Associations between neurodevelopmental genes, neuroanatomy, and ultra high risk symptoms of psychosis in 22q11.2 deletion syndrome

Item Type Article

Authors Thompson, Carlie A.; Karelis, Jason; Middleton, Frank A.; Gentile, Karen; Coman, Ioana L.; Radoeva, Petya D.; Mehta, Rashi; Fremont, Wanda P.; Antshel, Kevin M.; Faraone, Stephen V.; Kates, Wendy R.

Citation Thompson CA, Karelis J, Middleton FA, Gentile K, Coman IL, Radoeva PD, Mehta R, Fremont WP, Antshel KM, Faraone SV, Kates WR. 2017. Associations Between Neurodevelopmental Genes, Neuroanatomy, and Ultra High Risk Symptoms of Psychosis in 22q11.2 Deletion Syndrome. Am J Med Genet Part B 174B:295–314

DOI 10.1002/ajmg.b.32515

Publisher Wiley

Rights Attribution-NonCommercial-NoDerivatives 4.0 International

Download date 01/10/2021 18:46:16

Item License http://doi.wiley.com/10.1002/tdm_license_1 Link to Item http://hdl.handle.net/20.500.12648/1796 RESEARCH ARTICLE

Neuropsychiatric Genetics Associations Between Neurodevelopmental Genes, Neuroanatomy, and Ultra High Risk Symptoms of Psychosis in 22q11.2 Deletion Syndrome Carlie A. Thompson,1 Jason Karelis,1 Frank A. Middleton,1,2 Karen Gentile,2 Ioana L. Coman,3 Petya D. Radoeva,4 Rashi Mehta,5 Wanda P. Fremont,1 Kevin M. Antshel,1,6 Stephen V. Faraone,1 and Wendy R. Kates1* 1Department of Psychiatry and Behavioral Sciences, SUNY Upstate Medical University, Syracuse, New York 2Department of Neuroscience, SUNY Upstate Medical University, Syracuse, New York 3Department of Computer Science, SUNY Oswego, Oswego, New York 4Department of Psychiatry, University of Washington, Seattle, Washington 5Department of Radiology, SUNY Upstate Medical University, Syracuse, New York 6Department of Psychology, Syracuse University, Syracuse, New York

Manuscript Received: 14 April 2016; Manuscript Accepted: 7 November 2016

22q11.2 deletion syndrome is a neurogenetic disorder resulting in the deletion of over 40 genes. Up to 40% of individuals with How to Cite this Article: 22q11.2DS develop schizophrenia, though little is known about Thompson CA, Karelis J, Middleton FA, the underlying mechanisms. We hypothesized that allelic varia- Gentile K, Coman IL, Radoeva PD, Mehta tion in functional polymorphisms in seven genes unique to the R, Fremont WP, Antshel KM, Faraone SV, deleted region would affect lobar brain volumes, which would Kates WR. 2017. Associations Between predict risk for psychosis in youth with 22q11.2DS. Participants Neurodevelopmental Genes, included 56 individuals (30 males) with 22q11.2DS. Anatomic Neuroanatomy, and Ultra High Risk MR images were collected and processed using Freesurfer. Symptoms of Psychosis in 22q11.2 Deletion Participants were genotyped for 10 SNPs in the COMT, Syndrome. DGCR8, GNB1L, PIK4CA, PRODH, RTN4R, and ZDHHC8 genes. All subjects were assessed for ultra high risk symptoms Am J Med Genet Part B 174B:295–314. of psychosis. Allelic variation of the rs701428 SNP of RTN4R was significantly associated with volumetric differences in gray matter of the lingual gyrus and cuneus of the occipital lobe. 2008]. The typically deleted region (TDR) is 3 Mb, found in Moreover, occipital gray matter volumes were robustly associ- approximately 87% of cases [Shaikh et al., 2000]. Low-copy ated with ultra high risk symptoms of psychosis in the presence DNA repeats (LCRs) specific to chromosome 22 have been impli- of the G allele of rs701428. Our results suggest that RTN4R, a cated in the formation of 22q11.2 deletions and are found in relatively under-studied gene at the 22q11 locus, constitutes a proximity to the end-points of the TDR [Shaikh et al., 2000]. It susceptibility gene for psychosis in individuals with this syn- has been hypothesized that homologous recombination errors drome through its alteration of the architecture of the brain. between these LCRs during meiosis results in the deletions (and Ó 2017 Wiley Periodicals, Inc. duplications) found in the sequences of individuals with 22q11.2DS [Edelmann et al., 1999]. Key words: velo-cardio-facial syndrome; RTN4R; axonal development; Freesurfer; schizophrenia Conflicts of interest: The authors have no conflicts of interest to declare. Grant sponsor: National Institutes of Health; Grant number: MH064824. Correspondence to: Wendy R. Kates, Ph.D., Department of Psychiatry and Behavioral INTRODUCTION Sciences, State University of New York at Upstate Medical University, 750 East Adams Street, Syracuse, New York. 22q11.2 deletion syndrome (22q11.2DS) is the most common E-mail: [email protected] micro-deletion syndrome found in humans [Edelmann et al., Article first published online in Wiley Online Library 1999]. The size of the region deleted can vary but almost always (wileyonlinelibrary.com): 31 January 2017 spans a shared minimal deletion region of 1.5 Mb [Maynard et al., DOI 10.1002/ajmg.b.32515

Ó 2017 Wiley Periodicals, Inc. 295 296 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

Individuals with 22q11.2DS have increased susceptibility to glutamate [Bender et al., 2005]. Alterations in glutamate signaling behavioral, anxiety and mood disorders, and psychosis [Schneider are very well-established as risk factors for psychosis [Moghaddam et al., 2014]. In particular, they have over a 30% chance of and Javitt, 2012]. We examined the SNP, rs4819756 (A/G), as noted developing schizophrenia, which is considerably higher than the in Table I. worldwide population risk of less than 1% [Murphy et al., 1999; The ZDDHC8 gene, located in the distal third of the 1.5 Mb- Saha et al., 2005; Green et al., 2009; van Os and Kapur, 2009]. minimal deletion region [Maynard et al., 2008], is broadly Consequently, there has been considerable interest in determining expressed in many regions of the adult human brain [Mukai which genes in the TDR might contribute to the elevated risk for et al., 2004], particularly in the olfactory bulb, neocortex, and schizophrenia. Several genes for which individuals with 22q11.2DS cerebellum [Maynard et al., 2008], as well as the adult mouse brain are hemizygous have been linked to the development of schizo- [Mukai et al., 2004]. Among 72 known SNPs located within the 1.5- phrenia in both individuals with this syndrome and the population Mb deleted region, a polymorphism in ZDDHC8, known as at large [Chen et al., 2004a,b; Mukai et al., 2004; Gothelf et al., 2005, rs175174, had the greatest association with schizophrenia [Mukai 2011; Budel et al., 2008; Jungerius et al., 2008; Vorstman et al., 2009; et al., 2004]. This polymorphism mediates the alternative splicing Kempf et al., 2008]. However, data supporting the association of ZDDHC8 such that intron 4 is included in the final mRNA between individual genetic variants and risk for psychosis are not transcript; when this intron is retained, a premature stop codon is consistent, most likely due to the relatively distal relationship introduced into the growing amino acid chain that could terminate between genetic variants and psychopathological behaviors translation and result in a truncated protein with diminished [Meyer-Lindenberg and Weinberger, 2007], and the evidence activity [Mukai et al., 2004]. that multiple genes most likely confer small effects on risk for The Reticulon 4 receptor, also known as the Nogo-66 receptor psychiatric disease [Manolio et al., 2009]. It has been suggested that (NgR), is a glycosylphosphatidylinositol (GPI)-linked protein a more fruitful approach would include the examination of inter- encoded by the RTN4R gene, which is located within the distal mediate phenotypes. In this context, intermediate phenotypes are third of the minimal 1.5-Mb deletion region [Maynard et al., 2008]. biologically—based traits or mechanisms through which genes Through interacting with Nogo-66 (Neurite Outgrowth Inhibitor might affect behavior [Meyer-Lindenberg and Weinberger, 66), the receptor plays a significant role in the inhibition of myelin- 2007]. Since neuroanatomic structure is considered such an inter- mediated axonal growth [Fournier et al., 2001]. Nogo-66 localizes mediate phenotype, we sought to examine the extent to which to axons and binds to oligodendrocyte-myelin glycoprotein allelic variants of candidate genes alter neuroanatomic structure to (OMgp), myelin-associated glycoprotein (MAG), and Nogo-A increase the risk for psychosis in individuals with 22q11.2DS. Here (RTN4), all of which inhibit axonal sprouting. The Nogo-66 we describe briefly the functions of the seven candidate genes at the receptor, together with other GPI-linked axonal proteins, is re- 22q11.2 locus that we examined, all of which are expressed in brain quired for Nogo-66 to retain its inhibitory functions [Fournier and implicated in schizophrenia. Table I summarizes studies that et al., 2001]. Nogo-66 collapses the axonal growth cones found in have examined the clinical and, if available, neuroanatomic studies the dorsal root ganglion and inhibits neurite outgrowth [Fournier that justify the inclusion of each gene in the current study. et al., 2001]. Due to its potential role in plasticity and neuronal The COMT gene is located in the 1.5 critical deletion region that regeneration, the Reticulon 4 receptor is being researched as a is consistently deleted in cases of 22q11.2DS. Its protein product, possible drug-target for spinal and cerebrovascular injury [Baptiste the COMT enzyme (EC 2.1.1.6), catabolizes catecholamines— and Fehlings, 2007]. We have previously reported that the rs701428 dopamine, norepinephrine, and epinephrine. The COMT gene SNP in RTN4R is associated with alterations in white matter encodes both membrane-bound and soluble COMT enzyme microstructure the anterior limb of the internal capsule (ALIC) (MB-COMT and S-COMT, respectively), with the former being in individuals with 22q11.2DS (Table I). the predominant form in the brain [Bertocci et al., 1991; Chen et al., PIK4CA is an enzyme that helps to regulate signal transduction 2004a]. The COMT enzyme, along with monoamine oxidase A in neurons and synaptic transmission. It is expressed in gray (MAOA), has a particularly strong influence on dopamine metab- matter, though at lesser levels in the adult human brain than in olism, particularly in the prefrontal cortex [Tunbridge et al., 2004; fetal brain, potentially indicative of the neurodevelopmental role Tunbridge et al., 2006], where there is a relatively low concentra- PIK4CA could have [Vorstman et al., 2009]. As noted in Table I, tion of dopamine transporters and thus a need for an alternate way PIK4CA is associated with the development of schizophrenia in to clear dopamine from synapses. We examined a functional both individuals with and without 22q11.2DS [Vorstman et al., polymorphism at codon 158 of the MB-COMT enzyme, consisting 2009]. of the substitution of valine with methionine, that results in a DiGeorge Critical Region gene 8 (DGCR8), also known as Pasha, change from high to low enzymatic activity. As noted in Table I, this is essential to the processing of micro RNA molecules (miRNAs), single nucleotide polymorphism (SNP; rs4680) alters functions which are single-stranded mRNA segments that are typically 22- governed by the prefrontal cortex and has been implicated, albeit nucleotides in length [Sun et al., 2009; Brzustowicz and Bassett, inconsistently, in schizophrenia. 2012; Sellier et al., 2014; Zhao et al., 2015], and which have been PRODH, which codes for proline dehydrogenase, is also implicated in schizophrenia. They are the component of the RNA- located within the 1.5-mb critical deletion region of 22q11.2DS induced silencing complex (RISC) that enables the recognition of [Squarcione et al., 2013]. Proline is an intermediate in the biosyn- specific mRNA segments [Brzustowicz and Bassett, 2012; Forstner thesis of glutamate [Phang et al., 2001], and proline dehydrogenase et al., 2013; Sellier et al., 2014; Zhao et al., 2015]. It has been catalyzes the rate-limiting step in the conversion of proline to proposed that the miRNA/RISC system of post-transcriptional HMSNE AL. ET THOMPSON

TABLE I. Studies Involving SNPs of Interest and Relevant Findings

Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary DGCR8 Zhou et al. SCZ: 256 C: 252 (Chinese) SCZ: 28.44 12.51 Did not observe significant difference in genotype between SCZ and C rs3757 (A/G) [2013] AA: 11, AG: 76, GG: 169 AA:3,AG: 73, GG: 176 C: 30.73 15.5 but did note a significantly heightened risk of SCZ due to the recessive AA genotype.

GNB1L Williams et al. UK case-control: UK case-control: UK case-control: Though a number of SNPs were analyzed, rs5746832 and rs2269726 rs5746832 (A/G) [2008] SCZ: 662, SCZaff spectrum C: 1416 SCZ: 44 15 resulted in the strongest signals. In the UK case-control study, rs2269726(C/T) phenotype: 233 1958 birth cohort control C: 42.4 11.1 observed a male-specific genotype association in TBX1/GNB1L that German case-control: sample: 1421 (rs2269726 German case-control: was replicated in two other case-control studies. The Bulgarian trio SCZ: 513 only) SCZ: 38 sample also reinforced this finding. The 22q11.2DS sample also Bulgarian SCZ trio: German case-control: C: 49 provided evidence of a male-specific association between the SNP SCZ: 480 + parents C: 1330 Bulgarian trio: and psychosis. Genetic markers implicated in psychosis were 22q11.2DS: 83 (22 with SCZ: 33 8 associated with GNB1L expression alterations (using allele specific psychotic episodes) Fathers: 61 9 expression analysis). Mothers: 58 9

Ishiguro et al. rs5746832: rs5746832: SCZ: 48.9 14.5 Using Australian (10 SCZ, 10 C) and Japanese (43 SCZ, 11 C) post- [2010] SCZ: AA: 501, AG: 958, GG:430 AA: 484, AG: 916, GG:476 C: 49.0 14.3 mortem brain samples, observed reduced GNB1L expression levels rs2269726: rs2269726: in the prefrontal cortex of SCZ samples (these patients were not SCZ: TT: 338, TC: 896, CC: 671 TT: 309, TC: 911, CC: 686 genotyped for either SNP). No other gene expression levels differed significantly between cases and controls.No significant association between genotype and SCZ was observed. Did not observe any male- specific associations between genotype and SCZ as Williams et al. observed.

Li et al. [2011] BPD: 1135, MDD: 1135, SCZ: 1135 C: 1135 BPD: 36.6 Both SNPs were associated with BPD with rs5746832 demonstrating a (Chinese Han) rs5746832: MDD: 35.1 stronger association (p¼0.0001). Both SNPs were also associated rs5746832: A: 1175, G: 997 SCZ: 35.4 with SCZ but neither were linked to MDD. BPD: A: 1372, G:820 rs2269726: MDD: A: 1273, G: 932 C: 1272, T: 862 SCZ: A: 1059, G: 985 rs2269726: BPD: C: 1394, T: 758 MDD: C: 1383, T: 851 SCZ: C: 1296, T: 742

PIK4CA Jungerius et al. SCZ: 310 C: 880 (from two cohorts) N/A In the first stage of the study, rs165793 was the only SNP to survive rs165793 (A/G) [2008] rs2072513: rs2072513: correction following analyses comparing allele frequencies of 741 rs165862 (G/T) C: 271, T: 349 G: 584; A: 1192 SNPs within 138 myelin-related genes in a portion of both SCZ and C rs2072513 (C/T) rs165862: rs165862: groups. When more controls were added, association analyses G: 351, T: 269 G: 839; T: 945 revealed four SNPs from PIK4CA (including rs2072513, rs165862, rs165793: rs165793: rs16579) that passes correction. Observed protective TTA haplotype. G: 540, A:80 C: 1208; T: 322 In conclusion, this study demonstrated that in the Dutch population, PIK4CA is significantly associated with risk of SCZ.

(Continued) 297 298 TABLE I. (Continued) Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary Vorstman et al. 22q11.2DS SCZ: 32 22q11.2DS C (no SCZ): 47 SCZ: 40.0 8.7 The three PIK4CA variants were significantly associated with SCZ in an [2009] rs2072513: rs2072513: C: 30.0 9.4 adult 22q11.2DS sample. This study corroborated findings from the C: 15; T:17 C: 13; T:32 Jungerius et al. [2008] study such that a protective TTA haplotype rs165862: rs165862: was again revealed. The G-allele was observed significantly more in G: 22; T:10 G: 22; T:25 SCZ cases; IQ between rs165793 G- and A-allele carriers did not rs165793: rs165793: differ significantly, indicating that the variant more likely impacts G: 32; A 0 G: 36; A:11 brain function related to 22q11.2DS.

Ikeda et al. 22q11.2DS with psychotic 22q11.2DS without psychotic Cases: 31.0 11.0 This study could not genotype rs165862 and used a proxy SNP instead [2010] episode: 83 SCZ, BPD with episode: 59 C: 31.0 11.0 (rs165863). Did not find any evidence that rs165793 was psychotic features, atypical rs2072513: associated with SCZ despite earlier studies demonstrating psychosis (specifically SCZ) C: 13; T:34 relationships between the pathology and SNP in both 22q11DS and rs2072513: rs165793: general populations. C: 7 (6); T: 13 (10) G: 38; A:9 rs165793: G: 17 (14); A 3 (2)

RTN4R Meng et al. Han Chinese SCZ: 707 C: 689 SCZ: 46.98 13.57 Using both a case-control study and transmission disequilibrium test rs701428 (A/G) [2007] GG: 224, AG: 337, AA: 100 GG: 194, AG: 364, AA: 97 C: 33.61 9.01 (TDT), no significant differences were observed between SCZ and C TDT study: 372 family trios (SCZ TDT probands: group in either allele frequencies or genotype. and biological parents) 24.88 6.82

Budel et al. Case-control: Case-control: N/A Rs701428 was weakly associated with SCZ in the Caucasian group; the [2008] SCZ Caucasian: 336 C Caucasian: 300 A-allele (minor) was more prevalent in the SCZ group. SCZ African-American: 196 C African-American: 100 Frequency of minor A-allele: Frequency of minor A-allele: 0.402 0.344 Chinese Trios: Chinese Trios:

SCZ: 621 C: 501 GENETICS MEDICAL OF JOURNAL AMERICAN

Perlstein et al. 22q11.2DS: 99 C: 21 22q11.2DS: 18.0 1.6 rs701428 was significantly associated with differences in DTI [2014] rs701428: SIB: 26 C: 18.1 1.6 metrics (FA and RD) in the ALIC between the control and the A: 11; G: 28 Total: 47 (groups did not differ 22q11.2DS group. Various DTI metrics were also associated with in DTI metrics and positive prodromal symptoms of psychosis in the patient group. were combined accordingly) Increased FA and decreased RD of the ALIC were also associated No genetic data on controls. with the rs701428 G allele in the hemizygous 22q11.2DS group.

Isobe et al. No patient sample used Healthy individuals: 50 AA/AG: 25.7 6.6 In this healthy sample, variation at rs701428 was associated with [2015] AA/AG:36 GG: 24.1 5.3 changes to the corpus callosum, such that participants carrying the GG:14 A allele had larger total corpus callosum volumes than G homozygous counterparts. A allele carries also had lower RD in the central portion of the corpus callosum.

(Continued) B PART HMSNE AL. ET THOMPSON TABLE I. (Continued) Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary ZDHHC8 Mukai et al. “Extended sample of 389 families N/A Afrikaner probands: rs175174 modifies the retention of intron 4, regulating transcription. rs175174 (A/G) [2004] from the US and South Africa +19 years Observed a gender-specific role, such that the A allele was that included the families significantly over transmitted to females and not males. described in our original study”

Afrikaner parent-proband trio: 93 probands (+19y), met DSM-VI criteria for SCZ/SCZaff + biological parents

Chen et al. Case-control: Case-control: Case control: ZDHHC8 SNPs rs175174 and rs175179 were in linkage disequilibrium [2004b] SCZ: 465 C: 467 SCZ: 46.39 12.87 with one another. Both SNPs were associated with schizophrenia, GG: 223, GA: 200, AA:42 GG: 163, GA: 234, AA: 70 C: 29.10 9.98 however, in contrast with the Mukai et al. [2004] study, the G allele Family-based: Family-based: Family-based: (not A) was the observed schizophrenia susceptibility allele. SCZ: 158 Biological parents of family- SCZ: 23.68 6.60 Estimated allele frequencies (%): based SCZ sample G: 69.0, A: 31.0

Saito et al. Japanese SCZ population: 561 Japanese C population: 529 SCZ (F): 49.6 16.4 Did not observe any association between rs175174 and SCZ. No [2005] GG: 238, GA: 245, AA: 78 C: 529 SCZ (M): 47.0 14.9 gender-specific role was observed either. C (F): 39.7 15.4 C (M): 34.9 12.4

Glaser et al. Bulgarian proband-parent trio: Case-control panel (genotype Bulgarian proband- No association between rs175174 and SCZ was observed in any of the [2005] 474 probands and parents frequency): parent trio: samples. Additionally, there was no evidence for sex-related Provides “transmission pattern of Germany C: 789 Proband: 33 8 transmission differences or transmission distortion in any of the A-allele” AA:.39,AG: .45, GG: .16 Father: 61 9 groups. Case-control panel (genotype Poland C: 311 Mother: 58 9 frequency): AA:.35,AG: .47, GG: .18 Case-control panel: Germany SCZ: 606 Sweden C: 153 Germany: AA: .39, AG: .47, GG:.14 AA:.39,AG: .48, GG: .13 SCZ: 37 11 Poland SCZ: 280 C: 43 16 AA: .36, AG: .47, GG: .17 Poland: Sweden SCZ: 142 SCZ: 32 11 AA: .37, AG: .49, GG: .14 C: 44 10 Sweden: SCZ: 44 17 C: 44 16

Otani et al. Japanese population: BPD-C: 298 BPD: 52.5 13.7 No association was found between ZDHHC8 SNPs (including rs175174) [2005] BPD: 171 GG: 106, GA: 154, AA: 38 BPD-C: 50.6 17.7 and BPD or SCZ, corroborating Saito et al. [2005] study involving GG: 65, GA: 82, AA: 24 SCZ-C: 497 SCZ: 45.6 13.9 another Japanese population. Did not identify any gender effects SCZ: 407 GG: 179, GA: 254, AA: 64 SCZ-C: 53.0 14.7 associating ZDHHC8 with SCZ or BPD. GG: 157, GA: 194, AA:56

(Continued) 299 300 TABLE I. (Continued) Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary Faul et al. German SCZ proband-parent triad: Case-control: German probands: Results of the triad portion of the study indicated that the G-allele was [2005] 204 + biological parents C:186 32.4 8.8 preferentially transmitted to females, while the A-allele was AA: 83, AG: 95, GG:26 AA: 71, AG: 93, GG: 22 Case-control SCZ: transmitted to males when parents are heterozygous. The case- Case-control: 39.2 13.5 control portion of the study however, did not support a gender- SCZ: 433 C: 30.0 10.3 specific role in transmission. AA: 154, AG: 219, GG:60

Ota et al. SCZ: 282 C: 379 SCZ: 36.85 10.64 No association was observed between genotype and schizophrenia. [2013] Minor G-allele frequency: 0.4344 Minor G-allele frequency: C: 41.29 15.34 Variation at rs175174 was significantly associated with frontal, SCZ with MRI: 138 0.4393 cerebellar and parietal lobe GMV. Specifically, homozygous G-allele AA: 52, AG: 60, GG:26 carriers showed reduced GMV in the frontal lobe compared to A-allele carriers. Carriers of the G-allele also presented with decreased GMV in cerebellar hemispheres compared to homozygous A-allele counterparts. Participants carrying the A-allele compared to the homozygous G genotype had reductions in GMV in the posterior parts of the brain.

PRODH Kempf et al. Genetic study: Genetic study: Functional Haplotype The rs4819756 A-allele (minor allele) was negatively associated with rs4819756 (A/G) [2008] SCZ probands: 303 SCZ proband unaffected Demographics: SCZ. Observed a significant positive association between SCZ and a siblings/parents + 370 C VBM / Nback: discovered risk haplotype including the rs4819756 major allele, in Did not include specific allelic Neuroimaging control data set: Reference: 30.9 8.0 addition to two other SNPs on the PRODH gene (rs2870983 major distribution data for either Healthy Controls / 31.7 9.1 allele and rs450046 minor allele), creating the risk haplotype GCC. A group Risk: 38.4 12.4 / negative association (protective haplotype) was also observed with 41.7 9.9 the rs4819756 minor allele, and the rs2870983 and rs450046 Protective: 29.7 7.8 reference alleles (ACT). Reductions in GMV in the control group were / 31.7 9.6 associated with the risk haplotype, though they were confined to the Negative Control neostriatum.

Haplotype GENETICS MEDICAL OF JOURNAL AMERICAN Demographics: VBM / Nback: Reference: 32.5 10.1 / 32.5 11.2 Risk: 27.0 7.3 / 27.5 8.3 Protective: 33.0 7.7 / 32.7 8.3

Ota et al. SCZ: 192 C: 179 SCZ: 35.68 10.38 rs4819756 was not associated with SCZ. The SNP did differentiate from [2014] GG: 108, GA: 55, AA:11 GG: 83, GA: 68, AA: 24 C: 38.29 12.88 the predicted Hardy-Weinberg equilibrium value for both study group. Additionally, rs4819756 was not associated with any neuroanatomic measurements.

(Continued) B PART HMSNE AL. ET THOMPSON TABLE I. (Continued) Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary PRODH x COMT Radoeva et al. 22q11.2DS: 46 (42 F, 45 M) N/A Females: 13.5 3.5 Did not observe any effects between COMT or PRODH variants on ASD [2014] rs4819756: A: 34, G: 49 Males: 13.8 3.4 diagnosis in a 22q11.2DS population. Did observe an interaction, rs4680: A: 35, G:52 such that the low-activity alleles (A-A) of both SNPs, were more likely rs4819756-rs4680: A-A: 12, to result in an ASD diagnosis. G-G:28

COMT Murphy et al. 22q11.2DS: 48 (15 with SCZ: 12 22q11.2DS: Results did not support notion that the Met-allele (low activity) was rs4680 (A/G) [1999] psychosis, 35 without C: 316 31.0 10.0 associated with SCZ. psychosis) SCZ: 34.0 12.0 Non-22q11.2DS: 2 (may have C: 40.0 12.0 smaller deletion, point deletion, other form of mutation)

Egan et al. SCZ: 175 HS: 219 SCZ: 36.1 8.5 COMT genotype was not associated with general intelligence, but was [2001] Family-based trios: 104 C: 55 HS: 35.8 8.8 associated with performance on executive cognition related to the Listed demographics by C: 33.9 9.2 prefrontal cortex, independent of psychiatric diagnosis. The Met genotype, no raw genotype allele was associated better performance cognitively. fMRI was used data to examine genotype on prefrontal physiology while completing a working memory task. Again, the Val allele load was not as efficient as the Met in predicting an efficient physiological response. In a family-based sample, the Val allele was transmitted significantly more in SCZ offspring. Taken together, these data indicate that the Val allele increases risk for SCZ.

Shifman et al. Ashkenazi Jew SCZ C (Ashkenazi Jew + patients N/A Observed a moderately significant association between Val (G/G- [2002] sample with other diseases sample): genotype) and schizophrenia in a population of male Ashkenazi Female: 262 Female: 706 Jews. Allelic and genotypic frequency differences were observed on % G/G : 26.3, % A/G : 54.2, %G/G: 25.9, % A/G : 49.7, a gender basis. % A/A : 19.5 % A/A : 24.4 Male: 458 Male: 2264 % G/G : 31.9, % A/G : 47.8, %G/G: 25.8, % A/G : 52.6, % A/A : 20.3 % A/A : 21.6

Goldberg et al. SCZ: 74 HS: 108 SCZ: 37 8 Homozgyous Met participants had the highest n-back scores, while Val [2003] C: 68 HS: 37 9 homozygotes had the lowest. Observed similar results using 1- and C: 35 10 2- back conditions; siblings had significantly lower performance than unrelated controls. No genotype association was seen on Continuous Performance Test.

Kremer et al. Palestinian Arab SCZ triad (case- C: 77 SCZ: 34.8 13.6 The case-control study revealed an association between the Val (high [2003] control and family-based): GG: 14, GA: 39, AA:24 C: 39.5 15.5 activity) allele and schizophrenia. The case-control study also SCZ probands (from 248 nuclear No familial genotype exposed a sex effect; a weak association was observed between the families: 276 information presented. Val allele and the conferred risk of SCZ in females. GG: 77, GA: 136, AA:63

(Continued) 301 TABLE I. (Continued) 302 Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary Sazci et al. SCZ: 297 C: 341 SCZ: 41.22 9.43 There was a significant difference in both genotype and allele [2004] GG: 88, GA: 146, AA:63 GG: 97, GA: 205, AA:39 C: 40.94 8.11 frequenciesbetweenSCZandC(SCZpatientspresentedwiththe Met allele more frequently). Results indicated a significant association between the homozygous Met (low activity) genotype and SCZ. Observed a difference in allelic distribution and SCZ based on sex (Met allele was observed more frequently in women with SCZ).

Fan et al. SCZ: 862 C: 928 SCZ: 39.8 12.0 Using an association study, found no significant association [2005] GG: 491, GA: 316, AA:55 GG: 505, GA: 358, AA: 65 C: 38.3 11.7 between the Val (high activity) allele and SCZ. Participants with SCZ in the association study did present with the Val allele more frequently, however this relationship was not significant. A meta- analysis was also conducted and concluded that heterogeneity among study groups was significant. No significant associated was observed between SCZ and the Val allele in an Asian nor European population.

Gothelf et al. 22q11.2DS: 24 C (age, gender, ethnicity and 22q11.2DS T1: Observed reductions in prefrontal cortex volume accompanied by [2005] G: 11, A:13 IQ-matched with idiopathic 13.3 3.7 impaired cognition in 22q11.2DS participants with the Met allele, developmental disabilities): 22q11.2DS T2: who already suffer from COMT deficiency by way of hemizygosity 23 18..1 3.4 due to the deletion. The low-activity Met allele was also associated No genotyping data for C C T1: 12.5 3.1 with the emergence of various psychotic symptoms during C T2: 17.9 3.3 adolescence.

Smolka et al. N/A C: 35 C: 40.6 7.6 Activation of the brain from pleasant stimuli was not related to [2005] GG: 10, GA: 16, AA:9 genotype, but activation resulting from negative stimuli was. Specifically, a positive correlation was observed between the number of Met alleles and the limbic system, connected prefrontal regions, and visuospatial attention system. Genotype accounted for

38% of interindividual variance observed in BOLD response due to GENETICS MEDICAL OF JOURNAL AMERICAN unpleasant stimuli.

Kates et al. 22q11.2DS: 58 N/A F (26): 11.0 2.5 Results suggested that rs4680 may impact neuroanatomy, specifically [2006b] G: 32, A: 26 M (32): 11.1 2.9 in the dorsal and orbital prefrontal cortices of children with 22q11.2DS. Females with the Met allele and males with the Val allele presented with reduced orbital frontal volumes and larger dorsal prefrontal volumes than female/Val and male/Met counterparts.

Ohnishi et al. SCZ (Japanese): 47 C (Japanese): 67 SCZ GG: 45.98 15.29 A reduction of volume was observed in limbic and paralimbic systems, [2006] GG: 19, GA: 22, AA:6 GG: 38, GA: 25, AA:13 SCZ Met carrier: neocortical areas, and subcortical regions of participants with SCZ. 43.05 10.57 Compared to Met homozygotes (A/A), Val homozygotes (G/G) GA and AA were combined to C GG: 41.47 13.42 presented with volumetric reductions in the left anterior cingulate ‘Met-carrier group’ in both C Met carrier: cortices and right middle temporal gyrus that reached significance. samples because of small 39.26 10.6 Within the SCZ group, in comparison to the Met homozygous SCZ, Val number of Met-homozygotes homozygotes presented withreduced volumes in the bilateral

anterior cingulate cortices, left amygdala-uncus, right middle B PART temporal gyrus, and left thalamus that again, reached significance. (Continued) HMSNE AL. ET THOMPSON TABLE I. (Continued) Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary Heinz and N/A (Review) N/A (Review) N/A (Review) Examined impact of rs4680 on central processing during various tasks Smolka (pertaining to working memory, attentional control and emotion); [2006] utilized fMRI. Met carriers showed a prefrontal cortex-centered response during working memory tasks, and a cingulate cortex- focused response during attentional control tasks. The authors conclude that processing speed in the prefrontal cortex and cingulate may become more efficient as a result of the Met allele. A positive correlation was observed between reactivity to unpleasant visual stimuli (emotional task) at number of Met alleles (both in the amygdala and in limbic and paralimbic nodes). The authors suggest that this result may be the result of reduced emotional resilience to various states of negative mood in individuals containing a higher Met load. In conclusion, while the Met allele may be advantageous to carriers during both attention-related and working memory tasks, the Val allele is preferable in emotional response pertaining to negative stimuli.

McIntosh et al. Genetic high-risk (HR) SCZ: 78 C: 15 HR-NS (GG, GA, AA): In a sample of participants who were considered genetically at high [2007] HR-NS (No Symptoms): 35 GG:3,GA:8,AA:4 20.9 2.8, risk for SCZ (had a previous diagnosis of SCZ in addition to having GG:6,GA: 17, AA:12 21.6 2.9, family member with SCZ), the Val allele was associated with a higher HR-S (with Symptoms): 32 22.7 3.1 risk of SCZ. Reductions in the density of gray matter were also GG:5,GA: 14, AA:13 HR-S (GG, GA, AA): observed, specifically in the anterior cingulate cortex, of participants HR-SCZ (SCZ): 11 18.9 2.9, with the Val allele. GG:7,GA:3,AA:1 21.5 2.9, 21.6 2.8 HR-SCZ (GG, GA, AA): 19.4 2.1, 18.8 3.7, 16.3 (incalculable) C(GG, GA, AA): 22.0 1.6, 21.3 3.0, 21.9 2.2

Bassett et al. 22q11.2DS: 73 N/A 22q11.2DS: In the SCZ subjects, the Met allele was not significantly more [2007] G: 36, A:37 33.8 10.1 prevalent. The Met allele was associated with significantly worse SCZ (within the 22q11.2DS performance on several frontal cognitive tests and did result in sample): 33 more severe excitement symptoms (e.g. poor impulse control, G:13, A:20 uncooperativeness). Though COMT may be involved in frontal functioning within this sample, it has no association with schizophrenia prevalence.

(Continued) 303 TABLE I. (Continued) 304 Pathology/Patient (N) Average Age Gene/Allele Study (Year) per genotype Control (N)/genotype (Years) Summary Gothelf et al. 22q11.2DS (Total sample, C (22q11.2DS without 22q11.2DS (all): Observed an association between the low-activity Met allele in subjects [2007a] including controls): 55 psychiatric diagnosis): 24 16.6 10.2 with ADHD, and in those with OCD, such that the Met allele was G: 44%, A: 56% G: 16, A:8 With ADHD: observed more frequently in participants with these psychiatric With ADHD: 23 16.68 13.0 diagnoses. 4/5 of the SCZ/SZaff participants carried the Met allele, G:6,A:17 With OCD: however, this did not reach significance compared to controls. With OCD: 14 20.87 13.43 G:3,A:11 With SCZ/SZaff: With SCZ/SZaff: 5 27.09 12.68 G:1,A: 4 C: 15.37 6.87 Note: there were several comorbid diagnoses, which is why n does not add up to 55.

van Amelsvoort 22q11.2DS: 26 N/A 22q11.2DS Val: Significantly larger frontal lobe volumes were observed in Val- et al. G: 14, A:12 30.3 10.6 hemizygous subjects; grey matter density was also increased [2008] 12 people from sample had SCZ 22q11.2DS Met: (cerebellum, brainstem, and parahippocampal gyrus) and white G:5,A:7 37.33 10.6 matter density was reduced (cerebellum) in Val-hemizygotes. COMT genotype did not significantly impact neurocognitive performance.

Honea et al. N/A C: 151 C: 32.9 9.7 Variation at rs4680 resulted in volumetric differences to gray matter [2009] GG: 38, GA: 78, AA:35 in frontal and medial temporal areas. Both Val carriers and homozygotes presented with reduced volume in the left hippocampal and parahippocampal gyrus compared to Met homozygotes.

Gothelf et al. 22q11.2DS: 19 C: 18 22q11.2DS: Having greater reduction in the left dorsal prefrontal cortical (dPFC) [2011] G:8,A: 11 Did not genotype controls Time 1:13.05 3.96 GMV was predictive of psychotic symptoms (at Time 2) in Time 2: 17.92 3.81 22q11.2DS subjects. The subject group presented with significantly C: reduced dPFC at both Times 1 and 2 compared to controls.

Time 1: 13.40 4.04 Reduction was greater in this area in Met individuals compared to GENETICS MEDICAL OF JOURNAL AMERICAN Time 2: 18.31 4.48 Val throughout the time period. Morphometric GM and WM changes in several regions predicted risk for psychotic symptoms (using Leave-out-one Multivariate pattern analysis). In addition to the dPFC, the dorsal cingulum and medial prefrontal cortex were also found to be predictive.

Wang et al. N/A C: 446 (Han Chinese) Age range: 19-21 Both larger hippocampal volumes and better working memory [2013] Subjects with WM task data: GG (WM data; MRI): capabilities were observed in participants with the Val allele. Val GG: 248, GA: 168, AA:27 20.45 0.86; homozygotes performed the best on working memory tasks, Subjects with MRI data: 20.46 0.89 followed by heterozygotes. GG: 173, GA: 128, AA:19 GA (WM data; MRI): 20.32 0.86; 20.34 0.89 AA (WM data; MRI): 20.47 1.0; 20.56 0.92 B PART (Continued) THOMPSON ET AL. 305

regulation of gene expression may act as a buffer for the effects of mutations affecting gene regulation (i.e., mutations in transcrip- tion factors or the sequences of cis-binding sites) in that this RISC system serves to maintain levels of mRNA transcripts within a physiologic range [Brzustowicz and Bassett, 2012]. DGCR8 deficiency in mice has been associated with alterations in the structural morphology of dendritic spines in prefrontal cortex as well as altered prefrontal cortical short-term electrophys- iology and plasticity with impairment in the development of excitatory synapses [Stark et al., 2008; Fenelon et al., 2011; Scho- field et al., 2011; Zhao et al., 2015]. Furthermore, DGCR8 defi- ciency has been linked to a reduction in the elaboration of the dendritic network in the hippocampus [Stark et al., 2008], and has been found to be necessary for neurogenesis in the hippocampi of adult mice [Ouchi et al., 2013]. In mice, DGCR8 was found to regulate 59 miRNAs in prefrontal cortex and 30 in hippocampus [Stark et al., 2008]. DGCR8-/- mice showed downregulation of

volumetric reductions in temporal regionsparahippocampus, (right right hippocampus, fusiform, right bilateral middletemporal gyri), and right superior occipital regions (calcarinea and lingual right regions), parietal regions (precuneus).overlapped The with posterior the cingulate right also parietalsupplementary region. motor Volumetric areas increases and in thewere parietal also and observed occipital in cortices Val homozygotes. 25 mature miRNAs in hippocampus and prefrontal cortex [Zhao Using tensor-based morphometry, Val homozygotes presented with et al., 2015]. These findings suggest that proper miRNA processing is essential to the regulation of genes that are involved in proper prefrontal cortical circuitry development [Schofield et al., 2011; 17.93

Forstner et al., 2013] and in hippocampal circuitry development.

19.85 Guanine Nucleotide Binding Protein (G protein) Beta Polypep- ctive; VBM, voxel based morphometry; WM, white matter (Years) Summary tide 1-Like (GNB1L), located within the 1.5-Mbp critical region of r disorder; C, control; DTI, diffusion tensor imaging; EA, European American; FA, fractional anisotropy; GM, gray Average Age birth: 261.4 MRI: 294. 2 not indicated 22q11.2DS [Zhang et al., 2009; Sun et al., 2015] encodes a G- ) Gestational age at Gestational age at Time of buccal swab protein beta-subunit-like polypeptide [Gong et al., 2000; Funke et al., 2001] that contains six WD40 repeats, without evidence of homology to previously characterized proteins [Funke et al., 2001].

Continued ( One study has found that GNB1L is regulated by a cis-acting genetic 0

:61 variant within the 3 -region of the gene [Sun et al., 2015]. This gene AA is expressed prominently in cerebral cortex, hippocampus, and )/genotype TABLE I. N cerebellum [Riaz, 2015], and is linked to the canonical Wnt : 137,

GA signaling pathway, which has a known role in neurodevelopment [Riaz, 2015]. : 69, In the present study, we have acquired high resolution, anatomic GG MRI scans of the brain on 56 young adults with 22q11.2DS, and have genotyped them for the candidate genes described above,

) for which each participant was hemizygous by virtue of his/her N genetic diagnosis. We measured cortical volume in each participant and examined the extent to which genetic variants, in the presence of volumetric alterations, affected risk for psychosis. In line with our previous findings, we hypothesized that genetic variants of the

per genotype Control ( COMT and RTN4R genes, in the presence of volumetric alterations, would predict risk for psychosis in young adults with this genetic Pathology/Patient ( syndrome. N/A Neonates: 272 MATERIALS AND METHODS Participants

et al. [2014] The genetic, imaging, and psychiatric data presented in this study Knickmeyer are derived from a subsample of participants enrolled in a longi- tudinal study of biomarkers for psychosis in 22q11.2DS [Kates et al., 2004]. This subsample consists of participants with 22q11.2DS who returned for the fourth time point of this longi- tudinal study. Participants were recruited through the SUNY Upstate International Center for Evaluation, Treatment, and matter; GMV, gray matter volume; HS, healthy sibling; NBACK, a working memory task; RD, radial diffusivity; SCZ: schizophrenia; SCZaff, schizoaffe AfA, African American; ALIC, anterior limb of the internal capsule; ASD, autism spectrum disorder; BOLD, blood oxygen level-dependent; BPD, bipola Gene/Allele Study (Year) Study of Velo-Cardio-Facial Syndrome at SUNY Upstate Medical 306 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

University and the surrounding community. Fluorescence in situ Germany). An ultrafast gradient echo 3D sequence (MPRAGE) hybridization (FISH) was used to confirm a deletion at 22q11.2 in was used with PAT k-space-based algorithm GRAPPA and the the patient group. Participants provided informed consent prior to following parameters: echo time ¼ 3.31 ms; repetition time ¼ any data collection. The Institutional Review Board at SUNY 2530 ms; matrix size 256 256; field of view ¼ 256 mm, slice Upstate Medical University approved all procedures of this study. thickness ¼ 1 mm [McCarthy et al., 2015].\ Since genetic data were only collected for the patient group, this report is specifically focused on associations between genotypes, neuroanatomic variation and psychiatric status in individuals with Image Analysis 22q11.2DS. As shown in Table II, fifty-six individuals (30 males) The resulting images were processed and analyzed using Freesurfer were diagnosed with a microdeletion at 22q11 with an average age (FS; freesurfer.net). FS is a software suite used for both the of 20.88 (SD ¼ 2.31). All participants ranged in age from 17 to 25. processing and analysis of brain MR images. The software was Two participants were excluded from the PIK4CA (rs2072513 and operated with Ubuntu 12.04 on a Dell Optiplex machine. The first rs165862) analyses after failed attempts to genotype the SNPs of step in the FS pipeline was used to remove non-brain tissue. Due to interest. inaccuracies in this FS module, the brainmask was then imported into 3DSlicer and subjected to manual edits using the following MRI Acquisition cited protocol [Subramaniam et al., 1997]. The resulting skull stripped brains were then aligned along the anterior and posterior T1 weighted magnetic resonance imaging scans of the brain were commissure axis using a cubic spline transformation. The final acquired in the sagittal plane on a 3T Siemens Magnetom Trio Tim images remained the same resolution as the original data, isotropic scanner (syngo MR B17, Siemens Medical Solutions, Erlangen, voxels (1 mm3). Following the initial preprocessing steps, the images were sub- mitted to automated brain segmentation using FS. Images under- TABLE II. Participant Demographics went two reconstruction streams that comprise the automated brain segmentation process that FS offers (http://surfer.nmr.mgh. Demographic variable N harvard.edu/fswiki/recon-all). Data were then manually inspected Gender (M/F) 30/26 for inaccuracies in segmentation. The detailed manual intervention Age (SD) 20.88 (2.31) process we followed has been previously described [McCarthy Full Scale IQ 74.5 (11.8) et al., 2015]. Following manual intervention, the imaging data Positive prodromal symptoms score (SD) 3.61 (6.11) were run through the reconstruction stage again. This process was Ethnicity (% not hispanic/% hispanic) 92.9%/7.1% repeated up to four times contingent upon the number and severity Race (N) of errors in segmentation. After all necessary edits were made, the Asian 1 data were run through a final reconstruction step. (Supplementary Caucasian 51 More than one race 2 Fig. S1 visually compares the rendered view of cortical reconstruc- Unknown 2 tion of the averaged brains used to create the Freesurfer template to Gene/SNP (N) the cortical reconstruction of three individuals in our sample.) COMT (rs4680) A-24; G-32 Cortical thickness, surface area, and volume measurements were PRODH (rs4819756) A-23; G-33 extracted for 34 regions of interest (ROIs) in each hemisphere ZDHHC8 (rs175174) A-28; G-28 [Desikan et al., 2006]. The data for each hemisphere was summed RTN4R (rs701428) A-17; G-39 to provide values representative of the whole brain. PIK4CA (rs2072513) C-18; T-37 PIK4CA (rs165862) G-29; T-26 PIK4CA (rs165793) A-13; G-43 Genotyping DGCR8 (rs3757) A-14; G-42 Blood was collected in PAXGene Blood DNA Tubes, and DNA was GNB1L (rs5746832) A-26; G-30 GNB1L (rs2269726) C-26; T-29 extracted using the PAXGene Blood DNA Kit. The genotyping of Presence of a psychiatric diagnosis (N [%]) the rs701428 SNP of the Nogo-66 Receptor gene was performed as Mood disorder 19 (33.9) described previously [Perlstein et al., 2014]. The genotyping of the Anxiety disorder 10 (17.9) PIK4CA SNPs rs165793, rs165862, and rs2072513, ZDHHC8 SNP ADHD 20 (35.7) rs175174, GNB1L SNPs rs5746832 and rs2269726, and PRODH Prodromal/overt psychosis 16 (28.6) SNP rs4819756 were also completed via direct sequencing. The Any psychiatric diagnosis 39 (69.6) majority of COMT rs4860 samples were genotyped as previously Medication usage (N[%]) described [Coman et al., 2010], but some samples were genotyped Stimulants/straterra 10 (17.9) via an alternative method. The DNA region containing the poly- Anti-depressants/anti-anxiety 14 (25) morphism was amplified via PCR though the primers and con- Mood stabilizers 3 (5.4) ditions for each SNP varied. Each reaction had a volume of 25 ml Antipsychotics 8 (14.3) m m Any medication 23 (41.1) and included 1 lofa10 M mixture of forward and reverse primers, 5 ml Taq 5x Master Mix (New England Biolabs, M0285S), nuclease-free water and genomic DNA. Each PCR began with an THOMPSON ET AL. 307 initial denaturation step at 95˚C for 2 min, and 30–40 cycles at 95˚C Psychiatric Interview for 30 sec, followed by 30 sec at the unique annealing temperature Psychiatric interviews were administered with the Structured for the particular reaction, and 30 sec at 68˚C. The primers and Clinical Interview for DSM-IV-TR axis I disorders [SCID; First conditions for each reaction are described in Table III. Following et al., 2002], by two doctoral level clinicians. The Structured PCR, the amplified product was visualized on a 2% agarose gel to Interview for Prodromal Syndromes [SIPS; Miller et al., 2003] confirm the product size was correct for the SNP of interest. was administered to all participants to determine if ultra high risk COMT Genotyping of rs4680 and DGCR8 rs3757 were slightly symptoms of psychosis were present. The same trained, doctoral- different than the other SNPs. For the COMT SNP, amplification level psychiatrist or clinical psychologist who administered the was carried out with an initial denaturation step at 95˚C for 2 min, SCID also administered the SIPS to all participants. Inter-rater followed by 30 cycles of the following: 94˚C for 1 min, 62˚C for reliability, which was based on five consecutive SIPS interviews and 1 min, and 72˚C for 1 min [Inoue et al., 2005; Baclig et al., 2012]. was assessed with intra-class correlation coefficients, was 0.90. Only Because two bands were present on the gel following amplification, the Positive Symptom subscale was used for the present analyses. we excised the correct product from the gel and continue purifica- tion using the MinElute Gel Extraction Kit (Qiagen, Cat# 28004). The manufacturers protocol was followed precisely. The amplicon Statistical Analysis containing the SNP was then purified using the Clontech DNA Since cortical volume is the product of surface area and cortical Amplification Cleanup Kit (Ref. 636975). Purified DNA was then thickness, we initially investigated the effect of allelic variation on submitted for sequencing to GeneWiz (http://www.genewiz.com). cortical gray and white matter volumes of each candidate SNP, Since the participants in this study did have some form of micro- using multivariate analysis of variance (MANOVA). In order to deletion on chromosome 22, they were hemizygous for one of two reduce the data for statistical analyses, regions of interest within alleles for each SNP. each lobar region of the brain were summed (based on the Desikan Genotyping for the DGCR8 rs3757 SNP was performed in Atlas) in order to create volumes overall frontal, parietal, temporal, duplicate on 384 well plates using a commercial validated TaqMan occipital lobes, cingulate, insula. Since we conducted 10 MAN- Genotyping Assay (Thermo Fisher catalog #4351379) with Taq- OVA’s (i.e., one for each candidate SNP), we only report results for 1 Man Genotyping Master Mix (Thermo Fisher catalog #4371353). which Wilks’ Lambda values were significant at P < 0.005 (which This was prepared and run according to the recommended pro- represent a Bonferroni correction based on the 10 SNPs). We also 1 cedures in the TaqMan Genotyping Master Mix Protocol (Ap- conducted followup analyses to determine whether cortical thick- plied Biosystems publication #4371131, Rev. B) using a BioRad ness or surface area (or both) accounted for the significant SNP- CFX 384 Real Time PCR System with end-point allele discrimina- lobar volume associations that we found. Finally, if allelic variation tion performed with the BioRad CFX Manager software. of a particular SNP was found to significantly affect the volume of a

TABLE III. Oligonucleotide Primers and Conditions Used in PCR Amplification

Annealing Number of Product size Gene SNP Polarity Nucleotide sequence (50 to 30) temperature (˚C) cycles (bp) Allele PIK4CA rs165793 Forward AGT CAC AAG GGC ATG AGA CAC 60 30 374 A/G Reverse TCT AGC AGT GCA GGG AAC TTG rs165862 Forward TTG AGC AGC TGA CAT TGA CAC 60 30 243 G/T Reverse TGG TCC AGG TGA GGA TAT TTG rs2072513 Forward CAG CAC TAT CCT TTC CTC GTG 55 30 397 C/T Reverse AGA GCA GCC AGC AGA TGT TTA RTN4R rs701428 Forward CCT GTG CAT ATT TCT GCC TGT 55 30 353 A/G Reverse GAT GAG GGA CCC TGT TCC TAC ZDHHC8 rs175174 (V26) Forward AGG GAG ATT GAG CCT CAG AAG 55 30 379 A/G Reverse CAA CAC AGC TGC ACA AGC TAC PRODH rs4819756 Forward AGG GGT ATA GCC AGC AAA GAG 62 40 443 A/G Reverse GGC CTT TGT TAA GGA GAC TGC COMT rs4680 (val/met) Forward ACT GTG GCT ACT CAG CTG TG 62 30 169 A/G Reverse CCT TTT TCC AGG TCT GAC AAa GNB1L rs2269726 Forward AGG CGG TGA TTT CAA GAC TTT 54 40 309 C/T Reverse CAG GAG CCT AGA GCA GTG TTG rs5746832 Forward GTC AGG AGG TCA AGT GTG CAT 52 40 382 A/G Reverse GGG AAG TCC ACA TAG GAA AGC DGCR8 rs3757b

aInoue et al. [2005]. bInformation unavailable; please see text. 308 AMERICAN JOURNAL OF MEDICAL GENETICS PART B specific lobe, we repeated the analyses, with the inclusion of age, indicated that this association was driven by occipital lobe volume gender, full-scale IQ, presence of a lifetime comorbid psychiatric (F[df: 1,56] ¼ 9.62; P ¼ 0.003; eta2 ¼ 0.149). This association diagnosis and medication use as covariates. Moreover, for signifi- remained significant after separately including age (age effect: cant SNP-lobar associations, we followed up with a MANOVA to P ¼ 0.67), full scale IQ (IQ effect: P ¼ 0.001), gender (gender determine which lobar subregions accounted for the finding. effect: P ¼ 0.01), presence of a lifetime comorbid psychiatric We then used a multivariate, zero-inflated Poisson (ZIP) regres- diagnosis (comorbid effect: P ¼ 0.18), and medication usage (med- sionmodeltoexaminetheeffectofSNPandlobarcorticalvolumeon ication effect: P ¼ 0.14) in the model as covariates. Follow-up positive symptom scores. We selected this model due to the distri- analyses indicated that allelic variation in both cortical thickness bution of our SIPS scores. Both SNP and lobar volume were (F[df: 1,56] ¼ 10.87; P ¼ 0.002; eta2 ¼ 0.168) and surface area examined as independent variables. We only performed the multi- (F[df: 1,56] ¼ 4.35; P ¼ 0.042; eta2 ¼ 0.042) contributed to the variate ZIP regression analyses on lobar volumes that had previously association between allelic variation of this SNP and volume. passed the Bonferroni correction threshold in our MANOVAs. All Notably, however, the effect size for cortical thickness was more data were analyzed using SPSS v. 23 and STATA v. 12. than twice as large as surface area. We did not observe any significant associations between SNP variants and cortical white matter volumes. RESULTS As expected, the MANOVA that was conducted to determine Associations Between Allelic Variation and which subregions of the occipital lobe drove the association between this RTN4R SNP and occipital lobe volume was significant (Wilks’ Neuroanatomic Structure Lambda ¼ 0.77; P ¼ 0.01). Planned, follow-up univariate analyses After correcting for the analysis of all 10 SNP’s, only allelic variation indicated that the association was driven by volumes of the cuneus of RTN4R, rs701428 significantly affected cortical volumes (Wilks’ (F[df: 1,56] ¼ 10.55; P ¼ 0.002; eta2 ¼ 0.163) and the lingual gyrus Lambda ¼ 0.69; P ¼ 0.002). Planned, follow-up univariate analyses (F[df: 1,56] ¼ 11.38; P ¼ 0.001; eta2 ¼ 0.174) (See Fig. 1).

FIG. 1. This figure depicts the mean and data distribution of rs701428 alleles A and G for the cuneus and lingual gyrus. Significant regions of interest are also visualized on the right and left hemispheres of the medial surface of the brain, with the cuneus represented in dark pink and the lingual gyrus represented in light pink. THOMPSON ET AL. 309

Associations Between Allelic Variation, Gothelf et al., 2007b], temporal [Eliez et al., 2001; Chow et al., 2002; Neuroanatomic Structure and Positive Ultra Campbell et al., 2006], and anterior cingulate cortices [Dufour et al., 2008]. While the relationship between several of the genes we High Risk Symptoms examined (e.g., COMT, PRODH, and ZDHHC8) and neuroana- RTN4R, rs 701428 did not, independently, predict positive ultra tomic volume has been explored in both the general population high risk symptoms scores on the SIPS (z ¼ 0.63; P ¼ 0.526). In and 22q11.2DS, very few studies have attempted to examine the contrast, occipital lobe volumes significantly predicted SIPS scores effect of RTN4R on morphological changes in the brain. RTN4R (z ¼3.22; P ¼ 0.001). Accordingly, we conducted a multivariate (also known as the Nogo-66 receptor gene)has previously been regression analysis to determine the effect of one variable in the implicated in schizophrenia [Budel et al., 2008]. That is, previous presence of the other. When both RTN4R, rs701428 and occipital evidence suggests that the A allele predisposes individuals in the volume were included in the regression model, we observed that general population with the greatest risk of developing schizophre- both were significantly associated with ultra high risk symptoms nia [Budel et al., 2008]. Functionally, it is associated with the of psychosis (RTN4R: z ¼ 3.45; P ¼ 0.001; occipital lobe volume: inhibition of axonal growth [Fournier et al., 2001] and synaptic z ¼4.42; P < 0.001). We followed up this analysis by running plasticity [McGee et al., 2005]. Mouse orthologs have demon- separate models testing the association between occipital lobe strated that the gene is expressed throughout development, from volume and SIPS scores for individuals with the rs701428/A allele embryogenesis through adulthood, where it continues to be and those with the G allele, respectively. For individuals with the A expressed in the forebrain and other tissues outside the brain allele, the association between occipital lobe volumes and SIPS [Maynard et al., 2003]. The Nogo-66 receptor gene is critical in scores was not significant (z ¼1.03; P ¼ 0.305). However, for myelin-associated axonal growth and, as noted above, we have individuals with the G allele, the association was highly significant previously demonstrated that allelic variation in a functional (z ¼5.08; P < 0.001). polymorphism (rs701428) of this gene affects white matter micro- As noted above, none of the SNP’s that we examined signifi- structure [Perlstein et al., 2014]. Our current observation of a cantly affected cortical white matter volumes. However, in light of robust association between allelic variation of this SNP and occip- the results reported above, and our previous observations of an ital gray matter volume extend our previous findings. association between this RTN4R SNP, white matter microstruc- Although the role of RTN4R in myelin development and plasticity ture, and ultra high risk symptoms of psychosis [Perlstein et al., of white matter microstructure has been well documented, recent 2014], we conducted an exploratory analysis of the association reports also suggest that this gene is also expressed in neurons between this SNP, white matter occipital volume, and symptoms of [Petrinovic et al., 2010; Zatorre et al., 2012]. These findings suggest psychosis. When both RTN4R, rs701428 and white matter occipital that this gene may lead to structural modifications of the brain volume were included in the regression model, we again observed through its effect on both white matter and gray matter development that both were significantly associated with ultra high risk symp- [Zatorre et al., 2012]. This molecular “coupling” of white matter and toms of psychosis (RTN4R: z ¼ 2.55; P ¼ 0.01; occipital lobe gray matter [Zatorre et al., 2012] may explain, in part, associations volume: z ¼4.72; P < 0.001). Follow-up analysis indicated that that have been reported between decreased FA and gray matter the association between white matter occipital volume and symp- reductions in neuroanatomically similar regions in schizophrenia toms of psychosis was not significant for individuals with the A samples [Douaud et al., 2007; Spoletini et al., 2009]. In addition, allele (z ¼1.12; P ¼ 0.26), but the association was highly signifi- alterations in axonal integrity could lead to changes in neuronal cant for individuals with the G allele (z ¼5.39; P < 0.001). integrity [Dutta and Trapp, 2007], supporting the notion that the Accordingly, in this sample, both gray and white matter occipital variance in gray matter volume could also be a function, in part, of lobe volumes predicted ultra high risk symptoms of psychosis in aberrant white matter development due to genetic variation in the presence of the G allele of RTN4R, rs701428. candidate genes. In rodent studies, absent or reduced dosages of NgR1, which is the DISCUSSION protein encoded by RTN4R, have been associated with impairments in spatial working memory [Budel et al., 2008; Karlen et al., 2009], To our knowledge, this is the first study to investigate the effects of which functional imaging studies suggest is subserved by the dorsal allelic variation due to SNPs in multiple psychiatric risk genes and stream, a neural circuit consisting of occipital-parietal connections neuroanatomic structure on ultra high risk symptoms of psychosis [Haxby et al., 1991; Ungerleider et al., 1998]. Both the cuneus and the in a 22q11.2DS sample. In our study, allelic variation resulting from lingual gyrus, the specific occipital regions that we found to be RTN4R, rs701428 impacted occipital gray matter volume. More- differentially affected by the RTN4R SNP that we examined, have over, we found that both gray and white matter occipital lobe been implicated in visual spatial functions of location identification volumes predicted ultra high risk symptoms of psychosis only in [Haxby et al., 1991] motion responsiveness [Sunaert et al., 1999] and the presence of the G allele of this SNP. directional discrimination [Cornette et al., 1998]. Importantly, def- Structural abnormalities within the brain are frequently icits in visual spatial processing have been well-described in individ- reported in 22q11.2DS, but the mechanism by which these changes uals with 22q11.2DS [Bearden et al., 2001; Simon et al., 2008; Wong occur is still largely unknown. The most frequently reported et al., 2014]. Accordingly, allelic variation in SNP rs701428 may lead changes include increased insular volumes [Simon et al., 2005] to myelin-associated alterations in occipital-parietal circuits that accompanied by reductions in the orbitofrontal [Kates et al., comprise the dorsal stream, potentially accounting for the visual 2011b], parieto-occipital [Kates et al., 2001; Bearden et al., 2007; spatial deficits that characterize individuals with 22q11.2DS. 310 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

As noted above, the prevalence of psychosis is as high as 30–40% one of the last steps in white matter development, potentially in individuals with 22q11.2DS [Schneider et al., 2014]. The rela- explaining, in part, why idiopathic schizophrenia develops during tionship between brain tissue and psychosis has been extensively late adolescence or early adulthood. Accordingly, this notion may researched in 22q11.2DS [Kates et al., 2006a,b, 2011a,b]. However, also explain, in part, the pathophysiology of psychosis in individuals as previously noted, there have been far fewer studies that have with 22q11.2DS as well. examined the relationship between particular genes and psychosis A limitation of the current study lies within the relatively small in individuals with this syndrome. We observed that for individuals sample size. The resulting genetic subgroups were also relatively with the A allele for rs701428, occipital volumes were not associated small. Since much of these data are novel, studies of similar nature with UHR symptoms of psychosis, whereas for individuals with the should be conducted using larger sample sizes. Additionally, G allele, the association was robust. The finding that occipital lobe although our sample consists of participants in late adolescence volume predicts UHR symptoms in the presence of the G allele is and early adulthood, the younger participants may still be at risk for consistent with our previously published DTI study, in which we developing psychosis. Because we did not genotype ancestral concluded that the G allele increased risk for psychosis through its informative SNPs, we cannot rule out population stratification alteration of white matter microstructure [Perlstein et al., 2014]. In as an additional artifact that influences our results. contrast, Budel et al. [2008] found that in samples of individuals The findings from this study suggest that allelic variation of with idiopathic schizophrenia, the A allele of re701428 was the risk RTN4R, a relatively under-studied gene within the 22q11 locus, allele. Accordingly, more definitive studies with larger samples of may alter macroscopic volumes of posterior regions of the brain. individuals with 22q11.2DS are clearly indicated (and are currently Additionally, our research suggests that the G allele of RTN4R, being conducted by the International Consortium For Brain and rs401728 is a risk allele in individuals with this genetic syndrome, Behavior in 22q11.2 Deletion Syndome). in that we observed that occipital lobe volumes were robustly Previous examinations of candidate genes that may contribute associated with UHR symptoms of psychosis in the presence of to the pathophysiology of risk for psychosis in 22q11.2DS have this SNP’s G allele. Given the role of this gene in the regulation focused on genes that are associated with dopamine signaling myelin-mediated axonal growth, additional studies should in- (COMT), glutamate metabolism (PRODH), neuronal architecture corporate diffusion tensor imaging to determine whether alter- (ZDHHC8 and DGCR8), and microRNA processes (DGCR8) ations to the white matter microstructure are in fact driving [Karayiorgou et al., 2010], all of which we examined as well. macroscopic brain volume changes. Moreover it would be im- Although we did not observe any significant effects of allelic portant to explore the extent to which interactions among the variation in these SNP’s on neuroanatomic volume, and therefore genes within and beyond this region are influencing neurodevel- did not explore the extent to which these SNPs were associated with opment and risk for psychosis. As noted above, this may bear risk for psychosis, our findings do not negate the potential contri- responsibility for some of the phenotypic variability observed in bution of these genes. Since these SNPs have been previously linked 22q11.2DS. (albeit inconsistently) to either neuroanatomic structure [Kates et al., 2006b; Ota et al., 2013], psychosis [Liu et al., 2002; Chen et al., ACKNOWLEDGMENTS 2004b; Jungerius et al., 2008; Vorstman et al., 2009], or both [Gothelf et al., 2005; McIntosh et al., 2007; Kempf et al., 2008], This work was supported by funding from the National Institutes our negative results could stem from the small size of our sample. of Health grant (MH064824) to W.R.K. This may be the case in particular for SNPs in which the minor allele frequency is relatively small. Moreover, in the case of the COMT polymorphism, bioavailability of the protein product as well as REFERENCES phenotypic features may be influenced by gender [Gogos et al., 1998; Kates et al., 2006b], reducing the power to detect allelic Baclig MO, Predicala RZ, Mapua CA, Lozano-Kuhne JP, Daroy ML, differences to an even greater extent. Finally, there is evidence that Natividad FF, Javier FO. 2012. Allelic and genotype frequencies of gene—gene interactions may influence phenotypes in 22q11.2DS, catechol-O-methyltransferase (Val158Met) and CYP2D6 10 (Pro34Ser) such as interactions between COMT and PRODH polymorphisms single nucleotide polymorphisms in the philippines. Int J Mol Epidemiol Genet 3:115–121. [Radoeva et al., 2014; de Koning et al., 2015], potentially explaining our negative results to some extent. Baptiste DC, Fehlings MG. 2007. Update on the treatment of spinal cord To the extent that the pathophysiology of psychosis in 22q11.2DS injury. Prog Brain Res 161:217–233. is due, in part, to alterations in neuronal architecture (as noted Bassett AS, Caluseriu O, Weksberg R, Young DA, Chow EW. 2007. above), our results extend our understanding of the specific genes Catechol-O-methyl transferase and expression of schizophrenia in 73 adults with 22q11 deletion syndrome. Biol Psychiatry 61:1135–1140. that may be contributing to these alterations, and suggest that RTN4R constitutes a susceptibility gene for risk for psychosis in Bearden CE, van Erp TG, Dutton RA, Tran H, Zimmermann L, Sun D, individuals with this syndrome. RTN4R may be contributing to the Geaga JA, Simon TJ, Glahn DC, Cannon TD, Emanuel BS, Toga AW, Thompson PM. 2007. Mapping cortical thickness in children with pathophysiology of psychosis through its potential alteration of 22q11.2 deletions. Cereb Cortex 17:1889–1898. axonal developmental processes. Specifically, Budel et al. [2008] Bearden CE, Woodin MF, Wang PP, Moss E, McDonald-McGinn D, suggest that RTN4R variants result in loss-of-function, thus reduc- Zackai E, Emannuel B, Cannon TD. 2001. The neurocognitive phenotype ing thecapacity of theprotein to mediate myelin inhibition of axonal of the 22q11. 2 deletion syndrome: Selective deficit in visual-spatial growth.They arguethattheprocess ofmyelin-mediated inhibition is memory. J Clin Exp Neuropsychol 23:447–464. THOMPSON ET AL. 311

Bender H, Almashanu S, Steel G, Hu C, Lin W, Willis A, Pulver A, Valle D. genotype on frontal lobe function and risk for schizophrenia. Proc Natl 2005. Functional consequences of PRODH missense mutations. Am J Acad Sci USA 98:6917–6922. Hum Genet 76:409–420. Eliez S, Blasey CM, Schmitt EJ, White CD, Hu D, Reiss AL. 2001. Veloc- Bertocci B, Miggiano V, Da Prada M, Dembic Z, Lahm HW, Malherbe P. ardiofacial syndrome: Are structural changes in the temporal and mesial 1991. Human catechol-O-methyltransferase: Cloning and expression of temporal regions related to schizophrenia? Am J Psychiatry 158:447–453. the membrane-associated form. Proc Natl Acad Sci USA 88:1416–1420. Fan J, Zhang C, Gu N, Li X, Sun W, Wang H, Feng G, Clair DS, He L. 2005. Brzustowicz LM, Bassett AS. 2012. MiRNA-mediated risk for schizophre- Catechol-O-methyltransferase gene Val/Met functional polymorphism nia in 22q11. 2 deletion syndrome. Front Genet 3:291. and risk of schizophrenia: A large-scale association study plus meta- analysis. Biol Psychiatry 57:139–144. Budel S, Padukkavidana T, Liu BP, Feng Z, Hu F, Johnson S, Lauren J, Park JH, McGee AW, Liao J, Stillman A, Kim JE, Yang BZ, Sodi S, Gelernter J, Faul T, Gawlik M, Bauer M, Jung S, Pfuhlmann B, Jabs B, Knapp M, Stober Zhao H, Hisama F, Arnsten AF, Strittmatter SM. 2008. Genetic variants G. 2005. ZDHHC8 as a candidate gene for schizophrenia: Analysis of a of nogo-66 receptor with possible association to schizophrenia block putative functional intronic marker in case-control and family-based myelin inhibition of axon growth. J Neurosci 28:13161–13172. association studies. BMC Psychiatry 5:35. Campbell LE, Daly E, Toal F, Stevens A, Azuma R, Catani M, Ng V, van Fenelon K, Mukai J, Xu B, Hsu PK, Drew LJ, Karayiorgou M, Fischbach GD, Amelsvoort T, Chitnis X, Cutter W,Murphy DG, Murphy KC. 2006. Brain Macdermott AB, Gogos JA. 2011. Deficiency of Dgcr8, a gene disrupted by and behaviour in children with 22q11.2 deletion syndrome: A volumetric the 22q11.2 microdeletion, results in altered short-term plasticity in the and voxel-based morphometry MRI study. Brain 129:1218–1228. prefrontal cortex. Proc Natl Acad Sci USA 108:4447–4452. Chen J, Lipska BK, Halim N, Ma QD, Matsumoto M, Melhem S, Kolachana Forstner AJ, Degenhardt F, Schratt G, Nothen€ MM. 2013. MicroRNAs as the BS, Hyde TM, Herman MM, Apud J. 2004a. Functional analysis of cause of schizophrenia in 22q11. 2 deletion carriers, and possible impli- genetic variation in catechol-O-methyltransferase (COMT): Effects on cations for idiopathic disease: A mini-review. Front Mol Neurosci 6:47. mRNA, protein, and enzyme activity in postmortem human brain. Am J Fournier AE, GrandPre T, Strittmatter SM. 2001. Identification of a Hum Genet 75:807–821. receptor mediating nogo-66 inhibition of axonal regeneration. Nature Chen WY, Shi YY, Zheng YL, Zhao XZ, Zhang GJ, Chen SQ, Yang PD, He L. 409:341–346. 2004b. Case-control study and transmission disequilibrium test provide First M, Spitzer R, Gibbon M, Williams J. 2002. Structured clinical consistent evidence for association between schizophrenia and genetic interview for DSM-IV-TR axis I disorders, research version, non-patient variation in the 22q11 gene ZDHHC8. Hum Mol Genet 13:2991–2995. edition (SCID-I/NP). New York: New York State Psychiatric Institute. Chow EW, Zipursky RB, Mikulis DJ, Bassett AS. 2002. Structural brain FunkeB,PanditaRK,MorrowBE.2001.Isolationandcharacterizationof abnormalities in patients with schizophrenia and 22q11 deletion syn- a novel gene containing WD40 repeats from the region deleted in Velo- drome. Biol Psychiatry 51:208–215. cardio-facial/DiGeorge syndrome on chromosome 22q11. Genomics Coman IL, Gnirke MH, Middleton FA, Antshel KM, Fremont W, Higgins 73:264–271. AM, Shprintzen RJ, Kates WR. 2010. The effects of gender and catechol Glaser B, Schumacher J, Williams HJ, Jamra RA, Ianakiev N, Milev R, O-methyltransferase (COMT) Val108/158Met polymorphism on emo- Ohlraun S, Schulze TG, Czerski PM, Hauser J. 2005. No association tion regulation in velo-cardio-facial syndrome (22q11. 2 deletion syn- between the putative functional ZDHHC8 single nucleotide polymor- drome): An fMRI study. Neuroimage 53:1043–1050. phism rs175174 and schizophrenia in large european samples. Biol Cornette L, Dupont P, Rosier A, Sunaert S, Van Hecke P, Michiels J, Psychiatry 58:78–80. Mortelmans L, Orban GA. 1998. Human brain regions involved in Gogos JA, Morgan M, Luine V, Santha M, Ogawa S, Pfaff D, Karayiorgou direction discrimination. J Neurophysiol 79:2749–2765. M. 1998. Catechol-O-methyltransferase-deficient mice exhibit sexually de Koning MB, van Duin ED, Boot E, Bloemen OJ, Bakker JA, Abel KM, dimorphic changes in catecholamine levels and behavior. Proc Natl Acad van Amelsvoort TA. 2015. PRODH rs450046 and proline x COMT Sci USA 95:9991–9996. Val158Met interaction effects on intelligence and startle in adults with 22q11 deletion syndrome. Psychopharmacology (Berl) 232:3111–3122. Goldberg TE, Egan MF, Gscheidle T, Coppola R, Weickert T, Kolachana BS, Goldman D, Weinberger DR. 2003. Executive subprocesses Desikan RS, Segonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, in working memory: Relationship to catechol-O-methyltransferase Buckner RL, Dale AM, Maguire RP, Hyman BT. 2006. An automated Val158Met genotype and schizophrenia. Arch Gen Psychiatry 60:889– labeling system for subdividing the human cerebral cortex on MRI scans 896. into gyral based regions of interest. Neuroimage 31:968–980. Gong L, Liu M, Jen J, Yeh ET. 2000. GNB1L, a gene deleted in the critical Douaud G, Smith S, Jenkinson M, Behrens T, Johansen-Berg H, Vickers J, region for DiGeorge syndrome on 22q11, encodes a G-protein beta- James S, Voets N, Watkins K, Matthews PM, James A. 2007. Anatomi- subunit-like polypeptide. Biochim Biophys Acta 1494:185–188. cally related grey and white matter abnormalities in adolescent-onset schizophrenia. Brain 130:2375–2386. Gothelf D, Eliez S, Thompson T, Hinard C, Penniman L, Feinstein C, Kwon H, Jin S, Jo B, Antonarakis SE. 2005. COMT genotype predicts longi- Dufour F, Schaer M, Debbane M, Farhoumand R, Glaser B, Eliez S. 2008. tudinal cognitive decline and psychosis in 22q11. 2 deletion syndrome. Cingulate gyral reductions are related to low executive functioning and Nat Neurosci 8:1500–1502. psychotic symptoms in 22q11. 2 deletion syndrome. Neuropsychologia 46:2986–2992. Gothelf D, Hoeft F, Ueno T, Sugiura L, Lee AD, Thompson P, Reiss AL. 2011. Developmental changes in multivariate neuroanatomical patterns Dutta R, Trapp BD. 2007. Pathogenesis of axonal and neuronal damage in that predict risk for psychosis in 22q11. 2 deletion syndrome. J Psychiatr multiple sclerosis. Neurology 68:S22–S31; discussion S43-S54. Res 45:322–331. Edelmann L, Pandita RK, Morrow BE. 1999. Low-copy repeats mediate the Gothelf D, Michaelovsky E, Frisch A, Zohar AH, Presburger G, Burg M, common 3-mb deletion in patients with velo-cardio-facial syndrome. Aviram-Goldring A, Frydman M, Yeshaya J, Shohat M, Korostishevsky Am J Hum Genet 64:1076–1086. M, Apter A, Weizman A. 2007a. Association of the low-activity COMT Egan MF, Goldberg TE, Kolachana BS, Callicott JH, Mazzanti CM, Straub 158Met allele with ADHD and OCD in subjects with velocardiofacial RE, Goldman D, Weinberger DR. 2001. Effect of COMT Val108/158 met syndrome. Int J Neuropsychopharmacol 10:301–308. 312 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

Gothelf D, Penniman L, Gu E, Eliez S, Reiss AL. 2007b. Developmental Kates WR, Burnette CP, Jabs EW, Rutberg J, Murphy AM, Grados M, trajectories of brain structure in adolescents with 22q11. 2 deletion Geraghty M, Kaufmann WE, Pearlson GD. 2001. Regional cortical white syndrome: A longitudinal study. Schizophr Res 96:72–81. matter reductions in velocardiofacial syndrome: A volumetric MRI analysis. Biol Psychiatry 49:677–684. Green T, Gothelf D, Glaser B, Debbane M, Frisch A, Kotler M, Weizman A, Eliez S. 2009. Psychiatric disorders and intellectual functioning through- Kates WR, Miller AM, Abdulsabur N, Antshel KM, Conchelos J, Fremont out development in velocardiofacial (22q11. 2 deletion) syndrome. J Am W, Roizen N. 2006a. Temporal lobe anatomy and psychiatric symptoms Acad Child Adolesc Psychiatry 48:1060–1068. in velocardiofacial syndrome (22q11. 2 deletion syndrome). J Am Acad Child Adolesc Psychiatry 45:587–595. Haxby JV, Grady CL, Horwitz B, Ungerleider LG, Mishkin M, Carson RE, Herscovitch P, Schapiro MB, Rapoport SI. 1991. Dissociation of object Kates WR, Antshel KM, AbdulSabur N, Colgan D, Funke B, Fremont W, and spatial visual processing pathways in human extrastriate cortex. Proc Higgins AM, Kucherlapati R, Shprintzen RJ. 2006b. A gender-moderated Natl Acad Sci USA 88:1621–1625. effect of a functional COMT polymorphism on prefrontal brain mor- phology and function in velo-cardio-facial syndrome (22q11. 2 deletion Heinz A, Smolka MN. 2006. The effects of catechol O-methyltransferase syndrome). Am J Med Genet Part B Neuropsychiatr Genet 141B: genotype on brain activation elicited by affective stimuli and cognitive 274–280. tasks. Rev Neurosci 17:359–368. Kempf L, Nicodemus KK, Kolachana B, Vakkalanka R, Verchinski BA, Honea R, Verchinski BA, Pezawas L, Kolachana BS, Callicott JH, Mattay Egan MF, Straub RE, Mattay VA, Callicott JH, Weinberger DR. 2008. VS, Weinberger DR, Meyer-Lindenberg A. 2009. Impact of interacting Functional polymorphisms in PRODH are associated with risk and functional variants in COMT on regional gray matter volume in human protection for schizophrenia and fronto-striatal structure and function. brain. Neuroimage 45:44–51. PLoS Genet 4:e1000252. Ikeda M, Williams N, Williams HJ, Smith R, Monks S, Owen MJ, Murphy Knickmeyer RC, Wang J, Zhu H, Geng X, Woolson S, Hamer RM, KC, O’Donovan MC. 2010. Failure to confirm association between Konneker T, Lin W, Styner M, Gilmore JH. 2014. Common variants PIK4CA and psychosis in 22q11. 2 deletion syndrome. Am J Med Genet in psychiatric risk genes predict brain structure at birth. Cereb Cortex Part B 153B:980–982. 24:1230–1246. Inoue H, Shibuta K, Matsuyama A, Yoshinaga K, Sadanaga N, Ueo H, Kremer I, Pinto M, Murad I, Muhaheed M, Bannoura I, Muller D, Schulze Barnard GF, Mori M. 2005. Genetic susceptibility of catechol-O-meth- T, Reshef A, Blanaru M, Gathas S. 2003. Family-based and case-control yltransferase polymorphism in Japanese patients with breast cancer. study of catechol-O-methyltransferase in schizophrenia among pales- Oncol Rep 14:707–712. tinian arabs. Am J Med Genet Part B Neuropsychiatr Genet 119B:35–39. Ishiguro H, Koga M, Horiuchi Y, Noguchi E, Morikawa M, Suzuki Y, Arai Li Y, Zhao Q, Wang T, Liu J, Li J, Li T, Zeng Z, Feng G, He L, Shi Y. 2011. M, Niizato K, Iritani S, Itokawa M, Inada T, Iwata N, Ozaki N, Ujike H, Association study between GNB1L and three major mental disorders in Kunugi H, Sasaki T, Takahashi M, Watanabe Y, Someya T, Kakita A, chinese han populations. Psychiatry Res 187:457–459. Takahashi H, Nawa H, Arinami T. 2010. Supportive evidence for reduced expression of GNB1L in schizophrenia. Schizophr Bull 36:756–765. Liu H, Abecasis GR, Heath SC, Knowles A, Demars S, Chen YJ, Roos JL, Rapoport JL, Gogos JA, Karayiorgou M. 2002. Genetic variation in the Isobe M, Tanigaki K, Muraki K, Miyata J, Takemura A, Sugihara G, 22q11 locus and susceptibility to schizophrenia. Proc Natl Acad Sci USA Takahashi H, Aso T, Fukuyama H, Hazama M. 2015. Polymorphism 99:16859–16864. within a neuronal activity-dependent enhancer of NgR1 is associated with corpus callosum morphology in humans. Mol Neuropsychiatry Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindorff LA, Humter DJ, 1:105–115. McCarthy MI, Ramos EM, Cardon LR, Chakravarti A, Cho JH, Gutt- macher AE, Kong A, Kruglyak L, Mardis E, Rotimi CN, Klatkin M, Valle Jungerius B, Hoogendoorn M, Bakker S, Van’t Slot R, Bardoel A, Ophoff R, D, Whittemore AS, Boehnke M, Clark AG, Eichler EE, Gibson G, Haines Wijmenga C, Kahn R, Sinke R. 2008. An association screen of myelin- JL, Mackay TFC, MCCarroll SA, Visscher PM. 2009. Finding the missing related genes implicates the chromosome 22q11 PIK4CA gene in schizo- heritability of complex diseases. Nature 461:747–753. phrenia. Mol Psychiatry 13:1060–1068. Maynard T, Meechan D, Dudevoir M, Gopalakrishna D, Peters A, Heindel Karlen A, Karlsson TE, Mattsson A, Lundstromer K, Codeluppi S, Pham C, Sugimoto T, Wu Y, Lieberman J, LaMantia A. 2008. Mitochondrial TM, Backman CM, Ogren SO, Aberg E, Hoffman AF, Sherling MA, localization and function of a subset of 22q11 deletion syndrome Lupica CR, Hoffer BJ, Spenger C, Josephson A, Brene S, Olson L. 2009. candidate genes. Mol Cell Neurosci 39:439–451. Nogo receptor 1 regulates formation of lasting memories. Proc Natl Acad Sci USA 106:20476–20481. Maynard TM, Haskell GT, Peters AZ, Sikich L, Lieberman JA, LaMantia AS. 2003. A comprehensive analysis of 22q11 gene expression in the Karayiorgou M, Simon TJ, Gogos JA. 2010. 22q11.2 microdeletions: developing and adult brain. Proc Natl Acad Sci USA 100:14433–14438. Linking DNA structural variation to brain dysfunction and schizophre- nia22q11.2 microdeletions: Linking DNA structural variation to brain McCarthy CS, Ramprashad A, Thompson C, Botti J, Coman IL, Kates WR. dysfunction and schizophrenia. Nat Rev Neurosci 11(6):402–416. 2015. A comparison of FreeSurfer-generated data with and without manual intervention. Front Neurosci 9:379. Kates WR, Antshel KM, Faraone SV, Fremont WP, Higgins AM, Shprint- zen RJ, Botti J, Kelchner L, McCarthy C. 2011a. Neuroanatomic pre- McGee AW, Yang Y, Fischer QS, Daw NW, Strittmatter SM. 2005. dictors to prodromal psychosis in velocardiofacial syndrome (22q11. 2 Experience-driven plasticity of visual cortex limited by myelin and deletion syndrome): A longitudinal study. Biol Psychiatry 69:945–952. nogo receptor. Science 309:2222–2226. Kates WR, Bansal R, Fremont W, Antshel KM, Hao X, Higgins AM, Liu J, McIntosh AM, Baig BJ, Hall J, Job D, Whalley HC, Lymer GKS, Moorhead Shprintzen RJ, Peterson BS. 2011b. Mapping cortical morphology in TWJ, Owens DG, Miller P, Porteous D. 2007. Relationship of catechol- youth with velocardiofacial (22q11. 2 deletion) syndrome. J Am Acad O-methyltransferase variants to brain structure and function in a Child Adolesc Psychiatry 50:272–282. e2. population at high risk of psychosis. Biol Psychiatry 61:1127–1134. Kates WR, Burnette CP, Bessette BA, Folley BS, Strunge L, Jabs EW, Meng J, Shi Y, Zhao X, Guo S, Wang H, Zheng Y, Tang R, Feng G, Gu N, Liu Pearlson GD. 2004. Frontal and caudate alterations in velocardiofacial H. 2007. No association between the genetic polymorphisms in the syndrome (deletion at chromosome 22q11.2). J Child Neurol RTN4R gene and schizophrenia in the chinese population. J Neural 19:337–342. Transm 114:249–254. THOMPSON ET AL. 313

Miller TJ, McGlashan TH, Rosen JL, Cadenhead K, Ventura J, McFarlane Sazci A, Ergul E, Kucukali I, Kilic G, Kaya G, Kara I. 2004. Catechol-O- W, Perkins DO, Pearlson GD, Woods SW. 2003. Prodromal assessment methyltransferase gene Val108/158Met polymorphism, and susceptibil- with the structured interview for prodromal syndromes and the scale of ity to schizophrenia: Association is more significant in women. Mol prodromal symptoms: Predictive validity, interrater reliability, and Brain Res 132:51–56. training to reliability. Schizophr Bull 29:703. Schneider M, Debbane M, Bassett AS, Chow EW, Fung WL, van den Moghaddam B, Javitt D. 2012. From revolution to evolution: The gluta- Bree M, Owen M, Murphy KC, Niarchou M, Kates WR, Antshel KM, mate hypothesis of schizophrenia and its implication for treatment. Antshel KM, Fremont W, McDonald-McGinn DM, Gur RE, Zackai Neuropsychopharmacology 37:4–15. EH, Vorstman J, Duijff SN, Klaassen PW, Swillen A, Gothelf D, Green T, Weizman A, Van Amelsvoort T, Evers L, Boot E, Shashi V, Hooper Mukai J, Liu H, Burt RA, Swor DE, Lai W, Karayiorgou M, Gogos JA. 2004. SR, Bearden CE, Jalbrzikowski M, Armando M, Vicari S, Murphy DG, Evidence that the gene encoding ZDHHC8 contributes to the risk of Ousley O, Campbell LE, Simon TJ, Eliez S; International Consortium schizophrenia. Nat Genet 36:725–731. on Brain and Behavior in 22q11.2 Deletion Syndrome. 2014. Psychi- Murphy KC, Jones LA, Owen MJ. 1999. High rates of schizophrenia in atric disorders from childhood to adulthood in 22q11. 2 deletion adults with velo-cardio-facial syndrome. Arch Gen Psychiatry 56: syndrome: Results from the international consortium on brain and 940–945. behavior in 22q11. 2 deletion syndrome. Am J Psychiatry 171:627– 639. Ohnishi T, Hashimoto R, Mori T, Nemoto K, Moriguchi Y, Iida H, Noguchi H, Nakabayashi T, Hori H, Ohmori M, Tsukue R, Anami K, Schofield CM, Hsu R, Barker AJ, Gertz CC, Blelloch R, Ullian EM. 2011. Hirabayashi N, Harada S, Arima K, Saitoh O, Kunugi H. 2006. The Monoallelic deletion of the microRNA biogenesis gene Dgcr8 produces association between the Val158Met polymorphism of the catechol-O- deficits in the development of excitatory synaptic transmission in the methyl transferase gene and morphological abnormalities of the brain in prefrontal cortex. Neural Dev 6:1. chronic schizophrenia. Brain 129:399–410. Sellier C, Hwang VJ, Dandekar R, Durbin-Johnson B, Charlet- Ota VK, Bellucco FT, Gadelha A, Santoro ML, Noto C, Christofolini DM, Berguerand N, Ander BP, Sharp FR, Angkustsiri K, Simon TJ, AssuncS~ao IB, Yamada KM, Ribeiro-dos-Santos AK,^ Santos S. 2014. Tassone F. 2014. Decreased DGCR8 expression and miRNA dysre- PRODH polymorphisms, cortical volumes and thickness in schizophre- gulation in individuals with 22q11. 2 deletion syndrome. PLoS ONE nia. PLoS ONE 9:e87686. 9:e103884. Ota VK, Gadelha A, AssuncS~ao IB, Santoro ML, Christofolini DM, Bellucco Shaikh TH, Kurahashi H, Saitta SC, O’Hare AM, Hu P, Roe BA, Driscoll FT, Santos-Filho AF, Ottoni GL, Lara DR, Mari JJ. 2013. ZDHHC8 gene DA, McDonald-McGinn DM, Zackai EH, Budarf ML, Emanuel BS. 2000. may play a role in cortical volumes of patients with schizophrenia. Chromosome 22-specific low copy repeats and the 22q11.2 deletion Schizophr Res 145:33–35. syndrome: Genomic organization and deletion endpoint analysis. Hum Mol Genet 9:489–501. Otani K, Ujike H, Tanaka Y, Morita Y, Kishimoto M, Morio A, Uchida N, Nomura A, Kuroda S. 2005. The ZDHHC8 gene did not associate with Shifman S, Bronstein M, Sternfeld M, Pisante-Shalom A, Lev-Lehman E, bipolar disorder or schizophrenia. Neurosci Lett 390:166–170. Weizman A, Reznik I, Spivak B, Grisaru N, Karp L. 2002. A highly significant association between a COMT haplotype and schizophrenia. Ouchi Y, Banno Y, Shimizu Y, Ando S, Hasegawa H, Adachi K, Iwamoto T. Am J Hum Genet 71:1296–1302. 2013. Reduced hippocampal neurogenesis and working memory deficits in the DGCR8-deficient mouse model of 22q11.2 deletion-associated Simon TJ, Ding L, Bish JP, McDonald-McGinn DM, Zackai EH, Gee J. schizophrenia can be rescued by ICF2. J Neurosci 33(22):9408–9419. 2005. Volumetric, connective, and morphologic changes in the brains of children with chromosome 22q11. 2 deletion syndrome: An integrative Perlstein MD, Chohan MR, Coman IL, Antshel KM, Fremont WP, Gnirke study. Neuroimage 25:169–180. MH, Kikinis Z, Middleton FA, Radoeva PD, Shenton ME, Kates WR. 2014. White matter abnormalities in 22q11. 2 deletion syndrome: Simon TJ, Wu Z, Avants B, Zhang H, Gee JC, Stebbins GT. 2008. Atypical Preliminary associations with the nogo-66 receptor gene and symptoms cortical connectivity and visuospatial cognitive impairments are related of psychosis. Schizophr Res 152:117–123. in children with chromosome 22q11. 2 deletion syndrome. Behav Brain Funct 4:1. Petrinovic MM, Duncan CS, Bourikas D, Weinman O, Montani L, Schroeter A, Maerki D, Sommer L, Stoeckli ET, Schwab ME. 2010. Smolka MN, Schumann G, Wrase J, Grusser SM, Flor H, Mann K, Braus Neuronal nogo-A regulates neurite fasciculation, branching and exten- DF, Goldman D, Buchel C, Heinz A. 2005. Catechol-O-methyltransfer- sion in the developing nervous system. Development 137:2539–2550. ase val158met genotype affects processing of emotional stimuli in the amygdala and prefrontal cortex. J Neurosci 25:836–842. Phang JM, Hu C, Valle D. 2001. Disorders of proline and hydroxyproline metabolism. Metab Mol Bases Inherit Dis 2:1820–1838. Spoletini I, Cherubini A, Di Paola M, Banfi G, Rusch€ N, Martinotti G, Bria P, Rubino IA, Siracusano A, Caltagirone C. 2009. Reduced fronto- Radoeva PD, Coman IL, Salazar CA, Gentile KL, Higgins AM, Middleton temporal connectivity is associated with frontal gray matter density FA, Antshel KM, Fremont W, Shprintzen RJ, Morrow BE, Kates WR. reduction and neuropsychological deficit in schizophrenia. Schizophr 2014. Association between autism spectrum disorder in individuals with Res 108:57–68. velocardiofacial (22q11.2 deletion) syndrome and PRODH and COMT genotypes. Psychiatr Genet 24:269–272. Squarcione C, Torti MC, Di Fabio F, Biondi M. 2013. 22q11 deletion syndrome: A review of the neuropsychiatric features and their neurobi- Riaz Z. 2015. Characterisation of GNB1L, a novel gene on chromosome ological basis. Neuropsychiatry Dis Treat 9:1873–1884. 22qll, implicated in schizophrenia and autism (Doctoral dissertation, St. George’s, University of London). Stark KL, Xu B, Bagchi A, Lai W, Liu H, Hsu R, Wan X, Pavlidis P, Mills AA, Karayiorgou M. 2008. Altered brain microRNA biogenesis contributes Saha S, Chant D, Welham J, McGrath J. 2005. A systematic review of the to phenotypic deficits in a 22q11-deletion mouse model. Nat Genet prevalence of schizophrenia. PLoS Med 2:e141. 40:751–760. Saito S, Ikeda M, Iwata N, Suzuki T, Kitajima T, Yamanouchi Y, Kinoshita Subramaniam B, Hennessey JG, Rubin MA, Beach LS, Reiss AL. 1997. Y, Takahashi N, Inada T, Ozaki N. 2005. No association was found Software and methods for quantitative imaging in neuroscience: between a functional SNP in ZDHHC8 and schizophrenia in a Japanese The Kennedy Krieger Institute. Neuroinformatics: An Overview of case-control population. Neurosci Lett 374:21–24. the Human Brain Project 1:335. 314 AMERICAN JOURNAL OF MEDICAL GENETICS PART B

Sun Y, Tao Y, Wang J, Saffen D. 2015. The schizophrenia/bipolar disorder associated with better working memory and larger hippocampal volume candidate gene GNB1L is regulated in human temporal cortex by a cis- in healthy chinese. Genes Brain Behav 12:323–329. acting element located within the 30-region. Neurosci Bull 31:43–52. WilliamsNM,GlaserB,NortonN,WilliamsH,PierceT,MoskvinaV,Monks Sun G, Yan J, Noltner K, Feng J, Li H, Sarkis DA, Sommer SS, Rossi JJ. 2009. S,DelFaveroJ,GoossensD,RujescuD,GieglingI,KirovG,CraddockN, SNPs in human miRNA genes affect biogenesis and function. Rna Murphy KC, O’Donovan MC, Owen MJ. 2008. Strong evidence that GNB1L 15:1640–1651. is associated with schizophrenia. Hum Mol Genet 17:555–566. Sunaert S, Van Hecke P, Marchal G, Orban GA. 1999. Motion-responsive Wong LM, Riggins T, Harvey D, Cabaral M, Simon TJ. 2014. Children with regions of the human brain. Exp Brain Res 127:355–370. chromosome 22q11. 2 deletion syndrome exhibit impaired spatial working memory. Am J Intellect Dev Disabil 119:115–132. Tunbridge E, Burnet PW, Sodhi MS, Harrison PJ. 2004. Catechol-o- methyltransferase (COMT) and proline dehydrogenase (PRODH) Zatorre RJ, Fields RD, Johansen-Berg H. 2012. Plasticity in gray and white: mRNAs in the dorsolateral prefrontal cortex in schizophrenia, bipolar Neuroimaging changes in brain structure during learning. Nat Neurosci disorder, and major depression. Synapse 51:112–118. 15:528–536. Tunbridge EM, Harrison PJ, Weinberger DR. 2006. Catechol-o-methyl- Zhang D, Cheng L, Qian Y, Alliey-Rodriguez N, Kelsoe JR, Greenwood T, transferase, cognition, and psychosis: Val 158 met and beyond. Biol Nievergelt C, Barrett TB, McKinney R, Schork N. 2009. Singleton Psychiatry 60:141–151. deletions throughout the genome increase risk of bipolar disorder. Mol Psychiatry 14:376–380. Ungerleider LG, Courtney SM, Haxby JV. 1998. A neural system for human visual working memory. Proc Natl Acad Sci USA 95:883–890. Zhao D, Lin M, Chen J, Pedrosa E, Hrabovsky A, Fourcade HM, Zheng D, Lachman HM. 2015. MicroRNA profiling of neurons generated using van Amelsvoort T, Zinkstok J, Figee M, Daly E, Morris R, Owen MJ, induced pluripotent stem cells derived from patients with schizophrenia Murphy KC, De Haan L, Linszen D, Glaser B. 2008. Effects of a functional and schizoaffective disorder, and 22q11. 2 del. PLoS ONE 10:e0132387. COMT polymorphism on brain anatomy and cognitive function in adults with velo-cardio-facial syndrome. Psychol Med 38:89–100. Zhou Y, Wang J, Lu X, Song X, Ye Y, Zhou J, Ying B, Wang L. 2013. Evaluation of six SNPs of MicroRNA machinery genes and risk of van Os J, Kapur S. 2009. Schizophrenia. Lancet 374:635–645. schizophrenia. J Mol Neurosci 49:594–599. Vorstman JA, Chow EW, Ophoff RA, van Engeland H, Beemer FA, Kahn RS, Sinke RJ, Bassett AS. 2009. Association of the PIK4CA schizophrenia- susceptibility gene in adults with the 22q11. 2 deletion syndrome. Am J SUPPORTING INFORMATION Med Genet Part B Neuropsychiatr Genet 150B:430–433. Wang Y, Li J, Chen C, Zhu B, Moysis R, Lei X, Li H, Liu Q, Xiu D, Liu B. Additional supporting information may be found in the online 2013. COMT rs4680 met is not always the ‘smart allele’: Val allele is version of this article.