The Methylazoxymethanol Acetate Rat Model: Molecular and Epigenetic Effect in the Developing Prefrontal Cortex

Abstract

Gulchina and colleagues describe molecular and epigenetic changes in the developing prefrontal cortex (PFC) of the rats exposed to methyl azoxymethanol acetate (MAM). They found an n-methyl-D-aspartate receptor (NMDAR) hypofunction present in the PFC of juvenile MAM rats which was associated with abnormal epigenetic regulation of the Grin2b gene. These changes may be related to early cognitive impairments observed in MAM rats and schizophrenia patients.

Graphical Abstract

This Editorial highlights an article by Gulchina and colleagues in the current issue of the Journal of Neurochemistry, in which the authors describe molecular and epigenetic changes in the developing prefrontal cortex (PFC) of the rats exposed to MAM. They found an NMDAR hypofunction present in the PFC of juvenile MAM rats which was associated with abnormal epigenetic regulation of the Grin2b gene. These changes may be related to early cognitive impairments observed in MAM rats and schizophrenia patients.

Schizophrenia is a developmental disorder that typically manifests in late adolescence or early adulthood. While genetic predisposition clearly plays a major role in vulnerability to schizophrenia, genetics alone cannot account for the susceptibility for the disease. A more updated view is that a combination of genetic and environmental factors is necessary for disease manifestation. In fact, prenatal maternal immune activation (MIA), obstetric complications, adolescent cannabis use, stress, urbanicity, migration, and other factors can interact with predisposing genetics to increase risk for illness (Millan et al., 2016). Whether these disparate risk factors converge onto specific pathways is under debate. One hypothesis is that glutamate N-methyl-D-aspartate receptor (NMDAR) hypofunction could be a major convergence point for schizophrenia (Snyder & Gao, 2013).

Many studies have indicated that NMDAR antagonists, such as phencyclidine (PCP) and ketamine, can induce or exacerbate schizophrenia-like positive, negative, and cognitive symptoms.(Hu ^{et al.}, 2015) These observations broadly posit the involvement of NMDAR in schizophrenia, but the exact mechanisms remain elusive. One possibility is that NMDAR dysfunction during development contributes to the onset and the progression of schizophrenia symptoms (Snyder & Gao, 2013). However, this view has yet to be widely replicated in neurodevelopmental models of schizophrenia.

Several neurodevelopmental models of schizophrenia have been developed in recent years, and among these the methylazoxymethanol acetate (MAM) model has provided a translatable framework for clinical research. The MAM model is based on a single administration of the mitotoxin MAM to pregnant rats at gestational day 17, which produces in the offspring many characteristics that are consistent with clinical observations. For example, the MAM model recapitulates behavioral deficits (e.g. prepulse inhibition of startle, reversal learning, latent inhibition, social interaction), pharmacological responses (e.g. increased locomotion to amphetamine and to PCP), and anatomical changes (e.g. thinning of limbic cortices, loss of parvalbumin interneurons) similar to that observed in schizophrenia (Modinos et al., 2015). Furthermore, consistent with the onset of psychosis in schizophrenia patients, most of the alterations related to the positive symptoms observed in the MAM rats become evident only after puberty, whereas social withdrawal and cognitive deficits are present both before and after puberty (Gomes et al., 2016). Thus, this model offers the possibility to investigate the link between neurodevelopmental disruption and the transition into a schizophrenia-like phenotype in the adult.

The exact molecular mechanisms by which MAM induces a schizophrenia-like phenotype are still unknown, but at system level, MAM produces disruption of brain development (especially the hippocampal-midbrain-striatal circuits) believed to be fundamental to psychotic disorders (Modinos et al., 2015). Hence, the MAM model presents considerable face and predictive validities, and has been utilized to understand the mechanisms of current antipsychotic treatments and novel therapeutic approaches (Modinos et al., 2015). In contrast, the molecular mechanism of this model is essentially understudied, particularly with reference to age-specific molecular characterization, rendering the validity questioned by some groups. Fortunately, such gap in knowledge has been receiving increasing attention and inroads are being made into this question (Figure 1). For example, proteomic and metabolomic analyses identified major deficits in hippocampal glutamatergic network in MAM rats, analogous to that of schizophrenia patients (Hradetzky et al., 2012). However, fewer studies have directly examined prefrontal glutamatergic transmission in a developmental manner, which is thought to underlie cognitive deficits in schizophrenia (Snyder & Gao, 2013).

Figure 1.

Increasing support for the validity of MAM model of schizophrenia. The precise mechanism in which a single injection of MAM at gestational day 17 (E17) causes schizophrenia-like phenotypes in adults remains to be elucidated. Prenatally, the DNA alkylating agent MAM may selectively alter expression of genes that are implicated in normal brain development, either directly through genetic or indirectly through epigenetic mechanisms. Other factors such as immune activation prenatally may act in a similar manner to contribute to the initiation of the deficits. Gulchina and colleagues add further molecular evidence to the MAM model (see highlighted), indicating that glutamatergic neurotransmission in the PFC may be affected in a developmental manner. Thus, one hypothesis is that the “construct” validity of MAM takes place in early development, when animal are vulnerable to stress and other environmental stimuli.

In the current issue, Gulchina and colleagues (2017) describe several molecular and epigenetic changes in the developing prefrontal cortex (PFC) of the rats exposed to MAM, and further validate the relevance of this model to human disease. Evaluating if an NMDAR hypofunction would be present in early postnatal development, the authors first surveyed the expression of multiple NMDAR subunits and report a selective decrease in the NR2B subunit in the PFC of MAM rats at postnatal day 21. This change was associated with functional deficits in the glutamate synapses indicated by whole-cell recordings from pyramidal neurons showing a decrease in both spontaneous and evoked NMDA-mediated excitatory currents.

It is well-known that during early postnatal development, NMDARs switch their composition from primarily NR2B-containing to NR2A-containing subunits. This subunit exchange occurs throughout the brain and, at the gene level, is regulated by a transcriptional repressor, the repressor element 1-silencing transcription factor (REST), through epigenetic remodeling of Grin2b gene (NR2B-encoding) (Rodenas-Ruano et al., 2012). Thus, to further gain mechanistic insights regarding how the NR2B subunit is selectively affected, Gulchina and colleagues (2017) examined if epigenetic mechanisms could contribute to the hypofunction of NR2B-NMDARs in the mPFC of juvenile MAM rats. They found that the NR2B protein loss and NR2B-NMDAR dysfunction correlate with increased enrichment in the transcriptional repressor REST and with increased repressive histone marker H3K27me3 at the promoter region of Grin2b. Taken together, the authors conclude that the histone methylation-mediated hyper-repression of Grin2b in MAM rats contributes to the protein loss in synaptic NR2B and NMDAR hypofunction that may underlie early cognitive impairments observed in MAM rats and schizophrenia patients.

Evidence suggests that maladaptive changes in NMDAR subunit composition and function may have negative consequences for activity-dependent cortical development (Paoletti et al., 2013) and are likely to provide vulnerability to the deleterious effects of stress and other environmental influences that could trigger pathological circuit deficits. In this context, we found recently that MAM rats examined peripubertally show higher anxiety and stress responsivity compared to controls (Du & Grace, 2013). Interestingly, the administration of the anti-anxiety agent diazepam peripubertally at a dose sufficient to attenuate anxiety prevented the schizophrenia phenotype in the adult MAM rats (Du & Grace, 2013). Thus, our current hypothesis is that MAM itself may not “cause” schizophrenia-like phenotypes through a direct molecular mechanism, but rather confer increased vulnerability to adverse environmental stimuli (Gomes & Grace, 2017).

It is still unknown if the disrupted stress responsivity during early development observed in MAM rats is caused by NMDAR hypofunction or increased expression of repressive epigenetic markers. Interestingly, increased activation of REST-mediated gene regulation in the PFC during postnatal development is also involved in stress vulnerability (Uchida et al., 2010). REST is estimated to regulate over 2000 genes, and its target specificity varies by cell type, age, and disease states (Bruce et al., 2004). By far, the factors that determine the specificity of the interaction between REST and its targets are unknown. One attractive possibility is that an age-dependent epigenetic signature may in part contribute to the affinity and specificity of REST interaction (Hwang et al., 2017). Therefore, it is critical to characterize the epigenetic landscape in MAM rats in a developmentally-relevant manner. The REST complex serves as host for a multitude of epigenetic regulators, and given their frequent crosstalk, enzymes such as a variety of histone deacetylases (HDACs) may also be involved in the observed process. Thus, to further map the molecular mechanism that confers the observed REST specificity for Grin2b, future studies need to increase the pool of assayed lysine sites and epigenetic regulators. Due to the crucial role of NR2B subunits in cognition, a loss of NR2B protein and subsequent NMDAR dysfunction may underlie cognitive impairments in the MAM rats and schizophrenia patients. Thus, future research should also evaluate the behavioral relevance of the observed changes to schizophrenia symptoms, especially cognitive function.

In a broader context, the observed increase in REST binding and histone hyper-methylation at the promoters of NR2B subunit reflect an overall dysregulated epigenetic profile. Epigenetics is a broad term describing changes to chromatin that alter gene transcription without changing gene sequence. Two major epigenetic mechanisms are DNA methylation and a variety of histone modifications, including histone acetylation and methylation. During the prenatal and early postnatal period, the global epigenome is specifically sensitive to environmental stimuli, essentially forming waves of epigenetic plasticity that are concomitant with the critical period of cortical development (Nagy & Turecki, 2012). In schizophrenia, evidence from postmortem studies suggests that early life environmental impact can leave long-lasting maladaptive epigenetic imprints that may contribute to the pathology. For example, several risk genes of schizophrenia are highly regulated during cortical development. Their expression levels typically increase slowly until early adolescence, a period characterized by dynamic changes in promoter-bound DNA methylation and histone modification (Akbarian, 2014). This line of work broadly indicates that adverse events during heightened epigenetic vulnerability could lead to schizophrenia—a hypothesis recently receiving increasing support from animal models. Indeed, maternal care, MAM, and MIA can induce widespread changes in the GABAergic transcriptome (Akbarian, 2014). This type of epigenetic vulnerability likely spans other molecular pathways and developmental stages, and is potentially heritable for generations (Perez et al., 2016). As Gulchina’s work suggests, glutamatergic synapses and the juvenile stage of development could be involved in heightened epigenetic vulnerability as well, a novel research avenue that warrants extensive exploration.

To sum up, the study by Gulchina et al. (2017) provides the first evidence, combined with functional validation, that juvenile MAM rats display prefrontal NMDAR deficits relevant to schizophrenia, further lending validity to this model. More importantly, this study extends the research scope of the MAM model to NMDAR dysfunction at early developmental stages, opening exciting venues for future investigation.

List of abbreviations

MAM

Methylazoxymethanol acetate

NMDA

n-methyl-D-asparate

PFC

prefrontal cortex

PCP

phencyclidine

REST

repressor element 1-silencing transcription factor