Hiding inside the host: development and application of "Neospora caninum" bradyzoite in vitro culture

The bradyzoite stage of the apicomplexan parasite Neospora caninum
represents a hypobiotic, slowly proliferating and tissue cyst forming stage, which
can survive in the immunocompetent host for several years. Tissue cysts are
orally infectious. In addition, as immunocompetence gets impaired such as in the
pregnant cow, bradyzoites transform into the rapidly proliferating and more
virulent tachyzoites, which break out of the tissue cysts, cross the placenta and
infect the unborn foetus, causing abortion, stillbirth or the birth of weak calves.
Thus, the bradyzoite stage is epidemiologically important, since it plays a crucial
role in both oral and transplacental transmission.
During the thesis work, an in vitro cultivation model for the generation of
bradyzoites containing tissue cysts was developed that allowed to study this
parasite stage more closely.
In preliminary studies, different cell lines like Vero cells, fibroblasts and
neuroblastoma cells were infected with several Neospora caninum isolates and
stage conversion was induced using pH-stress, the macrolide antibiotic tylosin
and heat stress. However, the number of parasites expressing the bradyzoitespecific
antigen BAG1 remained very low. Besides, the host cells tended not to
tolerate such culture conditions that rendered a cultivation period over several
days impossible.
As an alternative approach to induce stage conversion, organotypic brain slice
cultures of rat cortical tissue were used for infection. This cultures provide a
three-dimensional array of central nervous tissue with the original complex
mixture of neuronal cells. The rationale behind this approach was that cofactors
that could contribute to stage conversion were more likely to be provided in such
a system than within a single cell. Organotypic brain slice cultures were infected
with 2x106 and 2x107 tachyzoites. To some cultures, 100 units of recombinant
mouse IFN-γ were added. IFN-γ, a key cytokine in the host immune response to
Neospora caninum infection has shown to trigger stage conversion in
Toxoplasma gondii infected macrophages and microglia. However, in none of the Neospora caninum infected cultures, IFN-γ-treated or untreated cultures,
parasites that expressed the bradyzoite marker BAG1 could be detected, neither
a cyst wall was found by transmission electron microscopy.
In order to monitor the kinetics of parasite proliferation in this system, a
quantitative real-time PCR was developed using a dual fluorescent hybridization
probe system and the LightCycler™ instrument for online detection of amplified
DNA. Treatment of the infected cultures with IFN-γ resulted in an inhibition of
parasite proliferation compared to the untreated cultures. Moreover, smaller
pseudocysts were found after IFN-γ-treatment. In addition, neuronal cytoskeletal
elements, namely glial acid protein filaments as well as actin microfilament
bundles largely colocalised with the pseudocyst periphery.
Although the use of organotypic brain slice cultures did not succeed in stage
conversion of Neospora caninum, a model was established that can be used to
gain more information on the cerebral phase of Neosporosis.
Since inhibitors of the mitochondrial respiratory chain were found to promote
stage conversion in Toxoplasma gondii, sodium nitroprusside, an exogenous
donor of nitric oxide that inhibits the respiratory chain at cytochrome oxidase was
tested for stage conversion in Neospora caninum. Additionally, long term
cultures of murine epidermal keratinocytes were used as host cells, because of
their very strong cell-substrate adhesion. Sodium nitroprusside was daily added
at a concentration of 70 μM to infected Vero cells, fibroblasts, neuroblastoma
cells and keratinocytes for up to 8 days. Keratinocytes were the only host cells to
withstand this treatment for a period of at least eight days. In these cultures,
sodium nitroprusside strongly inhibited Neospora caninum proliferation as
assessed by quantitative real-time PCR and induced the expression of BAG1
antigen from day 3 onwards. This data suggests that the inhibition of parasite
proliferation is closely linked to the expression of bradyzoite-specific markers. At
day 8, around 60% of the parasitophorous vacuoles contained BAG1-positive
parasites. Another marker, which was used, was mAbCC2 that is directed
against a Toxoplasma gondii cyst wall protein. After 8 days, around 60% of all
parasitophorous vacuoles exhibited a peripheral labelling with this antibody that
indicated the formation of a cyst wall. Inspection by transmission electron microscopy suggested that the majority of the intracellular compartments
occupied by Neospora parasites exhibited features that were indicative for
tachyzoite-to-bradyzoite stage conversion. Vacuoles contained 1-5 intracellular
parasites, featuring electron dense rhoptries, many micronemes that were
predominantly located at the anterior end, and large and small dense granules.
In addition, amylopectin granules, that are characteristic for the bradyzoite stage,
were found. The majority of the parasitophorous vacuoles contained an
accumulation of electron dense granular material at the periphery that indicated
the formation of a cyst wall, which varied considerably in thickness between 0.1-
1 μM.
Although this developed in vitro culture system was efficient in terms of NcBAG1-
expression, it was less suitable in order to obtain larger amounts of Neospora
caninum bradyzoites that are required for biochemical and molecular studies.
The separation of bradyzoites from keratinocytes was difficult, due to the fact,
that the tissue cysts were surrounded by keratin filament bundles, that hindered
the liberation of the parasites. The culture conditions were therefore adapted to
Vero cells by decreasing the concentration of sodium nitroprusside to a minimum
level of 17 μM, which was tolerated by the less adhesive cell type. The efficiency
of tachyzoite-to-bradyzoite conversion was similar to that obtained with 70 μM
sodium nitroprusside in keratinocytes as host cells, but resulted in a higher
number of NcBAG1-positive individual zoites, due to the increased size of
NcBAG1-positive vacuoles. Additionally, the modified system made it now
possible to purify bradyzoites out of the host cells. Furthermore, sodium
nitroprusside-treatment of infected Vero cells lead to a down regulation of the
major tachyzoite surface antigens NcSAG1 and NcSRS2, as assessed by
immunofluorescence and immunoblotting. The expression and localisation of
dense granule proteins in bradyzoites was also analysed by immunogold
transmission electron microscopy and immunofluorescence and showed that the
localisation of these proteins shifted towards the periphery of the cysts,
compared to tachyzoites where dense granule proteins were found at the
anterior and posterior end of the parasites. These results implied an involvement
of dense granuIe proteins in the formation and modification of the cyst wall. In addition, using purified tachyzoites and bradyzoites, the adhesion and the
invasion of these two stages to Vero cells was comparatively assessed and
demonstrated that tachyzoites were more invasive than bradyzoites. However,
removal of sialic acid by sialidase from the Vero cell surface and parasite surface
enhanced the bradyzoite invasion rate from 25% to 46% and 15% to 36%
respectively, whereas these treatments had no effect on the tachyzoite invasion
rate. Thus, sialic acid plays an important role in the invasion of host cells by
bradyzoites.