A Mutation in NPAS3 Segregates with Mental Illness in a Small Family

LETTERS TO THE EDITOR A mutation in NPAS3 segregates with mental illness in a small family

Molecular Psychiatry (2014) 19, 7–8; doi:10.1038/mp.2012.192; probands (26 with schizophrenia, 8 with schizoaffective disorders) 9 published online 22 January 2013 from a large genetic study of US pedigrees were selected for sequencing based on extensive family loading for schizophrenia (a sib with schizophrenia or schizoaffective disorder, and an Neuronal PAS domain protein 3 (NPAS3) is a member of the basic additional first degree (N ¼ 26) or second degree (N ¼ 8) family helix-loop-helix (bHLH) family of transcription factors. It contains a member with schizophrenia, schizoaffective disorder, bipolar bHLH domain involved in DNA binding, a PAS domain (acronym disorder, schizotypal personality disorder or psychotic disorder derived from the three initial proteins in which this motif was NOS). NPAS3 exons and 30 and 50 flanking regions were sequenced found) that alters the affinity for binding-partners in response to using PCR and capillary electrophoresis (ABI 3730). We detected a cellular signaling, and a transcriptional activation domain.1 bHLH G to A mutation in exon 8 at bp 910 (c.910 G4A) in a proband transcription factors regulate multiple processes, including neuro- with schizophrenia (Figure 1a). The mutation results in a valine to genesis, metabolism and circadian rhythms.2,3 Population studies isoleucine coding change at amino acid 304 (Figures 1b and c). We have associated common genetic variations of NPAS3 with schizo- obtained DNA from all available family members, and sequenced phrenia,4 bipolar disorder5 and antipsychotic efficacy.6 Two PCR products of the region to establish the presence of the nonsynonymous substitutions (A522P and A742V) have been mutation. The mother and a sib of the proband with schizophrenia previously associated with schizophrenia,7 and a balanced (9:14) have the mutation, as does a young sib (age 20 at ascertainment) translocation that disrupts NPAS3 was detected in a small family in with major depression. The father with schizotypal personality which the translocation segregates with schizophrenia.8 Like the disorder and a sister with major depression (age 30 at ascertain- effect of the 1:11 translocation on DISC1, the translocation through ment), do not have the mutation (Figure 1d). Unfortunately, the NPAS3 may result in a loss of gene and protein expression, or it family is unavailable for a reassessment of phenotype or a search could produce a truncated NPAS3 protein with a dominant nega- for more family members. tive effect. Both possibilities imply a loss of NPAS3 function. We used an allele-specific real-time PCR assay to screen for the We have identified a novel mutation in NPAS3 that segregates mutation in 600 probands with schizophrenia and 600 controls with mental illness in a small Caucasian US family. Thirty-four from the NIMH Genetics Initiative. The mutation was not detected

a d + -

+ - + + e

b GFP NPAS3 NPAS3-V304I

f 250 200 c V->I 304 150 ** 100

1 HLH PAS PAS 933 Neurite length 50 52-109 158-210 331-411 0

GFP NPAS3

NPAS3-V304I Figure 1. NPAS3 mutation. (a) Genotype. Proband (top) vs control (bottom). G/A heterozygosity in proband vs homozygous control. (b) Sequence. The G to A mutation creates a nonsynonymous mutation from valine to isoleucine at amino acid 304. (c) Location. The mutation in NPAS3 (red line) occurs between the PAS domains. (d) Pedigree with NPAS3 mutation. Filled ¼ schizophrenia. Diagonal lines ¼ major depression. Stippled ¼ schizotypal personality disorder. þ¼mutation present. The NPAS3-V304I mutation segregates with mental illness in this family. (e) NPAS3-V304I decreases neurite outgrowth. Primary mouse cortical neurons were cotransfected using lipofectamine 2000 at in vitro day 2 (2 DIV) with GFP and constructs encoding NPAS or NPAS3-V304I for 24 h. Images were analyzed using Image Pro plus. Representative cells transfected with GFP alone (left panel), WT NPAS3 (middle panel) or mutant NPAS3-V304I (right panel). (f) Quantification of longest neurite length for each condition, demonstrating that NPAS3-V304I significantly decreases longest neurite length. Each experiment was performed with four wells per condition; 10 individual cells were analyzed per well. Results represent four independent experiments, expressed as mean±s.e. using Kruskal À Wallis One-Way analysis of variance on ranks with Dunn’s post-hoc test (**Po0.001). Letters to the Editor 8 in either affected or unaffected individuals. The mutation REFERENCES (rs146677388) was detected in 2 of 2184 genomes (both of 1 Brunskill EW, Witte DP, Shreiner AB, Potter SS. Mech Dev 88: 237–241 1999. African-American individuals) examined in the 1000 Genome 2 Pieper AA, Wu X, Han TW, Estill SJ, Dang Q, Wu LC et al. Proc Natl Acad Sci USA Project (MAF ¼ 0.09%). 2005; 102: 14052–14057. The segregation of V304I with psychiatric disease in the small 3 Sha L, MacIntyre L, Machell JA, Kelly MP, Porteous DJ, Brandon NJ et al. Mol US family is intriguing. The family is too small to reach a definitive Psychiatry 2012; 17: 267–279. conclusion about the etiologic role of NPAS3-V304I, and detection 4 Pickard BS, Christoforou A, Thomson PA, Fawkes A, Evans KL, Morris SW et al. Mol of the mutation in two individuals of unknown phenotype does Psychiatry 2009; 14: 874–884. 5 Ferreira MA, O’Donovan MC, Meng YA, Jones IR, Ruderfer DM, Jones L et al. Nat not exclude an etiologic role. Nonetheless, the biological effects of 10 Genet 2008; 40: 1056–1058. valine to isoleucine mutations in amyloid precursor protein and 6 Lavedan C, Licamele L, Volpi S, Hamilton J, Heaton C, Mack K et al. Mol Psychiatry 11 transthyretin, and the location of the mutation in a region likely 2009; 14: 804–819. 12 related to DNA binding capacity and protein stability, support an 7 Macintyre G, Alford T, Xiong L, Rouleau GA, Tibbo PG, Cox DW. Schizophr Res 2010; etiologic role for NPAS3-V304I. We therefore attempted to 120: 143–149. determine if the mutation changes NPAS3 function. Constructs 8 Kamnasaran D, Muir WJ, Ferguson-Smith MA, Cox DW. JMedGenet2003; 40:325–332. consisting of GFP, full length NPAS3 complementary DNA, or 9 DeLisi LE, Shaw SH, Crow TJ, Shields G, Smith AB, Larach VW et al. Am J Psychiatry full length NPAS3-V304I cDNA were cloned into the pCMV6 2002; 159: 803–812. mammalian expression vector and demonstrated to have 10 Mullan M, Tsuji S, Miki T, Katsuya T, Naruse S, Kaneko K et al. Neurobiol Aging 1993; 14: 407–419. comparable expression levels by western blotting (data not 11 Jenne DE, Denzel K, Bla¨tzinger P, Winter P, Obermaier B, Linke RP et al. Proc Natl shown). These constructs were transfected into mouse primary Acad Sci USA 1996; 93: 6302–6307. neurons at in vitro day 2 (2 DIV) such that cells expressed GFP, GFP 12 Chapman-Smith A, Whitelaw ML. J Biol Chem 2006; 281: 12535–12545. and NPAS3 or GFP and NPAS3-V304I. Cells were imaged and 13 Ozeki Y, Tomoda T, Kleiderlein J, Kamiya A, Bord L, Fujii K et al. Proc Natl Acad Sci assayed for neurite outgrowth. There was a significant decrease in USA 2003; 100: 289–294. neurite outgrowth with NPAS3-V304I compared with GFP or GFP plus NPAS3 (Figure 1e and f), supporting our hypothesis of the biological relevance of V304I. We recognize that overexpression of a protein does not reflect a normal physiological state, and only provides a limited model of PTSD risk is associated with endogenous protein function. However, neurite outgrowth has been shown to be a sensitive measure of neuronal perturbation used to determine the potential relevance of other genes linked to BDNF Val66Met and BDNF schizophrenia, including DISC1.13 Therefore, a difference in neurite outgrowth between NPAS3 and NPAS3-V304I provides evidence overexpression for a functional role of the mutation. Identification of NPAS3-V304I helps strengthen the hypothesis that NPAS3 is involved in schizophrenia pathogenesis. Searching Molecular Psychiatry (2014) 19, 8–10; doi:10.1038/mp.2012.180; for other NPAS3 mutations, exploring the relationship between published online 15 January 2013 NPAS3 variation and human brain structure and function, and developing animal models of NPAS3 variants will further define the functional properties of normal and variant NPAS3 and possibly provide insights into the pathogenesis of schizophrenia Brain-derived neurotrophic factor (BDNF), which regulates neuronal and related illnesses. survival, growth, differentiation and synapse formation, broadly regulates the stress response.1 A common single-nucleotide polymorphism in the BDNF gene (Val66Met) influences hippo- CONFLICT OF INTEREST campal volume,2 memory3 and appears related to susceptibility to The authors declare no conflict of interest. a variety of neuropsychiatric disorders, including posttraumatic stress disorder (PTSD).1 Therefore, we hypothesized that BDNF may be an important mediator of the abnormal stress response ACKNOWLEDGEMENTS contributing to PTSD. We analyzed genetic variation in 461 and This work was supported by The Brain and Behavior Research Foundation Young the plasma levels of BDNF in 68 US. Army Special Operations Investigator Award and the Milton and Tamar Maltz Family Foundation, and the soldiers deployed during the Iraq and Afghanistan wars. A Hatten S. Yoder III Memorial Fellowship. probable-PTSD diagnosis was made if subjects endorsed Diagnostic and Statistical Manual of Mental Disorders-IV PTSD LYu1,6, N Arbez1, LG Nucifora1, GL Sell2, LE DeLisi3, CA Ross1,2,4,5, criteria (at least one re-experiencing, three avoidance and two RL Margolis1,4,5 and FC Nucifora Jr1 hyper-arousal symptoms) and had a total PTSD check list (PCL) 1Division of Neurobiology, Department of Psychiatry and Behavioral score of 50 or above (see Supplementary Methods). We found that Sciences, Johns Hopkins University School of Medicine, Baltimore, 42 (9%) participants met criteria for probable-PTSD, while the MD, USA; remaining 419 (91%) subjects did not. In those with probable- 2The Solomon H. Snyder Department of Neuroscience, Johns Hopkins PTSD, the frequency of Met/Met was nearly threefold higher than University School of Medicine, Baltimore, MD, USA; in the controls. The allelic frequency of Met was twofold higher in 3Veterans Affairs, Boston Healthcare System, Harvard Medical School, individuals with probable-PTSD than in controls (Supplementary Brockton, MA, USA; Information). BDNF levels in plasma of the subjects with PTSD 4Department of Neurology, Johns Hopkins University School of were significantly higher than those in the non-PTSD controls. Medicine, Baltimore, MD, USA and These results agree with evidence suggesting that BDNF is 5Program in Cellular and Molecular Medicine, Johns Hopkins involved in the development of PTSD.1 Met carriers have a University School of Medicine, Baltimore, MD, USA smaller hippocampus volume, relative to controls2 and perform E-mail: [email protected] more poorly than controls (Val/Val) on memory tasks.4 Veterans 6Present address: Division of Molecular Genetics, Department of with psychotic PTSD carry more Met alleles than non-psychotic Pediatrics, Columbia University Medical Center, New York, NY, USA veterans with PTSD or veterans without PTSD,5 even though a

Molecular Psychiatry (2014), 7 – 13 & 2014 Macmillan Publishers Limited