Hypolocomotive behaviour associated with increased microglia in a prenatal immune activation model with relevance to schizophrenia

Highlights

• PPI was not significantly disrupted in MIA rats but startle response seemed lower.

• Poly I:C animals showed significant hypolocomotive behaviour compared to controls.

• Characteristics of microglia activation were observed in the MIA brain.

• In the chronic stage, however, the microglia were not in a reactive state.

• Supporting evidence for a role played by microglia in psychiatric illness.

Abstract

Over the past decade a neurodevelopmental animal model with high validity for schizophrenia has been developed based on the environmental risk factor known as maternal immune activation (MIA). The immunological basis of this model, together with extensive data from clinical and preclinical context, suggests the involvement of an aberrant neuro-immune system in the pathophysiology of schizophrenia. The goal of this study was to examine microglia activation in adult behaviourally phenotyped MIA offspring. MIA was induced in pregnant rats using viral mimetic Poly I:C at gestational day 15. Adult offspring were behaviourally phenotyped at postnatal days (PND) 56, 90 and 180 through the evaluation of prepulse inhibition (PPI) of the acoustic startle and spontaneous locomotion. Finally, the presence of activated microglia in brain regions associated with schizophrenia was evaluated using post-mortem immunohistochemistry against OX-42 (CD11b) and ED-1 (CD68). Although a deficit in PPI could not be replicated despite the high number of animals tested, we found an overall decrease in basal startle response and spontaneous locomotion in offspring born to Poly I:C- compared to saline-treated dams, accompanied by increased microglial density with characteristics of non-reactive activation in the chronic stage of the model. These findings provide additional evidence for a role played by microglial activation in schizophrenia-related pathology in general and psychomotor slowing in particular, and warrant extensive research on the underlying mechanism in order to establish new drug targets for the treatment of schizophrenia patients with an inflammatory component.

Introduction

Schizophrenia is one of the most prevalent (1%) and disabling mental disorders affecting mankind [1]. It has been hypothesised to have a neurodevelopmental course during which a perinatal interaction of environmental and (epi)genetic factors results in a profound disruption of human functioning after a latent period of about 20 years [2]. While the underlying pathophysiological mechanisms have not been fully elucidated, extensive data from both clinical and preclinical trials suggests the involvement of an aberrant immune system [3], [4]. First, many of the neurodevelopmental risk factors, both genetic and environmental, are associated with the immune system, exemplified by MHC-gene and Interleukine-1 (IL-1), IL-2 and IL-4 gene polymorphisms, prenatal infection and stress [2], [3], [5], [6]. Second, the innate as well as the adaptive immune system of schizophrenia patients show signs of overactivation, including in cytokine values, B-lymphocyte populations, antibody production and differentiation and activation of T-lymphocytes [3]. Apart from peripheral immune alterations, a significant increase in activated microglia has been demonstrated in the frontotemporal brain regions of schizophrenic patients in three post-mortem as well as three in-vivo PET studies [7], [8], [9], [10], [11], [12]. Microglia are the resident macrophages of the brain and are the first in line to respond to pathological changes. Although microglia function to restore homeostasis, upon overactivation, the release of cytotoxic factors (e.g. pro-inflammatory cytokines, kynurenines, nitric oxide and reactive oxygen species) aimed at destroying pathogens can have detrimental neurotoxic effects [13]. Clinical trials now focus on the investigation of brain inflammation as a potential new drug target for the treatment of schizophrenia by adding anti-inflammatory medication to the treatment regimen of schizophrenia patients. The results of these trials seem very promising but are not conclusive to date [14], [15], [16], [17].

Although extreme care is required when modelling a psychiatric disorder in rodents, in the past decade a neurodevelopmental animal model with behavioural, morphological and neurochemical characteristics related to schizophrenia has been developed. This is based on the environmental risk factor known as maternal immune activation (MIA) [18]. The hypothesis of this model states that MIA disturbs normal brain development in offspring due to the induction of immune effectors in the pregnant mother that interfere with growth and differentiation of the foetal brain during gestation [19], [20]. The most commonly used cytokine inducer is Poly I:C, a synthetic analogue of viral double-stranded RNA which binds to the Toll-like receptor 3 (TLR-3), mimicking the acute phase response following a viral infection by inducing interferons and cytokines [21]. In addition to acutely upregulating pro-inflammatory cytokines in the maternal and foetal compartment, MIA also induces immune alterations that persist into later life and have the potential to mediate pathology. Recently, MIA by Poly I:C was shown to lead to long-lasting, region- and age-specific changes in brain en serum cytokines in mice offspring, similar to those reported in schizophrenia patients. For example, it was observed that predominantly pro-inflammatory cytokines were elevated at birth in frontal and cingulate cortices, after which they decreased during periods of synaptogenesis and plasticity, before increasing again around adulthood [22]. Additionally, Juckel et al. reported an increase in microglial activation in the hippocampus and striatum of Poly I:C mice at postnatal day 30 [22]. Moreover, an increase in the number of activated microglia has also been found in the brain of both juvenile and adult offspring of LPS-treated dams [23], [24].

Sensorimotor gating and locomotor activity are two frequently evaluated behaviours in the MIA model. Sensorimotor gating or the ability to filter incoming sensory information can be evaluated by measuring prepulse inhibition (PPI) of the acoustic startle response (ASR) [25]. This behaviour is particularly relevant to schizophrenia since PPI is impaired in both schizophrenia patients and people at high risk of developing schizophrenia [26]. Secondly, the locomotion test is a frequently used behavioural test in acute animal models of schizophrenia. Since dopamine agonists are known to induce schizophrenia-related symptoms in healthy subjects, amphetamine was initially used to model acute symptoms of schizophrenia in rodents, resulting in hyperlocomotion and stereotyped behaviour. In addition, NMDA antagonists have also been widely utilised since the introduction of the NMDA hypofunction theory of schizophrenia [27]. Amphetamine and NMDA antagonist challenges are frequently carried out in neurodevelopmental models of schizophrenia in order to evaluate the dopaminergic and glutamatergic state of the animal. However, results are often not univocal and both sensitisation and resistance to these challenges have been observed [28], [29], [30], [31], [32]. Measuring spontaneous locomotion may provide information regarding an altered endogenous dopaminergic and glutamatergic tone in the animals. However, this has not been systematically studied in the MIA models.

The goal of this study was to investigate whether microglia activation in male and female offspring is a neurobiological correlate to the altered phenotype in the chronic phase of the rat MIA model with relevance to schizophrenia. By including both male and female offspring, the present study also allows to investigate potential differences in sensitivity of females compared to males for developing schizophrenia-related behaviour as observed in the clinic. Behavioural phenotyping was performed by sensorimotor gating analysis and the evaluation of spontaneous locomotion.