Concise Review: the Promise of Human Induced Pluripotent Stem Cell-Based Studies of Schizophrenia

EMBRYONIC STEM CELLS/INDUCED PLURIPOTENT STEM CELLS

Concise Review: The Promise of Human Induced Pluripotent Stem Cell-Based Studies of Schizophrenia

KRISTEN J. BRENNAND,FRED H. GAGE Salk Institute for Biological Studies, Laboratory of Genetics, La Jolla, CA, USA

Key Words. Schizophrenia • Stem Cells • Neurons • Pharmacology • Genetics

ABSTRACT Schizophrenia (SCZD) is a heritable developmental disor- abnormalities and anatomical perturbations consistent with der. Although the molecular mechanism of disease remains human disease. The advent of human induced pluripotent unclear, insights into the disorder have been made through stem cells (hiPSCs) makes it possible to study SCZD using a vast array of experimental techniques. Together, mag- live human neurons with a genetic predisposition toward netic resonance brain imaging, pharmacological, and post- SCZD, even without knowledge of the genes interacting to mortem pathological studies have observed decreased brain produce the disease state. SCZD hiPSC neurons show cellu- volume, aberrant neurotransmitter signaling, reduced den- lar defects comparable to those identified in post-mortem dritic arborization, and impaired myelination in SCZD. Ge- human and mouse studies, and gene expression changes are nome-wide association studies (GWAS) have identified consistent with predictions made by GWAS. SCZD hiPSC common single nucleotide polymorphisms as well as rare neurons represent a new tool to look beyond phenotype copy number variants that contribute to SCZD, while and begin to dissect the molecular mechanisms of SCZD. mouse models of candidate SCZD genes show behavioral STEM CELLS 2011;29:1915–1922 Disclosure of potential conflicts of interest is found at the end of this article.

followed by hiPSC differentiation to neurons, produces a near INTRODUCTION limitless source of live human neurons, genetically identical to those present in patients, with which this disorder can be Schizophrenia (SCZD) is a neurological disorder characterized studied. This review will summarize the major findings of by three severe classes of symptoms: positive (hallucinations human and mouse studies of SCZD and compare these to and delusions), negative (inability to speak, express emotion, early reports of SCZD hiPSC neurons. or find pleasure), and cognitive (deficits in attention, memory, and planning) [1, 2]. The current treatment regimen for SCZD involves chronic treatment with powerful antipsychotics, the strong unpleasant side effects of which often result in cessa- INSIGHTS FROM HUMAN STUDIES tion of treatment [1]. The life expectancy of patients with SCZD is up to 10 years less than the general population [3]; Human studies to date have used three major approaches: 40% of schizophrenics suffer from substance abuse [4], 20% brain imaging, post-mortem pathology, and genome-wide are homeless [5], and 10% ultimately commit suicide [6, 7]. association studies (GWAS). It is estimated that 1.1% of the global population over 18 years of age has SCZD, including 3 million Americans [1]. Observations from Brain Imaging Human studies of SCZD have relied primarily on brain Magnetic resonance imaging (MRI) can be used to estimate imaging, pharmacology, post-mortem pathology, and genetic the volume of various brain regions, although these studies studies of patient lymphocytes. Animal studies are limited in must be interpreted in the context of the high variability of two important ways: (a) they do not reflect the complex brain measures across individuals [10]. Although group-aver- genetic interactions that result in the vast majority of cases of age differences exist between SCZD patients and controls, SCZD, and (b) it is difficult to evaluate classic symptoms of anatomic MRI differences are not adequate for diagnosis. For SCZD such as hallucinations, delusions, and disorganized analogy, although men tend to be taller than women on aver- speech in mice. Although many insights have been made age, height is not sufficient to determine sex; male/female through these methods, the molecular and cellular defects that height differences are twice as predictive of gender as most contribute to disease initiation and progression in neurons MRI neuroimaging experiments are predictive of a diagnosis remain unknown. Recently, a third approach has been of SCZD. described to study SCZD [8, 9]. Reprogramming of patient In first-episode patients with SCZD, MRI studies consis- fibroblasts to human induced pluripotent stem cells (hiPSCs), tently observe decreased whole brain volume (and increased

Author contributions: K.J.B. and F.H.G.: wrote the review together. Correspondence: Fred H. Gage, Ph.D., Salk Institute for Biological Studies, Laboratory of Genetics, 10010 North Torrey Pines Road, La Jolla, CA 92037, USA. Telephone: 858-453-4100; Fax: 858-597-0824;; e-mail: [email protected] Received August 5, 2011; accepted for publication September 25, 2011; ﬁrst published online in STEM CELLS EXPRESS October 18, 2011. VC AlphaMed Press 1066-5099/2011/ $30.00/0 doi: 10.1002/stem.762 STEM CELLS 2011;29:1915–1922 www.StemCells.com 1916 hiPSC-Based Studies of Schizophrenia

Figure 1. Average high-resolution magnetic resonance images (MRI scans) showing gray matter loss in adolescents with schizophrenia. Severe loss is observed (red and pink; up to 5% annually) in parietal, motor, and temporal cortices, whereas inferior frontal cortices remain stable (blue; 0–1% loss). Adapted from Ref. [15].

campus [20] while during working memory tasks, SCZD patients show reduced activation of the cortex [21]. The molecular mechanism of these changes has not been explained. Although brain imaging has associated clinical symptoms to the general brain areas affected, the relationship between brain imaging, cellular pathology, and the molecular mechanism of SCZD remains unknown.

Observations from Pharmacological Studies The accidental discovery that Chlorpromazine (Thorazine) reduces psychotic symptoms, likely by functioning as an antagonist of dopamine (DA) receptors, provided the basis for Figure 2. Weekly change in clinical assessment of psychosis, measured the ‘‘DA hypothesis’’ of SCZD. This hypothesis proposed that using the clinical PANSS. Adapted from Ref. [30]. Abbreviation: PANSS: excessive activation of D2 receptors was the cause of (the positive and negative symptom scale. **p < 0.01; ***p < 0.001. positive symptoms of) SCZD. By positron emission tomogra- phy (PET) imaging, researchers determine where given biological molecules, such as DA, bind in the brain. Such work ventricular volume) [12–14]. Specifically, the greatest has correlated DA receptor levels with the positive symptoms decreases are observed in the gray matter of the hippocampus, of SCZD (psychosis and delusions) [22], and strong evidence basal ganglia, and thalamus (Fig. 1) [14–16]. In chronically ill now links SCZD with increased DA synthesis, DA release, patients, the volume loss is most pronounced in the frontal and resting-state synaptic DA concentrations [23]. Further and temporal gray matter areas of the cortex [17], but whether support of this hypothesis is the psychosis-inducing effects of these changes result from disease or long-term treatment with DA receptor agonists such as amphetamine. antipsychotic medications is unclear. Longitudinal studies The DA hypothesis is now believed to be overly simplis- have observed progressive decrease in brain volume and tic. Among the DA antagonists, Clozapine is uniquely effec- increase in lateral ventricle volume for at least 20 years after tive for treatment-resistant SCZD [24]. This ability may occur the onset of symptoms. through a DA-independent mechanism, a hypothesis supported White matter is produced when oligodendrocytes wrap by anecdotal evidence that Clozapine effectively treats the axons in sheaths of myelin, which increases the speed of neu- psychotic symptoms associated with Parkinson’s disease (PD) rotransmission as well as the timing and synchrony of neuronal without exacerbating the tremors, rigidity, and bradykinesis firing patterns [18]. Beyond the well-characterized changes in caused by the loss of DA neurons in PD [25]. Clozapine is an gray matter, some studies also identify changes in the white antagonist of D1, D2, D3, and D4 DA receptors, a1- and a2- matter of the brain in SCZD, particularly in the prefrontal and adrenergic receptors, 5-hydroxytryptamine (5-HT) (5-HT2) se- temporal lobes and corpus collusum [19]. Longitudinal brain rotonin receptor, H1 histamine receptor, and M1, M2, M3, and imaging studies of childhood onset cases of SCZD have M5 receptors; an agonist of the M4 muscarinic receptor [26]; a observed both progressive loss of cortical gray matter during modulator of glutamatergic neurotransmission [27]; and an in- adolescence and delayed white matter development [11]. hibitor of GABAA receptor neurotransmission [28]; its com- Changes in blood flow and blood oxygenation in the brain plex pharmacological properties hint at the roles of others neu- are closely linked to neural activity. Functional MRI (fMRI) ronal cell types in SCZD. measures the change in blood flow as an indicator of brain ac- Glutamate (GLU)-blocking drugs such as ketamine induce tivity, and fMRI comparisons of control and SCZD patients many of the symptoms, including hallucinations and cognitive have revealed brain activity changes in the dorsolateral pre- deficits, associated with SCZD [29], whereas the GLU recep- frontal cortex. At rest, SCZD patients show cortical hyperac- tor (GLUR2/3) agonist LY2140023 has been shown to ameli- tivity and hyperconnectivity between the cortex and hippo- orate the symptoms of SCZD in recent clinical trials by Eli Brennand and Gage 1917

Lilly [30] (Fig. 2). The authors suggest that LY2140023 may Although studies of post-mortem brain tissue have consis- work by reducing the presynaptic release of GLU at limbic tently failed to detect alterations in DISC1 levels in typical synapses where these receptors are expressed. In summary, SCZD patients, they do observe increased NRG1 in the hippo- pharmacology of SCZD is intricate but evidence supports a campus [48] and prefrontal cortex [49, 50] and decreased role for altered DA and GLU neurotransmitter activity. expression of its receptor ERBB4 [51–53], suggesting that NRG1 and ERBB4 may be affected even in the absence of Observations from Postmortem Studies detectable genetic lesions. Although brain imaging identified decreased brain volume in Common polygenic variation has been shown to contrib- SCZD, post-mortem studies of brains from patients with ute to SCZD. Studies of single nucleotide polymorphisms SCZD have revealed no widespread neuronal loss or even a (SNPs) have estimated that thousands of alleles of very small glial response to a potential neuronal injury. Instead, patho- effect account for nearly 30% of the genetic variance of logical studies have observed three major changes in SCZD SCZD [54–56]. Much of this common polygenic variation brain tissue: increased density of pyramidal neurons, aberra- encompasses the major histocompatibility complex (MHC) tions in interneurons, and decreased oligodendrocytes. region at 6p22-p21 that contains over 200 genes. Although In the cortex, pathological changes are consistent with MHC genes have been implicated in immune diseases such as decreased neuronal connectivity in SCZD. Post-mortem studies type I diabetes, multiple sclerosis, Crohn’s disease, and rheu- have observed an increased density of cortical pyramidal neurons matoid arthritis, they also contribute to synaptic maturation without changes in cell number [31, 32]. Neuronal soma are [57–59]; therefore, this genetic evidence should not be pre- smaller [32], dendrites are shorter with reduced arborizations sumed to be proof of immune abnormalities in SCZD. [33], and there is reduced dendritic spine density [34, 35]. The Other common variants now well-implicated in SCZD disturbances in prefrontal cognitive functioning in SCZD may be include transcription factor 4 (TCF4) on chromosome 18q21, mediated by a process which involves atrophy of neuronal proc- zinc finger protein 804A (ZNF804A) on 2q32.1, and neuro- esses or synapses but stops short of actual neuronal loss [31]. granin (NRGN) on 11q24.2 [55, 60]. TCF4 is a neuronal tran- Changes in RNA and protein levels of a number of key scription factor essential for neurogenesis [61], and NRGN genes involved in neurotransmission have been observed in encodes a postsynaptic protein kinase substrate that binds post-mortem SCZD brain tissue. GLUR expression is altered calmodulin and is enriched in CA1 pyramidal neurons in the in SCZD; expression is decreased in the hippocampus and hippocampus [62], and ZNF804A is associated with altered increased in the cortex [36]. Specifically in GABAergic inter- neuronal connectivity in the dorsolateral prefrontal cortex [63]. neurons, one finds decreased GAD67 and calcium-binding pro- CNVs are large deletions or duplications in the genome. teins in parvalbumin- and calbindin-positive GABAergic neu- So far, only rare (<1% of cases) and large CNVs (>100 kb) rons. Changes in parvalbumin-positive GABAergic neurons have been shown to confer high risk of SCZD. SCZD CNVs are particularly relevant as they are thought to produce gamma are highly penetrant and account for up to 20% of SCZD oscillations, which synchronize pyramidal neuron firing, an ac- cases. CNVs identified to date disrupt genes, such as ERBB4 tivity that is impaired in SCZD. It is unclear whether decreased [64] or NRXN1 [65], or regions, including 1q21.1, 15q11.2, GABAergic inhibitory activity is a cell-autonomous cause of 15q13.3, 16p11.2, 22q11.2 [64, 66-68]. By comparing the SCZD or if it results from decreased glutamatergic input on DNA of both parents to the SCZD patient (proband), it was GABAergic inhibitory neurons in the disease state. shown that CNV mutations frequently occur de novo and are Finally, a number of studies have observed both fewer oli- not inherited from either parent. Given that the CNV mutation godendrocytes and decreased expression of myelin genes in rate far exceeds the rate for nucleotide substitutions [69] and post-mortem SCZD brains. This finding may be consistent that current CNV assays detect only the largest CNVs, consti- with decreases in white matter observed in SCZD. tuting just 5% of total CNVs, researchers may be underesti- mating the contribution of CNVs to SCZD [70]. GWAS of SCZD The genetic basis of SCZD does not necessarily conform A complex genetic psychiatric disorder, SCZD has a large to classical disease boundaries. Many CNVs that confer high inherited component with an estimated heritability of 80–85% risk of SCZD have been implicated in autism, mental retarda- [37, 38]. With tens of thousands of SCZD patients genotyped tion, and epilepsy, while many SNPs associated with SCZD to date, it is now widely accepted that the complicated herit- are shared with bipolar disorder [64, 66-68]. It is likely that ability of SCZD results from polygenic inheritance of a com- numerous disruptions in any number of key neurodevelop- bination of inherited common polymorphisms and both inher- mental pathways may be sufficient to produce a diseased state ited and de novo rare copy number variations (CNVs). that could ultimately manifest as SCZD. Because a large number of markers collectively account for risk of SCZD, the risk of each one is so small that it can be difficult to detect by GWAS. To date, most of the heritable variance of SCZD remains unaccounted for. INSIGHTS FROM MOUSE MODELS Early genetic studies of SCZD focused on traditional pedi- gree analyses. In a family identified in northern Scotland, Although causal mutations for SCZD have not been identified, more than half of the members suffer from mental illness, several genetic mouse models of SCZD have been developed. generally SCZD, and it was found that a balanced chromo- Of these, mice with decreased activity of DISC1, NRG1, somal translocation (1:11) segregated with disease. The dis- ERBB4, or the 22q11 genes show behavioral abnormalities rupted gene on chromosome 1 was subsequently termed Dis- and anatomical perturbations that may be relevant to SCZD. rupted-in-Schizophrenia-1 (DISC1) [39]. The genetic association between Neuregulin-1 (NRG1) and SCZD was DISC1 Mice Recapitulate Aspects of the SCZD first identified in association studies of Icelandic families [40] Phenotype and subsequently confirmed in follow-up studies in multiple DISC1-mutant animals demonstrate behavioral abnormalities populations in Scotland, Ireland, Netherlands, Taiwan, Korea, such as decreased learning and memory [71–73], decreased and China [41–47]. Mutations in DISC1 are much more pene- sociability [71, 72], depression [71, 72], hyperactivity [72, 73], trant that those in NRG1 but are also exceedingly rare. and aggression [72], which are consistent SCZD. Mice with www.StemCells.com 1918 hiPSC-Based Studies of Schizophrenia

size, and activity of excitatory synapses on GABAergic interneurons [88, 89], renewing support for the hypothesis that SCZD results, at least in part, due to reduced excitatory glutamatergic input onto GABAergic inhibitory neurons. 22q11 Models Impaired Long-Range Synchrony of Neural Activity Mouse models of 22q11.2 represent the ﬁrst studies of a CNV associated with SCZD; SCZD develops in about 20–25% of individuals with a chromosome 22q11.2 microdeletion. Mice with disruptions of 22q11.2 genes have fewer cortical neurons with slightly smaller spines, altered short- and long-term synaptic plasticity, enhanced neurotransmitter release, altered calcium kinetics in CA3 presynaptic terminals, and impaired long-range synchrony between the hippocampus and prefrontal cortex [90–92].

INSIGHTS FROM OLFACTORY NEURAL PRECURSORS

Olfactory neural precursors (ONPs) can be expanded following exfoliation of the nasal cavity via a noninvasive method [93]. ONPs are capable of self-renewal as well as differentiation to mature electrophysiologically active neurons. ONPs Figure 3. Representative phenotypes present in dominant negative Disrupted-in-Schizophrenia-1 (DISC1) mouse models of SCZD. Top, from SCZD patients and controls have been generated and increased ventricular volume in dnDISC1 mice. Middle, reduced den- banked [93] to ask whether genetic, structural, and/or func- dritic arborization in dnDISC1 mice. Bottom, reduced frequency of tional abnormalities are present in SCZD neurons. By similar spontaneous inhibitory postsynaptic currents in dnDISC1 mice. methods, a second group generated ONPs from control and Adapted from Refs. [71] and [72]. Abbreviation: SCZD, SCZD patients and performed gene expression comparisons, schizophrenia which identified differences in neurodevelopmental pathways associated with cell migration and axon guidance [94]. reduced DISC1 activity during development have reduced neu- ONPs are one source of live human neurons for the study rite outgrowth [74], reduced cortical migration [74], reduced of SCZD. Although differences between SCZD and control dendritic complexity [71, 73], reduced hippocampal synaptic ONPs have been identified, it is unclear whether SCZD ONPs transmission [71, 73], and slightly enlarged ventricles [72] but will recapitulate all of the neuronal defects present in brain no significant structural defects or signs of neurodegeneration regions such as the cortex or hippocampus. Additionally, (Fig. 3). Conversely downregulation of DISC1 in adulthood ONPs cannot yet be used as a source of cells from the neural causes accelerated neural differentiation and increased neural lineages specifically implicated in SCZD: glutamatergic neu- excitability in newborn neurons [75]. DISC1 seems to function rons, GABAergic neurons, dopaminergic neurons, and as a molecular scaffold; it interacts with multiple proteins, oligodendrocytes. including the centrosomal protein NUDEL [74, 75] required for neurite outgrowth and neuronal migration [76, 77], as well as phosphodiesterase 4B [78], a key regulator of cyclic adeno- INSIGHTS FROM hiPSC NEURONS sine monophosphate (cAMP), linked to learning, memory, and mood [79–81]. It remains unknown which of these functions is responsible for SCZD pathogenesis. Chiang et al. [9] first published the generation of hiPSCs from SCZD patients with a DISC1 mutation but did not char- NRG1 and ERBB4 Mice Demonstrate Role of Exci- acterize neurons differentiated from these hiPSCs. We then demonstrated that SCZD hiPSC neurons had defects in neuro- tatory Glutamatergic Input onto GABAergic Inhibi- nal connectivity and gene expression, which could be amelio- tory Neurons in SCZD rated following treatment with the antipsychotic Loxapine [8]. Although mice lacking both copies of either the NRG1 or Specifically, we observed reduced neuronal connectivity, ERBB4 genes are embryonically lethal due to cardiac defects, reduced outgrowths from soma, and reduced postsynaptic heterozygous null NRG1 and ERBB4 animals have behavioral density protein 95 (PSD95) dendritic protein levels, all of abnormalities such as hyperactivity, increased aggression, and which are cellular phenotypes previously described in post- deficiencies in prepulse inhibition, a measure of sensory gating mortem SCZD brain tissue as well as animal models of that is abnormal in SCZD [42, 82, 83]. Although cell layers in SCZD (Fig. 4). Additionally, we observed gene expression the cerebral cortex, hippocampus, and cerebellum develop nor- differences in SCZD neurons relative to controls, 25% of mally in the mutant mice, there are defects in neurite outgrowth which had been previously implicated in SCZD, with signifi- and arborization neuronal migration [84, 85] and impaired syn- cant perturbations in genes associated with the WNT pathway, aptic maturation and function [83, 86, 87]. Treatment with the cAMP signaling, and GLU receptor expression. We hypothe- antipsychotic drug Clozapine reverses the behavioral and spine size that studies of SCZD hiPSC neurons from an increased defects in these mice [83]. ERBB4 is enriched in GABAergic number of patients might identify core pathways of genes interneurons, while NRG1 is primarily localized at synapses in contributing to SCZD. Pedrosa et al. have also generated excitatory glutamatergic neurons. NRG1 increases the number, SCZD hiPSCs from three patients and report that SCZD Brennand and Gage 1919

Figure 4. Decreased neuronal connectivity in schizophrenia (SCZD) hiPSC neurons. (A): Representative images of reduced trans-neuronal labeling (red) in SCZD hiPSC neurons relative to controls. (B): Graph showing decreased neurites in SCZD hiPSC neurons. (C): Graph showing decreased PSD95 protein levels relative to the dendritic marker MAP2AB for control and SCZD hiPSC neurons. Adapted from Ref. [8]. Abbreviations: hiPSCs, human induced pluripotent stem cells; MAP2AB, microtubule-associated protein 2AB; PSD95, postsynaptic density protein 95. ***p < 0.001. hiPSC neurons express a number of transcription factors, ment brain imaging, pathological, pharmacological, genetic, chromatin remodeling proteins, and synaptic proteins relevant and animal studies of SCZD. to SZCD pathogenesis, independently validating the potential It will be critical that researchers carefully consider which utility of hiPSC neurons in modeling SCZD [95]. specific subtypes of neurons should be compared using hiPSC Although hiPSC-based studies show exciting promise for neurons studies. We propose that the field begin to characterize the study of SCZD, they remain limited by three types of varia- cell-autonomous defects in midbrain DA (mDA) and cortical bilities: neuron-to-neuron, hiPSC-to-hiPSC, and patient-to- glutamatergic (cGLU) and GABAergic SCZD hiPSC neurons. patient. Interneuron variability is countered by studying more An efficient protocol can now differentiate pluripotent homogeneous populations of hiPSC neurons, generated by stem cells to populations consisting of approximately 20% directed differentiation protocols to specific neuronal subtypes DA neurons [96] by recapitulating developmental cues found and subsequent purification by fluorescence-activated cell sort- in the ventral midline when SHH, FGF8, and WNT1 initiate ing (FACS). Inter-hiPSC variation is addressed by comparing DA differentiation; immature mDA neurons express NURR1, multiple hiPSC lines per patient, particularly as it is well-estab- EN1/2, and LMX1A/B, whereas mature mDA neurons also lished that genetic and epigenetic differences exist between express tyrosine hydoxylase and aromatic L-amino acid decar- hiPSC lines. Finally, interpatient variability can be tackled by boxylase [97]. Studies using mouse ESCs have shown that studying an increased number of patients and controls, and par- NODAL antagonists (LEFTYA) induce expression of the fore- ticularly by selecting homogenous patient cohorts characterized brain marker brain-factor 1 (BF1, FOXG1) and subsequent by common clinical endophenotypes, pharmacological treatment with WNT antagonists causes regional specification responses, or genetic mutations. Using hiPSC neurons, research- toward cortical fate [98]. Genes such as EMX1, FEZF2, and ers can now begin to dissect the molecular mechanism of phar- FEZ are expressed in immature cGLU neurons [99–101], while macological response and screen for new drugs to improve cel- mature cGLU neurons express CTIP2 and OTX1 [99, 102, lular phenotypes in SCZD neurons from patients with clear 103]. Subsequent culture with FGF2 has been shown to clinical pharmacological nonresponsiveness. increase GABAergic differentiation [104]; GABA neurons can be identified by expression of key GABAergic markers, such as glutamate decarboxylase (GAD65/67), DARPP32, ARPP21, CALBINDIN, or CALRETININ [105], although few if any re- MOVING FORWARD:FUTURE hiPSC STUDIES gional markers of basal ganglia identity have been identified. OF SCZD Future molecular studies of hiPSC neurons should incor- porate SNP, CNV, and gene expression data. Studies of quan- Brain-imaging studies have identified structural changes in titative trait loci (eQTLs) will determine how genetic lesions SCZD brains but cannot resolve which neuronal cell types are affect gene expression in SCZD neurons. Current eQTL stud- affected in SCZD. Pharmacological studies have identified a ies can only compare a limited supply of heterogeneous post- role for DA and GLU; however, chronic antipsychotic treat- mortem brain tissue confounded by variables such as patient ment alters brain structure and neural activity, confounding treatment history, drug/alcohol abuse, and poverty. Ideally, studies of human patients affected with SCZD. GWAS studies hiPSC-based eQTL studies would produce a renewable supply have yet to account for most of the heritable variance of of more homogeneous cell populations. As hiPSC generation, SCZD. Although animal models have recapitulated aspects of neuronal differentiation and subtype purification are stream- the behavioral and cellular phenotypes of SCZD, they lack lined and made more efficient, it will become possible to gen- the ability to define the complex interacting genetic factors erate any defined neuronal subtypes from hiPSCs generated that contribute to disease. hiPSC-based studies will comple- from hundreds of patients with known genetic backgrounds. www.StemCells.com 1920 hiPSC-Based Studies of Schizophrenia

Currently, laborious single-cell electrophysiological analy- cohorts, one can now test whether the severity of clinical out- sis is the best method to establish the maturity and assess syn- come is predictive of the magnitude of cellular phenotype, if aptic function of the SCZD hiPSC neurons. Electrophysiologi- genetic lesions correlate to neuronal gene expression differen- cal characterization can verify that hiPSC neurons have ces or if clinical pharmacological response is predictable by membrane potentials, undergo induced action potentials, and hiPSC neuronal drug response. Beyond correlating genetic show evidence of spontaneous synaptic activity. However, to lesions to the disease state, we can assay the expression or ac- study functional synaptic activity defects contributing to tivity of genes identified through GWAS, both in the specific SCZD, and detect significant effects in heterogeneous patient patient in whom the lesion was identified and across cohorts populations, it will be necessary to increase both the number of SCZD patients. By carefully selecting from patients with of patients and neurons that can be studied. We believe that well-characterized diagnosis, MRI brain scans, genotyping developing more efficient and higher throughput synaptic data, and clinical treatment history, hiPSCs can be generated analyses needs to become a priority of the field. specifically from SCZD patients with extreme endopheno- Beyond neurons, hiPSCs can of course also be differenti- types, for example, patients showing drastically altered brain ated to any other cell type implicated in SCZD, particularly volumes by MRI or clear treatment resistance. Neurons oligodendrocytes [106, 107]. By considering data that myelin derived from these patients will allow testing of the genetic dysfunction contributes to SCZD, comparisons of cocultures causes of each endophenotype. With hiPSCs, we can finally of control and SCZD neurons and oligodendrocytes should move beyond phenotyping of SCZD patients and begin to de- reveal whether reduced myelination in SCZD neurons is a velop well-controlled experiments to test the specific molecu- cell-autonomous effect. If studies of hiPSC-derived cells lar and cellular effects of disease and pharmacological treat- reveal aberrant oligodendrocyte activity in SCZD, this may ment and response. It is time to begin correlating clinical, indicate a new point of therapeutic intervention in SCZD. genetic, and pharmacological studies in vitro. Moving forward, hiPSC studies must make several critical advances. Future studies should focus on defined neuronal subtypes. More efficient hiPSC generation and neuronal differentiation will ultimately permit eQTL studies of SCZD. ACKNOWLEDGMENTS More scalable assays of synaptic function will allow characterization of increased numbers of control and patient neu- K.J.B. is supported by a training grant from the California Insti- rons. New studies will need to recruit better-characterized tute for Regenerative Medicine. The Gage Laboratory is par- patient populations with well-defined clinical endophenotypes, tially funded by CIRM Grant RL1-00649-1, The Lookout and pharmacological responses, and/or genetic lesions. Mathers Foundation, the Helmsley Foundation as well as Sanofi- Aventis.

CONCLUSION DISCLOSURE OF POTENTIAL It is now time to begin to synthesize the disparate ﬁelds of CONFLICT OF INTEREST brain imaging, neurobiology, and genetics. By generating hiPSC neurons from more and better-characterized patient The authors indicate no potential conﬂict of interests.

13 Steen RG, Mull C, McClure R et al. Brain volume in first-episode REFERENCES schizophrenia: Systematic review and meta-analysis of magnetic resonance imaging studies. Br J Psychiatry 2006;188:510–518. 14 Wright IC, Rabe-Hesketh S, Woodruff PW et al. Meta-analysis of re- 1 Association AP. Diagnostic and Statistical Manual of Mental Disor- gional brain volumes in schizophrenia. Am J Psychiatry 2000;157:16–25. ders: DSM-IV. Washington, DC: American Psychiatric Press, 15 Thompson PM, Vidal C, Giedd JN et al. Mapping adolescent brain 1994:886. change reveals dynamic wave of accelerated gray matter loss in very 2 Carpenter WT, Jr., Buchanan RW. Schizophrenia. New Engl J Med early-onset schizophrenia. Proc Natl Acad Sci USA 2001;98: 1994;330:681–690. 11650–11655. 3 Hannerz H, Borga P, Borritz M. Life expectancies for individuals 16 Ellison-Wright I, Glahn DC, Laird AR et al. The anatomy of first- with psychiatric diagnoses. Public Health 2001;115:328–337. episode and chronic schizophrenia: An anatomical likelihood estima- 4 Regier DA, Farmer ME, Rae DS et al. Comorbidity of mental disor- tion meta-analysis. Am J Psychiatry 2008;165:1015–1023. ders with alcohol and other drug abuse. Results from the Epidemio- 17 Hulshoff Pol HE, Kahn RS. What happens after the first episode? A logic Catchment Area (Eca) Study. JAMA 1990;264:2511–2518. review of progressive brain changes in chronically ill patients with 5 Brown S, Inskip H, Barraclough B. Causes of the excess mortality of schizophrenia. Schizophr Bull 2008;34:354–366. schizophrenia. Br J Psychiatry 2000;177:212–217. 6 Radomsky ED, Haas GL, Mann JJ et al. Suicidal behavior in patients 18 Fields RD, Stevens-Graham B. New insights into neuron-glia com- with schizophrenia and other psychotic disorders. Am J Psychiatry munication. Science 2002;298:556–562. 1999;156:1590–1595. 19 Kubicki M, McCarley R, Westin CF et al. A review of diffusion ten- 7 Caldwell CB, Gottesman II. Schizophrenia—A high-risk factor for suicide: sor imaging studies in schizophrenia. J Psychiatr Res 2007;41:15–30. Clues to risk reduction. Suicide Life Threat Behav 1992;22:479–493. 20 Wexler BE, Bell MD. Cognitive remediation and vocational rehabili- 8 Brennand KJ, Simone A, Jou J et al. Modelling schizophrenia using tation for schizophrenia. Schizophr Bull 2005;31:931–941. human induced pluripotent stem cells. Nature 2011;473:221–225. 21 Whitfield-Gabrieli S, Thermenos HW, Milanovic S et al. Hyperactiv- 9 Chiang CH, Su Y, Wen Z et al. Integration-free induced pluripotent ity and hyperconnectivity of the default network in schizophrenia stem cells derived from schizophrenia patients with a DISC1 muta- and in first-degree relatives of persons with schizophrenia. Proc Natl tion. Mol Psychiatry 2011;16:358–360. Acad Sci USA 2009;106:1279–1284. 10 Lange N, Giedd JN, Castellanos FX et al. Variability of human brain 22 Kessler RM, Woodward ND, Riccardi P et al. Dopamine D2 receptor structure size: Ages 4–20 years. Psychiatry Res 1997;74:1–12. levels in striatum, thalamus, substantia nigra, limbic regions, and cor- 11 Giedd JN, Rapoport JL. Structural MRI of pediatric brain develop- tex in schizophrenic subjects. Biol Psychiatry 2009;65:1024–1031. ment: What have we learned and where are we going?Neuron 2010; 23 Weinberger DR. Implications of normal brain development for the 67:728–734. pathogenesis of schizophrenia. Arch Gen Psychiatry 1987;44:660–669. 12 Vita A, De Peri L, Silenzi C et al. Brain morphology in first-episode 24 Kane J, Honigfeld G, Singer J et al. Clozapine for the treatment-re- schizophrenia: A meta-analysis of quantitative magnetic resonance sistant schizophrenic. A double-blind comparison with chlorproma- imaging studies. Schizophr Res 2006;82:75–88. zine. Arch Gen Psychiatry 1988;45:789–796. Brennand and Gage 1921

25 Klein C, Gordon J, Pollak L et al. Clozapine in Parkinson’s disease 51 Hakak Y, Walker JR, Li C et al. Genome-wide expression analysis psychosis: 5-year follow-up review. Clin Neuropharmacol 2003;26: reveals dysregulation of myelination-related genes in chronic schizo- 8–11. phrenia. Proc Natl Acad Sci USA 2001;98:4746–4751. 26 Nord M, Farde L. Antipsychotic occupancy of dopamine receptors in 52 Silberberg G, Darvasi A, Pinkas-Kramarski R et al. The involvement schizophrenia. CNS Neurosci Ther 2011;17:97–103. of ErbB4 with schizophrenia: Association and expression studies. 27 Gray L, van den Buuse M, Scarr E et al. Clozapine reverses schizo- Am J Med Genet B Neuropsychiatr Genet B 2006;141:142–148. phrenia-related behaviours in the metabotropic glutamate receptor 5 53 Sugai T, Kawamura M, Iritani S et al. Prefrontal abnormality of knockout mouse: Association with N-methyl-D-aspartic acid receptor schizophrenia revealed by DNA microarray: Impact on glial and neu- up-regulation. Int J Neuropsychopharmacol 2009;12:45–60. rotrophic gene expression. Ann N Y Acad Sci 2004;1025:84–91. 28 Michel FJ, Trudeau LE. Clozapine inhibits synaptic transmission at 54 Shi J, Levinson DF, Duan J et al. Common variants on chromosome GABAergic synapses established by ventral tegmental area neurones 6p22.1 are associated with schizophrenia. Nature 2009;460:753–757. in culture. Neuropharmacology 2000;39:1536–1543. 55 Stefansson H, Ophoff RA, Steinberg S et al. Common variants con- 29 Krystal JH, Karper LP, Seibyl JP et al. Subanesthetic effects of the ferring risk of schizophrenia. Nature 2009;460:744–747. noncompetitive NMDA antagonist, ketamine, in humans. Psychoto- 56 Purcell SM, Wray NR, Stone JL et al. Common polygenic variation mimetic, perceptual, cognitive, and neuroendocrine responses. Arch contributes to risk of schizophrenia and bipolar disorder. Nature Gen Psychiatry 1994;51:199–214. 2009;460:748–752. 30 Patil ST, Zhang L, Martenyi F et al. Activation of mGlu2/3 receptors 57 Huh GS, Boulanger LM, Du H et al. Functional requirement for class as a new approach to treat schizophrenia: A randomized Phase 2 I MHC in CNS development and plasticity. Science 2000;290: clinical trial. Nat Med 2007;13:1102–1107. 2155–2159. 31 Selemon LD, Goldman-Rakic PS. The reduced neuropil hypothesis: 58 Goddard CA, Butts DA, Shatz CJ. Regulation of CNS synapses by A circuit based model of schizophrenia. Biol Psychiatry 1999;45: neuronal MHC class I. Proc Natl Acad Sci USA 2007;104: 17–25. 6828–6833. 32 Rajkowska G, Selemon LD, Goldman-Rakic PS. Neuronal and glial 59 Boulanger LM. MHC class I in activity-dependent structural and somal size in the prefrontal cortex: A postmortem morphometric functional plasticity. Neuron Glia Biol 2004;1:283–289. study of schizophrenia and Huntington disease. Arch Gen Psychiatry 60 O’Donovan MC, Craddock N, Norton N et al. Identification of loci 1998;55:215–224. associated with schizophrenia by genome-wide association and fol- 33 Black JE, Kodish IM, Grossman AW et al. Pathology of layer V py- low-up. Nat Genet 2008;40:1053–1055. ramidal neurons in the prefrontal cortex of patients with schizophre- 61 Gulacsi AA, Anderson SA. Beta-catenin-mediated Wnt signaling reg- nia. Am J Psychiatry 2004;161:742–744. ulates neurogenesis in the ventral telencephalon. Nat Neurosci 2008; 34 Garey LJ, Ong WY, Patel TS et al. Reduced dendritic spine density 11:1383–1391. on cerebral cortical pyramidal neurons in schizophrenia. J Neurol 62 Huang FL, Huang KP, Boucheron C. Long-term enrichment enhan- Neurosurg Psychiatry 1998;65:446–453. ces the cognitive behavior of the aging neurogranin null mice with- 35 Glantz LA, Lewis DA. Decreased dendritic spine density on prefron- out affecting their hippocampal LTP. Learn Mem 2007;14:512–519. tal cortical pyramidal neurons in schizophrenia. Arch Gen Psychiatry 63 Esslinger C, Walter H, Kirsch P et al. Neural mechanisms of a ge- 2000;57:65–73. nome-wide supported psychosis variant. Science 2009;324:605. 36 Meador-Woodruff JH, Healy DJ. Glutamate receptor expression in 64 Walsh T, McClellan JM, McCarthy SE et al. Rare structural variants schizophrenic brain. Brain Res Brain Res Rev 2000;31:288–294. disrupt multiple genes in neurodevelopmental pathways in schizo- 37 Cardno AG, Gottesman II. Twin studies of schizophrenia: From phrenia. Science 2008;320:539–543. bow-and-arrow concordances to star wars Mx and functional genomics. Am J Med Genet 2000;97:12–17. 65 Kirov G, Gumus D, Chen W et al. Comparative genome hybridiza- tion suggests a role for NRXN1 and APBA2 in schizophrenia. 38 Sullivan PF, Kendler KS, Neale MC. Schizophrenia as a complex Human Mol Genet 2008;17:458–465. trait: Evidence from a meta-analysis of twin studies. Arch Gen Psy- chiatry 2003;60:1187–1192. 66 Mefford HC, Sharp AJ, Baker C et al. Recurrent rearrangements of 39 Ekelund J, Hovatta I, Parker A et al. Chromosome 1 loci in Finnish chromosome 1q21.1 and variable pediatric phenotypes. New Engl J schizophrenia families. Human Mol Genet 2001;10:1611–1617. Med 2008;359:1685–1699. 40 Stefansson H, Sigurdsson E, Steinthorsdottir V et al. Neuregulin 1 67 Stefansson H, Rujescu D, Cichon S et al. Large recurrent microdele- and susceptibility to schizophrenia. Am J Human Genet 2002;71: tions associated with schizophrenia. Nature 2008;455:232–236. 877–892. 68 International Schizophrenia Consortium.Rare chromosomal deletions 41 Kim JW, Lee YS, Cho EY et al. Linkage and association of schizo- and duplications increase risk of schizophrenia. Nature 2008;455: phrenia with genetic variations in the locus of neuregulin 1 in Ko- 237–241. rean population. Am J Med Genet B Neuropsychiatr Genet B 2006; 69 Zhang F, Gu W, Hurles ME et al. Copy number variation in human 141:281–286. health, disease, and evolution. Annu Rev Genomics Hum Genet 42 Corvin AP, Morris DW, McGhee K et al. Confirmation and refine- 2009;10:451–481. ment of an ‘at-risk’ haplotype for schizophrenia suggests the EST 70 Conrad DF, Pinto D, Redon R et al. Origins and functional impact of cluster, Hs 97362, as a potential susceptibility gene at the Neuregu- copy number variation in the human genome. Nature 2010;464: lin-1 locus. Mol Psychiatry 2004;9:208–213. 704–712. 43 Li T, Stefansson H, Gudfinnsson E et al. Identification of a novel 71 Li W, Zhou Y, Jentsch JD et al. Specific developmental disruption neuregulin 1 at-risk haplotype in Han schizophrenia Chinese patients, of disrupted-in-schizophrenia-1 function results in schizophrenia- but no association with the Icelandic/Scottish risk haplotype. Mol related phenotypes in mice. Proc Natl Acad Sci USA 2007;104: Psychiatry 2004;9:698–704. 18280–18285. 44 Liu CM, Hwu HG, Fann CS et al. Linkage evidence of schizophrenia 72 Pletnikov MV, Ayhan Y, Nikolskaia O et al. Inducible expression of to loci near neuregulin 1 gene on chromosome 8p21 in Taiwanese mutant human DISC1 in mice is associated with brain and behavioral families. Am J Med Genet B Neuropsychiatr Genet B 2005;134: abnormalities reminiscent of schizophrenia. Mol psychiatry 2008;13: 79–83. 173–186, 115. 45 Stefansson H, Sarginson J, Kong A et al. Association of neuregulin 1 73 Kvajo M, McKellar H, Arguello PA et al. A mutation in mouse with schizophrenia confirmed in a Scottish population. Am J Human Disc1 that models a schizophrenia risk allele leads to specific altera- Genet 2003;72:83–87. tions in neuronal architecture and cognition. Proc Natl Acad Sci 46 Tang JX, Chen WY, He G et al. Polymorphisms within 50 end of the USA 2008;105:7076–7081. Neuregulin 1 gene are genetically associated with schizophrenia in 74 Kamiya A, Kubo K, Tomoda T et al. A schizophrenia-associated the Chinese population. Mol Psychiatry 2004;9:11–12. mutation of DISC1 perturbs cerebral cortex development. Nat Cell 47 Bakker SC, Hoogendoorn ML, Selten JP et al. Neuregulin 1: Genetic Biol 2005;7:1167–1178. support for schizophrenia subtypes. Mol Psychiatry 2004;9: 75 Duan X, Chang JH, Ge S et al. Disrupted-In-Schizophrenia 1 regu- 1061–1063. lates integration of newly generated neurons in the adult brain. Cell 48 Law AJ, Lipska BK, Weickert CS et al. Neuregulin 1 transcripts are 2007;130:1146–1158. differentially expressed in schizophrenia and regulated by 50 SNPs 76 Sasaki S, Mori D, Toyo-oka K, et al. Complete loss of Ndel1 results associated with the disease. Proc Natl Acad Sci USA 2006;103: in neuronal migration defects and early embryonic lethality. Mol 6747–6752. Cell Biol 2005;25:7812–7827. 49 Hashimoto R, Straub RE, Weickert CS et al. Expression analysis of 77 Shu T, Ayala R, Nguyen MD et al. Ndel1 operates in a common neuregulin-1 in the dorsolateral prefrontal cortex in schizophrenia. pathway with LIS1 and cytoplasmic dynein to regulate cortical neu- Mol Psychiatry 2004;9:299–307. ronal positioning. Neuron 2004;44:263–277. 50 Chong VZ, Thompson M, Beltaifa S et al. Elevated neuregulin-1 and 78 Millar JK, Pickard BS, Mackie S et al. DISC1 and PDE4B are inter- ErbB4 protein in the prefrontal cortex of schizophrenic patients. acting genetic factors in schizophrenia that regulate cAMP signaling. Schizophr Res 2008;100:270–280. Science 2005;310:1187–1191. www.StemCells.com 1922 hiPSC-Based Studies of Schizophrenia

79 Morris JA, Kandpal G, Ma L et al. DISC1 (Disrupted-In-Schizophre- 93 Benitez-King G, Riquelme A, Ortiz-Lopez L et al. A non-invasive nia 1) is a centrosome-associated protein that interacts with MAP1A, method to isolate the neuronal linage from the nasal epithelium from MIPT3, ATF4/5 and NUDEL: Regulation and loss of interaction schizophrenic and bipolar diseases. J Neurosci Methods 2011;201: with mutation. Human Mol Genet 2003;12:1591–1608. 35–45. 80 Brandon NJ, Handford EJ, Schurov I et al. Disrupted in Schizophre- 94 Matigian N, Abrahamsen G, Sutharsan R et al. Disease-specific, neu- nia 1 and Nudel form a neurodevelopmentally regulated protein com- rosphere-derived cells as models for brain disorders. Dis Model plex: Implications for schizophrenia and other major neurological Mech 2010;3:785–798. disorders. Mol Cell Neurosci 2004;25:42–55. 95 Pedrosa E, Sandler V, Shah A et al. Development of patient-specific 81 Ozeki Y, Tomoda T, Kleiderlein J et al. Disrupted-in-Schizophrenia- neurons in schizophrenia using induced pluripotent stem cells. J Neu- 1 (DISC-1): Mutant truncation prevents binding to NudE-like rogenet 2011;25:88–103. (NUDEL) and inhibits neurite outgrowth. Proc Natl Acad Sci USA 96 Chambers SM, Fasano CA, Papapetrou EP et al. Highly efficient 2003;100:289–294. neural conversion of human ES and iPS cells by dual inhibition of 82 Li B, Woo RS, Mei L et al. The neuregulin-1 receptor erbB4 con- SMAD signaling. Nat Biotechnol 2009;27:275–280. trols glutamatergic synapse maturation and plasticity. Neuron 2007; 97 Ang SL. Transcriptional control of midbrain dopaminergic neuron 54:583–597. development. Development 2006;133:3499–3506. 83 Barros CS, Calabrese B, Chamero P et al. Impaired maturation of 98 Watanabe K, Kamiya D, Nishiyama A et al. Directed differentiation dendritic spines without disorganization of cortical cell layers in of telencephalic precursors from embryonic stem cells. Nat Neurosci mice lacking NRG1/ErbB signaling in the central nervous system. 2005;8:288–296. Proc Natl Acad Sci USA 2009;106:4507–4512. 99 Arlotta P, Molyneaux BJ, Chen J et al. Neuronal subtype-specific 84 Lopez-Bendito G, Cautinat A, Sanchez JA et al. Tangential neuronal genes that control corticospinal motor neuron development in vivo. migration controls axon guidance: A role for neuregulin-1 in thala- Neuron 2005;45:207–221. mocortical axon navigation. Cell 2006;125:127–142. 100 Molyneaux BJ, Arlotta P, Hirata T et al. Fezl is required for the birth 85 Krivosheya D, Tapia L, Levinson JN et al. ErbB4-neuregulin signal- and specification of corticospinal motor neurons. Neuron 2005;47: ing modulates synapse development and dendritic arborization 817–831. through distinct mechanisms. J Biol Chem 2008;283:32944–32956. 101 Chen B, Schaevitz LR, McConnell SK. Fezl regulates the differentia- 86 Pitcher GM, Beggs S, Woo RS et al. ErbB4 is a suppressor of long- tion and axon targeting of layer 5 subcortical projection neurons in term potentiation in the adult hippocampus. Neuroreport 2008;19: cerebral cortex. Proc Natl Acad Sci USA 2005;102:17184–17189. 139–143. 102 Frantz GD, Weimann JM, Levin ME et al. Otx1 and Otx2 define 87 Chen YJ, Zhang M, Yin DM et al. ErbB4 in parvalbumin-positive layers and regions in developing cerebral cortex and cerebellum. J interneurons is critical for neuregulin 1 regulation of long-term Neurosci 1994;14:5725–5740. potentiation. Proc Natl Acad Sci USA 2010;107:21818–21823. 103 Weimann JM, Zhang YA, Levin ME et al. Cortical neurons require 88 Wen L, Lu YS, Zhu XH et al. Neuregulin 1 regulates pyramidal neu- Otx1 for the refinement of exuberant axonal projections to subcorti- ron activity via ErbB4 in parvalbumin-positive interneurons. Proc cal targets. Neuron 1999;24:819–831. Natl Acad Sci USA 2010;107:1211–1216. 104 Barberi T, Klivenyi P, Calingasan NY et al. Neural subtype specifi- 89 Ting AK, Chen Y, Wen L et al. Neuregulin 1 promotes excitatory cation of fertilization and nuclear transfer embryonic stem cells and synapse development and function in GABAergic interneurons. J application in parkinsonian mice. Nat Biotechnol 2003;21: Neurosci 2011;31:15–25. 1200–1207. 90 Fenelon K, Mukai J, Xu B et al. Deficiency of Dgcr8, a gene dis- 105 Aubry L, Bugi A, Lefort N et al. Striatal progenitors derived from rupted by the 22q11.2 microdeletion, results in altered short-term human ES cells mature into DARPP32 neurons in vitro and in quino- plasticity in the prefrontal cortex. Proc Natl Acad Sci USA 2011; linic acid-lesioned rats. Proc Natl Acad Sci USA 2008;105: 108:4447–4452. 16707–16712. 91 Earls LR, Bayazitov IT, Fricke RG et al. Dysregulation of presynap- 106 Keirstead HS, Nistor G, Bernal G et al. Human embryonic stem cell- tic calcium and synaptic plasticity in a mouse model of 22q11 dele- derived oligodendrocyte progenitor cell transplants remyelinate and tion syndrome. J Neurosci 2010;30:15843–15855. restore locomotion after spinal cord injury. J Neurosci 2005;25: 92 Sigurdsson T, Stark KL, Karayiorgou M et al. Impaired hippocam- 4694–4705. pal-prefrontal synchrony in a genetic mouse model of schizophrenia. 107 Hu BY, Du ZW, Zhang SC. Differentiation of human oligodendro- Nature 2010;464:763–767. cytes from pluripotent stem cells. Nat Protoc 2009;4:1614–1622.