- Email: [email protected]
Ultrastructural observations in experimental hydrocephalus in the rabbit
Journal o f the Neurological Sciences, 1979, 43:333-344 © Elsevier/North-HollandBiomedicalPress
333
ULTRASTRUCTURAL OBSERVATIONS IN EXPERIMENTAL HYDROCEPHALUS IN THE RABBIT
G. GOPINATH, R. BHATIAand P. G. GOPINATH Departments of .4natomy, Neurosurgery and Nuclear Medicine, All India Institute of Medical Sciences, New Delhi 110016 (India)
(Received 27 June, 1979) (Accepted 10 July, 1979)
SUMMARY Changes in the ependyma and periventricular brain tissues of the lateral, 3rd and 4th ventricles and the cervical spinal canal were studied electron-microscopically in young rabbits on the 9th day after injecting kaolin into the cisterna magna. The ependyma of the lateral ventricle overlying the white matter was notably stretched causing increased egress of CSF and disorganisation of the normal architecture of the white matter and capillaries. The neurons and glial cells close to the white matter showed edematous changes. The changes in the ependymal lining and the underlying grey matter were less severe in the dorsal part of the 3rd and the 4th ventricle. The ventral part of the 3rd ventricle was the least affected. The height and the arrangement of the ependymal cells, the surrounding grey matter with narrow interstitial spaces and the absorbing tanycytes seemed to be factors which were responsible for the minimal changes in these regions. The changes appeared to be reversible if the CSF pressure was relieved at this stage. The spinal canal remained unaffected in the majority of our hydrocephalic animals, which could probably be attributed to the type of animal and the degree of hydrocephalus.
INTRODUCTION Available literature on experimental hydrocephalus produced in various species of animals indicates that there is an excessive transependymal migration of cerebrospinal fluid (CSF). Defects in the ependymal lining have been demonstrated by several investigators (Hochwald et al. 1969; Clark and Milhorat 1970; Milhorat et al. 1970; Lux et al. 1970; Ogata et al. 1972; Rubin et al. 1976) while others (Page 1975; Go et al. 1976; Torvik et al. 1976; Lindberg et al. 1977) have not observed defects or breaks in the ependymal lining even with severe hydrocephalus. Widening of the intercellular
334 junctions of the ependyma was observed by Milhorat et al. (1970) as early as 3 h after the experimental production of hydrocephalus. Based on the observations of Brightman and Reese (1969) that the gap junctions of normal ependyma allow passage of fluid and proteins from the ventricle, Sahar et al. (1969b) suggested that there might be an increased rate of transport of fluid across the normal gap junctions in conditions causing raised intracranial pressure. Regional variations in the morphological characteristics of the ependyma have been described by kuppa and Feustel (1971), Scott et al. (1974), Page (1975) and several others. Different functions have been ascribed to the ependymal cells based on the morphological characteristics (Anand Kumar and Knowles 1967; Knigge and Scott 1970; Scott and Knigge 1970; Kobayashi et al. 1972; Roderiguez 1972). Most investigators have concentrated their attention on the lateral ventricle following the experimental production of hydrocephalus, assuming that the changes seen here represent the entire ventricular system. Page (1975) and Go et al, (1976) from their scanning electron-microscopic studies of the ependymal lining of the ventricles in the hydrocephalic rabbit concluded that it was the lateral ventricles alone that were affected to any degree. The changes in the sub-ependymal regions adjacent to the dilated ventricles have been described by several investigators. Edema due to transependymal migration of the CSF was more pronounced in the white matter compared to the grey matter (Ogata et al. 1972). Marked disorganisation of the periventricular white matter characterised by extracellular edema, subependymal tissue destruction with phagocytosis, reactive astrocytosis as also primary axonal damage has been reported by Hochwald et al. (1969), Clark and Milhorat (1970), Milhorat et al. (1970), Weller et al. (1971), Ogata et al. (1972) and Rubin et al. (1976). These changes were almost entirely confined to the periventricular white matter, the grey matter remaining virtually unscathed (Ogata et al. 1972). In contrast Torvik et al. (1976) and Torvik and Stenwig (1977) in both light and electron-microscopic studies observed only subependymal sponginess with no tissue destruction or phagocytosis even in animals with severe hydrocephalus. The capacity of the spinal subarachnoid space to absorb CSF has been demonstrated by Coben and Smith (1969) and Hammerstad (1969). Bradbury and Lathem (1965) have also shown that in normal rabbits there is a small caudally directed flow of CSF through the central canal and which comes out into the sacral subarachnoid space. Becker et al. (1972) perfused the dilated central canal in kaolininduced hydrocephalic cats and the dye introduced at the obex escaped from the filum terminale into the subarachnoid space. Studies with tracer substances (Eisenberg et al. 1974a,b) again confirmed these observations. Several workers have observed dilatation of the central canal with the formation of clefts in the posterior columns in animals with extreme hydrocephalus (Becker et al. 1972; Dohrmann 1972; Eisenberg et al. 1974a; Torvik and Murthy 1977). In moderate hydrocephalus there was little or no dilatation of the central canal (Weller et al. 1971 ; Torvik and Murthy 1977) with no significant change in the lining of the canal (James et al. 1978).
335 MATERIALS AND METHODS Of the 28 3-6-week-old rabbits injected with 0.2-0.5 ml of a thick suspension of kaolin into the cisterna magna, 18 developed hydrocephalus. Thirteen of the animals were perfused with neutral formalin at varying intervals, from the 5th to the 38th day after injection, and the brain and the cervical spinal cord were processed for lightmicroscopic studies. Five animals were perfused with Karnovsky's fixative on the 9th postoperative day. After slow intracardiac perfusion the brain and the spinal cord were removed and the extent of the hydrocephalus was assessed. Blocks of tissues from the dorsolateral wall of the lateral ventricle, dorsal and ventral regions of the 3rd ventricle, floor of the caudal 4th ventricle and the cervical spinal canal were further fixed in the same fixative, washed thoroughly with cacodylate buffer, post-fixed in 1 buffered osmium tetroxide, dehydrated in acetone series and embedded in araldite. Sections were examined and photographed using a Philips 300 electron microscope. Littermates were used as control. RESULTS The changes around the brainstem following kaolin injection were similar to those reported by Torvik et al. (1976) and Torvik and Murthy (1977). Kaolin within the cisterna magna caused fibrosis and inflammation but no granulomatous arachnoiditis was observed. While there was progressive ventricular dilatation in all the animals there was considerable variation in the degree of hydrocephalus. The majority of the animals showed an initial dilatation of the 3rd ventricle which progressed rapidly till the end of the 3rd week. The lateral ventricle, affected later, continued to enlarge up to the end of the period studied. The light-microscopic changes were similar to the observations reported by Torvik et al. (1976). The ependyma of the lateral ventricle and the adjacent white matter showed the most changes. Ependymal stretching and the subependymal sponginess were evident. In contrast, the 3rd and the 4th ventricles were less affected. The cervical spinal canal of the control animals showed a variable appearance. The younger animals showed a round to oval patent spinal canal while the older animals had slit-like spinal canals which appeared completely obliterated. Some of the hydrocephalic animals showed a slightly dilated spinal canal which was comparable to the normal younger animals but there was no subependymal sponginess. In two animals of 28 and 36 post-injection days the canal dilatation was more marked and kaolin and inflammatory cells were present in the 4th ventricle and the dilated spinal canal. Slight subependymal sponginess and vacuolation of the posterior funiculus were seen in these animals. In the animals killed for electron microscopy there was no obvious spinal canal dilatation. Electron-microscopic observations of the normal ependyma of the dorsal part of the lateral ventricle revealed cuboidal cells with oblique intercellular junctions, where the cell membranes at times were interdigitating with each other. Junctional speciali-
Fig. 1. Cuboidal cells lining the dorsolateral wall of the lateral ventricle of a normal animal. Tight junctions are seen in the ependyma towards the ventricle. ,~ 12,600. Fig. 2. Pseudostratified ependyma with apical microvilli lining the ventral region of the 3rd ventricle in a normal animal :~ 11,250.
Fig. 3. Tall ependymal cells lining the dorsal part of the 3rd ventricle, 4th ventricle and the cervical spinal canal in a normal animal, x 10,350. Fig. 4. Dorsolateral wall of the lateral ventricle in a hydrocephalic rabbit. Only a thin rim ofependymal cytoplasm separates the ventricle from the underlying white matter. A gap junction is apparent and widened fluid spaces (S) are seen in the subependymal region, x 40,000.
338
Fig. 5. Ventral wall of the 3rd ventricle from an experimental animal. Ependymais stretched,thepseudostratification absent. >; 14,960. Fig. 6. Ependymal lining from a mildly dilated cervical canal, x 9,000.
339
~ii~ ~
...... . . . . . . i
¸i,¢ ¸
!!~i! i ~i i:~ !~!i: '~ ii i/!~i i~j,~
.i ,ii ~
/
Fig. 7. Oligodendroglial processes separated by the surrounding fluid spaces are seen extending to the myelinated fibres which appear normal, x 14,960. Fig. 8. Capillary of the white matter showing rarefied endothelial cytoplasm. Arrow indicates space between endothelium and the glial process, x 13,600
340
Fig. 9. A neuron closely related to the white matter shows disruption of the cell membrane and cytoplasm. Arrow indicates a perineuronal glial cell with vacuolated cytoplasm. ~ 6,600. Fig. 10. A neuron from the subependymal region of the 3rd ventricle (ventral part). Cytoplasmic disruption is seen, but the perineuronal glial cell is inact. "< 6,600.
,
r
341 sations like zonula occludens and adherentes were seen towards the lumen of the ventricle. Occasionally, intercellular clefts were seen below the specialised junctions. The luminal surface of the cells showed both microvilli and cilia (Fig. 1). The dorsal part of the 3rd ventricle, the floor of the 4th ventricle and the spinal canal were lined with tall cylindrical cells which had both cilia and microvilli (Fig. 3). The ventral part of the 3rd ventricle was lined by pseudostratified epithelium with few cilia scattered among the microvilli (Fig. 2). The plasma membranes of the adjacent cells made contacts in a vertical plane with the specialised junctions towards the apical surface in the 3rd and the 4th ventricles and the spinal canal (Figs. 2 and 3). In the hydrocephalic animals the ependyma of the area of the lateral ventricle studied was stretched, leaving only a thin rim of cytoplasm lining the ventricle. Microvilli and cilia were very sparse. Due to the ependymal stretching the intercellular junctions appeared parallel to the ventricular surface. Uninterrupted communicating channels between the ventricle and the subependymal fluid spaces were observed. Dilatation of the abluminal part of the intercellular regions was frequently seen (Fig.
4). The white matter underlying the lateral ventricle was edematous with clear fluid spaces. Myelinated axons could be seen dispersed among fluid spaces and astroglial processes were extending to these fibres (Fig. 7). The endothelial cells of the capillaries in the white matter appeared edematous and occasionally spaces were seen between the endothelium and the glial feet (Fig. 8). The neurons adjacent to the white matter showed discontinuity of the cell membrane with disruption of the cytoplasm. The nucleus and the organelles in the scanty cytoplasm retained the normal appearance. The perineuronal glial cells of the affected neurons had vacuolated cytoplasm (Fig. 9). The ependymal stretching and the subependymal fluid collection of the dorsal part of the 3rd ventricle and the 4th ventricle was not so severe as in the lateral ventricle. The neurons close to the subependymal regions only showed changes resulting from fluid entry into the cytoplasm. Glial cells and capillaries appeared normal. The pseudostratification of the ependymal lining of the ventral part of the 3rd ventricles was no longer apparent in the animals with hydrocephalus (Fig. 5). Other changes were minimal. Only the neurons close to the ependyma were affected (Fig. 10). No significant change was observed in the ependymal lining or the subependymal regions of the cervical spinal canal of the 9th-day hydrocephalic animals (Fig. 6). DISCUSSION The present study confirmed earlier observations that in kaolin-induced hydrocephalic rabbits the most affected region was the lateral ventricle (Torvik et al. 1976; Torvik and Stenwig 1977). The increased intraventricular pressure resulted in the extreme attenuation of the ependymal lining overlying the white matter in the lateral ventricle. The ependyma over the grey matter of the lateral ventricle was not as
342 severely affected. Examination of the ependyma of the 3rd ventricle, 4th ventricle and the cervical spinal canal showed only moderate changes suggesting that the grey matter adjacent to these regions had a protective (buttressing) effect. The relatively wider interstitial spaces of the white matter, measuring 500-1000A compared to 200-500A of the grey matter (Dobbing 1963 ; Lee 1971), permitted fluid accumulation and thereby stretching of the overlying ependyma producing the regional variations. The ependymal stretching resulted in the communicating channels permitting free passage of CSF into the subependymal region. Widened intercellular junctions in the ependyma have been observed by Milhorat et al. (1970) overlying the white matter in the lateral ventricle of the monkeys with acute hydrocephalus. The increased intraventricular pressure also enhanced the rate of transport of CSF across attenuated ependymal lining (Sahar et al. 1969a,b). The fluid egress into the white matter caused the dispersal of the myelinated fibres and edematous changes in the neurons adjacent to the white matter and the endothelial cells of the capillaries. The endothelial changes with lack of pinocytotic vesicles indicated passive entry of the fluid into the vascular compartment. There was no indication of axonal or cellular degeneration at this period. In fact, even in animals with severe hydrocephalus no evidence of tissue damage was observed by Torvik and Stenwig (1977). The initial dilatation of the 3rd ventricle could be because of its ventral and dependent location. The changes observed in the ependyma and subependymal regions were, however, minimal as also observed by Page (1975). In addition to the underlying grey matter (vide supra) the other probable factor is the presence of a thick ependymal lining with the presence of absorbing tanycytes in the region of the median eminence (Anand Kumar and Knowles 1967; Wagner and Pilgrim 1974). The lack of changes in the cervical spinal canal regions could be a reflection of the amount and localisation of the kaolin (Torvik and Murthy 1977), the type of animal used (James et al. 1977) and the degree of hydrocephalus. In the absence of gross destruction of the ependyma and periventricular tissues, it is reasonable to suggest that at least in hydrocephalus of short duration the changes are to a large extent reversible if treated early. The brunt of raised intraventricular pressure is borne by the white matter with relative sparing of the deeper structures like the basal ganglia and thalamus. This probably accounts for the preservation of intellectual capacity despite grossly dilated ventricles as commonly seen in clinical practice. This is comparable to the reversibility of changes produced by protein-calorie malnutrition in the early neo-natal period, when treated before the critical period of brain development (Dobbing 1968). Early treatment of hydrocephalus is essential to prevent the loss of neural tissue that occurs in long-standing human hydrocephalus (Russell 1949). ACKNOWLEDGEMENTS The authors are grateful to Dr. T. C. Anand Kumar for the facilities of the electronmicroscopic laboratory and to Mr. S. C. Sharma and Mr. Rampal for technical assistance.
343 REFERENCES Anand Kumar, T. C. and F. Knowles (1967)A system linking the third ventricle with the pars tuberalis of the rhesus monkey, Nature (Lond.), 215: 54-55. Becker, D. P., J. A. Wilson and W. G. Watson (1972) The spinal cord central canal - - Response to experimental hydrocephalus and canal occlusion, J. Neurosurg., 36: 416-424. Bradbury, N. W. B. and W. Lathem (1965) A flow of cerbrospinal fluid along the central canal of the spinal cord of the rabbit and communications of this canal and the sacral subarachnoid space, J. Physiol. (Lond.), 181 : 785-800. Brightman, M. W. and T. S. Reese 0969) Junctions between intimately opposed cell membranes in the vertebrate brain, J. Cell Biol., 40: 648-672. Clark, R. G. and T. H. Milhorat (1970) Experimental hydrocephalus - - Light microscopic findings in acute and subacute obstructive hydrocephalus in the monkey, J. Neurosurg., 32: 400-413. Coben, L. A. and Smith, K. R. (1969) Iodide transfer at four cerebrospinal fluid sites in the dog - Evidence for spinal iodide carrier transport, Exp. Neurol., 23 : 76-90. Dobbing, J. (1963) The blood-brain barrier and some recent developments, Guy's Hosp. Rep., 112: 267-286. Dobbing, J. (1968) Vulnerable period in developing brain. In A. N. Davison and J. Dobbing (Eds.), Applied Neuro-chemistry, Blackwell, Oxford, pp. 287-316. Eisenberg, H. M., J. E. McLennan and K. Welch (1974a)Ventricular perfusion in cats with kaolininduced hydrocephalus, J. Neurosurg., 41: 20-28. Eisenberg, H. M., J. E. McLennan and K. Welch (1974b) Radioisotope ventriculography in cats with kaolin-induced hydrocephalus, Radiology, 110: 399-402. Go, K. G., I. Stokroos, E. H. Blaauw, F. Zviderveen and I. Molenaar (1976) Changes of the ventricular ependyma and choroid plexus in experimental hydrocephalus, as observed by scanning electronmicroscopy, Acta neuropath. (Bed.), 34: 55-64. Hammerstad, J. P., A. V. Lorenz and R. W. P. Cutler (1969) Iodide transport from the spinal arachnoid fluid in the cat, Amer. J. PhysioL, 216: 353-358. Hochwald, G. M., R. Sahar, A. R. Sadik and J. Ransohoff (1969) Cerebrospinal fluid production and histological observations in animals with experimental obstructive hydrocephalus, Exp. Neurol., 25: 190-199. James, A. E., W. J. Flor, G. R. Novak, E. Strecker and B. Burns (1978) Evaluation of the central canal of the spinal cord in experimentally induced hydrocephalus, J. Neurosurg., 48: 970-974. Knigge, K. M. and D. E. Scott (1970) Structure and function of median eminence, Amer. J. Anat., 129 : 223-244. Kobayashi, H., M. Wada, H. Uemura and M. Ueck (1972) Uptake of peroxidase from the third ventricle by ependymal cells of the median eminence, Z. Zellforsch., 127:545-551. Lee, J. C. (1971) Evolution in the concept of the blood-brain barrier phenomenon. In H. M. Zimmerman (Ed.), Progress in Neuropathology, Grune and Stratton, New York, pp. 84-145. Lindberg, L. A., L. Vasenius and S. Talanti (1977) The surface fine structure of the ependymal lining of the lateral ventricle in rats with hereditary hydrocephalus, Cell Tiss. Res., 179:121-129. Luppa, H. and G. Feustel (1971) Localisation and characterisation of hydrolytic enzymes on the 3rd ventricle lining in the region of the recessus infundibularis of the rat - - A study of the function of the ependyma, Brain Res., 29: 253-270. Lux, Jr., W. E., G. E. Hochwald and J. Ransohoff (1970) Effect of pressure in experimental chronic hydrocephalus, Arch. Neurol. (Chic.), 23: 475-479. Milhorat, T. H., R. G. Clark, M. K. Hammock and P. P. McGrath (1970) Structural, ultrastructural and permeability changes in the ependyma and surrounding brain favouring equilibration in progressive hydrocephalus, Arch. Neurol. (Chic.), 22: 397-407. Ogata, J., G. M. Hochwald, H. Gravioto and J. Ransohoff, (1972) Light and electronmicroscopic studies of experimental hydrocephalus - - Ependymal and subependymal areas, Acta neuropath. (Berl.), 21: 213-223. Page, R. B. (1975) Scanning electron microscopy of the ventricular system in normal and hydrocephalic rabbits - - Preliminary report and atlas, J. Neurosurg., 42: 646-664. Rodriguez, E. M. (1972) Comparative and functional morphology of the median eminence. In: K. M. Knigge, D. E. Scott and A. Weindl (Eds.), Brain Endocrine Interaction - - Structure and Function, Karger, Basel, pp. 319-334. Rubin, R. C., G. M. Hochwald, M. Tiell, H. Mizutani and N. Ghatak (1976) Hydrocephalus, Part 1
344 (Histological and ultrastructural changes in the pre-shunted cortical mantle), Surg. Neurol., 5: 109-114. Russell, D. S. (1949) Observations on the Pathology of Hydrocephalus (MRC Special Report Series, No. 265), HM Stationary Office, London, pp. 119. Sahar, A., G. M. Hochwald, A. K. Sadik and J. Ransohoff (1969a) Cerebrospinal fluid absorption in animals with experimental obstructive hydrocephalus, Arch. Neurol. (Chic.), 21: 638-644. Sahar, A., G. M. Hochwald and J. Ransohoff (1969b) Alternate pathway for cerebrospinal fluid absorption in animals with experimental obstructive hydrocephalus, Exp. Neurol., 25: 200-206. Scott, D. E. and K. M. Knigge (1970) Ultrastructural changes in the median eminence of the rat following differentiation of the basal hypothalamus, Z. Zellforsch., 105 : 1-32. Scott, D. E., G. P. Kozlowski and M. N. Sheridan (1974) Scanning electron microscopy in the ultrastructural analysis of the mammalian cerebral ventricular system, Int. Rev. CytoL, 37: 349-388. Torvik, A. and V. A. Murthy (1977) The spinal cord central canal in kaolin-induced hydrocephalus, J. Neurosurg., 47: 397-402. Torvik, A. and A. E. Stenwig (1977) Pathology of the experimental obstructive hydrocephalus - Electron microscopic observations, Aeta neuropath. (Berl.), 38: 21-26. Torvik, A., R. Bhatia and R. Nyberg-Hansen (1976) The pathology of experimental obstructive hydrocephalus, Neuropath. appl. Neurobiol., 2: 41-52. Wagner, H. J. and Ch. Pilgrim (1974) Extracellular and transcellular transport of horseradish peroxidase (HRP) through the hypothalamic tanycyte ependyma, Cell Tiss. Res., 152: 477-491. Weller, R. O., H. Wisniewski, K. Shulman and R. D. Terry (1971) Experimental hydrocephalus in young dogs - - Histological and ultrastructural study of the brain tissue damage, J. Neuropath. exp. Neurol., 30: 613-626.
333
ULTRASTRUCTURAL OBSERVATIONS IN EXPERIMENTAL HYDROCEPHALUS IN THE RABBIT
G. GOPINATH, R. BHATIAand P. G. GOPINATH Departments of .4natomy, Neurosurgery and Nuclear Medicine, All India Institute of Medical Sciences, New Delhi 110016 (India)
(Received 27 June, 1979) (Accepted 10 July, 1979)
SUMMARY Changes in the ependyma and periventricular brain tissues of the lateral, 3rd and 4th ventricles and the cervical spinal canal were studied electron-microscopically in young rabbits on the 9th day after injecting kaolin into the cisterna magna. The ependyma of the lateral ventricle overlying the white matter was notably stretched causing increased egress of CSF and disorganisation of the normal architecture of the white matter and capillaries. The neurons and glial cells close to the white matter showed edematous changes. The changes in the ependymal lining and the underlying grey matter were less severe in the dorsal part of the 3rd and the 4th ventricle. The ventral part of the 3rd ventricle was the least affected. The height and the arrangement of the ependymal cells, the surrounding grey matter with narrow interstitial spaces and the absorbing tanycytes seemed to be factors which were responsible for the minimal changes in these regions. The changes appeared to be reversible if the CSF pressure was relieved at this stage. The spinal canal remained unaffected in the majority of our hydrocephalic animals, which could probably be attributed to the type of animal and the degree of hydrocephalus.
INTRODUCTION Available literature on experimental hydrocephalus produced in various species of animals indicates that there is an excessive transependymal migration of cerebrospinal fluid (CSF). Defects in the ependymal lining have been demonstrated by several investigators (Hochwald et al. 1969; Clark and Milhorat 1970; Milhorat et al. 1970; Lux et al. 1970; Ogata et al. 1972; Rubin et al. 1976) while others (Page 1975; Go et al. 1976; Torvik et al. 1976; Lindberg et al. 1977) have not observed defects or breaks in the ependymal lining even with severe hydrocephalus. Widening of the intercellular
334 junctions of the ependyma was observed by Milhorat et al. (1970) as early as 3 h after the experimental production of hydrocephalus. Based on the observations of Brightman and Reese (1969) that the gap junctions of normal ependyma allow passage of fluid and proteins from the ventricle, Sahar et al. (1969b) suggested that there might be an increased rate of transport of fluid across the normal gap junctions in conditions causing raised intracranial pressure. Regional variations in the morphological characteristics of the ependyma have been described by kuppa and Feustel (1971), Scott et al. (1974), Page (1975) and several others. Different functions have been ascribed to the ependymal cells based on the morphological characteristics (Anand Kumar and Knowles 1967; Knigge and Scott 1970; Scott and Knigge 1970; Kobayashi et al. 1972; Roderiguez 1972). Most investigators have concentrated their attention on the lateral ventricle following the experimental production of hydrocephalus, assuming that the changes seen here represent the entire ventricular system. Page (1975) and Go et al, (1976) from their scanning electron-microscopic studies of the ependymal lining of the ventricles in the hydrocephalic rabbit concluded that it was the lateral ventricles alone that were affected to any degree. The changes in the sub-ependymal regions adjacent to the dilated ventricles have been described by several investigators. Edema due to transependymal migration of the CSF was more pronounced in the white matter compared to the grey matter (Ogata et al. 1972). Marked disorganisation of the periventricular white matter characterised by extracellular edema, subependymal tissue destruction with phagocytosis, reactive astrocytosis as also primary axonal damage has been reported by Hochwald et al. (1969), Clark and Milhorat (1970), Milhorat et al. (1970), Weller et al. (1971), Ogata et al. (1972) and Rubin et al. (1976). These changes were almost entirely confined to the periventricular white matter, the grey matter remaining virtually unscathed (Ogata et al. 1972). In contrast Torvik et al. (1976) and Torvik and Stenwig (1977) in both light and electron-microscopic studies observed only subependymal sponginess with no tissue destruction or phagocytosis even in animals with severe hydrocephalus. The capacity of the spinal subarachnoid space to absorb CSF has been demonstrated by Coben and Smith (1969) and Hammerstad (1969). Bradbury and Lathem (1965) have also shown that in normal rabbits there is a small caudally directed flow of CSF through the central canal and which comes out into the sacral subarachnoid space. Becker et al. (1972) perfused the dilated central canal in kaolininduced hydrocephalic cats and the dye introduced at the obex escaped from the filum terminale into the subarachnoid space. Studies with tracer substances (Eisenberg et al. 1974a,b) again confirmed these observations. Several workers have observed dilatation of the central canal with the formation of clefts in the posterior columns in animals with extreme hydrocephalus (Becker et al. 1972; Dohrmann 1972; Eisenberg et al. 1974a; Torvik and Murthy 1977). In moderate hydrocephalus there was little or no dilatation of the central canal (Weller et al. 1971 ; Torvik and Murthy 1977) with no significant change in the lining of the canal (James et al. 1978).
335 MATERIALS AND METHODS Of the 28 3-6-week-old rabbits injected with 0.2-0.5 ml of a thick suspension of kaolin into the cisterna magna, 18 developed hydrocephalus. Thirteen of the animals were perfused with neutral formalin at varying intervals, from the 5th to the 38th day after injection, and the brain and the cervical spinal cord were processed for lightmicroscopic studies. Five animals were perfused with Karnovsky's fixative on the 9th postoperative day. After slow intracardiac perfusion the brain and the spinal cord were removed and the extent of the hydrocephalus was assessed. Blocks of tissues from the dorsolateral wall of the lateral ventricle, dorsal and ventral regions of the 3rd ventricle, floor of the caudal 4th ventricle and the cervical spinal canal were further fixed in the same fixative, washed thoroughly with cacodylate buffer, post-fixed in 1 buffered osmium tetroxide, dehydrated in acetone series and embedded in araldite. Sections were examined and photographed using a Philips 300 electron microscope. Littermates were used as control. RESULTS The changes around the brainstem following kaolin injection were similar to those reported by Torvik et al. (1976) and Torvik and Murthy (1977). Kaolin within the cisterna magna caused fibrosis and inflammation but no granulomatous arachnoiditis was observed. While there was progressive ventricular dilatation in all the animals there was considerable variation in the degree of hydrocephalus. The majority of the animals showed an initial dilatation of the 3rd ventricle which progressed rapidly till the end of the 3rd week. The lateral ventricle, affected later, continued to enlarge up to the end of the period studied. The light-microscopic changes were similar to the observations reported by Torvik et al. (1976). The ependyma of the lateral ventricle and the adjacent white matter showed the most changes. Ependymal stretching and the subependymal sponginess were evident. In contrast, the 3rd and the 4th ventricles were less affected. The cervical spinal canal of the control animals showed a variable appearance. The younger animals showed a round to oval patent spinal canal while the older animals had slit-like spinal canals which appeared completely obliterated. Some of the hydrocephalic animals showed a slightly dilated spinal canal which was comparable to the normal younger animals but there was no subependymal sponginess. In two animals of 28 and 36 post-injection days the canal dilatation was more marked and kaolin and inflammatory cells were present in the 4th ventricle and the dilated spinal canal. Slight subependymal sponginess and vacuolation of the posterior funiculus were seen in these animals. In the animals killed for electron microscopy there was no obvious spinal canal dilatation. Electron-microscopic observations of the normal ependyma of the dorsal part of the lateral ventricle revealed cuboidal cells with oblique intercellular junctions, where the cell membranes at times were interdigitating with each other. Junctional speciali-
Fig. 1. Cuboidal cells lining the dorsolateral wall of the lateral ventricle of a normal animal. Tight junctions are seen in the ependyma towards the ventricle. ,~ 12,600. Fig. 2. Pseudostratified ependyma with apical microvilli lining the ventral region of the 3rd ventricle in a normal animal :~ 11,250.
Fig. 3. Tall ependymal cells lining the dorsal part of the 3rd ventricle, 4th ventricle and the cervical spinal canal in a normal animal, x 10,350. Fig. 4. Dorsolateral wall of the lateral ventricle in a hydrocephalic rabbit. Only a thin rim ofependymal cytoplasm separates the ventricle from the underlying white matter. A gap junction is apparent and widened fluid spaces (S) are seen in the subependymal region, x 40,000.
338
Fig. 5. Ventral wall of the 3rd ventricle from an experimental animal. Ependymais stretched,thepseudostratification absent. >; 14,960. Fig. 6. Ependymal lining from a mildly dilated cervical canal, x 9,000.
339
~ii~ ~
...... . . . . . . i
¸i,¢ ¸
!!~i! i ~i i:~ !~!i: '~ ii i/!~i i~j,~
.i ,ii ~
/
Fig. 7. Oligodendroglial processes separated by the surrounding fluid spaces are seen extending to the myelinated fibres which appear normal, x 14,960. Fig. 8. Capillary of the white matter showing rarefied endothelial cytoplasm. Arrow indicates space between endothelium and the glial process, x 13,600
340
Fig. 9. A neuron closely related to the white matter shows disruption of the cell membrane and cytoplasm. Arrow indicates a perineuronal glial cell with vacuolated cytoplasm. ~ 6,600. Fig. 10. A neuron from the subependymal region of the 3rd ventricle (ventral part). Cytoplasmic disruption is seen, but the perineuronal glial cell is inact. "< 6,600.
,
r
341 sations like zonula occludens and adherentes were seen towards the lumen of the ventricle. Occasionally, intercellular clefts were seen below the specialised junctions. The luminal surface of the cells showed both microvilli and cilia (Fig. 1). The dorsal part of the 3rd ventricle, the floor of the 4th ventricle and the spinal canal were lined with tall cylindrical cells which had both cilia and microvilli (Fig. 3). The ventral part of the 3rd ventricle was lined by pseudostratified epithelium with few cilia scattered among the microvilli (Fig. 2). The plasma membranes of the adjacent cells made contacts in a vertical plane with the specialised junctions towards the apical surface in the 3rd and the 4th ventricles and the spinal canal (Figs. 2 and 3). In the hydrocephalic animals the ependyma of the area of the lateral ventricle studied was stretched, leaving only a thin rim of cytoplasm lining the ventricle. Microvilli and cilia were very sparse. Due to the ependymal stretching the intercellular junctions appeared parallel to the ventricular surface. Uninterrupted communicating channels between the ventricle and the subependymal fluid spaces were observed. Dilatation of the abluminal part of the intercellular regions was frequently seen (Fig.
4). The white matter underlying the lateral ventricle was edematous with clear fluid spaces. Myelinated axons could be seen dispersed among fluid spaces and astroglial processes were extending to these fibres (Fig. 7). The endothelial cells of the capillaries in the white matter appeared edematous and occasionally spaces were seen between the endothelium and the glial feet (Fig. 8). The neurons adjacent to the white matter showed discontinuity of the cell membrane with disruption of the cytoplasm. The nucleus and the organelles in the scanty cytoplasm retained the normal appearance. The perineuronal glial cells of the affected neurons had vacuolated cytoplasm (Fig. 9). The ependymal stretching and the subependymal fluid collection of the dorsal part of the 3rd ventricle and the 4th ventricle was not so severe as in the lateral ventricle. The neurons close to the subependymal regions only showed changes resulting from fluid entry into the cytoplasm. Glial cells and capillaries appeared normal. The pseudostratification of the ependymal lining of the ventral part of the 3rd ventricles was no longer apparent in the animals with hydrocephalus (Fig. 5). Other changes were minimal. Only the neurons close to the ependyma were affected (Fig. 10). No significant change was observed in the ependymal lining or the subependymal regions of the cervical spinal canal of the 9th-day hydrocephalic animals (Fig. 6). DISCUSSION The present study confirmed earlier observations that in kaolin-induced hydrocephalic rabbits the most affected region was the lateral ventricle (Torvik et al. 1976; Torvik and Stenwig 1977). The increased intraventricular pressure resulted in the extreme attenuation of the ependymal lining overlying the white matter in the lateral ventricle. The ependyma over the grey matter of the lateral ventricle was not as
342 severely affected. Examination of the ependyma of the 3rd ventricle, 4th ventricle and the cervical spinal canal showed only moderate changes suggesting that the grey matter adjacent to these regions had a protective (buttressing) effect. The relatively wider interstitial spaces of the white matter, measuring 500-1000A compared to 200-500A of the grey matter (Dobbing 1963 ; Lee 1971), permitted fluid accumulation and thereby stretching of the overlying ependyma producing the regional variations. The ependymal stretching resulted in the communicating channels permitting free passage of CSF into the subependymal region. Widened intercellular junctions in the ependyma have been observed by Milhorat et al. (1970) overlying the white matter in the lateral ventricle of the monkeys with acute hydrocephalus. The increased intraventricular pressure also enhanced the rate of transport of CSF across attenuated ependymal lining (Sahar et al. 1969a,b). The fluid egress into the white matter caused the dispersal of the myelinated fibres and edematous changes in the neurons adjacent to the white matter and the endothelial cells of the capillaries. The endothelial changes with lack of pinocytotic vesicles indicated passive entry of the fluid into the vascular compartment. There was no indication of axonal or cellular degeneration at this period. In fact, even in animals with severe hydrocephalus no evidence of tissue damage was observed by Torvik and Stenwig (1977). The initial dilatation of the 3rd ventricle could be because of its ventral and dependent location. The changes observed in the ependyma and subependymal regions were, however, minimal as also observed by Page (1975). In addition to the underlying grey matter (vide supra) the other probable factor is the presence of a thick ependymal lining with the presence of absorbing tanycytes in the region of the median eminence (Anand Kumar and Knowles 1967; Wagner and Pilgrim 1974). The lack of changes in the cervical spinal canal regions could be a reflection of the amount and localisation of the kaolin (Torvik and Murthy 1977), the type of animal used (James et al. 1977) and the degree of hydrocephalus. In the absence of gross destruction of the ependyma and periventricular tissues, it is reasonable to suggest that at least in hydrocephalus of short duration the changes are to a large extent reversible if treated early. The brunt of raised intraventricular pressure is borne by the white matter with relative sparing of the deeper structures like the basal ganglia and thalamus. This probably accounts for the preservation of intellectual capacity despite grossly dilated ventricles as commonly seen in clinical practice. This is comparable to the reversibility of changes produced by protein-calorie malnutrition in the early neo-natal period, when treated before the critical period of brain development (Dobbing 1968). Early treatment of hydrocephalus is essential to prevent the loss of neural tissue that occurs in long-standing human hydrocephalus (Russell 1949). ACKNOWLEDGEMENTS The authors are grateful to Dr. T. C. Anand Kumar for the facilities of the electronmicroscopic laboratory and to Mr. S. C. Sharma and Mr. Rampal for technical assistance.
343 REFERENCES Anand Kumar, T. C. and F. Knowles (1967)A system linking the third ventricle with the pars tuberalis of the rhesus monkey, Nature (Lond.), 215: 54-55. Becker, D. P., J. A. Wilson and W. G. Watson (1972) The spinal cord central canal - - Response to experimental hydrocephalus and canal occlusion, J. Neurosurg., 36: 416-424. Bradbury, N. W. B. and W. Lathem (1965) A flow of cerbrospinal fluid along the central canal of the spinal cord of the rabbit and communications of this canal and the sacral subarachnoid space, J. Physiol. (Lond.), 181 : 785-800. Brightman, M. W. and T. S. Reese 0969) Junctions between intimately opposed cell membranes in the vertebrate brain, J. Cell Biol., 40: 648-672. Clark, R. G. and T. H. Milhorat (1970) Experimental hydrocephalus - - Light microscopic findings in acute and subacute obstructive hydrocephalus in the monkey, J. Neurosurg., 32: 400-413. Coben, L. A. and Smith, K. R. (1969) Iodide transfer at four cerebrospinal fluid sites in the dog - Evidence for spinal iodide carrier transport, Exp. Neurol., 23 : 76-90. Dobbing, J. (1963) The blood-brain barrier and some recent developments, Guy's Hosp. Rep., 112: 267-286. Dobbing, J. (1968) Vulnerable period in developing brain. In A. N. Davison and J. Dobbing (Eds.), Applied Neuro-chemistry, Blackwell, Oxford, pp. 287-316. Eisenberg, H. M., J. E. McLennan and K. Welch (1974a)Ventricular perfusion in cats with kaolininduced hydrocephalus, J. Neurosurg., 41: 20-28. Eisenberg, H. M., J. E. McLennan and K. Welch (1974b) Radioisotope ventriculography in cats with kaolin-induced hydrocephalus, Radiology, 110: 399-402. Go, K. G., I. Stokroos, E. H. Blaauw, F. Zviderveen and I. Molenaar (1976) Changes of the ventricular ependyma and choroid plexus in experimental hydrocephalus, as observed by scanning electronmicroscopy, Acta neuropath. (Bed.), 34: 55-64. Hammerstad, J. P., A. V. Lorenz and R. W. P. Cutler (1969) Iodide transport from the spinal arachnoid fluid in the cat, Amer. J. PhysioL, 216: 353-358. Hochwald, G. M., R. Sahar, A. R. Sadik and J. Ransohoff (1969) Cerebrospinal fluid production and histological observations in animals with experimental obstructive hydrocephalus, Exp. Neurol., 25: 190-199. James, A. E., W. J. Flor, G. R. Novak, E. Strecker and B. Burns (1978) Evaluation of the central canal of the spinal cord in experimentally induced hydrocephalus, J. Neurosurg., 48: 970-974. Knigge, K. M. and D. E. Scott (1970) Structure and function of median eminence, Amer. J. Anat., 129 : 223-244. Kobayashi, H., M. Wada, H. Uemura and M. Ueck (1972) Uptake of peroxidase from the third ventricle by ependymal cells of the median eminence, Z. Zellforsch., 127:545-551. Lee, J. C. (1971) Evolution in the concept of the blood-brain barrier phenomenon. In H. M. Zimmerman (Ed.), Progress in Neuropathology, Grune and Stratton, New York, pp. 84-145. Lindberg, L. A., L. Vasenius and S. Talanti (1977) The surface fine structure of the ependymal lining of the lateral ventricle in rats with hereditary hydrocephalus, Cell Tiss. Res., 179:121-129. Luppa, H. and G. Feustel (1971) Localisation and characterisation of hydrolytic enzymes on the 3rd ventricle lining in the region of the recessus infundibularis of the rat - - A study of the function of the ependyma, Brain Res., 29: 253-270. Lux, Jr., W. E., G. E. Hochwald and J. Ransohoff (1970) Effect of pressure in experimental chronic hydrocephalus, Arch. Neurol. (Chic.), 23: 475-479. Milhorat, T. H., R. G. Clark, M. K. Hammock and P. P. McGrath (1970) Structural, ultrastructural and permeability changes in the ependyma and surrounding brain favouring equilibration in progressive hydrocephalus, Arch. Neurol. (Chic.), 22: 397-407. Ogata, J., G. M. Hochwald, H. Gravioto and J. Ransohoff, (1972) Light and electronmicroscopic studies of experimental hydrocephalus - - Ependymal and subependymal areas, Acta neuropath. (Berl.), 21: 213-223. Page, R. B. (1975) Scanning electron microscopy of the ventricular system in normal and hydrocephalic rabbits - - Preliminary report and atlas, J. Neurosurg., 42: 646-664. Rodriguez, E. M. (1972) Comparative and functional morphology of the median eminence. In: K. M. Knigge, D. E. Scott and A. Weindl (Eds.), Brain Endocrine Interaction - - Structure and Function, Karger, Basel, pp. 319-334. Rubin, R. C., G. M. Hochwald, M. Tiell, H. Mizutani and N. Ghatak (1976) Hydrocephalus, Part 1
344 (Histological and ultrastructural changes in the pre-shunted cortical mantle), Surg. Neurol., 5: 109-114. Russell, D. S. (1949) Observations on the Pathology of Hydrocephalus (MRC Special Report Series, No. 265), HM Stationary Office, London, pp. 119. Sahar, A., G. M. Hochwald, A. K. Sadik and J. Ransohoff (1969a) Cerebrospinal fluid absorption in animals with experimental obstructive hydrocephalus, Arch. Neurol. (Chic.), 21: 638-644. Sahar, A., G. M. Hochwald and J. Ransohoff (1969b) Alternate pathway for cerebrospinal fluid absorption in animals with experimental obstructive hydrocephalus, Exp. Neurol., 25: 200-206. Scott, D. E. and K. M. Knigge (1970) Ultrastructural changes in the median eminence of the rat following differentiation of the basal hypothalamus, Z. Zellforsch., 105 : 1-32. Scott, D. E., G. P. Kozlowski and M. N. Sheridan (1974) Scanning electron microscopy in the ultrastructural analysis of the mammalian cerebral ventricular system, Int. Rev. CytoL, 37: 349-388. Torvik, A. and V. A. Murthy (1977) The spinal cord central canal in kaolin-induced hydrocephalus, J. Neurosurg., 47: 397-402. Torvik, A. and A. E. Stenwig (1977) Pathology of the experimental obstructive hydrocephalus - Electron microscopic observations, Aeta neuropath. (Berl.), 38: 21-26. Torvik, A., R. Bhatia and R. Nyberg-Hansen (1976) The pathology of experimental obstructive hydrocephalus, Neuropath. appl. Neurobiol., 2: 41-52. Wagner, H. J. and Ch. Pilgrim (1974) Extracellular and transcellular transport of horseradish peroxidase (HRP) through the hypothalamic tanycyte ependyma, Cell Tiss. Res., 152: 477-491. Weller, R. O., H. Wisniewski, K. Shulman and R. D. Terry (1971) Experimental hydrocephalus in young dogs - - Histological and ultrastructural study of the brain tissue damage, J. Neuropath. exp. Neurol., 30: 613-626.