Cowdria ruminantium (Rickettsiales) causes heartwater, a disease of high mortality in
susceptible ruminants. A description of the isolation of new Cowdria isolates by
different methods and from different vectors and geographical locations in Kenya is
given. These included Amblyomma variegatum, A. gemma and A. lepidum. Isolates
from the later two species were also by feeding adults moulted from nymphs collected
in the field and is the first report on transtadial transmission by A. gemma ticks. A
spectrum of virulence ranging from highly virulent to mildly virulent for sheep was
found among the Cowdria isolates. The majority of isolates were highly virulent.
There was a range of mouse infectivity among the isolates from inapparent to lethal.
The Asembo and Baragoi isolates were pathogenic and lethal, the Kiswani, was
infective and non pathogenic for Balb/C mice while the other 8 were avirulent or
retractile to mice inducing only antibody production in various proportions of mice.
There was a difference in the infectivity for neutrophils both in the frequency of
infected cultures and in their level ofinfection. The different isolates were classified as
of low infectivity where even the few positive cultures rarely reached 1% infection
rate, medium infectivity if a good number of cultures regularly attained 1% infected
neutrophils, or high infectivity if a large proportion of the cultures became positive
regularly attained 1% infected neutrophils and at least some of them attained more
than 10% infected neutrophils. The isolates also had a range ofinfectivity for the brain
endothelial cells, from no detectable colonies to greater than 16% infected endothelial
cells in individual animals. A diagnostic index, that is the ease by which diagnosis
could be made at post mortem (% endothelial cells where colonies were found), for
each isolate by brain crush smears was formulated and virulent isolates generally had a
high diagnostic index whereas low virulent isolates had lower diagnostic indices, the
correlation between RI and DI being highly significant (p< 0.001 Spearmans
correl?+:on coefficient rs = 0.957). All isolates were confirmed to be Cowdria by PCR
and western blots. A one way cross immunity was performed with all isolates using
animals that had been treated or that had naturally recovered during the infection
studies. A spectrum of protection was seen ranging from zero to complete protection.
A two way cross immunity was performed with five isolates which had been selected
on the basis of their differences in virulence, (two milder and three virulent for sheep),
tick source and diverse geographical origin, four isolates were found to confer
different degrees of overall protection against the other isolates. In the order of their
ascending virulence, Bamba, Asembo, Baragoi and Suswa, gave, 18.8%, 56.1%,
98.8% and 67.0% cross protection respectively. Therefore the least virulent gave the
least overall protection while the second most virulent (the Baragoi isolate) gave the
broadest overall protection against all the other isolates. This supports the suggestion
that immunogenicity is closely related to the pathogenicity and virulence of Cowdria.
It also implies that total immunogenicity may be found in one rather than a cocktail of
stocks. Neither of the low virulence nor the high virulence isolates may be an
acceptable vaccine. However, the one way trial indicated that one of the intermediate
virulence isolates the Marigat isolate should be evaluated as it was protective against
the Baragoi isolate with a 98.5% reduction in the reaction index and be safer for use
in sheep. The two main vector ticks in Kenya, A variegatum and A gemma were
examined for their susceptibility to infection with Cowdria (Asembo and Bamba)
isolated from both species respectively. However infection rates were very low for
both isolates to both species and no conclusions could be drawn, perhaps because the
Cowdria isolates selected were both the milder ones.
The author concludes that the agent of heartwater is endemically widespread in many
districts in Kenya and poses a potential threat of outbreaks to areas newly invaded by
vector ticks and also to areas where immunity due to the local agent may not protect
against an invading one. The author recommends that the Baragoi or Marigat isolate
should be adopted for possible vaccine development in Kenya.