The underpinnings of many mental illnesses are multifactorial, arising from a combination of genetic susceptibilities and environmental factors (Tsuang, 2000; Moffitt et al., 2005; van Os et al., 2010). In psychiatric disease, it is therefore essential to develop an integrated understanding of the etiopathology, genetics and resultant modifications to key brain circuitry to aid our future battles to alleviate, or eliminate, patient suffering with novel therapeutic approaches. The research community is fortunate to have a growing methodological armory with which to fight these battles, including technological advances such as next-generation sequencing (Williams et al., 2009), alongside an ever-expanding experimental toolkit for functionally dissecting brain circuitry (Arenkiel & Ehlers, 2009).
The disrupted in schizophrenia 1 (DISC1) gene was originally identified in a unique Scottish pedigree, in which the gene is disrupted by an inherited balanced chromosomal translocation between chromosomes 1 and 11 [t(1;11)(q42.1;q14.3)]. Within this family, segregation is observed between the translocation event and a spectrum of psychiatric disorders, including major depression, schizophrenia, and bipolar disease (Jacobs et al., 1970; St Clair et al., 1990; Millar et al., 2000; Blackwood et al., 2001; Muir et al., 2008; Thomson et al., 2013). The molecular consequences of the translocation are likely to be very complex, as transcripts and abnormal proteins resulting from fusions between DISC1 and a gene on chromosome 11, known as Boymaw or FP1, have recently been identified (Brandon & Sawa, 2011; Eykelenboom et al., 2012). To date, however, the focus has been largely restricted to the consequences of DISC1 truncation alone rather than functional outcomes related to translocation-generated fusion proteins. Expression of DISC1 is also complex, and a striking number of splice forms (> 50) have been described (Nakata et al., 2009; Thomson et al., 2013). Which transcripts are actually translated into proteins is, however, not as yet fully understood.
Full-length human DISC1 encodes an intracellular protein consisting of 854 amino acids with a molecular mass of ~100 kDa. There is good evidence that the protein forms oligomeric species (Brandon et al., 2004; Narayanan et al., 2011), and a self-association domain has been identified biochemically (Kamiya et al., 2005). Furthermore, it has been proposed that assembly of DISC1 into large aggregates may play a role in its disease biology (Korth, 2012). This idea has been supported by the description of DISC1 aggresomes, which are intracellular entities that negatively impact on cellular transport processes (Atkin et al., 2011).
DISC1 expression is highest during central nervous system (CNS) development in both humans and rodents, and gradually decreases during life (Austin et al., 2004; Nakata et al., 2009). In adult mice, expression is seen in a broad range of brain areas, including the olfactory bulb, cortex, hippocampus, hypothalamus, cerebellum, and brainstem (Schurov et al., 2004). In some CNS areas, expression of DISC mRNA is seen only during development, with little if any mRNA being detectable in adulthood. Examples of such brain areas include the bed nucleus of the stria terminalis and the reticular thalamic nucleus (Austin et al., 2004). Although some earlier reports suggested that DISC1 was predominantly expressed in neurons and was largely absent from glia, more recent work has indicated that DISC1 expression may also occur in multiple classes of glial cell in both rodent and human tissue (Seshadri et al., 2010; Kuroda et al., 2011; Ma et al., 2013). Indeed, DISC1 has been implicated in cellular functions of both oligodendrocytes and astrocytes (Katsel et al., 2011; Ma et al., 2013).
As recently discussed elsewhere (Brandon & Sawa, 2011), despite its name, ‘disrupted in schizophrenia 1’, DISC1 may not be a key risk factor for schizophrenia. Instead, DISC1 disruption may confer a genetic risk at the level of endophenotypes or brain circuitry that underlies a number of major mental disorders. Indeed, at the genetic level, risk variants shared between multiple psychiatric disorders, such as schizophrenia, bipolar disorder, and autism, have been identified (Owen et al., 2007, 2011; Cuthbert & Insel, 2010). Important processes from where endophenotypes spanning a range of psychiatric disease might arise include the initial development of the nervous system and its subsequent refinement in the early years of life, and the functionality of synaptic connections within the circuits of the CNS. Both of these factors have been strongly implicated in the biology of DISC1 (Brandon & Sawa, 2011). This short review will specifically consider the latter, namely DISC1's roles in synaptic function, although it must always be borne in mind that certain DISC1-related functional alterations identified at synapses might have their genesis in the prior developmental program of the CNS, rather than reflecting a specific ongoing action of the protein in adult synaptic functionality.
Although much about the biology and pathophysiology of DISC1 still remains elusive, it is well established that its major role is as a scaffold protein, serving to co-locate other important signaling molecules and place them in the necessary cellular locations for them to perform their various roles (Brandon, 2007; Thomson et al., 2013). Within the neuronal population, DISC1 expression occurs in multiple cell types, including both glutamatergic and GABAergic cells. Furthermore, DISC1 can be found in multiple cellular compartments, and appears to perform functions at multiple cellular loci, e.g. the nucleus, the centrosome, the primary cilium, and the mitochondrion (Brandon & Sawa, 2011; Thomson et al., 2013).
In support of roles related to synaptic functionality, DISC1 is highly enriched in postsynaptic density (PSD) fractions, and has been also shown to be present in dendritic spines by the use of ultrastructural methods (Kirkpatrick et al., 2006; Clapcote et al., 2007; Hayashi-Takagi et al., 2010; Carlisle et al., 2011; Wang et al., 2011; Paspalas et al., 2013). Furthermore, characterisation of the DISC ‘interactome’ has identified a number of proteins with known roles in the functional biology of synapses (Camargo et al., 2007). These interacting proteins include molecules that are themselves highly concentrated in PSD fractions, e.g. Traf2 and NcK-interacting kinase (Camargo et al., 2007; Wang et al., 2011).
So how does one set about investigating the potential roles of DISC1 in synaptic function? Unlike workers studying other proteins with pivotal roles in synaptic physiology, including receptors, ion channels, transporters, and enzymes such as kinases; the DISC1 investigator is not much helped by pharmacology. As DISC1 is a scaffold protein, there are no drugs that specifically inhibit or activate DISC1 function. Thus, recording synaptic function and testing the outcome of acutely changing DISC1 function pharmacologically is not currently an approach available to the DISC1 investigator. Instead, most investigations have relied on molecular manipulations that, in some way, modify the amount or type of DISC1 expressed in rodent tissues.
Manipulations used to modify DISC expression include the use of transgenic mice that express various forms of DISC1. These include mice in which various genetic engineering strategies have been employed to express truncation mutants designed to approximate the translocation found in the Scottish pedigree (Hikida et al., 2007; Pletnikov et al., 2008; Shen et al., 2008). In one case, expression of truncated DISC1 is specifically directed to astrocytes (Ma et al., 2013). A mouse has also been produced with a deletion of exons 2 and 3; these exons encode motifs that are important for many key protein–protein interactions, and, consequently, this mouse may have lost many functions of DISC1 (Kuroda et al., 2011). It is also noteworthy that a number of strains of laboratory mice are incapable of making full-length DISC1, owing to a 25-bp deletion in exon 6, first identified in 129S6/SvEv; this results in a string of 13 novel amino acids followed by a premature stop codon within exon 7 (Clapcote & Roder, 2006; Koike et al., 2006; Kvajo et al., 2008; Ritchie & Clapcote, 2013). Crossing these mice into a line that produces full-length DISC1, C57BL/6J, allows comparisons between normal mice and those with truncated DISC1. A mouse that expresses a C-terminal fragment of DISC1 under the control of tamoxifen induction has also been used to probe aspects of DISC1 physiology (Li et al., 2007). In addition, two mouse lines with point mutations in DISC1 (L100P and Q31L) derived from ethyl-nitrosourea treatment (Clapcote et al., 2007) have been investigated in some detail.
In addition to the use of various genetically modified mice, DISC1 expression has been induced or suppressed with chemical transfection or viral transduction of DISC1 expression constructs, or various DISC1-directed RNA interference (RNAi) reagents. A recent methodological approach that has been exploited successfully in studies of DISC1 biology is in utero gene transfer. This has been used to both knock down DISC1 expression and to overexpress normal or truncated DISC1 during early CNS development (Kamiya et al., 2005; Meyer & Morris, 2009; Kubo et al., 2010; Niwa et al., 2010; Tomita et al., 2011; Maher & LoTurco, 2012). Given DISC1's role as a scaffold protein, another molecular approach that has been employed has been to disrupt the interaction between DISC1 and some of its binding partners by overexpressing peptides designed to disrupt specific DISC1 interaction sites (GIRDIN, Traf2 and NcK-interacting kinase, Lis1, NDE1, and FEZ1) (Taya et al., 2007; Enomoto et al., 2009; Kang et al., 2011; Wang et al., 2011).
As described above, a number of studies have indicated that manipulating the levels or nature of expressed DISC1 produces a range of effects on the development of the CNS. Many of these actions are likely to involve DISC1 binding partners, such as NDEL1, FEZ1, and GIRDIN (also known as KIAA1212). The most prevalent observations involve modifications to neurogenesis, neuronal migration, and integration of neurons into the neuronal parenchyma of their final destination. The last of these results in subpopulations of slightly misplaced neurons and alterations in mature neuronal morphology, such as changes in process branching (Kamiya et al., 2005; Duan et al., 2007; Faulkner et al., 2008; Shen et al., 2008; Enomoto et al., 2009; Kim et al., 2009; Kang et al., 2011; Kvajo et al., 2011). Although these effects are clear-cut, on the whole their global consequences are relatively subtle, e.g. slightly enlarged ventricles, a degree of cortical thinning, and small reductions in cell counts of certain neuronal populations – many of these have parallels in the human schizophrenia literature. Certainly, gross brain anatomy is not overtly disturbed in mice that lack full-length DISC1, whether as a result of genetic modification (Hikida et al., 2007; Pletnikov et al., 2008; Shen et al., 2008; Kuroda et al., 2011), cross-breeding of common laboratory strains (Koike et al., 2006), or DISC1 knockdown by in utero delivery of short hairpin RNA (Koike et al., 2006). Consequently, although interfering with DISC1 function produces developmental changes, well-established synaptic pathways still develop and appear to show fundamental aspects of synaptic function.
With regard to evidence for changes in synaptic connectivity, there are certainly reports of a reduction in dendritic spine density when DISC1 is manipulated. For example, in the ethyl-nitrosourea-derived Q31L and L100P point mutation mice, spine density was approximately 15–20% lower in both the hippocampus and the frontal cortex, the latter being paralleled by a decreased number of neurons and a propensity for the neuronal population to be located slightly deeper in the cortex (Clapcote et al., 2007). As well as changes to spine numbers, manipulating DISC1 has significant effects on spine size in cultured neurons. Short-term knockdown of DISC1 (2 days) increases both the number of spines and their size, whereas continuing inhibition of DISC1 expression up to 6 days results in fewer and smaller spines (Hayashi-Takagi et al., 2010). This latter observation also appears to translate to the in vivo situation when small interfering RNA (siRNA) is injected into the medial prefrontal cortex of young rats (Hayashi-Takagi et al., 2010). Additional evidence suggests that this effect on spine size is mediated by limiting the actions of Kalirin7 on Rac1 (Hayashi-Takagi et al., 2010).
Electrophysiological recording still remains the gold standard methodology for investigating alterations in synaptic function and related network activities. So far, there have been a relatively small number of neurophysiological investigations of the functional impacts of manipulating DISC1. These data frequently appear as fractional parts of studies in which other experimental approaches are also employed. Certainly, so far, the neurophysiological literature related to DISC1 is in its infancy as compared with the in-depth work performed on many other models of CNS disease. For example, we are not aware of any published studies of DISC1-related electrophysiological work performed in vivo.
The reported concentration of DISC1 expression in the adult PSD fraction (Kirkpatrick et al., 2006; Clapcote et al., 2007; Hayashi-Takagi et al., 2010; Carlisle et al., 2011; Wang et al., 2011), along with the known functions of some of the proteins for which it acts as a scaffold (Camargo et al., 2007; Brandon & Sawa, 2011), have led most workers so far to focus their electrophysiological investigations on fast excitatory signaling mediated by postsynaptic ionotropic glutamate receptors. All such work to date has focused on recordings made in hippocampal or cortical neurons, either in brain slices or in primary cultures. Notably, the limited number of electrophysiological studies performed to date have employed tissue derived from a number of different models, each of which has different underlying genetics. For this reason, it is frequently hard to compare findings between groups, which can, in some cases, at first sight appear contradictory. We believe that, as more studies are performed in the future, the key DISC1-related neurophysiological phenotypes will become clearer, and the work performed to date will provide useful indicators as to what should be examined.