Molecular Psychiatry (2009) 14,18–29 & 2009 Nature Publishing Group All rights reserved 1359-4184/09 $32.00 www.nature.com/mp FEATURE REVIEW The -binding 1 : features and networks AY Guo1,4, J Sun1,4, BP Riley1,2, DL Thiselton1, KS Kendler1,2 and Z Zhao1,2,3 1Department of Psychiatry and Virginia Institute for Psychiatric and Behavior Genetics, Virginia Commonwealth University, Richmond, VA, USA; 2Department of Human and Molecular Genetics, Virginia Commonwealth University, Richmond, VA, USA and 3Center for the Study of Biological Complexity, Virginia Commonwealth University, Richmond, VA, USA

The dystrobrevin-binding protein 1 (DTNBP1) gene has been one of the most studied and promising susceptibility since it was first reported to be associated with schizophrenia in the Irish Study of High Density Schizophrenia Families (ISHDSF). Although many studies have been performed both at the functional level and in association with psychiatric disorders, there has been no systematic review of the features of the DTNBP1 gene, protein or the relationship between function and phenotype. Using a bioinformatics approach, we identified the DTNBP1 gene in 13 vertebrate species. The comparison of these genes revealed a conserved gene structure, protein-coding sequence and domain, but a diverse noncoding sequence. The molecular evolutionary analysis suggests the DTNBP1 gene probably originated in chordates and matured in vertebrates. No signature of recent positive selection was seen in any primate lineage. The DTNBP1 gene likely has many more alternative transcripts than the current three major isoforms annotated in the NCBI database. Our examination of risk haplotypes revealed that, although the frequency of a single nucleotide polymorphism (SNP) or haplotype might be significantly different in cases from controls, difference between major geographic populations was even larger. Finally, we constructed the first DTNBP1 interactome and explored its network features. Besides the biogenesis of lysosome-related organelles complex 1 and -associated protein complex, several molecules in the DTNBP1 network likely provide insight into the role of DTNBP1 in biological systems: retinoic acid, b-estradiol, and tumour necrosis factor. Studies of these subnetworks and pathways may provide opportunities to deepen our understanding of the mechanisms of action of DTNBP1 variants. Molecular Psychiatry (2009) 14, 18–29; doi:10.1038/mp.2008.88; published online 29 July 2008 Keywords: DTNBP1; schizophrenia; splicing; haplotype; protein–protein interaction; gene network

Introduction In 2002, Straub et al.6 first identified the DTNBP1 gene as a putative schizophrenia susceptibility gene The human dystrobrevin-binding protein 1 (DTNBP1) by undertaking systematic linkage disequilibrium gene spans B140 kb on 6p22.3 and has mapping across a linkage region on 6p in the 270 10 exons. So far, it has not yet been classified into any multiply affected pedigrees from the Irish Study of known gene family. Dysbindin, a coiled-coil-contain- High Density Schizophrenia Families (ISHDSF). ing protein encoded by DTNBP1, was initially found Reanalysis of this data identified a single high-risk to interact with a- and b-dystrobrevin (DTNA and haplotype containing 8 single nucleotide polymorph- DTNB) in the muscle and of mice.1 DTNA and isms (SNPs) covering 30 kb.7 As of 19 April 2008, 45 DTNB are members of the dystrophin-associated follow-up association studies, 18 of which had protein complex (DPC), which links the cytoskeleton positive results, are annotated in the Schizophrenia- to the extracellular matrix and serves as a scaffold for Gene database.8 One meta-analysis has been pub- signaling .1,2 Dysbindin is also an essential lished.9 So far, DTNBP1 has been one of the most component of the biogenesis of lysosome-related prominent schizophrenia susceptibility genes.10,11 organelles complex 1 (BLOC-1) and interacts with DTNBP1 has been associated with other phenotypes all seven other components of BLOC-1.3–5 including intelligence, schizoaffective disorder, bipo- lar disorder and the Hermansky–Pudlak syndrome Correspondence: Dr Z Zhao, Department of Psychiatry, Virginia type 7.3,12,13 Commonwealth University, 800 East Leigh Street, Suite 118, DTNBP1 association studies have been frequently Richmond, VA 23298, USA. cited and reviewed,10,14–16 and will not be discussed E-mail: [email protected] 4These authors contributed equally to this work. here. However, its gene feature, expression and Received 13 May 2008; revised 24 June 2008; accepted 2 July protein’s interactions with other molecules in cellular 2008; published online 29 July 2008 systems have largely remained speculative.14 For DTNBP1 gene features and networks AY Guo et al 19 Table 1 DTNBP1 gene in 13 species

Species Gene ID DNA CDS Amino Dysbindin Major acid domain source

Length Repeat a Length Identityb Length Identityb Length Identityb (kb) (%) (bp) (%) (aa) (%) (aa) (%)

Human 84062 140.2 52.8 1056 351 121 NCBI Chimp ENSPTRG00000017744 145.9 50.8 1002 99.0 333 99.4 121 99.2 Ensembl Orangutanc 143.5 49.5 1050 98.1 349 97.7 121 100.0 Genome Rhesus ENSMMUG00000015363 147.1 52.0 996 97.3 331 97.3 120 97.5 Ensembl Marmostc 137.4 45.8 1056 95.1 351 94.3 121 94.2 Genome Dog ENSCAFG00000009893 115.8 40.9 1026 79.7 341 78.6 121 78.5 Ensembl Pig 100049697 118.4 39.0 1029 83.4 342 84.6 121 80.2 NCBI Cow 506612 89.1 38.0 1029 84.3 342 85.2 121 84.3 NCBI Mouse 94245 80.0 28.3 1059 83.0 352 77.7 121 80.2 NCBI Rat 641528 90.4 30.9 1059 81.8 352 76.8 121 81.0 NCBI Opossum ENSMODG00000010762 207.0 1038 79.2 345 82.1 120 81.8 Ensembl Chicken 420840 68.7 7.8 1056 76.4 351 78.5 121 87.6 NCBI Zebrafish 394109 31.1 22.5 1089 67.9 362 63.2 124 68.6 NCBI

Abbreviations: chimp, chimpanzee; rhesus, rhesus macaque. aThe proportion (%) of repetitive sequences identified by the RepeatMasker. bIdentity was calculated by comparing with the human DTNBP1 sequence. cOrangutan and marmost DTNBP1 genes were predicted from their draft genomes (see Supplementary Materials and methods).

example, there are few changes (nonsy- extensive search for the DTNBP1 gene in several nonymous ) observed in the human popula- major databases including NCBI Gene, tion, none of which has been reported to be associated Ensembl and dbEST, as well as in all the available with schizophrenia. Recent studies revealed a re- genomes (see Supplementary Materials and methods). duced expression of DTNBP1 in the frontal cortex and Based on these sources, we identified human hippocampal formation of schizophrenia patients.17–19 DTNBP1 orthologs in 12 other vertebrates, including DTNBP1 may confer susceptibility to schizophrenia 4 non-human primates, 6 non-primate mammals, and via reduced expression mediated by its high risk 2 non-mammals (Table 1). The length of its amino haplotype, which might tag one or more cis-acting acid sequence is 351 in humans. Similar length and variants.17,20 These observations call for an investiga- high identity of the amino acids are observed in other tion of the functional elements or regulatory mechan- species (Table 1). A dysbindin domain is annotated isms which might affect the DTNBP1 expression and (positions 184–304 in NCBI human dysbindin isoform confer the illness. Importantly, protein–protein a, NP_115498). The domain is highly conserved interaction (PPI) analysis of DISC1 (Disrupted in among the vertebrates. For example, the identity Schizophrenia 1), another prominent schizophrenia is 87.6% between human and chicken dysbindin susceptibility gene, suggested that schizophrenia domains. susceptibility genes (for example, DTNBP1 and Gene structure is also similar across species. DISC1) may share common PPIs and affect common For example, 10 exons are annotated in the databases biological processes.21 If this is true, network analysis or have been consistently predicted in the 13 species. may reveal novel mechanisms of etiology and inter- However, gene length varies widely from 31.1 kb in vention not easily reached by more conventional zebrafish to 207.0 kb in opossums, probably as a result approaches (for example, single gene analysis). of variation in the extent of repeats. For example, Here, we apply bioinformatics and systems biology human DTNBP1 has 74 088 bp repetitive sequences, approaches to explore the features of DTNBP1, accounting for 52.8% of the sequence; conversely, including its molecular evolution and sequence mouse DTNBP1 has 22 648 bp repetitive sequences, conservation, gene structure, transcripts, haplotypes accounting for only 28.3% of the sequence (Supple- and interactome and pathways. mentary Table S1). Repeats account for a smaller proportion in chickens and zebrafish DTNBP1 genes, whose lengths are also shorter. Conserved DTNBP1 gene structure in vertebrates Figure 1 shows the aligned amino acid sequences and protein features in the 13 species. Overall, the So far, the DTNBP1 gene has only been experimen- coiled-coil region, dysbindin domain, and some other tally verified in mice1 and humans.6 We performed an functional sites are conserved among these species.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 20 The N-terminal region is more conserved than mentary Table S2; Supplementary Figure S2). When the C-terminal region. Phylogenetic trees were more genetically distant species (human, dog, mouse, reconstructed using the aligned amino acid sequences and opossum) were employed in a cross-species and CDS sequences, respectively. The protein and sequence comparison, all exons are again CDS trees show similar topology (Supplementary highly conserved, but only a few CNGs are present Figure S1). (Figure 2). Human and mouse sequences are typically used to identify evolutionary conserved regions and to eval- DTNBP1 gene likely originated in chordates and uate sequence conservation.22 Using VISTA tool matured in vertebrates (http://genome.lbl.gov/vista/), we found that the exons are highly conserved; however, we also identi- Besides the DTNBP1 genes found in these 13 species, fied 41 conserved nongenic regions (CNGs) covering an extensive BLAST search of human protein 6729 bp with an average identity of 72.9% (Supple- sequence identified hits in 12 other organisms

Figure 1 Alignment of the 13 DTNBP1 amino acid sequences. The multiple alignment was first generated by computer program Clustal W and then manually checked and refined. Intron positions are indicated by triangles. The predicted coiled-coil region is labeled by a thin line on the top of the alignment and dysbindin domain is labeled by a red thick line. Two nonsynonymous SNPs (rs16876589: G214D, rs17470454: P272S) in isoform a are indicated.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 21

Figure 2 VISTA plot displaying DTNBP1 gene structure and evolutionary conserved regions in a comparison of human, dog, mouse and opossum DNA sequences. Conserved nongenic region (CNG, in red) is defined as an alignment at least 100- bp long and at least 70% identity. The plot used human gene and annotations as reference. In each gene line, a vertical bar denotes an exon, a dashed line denotes a shorter intronic region and a double forward slash denotes a longer intronic region than the human sequence. including 5 vertebrates and 4 invertebrates (Supple- selection in a lineage. A dN/dS ratio greater than 1 mentary Table S3). Using cutoff values for the BLAST suggests positive selection at the examined.26 output: identity at the amino acid level of > 30% and The results suggested an enrichment of a positive requiring the aligned sequences to cover > 50% of the selection signal in these genes. For DTNBP1,apositive human dysbindin amino acid sequence, significant selection signal was found in the European population hits are all from chordates. The most ancient species and a weak signal was also found in the African in our search results is amphioxus (Branchiostoma population in the HapMap data.25 However, no belcheri tsingtauense, accession no. AY280671), an positive selection was detected in the human and ancient fish classified as an invertebrate. However, chimpanzee lineages.24 Our test using human and four the most significant hits ( > 50% identity in the non-human primates further confirmed no positive BLAST output, > 80% coverage of the human amino selection in any primate lineage as the dN/dS ratio in acids) were only found in vertebrates. Moreover, the each pairwise primate lineage comparison was con- complete dysbindin domain was only observed in sistently smaller than 0.5 (Supplementary Table S4). vertebrates. Thus, the functional DTNBP1 gene Furthermore, no positive selection was detected at any probably originated in chordates and the current single codon site using PAML site models or branch- DTNBP1 gene structure likely matured in vertebrates. site models.26 As both the statistical methods tend to Given the simpler nervous system in invertebrate be liberal in detecting positive selection and the iHS chordates versus vertebrates, this suggests that algorithm implemented in the Haplotter only tests DTNBP1’s role arose with increasing central nervous three HapMap subpopulations,25,27 the DTNBP1gene system complexity. had probably not experienced substantial positive selection in recent evolutionary history.

Test recent positive selection in the human and other primate lineages Human DTNBP1 putative transcripts and expression Schizophrenia may be a maladaptive by-product of Currently, three major transcripts (isoforms a, b, and c; adaptive cognitive changes during evolution.23 A accession numbers NM_032122, NM_183040 and recent study24 of 76 genes associated with schizo- NM_183041) of human DTNBP1 are annotated in phrenia indicated that 28 of these genes may have the NCBI Entrez database and more than 160 mRNA been under recent positive selection by applying two or EST sequences are related to human DTNBP1. The methods: Haplotter, which tests selective sweeps in transcript isoform a is the reference transcript and has human populations using the HapMap allelic data,25 10 exons. The 10-exon gene structure and sequences and the ratio of nonsynonymous over synonymous are conserved in the 13 species we examined. substitution rates (dN/dS ratio), which tests positive However, the number of transcripts of human

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 22 DTNBP1 is likely more than 3. NCBI AceView transcripts a, d and e are detected in many clones (version April 2007),28 a comprehensive and well- from a large number of tissues, whereas short annotated alternative transcript variants database, transcripts are shown in a few clones and expressed annotated 16 human alternative transcripts whose in only one or a few tissues (Supplementary Table lengths vary from 268 to 1955 bp. Because the S5). Based on the current data, these alternative transcripts in AceView were annotated in a genome- transcripts, especially the short ones, must be wide fashion, we refined them as follows. We regarded as putative and require verification. Addi- retrieved all the mRNA and EST sequences of the tional experimental work in our group confirmed six human DTNBP1 gene from the NCBI UniGene of the eight transcripts assessed (manuscript in database, then performed a clustering analysis of all preparation). available mRNA and EST sequences using TGI Clustering Tools (http://compbio.dfci.harvard.edu/ SNPs in DTNBP1 gene tgi/software/), and finally compared the generated transcripts with the AceView transcripts based on the There are 558 nonredundant SNPs in the human original DTNBP1 mRNA and EST sequences. For the DTNBP1 gene region in the NCBI dbSNP data- 16 AceView transcripts (named a–p), we deleted base (build 129, http://www.ncbi.nlm.nih.gov/SNP/), transcript j and modified transcript o by deleting the including 196 HapMap SNPs. Only five SNPs are in last 200 bp in its 30 end. In addition, we added EST protein-coding regions, four of them (rs17470454, CD723892 as a partial transcript (details are shown in rs16876589, rs16876569 and rs16876571) nonsynon- Supplementary Materials and methods). This ymous. The minor allele frequencies of all these five resulted in a new set of 16 putative transcripts SNPs are very low in CEU and JPT þ CHB and 0 in (Figure 3). Among them, five (a, d, e, g and h) have YRI based on the HapMap data, indicating that these mRNA evidence and the others have only EST data. mutations occurred recently in non-African popula- Seven transcripts (k, l, m, n, o, p and CD723892) are tion and have not yet been fixed. Twenty-one SNPs short, which are likely partial transcripts. According are mapped in the CNGs (Supplementary Table S2). to the expression data in the AceView, the three There are 22 SNPs that have been reported either

Figure 3 transcripts of the human DTNBP1 gene. Sizes of the exons are approximately scaled. (a) Gene structure. The gray boxes indicate alternative splicing exons. Numbers 1–22 indicate the locations of the 22 SNPs that have been reported to be associated with schizophrenia. (b) Transcripts with available mRNAs. (c) Transcripts with available ESTs. In (b) and (c), black boxes denote coding exons and white boxes denote untranslated exons. For each transcript, NCBI mRNA or EST accession numbers are listed at the left and AceView transcript ID is listed at the right.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 23 Table 2 Comparison of DTNBP1 high-risk haplotype in the Irish study and the HapMap samples SNP 1 2 3 4 5 6 7 8 Frequency

Haplotype ISHDSF CEU JPT þ CHB YRI

Marmost A T G C G A G G Rhesus . C . G A . . . Orangutan . C ...... Chimp . C . A . . . A 1 . C T . . G . A 0.733 0.750 0.705 0.633 2GC.TA...0.058 0.108 0.005 0.025 3 . C T . . G A A 0.071 0.058 0.211 4 . C T T . G . A 0.010 5 G . . T A . . A 0.060 0.075 0.072 0.308 6 G C T T . G A . 0.015 0.008 Rare G . . T A . A .

Abbreviations: chimp, chimpanzee; rhesus, rhesus macaque; SNP, single nucleotide polymorphism. The SNPs from 1 to 8 are, respectively, rs1474605 (P1792), rs1018381 (P1578), rs2619522 (P1763), rs760761 (P1320), rs2005976 (P1757), rs2619528 (P1765), rs1011313 (P1325) and rs3213207 (P1635). Haplotypes 1 and 2 are the common and high-risk haplotypes in the ISHDSF. Haplotype frequencies were based on the ISHDSF study7 and the HapMap data (CEU, JPT þ CHB and YRI). The sum of the haplotype frequencies is less than 1 because some other haplotypes were not included.

individually or within a haplotype that is signifi- in JPT þ CHB sample. This difference is cantly associated with schizophrenia, including 2 mainly due to the preascertained SNPs selected in the SNPs in CNGs. The locations of the association SNPs HapMap project. These SNPs do not directly contain are displayed in Figure 3a and their details provided information about the underlying levels of nucleotide in Supplementary Table S6. In the mouse DTNBP1 diversity.25 Because iHS statistic in Haplotter tests the gene, there are 385 SNPs, including two non- extended haplotype homozygosity, not the number of synonymous mutations (rs46632574 and rs48618277) haplotypes, the observation of the similar number of and three synonymous mutations (rs51027077, haplotypes in CEU and JPT þ CHB is not directly rs50120298 and rs51319012). Overall, few SNPs related to their different positive selection signals could be found in the DTNBP1 protein-coding region. detected by Haplotter. Among the 22 SNPs from the association studies, 16 We re-examined the high-risk haplotype reported in have allele frequencies from HapMap. The ancestral the ISHDSF.7 Eight SNPs were employed in that alleles of the SNPs were inferred based on the high-risk haplotype: rs1474605 (P1792), rs1018381 maximum parsimony principle using a pipeline in (P1578), rs2619522 (P1763), rs760761 (P1320), Jiang and Zhao.29 The ancestral alleles of these SNPs rs2005976 (P1757), rs2619528 (P1765), rs1011313 are always major alleles with three exceptions: SNPs (P1325) and rs3213207 (P1635). Seven of them rs2619538, rs2619539 and rs742106 whose minor (excluding rs2005976) are available in the HapMap. alleles in JPT þ CHB are ancestral alleles (Supple- These seven SNPs result in the same haplotypes mentary Table S7). reported in ISHDSF7 (rs2005976 is redundant). We extracted the phased haplotypes and their frequencies from the HapMap and compared haplotype frequen- Haplotypes and high-risk haplotypes cies in the ISHDSF sample (Table 2). Interestingly, for We examined haplotype information for DTNBP and the high-risk haplotype (haplotype 2), its frequency in compared it to the reported high-risk haplotypes. CEU (0.108) is much higher than that in JPT þ CHB HapMap SNPs in the DTNBP1 gene were used to infer (0.005) or YRI (0.025). This confirmed the previous haplotype using the program PHASE.30 There are 44, suspicion of lower frequency in other samples.7 A 46, 56 haplotypes in CEU, JPT þ CHB and YRI, number of haplotypes were not observed in YRI or respectively, and 135 haplotypes in the whole JPT þ CHB sample. HapMap sample. Only a few haplotypes are shared Both van den Oord et al.7 and Mutsuddi et al.32 by subpopulations, for example, nine shared by at suggested that the ancestral haplotype was the most least two subpopulations and only two shared by common haplotype (haplotype 1, see Figure 1 in van three subpopulations. This nonsharing feature of den Oord et al.). We mapped the eight SNPs to four haplotypes between major geographic populations available primates (chimpanzee, orangutan, rhesus has been reported at other loci.31 Of note, the number macaque and marmost) and identified the corres- of haplotypes in CEU is nearly the same as that in ponding nucleotides at these sites. Figure 4a shows JPT þ CHB, though there are twice as many founder an evolutionary relationship of seven haplotypes and

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 24

Figure 4 Evolutionary tree and network of DTNBP1 8-SNP haplotypes. (a) The evolutionary tree reconstructed by the neighbor-joining method in MEGA 4.0. (b) The haplotype network generated by the median-joining algorithm in software NETWORK 4.2.01. Seven gray circles denote the seven haplotypes in van den Oord et al.7 Among them, haplotype 2 is the high-risk haplotype and haplotype 1 is the common haplotype. Circle areas are proportional to haplotype frequencies. Four octagon nodes denote the four non-human primates.

Table 3 Comparison of haplotypes (rs2619539-rs3213207-rs2619538) in Williams et al.34 and in the HapMap samples

Haplotype Cardiff sample Dublin sample CEU JPT þ CHB YRI

Cases Controls Cases Controls

GAA 0.13 0.14 0.11 0.11 0.117 0.011 0.183 CAAa 0.29 0.35 0.30 0.37 0.375 0.467 GGA 0.10 0.08 0.10 0.10 0.117 0.006 0.025 GAT 0.26 0.24 0.28 0.26 0.233 0.633 0.075 CATb 0.21 0.16 0.21 0.15 0.158 0.350 0.250 GGTa 0.00 0.03 0.00 0.02

aProtective haplotypes in the two studied samples (Cardiff and Dublin samples). bRisk haplotype in the two studied samples.

their outgroups (details were provided in the Supple- close to that in the control samples (Cardiff: 0.16; mentary Materials and methods). The evolutionary Dublin: 0.15). However, the frequencies in JPT þ CHB tree indicates that (1) haplotypes 2 and 5 are rare in a (0.350) and YRI (0.250) are higher than the controls branch separated from haplotypes 1, 3, 4 and 6 by and the cases (0.21 in both the samples). Of note, primates and (2) the common haplotype is genetically Numakawa et al.35 failed to detect a significant distant from primates. We further drew a haplotype association between the C-A-T haplotype and schizo- network (Figure 4b), which demonstrates the possible phrenia in a Japanese sample. The C-A-T risk mutational paths based on a median-joining algo- haplotype was associated with reduced cortical rithm.33 Both the evolutionary tree and haplotype expression of DTNBP1, poor spatial working memory, network indicate that the two haplotype groups are less severe manic-type symptoms, and early visual- separated by primates. The results suggest that (1) processing deficits in Irish samples.20,36–38 For the the high-risk and common haplotypes might have protective haplotype C-A-A, which is an ancestral undergone different evolutionary paths and, (2) haplotype, its frequencies in CEU and YRI are high contrary to the previous suggestions, the common (0.375 and 0.467, respectively), but it absents from the haplotype is not ancestral. JPT þ CHB. This absence is likely due to the low We also re-examined the risk haplotype reported in frequency of allele A (0.017) of SNP rs2619538 Williams et al.,34 which is based on three markers (Supplementary Table S7). The frequency of the (rs2619539, rs3213207 and rs2619538), one of which protective haplotype G-G-T, an entirely derived (rs3213207) was in the ISHDSF high-risk haplo- haplotype, is very low (0–0.03) in the Cardiff and type. One risk haplotype (C-A-T) and 2 protective Dublin samples and not found in any HapMap haplotypes (C-A-A and G-G-T) were reported in sample. two independent samples collected in Cardiff and The comparisons above suggest that, although a Dublin.34 Table 3 shows the frequencies of haplotypes SNP or haplotype might be found at significantly in these two samples and in the HapMap samples. For different frequencies in cases and controls, the the risk haplotype, its frequency in CEU (0.158) is difference between major geographic populations

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 25 might be even larger. Whether these population remaining 10 proteins, 7 could not be mapped to differences impact on differences in the case and any network and each of the 3 other proteins was control studies requires further examination. mapped to a small network with P-scorep3. Interest- ingly, is annotated as a functional feature in the DTNBP1 interactors. This is likely because cancer DTNBP1 protein interactions genes are widely annotated in the and have been associated with schizophrenia (for example, So far, only one schizophrenia susceptibility gene, TP53,43 AKT144). DISC1, has had its network extensively investi- As expected, frequent connections between pro- gated.21,39 Here, we present our detailed examination teins in group 2 were observed because most proteins of the network and pathway features for DTNBP1. are in the BLOC-1 complex. Interestingly, no direct To identify all proteins interacting with DTNBP1, interaction was found between any pair of proteins we collected and integrated all the available experi- within group 1 or between any pair of the remaining mentally verified human PPIs from four major PPI 10 proteins (we broadly named these 10 proteins as databases (details in Supplementary Materials and group 3). When DTNBP1 is removed, there is no methods). These protein pairs were considered as the direct interaction between any two groups, with one human protein interactome. A total of 31 proteins exception (a connection between ABI3 and were identified to directly interact with DTNBP1. The KIAA0408). To further examine this feature, we added genes encoding these 31 proteins are provided in proteins that directly interact with the 31 DTNBP1 Supplementary Table S10. Among them, 3 genes interactors (distance 1), that is, DTNBP1 interactors (BLOC1S3, DTNA and SYNE1) have phenotype with distance 2. The network shows that DTNBP1 annotations in the OMIM database and 10 genes serves as a superhub (a network node with more than (PLDN, CNO, MUTED, BLOC1S1, BLOC1S2, 15 links). Conversely, when DTNBP1 is intentionally BLOC1S3, DGCR6L, SNAPAP, RANBP5 and ZNF490) removed, the DTNBP1 interactors are only loosely have association studies in the SchizophreniaGene connected (Supplementary Figure S4). database.8 For these 10 genes, only 2 (BLOC1S3 and RANBP5) had positive results. We examined the distribution of characteristic GO terms associated DTNBP1’s networks/pathways with these genes. Fifty-seven GO terms associated with these 31 genes; 13 terms are associated with at We further explored the features of DTNBP1 path- least 5 genes and their hierarchical levels are at least 4 ways and protein interaction networks in the Inge- (Supplementary Table S11). The 13 GO terms are nuity GMN. We described the networks for the ‘’, ‘calcium ion binding’, ‘ binding’, proteins in group 2 because the analysis of proteins ‘nucleus’, ‘protein transport’, ‘’, ‘cytoske- in group 1 produced little additional information. leton’, ‘nucleoplasm’, ‘plasma membrane’, ‘intracellular’, Using the Ingenuity Network Analysis Tool, we ‘’, ‘cytosol’ and ‘transport’. Overall, extracted a complex subnetwork (details in Supple- these terms relate to subcellular localization, trans- mentary Materials and methods), which contains 140 port and signal transduction. The most frequently molecules including 125 single proteins, 4 protein observed GO term (23 genes) is cytoplasm, consistent complexes, 9 protein groups and 2 other molecules with the previous report of DTNBP1 subcellular (Supplementary Figure S5; Supplementary Table localization.40 The other two frequently observed GO S12). Moreover, in the network, 30 genes have terms are calcium ion binding (21 genes) and actin phenotype annotation in the OMIM database and 19 binding (18 genes). These proteins may be related to genes have been collected in the SchizophreniaGene mobility,41 which is consistent with database including 5 genes with at least 2 positive DTNBP1 being a synaptic and microtubular protein,42 results. As expected, the network includes two and part of DPC. Note that DTNBP1 and at least five of protein complexes, BLOC-1 and DPC, which involve its interactors (DTNA, PLDN, SNAPAP, RAB11A and DTNBP1. Fourteen molecules directly link to the SYNE1) have been implicated in neurodevelopmental components of these two complexes: retinoic acid processes. (RA), PRKACA, calmodulin (CaM), Nos, F-actin, Using the Ingenuity Systems Core Analysis, two ABCA1, SNAP25, SNAP23, Ap1, tumour necrosis groups stood out as being highly significant in the factor (TNF), PRX, actin, STX12 and EBAG9. To human Global Molecular Network (GMN; Figure 5a). interpret the results efficiently, we simplified the Group 1 (P-score = 28) contains 12 proteins. The P- network by limiting: (1) components of the BLOC-1 score is defined here as Àlog10P, where P is calculated and DPC complexes, (2) molecules directly linked to using Fisher’s exact test. The top functions of the the components of these two complexes (14 mole- proteins in group 1 are cell morphology, cellular cules), (3) the molecules directly linked these 14 development and cancer. Group 2 (P-score = 26) has molecules, and (4) any molecules having at least 15 10 proteins including all of the 8 components in links (that is, degreeX15) under the assumption that BLOC-1, DTNA and DTNB. The top functions in highly connected molecules tend to have important group 2 are cell morphology, cell-to-cell signaling and function.45 The simplified network is shown in Figure 5b. interaction, and developmental disorder. For the The features are described below.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 26

Figure 5 DTNBP1 interactome and extracted networks/pathways in the Ingenuity Global Molecular Network (GMN). (a) DTNBP1 interactome contains 31 proteins that directly interact with DTNBP1 in the human interactome. Using the Ingenuity System Core Analysis, these 31 proteins could be classified into group 1 (green, 12 proteins), group 2 (red, 9 proteins) and the others (10 proteins). Nodes in red indicate available phenotype annotation in the OMIM database and nodes in diamond indicate neurodevelopment annotation in the database. (b) Subnetwork for the proteins in group 2 and their direct interactors in GMN. Nodes in brown denote the proteins directly linked to the BLOC-1 and DPC complexes and node in gray denote the proteins in group 2 and their direct interactors. A solid line indicates a physical interaction, a dashed line with an arrow indicates a regulation relationship, and a solid line with an arrow indicates both a physical interaction and a regulation relationship.

Dysbindin in the BLOC-1 complex to contain at least eight components: DTNBP1, Dysbindin is an essential component of the BLOC-1 MUTED, PLDN, CNO, SNAPAP, BLOC1S1, BLOC1S2 complex.3 BLOC-1 is a 200-kDa ubiquitously and BLOC1S3. A number of the genes (DTNBP1, expressed soluble algometric protein complex known MUTED, PLDN, CNO and BLOC1S3) encoding these

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 27 proteins are defective in inbred mouse strains serving nia.58 Figure 5b depicts multiple paths from RA to as models of the Hermansky–Pudlak syndrome.3,5 DTNBP1; most are via regulation of expression. Figure 5b shows eight molecules that directly interact Figure 5b also shows b-estradiol, which link to with the BLOC-1 components: DTNA and DTNB that multiple molecules including Akt, TNF, insulin and interact with DTNBP1, EBAG9, SNAP23, PRKACA CaM. may be protective in schizophrenia as and SNAP25 that interact with SNAPAP, and STX12 men develop schizophrenia at an earlier age and with and F-actin that interact with PLDN. Six proteins greater severity than women.58,59 A recent study (STX12, SNAP23, PRKACA, EBAG9, SNAPAP, showed directly induces expression of RA SNAP25) are involved in membrane fusion of the biosynthetic enzymes in rats, suggesting a coordi- synaptic vesicle.46–49 In the fusion process, the nated role of estrogens and RA.60 SNARE complex plays a central role,46 which CaM, a calcium-binding protein, and PRKACA, a decreases the release of glutamate from astrocytes in cAMP-dependent protein kinase, are both involved in mice.50 A number of protein- and gene-expression long-term potentiation (LTP) of synaptic transmis- studies have presented evidence of DTNBP1 involve- sion.61 In , Nos is involved in synaptic ment in glutamate .19,35,42 There- long-term depression (LTD). LTP and its counterpart fore, the role of DTNBP1 and its interactors in the LTD have long been considered as a potential membrane fusion process of the synaptic vesicle is a mechanism for memory formation and learning.62 plausible mechanism by which risk of schizophrenia Upregulation of human CaM mRNA(s) in dorsolateral is increased through disruption of synaptic glutamate prefrontal cortex was reported in schizophrenics.63 neurotransmission.

Dysbindin in the DPC complex Conclusion Dysbindin interacts with DTNA in muscle and DTNB in the brain, both components of DPC.1 DPC is a The DTNBP1 gene is highly conserved in its structure multifunctional protein complex that has been stu- and functional domains, but not in its noncoding died extensively both genetically and biochemically sequences. This gene probably originated in chor- in muscle. Mutations in genes involved in DPC cause dates and matured in vertebrates, consistent with its several types of muscular dystrophies.51,52 DPC is a role in complex nervous systems. No strong evidence dynamic and strategically located cellular signaling of recent positive selection was found at the gene complex.53 In Figure 5b, most of the DPC components locus. The human DTNBP1 gene likely has many are shown including dystrophin (DMD) and it related more alternative transcripts than the current three proteins (UTRN, DTNA, DTNB and DRP2), dystrogly- major isoforms annotated in the NCBI database. Our cans (DAG1), (SNTA1, SNTB1, SNTB2, examination of risk haplotypes revealed large differ- SNTG1 and SNTG2), one intercellular binding part- ence between major geographic populations, which ner (F-actin) and two signaling molecules associated need to be taken into account in interpreting associa- with the complex: CaM and Nos (a group of proteins tion studies. Specifically, for the haplotypes based on including nNOS). complex was not the eight SNPs reported in the ISHDSF, we found that present, as previously reported.53 DTNA is highly (1) the seven haplotypes were separated as two groups expressed in whereas DTNB, which is by primate lineages, indicating different evolutionary abundantly expressed in the brain, kidney, lung and processes of the high-risk haplotype from the com- liver, is often considered a non-muscle protein.51 mon haplotype and (2) contrary to prior suggestion, Furthermore, DTNB is expressed at significant levels the most common haplotype is not the ancestral within the cerebral cortex and .54 How- haplotype. Finally, we constructed the first DTNBP1 ever, in brain, some of the individual components of interactome, which contains 31 interactors, and DPC are expressed presynaptically and others post- explored its network features in the Ingenuity GMN. synaptically. These data argue that the functions of The networks indicate that DTNBP1 is involved in DPC protein components in brain are different from both muscle and neuronal functions. DTNBP1 may those in muscle.55 Interestingly, to our knowledge, no confer its susceptibility to schizophrenia through its other gene in the DPC complex has been studied for impact on glutamate neurotransmission, RA signaling its association with schizophrenia. pathway, and synaptic LTP and LTD, which link to other biological pathways. Dysbindin and other molecules Besides BLOC-1 and DPC, several molecules in the network are noteworthy: RA, b-estradiol and CaM. RA Acknowledgments is a metabolic product of vitamin A (retinol) and an established signaling molecule involved in neuronal We regret that many DTNBP1 studies, especially patterning, neural differentiation and out- association studies, have not been cited in this review growth.56 Disruption of RA signaling has been because of our focus on gene feature and network implicated in the development of Alzheimer’s disease analysis. This work was supported by a NARSAD and possibly, Parkinson’s disease and depression.56,57 Young Investigator Award and a Jeffress Trust grant to Retinoids may play an etiological role in schizophre- ZZ and a research grant (R01MH41953) to KSK/BPR.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 28 References gene locus mediate risk for schizophrenia through reduced DTNBP1 expression. Hum Mol Genet 2005; 14: 1 Benson MA, Newey SE, Martin-Rendon E, Hawkes R, Blake DJ. 1947–1954. Dysbindin, a novel coiled-coil-containing protein that interacts 21 Camargo LM, Collura V, Rain JC, Mizuguchi K, Hermjakob H, with the dystrobrevins in muscle and brain. J Biol Chem 2001; 276: Kerrien S et al. Disrupted in Schizophrenia 1 Interactome: 24232–24241. evidence for the close connectivity of risk genes and a potential 2 Veroni C, Grasso M, Macchia G, Ramoni C, Ceccarini M, Petrucci TC synaptic basis for schizophrenia. Mol Psychiatry 2007; 12: 74–86. et al. beta-dystrobrevin, a kinesin-binding receptor, interacts with 22 Dermitzakis ET, Reymond A, Antonarakis SE. Conserved non- the extracellular matrix components pancortins. J Neurosci Res genic sequences—an unexpected feature of mammalian genomes. 2007; 85: 2631–2639. Nat Rev Genet 2005; 6: 151–157. 3 Li W, Zhang Q, Oiso N, Novak EK, Gautam R, O’Brien EP et al. 23 Brune M. Schizophrenia—an evolutionary enigma? Neurosci Hermansky–Pudlak syndrome type 7 (HPS-7) results from mutant Biobehav Rev 2004; 28: 41–53. dysbindin, a member of the biogenesis of lysosome-related 24 Crespi B, Summers K, Dorus S. Adaptive evolution of genes organelles complex 1 (BLOC-1). Nat Genet 2003; 35: 84–89. underlying schizophrenia. Proc Biol Sci 2007; 274: 2801–2810. 4 Nazarian R, Starcevic M, Spencer MJ, Dell’Angelica EC. Reinves- 25 Voight BF, Kudaravalli S, Wen X, Pritchard JK. A map of recent tigation of the dysbindin subunit of BLOC-1 (biogenesis of positive selection in the human genome. PLoS Biol 2006; 4:e72. lysosome-related organelles complex-1) as a dystrobrevin-binding 26 Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. protein. Biochem J 2006; 395: 587–598. Mol Biol Evol 2007; 24: 1586–1591. 5 Starcevic M, Dell’Angelica EC. Identification of snapin and three 27 MacCallum C, Hill E. Being positive about selection. PLoS Biol novel proteins (BLOS1, BLOS2, and BLOS3/reduced pigmenta- 2006; 4: e87. tion) as subunits of biogenesis of lysosome-related organelles 28 Thierry-Mieg D, Thierry-Mieg J. AceView: a comprehensive cDNA- complex-1 (BLOC-1). J Biol Chem 2004; 279: 28393–28401. supported gene and transcripts annotation. Genome Biol 2006; 6 Straub RE, Jiang Y, MacLean CJ, Ma Y, Webb BT, Myakishev MV 7(Suppl 1: S12): 11–14. et al. Genetic variation in the 6p22.3 gene DTNBP1, the human 29 Jiang C, Zhao Z. Mutational spectrum in the recent human genome ortholog of the mouse dysbindin gene, is associated with inferred by single nucleotide polymorphisms. Genomics 2006; 88: schizophrenia. Am J Hum Genet 2002; 71: 337–348. 527–534. 7 van den Oord EJ, Sullivan PF, Jiang Y, Walsh D, O’Neill FA, 30 Stephens M, Smith NJ, Donnelly P. A new statistical method for Kendler KS et al. Identification of a high-risk haplotype for the haplotype reconstruction from population data. Am J Hum Genet dystrobrevin binding protein 1 (DTNBP1) gene in the Irish study 2001; 68: 978–989. of high-density schizophrenia families. Mol Psychiatry 2003; 8: 31 Zhao Z, Yu N, Fu Y-X, Li W-H. Nucleotide variation and haplotype 499–510. diversity in a 10-kb noncoding region in three continental human 8 Allen NC, Bagade S, McQueen MB, Ioannidis JPA, Kavvoura FK, populations. Genetics 2006; 174: 399–409. Khoury MJ et al. Systematic meta-analyses and field synopsis of 32 Mutsuddi M, Morris DW, Waggoner SG, Daly MJ, Scolnick EM, genetic association studies in schizophrenia: the SzGene database. Sklar P. Analysis of high-resolution HapMap of DTNBP1 (Dysbindin) Nat Genet 2008; 40: 827–834. suggests no consistency between reported common variant associa- 9 Li D, He L. Association study between the dystrobrevin binding tions and schizophrenia. Am J Hum Genet 2006; 79: 903–909. protein 1 gene (DTNBP1) and schizophrenia: a meta-analysis. 33 Bandelt HJ, Forster P, Rohl A. Median-joining networks for Schizophr Res 2007; 96: 112–118. inferring intraspecific phylogenies. Mol Biol Evol 1999; 16: 37–48. 10 Riley B, Kendler KS. Molecular genetic studies of schizophrenia. 34 Williams NM, Preece A, Morris DW, Spurlock G, Bray NJ, Eur J Hum Genet 2006; 14: 669–680. Stephens M et al. Identification in 2 independent samples of a 11 Sun J, Kuo P-H, Riley BP, Kendler KS, Zhao Z. Candidate genes for novel schizophrenia risk haplotype of the dystrobrevin binding schizophrenia: a survey of association studies and gene ranking. protein gene (DTNBP1). Arch Gen Psychiatry 2004; 61: 336–344. Am J Med Genet B Neuropsychiatr Genet 2008; e-pub ahead 35 Numakawa T, Yagasaki Y, Ishimoto T, Okada T, Suzuki T, Iwata N of print. et al. Evidence of novel neuronal functions of dysbindin, a 12 Breen G, Prata D, Osborne S, Munro J, Sinclair M, Li T et al. susceptibility gene for schizophrenia. Hum Mol Genet 2004; 13: Association of the dysbindin gene with bipolar affective disorder. 2699–2708. Am J Psychiatry 2006; 163: 1636–1638. 36 Donohoe G, Morris DW, Clarke S, McGhee KA, Schwaiger S, 13 Zinkstok JR, de Wilde O, van Amelsvoort TA, Tanck MW, Baas F, Nangle JM et al. Variance in neurocognitive performance is Linszen DH. Association between the DTNBP1 gene and intelli- associated with dysbindin-1 in schizophrenia: a preliminary gence: a case–control study in young patients with schizophrenia study. Neuropsychologia 2007; 45: 454–458. and related disorders and unaffected siblings. Behav Brain Funct 37 Donohoe G, Morris DW, De Sanctis P, Magno E, Montesi JL, 2007; 3:19. Garavan HP et al. Early visual processing deficits in dysbindin- 14 O’Donovan MC, Williams NM, Owen MJ. Recent advances in the associated schizophrenia. Biol Psychiatry 2008; 63: 484–489. genetics of schizophrenia. Hum Mol Genet 2003; 12,SpecNo2 38 Corvin A, Donohoe G, Nangle JM, Schwaiger S, Morris D, Gill M. A R125–R133. dysbindin risk haplotype associated with less severe manic-type 15 Owen MJ, Williams NM, O’Donovan MC. The molecular genetics symptoms in psychosis. Neurosci Lett 2008; 431: 146–149. of schizophrenia: new findings promise new insights. Mol 39 Chubb JE, Bradshaw NJ, Soares DC, Porteous DJ, Millar JK. The Psychiatry 2004; 9: 14–27. DISC locus in psychiatric illness. Mol Psychiatry 2008; 13: 36–64. 16 Ross CA, Margolis RL, Reading SA, Pletnikov M, Coyle JT. 40 Di Pietro SM, Falcon-Perez JM, Tenza D, Setty SR, Marks MS, Neurobiology of schizophrenia. Neuron 2006; 52: 139–153. Raposo G et al. BLOC-1 interacts with BLOC-2 and the AP-3 17 Weickert CS, Straub RE, McClintock BW, Matsumoto M, complex to facilitate protein trafficking on endosomes. Mol Biol Hashimoto R, Hyde TM et al. Human dysbindin (DTNBP1) gene Cell 2006; 17: 4027–4038. expression in normal brain and in schizophrenic prefrontal cortex 41 Schweizer FE, Ryan TA. The synaptic vesicle: cycle of exocytosis and midbrain. Arch Gen Psychiatry 2004; 61: 544–555. and endocytosis. Curr Opin Neurobiol 2006; 16: 298–304. 18 Weickert CS, Rothmond DA, Hyde TM, Kleinman JE, Straub RE. 42 Talbot K, Cho DS, Ong WY, Benson MA, Han LY, Kazi HA et al. Reduced DTNBP1 (dysbindin-1) mRNA in the hippocampal Dysbindin-1 is a synaptic and microtubular protein that binds formation of schizophrenia patients. Schizophr Res 2008; 98: brain snapin. Hum Mol Genet 2006; 15: 3041–3054. 105–110. 43 Ni X, Trakalo J, Valente J, Azevedo MH, Pato MT, Pato CN et al. 19 Talbot K, Eidem WL, Tinsley CL, Benson MA, Thompson EW, Human p53 tumor suppressor gene (TP53) and schizophrenia: Smith RJ et al. Dysbindin-1 is reduced in intrinsic, glutamatergic case–control and family studies. Neurosci Lett 2005; 388: 173–178. terminals of the hippocampal formation in schizophrenia. J Clin 44 Thiselton DL, Vladimirov VI, Kuo PH, McClay J, Wormley B, Invest 2004; 113: 1353–1363. Fanous A et al. AKT1 is associated with schizophrenia across 20 Bray NJ, Preece A, Williams NM, Moskvina V, Buckland PR, Owen MJ multiple symptom dimensions in the Irish study of high density et al. Haplotypes at the dystrobrevin binding protein 1 (DTNBP1) schizophrenia families. Biol Psychiatry 2008; 63: 449–457.

Molecular Psychiatry DTNBP1 gene features and networks AY Guo et al 29 45 Jeong H, Mason SP, Barabasi AL, Oltvai ZN. Lethality and 55 Culligan K, Ohlendieck K. Diversity of the brain dystrophin- centrality in protein networks. Nature 2001; 411: 41–42. complex. J Biomed Biotechnol 2002; 2: 31–36. 46 Rizo J, Chen X, Arac D. Unraveling the mechanisms of synapto- 56 Maden M. Retinoic acid in the development, regeneration and tagmin and SNARE function in neurotransmitter release. Trends maintenance of the nervous system. Nat Rev Neurosci 2007; 8: Cell Biol 2006; 16: 339–350. 755–765. 47 Ruder C, Reimer T, Delgado-Martinez I, Hermosilla R, Engelsberg A, 57 Mey J, McCaffery P. Retinoic acid signaling in the nervous system Nehring R et al. EBAG9 adds a new layer of control on large dense- of adult vertebrates. Neuroscientist 2004; 10: 409–421. core vesicle exocytosis via interaction with Snapin. Mol Biol Cell 58 Palha JA, Goodman AB. Thyroid hormones and retinoids: a 2005; 16: 1245–1257. possible link between genes and environment in schizophrenia. 48 Ilardi JM, Mochida S, Sheng ZH. Snapin: a SNARE-associated Brain Res Rev 2006; 51: 61–71. protein implicated in synaptic transmission. Nat Neurosci 1999; 2: 59 Rao ML, Kolsch H. Effects of estrogen on brain development and 119–124. neuroprotection—implications for negative symptoms in schizo- 49 Sudhof TC. The synaptic vesicle cycle: a cascade of protein- phrenia. Psychoneuroendocrinology 2003; 28(Suppl 2): 83–96. protein interactions. Nature 1995; 375: 645–653. 60 Li XH, Kakkad B, Ong DE. Estrogen directly induces expression of 50 Zhang Q, Pangrsic T, Kreft M, Krzan M, Li N, Sul JY et al. Fusion- retinoic acid biosynthetic enzymes, compartmentalized between related release of glutamate from astrocytes. J Biol Chem 2004; 279: the epithelium and underlying stromal cells in rat uterus. 12724–12733. Endocrinology 2004; 145: 4756–4762. 51 Rees ML, Lien CF, Gorecki DC. Dystrobrevins in muscle and non- 61 Blitzer RD, Iyengar R, Landau EM. Postsynaptic signaling net- muscle tissues. Neuromuscul Disord 2007; 17: 123–134. works: cellular cogwheels underlying long-term plasticity. Biol 52 Rando TA. The dystrophin-glycoprotein complex, cellular signal- Psychiatry 2005; 57: 113–119. ing, and the regulation of cell survival in the muscular dystro- 62 Sullivan JM. Cellular and molecular mechanisms underlying phies. Muscle Nerve 2001; 24: 1575–1594. learning and memory impairments produced by cannabinoids. 53 Mehler MF. Brain dystrophin, neurogenetics and mental retarda- Learn Mem 2000; 7: 132–139. tion. Brain Res Rev 2000; 32: 277–307. 63 Hakak Y, Walker JR, Li C, Wong WH, Davis KL, Buxbaum JD et al. 54 Blake DJ, Nawrotzki R, Loh NY, Gorecki DC, Davies KE. beta- Genome-wide expression analysis reveals dysregulation of myeli- dystrobrevin, a member of the dystrophin-related protein family. nation-related genes in chronic schizophrenia. Proc Natl Acad Sci Proc Natl Acad Sci USA 1998; 95: 241–246. USA 2001; 98: 4746–4751.

Supplementary Information accompanies the paper on the Molecular Psychiatry website (http:// www.nature.com/mp)

Molecular Psychiatry