependent differences in delivery of acid hydrolases might be an expression for more widespread, progressive alterations in the amount, composition, and/or destination of materials moving by fast axonal transport in the course of neuronal maturation (Benowitz, '87). As in adult cats, noteworthy differences were found between PNS and CNS nodes of developing alpha-motor neurons in the ability to interfere with passage of axonally transported constituents (cf. Berthold et al., '88; Gatzinsky and Berthold, '90). A restriction in the movement of small and, as it seems, preferably anterogradely transported vesiculotubular organelles (see Hollenbeck and Swanson, 'go), which, if being lysosomes, can start a local degradation of axonally transported materials, mainly appears to take place at the PNS nodes. A nodalization process corresponding to that in the PNS, but with different structural manifestations, has also been demonstrated to take place in future large myelinated CNS axons, the PNP segments of which show a clear constriction at the end of the second week and attain a basically mature morphology at 4-6 weeks of age (Hildebrand, '71b). Although the development of the nodal axon constriction is reminiscent to that in the PNS, a concentration of lysosomes at nodes along the spinal cord portions of alpha-motor axons was seen first in the 6-week-old kittens, i.e., 3 weeks later than at the PNS

607

nodes. The underlying mechanisms for the implied lower ability of PNP segments along the CNS parts of both mature and developing motor axons to interfere with axonal transport, which is also exemplified by the generally lower organelle concentration in their axoplasm as compared to the PNP segments in the PNS (Berthold et al., '88), are not known. This difference probably depends on factors other than the mere degree of nodal constriction, which is approximately of the same magnitude in PNS and CNS parts of feline hind limb alpha-motor axons (Fabricius et al., '86; Berthold et al., '88). The concentration and organization of the transportpromoting microtubules most likely play an important role for axonal transport capacity. For example, Berthold ('82) has observed that the microtubular distribution differs between PNS nodes showing a high and a low axoplasmic density of organelles, respectively. No comparison between number and/or arrangement of microtubules appears to have been performed between nodes along the PNS and CNS parts of the same axon. However, in two recent studies, the internodal microtubular density in the spinal cord parts of small rat and frog motor and sensory axons was found to be approximately twice as high as that in their ventral and dorsal root parts, respectively (Saitua and Alvarez, '89; Lopez and Alvarez, '90). Since axonal microtubules consist of both stable and dynamic units, the latter undergoing a continuous assemblyldisassembly locally within the axon (Morris and Lasek, '84; Baas and Black, '90; Okabe and Hirokawa, 'go), the findings by Alvarez and colleagues suggest that some local mechanism contributes to the regulation of microtubular dynamics and organization. Evidence has now accumulated that ensheathing cells play a role in determining the structure and function of myelinated axons (Windebank et al., '85; Pannese et al., '88; de Waegh and Brady, '90; Price et al., '90). Thus Schwann cells seem to possess a capacity to exert a local influence on transport, assembly/disassembly, and organization of the axonal cytoskeleton (de Waegh and Brady, '90). As indicated by the regular presence of paranodal ASNs in large myelinated PNS axons and the scarcity under normal conditions of similar entities in the CNS (Hildebrand, '71a; Spencer and Thomas, '74; Gatzinsky et al., '88, '91), the Schwann cell-axon interplay is probably mediated differently from that between the astrocyte/oligodendrocyte and the axon. Perhaps Schwann cells and CNS glial cells differ in their ability to influence on the organization of the axonal cytoskeleton (cf. de Waegh and Brady, '881, which in turn could have bearings upon axonal transport capacity preferably in the nodal regions, where the axon crosssection area is reduced to less than 25% of internodal values (Rydmark, '81).

Possible functional implications of low intraaxonal contents of lysosomes during the early postnatal period During late foetal and early postnatal development, many neuronal populations undergo progressive changes in the character of their retrograde responses to axotomy and other metabolic or toxic injuries (Griffith and LaVelle, '71; Lieberman, '74). Presumably, the well-known higher vulnerability of immature mammalian neurons to axonal lesions to some extent can be attributed to the cellular machinery (particularly the cytoplasmic protein synthesizing system) involved in the normal metabolic response to stress and injury, which neither seems to be fully developed in imma-

608

K.P. GATZINSKY ET AL.

ture neurons nor can initiate andtor sustain the specific responses elicited by an equivalent injury to adult neurons (Dennis and Harris, ’80; Sheard et al., ’84). Immature neurons are, however, also highly dependent upon the influence exerted by the functional state of their axons and ensheathing cells, as well as by their peripherally located effector organs, which, via retrograde axonal transport of various endocytotically retrieved trophic factors, govern normal neuronal development, growth, and maintenance (for some comprehensive reviews, see Thoenen and Schwab, ’79; Oppenheim, ’81; Kristensson, ’84). A decrease in the supply of such factors, e.g., after axonal transection or due to insufficient contacts between axon terminals and target organs, can be deleterious to a variety of immature neuron types, including motor neurons (Oppenheim et al., ’88; Crews and Wigston, ’90; Kuno, ’90;McManaman et al., ’90; Weill and Greene, ’90). Conceivably, during the critical early postnatal period, when development of the nervous system is extra sensitive to external conditions, the infrequent presence of lysosomes in alpha-motor axons, as well as the suggested disability of their PNP segments to interfere with passage of axonally transported materials, may have a functional significance in that chemical signals can pass promptly and properly back and forth between the periphery and the perikaryon. However, as a consequence of such a more or less “unrestricted” axonal flow, various endocytotically imbibed exogenous substances, which do not fulfill trophic functions, but may instead be potentially toxic to the neuron, will be able to circumvent the bloodbrain barrier and reach the lower motor neuron perikarya in the CNS via retrograde axonal transport (cf. Kristensson, ’70; Kristensson et al., ’71; Broadwell and Brightman, ’76; Malmgren et al., ’77; Olsson and Kristensson, ’79). Thus the neuronal susceptibility of very young mammals to various noxious agents, which can spread to the CNS via retrograde axonal transport, e.g., along motor neurons (Kristensson, ’82; Griffin and Watson, ’88;Card et al., ’901, may to some degree reflect the implied postnatal disability particularly of the nodal regions in the PNS to take part in turnover of exogenous materials emanating from axon terminals outside the CNS.

ACKNOWLEDGMENTS This work was supported by the Swedish Medical Research Council (project no. 03157) and by grants from the Goteborg Medical Society, the Medical Faculty, University of Goteborg, and Stiftelsen Lars Hiertas Minne. We thank Dr. A. Mellstrom for valuable discussions and Marieanne Eriksson and Rita Grander for excellent technical assistance.

LITERATURE CITED Baas, P.W., and M.M. Black (1990) Individual microtubules in the axon consist of domains that differ in both composition and stability. J. Cell Biol. 111r495-509. Benowitz, L.I. (1987) Protein synthesis and fast axonal transport in developing and regenerating neurons. In R.S. Smith, and M.A. Bishy (eds): Axonal Transport. New York: Alan R. Liss, pp. 385406. Berthold, C.-H. (1968a) Ultrastructure of the node-paranode region of mature feline ventral lumbar spinal-root fibres. Acta Soc. Med. Upsal. 73(Suppl. 9)t37-78. Berthold, C.-H. (196%) Ultrastructure of postnatally developing feline Med. Upsal. 73:145-175. peripheral nodes of Ranvier. Acta SOC. Berthold.~,C.-H. (1973a) Histochemistrv of uostnatallv develouing -~ feline spinal roots. 11. Occurrence of acid phosphatase activity as studied by light and electron microscopical methods. Neurobiology 3:291-310. ~

“

I

Berthold, C.-H. (1973b) Local “demyelination” in developing feline nerve fibres. Neurobiology 3:339-352. Berthold, C.-H. (1974) A comparative morphological study of the developing node-paranode region in lumbar spinal roots. I. Electron microscopy. Neurobiology 4:82-104. Berthold, C.-H. (1982) Some aspects of the ultrastructural organization of peripheral myelinated axons in the cat. In D.G. Weiss (ed): Axoplasmic Transport. Berlin: Springer-Verlag,pp, 40-54. Berthold, C-H., and A. Mellstrom (1981) Postnatal development of nodeparanode regions in two hind limb nerves of the cat. Dev. Biol. 86:lll-116. Berthold, C.-H., and A. Mellstrom (1982) Distribution of peroxidase activity a t nodes of Ranvier after intramuscular injection of horseradish peroxidase in the cat. Neuroscience 7r45-54. Berthold, C.-H., and A. Mellstrom (1986) Peroxidase activity at consecutive nodes of Ranvier in the nerve to the medial gastrocnemic muscle after intramuscular administration of horseradish peroxidase. Neuroscience 19r1349-1362. Berthold, C.-H., 0. Corneliuson, and A. Mellstrom (1986) Peroxidase activity a t nodes of Ranvier in !umbosacral ventral roots and in the PNS-CNS transitional region after intramuscular administration of horseradish peroxidase. J. Neurocytol. 15253-260. Berthold, C.-H., 0. Corneliuson, and A. Mellstrom (1988) Peroxidase activity at CNS nodes of Ranvier and in initial axon segments of lumbosacral alpha-motor neurons after intramuscular administration of horseradish peroxidase. Brain Res. 456r293-301. Broadwell, R.D., and M.W. Brightman (1976) Entry of peroxidase into neurons of the ventral and peripheral nervous system from extracerebral and cerebral blood. J. Comp. Neurol. 166:257-284. Broadwell, R.D., and M.W. Brightman (1979) Cytochemistry of undamaged neurons transporting exogenous protein in vivo. J. Comp. Neurol. 185:31-74. Broadwell, R.D., and A.M. Cataldo (1984) The neuronal endoplasmic reticulum: Its cytochemistry and contribution to the endomembrane system. 11. Axons and terminals. J. Comp. Neurol. 230:231-248. Burke, R.E., P.L. Strick, K. Kanda, C.C. Kim, and B. Walmsley (1977) Anatomy of medial gastrocnemius and soleus motor nuclei in cat spinal cord. J. Neurophysiol. 4Ot667-680. Cameron, W.E., and H. Fang (1989) Morphology of developing motoneurons innervating the medial gastrocnemius of the cat. Dev. Brain Res. 49: 253-263. Card, J.P., L. Rinaman, J.S. Schwaber, R.R. Miselis, M.E. Whealy, A.K. Robbins, and L.W. Enquist (1990) Neurotropic properties of pseudorabies virus: Uptake and transneuronal passage in the rat central nervous system. J. Neurosci. 10:1974-1994. Clarke, P.G.H., and A.H. Martin (1985) Effects of de-efferentation on chick spinal motoneurons: peroxidase uptake, and activities of acid phosphatase and N-acetyl-B-glucosaminidase.Cell Biol. Int. Rep. 9r676. Crews, L.L., and D.J. Wigston (1990) The dependence of motoneurons on their target muscle during postnatal development of the mouse. J. Neurosci. lO:1643-1653. Decker, R.S. (1974) Lysosomal packaging in differentiating and degenerating anuran lateral motor column neurons. J. Cell Biol. 61t599-612. Dennis, M.J., and A.J. Harris (1980) Transient inability of neonatal rat motoneurons to reinnervate muscle. Dev. Biol. 74r173-183. de Waegh, S., and S.T. Brady (1988) Altered slow axonal transport in optic nerve of shiverer mutant mice. Soc. Neurosci. Abstr. I4:48.1 (Abstract). de Waegh, S., and S.T. Brady (1990) Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: the trembler as an in vivo model for Scwann cell-axon interactions. J. Neurosci. lOr18551865. Ellisman, M.H., C.A. Wiley, J.D. Lindsey, and C.C. Wurtz (1984) Structure and function of the cytoskeleton and endomembrane systems at the node of Ranvier. In J.C. Zagoren and S. Fedoroff (eds): The Node of Ranvier. New York: Academic Press, pp. 153-181. Fabricius, C., M. Rydmark, and C.-H. Berthold (1986) A morphometric study of myelinated nerve fibres of the intrafunicular part of cat motoneurons. Arch. Biol. 97t42. Fabricius, C., C.-H. Berthold, 0. Corneliuson, and K. Gatzinsky (1987) Peroxidase activity in developing cat alpha motoneurons after intramuscular HRP-injection. Neuroscience Suppl. 22:S794 (Abstract). Gatzinsky, K., and C.-H. Berthold (1988) Lysosomes in developing feline alpha-motor axons. J. Histochem. Cytochem. 36:959 (Abstract)

ACID PHOSPHATASE ACTIVITY AT NODES OF RANVIER Gatzinsky, K.P., and C.-H. Berthold (1990) Lysosomal activity at nodes of Ranvier during retrograde axonal transport of horseradish peroxidase in alpha-motor neurons of the cat. J. Neurocytol. 19r989-1002. Gatzinsky, K.P., C.-H. Berthold, and 0. Corneliuson (1988) Acid phosphatase activity at nodes of Ranvier in alpha-motor and dorsal root ganglion neurons of the cat. J. Neurocytol. I7t531-544. Gatzinsky, K.P., C.-H. Berthold, and M. Rydmark (1991) Axon-Schwann cell networks are regular components of nodal regions in normal large nerve fibres of cat spinal roots. Neurosci. Lett. 124r264-268. Gomori, G. (1952) Microscopic Histochemistry: Principles and Practice. Chicago: University of Chicago Press. Graham, R.C., Jr., and M.J. Karnovsky (1966) The early stages of adsorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: Ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem. 14:29 1-3 02. Griffin, J.W., and D.F. Watson (1988) Axonal transport in neurological disease. Ann. Neurol. 23:3-13. Griffith, A., and A. LaVelle (1971) Developmental protein changes in normal and chromatolytic facial nerve nuclear regions. Exp. Neurol. 33:360371. Hildebrand, C. (1971a) Ultrastructural and light-microscopic studies of the nodal region in large myelinated fibres of the adult feline spinal cord white matter. Acta Physiol. Scand. Suppl. 364t43-79. Hildebrand, C. (1971b) Ultrastructural and light-microscopic studies of the developing feline spinal cord white matter. I. The nodes of Ranvier. Acta Physiol. Scand. Suppl. 364:81-107. Hildebrand, C., and S. Skoglund (1971) Histochemical studies of adult and developing feline spinal cord white matter. Acta Physiol. Scand. Suppl. 364t145-173. Hollenbeck, P.J., and J.A. Swanson (1990) Radial extension of macrophage tubular lysosomes supported by kinesin. Nature 346t864-866. Holtzman, E., and A.B. Novikoff (1965) Lysosomes in the rat sciatic nerve followingcrush. J. Cell Biol. 27;651-669. Horcholle-Bossavit, G., L. Jami, D. Thiesson, and D. Zytnicki (1990) Postnatal development of peroneal motoneurons in the kitten. Dev. Brain Res. 54905-215. Janetzko, A., H. Zimmermann, and W. Volknandt (1989) Intraneuronal distribution of a synaptic vesicle membrane protein: antibody binding sites at axonal membrane compartments and trans-Golgi network and accumulation a t nodes of Ranvier. Neuroscience 3265-77. Kristensson, K. (1970) Transport of fluorescent protein tracer in peripheral nerves. Acta Neuropath. 16:293-300. Kristensson, K. (1982) Implications of axoplasmic transport for the spread of virus infections in the nervous system. I n D.G. Weiss, and A. Gorio (eds): Axoplasmic Transport in Physiology and Pathology. Berlin: SpringerVerlag, pp. 153-158. Kristensson, K. (1984) Retrograde signaling after nerve injury. I n J.S. Elam and P. Cancalon (eds): Axonal Transport in Neuronal Growth and Regeneration. New York Plenum Press, pp. 31-43. Kristensson, K., Y. Olsson, and J. Sjostrand (1971) Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve. Brain Res. 32:399-406. Kuno, M. (1990) Target dependence of motoneuronal survival: the current status. Neurosci. Res. 9t155-172. LaVail, J.H., and T.P. Margolis (1987) The anterograde axonal transport of wheat germ agglutinin as a model for transcellular transport in neurons. In R.S. Smith, and M.A. Bisby (eds): Axonal Transport. New York: Alan R. Liss, pp. 311-326. Lieberman, A.R. (1974) Some factors affecting retrograde neuronal responses to axonal lesions. In R. Bellairs, and E.G. Gray (eds): Essays on the Nervous System. Oxford: Clarendon Press, pp. 71-105. Lopez, J.M.A., and J. Alvarez (1940) The microtubular pattern changes at the spinal cord-root junction and reverts at the root-peripheral nerve junction in sensory and motor fibres of the rat. Eur. J. Neurosci. 2873-878. McManaman, J.L., R.W. Oppenheim. D. Prevette. and D. Marchetti (1990) Rescue of motoneuron8-from cell death by a purified skeletal muscle

609

polypeptide: effects of the ChAT development factor, CDF. Neuron 4t891-898. Malmgren, L.T., M.J. Lyon, and R.R. Gacek (1977) Localization of abductor and adductor fibers in the kitten recurrent laryngeal nerve: use of a variation of the horseradish peroxidase tracer technique. Exp. Neurol. 55t187-198. Maslinska, D., and E. Thomas (1975) Enzyme-histochemical studies of "retrograde" reaction in motor neurones of immature rats. Acta Neuropath. 33:317-323. Morris, J.R., and R.J. Lasek (1984) Monomer-polymer equilibria in the axon: direct measurement of tubulin and actin as polymer and monomer in axoplasm. J. Cell Biol. 98t2064-2076. Nilsson, I., and Berthold, C.-H. (1988) Axon classes and internodal growth in the ventral spinal root L7 of adult and developing cats. J. Anat. 156:7 1-96. Ochs, S. (1984) Basic properties of axoplasmic transport. In P.J. Dyck, P.K. Thomas, E.H. Lambert, R.P. Bunge (eds): Peripheral Neuropathy. Philadelphia: Saunders, pp. 453-476. Okabe, S., and N. Hirokawa (1990) Turnover of fluorescently labelled tubulin and actin in the axon. Nature 343t479-482. Olsson, T., and K. Kristensson (1979) Uptake and retrograde axonal transport of horseradish peroxidase in normal and axotomized motor neurons during postnatal development. Neuropath. Appl. Neurobiol. 5:377-387. Oppenheim, R.W. (1981) Neuronal cell death and some related regressive phenomena during neurogenesis: a selectivehistorical review and progress report. In W.M. Cowan (ed): Studies in Developmental Neurobiology. New York: Oxford University Press, pp. 74-133. Oppenheim, R.W., L.J. Haverkamp, D. Prevette, J.L. McManaman, and S.H. Appel(1988) Reduction of naturally occurring motoneuron death in vivo by a target-derived neurotrophic factor. Science 240:919-922. Pannese, E., M. Ledda, and S. Matsuda (1988) Nerve fibres with myelinated and unmyelinated portions in dorsal spinal roots. J. Neurocytol. 17t693700. Price, R.L., R.J. Lasek, and M.J. Katz (1990) Internal axonal cytoarchitecture is shaped locally by external compressive forces. Brain Res. 530:205214. Rydmark, M. (1981) Nodal axon diameter correlates linearly with internodal axon diameter in spinal roots of the cat. Neurosci. Lett. 24.247-250. Saitua, F., and J. Alvarez (1989) Microtubular packing varies along the course of motor and sensory axons: possible regulation of microtubules by environmental cues. Neurosci. Lett. 104r249-252. Schmied, R., and E. Holtzman (1987) A phosphatase activity and a synaptic vesicle antigen in multivesicular bodies of frog retinal photoreceptor terminals. J. Neurocytol. 16:627-637. Sheard, P., C.D. McCaig, and A.J. Harris (1984) Critical periods in rat motoneuron development. Dev. Biol. 102:21-31. Spencer, P.S., and P.K. Thomas (1974) Ultrastructural studies of the dying-back process. 11. The sequestration and removal by Schwann cells and oligodendrocytes of organelles from normal and diseased axons. J. Neurocytol. 3r763-783. Tatton, W.G., and E. Theriault (1988) Postnatal growth of medial gastrocnemius motoneurons in the kitten. Dev. Brain Res. 43t191-206. Thoenen, H., and M. Schwab (1979) Physiological and pathophysiological implications of the retrograde axonal transport of macromolecules. Adv. Pharm. Ther. 5t37-59. Ulfhake, B., and S. Cullheim (1988) Postnatal development of cat hind limb rnotoneurons. 111: Changes in size of motoneurons supplying the triceps surae muscle. J. Comp. Neurol. 278:103-120. Weill, C.L., and D.P. Greene (1990) The prevention of natural motoneuron cell death by dibutyryl-cyclic GMP in the spinal cord of White Leghorn chick embryos. Dev. Brain Res. 55r143-146. Windebank, A.J., P. Wood, R.P. Bunge, and P.J. Dyck (1985) Myelination determines the caliber of dorsal root ganglion neurons in culture. J. Neurosci. 5r1563-1569.