Acta
Acta neuropath. (Berl.) 35, 93 --107 (1976)
Neuropathologica
9 by Springer-Verlag1976
Ultrastructure of the Gunn Rat Substantia Nigra
I. Cytoplasmic Changes H. K. Batty and O. E. Millhouse Departments of Anatomy and Neurology, College of Medicine, University of Utah, 50 North Medical Drive, Salt Lake City, Utah 84132, U.S.A.
Summary. The substantia nigra of various aged hyperbilirubinemic (Gunn) rats was studied by means of electron microscopy. The cytological features observed in the neuronal somata were the presence of (1) complex membranous bodies (CMBs), (2) dilated cisternae of granular endoplasmic reticulum, (3) single membrane bound vacuoles and (4) enlarged mitochondria. Nearly every neuronal soma studied in two week old Gunn rats contained CMBs, which consisted of several layers of membrane that usually, but not always, surrounded small islands of cytoplasm. On occasion CMBs were seen to be directly connected with granular endoplasmic reticulum and, in a few instances, they were located within a cistern of endoplasmic reticulum. There were significantly fewer CMBs in the neuronal somata of adult Gunn rats. They were located peripherally in the somata or in the proximal portions of dendrites. Furthermore, in these animals the cytoplasm appeared normal and did not exhibit the features commonly seen in younger animals. Only a few hyperchromatic neurons were observed and no pronounced gliosis was evident. Therefore it is assumed that the majority of substantia nigra neurons recovered from the toxic effect of bilirubin or that the concentration of bilirubin deposited in the substantia nigra is not sufficient to be lethal. The hypothesis that is considered is that CMBs represent autophagic activity which results from exposure of neurons to bilirubin. The adjacent neurites and glia did not demonstrate the cytoplasmic changes that were characteristic of the neuronal somata.
Key words: Substantia n i g r a - Gunn rat - - Complex membranous bodies -
-
Bilirubin - - Electron microscopy.
INTRODUCTION The Gunn rat (Gunn, ]944) was used as a model to study the effects of bilirubin on the central nervous system. These rats lack the genetic information to code for the liver enzyme glucuronic acid transferase (Carbone and Grodsky, 1957; 7
Acta neuropath. (Berl.) ~d. 35
94
H.K. Batty and O. E. Millhouse
Schmid et al., 1958). As a result the animals develop a non-hemolytic hyperbilirubinemia (Johnson et al., 1959, 1961; Blanc and Johnson, 1959; Johnson, 1961 ; Gardner and Konigsmark, 1969). Bilirubin deposition within the brain has been commonly associated with the kernicterus syndrome (Ernster et al., 1957; Blanc and Johnson, 1959; Johnson et al., 1959; Blanc, 1961; Johnson, 1961). The lemon yellow stain seen in several areas of the Gunn rat brain (see Schutta and Johnson, 1967) until about the twentieth post-natal day corresponds to material of the same color which Waters et al. (1954) analyzed spectrophotometrically. On the basis of its spectrophotometric curve in comparison with the curve for bilirubin, the lemon yellow stain was determined by these authors to represent bilirubin. The bilirubin encephalopathy begins to develop in newborn Gunn rats 3--4 days old (Johnson et al., 1959; Blanc, 1961; Schutta and Johnson, 1967). Several explanations have been offered to account for the deposition of bilirubin in these young brains. Waters and Britton (1955) and Schutta and Johnson (1967) argue that it is due to the immaturity of the blood-brain-barrier. However others have suggested that it is correlated with low levels of serum albumin (Diamond and Schmid, 1966) and with blood pH (Odell, 1961, 1966). Schutta and Johnson (1967) studied the ultrastructure of the cerebellum in Gunn rats ranging in age from a few days to 18 months old. Their observations were substantiated by the subsequent investigation of Silberberg and Schutta (1967) on the cytological events that occur in cerebellar tissue cultures exposed to different concentrations of bilirubin. The present investigation has been concerned with the cytological changes of neurons and glia in the Gunn rat substantia nigra. It is an area known to be stained by bilirubin in young animals of this strain (Blanc, 1961). The substantia nigra is of especial interest because alterations within its neurons in man (for example, depletion of melanin stores in Parkinson's disease) are associated, perhaps etiologically, with disease. This investigation has been approached with the hope of eventually understanding how substantia nigra neurons react under a variety of metabolic alterations. One of the most cogent problems before contemporary neurobiologists is to understand the basis for the selective vulnerability of basal ganglia tissue in metabolic disorders. MATERIALS AND METHODS Thirty jaundiced female rats were used in this study. The youngest animals were two weeks old. This was the earliest age at which the tremulous movement characteristic of Gunn rats could be observed. Only those animals that demonstrated (t) jaundice shortly after birth and (2) tremor were studied. The oldest animals in the series were t 2-- 13 months old. The colony was originally obtained through the Department of Vivarial Sciences at the Mayo Clinic. The animals were maintained on a controlled high nutritive diet. Normal, healthy Sprague-Dawley rats, 2 weeks--13 months old, were also studied. The cytology of their substantia nigra was compared with that of Gunn rats. Both strains were handled in the same manner. For light microscopic study the animals were anesthetized with Pentabarbital Sodium (0.1 cc/100 g body weight) and perfused with 10 ~ neutral formalin (pH 7.2). The brains were dehydrated, embedded in paraffin and serially sectioned. The 6 ~ thick sections were stained with cresyl-violet.
Ultrastructure of the Gunn Rat Substantia Nigra
95
For electron microscopic study animals were perfused trans-cardially with a solution of 1 ~ glutaraldehyde, 1 ~ paraformaldehyde in a 0.12 M phosphate buffer at pH 7.4 (Wuerker and Palay, 1969). The animal was artifically respirated with a mixture of 95 oxygen and 5 ~ carbon dioxide up to the moment that the cannula was inserted into the heart. Prior to cannulation, 0.1 cc sodium-Heparin and 0.4 cc 1 ~ sodium nitrite were injected separately into the left ventricle. The tissue was post-fixed in a solution of 2 OsO4, 3 . 5 ~ dextrose in 0.12 M phosphate buffer 3 h. Following this, the tissue was dehydrated in increasing concentrations of methyl alcohol and, finally, in propylene oxide. Small blocks of tissue were embedded in Epon 812. In some cases post-fixation took place 1 --2 h after perfusion. Otherwise the perfused brains were left in situ 3--4 h and then stored in the original fixing solution overnight at 4 ~C. Post-fixation, dehydration and embedding were carried out the next day. Plastic sections, 1.5 ~zthick, were cut with glass knives by means of a Sorvall PorterBtum Mt-1 ultramicrotome. They were stained with a 1 ~ solution of cresyl-violet. Once the substantia nigra was recognized in these sections, the blocks were trimmed for taking thin sections. The thin sections were stained in 5 ~ uranyl acetate for 2.5 min and in lead citrate (Reynolds, 1963) for the same length of time. The tissue was examined and photographed with a R C A EMU-3H electron microscope at an accelerating voltage of 50 kV, with a 25 ~ objective aperture.
OBSERVATIONS
Light Microscopic Observations Neuronal somata containing densely staining cytoplasmic inclusions were indentified in 1 --2 ~z thick sections of epoxy embedded tissue. These inclusions occurred most commonly in 2 week old Gunn rats. With increasing age (6 weeks--13 months), they were observed less frequently. They were not constant in shape or size and were scattered through the soma. In a few instances the somata were hyperchromatic. Sections of the substantia nigra of normal, Sprague-Dawley rats, ranging in age 2 weeks--12 months, did not contain similar cytoplasmic inclusions. The neuropil did not reveal any observable modification in either experimental or control animals. Nor was there any indication of gliosis in adult Gunn rat substantia nigra. Furthermore, there was no noticeable difference in the numbers of neurons in comparing the substantia nigra of Gunn rats and control animals.y
Electron Microscopic Observations The m o s t p r o n o u n c e d structural effects observed in the substantia nigra o f hyperbilirubinemic G u n n rats were alterations o f the neuronal cytoplasm. In large part these changes were confined to the somata, but the proximal portions of dendrites were also involved. Only rarely were axons affected. The principal effects consisted o f the presence o f (1) complex m e m b r a n o u s bodies, (2) dilated cisternae of endoplasmic reticulum, (3) single m e m b r a n e b o u n d vacuoles and (4) enlarged mitochondria. Furthermore, there were accumulations o f osmiophilic granules, p r o b a b l y glycogen particles, in mitochondria and ergastoplasmic cisternae. This last point will be considered m o r e fully in an a c c o m p a n y i n g report.
1. Complex Membranous Bodies (CMBs). The size and shape of the CMBs varied considerably (Figs. 1--3, 5). Their m a j o r constituent was m e m b r a n e that was not noticeably different f r o m or thicker than cytoplasmic unit membranes. 7*
96
Fig. 1
H.K. Batty and O. E. Millhouse
Ultrastructure of the Gunn Rat Substantia Nigra
97
These bodies were seldom evenly round; instead they were ovoid, elongate or so irregular as to be unique. This was particularly true in the youngest (2 week old) brains examined (Fig. 1). Neurons of animals 12 weeks and older had whorls of membrane that were more concentric and of a more even density (Fig. 5). However irregularities in shape and periodicity of CMBs were not u n c o m m o n in mature brains. Although some of the CMBs consisted entirely of membranous whorls, the majority were composed of loosely wound membrane surrounding a bit of cytoplasm. Mitochondria, free ribosomes, and microtubules were identified within these pieces of cytoplasm (Figs. 1 --3). Occasionally an unevenly thick clear zone separated the apparently isolated cytoplasm from the surrounding membrane. Also clear vacuoles or vacuoles containing flocculent matter were associated with CMBs (Fig.6). N o t all the constituents of these vacuoles could be identified, although they were considered to be remnants of cytoplasm. The CMBs, in some examples, were directly connected with granular endoplasmic reticulum. However ribosomes were not attached to any of the membranes of CMBs. A few CMBs, consisting entirely of membrane, were located within an endoplasmic reticulum cistern (Fig. 2). Virtually every neuron within the substantia nigra of the young Gunn rats examined contained at least one or two CMBs. The amount of area occupied by CMBs was not constant from neuron to neuron. Profiles of several neuronal somata were occupied almost entirely by CMBs. In other somata, however, only a few CMBs were seen and these were separated from one another by relatively normal appearing cytoplasm. There was no indication of an increase in the population of either microtubules or neurofilaments. The CMBs were less frequently seen in adult than in 2 week old brains. With age CMBs became concentrated further from the nucleus. Also, their membranes were organized in more regular arrays (Fig. 5). Furthermore they consisted, in older animals, essentially of membrane, but some of them did have an amorphous core (Fig. 5).
2. Endoplasmic Reticulum. The granular endoplasmic reticulum in proximity to CMBs in 2--6 week old animals assumed a dilated, angular appearance (Figs.1--3). This dilatation was occasionally sufficient to make the ergastoplasmic channels look swollen and vacuolar. This was not characteristic of substantia nigra neurons either in older G u n n rats or in 2 - - 6 week old control brains (Fig.4). Additionally, there was no perceptible loss of ribosomes from the ergastoplasmic membrane except where it was continuous with CMBs. Even in young Gunn rats free ribosomes were arranged in rosettes and were not dispersed as single units (Figs. 1--3). Fig./. This section through a portion of a neuronal soma shows several examples of CMBs of variable size and shape. The membranes surround bits of cytoplasm containing mitochondria (m), ribosomes (r) and vacuoles (v) containing moderately dense flocculent material. The endoplasmic reticulum channels (c) appear distended. The astrocytic processes (As) along the neuronal soma and the adjacent group of neurites (n) do not show any indication of cytoplasmic alteration. 2 week old Gunn rat; 27200 •
Fig. 2. A CMB composed entirely of swirls of membrane appears to lie within a cistern of granular endoplasmic reticulum. Observe the rosette arrangement of the u n b o u n d ribosomes (r). 2 weeks old G u n n rat; 27200 • Fig. 3. The cytoplasm of this neuronal soma contains distended endoplasmic reticulum cisternae (c) and a relatively large CMB that encircles a mitochondrion (m). Smaller CMBs, apparently connected with the endoplasmic reticulum, are indicated by arrowheads. The axon terminals (at) synapsing on this ceil body look normal, as do the other structures in the neuropil. 2 week old G u n n rat; 27200 •
Ultrastructure of the Gunn Rat Substantia Nigra
99
Fig. 4. The substantia nigra of several 2 week old Sprague-Dawley rats was studied and the cytological features of their neurons were compared with those of the Gunn rat. Notice in this figure that in the Sprague-Dawley rats the granular endoplasmic reticulum, in the form of a typical Nissl body (Nb), does not appear distended as it does in same aged Gunn rats. Also compare the size of these mitochondria (m) with those of the Gunn rat as seen in the previous figures. 2 week old Sprague-Dawley rat; 27 200 •
3. Single Membrane Bound Vacuoles. Pale, single membrane bound vacuoles measuring 0.2--2 ~ in diameter were c o m m o n in young animals (Figs. 6 and 7). The average diameter measured in 125 randomly selected vacuoles was 0.9 ~. They were plentiful in the vicinity of the nucleus (Fig. 7), closely adjacent to the Golgi system. They were also present at other sites in the cytoplasm, frequently in conjunction with CMBs (Fig. 6). Although generally spherical, their surface was rippled and even deeply indented (Figs.6 and 7). Their lumina held weakly dense osmiophilic material reminiscent of the thin thread-like elements that were also found in dilated ergastoptasmic cisternae. What might be interpreted as cellular detritus was sometimes contained in the lumen. The limiting membrane of these vacuoles was not always complete; consequently there was continuity between the lumen and cytoplasm. The vacuolar membranes, moreover, were not usually as sharp and crisp as other cellular membranes (Figs. 6 and 7). Often they were indistinct as though cut tangentially. 4. Mitochondria. The mitochondria were enlarged by one of two processes. Those in 2 - - 6 week old brains showed an expansion of the matrix compartment (Figs. 1 and 3). However other mitochondria in these same animals, but more frequently in older ones, were engorged with osmiophilic granules that produced an ex-
100
H . K . Batty and O. E. Millhouse
Fig. 5. With increasing age, the CMBs are located more peripherally in the neuronal soma. They also assume a more regular appearance. This CMB, which is only partially shown, nearly fills a dendritic profile, leaving only a rim of normal appearing cytoplasm that contains microtubules (arrowheads) and segments of agranular endoplasmic reticulum. The thick layer of membranes of this CMB surrounds a dense and relatively amorphous core (de). Adult Gunn rat; 46000 x
p a n s i o n o f the i n t r a c r i s t a l c o m p a r t m e n t . M o r e t h a n 200 c o m p a r a b l e m i t o c h o n d r i a l profiles were m e a s u r e d in the s u b s t a n t i a n i g r a o f b o t h 2 week c o n t r o l a n d G u n n r a t brains. T h e average d i a m e t e r in the f o r m e r was 0.39 lJ. a n d in the latter, 1.3 ix. O n l y those m i t o c h o n d r i a l profiles l o c a t e d in s o m a t a a n d n o t cont a i n i n g granules were m e a s u r e d . I t deserves e m p h a s i s t h a t this o b s e r v a b l e m i t o c h o n d r i a l m a t r i x e n l a r g e m e n t was c h a r a c t e r i s t i c only o f 2 - - 6 week o l d G u n n rats. A similar event was n o t f o u n d
Fig. 6. Single membrane bound vacuoles (V) containing a flocculent material were commonly seen in the neuronal cytoplasm of 2 week old Gunn rats. The rippled limiting membrane was not always distinct but often appeared to be blurred (arrowheads). The vacuoles frequently joined a CMB. 27200 • Fig. 7. Clusters of small vacuoles (vs) were usually found adjacent to the nucleus (Nu), in the region of the Golgi system. Except for their size, they are structurally similar to the larger vacuoles shown in Figure 6. 2 week old Gunn rat; 27200 x
Ultrastructure of the Gunn Rat Substantia Nigra
Figs. 6 and 7
101
102
H . K . Batty and O. E. Millhouse
Fig. 8. Only a few hyperchromatic neurons were seen in the G u n n rat substantia nigra. The nucleus (Nu) and cytoplasm are equally dense. Several CMBs can be seen in the cytoplasm. Notice that the surface of the neuronal soma is extremely irregular. Astrocytie processes (As) that seem to be hypertrophied surround the neuron. 2 week old G u n n rat; 27200 x
Ultrastructure of the Gunn Rat Substantia Nigra
103
in animals 12 weeks and older. There was no evidence of a consistent increase or decrease in matrix density nor was mitochondrial hyperplasia observed. 5. Hyperchromatic Neurons. Approximately 3--5 ~ of the neurons examined in younger animals were hyperchromatic (Fig. 8); however, they were not apparent in animals older than 6 weeks. These neurons contained pleomorphic CMBs, electron opaque vacuoles and enlarged mitochondria. The hyperchromasia appeared to result from an increased density of the cytoplasmic matrix. No abnormal concentrations of ribosomes were recognized (Fig. 8). The nucleus and cytoplasm were nearly equally dense and the nuclear membrane was interrupted in a few instances. 6. Comments on the Neuropil. The neuropil around the affected neuronal somata was not structurally modified to any observable extent, whether in 2 week old or adult Gunn rat (Figs. l, 3 and 5). Synaptic terminals on the somata were unchanged (Fig.3). Fascicles of small dendrites and unmyelinated axons did not reflect the obvious structural alterations seen in adjacent somata, except for the presence of osmiophilic granules in mitochondria. Comparison of light and electron microscopic preparations of Gunn rat and control brains did not indicate any significant neuron loss in the substantia nigra of the mature Gunn rat. Furthermore, there was no pronounced gliosis. DISCUSSION The cytological alterations of neurons in the substantia nigra of Gunn rats are assumed to be the effect of bilirubin on neuronal tissue. However bilirubin was not identified within any of the cells examined. One possible reason is that the youngest brains studied were two weeks old. Bilirubin may have been removed from the tissue at an earlier age. Schutta and Johnson (1967) pointed out that both the generalized and localized yellow staining characteristic of bilirubin deposits was no longer visible in Gunn rat brains by the twentieth post-natal day. Also, there is a close correspondence between the structural changes we observed and those reported by Silberberg and Schutta (1967), who exposed cultures of rat cerebellum to specific concentrations of bilirubin. Furthermore, Chen et al. (1969) injected tritium labeled bilirubin into rabbits that had been previously subjected to anoxia in order to circumvent the blood-brain barrier. Although the label can be seen in their micrographs, it is not associated with a unique cellular inclusion, crystal or pigment molecule. Consequently, bilirubin, if present, may be sequestered or modified so that it cannot be demonstrated directly with conventional electron microscopic techniques. The varied appearance of CMBs may indicate that bilirubin exerts more than a single effect on neurons and neuronal metabolism. The more compact membranous whorls are similar to the "membranous cytoplasmic body" described by Terry and Weiss (1963) in Tay-Sachs disease. Samuels et al. (1965) studied the possible modes of formation of these membrane bodies and concluded that they arise from the spontaneous aggregation of gangliosides, cholesterol and phospholipids. By combining these three substances in the ratio in which they occur in the disease, Samuels et al. were able to generate the characteristic membranous
104
H.K. Batty and O. E. Millhouse
bodies extracellularly. They concluded that these membrane whorls were not necessarily derived from a specific organelle and they doubted that these bodies were lysosomes. Claireux et al. (1953) demonstrated that bilirubin forms a complex with a neuronal lipid which was not further specified. On the basis of this observation and the experiments of Samuels et al. (1965), Schutta and Johnson (1967) argued that such a bilirubin-lipid complex might produce a compound that forms the membranous whorls seen in Purkinje cells of the Gunn rat. It was assumed that this compound was not readily metabolized by neurons and, thus, accumulated in the cytoplasm in the form of membranous bodies. Lullmann-Rauch et al. (1972) also conclude that the membranous bodies they described in lung tissue following injections of chlorphentermine represent a non-digestible phospholipid that accumulates because of the interaction of the drug with certain phospholipids. However, the CMBs observed in Gunn rat neurons possibly represent more than the passive accumulation of a non-metabolizable molecular chain. Stern et al. (1972) have recently exposed long-term organotypic tissue cultures of rat spinal cord to various concentrations of sulfatide, ganglioside and cerebroside. Membranous bodies strikingly similar to those seen in Tay-Sachs and other storage diseases developed in the neurons in these cultures. Significantly, acid phosphatase activity was demonstrated in these cytoplasmic bodies, thus, proving them to be lysosomes. Lullmann-Rauch et al. (1972) also demonstrated acid phosphatase activity in similar appearing membrane whorls of rat pulmonary cells. Yet the majority of CMBs observed in Gunn rat neurons consisted of more than layers of membrane. The single, membrane bound vacuoles filled with flocculent material closely correspond in structure, content, and size to the autophagic vacuoles observed in cultured macrophages exposed to chloroquine (Fedorko et al., 1968). With time, these autophagic vacuoles were reported to increase in size, become irregular in shape, invaginate and to contain " . . . cytoplasmic components apparently in various stages of digestion" as well as membrane whorls, amorphous matter and lipid droplets. CMBs also bear close resemblance to the cytological structures reported in other types of cellular injury (Hruban et al., 1963; Colonnier, 1964; Barron et al., 1967, 1973; Hwang et al., 1974; Lullmann-Rauch et al., 1972; Lullmann-Rauch, 1974; Lullmann-Rauch and Pietschmann, 1974). Considering that CMBs consist of multilaminar membranes frequently surrounding segregated pieces of cytoplasm and considering, too, their structural similarity with the membranous bodies produced by Stern et al. (1972), the myeloid bodies of Barron et al. (1973), the autophagic vacuoles described by Fedorko et al. (1968) and the inclusions seen in examples of focal cytoplasmic degradation (Hruban et al., 1963), we are led to interpret the CMBs as lysosomes, representing autophagic activity. The origin of the membrane lamellae associated with late autophagic vacuoles or residual bodies has not been explained satisfactorily. Fedorko et al. (1968) interpreted their data to argue that chloroquine acts on the membranes of the Golgi system and agranular endoplasmic reticulum to produce a fusion of the membranes with one another. Recalling the experiments of Samuels et al. (1965),
Ultrastructure of the Gunn Rat Substantia Nigra
105
the CMBs could conceivably arise as a combination of bilirubin with some neuronal lipid. This novel molecular complex may trigger an autophagic response. As a consequence the membranous complex would become associated with autophagic vacuoles, and, predictably, acid hydrolase. Hwang et al. (1974) studied the formation of membranous whorls in liver parenchymal cells exposed to cycloheximide. Their biochemical experiments showed that during the time these whorls were being formed protein synthesis was decreased 40--60~. However at the same time, phospholipid synthesis was increased. They concluded that the membrane bodies result from a desynchronization in the synthesis of the protein and lipid components of membrane. Bilirubin may also precipitate an autophagic response by another means. Cowger (1971) has shown that bilirubin affects cells in a manner similar to that of lipophilic surfactants. Thus, it would not be unreasonable to expect that bilirubin partially exerts its effects on neurons by affecting lipids of the cytomembrane system, possibly altering membrane permeability. Should this occur, hydrolytic enzymes could escape into the surrounding cytoplasm and cause focal cytoplasmic degradation. It is not difficult to envision a marshalling of a sequence of adaptive responses which would contain the degradation. But containment is not the only point to consider: Novikoff (1967) points out that autophagic activity may be a means of assimilating the cytologically changed areas into the intracellular economy. Bilirubin is also known to act as a respiratory poison by uncoupling phosphorylation and inhibiting mitochondrial respiration (see Cowger, 1971). The morphological expression of this is seen as an increase in mitochondrial matrix (Tapley and Cooper, 1956; Lehninger, 1962, 1965). The resulting decrease in intracellular levels of ATP may also contribute to cytoplasmic injury. Because there is no evidence for either a precipitous loss of neurons or a significant proliferation of glia in the substantia nigra of adult Gunn rats, it appears that the majority of these neurons recover from exposure to bilirubin. However not all the neurons survived. A small proportion, 3--5 ~ , were hyperchromatic, a cytological feature generally regarded as characteristic of dying or dead neurons (Barron et al., 1974; Matthews, 1973). Although the reason(s) for this cell death has not been established, two points should be considered. One, the cell death may reflect the deposition of different concentrations of bilirubin in the substantia nigra so that only a few neurons were exposed to a lethal dose. Two, there is evidence that autophagocytosis is dependent upon the maintenance of a certain intracellular level of ATP (Trump et al., 1971 ; Shelburne et al., 1973). If autophagocytosis cannot occur, the consequence could be cell death. Regarding this, Cowger (1971) has pointed out that the effect of bilirubin on mitochondrial respiration p e r s e does not cause cell death. The cellular events observed by examining the substantia nigra of various aged animals indicate to us that neurons exposed to certain (but, as yet, unspecified) concentrations of bilirubin undergo extensive autophagocytosis, but providing that there are sufficient levels of ATP, the neuronal cytoplasm is renewed. Autophagocytosis, then, may be viewed as being important in (1) containing the toxic effects of bilirubin and paradoxically, (2) reconstituting the cytoplasm (Novikoff, 1967).
106
H . K . Batty and O. E. Millhouse
Acknowledgements. We are grateful to Drs. M. L. Grunnet, L. J. Stensaas, W. Stevens and, in particular, L. W. Jarcho for advice on the manuscript. Mesdames D. Lerdahl and S. Carter provided technical assistance as did Mr. K. Johnson. Mrs. J. English helped with the final preparation of this paper. Their exceptionally fine help is very much appreciated. This research was partially supported by grants GM-00958 (NIH) and 5 T01 NS05309 (NIH) and the Eleanor Roosevelt Cancer Foundation Research Institute. REFERENCES Aleu, F. P., Terry, R. D., Zellweger, H. : Electron microscopy of two cerebral biopsies in gargoylism. J. Neuropath. exp. Neurol. 24, 304--317 (1965) Barron, K. D., Doolin, P. F., Oldershaw, J. B.: Ultrastructural observations on retrograde atrophy of lateral geniculate body. Neuronal Alteration J. Neuropath. exp. Neurol. 26, 300--326 (1967) Barron, K. D., Means, E. D., Larsen, E. : Ultrastructure of retrograde degeneration in thalamus of rat. J. Neuropath. exp. Neurol. 32, 218--244 (t973) Blanc, W. A., Johnson, L.: Studies on Kernicterus. Relationships with sulfonamide intoxication, report on kernicterus in rats with glucuronyl transferase deficiency and review of pathogenesis J. Neuropath. exp. Neurol. 18, 165--187 (1959) Blanc, W. A. : Kernicterus in Gunn's strain of rat. In: Kernicterus (ed. A. Sass-Korstak), pp. 150-- 152. Toronto: University of Toronto Press 1961 Carbone, J. V., Grodsky, G. M. : Constitutional nonhemolytic hyperbilirubinemia in the rat: Defect of bilirubin conjugation. Proc. Soc. exp. Biol. (N.Y.) 94, 461--463 (1957) Chen, H., Tsai, D., Wang, C., Chen, Y. : An electron microscopic and radioautographic study on experimental kernicterus. Amer. J. Path. 56, 31--58 (1969) Claireaux, A. E., Cole, P. G., Lathe, G. H. : Icterus of the brain of the newborn. Lancet 1953 II, 1226--1230 Colonnier, M. : Experimental degeneration in the cerebral cortex. Amer. J. Anat. 98, 4 7 - 5 3 (1964) Cowger, M. L. : Mechanism of bilirubin toxicity on tissue culture cells: Factors that affect toxicity, reversibility by albumin, and comparison with other respiratory poisons and surfactants. Biochem. Med. 5, 1016 (1971) Diamond, I., Schmid, R. : Experimental bilirubin encephalopathy. The mode of entry of bilirubin-14C into the central nervous system. J. clin. Invest. 45, 6 7 8 - 6 8 9 (1966) Ernster, L., Herlin, L., Zetterstrom, R. : Experimental studies of the pathogenesis of kernicterus. Pediatrics 20, 647--652 (1957) Farrell, D. F., Baker, H. J., Herndon, R. M., Lindsey, J. R., McKhann, G. M. : Feline GM1 gangliosidosis: Biochemical and ultrastructural comparisions with the disease in man. J. Neuropath. exp. Neurol. 32, 1 (1973) Fedorko, M. E., Hirsch, J. G., Cohn, Z. A. : Autophagic vacuoles produced in vitro. II. Studies on the mechanism of autophagic vacuoles produced by chloroquine. J. Cell Biol. 38, 392--402 (1968) Gardner, W. A., Konigsmark, B. W. : Familial nonhemolytic jaundice: Bilirubinosis and encephalopathy. Pediatrics 43, 365--376 (1969) Gunn, C. K. : Hereditary acholuric jaundice in the rat. Canad. reed. Ass. J. 50, 230--237 (1944) Hruban, Z., Spargo, B., Swift, H., Wissler, R. W., Kleinfeld, R. G. : Focal cytoplasmic degradation. Amer. J. Path. 42, 657--683 (1963) Hwang, K. M., Yang, L. C., Carrico, C. K., Schulz, R.A., Schenkman, J.B., Sartorelli, A. C. : Production of membrane whorls in ra liver by some inhibitors of protein synthesis. J. Cell Biol. 62, 20--31 (1974) Johnson, L. : The effects of certain substances on bilirubinic levels and the occurrence of kernicterus in generally jaundiced rats. In: Kernicterus (ed. A. Sass-Korstak), pp. 208--218. Toronto: University of Toronto Press t961
Ultrastructure of the Gunn Rat Substantia Nigra
107
Johnson, L., Garcia, M. L., Figueroa, E., Sarmiento, F.: Kernicterus in rats lacking glucuronyl transferase. Am. J. Dis. Child. 101, 3 2 2 - 3 4 9 (1961) Kirkpatrick, J. B.: Chromatolysis in the hypoglossal nucleus of the rat: An electron microscopic analysis. J. comp. Neurol. 132, 189 212 (1968) Lehninger, A. L. : Water uptake and extrusion by mitocbondria in relation to oxidative phosphorylation. Physiol. Rev. 42, 467--517 (1962) Lehninger, A. L.: In: The Mitochondrion, molecular basis of structure and function. New York: W. A. Benjamin 1964 Ltillmann-Rauch, R. : Lipodosis-like alterations in hypothalamic neurosecretory cells of rats treated with chlorphentermine or iprindole. Cell Tiss. Res. 149, 587--590 (1974) Ltillmann-Rauch, R., Reil, G. H., Rossen, E., Seiler, K. U. : The ultrastructure of rat lung changes induced by an anorganic drug (chlorphentermine). Virchows Arch., Abt. B, Zellpatb. 11, 167--181 (1972) Ltillmann-Rauch, R., Pietschmann, N. : Lipodosis-like cellular alterations in lymphatic tissues of chlorphentermine-treated animals. Virchows Arch., Abt. B, Zellpath. 15, 295--308 (1974) Matthews, M. A. : Death of the central neuron: an electron microscopic study of thalamic retrograde degeneration following cortical ablation. J. Neurocytol. 2, 265--288 (1973) Novikoff, A. B. : Lysosomes in nerve cells. In: The neuron (ed. H. Hyden), pp. 255 377. Amsterdam-London-New York: Elsevier Publ. Co. 1967 Odell, G. B. : Factors influencing the binding of bilirubin by albumin in kernicterus. In: Kernicterus (ed. A. Sass-Korstak), pp. 199--207. Toronto: University of Toronto Press 1961 Reynolds, E. S. : The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17, 208--212 (1963) Samuels, S., Gonatas, N. K., Weiss, M. : Formation of membranous cytoplasmic bodies in Tay-Sachs disease: An in vitro study. J. Neuropath. exp. Neurol. 24, 256--264 (1965) Schmid, R., Axelrod, J., Hammaker, L., Swarm, R. L. : Congenital jaundice in rats, due to a defect in glucuronide formation. J. clin. Invest. 37, 1123 1130 (1958) Schutta, H. S., Johnson, L. : Bilirubin encephalopathy in the Gunn rat: A fine structural study of the cerebellar cortex. J. Neuropath. exp. Neurol. 26, 377--396 (1967) Shelburne, J. D., Arstila, A. U., Trump, B. F.: Studies on cellular autophagocytosis, cyclic A M P and dibutyryl cyclic A M P stimulated autophagy in rats. Amer. J. Path. 72, 521--540 (1973) Silberberg, D. H., Schutta, H. S. : The effects of unconjugated bilirubin and related pigments on cultures of rat cerebellum. J. Neuropath. exp. Neurol. 26, 572 583 (1967) Stern, J., Novikoff, A. B., Terry, R. D. : The induction of sulfatide, ganglioside, and cerebroside storage in organized nervous system cultures. In: Sphingolipids, sphinglolipidoses, and allied disorders (eds. B. W. Volk and S. M. Aronson), pp. 651--660. New York: Plenum Press 1972 Tapley, D. F., Cooper, G. : Effect of thyroxine on the swelling of mitochondria isolated from various tissues of the rat. Nature (Lond.) 167, 1119 (1956) Terry, R. D., Weiss, M.: Studies in Tay-Sachs disease. II. Ultrastructure of the cerebrum. J. Neuropath. exp. Neurol. 22, 18--55 (1963) Trump, B. F., Croker, B. P., Mergener, W. J. : The role of energy metabolism, ion, and water shifts in the pathogenesis of cell injury. In: Cell membranes, biological and pathological aspects (eds. G. W. Richter, D. G. Scarpelli), pp. 84--128. Baltimore: The Williams and Wilkins Co. 1971 Waters, W. J., Richert, D. A., Rawson, H. H. : Bilirubin encephalopathy. Pediatrics 13, 319--325 (1954) Waters, W. J., Britton, H. A. : Bilirubin encephalopathy: Preliminary studies related to production. Pediatrics 15, 45--48 (1955) Wuerker, R. B., Palay, S. L. : Neurofilaments and microtubules in anterior horn cells of the rat. Tissue and Cell 1, 387--402 (1969)
Received September 10, 1975; Accepted December 22, 1975
Acta neuropath. (Berl.) 35, 93 --107 (1976)
Neuropathologica
9 by Springer-Verlag1976
Ultrastructure of the Gunn Rat Substantia Nigra
I. Cytoplasmic Changes H. K. Batty and O. E. Millhouse Departments of Anatomy and Neurology, College of Medicine, University of Utah, 50 North Medical Drive, Salt Lake City, Utah 84132, U.S.A.
Summary. The substantia nigra of various aged hyperbilirubinemic (Gunn) rats was studied by means of electron microscopy. The cytological features observed in the neuronal somata were the presence of (1) complex membranous bodies (CMBs), (2) dilated cisternae of granular endoplasmic reticulum, (3) single membrane bound vacuoles and (4) enlarged mitochondria. Nearly every neuronal soma studied in two week old Gunn rats contained CMBs, which consisted of several layers of membrane that usually, but not always, surrounded small islands of cytoplasm. On occasion CMBs were seen to be directly connected with granular endoplasmic reticulum and, in a few instances, they were located within a cistern of endoplasmic reticulum. There were significantly fewer CMBs in the neuronal somata of adult Gunn rats. They were located peripherally in the somata or in the proximal portions of dendrites. Furthermore, in these animals the cytoplasm appeared normal and did not exhibit the features commonly seen in younger animals. Only a few hyperchromatic neurons were observed and no pronounced gliosis was evident. Therefore it is assumed that the majority of substantia nigra neurons recovered from the toxic effect of bilirubin or that the concentration of bilirubin deposited in the substantia nigra is not sufficient to be lethal. The hypothesis that is considered is that CMBs represent autophagic activity which results from exposure of neurons to bilirubin. The adjacent neurites and glia did not demonstrate the cytoplasmic changes that were characteristic of the neuronal somata.
Key words: Substantia n i g r a - Gunn rat - - Complex membranous bodies -
-
Bilirubin - - Electron microscopy.
INTRODUCTION The Gunn rat (Gunn, ]944) was used as a model to study the effects of bilirubin on the central nervous system. These rats lack the genetic information to code for the liver enzyme glucuronic acid transferase (Carbone and Grodsky, 1957; 7
Acta neuropath. (Berl.) ~d. 35
94
H.K. Batty and O. E. Millhouse
Schmid et al., 1958). As a result the animals develop a non-hemolytic hyperbilirubinemia (Johnson et al., 1959, 1961; Blanc and Johnson, 1959; Johnson, 1961 ; Gardner and Konigsmark, 1969). Bilirubin deposition within the brain has been commonly associated with the kernicterus syndrome (Ernster et al., 1957; Blanc and Johnson, 1959; Johnson et al., 1959; Blanc, 1961; Johnson, 1961). The lemon yellow stain seen in several areas of the Gunn rat brain (see Schutta and Johnson, 1967) until about the twentieth post-natal day corresponds to material of the same color which Waters et al. (1954) analyzed spectrophotometrically. On the basis of its spectrophotometric curve in comparison with the curve for bilirubin, the lemon yellow stain was determined by these authors to represent bilirubin. The bilirubin encephalopathy begins to develop in newborn Gunn rats 3--4 days old (Johnson et al., 1959; Blanc, 1961; Schutta and Johnson, 1967). Several explanations have been offered to account for the deposition of bilirubin in these young brains. Waters and Britton (1955) and Schutta and Johnson (1967) argue that it is due to the immaturity of the blood-brain-barrier. However others have suggested that it is correlated with low levels of serum albumin (Diamond and Schmid, 1966) and with blood pH (Odell, 1961, 1966). Schutta and Johnson (1967) studied the ultrastructure of the cerebellum in Gunn rats ranging in age from a few days to 18 months old. Their observations were substantiated by the subsequent investigation of Silberberg and Schutta (1967) on the cytological events that occur in cerebellar tissue cultures exposed to different concentrations of bilirubin. The present investigation has been concerned with the cytological changes of neurons and glia in the Gunn rat substantia nigra. It is an area known to be stained by bilirubin in young animals of this strain (Blanc, 1961). The substantia nigra is of especial interest because alterations within its neurons in man (for example, depletion of melanin stores in Parkinson's disease) are associated, perhaps etiologically, with disease. This investigation has been approached with the hope of eventually understanding how substantia nigra neurons react under a variety of metabolic alterations. One of the most cogent problems before contemporary neurobiologists is to understand the basis for the selective vulnerability of basal ganglia tissue in metabolic disorders. MATERIALS AND METHODS Thirty jaundiced female rats were used in this study. The youngest animals were two weeks old. This was the earliest age at which the tremulous movement characteristic of Gunn rats could be observed. Only those animals that demonstrated (t) jaundice shortly after birth and (2) tremor were studied. The oldest animals in the series were t 2-- 13 months old. The colony was originally obtained through the Department of Vivarial Sciences at the Mayo Clinic. The animals were maintained on a controlled high nutritive diet. Normal, healthy Sprague-Dawley rats, 2 weeks--13 months old, were also studied. The cytology of their substantia nigra was compared with that of Gunn rats. Both strains were handled in the same manner. For light microscopic study the animals were anesthetized with Pentabarbital Sodium (0.1 cc/100 g body weight) and perfused with 10 ~ neutral formalin (pH 7.2). The brains were dehydrated, embedded in paraffin and serially sectioned. The 6 ~ thick sections were stained with cresyl-violet.
Ultrastructure of the Gunn Rat Substantia Nigra
95
For electron microscopic study animals were perfused trans-cardially with a solution of 1 ~ glutaraldehyde, 1 ~ paraformaldehyde in a 0.12 M phosphate buffer at pH 7.4 (Wuerker and Palay, 1969). The animal was artifically respirated with a mixture of 95 oxygen and 5 ~ carbon dioxide up to the moment that the cannula was inserted into the heart. Prior to cannulation, 0.1 cc sodium-Heparin and 0.4 cc 1 ~ sodium nitrite were injected separately into the left ventricle. The tissue was post-fixed in a solution of 2 OsO4, 3 . 5 ~ dextrose in 0.12 M phosphate buffer 3 h. Following this, the tissue was dehydrated in increasing concentrations of methyl alcohol and, finally, in propylene oxide. Small blocks of tissue were embedded in Epon 812. In some cases post-fixation took place 1 --2 h after perfusion. Otherwise the perfused brains were left in situ 3--4 h and then stored in the original fixing solution overnight at 4 ~C. Post-fixation, dehydration and embedding were carried out the next day. Plastic sections, 1.5 ~zthick, were cut with glass knives by means of a Sorvall PorterBtum Mt-1 ultramicrotome. They were stained with a 1 ~ solution of cresyl-violet. Once the substantia nigra was recognized in these sections, the blocks were trimmed for taking thin sections. The thin sections were stained in 5 ~ uranyl acetate for 2.5 min and in lead citrate (Reynolds, 1963) for the same length of time. The tissue was examined and photographed with a R C A EMU-3H electron microscope at an accelerating voltage of 50 kV, with a 25 ~ objective aperture.
OBSERVATIONS
Light Microscopic Observations Neuronal somata containing densely staining cytoplasmic inclusions were indentified in 1 --2 ~z thick sections of epoxy embedded tissue. These inclusions occurred most commonly in 2 week old Gunn rats. With increasing age (6 weeks--13 months), they were observed less frequently. They were not constant in shape or size and were scattered through the soma. In a few instances the somata were hyperchromatic. Sections of the substantia nigra of normal, Sprague-Dawley rats, ranging in age 2 weeks--12 months, did not contain similar cytoplasmic inclusions. The neuropil did not reveal any observable modification in either experimental or control animals. Nor was there any indication of gliosis in adult Gunn rat substantia nigra. Furthermore, there was no noticeable difference in the numbers of neurons in comparing the substantia nigra of Gunn rats and control animals.y
Electron Microscopic Observations The m o s t p r o n o u n c e d structural effects observed in the substantia nigra o f hyperbilirubinemic G u n n rats were alterations o f the neuronal cytoplasm. In large part these changes were confined to the somata, but the proximal portions of dendrites were also involved. Only rarely were axons affected. The principal effects consisted o f the presence o f (1) complex m e m b r a n o u s bodies, (2) dilated cisternae of endoplasmic reticulum, (3) single m e m b r a n e b o u n d vacuoles and (4) enlarged mitochondria. Furthermore, there were accumulations o f osmiophilic granules, p r o b a b l y glycogen particles, in mitochondria and ergastoplasmic cisternae. This last point will be considered m o r e fully in an a c c o m p a n y i n g report.
1. Complex Membranous Bodies (CMBs). The size and shape of the CMBs varied considerably (Figs. 1--3, 5). Their m a j o r constituent was m e m b r a n e that was not noticeably different f r o m or thicker than cytoplasmic unit membranes. 7*
96
Fig. 1
H.K. Batty and O. E. Millhouse
Ultrastructure of the Gunn Rat Substantia Nigra
97
These bodies were seldom evenly round; instead they were ovoid, elongate or so irregular as to be unique. This was particularly true in the youngest (2 week old) brains examined (Fig. 1). Neurons of animals 12 weeks and older had whorls of membrane that were more concentric and of a more even density (Fig. 5). However irregularities in shape and periodicity of CMBs were not u n c o m m o n in mature brains. Although some of the CMBs consisted entirely of membranous whorls, the majority were composed of loosely wound membrane surrounding a bit of cytoplasm. Mitochondria, free ribosomes, and microtubules were identified within these pieces of cytoplasm (Figs. 1 --3). Occasionally an unevenly thick clear zone separated the apparently isolated cytoplasm from the surrounding membrane. Also clear vacuoles or vacuoles containing flocculent matter were associated with CMBs (Fig.6). N o t all the constituents of these vacuoles could be identified, although they were considered to be remnants of cytoplasm. The CMBs, in some examples, were directly connected with granular endoplasmic reticulum. However ribosomes were not attached to any of the membranes of CMBs. A few CMBs, consisting entirely of membrane, were located within an endoplasmic reticulum cistern (Fig. 2). Virtually every neuron within the substantia nigra of the young Gunn rats examined contained at least one or two CMBs. The amount of area occupied by CMBs was not constant from neuron to neuron. Profiles of several neuronal somata were occupied almost entirely by CMBs. In other somata, however, only a few CMBs were seen and these were separated from one another by relatively normal appearing cytoplasm. There was no indication of an increase in the population of either microtubules or neurofilaments. The CMBs were less frequently seen in adult than in 2 week old brains. With age CMBs became concentrated further from the nucleus. Also, their membranes were organized in more regular arrays (Fig. 5). Furthermore they consisted, in older animals, essentially of membrane, but some of them did have an amorphous core (Fig. 5).
2. Endoplasmic Reticulum. The granular endoplasmic reticulum in proximity to CMBs in 2--6 week old animals assumed a dilated, angular appearance (Figs.1--3). This dilatation was occasionally sufficient to make the ergastoplasmic channels look swollen and vacuolar. This was not characteristic of substantia nigra neurons either in older G u n n rats or in 2 - - 6 week old control brains (Fig.4). Additionally, there was no perceptible loss of ribosomes from the ergastoplasmic membrane except where it was continuous with CMBs. Even in young Gunn rats free ribosomes were arranged in rosettes and were not dispersed as single units (Figs. 1--3). Fig./. This section through a portion of a neuronal soma shows several examples of CMBs of variable size and shape. The membranes surround bits of cytoplasm containing mitochondria (m), ribosomes (r) and vacuoles (v) containing moderately dense flocculent material. The endoplasmic reticulum channels (c) appear distended. The astrocytic processes (As) along the neuronal soma and the adjacent group of neurites (n) do not show any indication of cytoplasmic alteration. 2 week old Gunn rat; 27200 •
Fig. 2. A CMB composed entirely of swirls of membrane appears to lie within a cistern of granular endoplasmic reticulum. Observe the rosette arrangement of the u n b o u n d ribosomes (r). 2 weeks old G u n n rat; 27200 • Fig. 3. The cytoplasm of this neuronal soma contains distended endoplasmic reticulum cisternae (c) and a relatively large CMB that encircles a mitochondrion (m). Smaller CMBs, apparently connected with the endoplasmic reticulum, are indicated by arrowheads. The axon terminals (at) synapsing on this ceil body look normal, as do the other structures in the neuropil. 2 week old G u n n rat; 27200 •
Ultrastructure of the Gunn Rat Substantia Nigra
99
Fig. 4. The substantia nigra of several 2 week old Sprague-Dawley rats was studied and the cytological features of their neurons were compared with those of the Gunn rat. Notice in this figure that in the Sprague-Dawley rats the granular endoplasmic reticulum, in the form of a typical Nissl body (Nb), does not appear distended as it does in same aged Gunn rats. Also compare the size of these mitochondria (m) with those of the Gunn rat as seen in the previous figures. 2 week old Sprague-Dawley rat; 27 200 •
3. Single Membrane Bound Vacuoles. Pale, single membrane bound vacuoles measuring 0.2--2 ~ in diameter were c o m m o n in young animals (Figs. 6 and 7). The average diameter measured in 125 randomly selected vacuoles was 0.9 ~. They were plentiful in the vicinity of the nucleus (Fig. 7), closely adjacent to the Golgi system. They were also present at other sites in the cytoplasm, frequently in conjunction with CMBs (Fig. 6). Although generally spherical, their surface was rippled and even deeply indented (Figs.6 and 7). Their lumina held weakly dense osmiophilic material reminiscent of the thin thread-like elements that were also found in dilated ergastoptasmic cisternae. What might be interpreted as cellular detritus was sometimes contained in the lumen. The limiting membrane of these vacuoles was not always complete; consequently there was continuity between the lumen and cytoplasm. The vacuolar membranes, moreover, were not usually as sharp and crisp as other cellular membranes (Figs. 6 and 7). Often they were indistinct as though cut tangentially. 4. Mitochondria. The mitochondria were enlarged by one of two processes. Those in 2 - - 6 week old brains showed an expansion of the matrix compartment (Figs. 1 and 3). However other mitochondria in these same animals, but more frequently in older ones, were engorged with osmiophilic granules that produced an ex-
100
H . K . Batty and O. E. Millhouse
Fig. 5. With increasing age, the CMBs are located more peripherally in the neuronal soma. They also assume a more regular appearance. This CMB, which is only partially shown, nearly fills a dendritic profile, leaving only a rim of normal appearing cytoplasm that contains microtubules (arrowheads) and segments of agranular endoplasmic reticulum. The thick layer of membranes of this CMB surrounds a dense and relatively amorphous core (de). Adult Gunn rat; 46000 x
p a n s i o n o f the i n t r a c r i s t a l c o m p a r t m e n t . M o r e t h a n 200 c o m p a r a b l e m i t o c h o n d r i a l profiles were m e a s u r e d in the s u b s t a n t i a n i g r a o f b o t h 2 week c o n t r o l a n d G u n n r a t brains. T h e average d i a m e t e r in the f o r m e r was 0.39 lJ. a n d in the latter, 1.3 ix. O n l y those m i t o c h o n d r i a l profiles l o c a t e d in s o m a t a a n d n o t cont a i n i n g granules were m e a s u r e d . I t deserves e m p h a s i s t h a t this o b s e r v a b l e m i t o c h o n d r i a l m a t r i x e n l a r g e m e n t was c h a r a c t e r i s t i c only o f 2 - - 6 week o l d G u n n rats. A similar event was n o t f o u n d
Fig. 6. Single membrane bound vacuoles (V) containing a flocculent material were commonly seen in the neuronal cytoplasm of 2 week old Gunn rats. The rippled limiting membrane was not always distinct but often appeared to be blurred (arrowheads). The vacuoles frequently joined a CMB. 27200 • Fig. 7. Clusters of small vacuoles (vs) were usually found adjacent to the nucleus (Nu), in the region of the Golgi system. Except for their size, they are structurally similar to the larger vacuoles shown in Figure 6. 2 week old Gunn rat; 27200 x
Ultrastructure of the Gunn Rat Substantia Nigra
Figs. 6 and 7
101
102
H . K . Batty and O. E. Millhouse
Fig. 8. Only a few hyperchromatic neurons were seen in the G u n n rat substantia nigra. The nucleus (Nu) and cytoplasm are equally dense. Several CMBs can be seen in the cytoplasm. Notice that the surface of the neuronal soma is extremely irregular. Astrocytie processes (As) that seem to be hypertrophied surround the neuron. 2 week old G u n n rat; 27200 x
Ultrastructure of the Gunn Rat Substantia Nigra
103
in animals 12 weeks and older. There was no evidence of a consistent increase or decrease in matrix density nor was mitochondrial hyperplasia observed. 5. Hyperchromatic Neurons. Approximately 3--5 ~ of the neurons examined in younger animals were hyperchromatic (Fig. 8); however, they were not apparent in animals older than 6 weeks. These neurons contained pleomorphic CMBs, electron opaque vacuoles and enlarged mitochondria. The hyperchromasia appeared to result from an increased density of the cytoplasmic matrix. No abnormal concentrations of ribosomes were recognized (Fig. 8). The nucleus and cytoplasm were nearly equally dense and the nuclear membrane was interrupted in a few instances. 6. Comments on the Neuropil. The neuropil around the affected neuronal somata was not structurally modified to any observable extent, whether in 2 week old or adult Gunn rat (Figs. l, 3 and 5). Synaptic terminals on the somata were unchanged (Fig.3). Fascicles of small dendrites and unmyelinated axons did not reflect the obvious structural alterations seen in adjacent somata, except for the presence of osmiophilic granules in mitochondria. Comparison of light and electron microscopic preparations of Gunn rat and control brains did not indicate any significant neuron loss in the substantia nigra of the mature Gunn rat. Furthermore, there was no pronounced gliosis. DISCUSSION The cytological alterations of neurons in the substantia nigra of Gunn rats are assumed to be the effect of bilirubin on neuronal tissue. However bilirubin was not identified within any of the cells examined. One possible reason is that the youngest brains studied were two weeks old. Bilirubin may have been removed from the tissue at an earlier age. Schutta and Johnson (1967) pointed out that both the generalized and localized yellow staining characteristic of bilirubin deposits was no longer visible in Gunn rat brains by the twentieth post-natal day. Also, there is a close correspondence between the structural changes we observed and those reported by Silberberg and Schutta (1967), who exposed cultures of rat cerebellum to specific concentrations of bilirubin. Furthermore, Chen et al. (1969) injected tritium labeled bilirubin into rabbits that had been previously subjected to anoxia in order to circumvent the blood-brain barrier. Although the label can be seen in their micrographs, it is not associated with a unique cellular inclusion, crystal or pigment molecule. Consequently, bilirubin, if present, may be sequestered or modified so that it cannot be demonstrated directly with conventional electron microscopic techniques. The varied appearance of CMBs may indicate that bilirubin exerts more than a single effect on neurons and neuronal metabolism. The more compact membranous whorls are similar to the "membranous cytoplasmic body" described by Terry and Weiss (1963) in Tay-Sachs disease. Samuels et al. (1965) studied the possible modes of formation of these membrane bodies and concluded that they arise from the spontaneous aggregation of gangliosides, cholesterol and phospholipids. By combining these three substances in the ratio in which they occur in the disease, Samuels et al. were able to generate the characteristic membranous
104
H.K. Batty and O. E. Millhouse
bodies extracellularly. They concluded that these membrane whorls were not necessarily derived from a specific organelle and they doubted that these bodies were lysosomes. Claireux et al. (1953) demonstrated that bilirubin forms a complex with a neuronal lipid which was not further specified. On the basis of this observation and the experiments of Samuels et al. (1965), Schutta and Johnson (1967) argued that such a bilirubin-lipid complex might produce a compound that forms the membranous whorls seen in Purkinje cells of the Gunn rat. It was assumed that this compound was not readily metabolized by neurons and, thus, accumulated in the cytoplasm in the form of membranous bodies. Lullmann-Rauch et al. (1972) also conclude that the membranous bodies they described in lung tissue following injections of chlorphentermine represent a non-digestible phospholipid that accumulates because of the interaction of the drug with certain phospholipids. However, the CMBs observed in Gunn rat neurons possibly represent more than the passive accumulation of a non-metabolizable molecular chain. Stern et al. (1972) have recently exposed long-term organotypic tissue cultures of rat spinal cord to various concentrations of sulfatide, ganglioside and cerebroside. Membranous bodies strikingly similar to those seen in Tay-Sachs and other storage diseases developed in the neurons in these cultures. Significantly, acid phosphatase activity was demonstrated in these cytoplasmic bodies, thus, proving them to be lysosomes. Lullmann-Rauch et al. (1972) also demonstrated acid phosphatase activity in similar appearing membrane whorls of rat pulmonary cells. Yet the majority of CMBs observed in Gunn rat neurons consisted of more than layers of membrane. The single, membrane bound vacuoles filled with flocculent material closely correspond in structure, content, and size to the autophagic vacuoles observed in cultured macrophages exposed to chloroquine (Fedorko et al., 1968). With time, these autophagic vacuoles were reported to increase in size, become irregular in shape, invaginate and to contain " . . . cytoplasmic components apparently in various stages of digestion" as well as membrane whorls, amorphous matter and lipid droplets. CMBs also bear close resemblance to the cytological structures reported in other types of cellular injury (Hruban et al., 1963; Colonnier, 1964; Barron et al., 1967, 1973; Hwang et al., 1974; Lullmann-Rauch et al., 1972; Lullmann-Rauch, 1974; Lullmann-Rauch and Pietschmann, 1974). Considering that CMBs consist of multilaminar membranes frequently surrounding segregated pieces of cytoplasm and considering, too, their structural similarity with the membranous bodies produced by Stern et al. (1972), the myeloid bodies of Barron et al. (1973), the autophagic vacuoles described by Fedorko et al. (1968) and the inclusions seen in examples of focal cytoplasmic degradation (Hruban et al., 1963), we are led to interpret the CMBs as lysosomes, representing autophagic activity. The origin of the membrane lamellae associated with late autophagic vacuoles or residual bodies has not been explained satisfactorily. Fedorko et al. (1968) interpreted their data to argue that chloroquine acts on the membranes of the Golgi system and agranular endoplasmic reticulum to produce a fusion of the membranes with one another. Recalling the experiments of Samuels et al. (1965),
Ultrastructure of the Gunn Rat Substantia Nigra
105
the CMBs could conceivably arise as a combination of bilirubin with some neuronal lipid. This novel molecular complex may trigger an autophagic response. As a consequence the membranous complex would become associated with autophagic vacuoles, and, predictably, acid hydrolase. Hwang et al. (1974) studied the formation of membranous whorls in liver parenchymal cells exposed to cycloheximide. Their biochemical experiments showed that during the time these whorls were being formed protein synthesis was decreased 40--60~. However at the same time, phospholipid synthesis was increased. They concluded that the membrane bodies result from a desynchronization in the synthesis of the protein and lipid components of membrane. Bilirubin may also precipitate an autophagic response by another means. Cowger (1971) has shown that bilirubin affects cells in a manner similar to that of lipophilic surfactants. Thus, it would not be unreasonable to expect that bilirubin partially exerts its effects on neurons by affecting lipids of the cytomembrane system, possibly altering membrane permeability. Should this occur, hydrolytic enzymes could escape into the surrounding cytoplasm and cause focal cytoplasmic degradation. It is not difficult to envision a marshalling of a sequence of adaptive responses which would contain the degradation. But containment is not the only point to consider: Novikoff (1967) points out that autophagic activity may be a means of assimilating the cytologically changed areas into the intracellular economy. Bilirubin is also known to act as a respiratory poison by uncoupling phosphorylation and inhibiting mitochondrial respiration (see Cowger, 1971). The morphological expression of this is seen as an increase in mitochondrial matrix (Tapley and Cooper, 1956; Lehninger, 1962, 1965). The resulting decrease in intracellular levels of ATP may also contribute to cytoplasmic injury. Because there is no evidence for either a precipitous loss of neurons or a significant proliferation of glia in the substantia nigra of adult Gunn rats, it appears that the majority of these neurons recover from exposure to bilirubin. However not all the neurons survived. A small proportion, 3--5 ~ , were hyperchromatic, a cytological feature generally regarded as characteristic of dying or dead neurons (Barron et al., 1974; Matthews, 1973). Although the reason(s) for this cell death has not been established, two points should be considered. One, the cell death may reflect the deposition of different concentrations of bilirubin in the substantia nigra so that only a few neurons were exposed to a lethal dose. Two, there is evidence that autophagocytosis is dependent upon the maintenance of a certain intracellular level of ATP (Trump et al., 1971 ; Shelburne et al., 1973). If autophagocytosis cannot occur, the consequence could be cell death. Regarding this, Cowger (1971) has pointed out that the effect of bilirubin on mitochondrial respiration p e r s e does not cause cell death. The cellular events observed by examining the substantia nigra of various aged animals indicate to us that neurons exposed to certain (but, as yet, unspecified) concentrations of bilirubin undergo extensive autophagocytosis, but providing that there are sufficient levels of ATP, the neuronal cytoplasm is renewed. Autophagocytosis, then, may be viewed as being important in (1) containing the toxic effects of bilirubin and paradoxically, (2) reconstituting the cytoplasm (Novikoff, 1967).
106
H . K . Batty and O. E. Millhouse
Acknowledgements. We are grateful to Drs. M. L. Grunnet, L. J. Stensaas, W. Stevens and, in particular, L. W. Jarcho for advice on the manuscript. Mesdames D. Lerdahl and S. Carter provided technical assistance as did Mr. K. Johnson. Mrs. J. English helped with the final preparation of this paper. Their exceptionally fine help is very much appreciated. This research was partially supported by grants GM-00958 (NIH) and 5 T01 NS05309 (NIH) and the Eleanor Roosevelt Cancer Foundation Research Institute. REFERENCES Aleu, F. P., Terry, R. D., Zellweger, H. : Electron microscopy of two cerebral biopsies in gargoylism. J. Neuropath. exp. Neurol. 24, 304--317 (1965) Barron, K. D., Doolin, P. F., Oldershaw, J. B.: Ultrastructural observations on retrograde atrophy of lateral geniculate body. Neuronal Alteration J. Neuropath. exp. Neurol. 26, 300--326 (1967) Barron, K. D., Means, E. D., Larsen, E. : Ultrastructure of retrograde degeneration in thalamus of rat. J. Neuropath. exp. Neurol. 32, 218--244 (t973) Blanc, W. A., Johnson, L.: Studies on Kernicterus. Relationships with sulfonamide intoxication, report on kernicterus in rats with glucuronyl transferase deficiency and review of pathogenesis J. Neuropath. exp. Neurol. 18, 165--187 (1959) Blanc, W. A. : Kernicterus in Gunn's strain of rat. In: Kernicterus (ed. A. Sass-Korstak), pp. 150-- 152. Toronto: University of Toronto Press 1961 Carbone, J. V., Grodsky, G. M. : Constitutional nonhemolytic hyperbilirubinemia in the rat: Defect of bilirubin conjugation. Proc. Soc. exp. Biol. (N.Y.) 94, 461--463 (1957) Chen, H., Tsai, D., Wang, C., Chen, Y. : An electron microscopic and radioautographic study on experimental kernicterus. Amer. J. Path. 56, 31--58 (1969) Claireaux, A. E., Cole, P. G., Lathe, G. H. : Icterus of the brain of the newborn. Lancet 1953 II, 1226--1230 Colonnier, M. : Experimental degeneration in the cerebral cortex. Amer. J. Anat. 98, 4 7 - 5 3 (1964) Cowger, M. L. : Mechanism of bilirubin toxicity on tissue culture cells: Factors that affect toxicity, reversibility by albumin, and comparison with other respiratory poisons and surfactants. Biochem. Med. 5, 1016 (1971) Diamond, I., Schmid, R. : Experimental bilirubin encephalopathy. The mode of entry of bilirubin-14C into the central nervous system. J. clin. Invest. 45, 6 7 8 - 6 8 9 (1966) Ernster, L., Herlin, L., Zetterstrom, R. : Experimental studies of the pathogenesis of kernicterus. Pediatrics 20, 647--652 (1957) Farrell, D. F., Baker, H. J., Herndon, R. M., Lindsey, J. R., McKhann, G. M. : Feline GM1 gangliosidosis: Biochemical and ultrastructural comparisions with the disease in man. J. Neuropath. exp. Neurol. 32, 1 (1973) Fedorko, M. E., Hirsch, J. G., Cohn, Z. A. : Autophagic vacuoles produced in vitro. II. Studies on the mechanism of autophagic vacuoles produced by chloroquine. J. Cell Biol. 38, 392--402 (1968) Gardner, W. A., Konigsmark, B. W. : Familial nonhemolytic jaundice: Bilirubinosis and encephalopathy. Pediatrics 43, 365--376 (1969) Gunn, C. K. : Hereditary acholuric jaundice in the rat. Canad. reed. Ass. J. 50, 230--237 (1944) Hruban, Z., Spargo, B., Swift, H., Wissler, R. W., Kleinfeld, R. G. : Focal cytoplasmic degradation. Amer. J. Path. 42, 657--683 (1963) Hwang, K. M., Yang, L. C., Carrico, C. K., Schulz, R.A., Schenkman, J.B., Sartorelli, A. C. : Production of membrane whorls in ra liver by some inhibitors of protein synthesis. J. Cell Biol. 62, 20--31 (1974) Johnson, L. : The effects of certain substances on bilirubinic levels and the occurrence of kernicterus in generally jaundiced rats. In: Kernicterus (ed. A. Sass-Korstak), pp. 208--218. Toronto: University of Toronto Press t961
Ultrastructure of the Gunn Rat Substantia Nigra
107
Johnson, L., Garcia, M. L., Figueroa, E., Sarmiento, F.: Kernicterus in rats lacking glucuronyl transferase. Am. J. Dis. Child. 101, 3 2 2 - 3 4 9 (1961) Kirkpatrick, J. B.: Chromatolysis in the hypoglossal nucleus of the rat: An electron microscopic analysis. J. comp. Neurol. 132, 189 212 (1968) Lehninger, A. L. : Water uptake and extrusion by mitocbondria in relation to oxidative phosphorylation. Physiol. Rev. 42, 467--517 (1962) Lehninger, A. L.: In: The Mitochondrion, molecular basis of structure and function. New York: W. A. Benjamin 1964 Ltillmann-Rauch, R. : Lipodosis-like alterations in hypothalamic neurosecretory cells of rats treated with chlorphentermine or iprindole. Cell Tiss. Res. 149, 587--590 (1974) Ltillmann-Rauch, R., Reil, G. H., Rossen, E., Seiler, K. U. : The ultrastructure of rat lung changes induced by an anorganic drug (chlorphentermine). Virchows Arch., Abt. B, Zellpatb. 11, 167--181 (1972) Ltillmann-Rauch, R., Pietschmann, N. : Lipodosis-like cellular alterations in lymphatic tissues of chlorphentermine-treated animals. Virchows Arch., Abt. B, Zellpath. 15, 295--308 (1974) Matthews, M. A. : Death of the central neuron: an electron microscopic study of thalamic retrograde degeneration following cortical ablation. J. Neurocytol. 2, 265--288 (1973) Novikoff, A. B. : Lysosomes in nerve cells. In: The neuron (ed. H. Hyden), pp. 255 377. Amsterdam-London-New York: Elsevier Publ. Co. 1967 Odell, G. B. : Factors influencing the binding of bilirubin by albumin in kernicterus. In: Kernicterus (ed. A. Sass-Korstak), pp. 199--207. Toronto: University of Toronto Press 1961 Reynolds, E. S. : The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17, 208--212 (1963) Samuels, S., Gonatas, N. K., Weiss, M. : Formation of membranous cytoplasmic bodies in Tay-Sachs disease: An in vitro study. J. Neuropath. exp. Neurol. 24, 256--264 (1965) Schmid, R., Axelrod, J., Hammaker, L., Swarm, R. L. : Congenital jaundice in rats, due to a defect in glucuronide formation. J. clin. Invest. 37, 1123 1130 (1958) Schutta, H. S., Johnson, L. : Bilirubin encephalopathy in the Gunn rat: A fine structural study of the cerebellar cortex. J. Neuropath. exp. Neurol. 26, 377--396 (1967) Shelburne, J. D., Arstila, A. U., Trump, B. F.: Studies on cellular autophagocytosis, cyclic A M P and dibutyryl cyclic A M P stimulated autophagy in rats. Amer. J. Path. 72, 521--540 (1973) Silberberg, D. H., Schutta, H. S. : The effects of unconjugated bilirubin and related pigments on cultures of rat cerebellum. J. Neuropath. exp. Neurol. 26, 572 583 (1967) Stern, J., Novikoff, A. B., Terry, R. D. : The induction of sulfatide, ganglioside, and cerebroside storage in organized nervous system cultures. In: Sphingolipids, sphinglolipidoses, and allied disorders (eds. B. W. Volk and S. M. Aronson), pp. 651--660. New York: Plenum Press 1972 Tapley, D. F., Cooper, G. : Effect of thyroxine on the swelling of mitochondria isolated from various tissues of the rat. Nature (Lond.) 167, 1119 (1956) Terry, R. D., Weiss, M.: Studies in Tay-Sachs disease. II. Ultrastructure of the cerebrum. J. Neuropath. exp. Neurol. 22, 18--55 (1963) Trump, B. F., Croker, B. P., Mergener, W. J. : The role of energy metabolism, ion, and water shifts in the pathogenesis of cell injury. In: Cell membranes, biological and pathological aspects (eds. G. W. Richter, D. G. Scarpelli), pp. 84--128. Baltimore: The Williams and Wilkins Co. 1971 Waters, W. J., Richert, D. A., Rawson, H. H. : Bilirubin encephalopathy. Pediatrics 13, 319--325 (1954) Waters, W. J., Britton, H. A. : Bilirubin encephalopathy: Preliminary studies related to production. Pediatrics 15, 45--48 (1955) Wuerker, R. B., Palay, S. L. : Neurofilaments and microtubules in anterior horn cells of the rat. Tissue and Cell 1, 387--402 (1969)
Received September 10, 1975; Accepted December 22, 1975
