MicroRNAs as the cause of schizophrenia in 22q11.2 deletion carriers, and possible implications for idiopathic disease: a mini-review

The 22q11.2 deletion is the strongest known genetic risk factor for schizophrenia. Research has implicated microRNA-mediated dysregulation in 22q11.2 deletion syndrome (22q11.2DS) schizophrenia-risk. Primary candidate genes are DGCR8 (DiGeorge syndrome critical region gene 8), which encodes a component of the microprocessor complex essential for microRNA biogenesis, and MIR185, which encodes microRNA 185. Mouse models of 22q11.2DS have demonstrated alterations in brain microRNA biogenesis, and that DGCR8 haploinsufficiency may contribute to these alterations, e.g., via down-regulation of a specific microRNA subset. miR-185 was the top-scoring down-regulated microRNA in both the prefrontal cortex and the hippocampus, brain areas which are the key foci of schizophrenia research. This reduction in miR-185 expression contributed to dendritic and spine development deficits in hippocampal neurons. In addition, miR-185 has two validated targets (RhoA, Cdc42), both of which have been associated with altered expression levels in schizophrenia. These combined data support the involvement of miR-185 and its down-stream pathways in schizophrenia. This review summarizes evidence implicating microRNA-mediated dysregulation in schizophrenia in both 22q11.2DS-related and idiopathic cases.


INTRODUCTION
The 22q11.2 deletion syndrome (22q11.2DS), also known as the velocardiofacial/DiGeorge syndrome, is a phenotypically heterogenous disease which is caused by a hemizygous microdeletion on the long arm of chromosome 22 in the region q11.2. The overall prevalence is 1 in 2,000-4,000 live births (Murphy et al., 1999;Botto et al., 2003;Robin and Shprintzen, 2005). The disorder is associated with a high risk for psychiatric disorder.
In particular, 22q11.2DS patients have an estimated 20-25% risk for schizophrenia or related psychotic disorders such as schizoaffective disorder (Murphy et al., 1999;Chow et al., 2006;Bassett and Chow, 2008;Philip and Bassett, 2011). The deletion is therefore the strongest known genetic risk factor for schizophrenia (odds ratio = 20.3; Levinson et al., 2011), and accounts for approximately 1-2% of all schizophrenia cases (Karayiorgou et al., 1995Bassett and Chow, 2008;International Schizophrenia Consortium, 2008;Stefansson et al., 2008). Individuals with 22q11.2DS have variable cognitive and behavioral deficits  including relative impairments in social judgment, motor skills, verbal learning, and executive functioning (Chow et al., 2006;Philip and Bassett, 2011). In addition, adults with a 22q11.2 microdeletion have a two-to threefold increase in the risk of generalized anxiety disorder compared to the general population (Philip and Bassett, 2011). The major clinical features of 22q11.2DS-related schizophrenia are largely indistinguishable from those of the idiopathic disease (Murphy et al., 1999;Chow et al., 2006;Bassett and Chow, 2008). Identification of schizophrenia-risk gene/s in the 22q11.2DS deletion region may therefore generate insights into the pathophysiology of schizophrenia in general (Earls et al., 2012).
The size of the 22q11.2 deletion varies. The majority of 22q11.2 deletions (around 90%) are 3 Mb in size and span approximately 60 known genes, while the remaining 10% are 1.5 Mb in size and encompass around 35 genes (Edelmann et al., 1999;Shaikh et al., 2000). Both the larger and the smaller 22q11.2 microdeletions usually result from non-allelic homologous recombination, which is mediated by flanking low-copy repeats (Edelmann et al., 1999). Although the 22q11.2DS phenotype is highly variable, its severity is not correlated with the size of the deletion. This suggests that the minimal 1.5 Mb deletion region is crucial in terms of etiology (Carlson et al., 1997;Karayiorgou et al., 2010).
Initial research to identify schizophrenia-risk genes in the 22q11.2 deletion region proved unsuccessful. The identification of heterozygous loss-of-function mutations in non-deleted schizophrenia patients would be the most obvious human genetic evidence for the involvement of a specific gene in disease susceptibility. Hopes were raised by the identification of heterozygous point mutations in the T-box 1 gene (TBX1), which encodes a T-box transcription factor, that resulted in the characteristic abnormal facies and cardiac defects of 22q11.2DS in patients without a 22q11.2 deletion (Yagi et al., 2003;Zweier et al., 2007). However, no such mutation has yet been identified in schizophrenia patients.
In contrast, recent research in the Df(16)A +/− mouse model has generated breakthroughs in our understanding of the underlying biological mechanisms of 22q11.2DS schizophrenia-risk. This engineered mouse strain carries a heterozygous chromosomal deletion which spans a segment syntenic to the human 22q11.2 locus. Df(16)A +/− mice show deficits in the synaptic connectivity of hippocampal neurons, including a lower density of dendritic spines and glutamatergic synapses (Mukai et al., 2008). In addition, Df(16)A +/− mice display hyperactive behavior and deficits in spatial working memory-dependent learning . Further characterization of this animal model has suggested that the 22q11.2 microdeletion results in alterations in the biogenesis of brain microRNAs Xu et al., 2010). Primary candidate genes in the region are the DiGeorge syndrome critical region gene 8 (DGCR8), which encodes a component of the microprocessor complex essential for microRNA biogenesis (Tomari and Zamore, 2005), and the MIR185 gene , which encodes microRNA 185. Both genes are located within the minimal 1.5 Mb deletion region at 22q11.2 .
The microRNAs are a class of 21-25-nucleotide small non-coding RNAs. They control the expression of their target genes by binding to target sites in messenger RNAs (mRNAs), typically in their 3 untranslated regions (He and Hannon, 2004;Meola et al., 2009). In most cases, microRNAs negatively regulate target gene expression through a combination of repression of mRNA translation and promotion of mRNA decay. Each microRNA usually controls up to several hundred target mRNAs, while one mRNA target can be synergistically regulated by multiple microR-NAs (Sathyan et al., 2007;Didiano and Hobert, 2008;Drew et al., 2011). This allows microRNAs to integrate different intracellular signals and to regulate various signaling pathways (Johnston and Hobert, 2003;Choi et al., 2007). Accumulating evidence suggests that microRNAs contribute to the basic mechanisms underlying brain development and plasticity ( Table 1; Fineberg et al., 2009;Schratt, 2009;Im and Kenny, 2012). Neural microRNAs play an important role at various stages of synaptic development, including dendritic arborization (Vo et al., 2005;Yu et al., 2008), synapse formation, and synapse maturation (Caygill and Johnston, 2008;Siegel et al., 2009). Arguably the two most extensively studied examples in the context of synapse development are miR-132 and miR-134. CREB-induced miR-132 promotes dendritogenesis and spine growth by down-regulating p250GAP (Wayman et al., 2008;Magill et al., 2010). miR-134 on the other hand is required for activity-dependent dendritic arborization and the restriction of spine growth by targeting Pumilio-2 and Lim-domain containing protein kinase (Limk1), respectively (Schratt et al., 2006;Fiore et al., 2009). Furthermore, investigation of a mouse model displaying conditional knock-out of the microRNA biogenesis enzyme Dicer (Schratt, 2009) revealed disrupted morphogenesis of the hippocampus and cortex (Davis et al., 2008), suggesting that undisturbed microRNA processing might be necessary for normal brain development (Xu et al., 2010). These data suggest the possible involvement of microRNA-dependent dysregulation in the pathogenesis of various psychiatric disorders (Forero et al., 2010;Xu et al., 2010), including schizophrenia (Beveridge et al., 2008).
The present review summarizes the various lines of evidence implicating microRNAs as the causal factor for schizophrenia in 22q11.2DS carriers and emerging evidence from expression studies and genome-wide association studies (GWAS) that these mechanisms may also be involved in the development of idiopathic schizophrenia.

THE ROLE OF DGCR8
Investigation of Dgcr8 +/− mice confirmed that heterozygous Dgcr8 deficiency was responsible for the reduced biogenesis of microRNAs observed in Df(16)A +/− mice Schofield et al., 2011). Dgcr8 +/− mice displayed 22q11.2DSassociated cognitive and behavioral deficits, and altered short-term plasticity in the prefrontal cortex (PFC; Stark et al., 2008). This indicates that DGCR8 heterozygosity, and the resulting alterations in microRNA expression, are sufficient to produce some of the neural deficits observed in 22q11.2DS (Schofield et al., 2011). On the neuronal cell level, Dgcr8 deficiency resulted in structural changes in dendritic spines and reduced dendritic complexity in the hippocampus . Schofield et al. (2011) identified alterations in the electrical properties of layer V pyramidal neurons in the medial PFC of Dgcr8 +/− mice, as well as a decrease in the complexity of the basal dendrites and reduced excitatory synaptic transmission. These functional results suggest that precise microRNA expression is critical for the development of PFC circuitry (Schofield et al., 2011), circuitry which has been reported to be altered in schizophrenia patients (Ursu et al., 2011).
Dgcr8 +/− mice also displayed a decrease in the number of cortical neurons, structural deficits in dendritic spines in the PFC, and alterations in synaptic potentiation and short-term plasticity (Fenelon et al., 2011). These alterations might influence functional connectivity (Schreiner et al., 2013), and could be implicated in the observed cognitive and behavioral deficits. In particular, they may explain observed alterations in prepulse inhibition , which have also been reported in schizophrenia patients (Powell et al., 2009). Ouchi et al. (2013) showed that heterozygous Dgcr8 deficiency in mice led to reduced progenitor cell proliferation and neurogenesis in the adult hippocampus. This is of particular interest since alterations in the anatomy, histology, and function of the hippocampus have been consistently reported in schizophrenia patients (Tamminga et al., 2010). Several schizophrenia-associated genes were down-regulated in the hippocampus of Dgcr8 +/− mice (Ouchi et al., 2013), including the insulin-like growth factor 2 (IGF2), which was recently found to play a crucial role in hippocampal functions such as memory consolidation and fear extinction (Agis-Balboa et al., Chen et al., 2011a). Interestingly, restoration of IGF2 expression in the hippocampus rescued the observed spatial working memory deficits in Dgcr8 +/− mice, suggesting that IGF2 contributes -at least in part -to the learning and spatial working memory deficits that are associated with 22q11.2DS-related schizophrenia (Ouchi et al., 2013).

Frontiers in Molecular Neuroscience
www.frontiersin.org     The question now arises as to which specific microRNAs are regulated by DGCR8. The investigation of Dgcr8 +/− mice identified 59 down-regulated microRNAs in the PFC and 30 downregulated microRNAs in the hippocampus . These down-regulated microRNAs include miR-185, which is also located in the minimal 1.5 Mb deletion region at 22q11.2.

THE ROLE OF MIR185
Studies of 22q11.2DS mouse models have identified miR-185 as the top-scoring down-regulated microRNA in schizophreniaassociated brain areas Benetti et al., 2009). A recent study by Xu et al. (2013) confirmed the drastic reduction in miR-185 expression levels in the hippocampus and PFC of Df(16)A +/− mice, and showed that this reduction contributed to deficits in dendritic complexity and spine development in hippocampal neurons. In addition, Dgcr8 deficiency resulted in an approximately 20% reduction in miR-185 expression in the hippocampus (Xu et al., 2013). This suggests that the pronounced reduction of miR-185 expression in Df(16)A +/− mice -a reduction which is much more pronounced than would be expected by the 50% decrease in gene dosage -may be due to the combined effect of the hemizygosity of the MIR185 gene and the impaired maturation of the pri-miR-185 transcript secondary to reduced Dgcr8 levels (Figure 1; Xu et al., 2013). The large reduction in miR-185 expression renders miR-185 unique among the genes that are affected by the 22q11.2 microdeletion (Xu et al., 2013).
A recent human study confirmed a down-regulation of MIR185 expression to 0.4× normal levels in the peripheral blood of patients with 22q11.2DS (de la Morena et al., 2013). This finding suggests that pronounced miR-185 down-regulation also occurs in patients with 22q11.2DS.
Previous research has shown that MIR185 is present or enriched in synapses (Lugli et al., 2008;Earls et al., 2012). This may indicate that MIR185 is of relevance to neural function, since a number of microRNAs have been shown to play a critical role in synaptic plasticity (Schratt, 2009).

Frontiers in Molecular Neuroscience
www.frontiersin.org FIGURE 1 | Dysregulation of microRNA biogenesis in 22q11.2DS animal models. The MIR185 and the DGCR8 genes are located within the minimal 1.5 Mb microdeletion region on chromosome 22q11.2 and the equivalent region of mouse chromosome 16. The microdeletion leads to a hemizygosity of MIR185 and DGCR8. The heterozygous Dgcr8 deficiency is responsible for the reduced biogenesis of a specific subset of microRNAs (red) observed in Df(16)A +/− mice . These down-regulated microRNAs include miR-185 (black). The pronounced reduction of miR-185 expression (indicated by a broader arrow) may be due to a combined effect of the hemizygosity of the MIR185 gene and the impaired maturation of the pri-miR-185 transcript secondary to reduced Dgcr8 levels (Xu et al., 2013). The resulting alterations in mature microRNA expression levels may lead to altered gene expression of target genes, which might produce some of the neural, cognitive, and behavioral deficits observed in 22q11.2DS. miRISC, microRNA-induced silencing complex. Earls et al. (2012) identified MIR185 as a regulator of sarco(endo)plasmic reticulum Ca(2+) ATPase (SERCA2) which maintains Ca 2+ levels in the endoplasmatic reticulum. The depletion of MIR185 contributes to SERCA2 upregulation and has been proposed as a mechanism leading to abnormal hippocampal synaptic plasticity in 22q11.2DS mouse models. The microRNA regulation of SERCA2 translation may also be implicated in the elevation of SERCA2 protein observed in the post-mortem brains of idiopathic schizophrenia patients (Earls et al., 2012). These results suggest that microRNA-mediated SERCA2 upregulation at central synapses might be a mechanistic link between 22q11.2DS and idiopathic schizophrenia (Earls et al., 2012).
Further support for the involvement of MIR185 in schizophrenia is provided by findings that two of its validated targets (RhoA, Cdc42; Liu et al., 2011) are associated with altered expression levels in schizophrenia (Hill et al., 2006;Ide and Lewis, 2010). Cdc42 (cell division cycle 42) is a member of the RhoGTPase family (Hill et al., 2006) and promotes dendritic spine formation (Irie and Yamaguchi, 2002;Tada and Sheng, 2006;Wegner et al., 2008) by regulating the polymerization of the actin cytoskeleton into filopodia (Nobes and Hall, 1995). Cdc42 is activated by Collybistin/ARHGEF9 (Reid et al., 1999;Reddy-Alla et al., 2010), which has recently been identified as a candidate blood biomarker in psychosis (Kurian et al., 2011). RhoA (Ras homologous member A) also belongs to the RhoGTPase family and regulates the destabilization of the actin cytoskeleton (Hill et al., 2006). The activation of RhoA leads to a reduction in the number of dendritic branches and the density of dendritic spines (Nakayama et al., 2000;Hill et al., 2006).

EXPRESSION OF microRNAs IN IDIOPATHIC SCHIZOPHRENIA
Post-mortem studies of human brain tissue have revealed alterations in microRNA expression in patients with schizophrenia. This research is reviewed elsewhere (Beveridge and Cairns, 2012). Briefly, numerous microRNAs have been implicated in the disorder across multiple studies, including 16 microRNAs with increased and 11 microRNAs with decreased expression (Beveridge and Cairns, 2012). Of particular interest in the context of the present review is the study by Moreau et al. (2011). This reported a significant overlap between microR-NAs dysregulated in human post-mortem brain tissue and microRNAs previously found to be altered in the PFC of a 22q11.2DS mouse model Moreau et al., 2011). This finding supports the hypothesis that findings in 22q11.2DS might be of relevance to idiopathic schizophrenia (Brzustowicz and Bassett, 2012).

Frontiers in Molecular Neuroscience
www.frontiersin.org

GWAS OF IDIOPATHIC SCHIZOPHRENIA
The involvement of microRNA-dependent dysregulation in schizophrenia is supported by the results of the large GWAS of schizophrenia performed by the Schizophrenia Psychiatric Genome-Wide Association Study (GWAS) Consortium (2011). In total 17,836 patients and 33,859 controls were investigated. A single-nucleotide polymorphism (SNP) in an intron of MIR137 was the second strongest finding (odds ratio = 1.12). Four other loci with genome-wide significance in this study contained predicted targets of MIR137 (TCF4, CACNA1C, CSMD1, C10orf26). All four genes have recently been validated as miR-137 targets (Kwon et al., 2013). The miRanda database lists 5,487 genes as targets of miR-137 (John et al., 2004). Interestingly, ZNF804A is listed as a validated target (Kim et al., 2012a). This gene has shown strong association with schizophrenia in previous studies (O'Donovan et al., 2008;Williams et al., 2011). Other promising targets include the ubiquitin ligase Mind bomb one (Mib1; Smrt et al., 2010) which plays an important role in neurogenesis and neurodevelopment (Itoh et al., 2003;Choe et al., 2007;Ossipova et al., 2009).
Research in post-mortem brain samples suggests that the functional effect of the miR-137 risk allele may result in a reduced miR-137 expression (Guella et al., 2013). Further down-stream this may be responsible for the reduced white matter integrity, smaller hippocampi, and larger lateral ventricles observed in schizophrenia patients with the miR-137 risk genotype (Lett et al., 2013).

CONCLUSION AND OUTLOOK
Strong evidence suggests that microRNA dysregulation is implicated in the development of schizophrenia in 22q11.2DS patients. This is consistent with the growing recognition of microRNAs as important regulators of gene expression. As microRNAs integrate different intracellular signals and regulate various signaling pathways (Johnston and Hobert, 2003;Choi et al., 2007), the dysregulation of specific microRNAs could lead to the heterogenous phenotype observed in 22q11.2DS.
As summarized above, emerging evidence from expression and genetic analyses suggests that the same microRNA-regulated pathways may also play a role in idiopathic schizophrenia. However, despite a number of systematic investigations of genes in the 22q11.2DS region and the ever increasing number of GWAS data sets Sullivan et al., 2012), no genetic study to date has identified common variation in DGCR8 or MIR185 as a risk factor for schizophrenia. This may simply reflect a lack of common functional variants at these loci. This hypothesis is supported by a recent study of genetic regulation of microRNA expression (Gamazon et al., 2012). Gamazon et al. (2012) systematically investigated the relationship between microRNA expression levels (as quantitative traits) and common genetic variation. In this study, no SNP had significant cis effects on miR-185 expression. A small number of SNPs have been reported to have significant cis effects on DGCR8 expression in human monocytes (Zeller et al., 2010), and fibroblasts (Dimas et al., 2009). However, these associations might be tissue-specific, since a recent study of five different human post-mortem brain regions failed to identify any SNP with significant cis effects on DGCR8 expression (Kim et al., 2012b).
A challenge for future research will be to identify and validate the target genes that are affected by microRNA dysregulation and their respective pathways in a more comprehensive manner . Such research will improve our understanding of how alterations in microRNA-regulated genetic networks contribute to the pathophysiology of both 22q11.2DS-related and idiopathic schizophrenia.
Idiopathic schizophrenia is a multifactorial disorder for which both genetic and environmental factors exert an impact on disease susceptibility (Sawa and Snyder, 2002). However, very few data are available concerning the influence of environmental factors on microRNA dysregulation. Recent studies in mice showed that environmental factors such as stress resulted in alterations of microRNA expression in the frontal cortex (Rinaldi et al., 2010). Future studies are therefore warranted to investigate the extent to which environmental factors are associated with microRNA dysregulation in schizophrenia.
Further research into the precise role of microRNAs in schizophrenia is important clinically, since modification of microRNA dysregulation would represent a novel therapeutic approach to this devastating and chronic disease. MicroRNAs are excellent candidates for therapy since they regulate multiple targets in various signaling pathways, thereby minimizing the risk of resistance development or compensatory mechanisms (Soriano et al., 2013). This view is supported by several recent studies and reviews, which have highlighted microRNAs as promising pharmacological targets in the treatment of complex diseases such as psychiatric disorders (Im and Kenny, 2012), cancer (Soriano et al., 2013), and diabetes (Mao et al., 2013).