Polioencephalomalacia (PE), also known as Cerebro- cortical Necrosis, is a neurological disease of several ruminant species including sheep and cattle, which is diagnosed on the basis of characteristic brain lesions. Prior to the present investigation it had been established that brain lesions similar to those of natural PE, may be produced in calves and sheep by the oral administration of the thiamine antagonist amprolium. The similarity of these encephalopathies at the light microscopic level was considered to be evidence that the natural disease may be a consequence of thiamine antimetabolite production in the rumen of affected animals. Support for this proposal was then further provided by the demonstration, in the rumen of affected sheep and cattle, of an enzyme capable of producing such antimetabolites, thiaminase type I.
In this thesis the mechanisms responsible for the brain lesions of natural PE are examined, and compared with those of amprolium poisoning encephalopathy, at the ultrastructural level. The distribution of thiaminase type I, the source of the enzyme, and reasons for its accumulation in the gut contents of sheep have also been investigated.
Of 21 sheep suffering from PE, diagnosed by identifi- cation of characteristic brain lesions, 19 were found to have thiaminase type I activity in their rumen contents. Two cases of PE, which did not have detectable thiaminase activity, indicated that factors other than impaired thiamine metabòlism may occasionally be responsible for the lesions.
Ultrastructural studies of the brains of PE cases revealed that the primary morphological change in the cerebral cortex is oedema of astrocytes, and that degene- ration of other cortical elements is probably secondary. No changes were observed consistently in tissues other than the brain.
Intoxication of pre -ruminant lambs with amprolium resulted in a complex syndrome after 3 to 4 weeks treatment, one feature of which was brain lesions resembling those of natural PE. Electron -microscopy of the cerebro- cortical changes demonstrated that the primary morphological change was again swelling of astrocytes. This similarity to PE was considered to provide further evidence for the implica- tion of thiamine antimetabolites in the aetiology of the natural disease.
In addition to the encephalopathy, amprolium- poisoned lambs had diarrhoea, in common with some cases of natural PE, but investigations failed to establish the mechanisms responsible. In the terminal stages of the intoxication these lambs also developed extensive systemic haemorrhages, which predominated in actively mobile tissues. The haemorr- hages were attributed to a thrombocytopenia demonstrated during the 2 to 3 days prior to the onset of neurological disturbance. Consequent examination of the bone- marrow revealed severe degeneration of megakaryocytes, and this was considered to account for the reduced numbers of circulating platelets. Cellular depopulation of the bone -marrow, also observed in the amprolium treated animals, was associated with reduced numbers of mitotic figures in
bone -marrow cells, suggesting that amprolium may have arrested cell division in this tissue. Bone- marrow changes have not been reported in natural PE.
The brain lesions of amprolium poisoning were preceded by decreased cerebro- cortical transketolase activity, indicating that they may be a consequence of impaired thiamine metabolism in this area of the brain. In spite of the differences observed between amprolium poisoning and natural PE, it was concluded from this study that the overall similarity of the 2 conditions supported the implication of thiaminase type I in the cause of the natural disease.
Thiaminase assays were carried out using a 14C- labelled- thiamine method devised by other workers, but modified by the author to improve its sensitivity. Faecal thiaminase estimations of clinically normal animals, in flocks involved in outbreaks of PE, demonstrated that this enzyme may be wide -spread. Over half of the lambs examined in one such flock showed faecal thiaminase activity on at least one occasion during a 5 week period. In three groups of sheep the administration of certain anthelmintics, which are known to react with thiamine in the 'thiaminase type I reaction', appeared to induce further cases of PE.
Of a small number of clinically normal sheep with ruminal fistulae, some showed intermittent ruminai thiam- inase which appeared as peaks of activity of 3 to 10 days duration, and in one animal peaks were repeated during a period of several months. Thiaminase was not detected in the faeces of these sheep even when ruminai enzyme activity was high, which indicates that faecal thiaminase studies
xi could under -estimate the 'thiaminase status' of flocks. It was concluded from this work, that PE is a clinical manifestation of a much more extensive condition, character- ised by the presence of thiaminase type I in the alimentary tract.
Samples taken from feedstuffs in use during outbreaks of PE showed no evidence of thiaminase, which suggests that in the outbreaks investigated the enzyme was probably not ingested by the sheep but synthesised in the alimentary tract. A subsequent search for the source of thiaminase in the rumen contents and faeces of sheep affected with PE resulted in the isolation of bacteria which produced thiaminase type I activity in vitro. After taxonomic studies they were considered to be strains of Bacillus thiaminolyticus. The thiaminases produced by these organisms were found to have similar co- factor specificities to the naturally- occurring enzymes and they mediated in the base - exchange reaction implicated in the aetiology of PE. It was thus proposed that thiaminase type I present in sheep may be synthesised by strains of B. thiaminolyticus. This represents the first reported account of the isolation of B. thiaminolyticus in Britain.
Attempts to confirm that these organisms are responsible for the presence of thiaminase, by the administration to sheep of ovine strains of B. thiaminolyticus, failed to influence ruminal activity of this enzyme. Subsequent studies demonstrated that these organisms are inhibited by a range of other bacterial species, and by a number of volatile fatty acids (VFAs).
Concurrent studies by other workers have resulted in the isolation of strains of Clostridium sporogenes from sheep affected with PE. It was found in the present investigations that these organisms, which produce thiamin - ase type I similar to that of B. thiaminolyticus, are also inhibited by the same bacterial species and VFAs as B. thiaminolyticus. It was therefore concluded that a similar range of influences on rumen function could control thiaminase type I production, and consequently the occurrence of PE, irrespective of which of these 2 bacterial species was involved.