We investigated the behavioral changes in Toxoplasma-infected mice that developed brain cysts in two environments with exposed versus non-exposed areas and observed that chronic infection modifies a series of specific aspects of unconditioned behaviors. In infected animals, exploratory locomotion was more common, faster and had higher initial acceleration. Strikingly, this locomotion was also organized differently relative to control groups: infected animals exhibited fewer and longer segments of ambulatory movement. Additionally, these animals did not display normal cautious behaviors when placed in a novel environment, and they also showed differential responses to unsafe areas with reduced unconditioned fear and more risky behaviors.
When we reduced the dimensionality of 37 behavioral features to 5 different factors using principal component analysis, we were able to cluster the behavioral differences between animals with brain cysts and the different control groups into two main factors. These factors mostly represented features related to the microstructure of exploratory behavior (F4) and risk/unconditioned fear (F5). Surprisingly, no significant differences were observed between experimental groups for the factor representing features of general locomotion (F1). This result indicates that the behavioral modifications in exploratory locomotion observed in mice with brain cysts likely reflected the altered structure of exploratory behavior, rather than representing hyperactivity.
The data presented here suggest that chronically infected animals show differential responses to exposed areas, away from normal defensive behavior, and are consistent with previous studies in infected rats 
. However, our data do not allow us to discriminate whether these behavioral alterations are caused by changes in the way environmental risk is evaluated, and/or by changes in producing/organizing the appropriate behavioral responses.
We observed that although cysts in chronically infected animals were not randomly distributed across different brain areas, there was no special frequency of cyst accumulation in specific brain regions (e.g. in the amygdala as has been reported in previous studies 
). However, we found evidence that particular combinations of cyst distribution in the brain biased the animals for specific behavioral phenotypes (changes in risk/unconditioned fear, described by F5). These results suggest that cyst accumulation in different areas of a particular circuit may lead to similar behavioral alterations and thus that the parasite may have experienced selective pressure to manipulate functional neuronal circuits rather than a specific area.
Another possibility is that the parasite would change the secretion or function of general neuromodulators. For example, altered concentrations of catecholamines and indolamines have been observed in whole brain extracts of Toxoplasma
-infected mice. In particular, dopamine levels have been reported to be higher in Toxoplasma
-infected mouse brains 
, and a recent study showed that brain cysts were able to produce tyrosine hydroxylase, an enzyme involved in dopamine biosynthesis, and also that Toxoplasma
-infected dopaminergic neurons showed an increase in dopamine synthesis and release 
. These observations suggest that dopamine could be involved in some of the behavioral modifications described here, namely in movement structure (F4), since these seem to be less dependent on cyst localization.
An interesting, and to our knowledge, previously unreported phenotype was observed in the infection control group. In this group, systemic parasitic contact did not result in clinical symptoms of infection (as indicated by lack of antibodies against Toxoplasma
, weight loss and brain cysts). The absence of detectable circulating antibodies against this parasite is intriguing. One possibility would be that the concentration of antibodies is below the detection levels of the method used and, alternatively, that mice in this experimental group used cell-mediated immune response mechanisms. In fact, cell-mediated immune reactions (involving CD4+
T cells and macrophages) are believed to be involved in the defense against intracellular parasites (reviewed in 
). Furthermore, there are cell-autonomous defense mechanisms implicated in resistance to Toxoplasma
infection, involving IFN-inducible immunity-related GTPases, which drive targeted destruction of parasite-containing vacuoles (reviewed in 
). Despite the absence of infection symptoms and brain cysts in this experimental group, contact with the parasite was sufficient to alter behavioral responses to exposed areas. This is consistent with previous studies showing that asymptomatic Campylobacter jejuni
-infected animals display an increase in closed arm entries in the elevated plus maze 
. Subsequent studies have demonstrated that peripheral infection, even in the absence of a measurable immune response, can activate viscerosensory pathways that interface with defensive brain networks 
. This suggests that in our infection control animals, information about transient Toxoplasma
infection could have been relayed to visceral sensory structures in the brain therefore leading to the observed behavioral changes. It is also of interest that the behavioral changes observed in this experimental group are generally opposite to those observed in animals with brain cysts. This further suggests that the behavioral alterations in the different groups arise by different mechanisms. In the “Brain cysts” group the behavioral modifications observed are related to cyst presence in the brain (and specific cyst localizations in the case of F5), whereas in the “No-cysts” group alterations may result from a systemic effect that produces changes in the brain.
The behavioral differences described here for infected animals are not easily explained by general debilitation due to Toxoplasma
infection, because animals with brain cysts show increased performance in some behavioral variables and decreased performance in others. Rather, these results suggest that chronically infected mice interact with their environment differently than non-infected animals. This change results in maladaptive behaviors, such as longer bouts of locomotion in exposed areas, or increased exploration of unsafe zones. These behaviors would render infected animals more vulnerable to predation or environmental risks, the latter resulting in enhanced capture probability. This is unlikely to be a general effect of parasite-altered behavior. Parasite-driven behavior modifications do not necessarily lead to increases in trappability or general predation, since other studies have shown that parasites can manipulate intermediate host behavior to actually decrease general predation (reviewed in 
One interesting hypothesis is that in addition to the selective pressure to increase the vulnerability of infected rodents to felid hosts (where sexual and asexual reproduction occur), there may have been selective pressures to increase transmission to host predators in general (including intermediate hosts, where asexual reproduction occurs). Several studies have suggested a number of mechanisms through which Toxoplasma
could increase the likelihood of predation by the definitive host (reviewed in 
). The behavioral modifications described here could serve to increase the overall capture probability observed in Toxoplasma
-infected rodents 
, and thus be of evolutionary importance for parasite transmission between a variety of different hosts. Even though there is no consensus over the actual contribution of asexual reproduction to Toxoplasma
population structure (reviewed in 
), some studies suggest that Toxoplasma
has been expanding largely clonally (asexually) for the past 10 000 years, with the interesting implication that expansion of intermediate host range could be one of the driving forces behind the success of this parasite 
. Some studies even argue that there is no reason why Toxoplasma
could not skip the definitive host altogether, using carnivorism and scavenging behaviors to move within the food chain 
. This would imply that the parasite would also be under selective pressure to manipulate behavior in intermediate hosts other than the rodent, including those that would not be normally predated by felids. In this respect, it is interesting to note that a series of studies have shown that chronic toxoplasmosis in humans can result in behavioral modifications, like increased activity, decreased reaction times and altered personality profiles 
The increase in host range and success of clonal expansion of Toxoplasma
populations could result from the acquisition of the ability to directly infect successive intermediate hosts after cyst ingestion 
. This bypass of sexual reproduction is absent from a parasite closely related phylogenetically to Toxoplasma
, Hammondia hammondi
. This cyst-forming parasite has a limited host range, with cats as definitive hosts and rodents as unique intermediate hosts. Interestingly, H. hammondi
cysts are rarely found in the brain 
, even though this parasite needs to reach the definitive host to complete its life cycle. Therefore, brain cyst formation during Toxoplasma
infection may not be the result of selective pressures to increase parasite transmission to the definitive host. Instead, it is interesting to postulate that the appearance of brain cysts is related to the increase in clonal expansion and intermediate host range observed in Toxoplasma
, by the modification of the structure and the risk of intermediate host behavior. However, given the controversy that exists regarding the relative roles of sexual and asexual reproduction for Toxoplasma
evolution, it is also possible that the impact of the behavioral alterations observed here is mainly to increase transmission to the definitive host. Further studies are needed to investigate these postulates.