- Email: [email protected]
Histopathological changes in the brain of mice given Clostridium perfringens type D epsilon toxin
J, COMP. PATH. 1984. Vat.
363
94.
HISTOPATHOLOGICAL CHANGES IN THE MICE GIVEN CLOSTRIDIUM PERFRINGENS EPSILON TOXIN
Department
of Pathology,
of Adelaide,
UniversiQ
Frame
Road,
Adelaide,
BRAIN TYPE
S.A. 5001,
OF D
Australia
INTRODUCTION
The epsilon toxin of Clostridium perfringens (welchii) Type D causes a neurological disorder of young iambs characterized by bilaterally symmetrical malacic lesions in the basal ganglia, internal capsule, thalamus and substantia nigra, known as focal symmetrical encephalomalacia (FSE) (Hartley, 1956). It occurs in those sheep which survive acute intoxication and thereby undergo a more protracted clinical course. Griner (1961) and Morgan and Kelly (1974) found mice to be useful experimental models for the ovine disease and produced brain lesions of similar nature and anatomical distribution to those found in FSE. The pathogenesis appears to be associated with toxin damage to the cerebral vascular endothelium (Gardener, 1973; Morgan and Kelly, 1974). The precise sequence of events, however, from the apparent initial diffuse cerebral oedema to the development of malacic foci in the brain is poorly understood, both in terms of the ultrastructural alterations which lead to necrosis and the apparent selective vulnerability of certain regions in the brain. This paper reports the results of experiments designed to elucidate the pathogenesis. MATERIALS
AND
METHODS
Studies were conducted on 8 to lo-week-old, outbred Swiss-white mice weighing 20 to 30 g. The toxin used was a partially purified prototoxin prepared from filtrates of broth cultures of ClostridiumperfringensType D. One gram of 0.25 per cent trypsin (Difco) was dissolved in 25 ml of phosphate buffered O-9 per cent sodium chloride solution and 1 ml of a final dilution of this 1 in 250 solution was incubated at 37°C for 45 min to permit enzyme activation of the toxin, then stored in aliquots at - 20°C until required. The trypsin-activated toxin was injected intraperitoneally into mice at a dose rate of 0.5 ml per mouse, after 1 in 300 dilution of toxin as a lethal dose, or after 1 in 3000 dilution as a sublethal prolonged course. The lethal doses killed 4 to 6 h after injection. The brains were fixed by a perfusion technique, Mice were anaesthetized by chloroform, the thorax opened rapidly, the right auricle incised and approximately 20 ml of heparinized saline (9 g NaCl per 1 containing 0.02 per cent heparin) was perfused into the left ventricle through a needle inserted into the apex of the heart. When this solution had been infused and clear fluid issued from the auricle, the injection of 20 ml of 10 per cent buffered formalin was commenced in a similar manner 002 l-9975/84/030363
+ 08 $03.00/O
0 1984 Academic
Press Inc. (London)
Limited
364
J.
W.
FINNIE
into the beating left ventricle. The brain was then rapidly removed, immediately immersed in fixative and 1 mm thick coronal slices cut. Tissues were processed to paraffin wax and 6 pm thick sections were cut and stained with haematoxylin and eosin (HE). Selected sections were also stained with the periodic acid-Schiff technique. Ex@eriment 1. Distribution, Severity and Frequency of Brain Lesions Two groups of mice were given sublethal injections of toxin and killed 24 h postinoculation. The first group contained 60 mice and received 3 injections over the 24 h period, at 8 h intervals, while the second group, comprising 30 mice, received only a single injection. Histological assessment of the severity of lesionswas made by a 0 to + -t + system of grading which is specified in Table 1. Experiment2. SequentialChanges in Brain Morphology Mice in one group were given a lethal injection of toxin and were killed from 1 to 6 h post-inoculation. Another group were given multiple sublethal injections and killed from 6 h to 7 days post-injection. Five mice were killed at each time interval. Brains from the first group were perfusion-fixed while those in the second group were immersion-fixed. RESULTS
Experiment 1. Distribution, Seuerity and Frequency of Brain Lesions The results of this study are summarized in Table 1. In mice given multiple sublethal doses of toxin, lesions were found most commonly in the corpus striatum, cerebral cortex, vestibular area, corpus callosum and callosal radiations, and corpus medullare cerebelli. Less frequently, lesions occurred in the thalamus, granular layer of the cerebellum, paraventricular area lateral to the lateral ventricles, anterior commissures, substantia nigra, the fimbria and alveus hippocampi and fornix (Figs 1 and 2). No malacic foci were found in the spinal cord. In the mice receiving a single dose of toxin, lesions only occurred in a few mice, but the granular layer of the cerebellum was the region where necrosis was seen most often. Experiment 2. SequentialChangesin Brain Morphology With good perfusion, the brain was firm, rubbery and yellowish in colour, with no blood in meningeal vessels. Affected mice showed no gross alterations apart from the occasional animal in which the cerebellum was partially herniated through the foramen magnum. Microscopically, in control brains, the neuropil was compact. Blood vessels were preserved in an open state and capillaries were in direct apposition to the neuropil. Astrocytic end-feet were not identifiable by light microscopy. In toxin-treated brains, at 1 h the most obvious change was the presence of clear spaces around most vessels, sometimes imparting a scalloped appearance around vesselsof capillary size. While blood vesselscould still be found readily with lumina fixed in the fully open state and in close apposition to the neuropil, many were collapsed and commonly contained erythrocytes which had not
EPSILON
TOXIN
IN TABLE
D~STRI~~“T~ON AND SEVERITY
OF BRAIN
Graded
365
BRAIN
1
LESIONS IN MICE AFTER THE INJECTION
Nwnbers of animals affected
severity
of lesions
Cerebral cor,ex Corpus callosum Corpus striatum Thalamus Paraventricular area of lateral wntricles Vestibular area Corpus medullare cercbeili Granular layer of cerebellum Substantia nigra Anterior commissure
MOUSE
+
++
++t
9 4
4 3
A
4
9 I 1
5 7
OF TYPE D TOXIN
++
3
1)
Numbers of animals affected
Graded severity of lesions +
(EXPT
t++ -
3
-
1 -
3 1
l-4 13 14 6
T -
-
1 7
-
6 14
1
-
I
4
6
1
11
1
-
1
2
2 I 1
4 1 1
1 2
4 -
-
Number of animals in,jected Number of animals with lesions Degree of srverity of lesions. + =Mild vacuolation of the + + =Moderate vacuolation, + + + =Severe vacuolation,
-
60
30
32
7
-
5 2 -
neuropil and macroglial reaction. with necrosis of a few glial cells. with necrosis of most glial elements.
been flushed out during perfusion and probably represented vascular stasis. Nearby vessels which were fully patent very rarely contained erythrocytes. Fine vacuolation of the neuropil was present in many areas and astrocytes were often surrounded by clear spaces and the nuclei of these cells were larger and paler than normal. Some neurones were also surrounded by small clear spaces but the neurones themselves appeared unaltered. The above changes became progressively more obvious and at 4 h the neuropil had a coarse, vacuolated appearance. In some areas, discrete vacuoles gave a distinctly bubbly appearance to the neuropil. Occasionally there were larger spaces in the white matter with considerable separation and some fragmentation of nerve fibres, and many astrocytes in these lesions now possessed a visible amount of eosinophilic cytoplasm. Vacuolation of the Purkinje cell layer of the cerebellum was conspicuous and the granular layer was less compactly arranged. A few small foci of necrosis were present in the granular layer, with pyknotic nuclei of granule cells disposed in a non-staining matrix (Fig. 3). At 6 h, perivascular spaces were distended but contained no plasma exudate and most vessels were markedly hyperaemic. In the white matter, capillary haemorrhage was sometimes observed and astrocytes were reactive and formed gemistocytes or, occasionally, were degenerate with nuclear pyknosis and karyorrhexis. At 12 h, small amounts of faintly eosinophilic plasma exudate were found
366
J.
W.
FINNIE
EPSILON
TOXIN
IN
MOUSE
367
BRAIN
adjacent to a few small blood vessels, and astrocytic nuclei were swollen, oval eosinophilic bodies vesicular and hypochromatic. Occasionally, representing swollen axons (spheroids or retraction bulbs) were visible in areas of white matter damage. At 18 h, fragmentation of glia in white matter lesions was more evident and in some of these areas, especially the corpus callosum, there were lakes of deeply eosinophilic plasma exudate which had a foamy, vacuolated appearance and stained strongly positive with the PAS technique. In the cerebral cortex, discrete foci of spongy change were sometimes found and neurones in these foci were shrunken and deeply acidophilic; neurones in adjacent unaffected areas of cortex, however, appeared normal. In areas of the corpus striatum (where grey and white matter are closely admixed) showing severe oedematization, some neurones had large vesicular nuclei and cytoplasmic vacuolation and a few were shrunken and deeply eosinophilic. At this time, especially in the vestibular tracts, axons not uncommonly showed irregular beaded swellings and were surrounded by an expanded, varicose, non-staining myelin sheath. With fragmentation and retraction of the axon and its sheath, ellipsoids formed and often enclosed an oval, eosinophilic swollen axon (digestion chambers). At a later stage, ellipsoids became less welldefined, leaving an irregular pale-staining focus containing eosinophilic axonal debris, at the periphery of which compound granular corpuscles (gitter cells) had appeared. At 24 h, white matter lesions had in some areas progressed to malacia, with the neuropil presenting a lace-like appearance. Fibres were widely separated by clear spaces containing a few small, round deeply basophilic nuclei, most of which represented pyknotic astrocytic nuclei with a few surviving oligodendrocytes. Extravasated erythrocytes were sometimes numerous in malacic foci and, at the periphery of these lesions, astrocytic nuclei were enlarged and hypochromatic, endothelial cells were swollen, and there was a slight increase in the number of microglial cells. No polymorphs or lymphocytes were seen in any malacic foci in the brain at any stage. A few necrotic areas had undergone complete liquefaction necrosis, with empty spaces devoid of any tissue and this change was most commonly observed in the corpus callosum, sometimes resulting in separation of the white matter from the overlying cortical grey matter. In brains examined at 7 days, the malacic foci were almost completely filled with gitter cells, which had small, round nuclei and a clear, vacuolated or finely granular cytoplasm. These lesions also contained fibrous astrocytes and axonal debris, and capillaries were seen to be invading the organizing focus at the perimeter.
Fig. 1. Bilaterally thalamus, receiving
symmetrical lesions in the corpus callosum, paraventricular fimbria ofthe hippocampus and a focal lesion in the cerebral multiple sublethal doses of toxin (Expt. I). HE x 7.
Fig. 2. Bilaterally sublethal
symmetrical doses of toxin
Fig. 3. Focal necrosis dose of toxin
malacic (Expt.
in the granular (Expt. 2). HE
foci in the vestibular 1). HE x 7. layer x 35.
of the cerebellum
area
of a mouse
of a mouse
cortext
area, corpus of a mouse
24 h after
4 h after
receiving
striatum, 24 h after
receiving a single
multiple lethal
368
J.
W.
FINNIE
Lesions consistently manifested 3 features. The malacic foci were often focal; they were commonly bilaterally symmetrical; and they were essentially confined to the white matter. DISCUSSION
In Experiment 1, the distribution of lesions was consistent with findings in mice by Griner (1961) and Morgan and Kelly (1974), and similar to the pattern of lesions in lambs described by Hartley (1956) and Griner ( 1961). In mice, however, more regions of the brain appeared to be susceptible than in lambs. The proportion of mice which developed lesions increased when multiple, rather than single, doses of toxin were given, which is in agreement with observations by Morgan and Kelly (1974). Morgan and Kelly (1974) considered that the dorso-lateral aspect of the corpus medullare cerebelli and the paraventricular areas lateral to the lateral ventricles were the sites where the initial lesions developed. The corresponding site in the present study may be in the granular layer of the cerebellum. This was the necrotic area in mice given single injections. The bilateral symmetry of lesions in FSE is one of the hallmarks of this condition (Hartley, 1956). Oedema in the brain causes increased intracranial pressure and may secondarily produce local lesions by compression of arteries (Lindenberg, 1955). B ecause the pressure exerted by oedema in this manner is symmetrical, it is to be expected that the lesions in this condition are also symmetrical. The early lesions induced in the brain of mice by epsilon toxin were essentially those of oedema, being more severe in the white matter. It has long been recognized that white matter is much more susceptible to oedema and it is here that oedema can lead to tissue destruction (Manz, 1974). The consistent and often extensive involvement of the more compactly arranged myelinated fibre tracts, such as the corpus callosum and internal capsule, in the present study was, however, unexpected as these tracts are relatively resistant to the destructive effects of oedema (Manz, 1974). Furthermore, Jubb and Kennedy (1970) stated that the permeability of vessels to trypan blue in lambs with acute enterotoxaemia was diffuse throughout the’ brain, but the heavily myelinated tracts were spared. These findings suggest that necrosis in these tracts may be due to direct vascular involvement and not to simple passive accumulation of oedema fluid. The internal capsule, for example, is supplied by end-arteries (Brierley, 1977). Lesions were commonly found in paraventricular areas lateral to the lateral ventricles and these regions are thought to be predisposed to ischaemia since they derive their main blood supply from a system of primitive penetrating arteries which are essentially end-arteries (De Reuck, Chattha and Richardson, 1972). The basal ganglia, corpus striatum and thalamus, regions often involved in intoxicated brains in the present study, are highly vulnerable to hypoxia and in experimental enterotoxaemia in lambs there is a profound systemic hypoxia (Gardner, 1973). This may have produced a secondary reduction in brain
EPSILON
TOXIN
IN
MOUSE
BRAIN
369
perfusion pressure due to impairment of cardio-respiratory function (Brierley, 1977). The brain is a relatively impoverished tissue in terms of capillary density and, in an idealized mid-intercapillary region, doubling the intercapillary distance would reduce to nil the oxygen tension by extending the length of the diffusion path (Bourke, Kimelberg, Nelson, Barron, Auen, Popp and Waldman, 1980). Since astrocytic processes are disposed as satellites around neurones and their end-feet cover most of the surface of capillaries, they are in a potentially strategic position to alter the supply of blood-borne nutrients to with astrocytic swelling, the dependent neurones. In oedematous conditions, diffusion path length between vessels and neurones may be exceeded in focal scattered areas of grey matter and result in cortical lesions of the type found in the present study. Similarly, focal necrosis in the cerebellar granular layer may be due to interference with the supply of nutrients to these cells by early and severe swelling of astrocytes. The bilaterally symmetrical malacic lesions in the vestibular area, commonly found in toxin-treated mice in the present study, also occur in Chastek’s paralysis of carnivores. The sequential pathogenesis of lesions in these latter animals is oedema, vascular dilatation, haemorrhage and necrosis Wubb and Kennedy, 1970), which is similar to the progression of events in the mice, but an explanation for the topographical distribution of lesions in Chastek’s paralysis has not been forthcoming.
SUMMARY
The distribution, severity and frequency of brain lesions produced in mice by the administration of C’lostridium perfringens Type D epsilon toxin were examined by light microscopy. The granular layer of the cerebellum was the area most frequently affected in mice given single doses of toxin. Sequential changes in brain morphology were examined from 1 h to 7 days after injection of toxin. Lesions progressed from an initial vasogenic oedema to malacic foci which commonly were focal and bilaterally symmetrical, with a predilection for white matter. The topographical distribution of these malacic areas is discussed.
REFERENCES
Bourke, R. L., Kimelberg, H. K., Nelson, L. R., Barron, K. D., Auen, E. L., Popp, A. J7 and Waldman, J. B. (1980). Biology of glial swelling in experimental brain oedema. In Advances in ..Meurology, Vol. 28, Raven Press,New York, p. 100. Brierley, J. B. (1977). Experimental hypoxic brain damage. Journal of Clinical Pathology, 30, SuppI. 11, 181-187. DeReuck, J., Chattha, A. S., and Richardson, E. P. (1972). Pathogenesisand evolution of periventricular leukomalacia in infancy. Archives of Neurology, 27, 229-236. Gardener, D. E. (1973). Pathology of Clastridium welchii Type D enterotoxaemia. II. Structural and ultrastructural alterations in the tissuesof lambs and mice. Journal of Comparative Pathology, 83, 509-524.
Griner, L. A. (1961). Enterotoxaemia of sheep.I. Effects of Clostridium perfringens Type
370
J.
W.
FINNIE
D toxin on the brains of sheep and mice. American Journal of Veterinary Research, 22, 429-442. Hartley, W. J. (1956). A focal symmetrical encephalomalacia of lambs. New Zealand Veterinary Journal, 4, 129-l 35. Jubb, K. V. F., and Kennedy, P. C. (1970). Pathology of Domestic Animals, Vol. 1. Academic Press, New York, pp. 115, 385. Lindenberg, R. (1955). Compression of brain arteries as pathogenic factor for tissue necroses and their areas of predilection. Journal of Neuropathology and Exberimental Neurology, 14, 223-243. Manz, H. J. (1974). The pathology of cerebral oedema. Human Patholou 5, 291-313. study of brain lesions Morgan, K. T., and Kelly, B. G. (1974). Ultrastructural produced in mice by the administration of Clostridium welchii Type D toxin. Journal of Comparative Pathology, 84, 181-191. [Received for publication,
May 1 Oth, 19831
363
94.
HISTOPATHOLOGICAL CHANGES IN THE MICE GIVEN CLOSTRIDIUM PERFRINGENS EPSILON TOXIN
Department
of Pathology,
of Adelaide,
UniversiQ
Frame
Road,
Adelaide,
BRAIN TYPE
S.A. 5001,
OF D
Australia
INTRODUCTION
The epsilon toxin of Clostridium perfringens (welchii) Type D causes a neurological disorder of young iambs characterized by bilaterally symmetrical malacic lesions in the basal ganglia, internal capsule, thalamus and substantia nigra, known as focal symmetrical encephalomalacia (FSE) (Hartley, 1956). It occurs in those sheep which survive acute intoxication and thereby undergo a more protracted clinical course. Griner (1961) and Morgan and Kelly (1974) found mice to be useful experimental models for the ovine disease and produced brain lesions of similar nature and anatomical distribution to those found in FSE. The pathogenesis appears to be associated with toxin damage to the cerebral vascular endothelium (Gardener, 1973; Morgan and Kelly, 1974). The precise sequence of events, however, from the apparent initial diffuse cerebral oedema to the development of malacic foci in the brain is poorly understood, both in terms of the ultrastructural alterations which lead to necrosis and the apparent selective vulnerability of certain regions in the brain. This paper reports the results of experiments designed to elucidate the pathogenesis. MATERIALS
AND
METHODS
Studies were conducted on 8 to lo-week-old, outbred Swiss-white mice weighing 20 to 30 g. The toxin used was a partially purified prototoxin prepared from filtrates of broth cultures of ClostridiumperfringensType D. One gram of 0.25 per cent trypsin (Difco) was dissolved in 25 ml of phosphate buffered O-9 per cent sodium chloride solution and 1 ml of a final dilution of this 1 in 250 solution was incubated at 37°C for 45 min to permit enzyme activation of the toxin, then stored in aliquots at - 20°C until required. The trypsin-activated toxin was injected intraperitoneally into mice at a dose rate of 0.5 ml per mouse, after 1 in 300 dilution of toxin as a lethal dose, or after 1 in 3000 dilution as a sublethal prolonged course. The lethal doses killed 4 to 6 h after injection. The brains were fixed by a perfusion technique, Mice were anaesthetized by chloroform, the thorax opened rapidly, the right auricle incised and approximately 20 ml of heparinized saline (9 g NaCl per 1 containing 0.02 per cent heparin) was perfused into the left ventricle through a needle inserted into the apex of the heart. When this solution had been infused and clear fluid issued from the auricle, the injection of 20 ml of 10 per cent buffered formalin was commenced in a similar manner 002 l-9975/84/030363
+ 08 $03.00/O
0 1984 Academic
Press Inc. (London)
Limited
364
J.
W.
FINNIE
into the beating left ventricle. The brain was then rapidly removed, immediately immersed in fixative and 1 mm thick coronal slices cut. Tissues were processed to paraffin wax and 6 pm thick sections were cut and stained with haematoxylin and eosin (HE). Selected sections were also stained with the periodic acid-Schiff technique. Ex@eriment 1. Distribution, Severity and Frequency of Brain Lesions Two groups of mice were given sublethal injections of toxin and killed 24 h postinoculation. The first group contained 60 mice and received 3 injections over the 24 h period, at 8 h intervals, while the second group, comprising 30 mice, received only a single injection. Histological assessment of the severity of lesionswas made by a 0 to + -t + system of grading which is specified in Table 1. Experiment2. SequentialChanges in Brain Morphology Mice in one group were given a lethal injection of toxin and were killed from 1 to 6 h post-inoculation. Another group were given multiple sublethal injections and killed from 6 h to 7 days post-injection. Five mice were killed at each time interval. Brains from the first group were perfusion-fixed while those in the second group were immersion-fixed. RESULTS
Experiment 1. Distribution, Seuerity and Frequency of Brain Lesions The results of this study are summarized in Table 1. In mice given multiple sublethal doses of toxin, lesions were found most commonly in the corpus striatum, cerebral cortex, vestibular area, corpus callosum and callosal radiations, and corpus medullare cerebelli. Less frequently, lesions occurred in the thalamus, granular layer of the cerebellum, paraventricular area lateral to the lateral ventricles, anterior commissures, substantia nigra, the fimbria and alveus hippocampi and fornix (Figs 1 and 2). No malacic foci were found in the spinal cord. In the mice receiving a single dose of toxin, lesions only occurred in a few mice, but the granular layer of the cerebellum was the region where necrosis was seen most often. Experiment 2. SequentialChangesin Brain Morphology With good perfusion, the brain was firm, rubbery and yellowish in colour, with no blood in meningeal vessels. Affected mice showed no gross alterations apart from the occasional animal in which the cerebellum was partially herniated through the foramen magnum. Microscopically, in control brains, the neuropil was compact. Blood vessels were preserved in an open state and capillaries were in direct apposition to the neuropil. Astrocytic end-feet were not identifiable by light microscopy. In toxin-treated brains, at 1 h the most obvious change was the presence of clear spaces around most vessels, sometimes imparting a scalloped appearance around vesselsof capillary size. While blood vesselscould still be found readily with lumina fixed in the fully open state and in close apposition to the neuropil, many were collapsed and commonly contained erythrocytes which had not
EPSILON
TOXIN
IN TABLE
D~STRI~~“T~ON AND SEVERITY
OF BRAIN
Graded
365
BRAIN
1
LESIONS IN MICE AFTER THE INJECTION
Nwnbers of animals affected
severity
of lesions
Cerebral cor,ex Corpus callosum Corpus striatum Thalamus Paraventricular area of lateral wntricles Vestibular area Corpus medullare cercbeili Granular layer of cerebellum Substantia nigra Anterior commissure
MOUSE
+
++
++t
9 4
4 3
A
4
9 I 1
5 7
OF TYPE D TOXIN
++
3
1)
Numbers of animals affected
Graded severity of lesions +
(EXPT
t++ -
3
-
1 -
3 1
l-4 13 14 6
T -
-
1 7
-
6 14
1
-
I
4
6
1
11
1
-
1
2
2 I 1
4 1 1
1 2
4 -
-
Number of animals in,jected Number of animals with lesions Degree of srverity of lesions. + =Mild vacuolation of the + + =Moderate vacuolation, + + + =Severe vacuolation,
-
60
30
32
7
-
5 2 -
neuropil and macroglial reaction. with necrosis of a few glial cells. with necrosis of most glial elements.
been flushed out during perfusion and probably represented vascular stasis. Nearby vessels which were fully patent very rarely contained erythrocytes. Fine vacuolation of the neuropil was present in many areas and astrocytes were often surrounded by clear spaces and the nuclei of these cells were larger and paler than normal. Some neurones were also surrounded by small clear spaces but the neurones themselves appeared unaltered. The above changes became progressively more obvious and at 4 h the neuropil had a coarse, vacuolated appearance. In some areas, discrete vacuoles gave a distinctly bubbly appearance to the neuropil. Occasionally there were larger spaces in the white matter with considerable separation and some fragmentation of nerve fibres, and many astrocytes in these lesions now possessed a visible amount of eosinophilic cytoplasm. Vacuolation of the Purkinje cell layer of the cerebellum was conspicuous and the granular layer was less compactly arranged. A few small foci of necrosis were present in the granular layer, with pyknotic nuclei of granule cells disposed in a non-staining matrix (Fig. 3). At 6 h, perivascular spaces were distended but contained no plasma exudate and most vessels were markedly hyperaemic. In the white matter, capillary haemorrhage was sometimes observed and astrocytes were reactive and formed gemistocytes or, occasionally, were degenerate with nuclear pyknosis and karyorrhexis. At 12 h, small amounts of faintly eosinophilic plasma exudate were found
366
J.
W.
FINNIE
EPSILON
TOXIN
IN
MOUSE
367
BRAIN
adjacent to a few small blood vessels, and astrocytic nuclei were swollen, oval eosinophilic bodies vesicular and hypochromatic. Occasionally, representing swollen axons (spheroids or retraction bulbs) were visible in areas of white matter damage. At 18 h, fragmentation of glia in white matter lesions was more evident and in some of these areas, especially the corpus callosum, there were lakes of deeply eosinophilic plasma exudate which had a foamy, vacuolated appearance and stained strongly positive with the PAS technique. In the cerebral cortex, discrete foci of spongy change were sometimes found and neurones in these foci were shrunken and deeply acidophilic; neurones in adjacent unaffected areas of cortex, however, appeared normal. In areas of the corpus striatum (where grey and white matter are closely admixed) showing severe oedematization, some neurones had large vesicular nuclei and cytoplasmic vacuolation and a few were shrunken and deeply eosinophilic. At this time, especially in the vestibular tracts, axons not uncommonly showed irregular beaded swellings and were surrounded by an expanded, varicose, non-staining myelin sheath. With fragmentation and retraction of the axon and its sheath, ellipsoids formed and often enclosed an oval, eosinophilic swollen axon (digestion chambers). At a later stage, ellipsoids became less welldefined, leaving an irregular pale-staining focus containing eosinophilic axonal debris, at the periphery of which compound granular corpuscles (gitter cells) had appeared. At 24 h, white matter lesions had in some areas progressed to malacia, with the neuropil presenting a lace-like appearance. Fibres were widely separated by clear spaces containing a few small, round deeply basophilic nuclei, most of which represented pyknotic astrocytic nuclei with a few surviving oligodendrocytes. Extravasated erythrocytes were sometimes numerous in malacic foci and, at the periphery of these lesions, astrocytic nuclei were enlarged and hypochromatic, endothelial cells were swollen, and there was a slight increase in the number of microglial cells. No polymorphs or lymphocytes were seen in any malacic foci in the brain at any stage. A few necrotic areas had undergone complete liquefaction necrosis, with empty spaces devoid of any tissue and this change was most commonly observed in the corpus callosum, sometimes resulting in separation of the white matter from the overlying cortical grey matter. In brains examined at 7 days, the malacic foci were almost completely filled with gitter cells, which had small, round nuclei and a clear, vacuolated or finely granular cytoplasm. These lesions also contained fibrous astrocytes and axonal debris, and capillaries were seen to be invading the organizing focus at the perimeter.
Fig. 1. Bilaterally thalamus, receiving
symmetrical lesions in the corpus callosum, paraventricular fimbria ofthe hippocampus and a focal lesion in the cerebral multiple sublethal doses of toxin (Expt. I). HE x 7.
Fig. 2. Bilaterally sublethal
symmetrical doses of toxin
Fig. 3. Focal necrosis dose of toxin
malacic (Expt.
in the granular (Expt. 2). HE
foci in the vestibular 1). HE x 7. layer x 35.
of the cerebellum
area
of a mouse
of a mouse
cortext
area, corpus of a mouse
24 h after
4 h after
receiving
striatum, 24 h after
receiving a single
multiple lethal
368
J.
W.
FINNIE
Lesions consistently manifested 3 features. The malacic foci were often focal; they were commonly bilaterally symmetrical; and they were essentially confined to the white matter. DISCUSSION
In Experiment 1, the distribution of lesions was consistent with findings in mice by Griner (1961) and Morgan and Kelly (1974), and similar to the pattern of lesions in lambs described by Hartley (1956) and Griner ( 1961). In mice, however, more regions of the brain appeared to be susceptible than in lambs. The proportion of mice which developed lesions increased when multiple, rather than single, doses of toxin were given, which is in agreement with observations by Morgan and Kelly (1974). Morgan and Kelly (1974) considered that the dorso-lateral aspect of the corpus medullare cerebelli and the paraventricular areas lateral to the lateral ventricles were the sites where the initial lesions developed. The corresponding site in the present study may be in the granular layer of the cerebellum. This was the necrotic area in mice given single injections. The bilateral symmetry of lesions in FSE is one of the hallmarks of this condition (Hartley, 1956). Oedema in the brain causes increased intracranial pressure and may secondarily produce local lesions by compression of arteries (Lindenberg, 1955). B ecause the pressure exerted by oedema in this manner is symmetrical, it is to be expected that the lesions in this condition are also symmetrical. The early lesions induced in the brain of mice by epsilon toxin were essentially those of oedema, being more severe in the white matter. It has long been recognized that white matter is much more susceptible to oedema and it is here that oedema can lead to tissue destruction (Manz, 1974). The consistent and often extensive involvement of the more compactly arranged myelinated fibre tracts, such as the corpus callosum and internal capsule, in the present study was, however, unexpected as these tracts are relatively resistant to the destructive effects of oedema (Manz, 1974). Furthermore, Jubb and Kennedy (1970) stated that the permeability of vessels to trypan blue in lambs with acute enterotoxaemia was diffuse throughout the’ brain, but the heavily myelinated tracts were spared. These findings suggest that necrosis in these tracts may be due to direct vascular involvement and not to simple passive accumulation of oedema fluid. The internal capsule, for example, is supplied by end-arteries (Brierley, 1977). Lesions were commonly found in paraventricular areas lateral to the lateral ventricles and these regions are thought to be predisposed to ischaemia since they derive their main blood supply from a system of primitive penetrating arteries which are essentially end-arteries (De Reuck, Chattha and Richardson, 1972). The basal ganglia, corpus striatum and thalamus, regions often involved in intoxicated brains in the present study, are highly vulnerable to hypoxia and in experimental enterotoxaemia in lambs there is a profound systemic hypoxia (Gardner, 1973). This may have produced a secondary reduction in brain
EPSILON
TOXIN
IN
MOUSE
BRAIN
369
perfusion pressure due to impairment of cardio-respiratory function (Brierley, 1977). The brain is a relatively impoverished tissue in terms of capillary density and, in an idealized mid-intercapillary region, doubling the intercapillary distance would reduce to nil the oxygen tension by extending the length of the diffusion path (Bourke, Kimelberg, Nelson, Barron, Auen, Popp and Waldman, 1980). Since astrocytic processes are disposed as satellites around neurones and their end-feet cover most of the surface of capillaries, they are in a potentially strategic position to alter the supply of blood-borne nutrients to with astrocytic swelling, the dependent neurones. In oedematous conditions, diffusion path length between vessels and neurones may be exceeded in focal scattered areas of grey matter and result in cortical lesions of the type found in the present study. Similarly, focal necrosis in the cerebellar granular layer may be due to interference with the supply of nutrients to these cells by early and severe swelling of astrocytes. The bilaterally symmetrical malacic lesions in the vestibular area, commonly found in toxin-treated mice in the present study, also occur in Chastek’s paralysis of carnivores. The sequential pathogenesis of lesions in these latter animals is oedema, vascular dilatation, haemorrhage and necrosis Wubb and Kennedy, 1970), which is similar to the progression of events in the mice, but an explanation for the topographical distribution of lesions in Chastek’s paralysis has not been forthcoming.
SUMMARY
The distribution, severity and frequency of brain lesions produced in mice by the administration of C’lostridium perfringens Type D epsilon toxin were examined by light microscopy. The granular layer of the cerebellum was the area most frequently affected in mice given single doses of toxin. Sequential changes in brain morphology were examined from 1 h to 7 days after injection of toxin. Lesions progressed from an initial vasogenic oedema to malacic foci which commonly were focal and bilaterally symmetrical, with a predilection for white matter. The topographical distribution of these malacic areas is discussed.
REFERENCES
Bourke, R. L., Kimelberg, H. K., Nelson, L. R., Barron, K. D., Auen, E. L., Popp, A. J7 and Waldman, J. B. (1980). Biology of glial swelling in experimental brain oedema. In Advances in ..Meurology, Vol. 28, Raven Press,New York, p. 100. Brierley, J. B. (1977). Experimental hypoxic brain damage. Journal of Clinical Pathology, 30, SuppI. 11, 181-187. DeReuck, J., Chattha, A. S., and Richardson, E. P. (1972). Pathogenesisand evolution of periventricular leukomalacia in infancy. Archives of Neurology, 27, 229-236. Gardener, D. E. (1973). Pathology of Clastridium welchii Type D enterotoxaemia. II. Structural and ultrastructural alterations in the tissuesof lambs and mice. Journal of Comparative Pathology, 83, 509-524.
Griner, L. A. (1961). Enterotoxaemia of sheep.I. Effects of Clostridium perfringens Type
370
J.
W.
FINNIE
D toxin on the brains of sheep and mice. American Journal of Veterinary Research, 22, 429-442. Hartley, W. J. (1956). A focal symmetrical encephalomalacia of lambs. New Zealand Veterinary Journal, 4, 129-l 35. Jubb, K. V. F., and Kennedy, P. C. (1970). Pathology of Domestic Animals, Vol. 1. Academic Press, New York, pp. 115, 385. Lindenberg, R. (1955). Compression of brain arteries as pathogenic factor for tissue necroses and their areas of predilection. Journal of Neuropathology and Exberimental Neurology, 14, 223-243. Manz, H. J. (1974). The pathology of cerebral oedema. Human Patholou 5, 291-313. study of brain lesions Morgan, K. T., and Kelly, B. G. (1974). Ultrastructural produced in mice by the administration of Clostridium welchii Type D toxin. Journal of Comparative Pathology, 84, 181-191. [Received for publication,
May 1 Oth, 19831