A comparative study of the bacteriology of the GI tract of horses

Lynsey Garrett Holder of the John Gilmour Studentship of the Equine Grass Sickness Fund

Date: 04/06/01

Lynsey Garrett is a final year student within the Department of Medical Microbiology at the University of Edinburgh, where she holds the John Gilmour Studentship awarded by the Equine Grass Sickness Fund. She is working on a comparative study of the bacteriology of the GI tract of horses suffering from Equine Grass Sickness (EGS) and healthy animals.

Earlier work carried out by Leonie Hunter, a former PhD student colleague, demonstrated a strong correlation between horses with EGS and the presence of Clostridium botulinum type C neurotoxin in the gut. One of the most significant features of this disease is intestinal stasis. We have therefore hypothesised that this is a result of damage to the enteric nervous system following a toxicoinfection by Clostridium botulinum type C.

The aims of the current PhD study are

1. to determine the normal microbiota of the GI tract of healthy horses by conventional bacteriological techniques, 2. to compare the microbiota from the healthy horse with those with EGS, both chronic (CGS) and acute (AGS), and with colic; to observe if changes in the gut bacterial populations occurred during stasis and 3. to develop molecular methods for detecting toxigenic clostridia in the horse GI tract and the environment.

To summarise, the following results have been obtained:

(1) The ratio of anaerobic bacteria to aerobic bacteria present in the faeces of the 22 horses investigated is approximately 1:1 in all groups studied, but mean counts were 10-100-fold higher in the animals with GI disease (EGS and colic) compared to the controls.

(2) There were 10-100-fold more Gram-positive anaerobic cocci observed in animals with GI disease compared to healthy controls. However, most of the seven species of anaerobic cocci identified were found in both control and diseased animals.

(3) The carriage of bacteria from the genus Clostridium was higher in diseased animals compared to the controls. Fifteen different Clostridium species were isolated from the GI disease groups, compared to only one species, Cl. bifermentans, in controls. This may have implications in EGS and colic.

(4) The mean faecal clostridial counts in CGS were 10-fold higher than in the controls and 100-fold higher than in the AGS cases. This contrasted with the mean counts for ileal samples, which showed a 10-fold increase for clostridia in AGS but 1000-fold reduction in CGS, when compared to the corresponding faecal counts. The results indicate an increase in bacterial populations in the GI tract of animals with GI stasis compared to the controls and the clostridial species are prominent in EGS and colic. Whether these results are the cause of GI stasis or a consequence remain to be determined.

Cl. botulinum type C, the organism of our main interest, is extremely fastidious and difficult to isolate using conventional techniques. The aim is to use a molecular technique known as the polymerase chain reaction (PCR) to identify Cl. botulinum type C from clinical samples. Several extraction methods have been investigated in order to determine the most efficient way of extracting DNA from the clinical samples. The maximum and minimum detection levels of the toxin genes must also be established before the technique can be used on clinical samples.

The clinical samples have first been screened for the presence of a consensus sequence common to all clostridia. Those that were positive were then screened for the type C toxin genes. Clostridium -positive samples have also been screened for Cl. botulinum type A, B and E genes in order to exclude their involvement in the disease. The toxin genes for Cl. botulinum type D are also being investigated as this type of organism is phenotypically indistinguishable from type C organisms (apart from the toxin produced). Phage conversion can result in type D organisms producing the type C toxin and vice versa.

PCR methodologies have been developed for pure culture and “spiked” faeces and soil but to date no positive results have been observed for any of the toxin genes in individual specimens. However, it is probable at this stage that the PCR method is not sensitive enough. The minimum sensitivity is 105 bacteria per gram. The samples being used also contain many potential PCR inhibitors and this may also be causing negative results. However new methods are now being implemented which will “clean-up” the clinical sample and positive results should be able to be detected more readily.

Hopefully through the use of molecular techniques, the mystery of the cause of Equine Grass Sickness will be solved.