Ultrastructural changes in the nucleolus of facial motor neurons following axotomy during an early critical period in development

THE JOURNAL OF COMPARATIVE NEUROLOGY 312:132-144 (1991)

Ultrastructural Changes in the Nucleolus of Facial Motor Neurons Following

Axotomy During an Early Critical Period in Development

PAMELA CLARK, KATHRYN J. JONES, AND ARTHUR LAVELLE Department of Anatomy and Cell Biology (P.C., K.J.J., A.L.) and Department of Physical

Therapy (K.J.J.), University of Illinois, Chicago, Illinois 60612

ABSTRACT In this study, the effects of axotomy on the ultrastructure of the nucleolus and associated

organelles were examined in fetal, newborn, and early postnatal facial motoneurons of the hamster. Golden hamsters used for this study were the 14-day fetus, newborn (0 days; < 6 hr) and 2, 4, 7, and 9 days postnatal ages, with 3 animals per group. For prenatal surgeries, pregnant hamsters were anesthetized and the facial nerves severed in the fetuses via electrocautery through the uterine wall and amniotic membrane. For postnatal surgeries, the animals were anesthetized and the right facial nerve exposed and severed at its exit from the stylomastoid foramen. At the appropriate postoperative times, the animals were re- anesthetized and perfused-fixed. The facial nuclear groups were dissected and processed for routine electron microscopy. Microbody and coiled body frequencies were determined from the number of neurons containing these structures per number of neurons sampled per animal in each experimental or control group and subjected to statistical analysis. Nucleolar reactive changes that occurred during this developmental sequence fell into two major categories. The first category displayed by most injured cells consisted of an initial compacting of fibrillar material and reduction in vacuolar space. The second category appeared to represent a progression from this first stage of nucleolar reactivity into degenerative changes involving a striking segregation of nucleolar components into five distinct regions. The incidence of microbodies increased as a result of axotomy, whereas the presence of coiled bodies decreased at the later postoperative stages in the older animals. With increasing age and nucleolar maturation, the nucleolar reactive pattern became less pronounced and severe, and neuronal survival predominated. It appears, therefore, that the two categories of nucleolar changes following axotomy during early development correlate with changes observed in nucleoli under conditions of rRNA downregulation. It is hypothesized from these results that a key step in the ability of neurons to survive axotomy and successfully regenerate at these early developmental stages occurs at some point in ribosomal RNA transcription and/or processing. Complementary information at the molecular level concerning changes in nucleolar synthetic activity and ribosome production will be necessary to test this hypothesis.

Key words: regeneration, axon injury, microbody, coiled body

The greater vulnerability of neurons in very young animals to the deleterious effects of axotomy was recog- nized over a century ago by Vulpian and Gudden (Romanes, '46). These early investigators reported the nearly total loss of neurons from spinal and cranial motor nuclei on examina- tion several weeks or months after axon iniurv. It has since

shown to diminish with increasing developmental age, both prenatally and postnatally (Brodal, '40; Hess, '56; LaVelle and LaVelle, '59).

In hamster facial neurons, the 2 weeks following birth represent a period of rapid growth and cytological matura-

" . I

been well established that peripheral, and many central, immature neurons degenerate in much greater numbers

rate and extent of neuronal degeneration have also been

AcceptedMay 14, 1991, Address reprint requests to Dr. Kathryn J. Jones, Dept. of Physical

Therapy (MIC 8981, University of Illinois, 1919 W. Taylor St., 4th fl., than mature (see review by Lieberman, '74). The Chicago, IL 60612.

o 1991 WILEY-LISS, INC.

133 NUCLEOLAR MORPHOLOGY IN AXOTOMIZED DEVELOPING MOTOR NEURONS

tion, which is accompanied by changes in the pattern of reaction to injury. During this period the immature multi- ple nucleoli are succeeded by one large nucleolus, nuclear growth is accompanied by the development of numerous infoldings or "invaginations" of the nuclear envelope, and somal growth is associated with the production of large amounts of Nissl substance (LaVelle and LaVelle, '58a,b, '70; Kinderman and LaVelle, '76a; LaVelle and Buschmann, '83; Clark et al., '90). This normal developmental sequence is paralleled by a gradual increase in cell survival following axon severance. For example, 14-day fetal neurons have nearly completely disappeared within 48 hours after mot- omy (LaVelle and LaVelle, '59). Transection at birth takes 4 to 6 days for complete disappearance of the nuclear group (LaVelle and LaVelle, '58a). At 7 days postnatal, about 40% survive axotomy; toward the end of the second postnatal week, nearly 90% of the neurons survive transection (LaVelle and LaVelle, '58b). Thus the temporal relationship between normal nucleolar and nuclear maturation and the ability of facial neurons to survive axotomy has led to the hypothesis that neuronal survivability following injury is related to an increased capacity for structural and functional maintenance obtained during the early postnatal time period (LaVelle and LaVelle, '84). The nucleolus, which plays an integral role in the level of protein synthetic activities within cells (Busch and Smetana, '701, is, therefore, a key element in the regenerative capacities of various neuronal types. In a previous study, we have identified some of the earliest nucleolar and nuclear changes in axotomized hamster facial neurons that occur during the later postnatal and adult stages (Jones and LaVelle, '86).

At the present time, however, no ultrastructural study of the nucleolar reaction following axotomy during the fetal or early postnatal stages in facial neurons or any other cell type has been done. Previous ultrastructural work identifying developmentally regulated aspects of the axon reaction in young, injured motoneurons have focused at two levels: perisomatic (Borke, '82) and cytoplasmic (Borke, '83; Hall and Borke, '88). Our present report focuses primarily on ultrastructural changes in nucleolar morphology following axotomy of the facial motor neurons during the early, critical period of growth and maturation described above. Ultrastructural details of early, normal nuclear development in hamster facial neurons, pertinent to the present study on axotomy, have appeared in a previous publication (Clark et al., '90). The objective was to determine how the reactive changes in nucleolar structure and associated structures correlated with the gradual increase in the ability of these neurons to survive injury. Since the change from no survival to increasing survival following axotomy takes place between the middle of the first and the begin- ning of the second postnatal weeks, our selection of operative ages was chosen to bracket this period. The results expand and complement the ultrastructural study of the axon reaction in facial neurons at later postnatal stages and through adulthood (Jones and LaVelle, '86). With fine structural events occurring with injury outlined through the lifespan of the cell, the model now provides an excellent potential for future studies directed at elucidating molecular events associated with neuronal survival and regeneration after injury.

MATERIALS AND METHODS The golden hamster (Mesocricetus auratus) was espe-

cially suited to this developmental study because of its ease

of breeding and its brief and consistent length of gestation period (16 days), which involves rapid prenatal development and relative immaturity at birth. The young used for this series were the 14-day fetus, newborn (0 days <6 hr), and 2, 4, 7, and 9 days postnatal ages. Closely timed pregnancies, requisite for this developmental study, were obtained by placing a male with a female in estrus, observ- ing coitus, and leaving them together for 1 hour. The ages of the resulting litters were determined from the midpoint of the time during which the animals were housed together. Three or more animals of approximately equal weight from each age group were eventually chosen for study.

The schedule of operative ages and postoperative survival times utilized in this study is presented in Table 1. Previous work has shown that facial neurons withstand the effects of injury for longer periods of time when axotomized at successively later developmental ages (LaVelle and LaVelle, '84). Therefore, with increasing age at operation, longer postoperative survival times were selected. This was done to obtain the greatest degree of retrograde reaction prior to any significant amount of cell degeneration or death. Also, at later ages the longest postoperative time was limited to 4 days, since surviving neurons continue to develop and their reactive pattern merges into that of the next operative age. This limitation helped to avoid difficulties that would otherwise be brought about by overlapping reaction patterns. All surgeries were done in accordance with NIH guidelines for the use of laboratory animals.

Prenatal surgical procedure On the 14th day of gestation, pregnant hamsters were

anesthetized with sodium pentobarbital (Nembutal, Abbott Laboratories, 50 mg/ml; 0.14 m1/100 gm body weight). At this time the fetal heads and external ears can still be seen by transillumination through the uterine wall and amniotic membrane. The facial nerves of the fetuses were severed near the stylomastoid foramen by electrocautery, using a Wapler cold cautery unit in accord with the procedure of LaVelle and LaVelle ('59), which should be consulted for details. In summary, a mother's back and abdomen were cleaned with 70% ethanol and shaved. Her back was next soaked with a 0.9% saline solution and placed on a cold plate-electrode. The approach was via a midline incision in the abdomen. Axotomy in utero was then carried out by inserting through the uterine wall a 0.6 mm diameter needle, which provided a single, hot point for controlled electrocautery. This method allowed a more selective in utero severance of axons from a specific population of peripheral neurons than had been utilized by previous investigators (Hooker and Nicholas, '30; Hess, '56).

After the postoperative interval of 17 hours, the mother was given an overdose of Nembutal and the abdomen was reopened. Fetuses were removed, chilled to immobility on ice, and perfused individually. Fetuses that were undersize or showed evidence of hemorrhage or coagulated areas larger than 2 mm in diameter were discarded. The unoper-

TABLE I. Schedule of Operative Ages and Survival Times

Age of operation

Fetal day 14 17 hours Newborn ( < 6 hours) 24 hours Postnatal day 2 24 hours Postnatal day 4 24 hours; 2 days Postnatal day I 2 days; 4 days Postnatal day 9 2 days; 4 days

Postoperative survival times

134 P. CLARK ET AL.

handled. Statistical analysis was accomplished by one-way analysis of variance (Sokal and Rohlf, '81).

ated side served as a control for each animal. In addition, at least one intact fetus was sacrificed as a control along with the experimental animals. The difficulty in identifying sham-operated fetuses at the time of sacrifice prevented their use as controls in the fetal experimental groups.

Postnatal axotomy Animals were weighed and sexed. Newborn and 2-day-old

hamsters were anesthetized by cooling on ice until immobi- lized. All other experimental animals were anesthetized with sodium pentobarbital (Nembutal, Abbott Laborato- ries, 50 mg/ml; .014 m1/10 gm body weight). All operations were performed with the aid of a dissection microscope. An area behind and below the right ear was cleaned with 70% ethanol and the nerve was exposed and severed, with an iridectomy scissors, near its exit from the stylomastoid foramen. A piece of the distal nerve was removed to prevent reapposition of the nerve ends and the skin incision closed with surgical silk. Upon revival, the animals were placed back with their littermates. The nonoperated left side served as a control for each animal. In addition, sham operations were performed that were identical except that the nerve was not cut. In all postnatal animals, nerve section resulted in a slackening of the ipsilateral facial musculature with a cessation of vibrissal movements and a drooping of the corner of the mouth. Since eye opening does not occur prior to 15 days after birth, the eye-blink reflex could not be tested. No functional recovery in the peripheral field of the severed neurons was noted at the time of sacrifice in any of the experimental animals studied.

Prior to perfusion-fixation, the fetuses were removed from mothers anesthetized by intraperitoneal injections of sodium pentobarbital (Nembutal, Abbott Laboratories; 0.14 ml/lOO gm body weight). For perfusion, the fetuses and newborn animals were anesthetized by chilling with ice, whereas all other postnatal animals were anesthetized by intraperitoneal injection with Nembutal (.014 m1/10 gm body weight). All brains were perfusion-fixed via the heart by the two-step procedure of Peters ('70). The brainstems were removed 1 hour after perfusion, a stabilization period that may diminish the possibility of postfixation artifact (Cammermeyer, '60) and immersed in concentrated fixative overnight. Further processing included isolation of the facial motor nuclear groups with the aid of a vibratome, embedding in synthetic resin, and the cutting and staining of thin sections for electron microscopy (cf. LaVelle and Buschmann, '83; Jones and LaVelle, '85). Sections (ca. 70 nm) were picked up on 200-mesh copper grids and sequen- tially stained by saturated, alcoholic uranyl acetate, and by Reynold's ('63) lead citrate stain. Between 10-15 electron micrographs of experimental or control neurons per animal at each of the operative ages and survival times were collected and analyzed.

Nucleolar microbody frequency was defined as the percentage of neurons with nucleolar profiles containing microbodies, and coiled body frequency was defined as the percentage of neurons with coiled bodies attached to the nucleolus or free within the nucleoplasm. The mean microbody frequency values from all early ( I 24 hours) postoperative animals were combined as one group (cf. Table 1). All 2-day postoperative animals were combined as a second group and those from all 4-day postoperative animals were combined as a third group (cf. Table 1). Coiled body frequency values from the three groups were similarly

RESULTS Ultrastructurally, the normal nucleolus of the adult

facial neuron is comprised of anastomosing, densely fibrillar strands or "nucleolonema" that range in diameter from about 0.1 p n to 0.2 pm. The strands are separated by irregularly shaped spaces or vacuoles containing variable numbers of granules. Some of the strands surround less densely stained areas or fibrillar centers. The individual fibrils making up either the nucleolonemal strands or the lighter centers are about 3.7 nm in diameter. The mature nucleolus is strikingly characterized by a large, centrally positioned cluster of ribonucleoprotein granules. This cluster can attain a diameter of 2 km, or about half the diameter of nucleolus itself. Heterochromatin granules, or nucleolus-associated chromatin, are unevenly distributed as a shell about the nucleolus proper. Occasionally another associated structure, the coiled body, is attached to a nucleolonemal strand at the periphery of the nucleolus.

Of course, the nucleolus undergoes changes involving different configurations and proportions of materials at different developmental stages. During the first 2 postnatal weeks, the facial neuronal nucleolus changes from a com- pactly organized structure, with highly segregated components, to one with an open reticulated organization (Clark et al., '90). It is not until after 15 days postnatal age that the centrally located aggregate of ribonucleoprotein particles, typical of the nucleolus in the hamster facial neuron, begins to form. Illustrations and more ultrastructural detail concerning the mature nucleolus and its later developmental stages may be found in previous publications (Kinderman and LaVelle, '?6a,b; Jones and LaVelle, '86).

Nucleolar ultrastructure following axotomy Previous observations with the light microscope showed

that facial nerve axotomy in hamsters during fetal life and at birth resulted in initial nucleolar condensation and/or loss, followed by eventual cell death (LaVelle and LaVelle, '84). By 7 days postnatal age, however, axotomy resulted in chromatolysis without any appreciable nucleolar alterations in surviving neurons (LaVelle and LaVelle, '84). The gradual increase in neuronal survival after injury, therefore, appeared to be associated with developmental changes in nucleolar c,ytology and reaction pattern after injury. Light microscopy, however, could not reveal the more subtle cytoarchitectural changes that occurred within the nucleoli during this early sequence of reactions. The reactive changes that occurred during this developmental sequence of the facial neurons fell into two major categories. Most of the electron micrographs displayed initial alterations that preceded the more severe condensation stages of the reaction and that were fairly consistent, regardless of the level of nucleolar maturation at the time of injury. Differences in developmental age, however, did affect the length of the postoperative period required before the changes became apparent, since the younger the operative age, the earlier the changes were observed. These initial reactions were placed in the first category. The second category consisted of a few neurons that exhibited the striking, segregative or condensation changes, representative of terminal degenerative stages, similar to those seen at the light microscopic level (LaVelle and LaVelle, '58a, '59).

NUCLEOLAR MORPHOLOGY IN AXOTOMIZED DEVELOPING MOTOR NEURONS 135

First category. The earliest reactive changes in nucleoli were found 17 hours after axotomy on fetal day 14 (Fig. lA,B) and 24 hours after axotomy on postnatal days 0, 2, and 4 (representative micrographs from the 2-day postnatal group are illustrated in Fig. lC,D). Note that for Figures 1-3, the control nucleoli are illustrated in A and C. Segrega- tion of nucleolar components into distinct fibrillar and granular areas was initially observed (Fig. 1D) and was

followed by a compacting of nucleolar components and an increase in microbodies (Fig. lB,D). These bodies, in both control and experimental nucleoli, ranged from 55 to 150 nm in diameter and were surrounded by translucent halos of variable width. The microbodies were found in both the fibrillar and granular regions of nucleoli, although they appeared more frequently in the granular component of the injured neurons than in that of the corresponding controls.

Fig. 1. Nucleolar changes within 24 hours after axotomy on fetal day 14 and postnatal day 2. Nucleolus associated chromatin (NC, Nc); granular material (GI; dense fibrillar material (F) with an associated fibrillar center (Fc); vacuolar space (V); coiled bodies (CB); and microbodies (arrowheads) within the fibrillar and granular regions. The

magnification of A-D is indicated by the length of the line (1 pn) in A. A. Nucleolus of a control fetal neuron. B. Nucleolus of a neuron 17 hours after axotomy on fetal day 14. C. Nucleolus of a control postnatal day 3 neuron. D. Nucleolus of a neuron 24 hours after axotomy on postnatal day 2.

136 P. CLARK ET AL.

Fig. 2. Nucleolar changes 2 days after axotomy on postnatal days 4 and 9. Nucleolus associatedchromatin (Nc); dense fibrillar material (F); vacuolar space 0; granular material (G); fibrillar centers (Fc); microbodies (arrowheads) within the fibrillar and granular regions. The magnification of A-D is indicated by the length of the line (1 pm) in A.

A. Nucleolus of a control postnatal day 6 neuron. B. Nucleolus of a neuron 2 days after axotomy on postnatal day 4. C. Nucleolus of a control postnatal day 11 neuron. D. Nucleolus of a neuron 2 days after axotomy on pastnatal day 9.

A similar nucleolar reaction was seen at 2 days after axotomy on postnatal day 4 (Fig. 2A,B). By contrast, 2 days after axotomy on postnatal days 7 and 9, little or no nucleolar alterations were evident, when compared with control neurons (representative micrographs from the 9-day postnatal group are illustrated in Fig. 2C,D). However, after the longer postoperative period of 4 days, nucleolar alterations, similar to that in the younger ages, were seen in neurons axotomized on postnatal days 7 and 9 (Fig. 3A-D).

The 7- and 9-day operated cells also showed signs of RER dispersion or chromatolysis (not illustrated), with some tendency toward nuclear eccentricity and flattening.

The general increase in nucleolar compactness was con- comitant with a reduction in vacuolar space. The nucleolonema of the injured neurons also appeared less well developed than normal, and the dense fibrillar portion took on a blocky appearance. This was in contrast to control nucleoli in which the interlocking network of nucleolonemal strands

NUCLEOLAR MORPHOLOGY IN AXOTOMIZED DEVELOPING MOTOR NEURONS

Fig. 3. Nucleolar changes 4 days after axotomy on postnatal days 7 and 9. Nucleolus associated chromatin (Nc); dense fibrillar material (F); vacuolar space (V); granular material (GI; fibrillar centers (Fc); microbodies (arrowheads) within the fibrillar and granular regions. The magnification of A-D is indicated by the length of the line (1 ym) in A.

were delineated by vacuolar regions (compare A and B in Figs. 2 and 3 and C and D in Fig. 3) . At all ages, the granular component of the injured neurons appeared more compact than normal and failed to become integrated into the nucleolonemal network.

Although microbody frequency was elevated within 24 hours after axotomy on fetal day 14 and postnatal days 0,2, and 4, the increase was not statistically significant (Fig. 4). Microbody frequency was unchanged 2 days following axotomy on postnatal days 4, 7, and 9. However, at the

137

A. Nucleolus of a control postnatal day 11 neuron. B. Nucleolus of a neuron 4 days after axotomy on postnatal day 7. C. Nucleolus of a control postnatal day 13 neuron. D. Nucleolus of a neuron 4 days after axotomy on postnatal day 9.

longest postoperative interval of 4 days, there was a statistically significant (p < 0.05) increase in microbody frequency after axotomy on postnatal days 7 and 9 (Fig. 4).

The number and size of fibrillar centers in the reacting nucleoli remained variable, although there appeared to be a trend toward a reduction in number with an increase in size when compared to control neurons (Figs. 2A-D, 3C,D). Because of this variability, counts of fibrillar numbers were not attempted. The largest fibrillar centers (75 nm) were found in a few nucleoli taken from neurons injured on

138 P. CLARK ET AL.

100

%

N a0 e

r 0 60 n S

w 40 i t h

M B

U

20

0

MICROBODY FREQUENCY

24 hours 2 days 4 days

Postoperative Times

Control Axotomy

Fig. 4. Changes in the microbody frequency dwtribution I 24 hours following axotomy on fetal day 14 and postnatal days 0,2, and 4 days; 2 days following axotomy at 4, 7, and 9 days; and 4 days following axotomy at 7 and 9 days postnatal. See Materials and Methods for additional details. Vertical lines = standard errors.

postnatal day 9, with a 2-day survival time (Fig. 2C,D), which was the only change observed in the otherwise normal appearing nucleoli of that experimental group.

This category was evident in a few electron micrographs collected from the injured facial nuclei of animals axotomized on fetal day 14 and postnatal days 0, 2 and 4. These neurons showed striking nucleolar changes overtly comparable to those seen at the light microscopic level (LaVelle and LaVelle, '58, '59). The representative nucleus, shown in Figure 5A and taken from an animal sacrificed 24 hours after axotomy on postnatal day 4, has undergone nucleolar segregation such that the various nucleolar components have separated into roughly spherical masses. The surrounding nucleoplasm appears otherwise largely unchanged, although the cytoplasm is slightly more electron dense than normal due to the presence of numerous free ribosomes. Figure 5B shows a neuron taken from an animal sacrificed 2 days after injury at 4 days. This neuron represents the more advanced stage of reaction, with further condensation of segregated nuclear constituents and an increased electron density of the nucleoplasm. There was also a further increase in the electron density of the cytoplasm.

The segregated nucleolus shown at higher magnification in Figure 6 was taken from an animal sacrificed 24 hours

Second category of nucleolar changes.

after axotomy on postnatal day 0. Five structural components, or regions, were found after nucleolar segregation. These were: (1) a dense, fine fibrillar component (df), (2) a light, fine fibrillar component (If), (3) a dense, coarse fibrillar component (dc), (4) a heterogeneous region (hr), and (5) vacuoles (V). These structural components, as seen in different nucleoli, were organized in a fairly consistent pattern. The dense, coarse fibrillar component (dc) usually comprised the largest spherical mass, as well as some additional smaller ones. This material was associated with one or more somewhat smaller masses of dense, fine fibrillar material. The latter had an electron density and internal fibrillar structure similar to that of the dense fibrillar material of a normal nucleolus. A small amount of light, fine fibrillar material, similar in character to the fibrillar centers of a normal nucleolus, was frequently associated with the dense, fine fibrillar material. The light fine fibrillar material could be found tightly interposed between the dense fine and the dense coarse fibrillar material. The heterogeneous region was composed of loosely arranged fibrillar and granular material. In addition, there were somewhat larger particles, which ranged in size from that of perichromatin granules (50 nm) to that of microbodies (55 to 150 nm). The small amount of vacuolar space, found in the segregated nucleoli, was associated with the heterogeneous region. This last region was loosely interposed between the masses of dense fine fibrillar material and the smaller masses of dense coarse fibrillar material, resulting in a fairly regular spatial arrangement (see also Fig. 5A).

Changes in nucleolus-associated structures following axotomy

With normal development, the nucleolus-associated chromatin becomes irregularly distributed about the periphery of the nucleolus (LaVelle and LaVelle, '84). During the period studied, we detected no integral alterations in this chromatin resulting from injury. Coiled bodies also were observed to be attached to the periphery of nucleoli in some of the axotomized neurons of the present study. Figure 7A,B shows low and high magnification views of a coiled body from a neuron axotomized on fetal day 14 with a 17-hour survival time. A high magnification view of a coiled body from a neuron axotomized on postnatal day 4 with a 24-hour survival time is also shown (Fig. 7 0 . The coiled bodies of these neurons structurally appeared no different from those seen in control neurons, in agreement with the results of Kinderman and LaVelle ('76b) on coiled bodies in hamster facial neurons axotomized at much later postnatal and adult ages.

No difference in coiled body frequency from control cells was found within 24 hours after axotomy on fetal day 14 or postnatal days 0, 2, and 4 (Fig. 8). In contrast, significant reductions in coiled body frequency (p < 0.05) were found 2 days after axotomy on postnatal days 4,7, and 9, and 4 days after axotomy on postnatal days 7 and 9 (Fig. 8). According to Kinderman and LaVelle ('76b), coiled bodies also appeared to be numerically decreased in axotomized adult facial neurons.

DISCUSSION During the first 2 weeks of postnatal life, hamster facial

neurons undergo a progressive, critical change in their


Fig. 5. Nucleolar segregation after axotomy. A. This neuron, axotomized onpostnatal day 4 with a 24-hour survival time, shows segregation of the nucleolar constituents (arrowheads) with an otherwise normal appearing nucleus. The cytoplasm has an increased electron- densitydue to the presence of many free ribosomes (asterisk). Length of line = 2 Fm. B. This neuron, axotomized on postnatal day 4 with a

2-day survival time, shows a further condensation of nucleolar and nuclear constituents (arrowheads). The nucleus appears shrunken, and there is an increase in the electron-density of the nucleoplasm as well as a further increase in the electron-density of the cytoplasm (asterisk). Note: Line in A indicates magnification in B.

pattern of reaction to injury. hotomy during late gestation and early in the first postnatal week resulted in an initial compacting and segregation of nucleolar components in most cells (first category changes). In a few cells severe condensation and segregation of nucleolar elements occurred (second category changes). As nucleolar maturation proceeded during the middle of the developmental span studied, the chromatolytic type of reaction appeared. This later reaction type was associated with an increasingly longer postoperative interval before the first category changes appeared and without the distinctive second cate-

gory changes. As shown in earlier studies (LaVelle and LaVelle, '58b, '59), there is a gradual increase in facial neuron survival after axotomy during the first 2 postnatal weeks.

In the first category, nucleoli became more compact than control nucleoli due to a reduction of vacuolar space and a condensation of granular material. The nucleolonema was not as extensively developed and the dense fibrillar material retained a blocky appearance. The number of microbodies appeared to increase somewhat in the younger injured neurons and then significantly so in the older ones. This set

140 P. CLARK ET AL.

Fig. 6. Five structural components found after nucleolar segregation resulting from axotomy. The segregated nucleolus, from a neuron axotomized on postnatal day 0 with a 24-hour survival time, has roughly spherical masses of a dense, fine fibrillar component (df and of a dense, course fihrillar component (dc). A light, fine fihrillar (If) component can be seen interposed between the df and dc components. The heterogeneous region (hr) is composed of loosely arranged fibrillar and granular material, including spherical particles of fibrillar material that range in size from 25 nm to 80 nm (arrowheads). Associated with the heterogeneous region are vacuoles (v). Length of line = 1 pm.

of alterations involved most of the neurons in any nuclear field on the affected side.

Several investigators have reported the formation of microbodies (or microspherules) during the course of segregation induced by various drug regimens (Unuma and Busch, '67; Recher et al., '71; Merski et al., '76; Daskal et al., '78). Their formation appears to involve the sequestra- tion of certain nucleolar fibrillar elements of high electron density (Daskal, '79). The presence of microbodies in pyramidal cells of the hamster cortex has been shown to correlate with two critical phases of development during which altered levels of neuronal metabolism were predicted (Buschmann and LaVelle, '83). In the present investiga- tion, microbodies were found in the nucleoli of control facial neurons with approximately equal frequency throughout the growth period, and so their presence did not appear to correlate with increasing levels of nucleolar maturation or stage specific metabolic demands. Microbody frequency was somewhat elevated after injury to fetal and early postnatal neurons and significantly increased 4 days after axotomy at older ages. The lack of statistically significant increases in

microbody frequency in the youngest injured neurons may be related to the rapid progression from early to late stages of nucleolar segregation. In contrast, the older neurons tend to be more resistant to the factors responsible for inducing nucleolar segregation, possibly because of greater amounts of active nucleolar DNA, and thus would remain in a state of early or partial segregation for a longer period. This would allow more consistent detection of microbodies, resulting in the significant increase in frequency observed after injury to the older animals.

The number and size of the fibrillar centers in the injured neurons also changed, such that there was a trend toward a reduction in number along with an increase in size. This is in contrast to the trend observed in control or normal neurons (Clark et al., '90) wherein fibrillar centers tended to increase in number and decrease in size. There is some evidence suggesting that fibrillar centers are not sites of active rDNA or RNA transcription (Dupuy-Coin et al., '86; Wachtler et al., '89). Various cytochemical tests indicate that the fibrillar centers contain rDNA and enzymes responsible for ribosomal RNA transcription, thus supplying evidence that the centers are the site of ribosomal genes (Derenzini et al., '90; Reeder, '90; Scheer and Benavente, '90). Previously, it has been suggested that these centers might also serve as reservoirs (Mirre and Knibiehler, '81) or sites of transfer of inactive rDNA Wagner-Capodano et al., '84; Cataldo et al., '881, or as storage sites of proteins in ribosome processing and transport (Knibiehler et al., '83). Such putative functions could be compatible with the observed reversal of the normal sequence of fibrillar center development, following axotomy, in the hamster neurons. Whereas the function(4 of the fibrillar centers is not entirely clear, recent stereological data suggest that nucleolar inactivation is accompanied by a decrease in number and/or increase in size of the fibrillar centers (Jordan and McGovern, '81). This correlation fits with both our observations and hypothesis concerning the state of nucleolar activity within each of the defined categories and suggests that an interesting future direction would be a stereological analysis of each of the individual components of the nucleolus following axotomy during development.

The second category of nucleolar changes was characterized by the segregation of nucleolar components into distinct, roughly spherical masses. These segregative changes were present in relatively few neurons from animals axotomized on fetal day 14 and postnatal days 0 , 2 and 4, but not at later postnatal stages. Initial compacting of nucleoli appeared to be the earliest changes. Yet, second category changes were also found at the earliest stages. These latter must follow the initial compacting stage and may involve neurons that were "born" closer to the time of injury (i.e., the most immature) and thus were more susceptible to injury within the same population. Therefore, this sequence of variability may be related to the time of origin of individual neurons. A similar sequence of nuclear changes was observed in fetal and neonatal hamsters, using light microscopy (LaVelle and LaVelle, '%a, '59). These earlier studies showed that normal nucleoli were replaced by sharply rounded profiles, which stained intensely with basic dye and by the Feulgen technique for DNA. Our present study shows that the segregated masses include at least five distinctly organized components.

The general pattern of structural change observed in the severely segregated nucleoli was similar to that reported in other systems (Bernhard et al., '65; Simard and Bernhard,


Fig. 7. Coiled body ultrastructure. A. A low magnification view of a neuron, axotomized on fetal day 14 with a 17-hour survival time, shows 3 immature nucleoli (Nu), one of which has an attached coiled body (arrowhead). Length of line = 2 pm. B. A high magnification view of the

coiled body (cb) in A. Length of line = 0.5 pm. C. A high magnification view of a nucleolus with an attached coiled body (arrowhead) from a neuron axotomized on postnatal day 4, with a 24-hour survival time. Length of line = 1 pm.

’66; Busch and Smetana, ’70; Recher et al., ’71; Torvik and Skjorten, ’74; Dimova et al., ’79; Benavente et al., ’87, ’88). A precise correlation between each of the residual nucleolar components found here, and those reported by others, could not be made. However, on the basis of immunolocalization studies with antibodies to RNA polymerase I (Hadjiolov et al., ’86), the light, fine fibrillar areas (10 in Figure 6 most likely represent “normal” fibrillar centers segregated from surrounding nucleolonemal strands or dense fibrillar material (df in Fig. 6). Further studies will be necessary to determine the exact composition of these five segregated components, which differ in electron density and ultrastructure. Nucleolar segregation is a well-established cellular phenomenon related to the inhibition of ribosomal RNA synthesis (for reviews see Daskal, ’79; Bouteille et al., ’82). It has been associated most frequently, although not exclu-

sively, with the effects of antimetabolic drugs capable of binding to rDNA in a way that interferes with its template activity in RNA synthesis (Daskal, ’79).

The question arises as to whether the two categories of nucleolar changes represent distinct events in or successive phases of a common mechanism of nucleolar inactivation. On the one hand, it could be that the first category of nucleolar changes reff ects an interference in the processing of rRNA particles, whereas the second category of nucleolar alterations reflects a block in rRNA transcription. On the other hand, it may be that the two categories represent differential degrees of a single process initiating with transcriptional inactivation. Of particular interest in this regard were the findings of the effects of injury on the dense fibrillar component and microbody frequency. Dimova et al. (‘79) showed pronounced alterations in the dense fibrillar

142

40

% 35-

N e 3O- U

0 25-

n s 2 0 -

i t h 10-

C B '-

1 5 -

P. CLARK ET AL.

~

COILED BODY FREQUENCY

0- 24 hours 2 days 4 days

Postoperative Times

Control Axotomy

Fig. 8. Changes in coiled body frequency 1 2 4 hours following axotomy on fetal day 14 and postnatal days 0, 2, and 4 days; 2 days following axotomy at 4,7, and 9 days; and 4 days following axotomy at 7 and 9 days postnatal. See Materials and Methods for additional details. Vertical lines = standard errors.

components and induction of microbodies in rat hepatocytes following inhibition of rRNA transcription by D-galactosamine. Also, axotomy-induced changes in the dense fibrillar component are very similar to the initial changes observed by Recher et al. ('71) after treating a cultured human cell line with actinomycin D for 30 minutes. Contin- ued treatment with this drug for an additional 30 minutes resulted in segregation of the nucleolar components similar to that seen in facial neurons exhibiting the second category of changes. Goldblatt and Sullivan ('70), in studies of rat hepatic cells, found that the extent of nucleolar segregation induced by treatment with actinomycin D and the degree to which it was reversible was dose dependent. On the basis of their findings, they suggested that a critical amount of actinomycin D must be bound to the nucleolar DNA (thus inactivating it) before typical segregation is induced. They further suggested that partial segregation reflects inactivation of a lesser amount of DNA. Thus evidence correlating both categories with downregulation of rRNA transcription exists in other cell types.

Connections between the nucleolus and nucleolus- associated chromatin have been observed, suggestive of metabolic interrelationships (Goessens, '84; Lafarga, '89). Axotomy, however, resulted in no alterations in nucleolus- associated chromatin in these early ages. The nucleolus- associated chromatin does become more thinly dispersed

about the nucleolus after axotomy at about 20 days postnatal and later ages (LaVelle and LaVelle, '58b; Jones and LaVelle, '86a). This is coincident with the nucleolar swelling that is characteristic of the injurious reaction at these later ages.

An additional change, involving a nucleolus-associated entity, the coiled body, was observed. Although coiled body ultrastructure was not affected at any time by injury, decreased numbers of these structures resulted from axotomy on postnatal day 4 or later. This finding is in agreement with previous work which indicated that axotomy of adult hamster facial neurons resulted in fewer coiled bodies (Kinderman and LaVelle, '76b). Most evidence indicates that the coiled bodies consist primarily of ribonucleoprotein, lack cytochemically detectable DNA, and probably are not involved in RNA transcription (Hardin et al., '69; Monneron and Bernhard, '69; Zareba-Kowalska, '89). If coiled bodies are involved in some aspect of rRNA processing or transport, as has been suggested on the basis of their morphological and cytochemical characteristics (Hervas et al., '80; Moreno Diaz de la Espina et al., '82; Seite et al., '82)' the reduced incidence of coiled bodies in injured neurons might be related to alterations in post-transcriptional regulation of rRNA.

Nucleolar morphology is a sensitive index of the metabolic status of a cell, particularly of the level of rRNA transcription and processing (Busch and Smetana, '70; Goldblatt and Sullivan, '70; Miller and Gonzalez, '76; Goessens, '84; Hadjiolov, '85). Axotomized adult hamster facial motor neurons undergo nucleolar hypertrophy with (1) a loosening of the nucleolonemal strands, (2) an increase in vacuolar space, and (3) a dispersion of granular preriboso- ma1 material (finderman and LaVelle, '76a; Jones and LaVelle, '86a). They withstand the deleterious effects of axotomy for very long periods of time (McLoon and LaVelle, '81). This contrasts with the condensation and segregative sequelae of nucleolar alterations in the very young neurons described here. Both the nucleolar changes, as well as the presence of dispersed, single ribosomes, suggest an inactivation of neuronal protein synthesis (Palay et al., '74) associated with axotomy in the younger neurons. These differences are indicative that the degree of nucleolar and RER maturation achieved at a given time during development plays a significant role in the type of reaction resulting from injury at that time. An interesting future direction could involve a stereological analysis of the cytoplasmic rRNA components and correlation with the nucleolar data. Config- urational changes in RNA-containing elements of the cytoplasm have been described in considerable details in many different axotornized systems (Lieberman, '741, but to date no quantitative approach to the study of these changes has been done. The results further demonstrate that opera- tional age must be strictly controlled. A newborn animal, for example, cannot be taken as any age during the first or second week of birth. Postoperative times should not be overly extended in time, since overlap or confluence of reactions with those peculiar to a later operative age can obscure or at most make an interpretation of reaction patterns difficult.

In conclusion, two categories of nucleolar changes following axotomy to facial motor neurons during early development have been described. These involved inceptive ultrastructural alterations as well as pronounced nucleolar segregation similar to that observed under conditions of rRNA downregulation. With increasing age and nucleolar


maturation, the nucleolar reactive pattern described above became less pronounced and neuronal survival predominated. The results lead us to hypothesize that a key step in the ability of neurons to survive axotomy and successfully regenerate at these early developmental stages occurs at some point in ribosomal RNA transcription and/or processing. Complementary information at the molecular level concerning changes in nucleolar synthetic activity and ribosome production will be necessary to test this hypothesis.

LITERATURE CITED Benavente, R., K.M. Rose, G. Reimer, B. Hugle-Dorr, and U. Scheer (1987)

Inhibition of nucleolar reformation after microinjection of antibodies to RNA polymerase I into mitotic cells. J. Cell Biol. 105:1483-1491.

Benavente, R., G. Reimer, K.M. Rose, B. Hugle-Dorr, and U. Scheer (1988) Nucleolar changes after microinjection of antibodies to RNA polymerase I in the nucleus of mammalian cells. Chromosoma 97:115-123.

Bernhard, W., C. Frayssinet, C. Lafarge, and E. Lebreton (1965) Lesions nucleolaires precoces provoquees par l'aflatoxine das les cellules hepa- tiques du Rat. C.R. Acad. Sci. 261:1785-1788.

Borke, R.C. (1982) Perisomatic changes in the maturing hypoglossal nucleus after axon injury. J. Neurocytol. 11t463-485.

Borke, R.C. (1983) Intrasomatic changes in the maturing hypoglossal nucleus after axon injury. J. Neurocytol. 12873-883.

Bouteille, M., D. Hernandez-Verdun, A. Dupuy-Coin and C.A. Bourgeois (1982) Nucleoli and nucleolar-related structures in normal, infected and drug treated cells. In E.G. Jordan and C.A. Cullis (eds): The Nucleolus. New York: Cambridge University Press.

Brodal, A. (1940) Modification of Gudden method for study of cerebral localization. Arch. Neur. Psychiat. Chic. 43t46-58.

Busch, H. and K. Smetana (1970) The Nucleolus. New York: Academic Press.

Buschmann, MB.T., and A. LaVelle (1983) Morphometry of nuclei, nuclear envelopes and nucleoli in aging hamster cerebrum. Neurobiol. Ageing 4: 197-202.

Cammermeyer, J. (1960) The postmortem origin and mechanism of neuronal hyperchromatosis and nuclear pyknosis. Exp. Neurol. 2379405.

Cataldo, C., C. Souchier, and A. Stahl(1988) Three-dimensional ultrastructure and quantitative analysis of the human Sertoli cell nucleolus. Biol. Cell 63:277-285.

Clark, P., K.J. Jones, and A. LaVelle (1990) An ultrastructural and morphometric analysis of nucleolar and nuclear changes during the early growth period in hamster facial neurons. J. Comp. Neurol. 302749-760.

Daskal, Y. (1979) Drug effects on nucleolar and extranucleolar chromatin. In S.T. Crooke, Y. Daskal and H. Busch (eds): Effects of Drugs on the Cell Nucleus. New York Academic Press, pp. 107-125.

Daskal, Y., C. Woodward, S.T. Crooke and H. Busch (1978) Comparative ultrastructural studies of nucleoli of tumor cells treated with adriarnycin and the newer anthracyclines, carminomycin and marcellomycin. Cancer Res. 38t467-473.

Dimova, R.N., K.C. Gajdardjieva, M.D. Dabeva, and A.A. Hadjiolov (1979) Early effects of D-galactosamine on rat liver nucleolar structures. Biol. Cell 35: 1-9.

Derenzini, M., M. Thiry, and G. Goessens (1990) Ultrastructural cytochem- istry of the mammalian cell nucleolus. J. Histochem. Cytochem. 38r1237- 1256.

Dupuy-Coin, A.-M., M.J . Pebusque, R. Seite and M. Bouteille (1986) Localization of transcription in nucleoli of rat sympathetic neurons: A quantitative ultrastructural autoradiography study. J. Submicrosc. Cy- tol. 18:21-27.

Goessens, G. (1984) Nucleolar structure. Int. Rev. Cytol. 87t107-158. Goldblatt, P., and R.J. Sullivan (1970) Sequential morphological alterations

in hepatic cell nucleoli induced by varying doses of actinomycin D. Cancer Res. 30:13461349.

Hadjiolov, A.A. (1985) The nucleolus and ribosome biogenesis. In: Cell Biology Monographs, Vol. 12. New York Springer-Verlag, pp. 268.

Hadjiolova, K., K.M. Rose, and U. Scheer (1986) Immunolocalization of nucleolar proteins after D-galactosamine-induced inhibition of transcription in rat hepatocytes. Exp. Cell Res. 165481493,

Hall, L.L., and R.C. Borke (1988) A morphometric analysis of the somata and

organelles of regenerating hypoglossal motoneurons from the rat. J. Neurocytol. 17:835-844.

Hardin, J.W., S.S. Spicer, and W.B. Green (1969) The paranucleolar structure, accessory body of Cajal, sex chromatin, and related structures in nuclei of rat trigeminal neurons: A cytochemical and ultrastructural study. Anat. Rec. 164:403-432.

Hervas, J.P., J. Villegas, D. Crespo, and M. Lafarga (1980) Coiled bodies in supraoptic nuclei of the rat hypothalamus during the postnatal period. Am. J. Anat. 159:447-454.

Hess, A. (1956) Reactions of mammalian fetal spinal cord, spinal ganglia and brain to injury. J. Exp. Zool. 13.2349-389.

Hooker, D., and J.S. Nicholas (1930) Spinal cord section in rat fetuses. J. Comp. Neurol. 503413467,

Jones, K.J., and A. LaVelle (1985) Changes in nuclear envelope invaginations in axotomized immature and mature hamster facial motoneurons. Dev. Brain Res. 21:241-249.

Jones, K.J., and A. LaVelle (1986a) Differential effects of axotomy on immature and mature hamster facial neurons: A time course study of initial nucleolar and nuclear changes. J. Neurocytol. 15: 197-206.

Jones, K.J., and A. LaVelle (198613) Ultrastructural changes in the nucleoplasm of hamster facial neurons during a postnatal maturation period. Brain Res. 377:119-126.

Jordan, E.G., and J.H. McGovern (1981) The quantitative relationship of the fibrillar centres and other nucleolar components to changes in growth conditions, serum deprivation and low doses of actinomycin D in cultured diploid human fibroblasts (strain MRC-5). J. Cell Sci. 5.2373- 389.

Kinderman, N.B., and A. LaVelle (1976a) Ultrastructural changes in the developing nucleolus following axotomy. Brain Res. 1081237-247.

Kinderman, N.B., and A. LaVelle (1976b) A nucleolus-associated coiled body. J. Neurocytol. 5~545-550.

Knibiehler, B., C. Mirre, and A. Stahl (1983) Multiplication of nucleolar fibrillar centres and absence of rDNA amplification in mouse oocytes during meiotic prophase I. Tissue Cell 15159-166.

Lafarga, M., M.T. Berciano, J.P. Hervas, and J. Villegas (1989) Nucleolar organization in granule cell neurons of the rat cerebellum. J. Neurocytol. 18t19-26.

LaVelle, A,, and MB.T. Buschmann (1983) Nuclear envelope invaginations in hamster facial motor neurons during development and aging. Dev. Brain Res. 10:171-175.

LaVelle, A,, and F.W. LaVelle (1958a) Neuronal swelling and chromatolysis as influenced by the state of cell development. Am. J. Anat. 102t219-241.

LaVelle, A., and F.W. LaVelle (1958b) The nucleolar apparatus and neuronal reactivity to injury during development. J. Exp. Zool. 137:285-315.

LaVelle, A., and F.W. LaVelle (1959) Neuronal reaction to injury during development: Severance of the facial nerve in utero. Exp. Neurol. 1:82-95.

LaVelle, A,, and F.W. LaVelle (1984) Neuronal reaction to injury during development. In S. Finger and C.R. Almli (eds): Early Brain Damage, Neurobiology and Behavior. New York: Academic Press, pp. 3-16.

Lieberman, A.R. (1974) Some factors affecting retrograde neuronal re- sponses to axonal lesions. In R. Bellairs and E.G. Gray (eds): Essays on the Nervous System. London: Clarendon, pp. 71-105.

McLoon, L.K., and A. LaVelle (1981) Long-term effects of regeneration and prevention of regeneration on nucleolar morphology after facial nerve injury during development. Exp. Neurol. 73:762-774.

Merski, J., Y. Daskal, and H. Busch (1976) Effects of adriamycin on ultrastructure of nucleoli in the heart and liver cells of the rat. Cancer Res. 36:1580-1584.

Miller, L., and F. Gonzales (1976) The relationship of ribosomal RNA synthesis to the formation of segregated nucleoli and nucleolus-like bodies. J. Cell Biol. 71r939-949.

Mirre, C., and B. Knibiehler (1981) Ultrastructural autoradiographic localization of the rRNA transcription sites in the quail nucleolar components using two RNA antimetabolites. Biol. Cell 4273-78.

Monneron, A,, and W. Bernhard (1969) Fine structural organization of the interphase nucleus in some mammalian cells. J. Ultrastruct. Res. 27:266-288.

Moreno Diaz Dela Espina, S., F.J. Medina, and M.C. Risueno (1982) Ultrastructural, cytochemical and autoradiographic characterization of coiled bodies in the plant cell nucleus. Biol. Cell 44:229-238.

Palay, S.L., S. Billings-Gagliardi, and V. Chan-Palay (1974) Neuronal perikarya with dispersed, single ribosomes in the visual cortex of Macaca nulatta. J. Cell Biol. 63:1074-1089.

144 P. CLARK ET AL.

nucleolaire: specifite d'action de certains antimetabolites. J. Cancer k463-479.

Sokal, R.R., and F.J. Rohlf (1981) Biometry. San Francisco: W.H. Freeman, pp. 859.

Torvik, A,, and F. Skjorten (1974) The effect of actinomycin D upon normal neurons and retrograde nerve cell reaction. J. Neurocytol. 3:87-97.

Unuma, T., and H. Busch (1967) Formation of microspherules in nucleoli of tumor cells treated with high doses of actinomycin D. Cancer Res. 27:1232-1242.

Vagner-Capodano, A.-M., A S . Henderson, S. Lissitzsky, and A. Stahl(1984) The relationships between ribosomal genes and fibrillar centers in thyroid cells cultivated in uitro. Biol. Cell 51:ll-22.

Wachtler, F., M. Hartung, M. Devictor, J. Wiegant, A. Stahl, and H.G. Schwarzacher (1989) Ribosomal DNA is located and transcribed in the dense fibrillar component of human Sertoli cell nucleoli. Exp. Cell Res. 184;61-71.

Zareba-Kowalska, A. (1989) Cytochemical observations of the coiled bodies in neurons of rat sympathetic ganglia. Histochemistry 91251-256.

Peters, A. (1970) The fixation of central nervous tissue and the analysis of electron micrographs of neuropil, with special reference to the cerebral cortex. In W.J.H. Nauta and S.O.E. Ebbesson (eds): Contemporary Research Methods in Neuroanatomy. New York Springer-Verlag, pp. 56-76.

Recher, L., L.G. Briggs and N.T. Perry (1971) A reevaluation of nuclear and nucleolar changes induced in uitro by actinomycin-D. Cancer Res. 31:140-151.

Reeder, R. (1990) rRNA synthesis in the nucleolus. Trends Gen. 6:390-394. Reynolds, E.S. (1963) The use of lead citrate at high pH as an electron-

opaque stain in electron microscopy. J. Cell Biol. 17.208-212. Romanes, G.J. (1946) Motor localization of the effects of nerve injury on the

ventral horn cells of the spinal cord. J. Anat. 80:117-131. Scheer, U. and R. Benavente (1990) Functional and dynamic aspects of the

mammalian nucleolus. BioEssays 12:14-21. Seite, R., M.J. Pebusque, and M. Vio-Cigna (1982) Argyrophilic proteins on

coiled bodies in sympathetic neurons identified by Ag-NOR procedure. Biol. Cell 46:97-100.

Simard, R., and W. Bernhard (1966) Le phenomene de la segregation

Documents

Ultrastructural changes in the nucleolus of facial motor neurons following axotomy during an early critical period in development