Assessment of the Neuropeptide S System in Anxiety Disorders

Assessment of the Neuropeptide S System in Anxiety Disorders Jonas Donner, Rita Haapakoski, Sini Ezer, Erik Melén, Sami Pirkola, Mònica Gratacòs, Marco Zucchelli, Francesca Anedda, Lovisa E. Johansson, Cilla Söderhäll, Christina Orsmark-Pietras, Jaana Suvisaari, Rocío Martín-Santos, Marta Torrens, Kaisa Silander, Joseph D. Terwilliger, Magnus Wickman, Göran Pershagen, Jouko Lönnqvist, Leena Peltonen, Xavier Estivill, Mauro D’Amato, Juha Kere, Harri Alenius, and Iiris Hovatta Background: The G protein-coupled receptor neuropeptide S receptor 1 (NPSR1) and its ligand neuropeptide S (NPS) form a signaling system mainly implicated in susceptibility to asthma and inflammatory disorders in humans and regulation of anxiety and arousal in rodents. We addressed here the role of NPS and NPSR1 as susceptibility genes for human anxiety disorders. Methods: We performed comprehensive association analysis of genetic variants in NPS and NPSR1 in three independent study samples. We first studied a population-based sample (Health 2000, Finland) of 321 anxiety disorder patients and 1317 control subjects and subsequently a Spanish clinical panic disorder sample consisting of 188 cases and 315 control subjects. In addition, we examined a birth cohort of 2020 children (Barn Allergi Miljö Stockholm Epidemiologi [BAMSE], Sweden). We then tested whether alleles of the most significantly associated single nucleotide polymorphisms alter DNA-protein complex formation in electrophoretic mobility shift assays. Finally, we compared acute stress responses on the gene expression level in wild-type and Npsr1⫺/⫺ mice. Results: We confirmed previously observed epidemiological association between anxiety and asthma in two population-based cohorts. Single nucleotide polymorphisms within NPS and NPSR1 associated with panic disorder diagnosis in the Finnish and Spanish samples and with parent-reported anxiety/depression in the BAMSE sample. Moreover, some of the implicated single nucleotide polymorphisms potentially affect transcription factor binding. Expression of neurotrophin-3, a neurotrophic factor connected to stress and panic reaction, was significantly downregulated in brain regions of stressed Npsr1⫺/⫺ mice, whereas interleukin-1 beta, an active stress-related immunotransmitter, was upregulated. Conclusions: Our results suggest that NPS-NPSR1 signaling is likely involved in anxiety. Key Words: Anxiety disorders, asthma, genetic association, lexicon, neuropeptide S, SNP europeptide S receptor 1 (NPSR1) and its ligand neuropeptide S (NPS) form a signaling system with multiple functions. The gene NPSR1 was originally identified as a susceptibility gene for asthma and atopy (1), and this association has been replicated in several populations (e.g., 2– 4) and in another immunologic phenotype, inflammatory bowel disease (5). The NPS-NPSR1 system also modulates neurobiological phenotypes. The gene NPSR1 was identified as a possible mediator of bedtime and sleepiness (6), and Npsr1 is highly expressed in rat brain regions regulating arousal, anxiety, learning, and memory (7). Accordingly, NPS modulates arousal and anxiety,

N

with a unique behavioral profile of simultaneously inducing wakefulness and exerting anxiolytic effects when centrally administered in rodents (8,9). Moreover, the NPS-NPSR1 system was suggested to mediate some behavioral effects of nicotine and caffeine (10,11) and to regulate ethanol and food intake (12,13). Interestingly, epidemiological comorbidity of anxiety disorders and asthma is well documented in both clinical and community based samples (14 –17). Patients suffering from both asthma and an anxiety disorder have poorer asthma control, increased functional impairment, and decreased quality of life compared with asthma patients without a psychiatric diagnosis (18 –20). Several possible explanations for the comorbidity exist, including shared genetic vulnerability, but the underlying mechanisms remain unknown

From the Research Program of Molecular Neurology (JD, IH), Biomedicum Helsinki; Department of Medical Genetics (JD, SE, LP, JK, IH) and Faculty of Medicine (JDT), University of Helsinki; Finnish Institute of Occupational Health (RH, HA); Folkhälsan Institute of Genetics (SE, JK); Department of Mental Health and Substance Abuse Services (SP, JS, JL, IH), National Institute for Health and Welfare; Department of Psychiatry (SP, JL), Helsinki University Central Hospital; and Public Health Genomics Unit (KS, LP), National Institute for Health and Welfare, and Institute of Molecular Medicine, Finland, Helsinki, Finland; Institute of Environmental Medicine (EM, MW, GP), Karolinska Institutet; Astrid Lindgren Children’s Hospital (EM) and Clinical Research Centre (MZ, FA, CO-P, MD, JK), Karolinska University Hospital; and Sachs’s Children’s Hospital (MW), Södersjukhuset, Stockholm; and Department of Biosciences and Nutrition (MZ, FA, LEJ, CS, CO-P, MD, JK), Karolinska Institutet, Huddinge, Sweden; Genes and Disease Program (MG, XE), Center for Genomic Regulation; Centro de Investigación Biomédica en Red en Epidemiología y Salud Pública (MG, XE); Neuropsychopharmacology Programme (RM-S), Institut Municipal d’Investigació Mèdica, and Department of Drug Abuse and Psychiatry (MT), Institut d’Atenció Psiquiàtrica, Hospital del Mar; Psychiatry Service (RM-S), Neurosciences Institute, Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi i Sunyer, Centro de Investigación Biomédica en Red de Salud Mental; and Experimental and Health Sciences Department (XE), Pompeu Fabra University, Barcelona, Spain; Istituto di Neurogenetica e Neurofarmacologia (FA), Consiglio Nazionale delle Ricerche, Monserrato, Italy; Department of Genetics and Development (JDT), Department of Psychiatry, Columbia Genome Center, Columbia University; and Division of Medical Genetics (JDT), New York State Psychiatric Institute, New York, New York; Wellcome Trust Sanger Institute (LP), Hinxton, United Kingdom; The Broad Institute of Massachusetts Institute of Technology and Harvard (LP), Cambridge, Massachusetts. Address correspondence to Juha Kere, M.D., Ph.D., Karolinska Institutet, Department of Biosciences and Nutrition, Hälsovägen 7, SE-14157, Huddinge, Sweden; E-mail: [email protected]. Received Oct 14, 2009; revised May 25, 2010; accepted May 28, 2010.

0006-3223/$36.00 doi:10.1016/j.biopsych.2010.05.039

BIOL PSYCHIATRY 2010;68:474 – 483 © 2010 Society of Biological Psychiatry

BIOL PSYCHIATRY 2010;68:474 – 483 475

J. Donner et al. (15). The NPS-NPSR1 system could represent a genetic link between these diseases. Motivated by the evidence implicating NPS and NPSR1 in regulation of anxiety-like behavior in rodents and by the comorbidity between asthma and anxiety disorders in humans, we aimed to elucidate the predisposing role of this system in human anxiety. Only one study has previously addressed the role of NPSR1 in human psychiatric disorders. It focused on one functional single nucleotide polymorphism (SNP; rs324981) and reported some evidence for male-specific underrepresentation of a genotype in panic disorder (PD) (21). We performed comprehensive association analysis testing of genetic variation in NPS and NPSR1 in three independent study samples from Finland, Spain, and Sweden. In the epidemiological Finnish and Swedish samples, we also evaluated the comorbidity of asthma and anxiety. We further tested functionally whether the most significantly associated SNPs alter transcription factor binding. In addition, we examined whether neuropeptide S receptor 1 knockout (Npsr1⫺/⫺) mice respond differently to stress compared with wild-type (WT) mice on a gene expression level.

Methods and Materials Study Samples Epidemiological Cohorts. The comorbidity of asthma and anxiety was evaluated in two epidemiological cohorts, the Finnish population-based Health 2000 sample and the Swedish Barn Allergi Miljö Stockholm Epidemiologi (BAMSE) birth cohort (Table 1). Subjects for the genetic association analyses were also derived from these cohorts. The Health 2000 sample consists of 6005 subjects interviewed with the Composite International Diagnostic Interview for the most common DSM-IV depressive, anxiety, and alcohol use disorders (22,23). Consensus diagnoses of asthma were based on assessments of asthma-related phenotypes of the subjects, including physician’s clinical examination, spirometry, and register data regarding use of health care services and reimbursed asthma medication. The BAMSE birth cohort included 4089 children from Stockholm (2,24). At age 8, parents completed questionnaires on their chil-

dren’s symptoms related to asthma and other allergic diseases, and EuroQol (EQ-5D) questions (25) were used to assess quality of life. To measure anxiety/depression, parents were asked: “Indicate which statement best describes your child’s health condition today,” and the reply options were 1) not anxious or depressed, 2) moderately anxious or depressed, or 3) extremely anxious or depressed. As only .1% of the children were extremely anxious or depressed, we combined them with the moderately anxious or depressed for the analyses. Asthma was defined as at least four episodes of wheezing during the last 12 months or at least one episode of wheezing during the same period if the child was receiving inhaled steroids (26). Samples for Genetic Association Analyses. Genetic variation in NPS and NPSR1 genes was tested for association to anxiety disorders or anxiety symptoms in three independent samples (Health 2000, a Spanish PD sample, and the BAMSE sample; Table 1). The Finnish sample was described previously (22,27). Briefly, it was derived from participants of the Health 2000 Study and consists of 321 adults diagnosed with an anxiety disorder (n ⫽ 282) or an anxiety disorder subthreshold diagnosis (n ⫽ 39) according to DSM-IV criteria. In addition to analyzing this combined sample, we also performed separate analyses of the major diagnostic subgroups. We initially used 653 control subjects matched for sex, age (⫾ 1 year), and hospital catchment area. Control subjects lacked anxiety or major mental disorders, had no missing data in anxiety screen questions, and had explicit negative diagnoses for all symptoms of anxiety. Most selected respondents had negative replies to every anxiety-related question in the Composite International Diagnostic Interview and a sum score of 0 in General Health Questionnaire-12, indicating absence of symptoms of psychic distress. Subsequently, we identified an additional set of control subjects (n ⫽ 664; age matching relaxed to ⫾ 2 years) from the same cohort, which was genotyped for SNPs showing the most significant evidence for association. In the genetic analyses, cases of each diagnostic subgroup were compared with the corresponding matched control group. The Spanish sample consisted of 188 adult Spanish Cauca-

Table 1. Characteristics of the Three Investigated Human Study Samples Health 2000 Country of Origin Ascertainment

Diagnostic Instruments and Criteria Anxiety Asthma

Investigated Anxiety Phenotypes Sample Sizes Epidemiological studies Genetic studies

Sex Distribution Mean Age ⫾ SD

Barcelona

BAMSE

Finland Cases and control subjects identified from the same population-based epidemiological cohort

Spain Clinical recruitment of psychiatry outpatients

Sweden Population-based birth cohort

M-CIDI Clinical examination, spirometry, and register data

SCID-CV N/A

12-month DSM-IV anxiety disorder diagnoses

Active DSM-IV panic disorder

EQ-5D ⱖ 4 episodes of wheezing during the last 12 months or ⱖ 1 episode with use of inhaled steroids Parent-reported moderate/extreme anxiety or depression

6005 subjects 321 anxiety disorder patients (panic disorder, generalized anxiety disorder, social phobia, agoraphobia, and phobia not otherwise specified). 653 (or 1317 for some markers) matched control subjects. 63% females, 37% males 49.8 ⫾ 12.7

N/A 188 panic disorder patients. 315 population control subjects.

2033 subjects 138 children with moderate/extreme anxiety or depression. 1882 children with no anxiety or depression.

75% females, 25% males 35.5 ⫾ 9.3

48% females, 52% males 8.4 ⫾ .5

BAMSE, Barn Allergi Miljö Stockholm Epidemiologi; EQ-5D, EuroQol 5D; M-CIDI, Composite International Diagnostic Interview; SCID-CV, Structured Clinical Interview for DSM-IV Disorders: Clinical Version

www.sobp.org/journal

476 BIOL PSYCHIATRY 2010;68:474 – 483 sian PD outpatients from the Psychiatry outpatient unit in Hospital del Mar, Barcelona. Diagnoses were independently assigned by two senior psychiatrists after the Structured Clinical Interview for DSM-IV Disorders: Clinical Version. Exclusion criteria were age ⬍ 18 years, organic brain syndromes, psychoactive substance abuse disorders (except nicotine abuse), comorbid DSM-IV Axis I disorders apart from anxiety disorders, life prevalence of mood disorder, severe organic or neurological pathology including partial epilepsy, and illiteracy. Control individuals consisted of 315 nonpsychiatrically screened blood donors, recruited from the Blood and Tissue Bank of the Catalan Health Service and group matched to cases for ethnicity. At age 8, children of the BAMSE study were invited to clinical testing, and blood samples were obtained. DNA was extracted from 2033 samples after exclusion of samples with too little blood, incomplete questionnaire data, or lack of parental consent to genetic analysis. Information concerning genetic power calculations and measures taken to avoid population stratification is available in Supplementary Methods and Materials in Supplement 1. Genotyping Single nucleotide polymorphism genotyping was performed using Sequenom MassARRAY technology with iPLEX chemistry (Sequenom, San Diego, California) in the Finnish and Swedish samples. In the Spanish sample, the SNPplex Multiplex Genotyping System (Applied Biosystems, Foster City, California) was used. Additional information on SNP selection, assay validation, and quality control is available in Supplementary Methods and Materials in Supplement 1. Statistical Analyses The epidemiological association between asthma and anxiety disorders in the Health 2000 study sample was estimated with simple bivariate and multivariate analyses (SPSS, Chicago, Illinois). Predictors of asthma were evaluated by logistic regression modeling with age, sex, length of education, and specific anxiety disorders as independent variables. In the BAMSE sample, the association between asthma and anxiety was evaluated with standard logistic regression (STATA, College Station, Texas), with adjustments for sex, heredity for allergic diseases, maternal smoking during pregnancy and/or at childbirth, breastfeeding duration, and maternal age at enrollment (previously identified as potential confounders). Testing of pointwise genetic association was performed as detailed earlier (27). Briefly, we used a conventional 2 ⫻ 2 contingency table

J. Donner et al. likelihood-ratio test of independence of SNP allele counts in cases and control subjects. Haplotype blocks were identified with the algorithm of Gabriel et al. (28) in Haploview (Broad Institute of Massachusetts Institute of Technology and Harvard, Cambridge, Massachusetts; http://www.broad.mit.edu/mpg/haploview). Specific haplotypes within the identified block windows were subsequently tested for association with Unphased (Frank Dudbridge, Cambridge, United Kingdom; http://www.mrc-bsu.cam.ac.uk/ personal/frank/software/unphased) (29). Empirical p values were estimated for all SNPs and haplotypes by evaluating 10,000 permutations of the dataset. Haplotype phylogenies were estimated using the maximum parsimony algorithm in MEGA (v.4.0.2; Sudhir Kumar and colleagues, Tempe, Arizona; http:// www.megasoftware.net/index.html) (30). Additional information on statistical methodology is available in Supplementary Methods and Materials in Supplement 1. Electrophoretic Mobility Shift Assays The human neuroblastoma cell line SH-SY5Y (American Tissue Culture Collection, Manassas, Virginia; http://www.atcc.org) was grown in Dulbecco’s Modified Eagle Medium with 10% fetal calf serum (Sigma-Aldrich, St. Louis, Missouri). Nuclear extracts were prepared according to standard protocols (31). Duplex probes encompassing NPSR1 polymorphisms were produced by annealing oligonucleotide pairs specific for each allele at the polymorphic site (2.5 ␮mol/L; Table S1 in Supplement 2). Binding reactions were performed in buffer containing 25 mmol/L HEPES pH 7.9 (Invitrogen, Carlsbad, California), 150 mmol/L potassium chloride (SigmaAldrich), 10% glycerol (Sigma-Aldrich), and 5 mmol/L dithiothreitol (Invitrogen), with 3 ␮g of nuclear extract, approximately .5 ng of [␣-32P] Klenow-labeled probe (New England Biolabs, Ipswich, Massachusetts), and 3 ␮g of Poly(deoxyinosinic-deoxycytidylic) (GE Healthcare, Buckinghamshire, United Kingdom) in a total volume of 20 ␮L. A 10-minute preincubation with either unlabeled duplexes in 200⫻ excess (competition experiments) or 2 ␮g of antibody (super shift assays) was performed before adding the [␣-32P] Klenow-labeled probe. Antibodies against sex determining region Ybox 5 (SOX-5) and runt-related transcription factor 2 (RUNX2) (sc20091X and sc-8566X, respectively; Santa Cruz Biotechnology, Santa Cruz, California) were used. Polyacrylamide gels (6%) were run at 4°C for 3.5 hours at 300 V, dried, and autoradiographed to visualize DNA-protein complexes. Experiments were repeated on two separate preparations of nuclear extracts and with two differ-

Figure 1. Genomic structure of NPSR1 (A) and NPS (B) and relative positions of the investigated single nucleotide polymorphisms (SNPs). Open circles below the SNP rsnumbers indicate that the marker was analyzed in the corresponding study sample. Closed circles indicate that the SNP was analyzed and that an association with p ⬍ .05 to either panic disorder (Finland and Spain) or EuroQol 5D questionnaire anxiety/depression (Finland and Sweden) was observed in an allelic likelihood-ratio test. Closed diamonds designate associations with p ⬍ .05 but to opposite alleles. The spans of haplotypes associating to panic disorder in the Finnish sample are illustrated with bars above the genomic structure of the genes, along with observed haplotypic p values. (Figures created using Locusview2.0 output [T. Petryshen, A. Kirby, and M. Ainscow, unpublished software]). agoraph., agoraphobia; ESP, Spain; FIN, Finland; NPS, neuropeptide S; NPSR1, neuropeptide S receptor 1; PD, panic disorder; SWE, Sweden.

www.sobp.org/journal

BIOL PSYCHIATRY 2010;68:474 – 483 477

J. Donner et al. ent sets of radiolabeled duplex probes. Prediction of transcription factor binding was done with SNPInspector and MatBase (Genomatix Software, Munich, Germany; http://www.genomatix.de).

effects within treatment groups was performed using unpaired t tests (SPSS). Additional methodological information is available in Supplementary Methods and Materials in Supplement 1.

Acute Immobilization Stress and Quantitative Real-Time Polymerase Chain Reation Analysis The Npsr1⫺/⫺ and WT male mice on a heterogeneous C57BL/6129/SvEvBrd background were obtained from Lexicon Genetics (The Woodlands, Texas). For acute immobilization stress, 12- to 16-week-old mice (four to six in each experimental group) were placed into ventilated 50-mL conical tubes, which were put into darkened plastic boxes to eliminate effects of vision, smell, and lighting. Experimental groups were 1) mice killed immediately after 1 hour restraint stress; 2) mice subjected to 1 hour restraint stress and allowed to recover for 1 hour in their home cages before killing; and 3) mice killed immediately after home cage removal, without additional handling, served as nonstressed control animals. Mice were killed by carbon dioxide inhalation, and blood was collected for serum corticosterone measurement. Brain regions were immediately dissected and stored for quantitative real-time polymerase chain reaction analysis of gene expression. The significance of genotype and stress treatment effects and their interaction was tested with factorial analysis of variance, and post hoc testing of genotype

Results Comorbidity of Anxiety and Asthma We evaluated the epidemiological association between asthma and anxiety in the Finnish Health 2000 cohort and in the Swedish BAMSE birth cohort. When analyzing specific DSM-IV anxiety disorders in the Finnish sample (PD, generalized anxiety disorder, social phobia, and agoraphobia), we noticed that subjects with pure agoraphobia had asthma significantly more often than subjects without pure agoraphobia (14.1% vs. 4.0%, ␹2 ⫽ 20.1, df ⫽ 1, p ⫽ 3.0*10⫺4). In further logistic regression modeling, female sex (odds ratio [OR] ⫽ 5.17, 95% confidence interval [CI95%] 2.23–12.00) and agoraphobia (OR ⫽ 1.64, CI95% 1.24 –2.17) remained the only significant explanators of asthma. At age 8, 7.5% of the children (n ⫽ 151) in the BAMSE study had asthma. Among them, 11.3% were reported by their parents as having presence of anxiety/depression compared with 6.5% among children with no asthma (crude p ⫽ .026, OR ⫽ 1.84, CI95% 1.08 –3.15, and adjusted p ⫽ .049, OR ⫽ 1.76, CI95% 1.00 –3.10).

Figure 2. Genomic structure, relative single nucleotide polymorphism positions, and linkage disequilibrium plots for NPSR1 (A) Finnish sample, (B) Spanish sample, (C) Swedish sample, and NPS (D) Finnish sample, (E) Spanish sample, (F) Swedish sample. The spans of analyzed haplotype blocks are shown with bars below the genomic structure. (Figure created using Locusview 2.0 output [T. Petryshen, A. Kirby, and M. Ainscow, unpublished software]). LOD, logarithm of odds; NPS, neuropeptide S; NPSR1, neuropeptide S receptor 1.

www.sobp.org/journal

478 BIOL PSYCHIATRY 2010;68:474 – 483

J. Donner et al.

Table 2. Panic Disorder, Single Marker Association Analysis Results with p ⬍ .05 in at Least One Diagnostic Group PD With and Without Agoraphobia Gene

SNP rs Number

Finnish Sample NPSR1 rs1819802 rs2125404 rs2168891 rs1963499 rs2530547 rs887020 rs2530548 rs2530552 rs411323 rs2530566 rs2530567 rs2531840 rs323917 rs324396 rs17199659 rs10263447 rs725902 NPS rs990310 rs1999635 rs11018195 Sample Size Spanish Sample NPS rs990310 rs11018195 Sample Size

SNP Position

Minor Allele

5=-upstream Predicted promoter Predicted promoter Predicted promoter Predicted promoter Predicted promoter Intron 1 Intron 1 Intron 1 Intron 1 Intron 1 Intron 1 Intron 2 Intron 2 Intron 4 Intron 4 Intron 4 Exon 2, NS Intron 2 Intron 2

Exon 2, NS Intron 2

PD Without Agoraphobia

MAF Cases

MAF Controls

Allelic LRT p Value

MAF Cases

MAF Controls

Allelic LRT p Value

G C A A T A G G T G G A G T C C G T A G

.356 .108 .074 .117 .431 .500 .380 .370 .262 .523 .278 .370 .084 .341 .181 .185 .319 .094 .089 .094 108

.413 .063 .055 .067 .358 .413 .285 .279 .320 .421 .215 .278 .046 .385 .109 .262 .238 .050 .037 .050 218 or 443a

.163 .047 .343 .040 .052 .019 .010 .012 .129 .005 .053 .011 .065 .270 .006 .018 .014 .021 .007 .022

.353 .110 .093 .122 .440 .487 .373 .367 .236 .520 .301 .367 .101 .311 .207 .207 .340 .075 .068 .075 75

.453 .058 .044 .058 .338 .394 .268 .268 .353 .404 .205 .266 .041 .412 .099 .258 .219 .057 .047 .057 148 or 300a

.040 .068 .048 .024 .024 .033 .014 .021 .011 .009 .016 .020 .019 .037 .001 .186 .003 .422 .377 .424

T G

.014 .013

.082 .083

.008 .008

188

.030 .030 315

128

.030 .030

.030 .029

315

Empirical p values based on 10,000 permutations are shown. MAF, minor allele frequency; LRT, likelihood-ratio test; NS, nonsynonymous; PD, panic disorder; SNP, single nucleotide polymorphism. a The rs numbers of SNPs that were genotyped in an additional Finnish control sample (n control subjects in total ⫽ 443) have been highlighted in bold. The remaining markers were genotyped only in the original set of Finnish control subjects (n ⫽ 218).

Specific Alleles and Haplotypes within NPSR1 and NPS Are Associated with Panic Disorder Results from the allele and haplotype association tests performed in the three studied samples are graphically summarized in Figure 1, and intermarker linkage disequilibrium plots for both genes are illustrated in Figure 2. We first tested individually 59 SNPs for association to anxiety disorders in the Finnish Health 2000 sample. Several SNPs in both NPSR1 and NPS showed evidence for association to PD specifically (Table 2 and Table S2 in Supplement 2). Exclusion of patients and control subjects with asthma (n ⫽ 4 and n ⫽ 21, respectively) verified that the observed associations with PD were not due to the confounding effect of asthma status (25 of 28 associations presented in Table 2 remained significant at p ⬍ .05). We subsequently split the Finnish sample into PD without agoraphobia and PD with agoraphobia and analyzed these subgroups separately due to the epidemiological comorbidity between asthma and pure agoraphobia in this sample (Table 2 and Table S2 in Supplement 2). However, the association signal is coming from both subgroups, without one being more significantly associated than the other. Because the most significant evidence for association was observed in the Finnish PD subsample, we then followed up these findings in a separate clinical sample consisting of 188 adult Spanish PD outpatients and 315 control individuals. Thirty-five SNPs were tested for association to PD and PD with agoraphobia. Two NPS SNPs in complete linkage disequilibrium (rs990310 and rs11018195) associated to PD with agoraphobia (Figure 1, Table 2). www.sobp.org/journal

Although they also associated to PD in the Finnish sample, the risk alleles were different. Results for PD without agoraphobia in the Spanish sample (only 60 cases) are shown in Table S2 in Supplement 2. Based on the individual SNP findings in PD, we additionally tested 37 specific haplotypes within eight haplotype blocks (Figure 2) for association in the Finnish sample. Blocks including specific haplotypes associating to PD and/or PD without agoraphobia (empirical p ⬍ .05) are described in detail in Figure S1 in Supplement 1. Haplotype phylogeny estimates supported the presence of one main risk haplotype per studied block. The NPS risk haplotype was tagged by the aforementioned rs990310, which causes a nonconservative amino acid substitution (S14L) within the signal peptide of the NPS precursor (Figure S1A in Supplement 1). A functional SNP (rs324981) divides haplotypes into two major clades within one of the analyzed NPSR1 blocks (Figure S1B in Supplement 1). Notably, only one of the haplotypes carrying the gain-of-function T-allele (Ile), resulting in a receptor 10 times more sensitive to NPS (32), confers increased PD susceptibility. No association (empirical p ⬍ .05) to 21 haplotypes within six haplotype blocks was detected in the Spanish sample. Further Evidence for Involvement of NPSR1 in Anxiety from a Swedish Birth Cohort Children in the Swedish BAMSE cohort were genotyped for SNPs in NPSR1 and NPS. We evaluated the genetic association between 28 SNPs and the same EQ-5D measure of anxiety/depression used in evaluation of comorbidity between asthma and anxiety. Five

BIOL PSYCHIATRY 2010;68:474 – 483 479

J. Donner et al. Table 3. EQ-5D Anxiety and Depression, Single Marker Association Results with p ⬍ .05

Minor Allele Frequencies

Gene Swedish Sample NPSR1

Sample Size Finnish Sample NPSR1 Sample Size

Moderate/Extreme Anxiety or Depression

No Anxiety or Depression

Allelic LRT p Value

SNP Position

Minor Allele

rs2125404 rs2168891 rs1963499 rs1379928a rs2609215a

Predicted promoter Predicted promoter Predicted promoter Intron 1 Intron 1

C A A C G

.153 .093 .147 .278 .126 138

.108 .059 .101 .205 .088 1882

.029 .038 .024 .006 .046

rs11770777 rs1859409b

Intron 1 3-downstream

C T

.246 .376 193

.187 .319 730

.012 .039

SNP rs Number

Empirical p values based on 10,000 permutations are shown. EQ-5D, EuroQol 5D; LRT, likelihood-ratio test; SNP, single nucleotide polymorphism. a Not genotyped in the Finnish sample. b Not genotyped in the Swedish sample.

SNPs in NPSR1 associated to this phenotype (Table 3). Notably, the same alleles of three of them (rs2125404, rs2168891, and rs1963499) were also implicated in PD in the Finnish sample (Table 2). The most significant SNP in the Swedish sample, rs1379928, was not genotyped in the Finnish or Spanish samples, but its proxy (rs2530567, HapMap [Utah residents with Northern and Western European ancestry; http://www.hapmap.org] CEU r2 ⫽ .90) also associated with PD in the Finnish sample. Excluding asthma patients (n ⫽ 151) did not significantly alter the results. As the EQ-5D was also included in the Health 2000 Study, we tested whether SNPs associated with the same measure also in this sample. Associations were not detected for the same SNPs as in the BAMSE sample, but two other SNPs showed evidence for association to this phenotype (Table 3). NPSR1 SNPs Affect Binding of Nuclear Proteins to DNA We evaluated with electrophoretic mobility shift assays whether NPSR1 SNPs rs2530548, rs1379928, rs2530566, rs17199659, and rs725902, showing the strongest evidence for genetic association (p ⬍ .01), are also of functional relevance. Qualitative and quantitative allelic differences in DNA-protein complex formation could be discerned for polymorphisms rs2530548 and rs2530566, suggesting that they may affect transcription factor binding to DNA (Figure 3). No allele-specific differences were observed for rs1379928, rs17199659, and rs725902 (data not shown). Competition experiments with unlabeled probes confirmed specificity of the interactions for rs2530548 and rs2530566. Interestingly, bioinformatic analysis predicted these polymorphisms to affect binding of transcription factors SOX-5 for rs2530548 and growth factor independent one transcription repressor, ovo-like 1, homeobox transcription factors A9 and B9, and RUNX2 for rs2530566. In an attempt to provide experimental support for the predictions, we selected SOX-5 and RUNX2 for testing in super shift assays based on suggested involvement in neuronal function and mental disorders (33,34). However, antibodies for these transcription factors did not shift the allele-specific binding patterns (Figure 3). Altered Gene Expression Responses in Stressed Npsr1ⴚ/ⴚ Mice One-hour immobilization stress increased the level of serum corticosterone, as expected [Mcontrols (SD) ⫽ 82 ng/mL (52),

Mimmobilized (SD) ⫽ 466 ng/mL (25); treatment main effects F (2,16) ⫽ 140.96, p ⫽ 6.92*10⫺11], but WT and Npsr1⫺/⫺ mice did not differ regarding this basic stress-related function. Experimentwide significant main effects of stress treatment on gene expression were observed for several genes in at least one brain region (Egr-1, Fos, Il1b, JunB, Klf10, and Tnf; Figure 4, Table S3 in Supplement 2). Results for genes showing at least nominally significant genotype main effects and genotype ⫻ stress treatment interactions are depicted in Figure 4. No expression of Npsr1 was detected in Npsr1⫺/⫺ mice, confirming successful deletion of the gene, and its expression did not change in response to stress in WT mice (Table S3 in Supplement 2). However, stress induced cortical upregulation of Nps 1 hour after immobilization in WT mice but not in their Npsr1⫺/⫺ littermates [Figure 4A; F (2,12) ⫽ 8.17, p ⫽ .006, nominally significant interaction]. Both Fos and JunB immediate early response genes showed cortical genotype main effects that, however, did not reach experiment-wide significance [Figure 4B,C; F (1,18) ⫽ 5.13, p ⫽ .036, and F (1,18) ⫽ 4.51, p ⫽ .048, respectively). We further observed increased cortical upregulation of Il1b in Npsr1⫺/⫺ mice compared with WT mice in response to stress [Figure 4D; genotype main effects F (1,18) ⫽ 17.62, p ⫽ .0005, experiment-wide significant]. Upregulation of Il1b in Npsr1⫺/⫺ mice was also seen in the hypothalamus immediately after immobilization [Figure 4E; genotype ⫻ stress treatment interaction F (2,18) ⫽ 4.15, p ⫽ .033, nominally significant], supporting the cortical finding. In contrast, both cortical and striatal expression of Ntf3 was decreased in Npsr1⫺/⫺ mice compared with WT mice after immobilization [Figure 4F,G; nominally significant genotype ⫻ stress treatment interactions F (2,18) ⫽ 5.09, p ⫽ .018, and F (2,29) ⫽ 4.25, p ⫽ .024, respectively, and experiment-wide significant striatal genotype main effects F (1,29) ⫽ 14.96, p ⫽ .0006].

Discussion We investigated the role of the asthma susceptibility gene NPSR1 and its ligand, the anxiolytic neuropeptide NPS, in human anxiety. First, we replicated previously observed comorbidity between asthma and anxiety disorders in the two epidemiological cohorts from which subjects for genetic analyses were identified. In the BAMSE cohort, we observed increased anxiety among children with asthma, as reported by others (17). In the Health 2000 cohort, www.sobp.org/journal

480 BIOL PSYCHIATRY 2010;68:474 – 483

Figure 3. Electrophoretic mobility shift assay of NPSR1 single nucleotide polymorphisms rs2530548 (with alleles A/G) and rs2530566 (with alleles A/G). Nuclear extracts from SH-SY5Y neuroblastoma cells were allowed to bind to radiolabeled probe pairs corresponding to the two alleles at each polymorphic site. Allele-specific differences (both qualitative and quantitative) in DNA-protein complex formation are highlighted by marking the unique or stronger band in either of the two lanes (alleles) for each single nucleotide polymorphism with arrows. Competition experiments with unlabeled probes in 200⫻ excess confirmed specificity of the interactions. The gel migration patterns were not altered by antibodies against transcription factors sex determining region Y-box 5 and runt-related transcription factor 2, bioinformatically predicted to bind to the respective polymorphic sites, in super shift assays. Comp, competition experiments; RUNX2, runt-related transcription factor 2; SOX-5, sex determining region Y-box 5.

we evaluated relationships between specific anxiety disorders and asthma and found comorbidity between asthma and agoraphobia. Similar results were previously observed in adolescents and were hypothesized to be attributable to conditioning by previous frightful episodes of asthma in a public place, parental overprotection of children with asthma, or shared genetic liability (35). Our results from genetic association analyses in a Finnish anxiety disorder sample and a Spanish follow-up PD sample suggest that NPSR1 and NPS may modulate predisposition mainly to PD, independently of asthma status. Some evidence for association to social phobia and agoraphobia was also seen. This distinction from other anxiety disorders, such as generalized anxiety disorder, is interesting, as PD and social phobia show some phenotypic similarity in the form of avoidance behavior. The associations discovered to PD were particularly intriguing, as it is the anxiety disorder most frequently co-occurring with asthma (15,36). This comorbidity appears bidirectional, with active asthma predicting subsequent PD www.sobp.org/journal

J. Donner et al. and vice versa (37). Asthma and PD also share symptoms such as a sensation of being smothered, choking, hyperventilation-induced dyspnea, and increased anxiety (38). In addition, genetic variation in NPSR1 associated with a measure of anxious depression in the BAMSE cohort. This unspecific phenotype is different from and cannot be directly compared with the DSM-IV anxiety disorder diagnoses used in the Finnish and Spanish samples. However, in the Finnish sample, it overlapped considerably with diagnoses of agoraphobia, generalized anxiety disorder, and social phobia. Moreover, parent-reported emotional problems (anxiety, shyness, and withdrawal) in 8-year old children have been specifically associated with a 2.6-fold risk for an anxiety disorder in early adulthood (39). Therefore, the association findings in the Swedish sample provide further support for involvement of the NPS-NPSR1 system in regulation of human anxiety. We further provide preliminary evidence that two of the five SNPs showing most significant evidence for association might affect transcription factor binding, although the nuclear proteins involved in allele-specific binding remain to be identified. The binding may still involve the tested SOX-5 and RUNX2 but depend on the formation of a protein complex with multiple participating factors. We also investigated whether molecular stress responses are altered in Npsr1⫺/⫺ mice on a gene expression level by examining a set of anxiety and stress-related genes, including several immediate early genes previously implicated as possible downstream target genes of NPS-mediated signaling (40). We found that stress increased cortical expression of Nps in WT mice but not in Npsr1⫺/⫺ mice, although the analysis did not reach experiment-wide significance. This finding suggests a feedback loop between NPSR1 receptor signaling and the expression of its anxiolytic ligand. Furthermore, experiment-wide significant differential stress responses were detected regarding induction of two anxiety-related genes, Ntf3 (downregulated in Npsr1⫺/⫺) and Il1b (upregulated in Npsr1⫺/⫺). Stress-related upregulation of the neurotrophic factor neurotrophin 3 was previously observed and may reflect a compensatory attempt to maintain neuroplasticity and brain function, avoiding behavioral maladaptations (41). Interestingly, in light of our human findings, mice overexpressing the neurotrophin 3 receptor tropomyosin receptor kinase C exhibit enhanced panic reaction (42), and dysregulation of this system may be involved in sensitization to chronic stress and onset of anxiety- and mood-related disorders (43). The role of proinflammatory cytokines such as interleukin 1 beta in stress is well established, and they may induce adaptive behavior allowing healing in response to trauma but induce anxiety or depression when dysregulated (44). Moreover, PD patients have elevated serum interleukin 1 beta levels (45), and interleukin 1 beta signaling may directly enhance fear memory (44). Our results should be interpreted in the context of some limitations. Although both the Finnish and Spanish samples revealed some associations in NPS, different alleles were implicated. Similar allelic discrepancies were detected for other asthma susceptibility genes (46) and may be attributable to interaction between the locus and other risk factors or population variation in interlocus correlation (47). Discrepancies might be due to differences in recruitment of both cases (population-based vs. clinical) and control subjects (psychiatrically screened, matched, population-based control subjects vs. nonpsychiatrically screened blood donor control subjects), differences in diagnostic interview and exclusion criteria (e.g., exclusion of patients with mood disorder comorbidity in the Spanish sample), and population genetic differences. Even though the NPS-NPSR1 system is strongly linked to regulation of rodent anxiety-related behavior, the effect sizes of predisposing

J. Donner et al.

BIOL PSYCHIATRY 2010;68:474 – 483 481

Figure 4. The effect of immobilization stress on brain expression levels of anxiety- and stress-related genes in wild-type and Npsr1 deficient mice (A-G). Results are shown for untreated control mice, mice killed immediately after immobilization stress, and for stressed mice killed after a 1 hour recovery period, with four to six mice per treatment group. Significant pairwise comparisons of genotype groups in unpaired t tests are indicated as follows: p ⬍ .05 (*) and p ⬍ .01 (**). Units are relative, and error bars indicate ⫹ 1 SD. Npsr1⫺/⫺, Npsr1 deficient mice.

genetic variants for human anxiety disorders may be small and the variants differ between populations. Finally, as we covered the investigated genes extensively by analyzing a large number of SNPs in three samples, we acknowledge that our genetic results should be viewed in light of issues related to multiple testing. Therefore, replication is needed to confirm the role of NPSR1 and NPS in predisposition to human anxiety-related traits. Recent evidence supports involvement of NPS-NPSR1 in regulation of fear and anxiety-related behavior. Specific neural circuitry involved in NPS action in the amygdala was identified (48,49). Moreover, injecting NPS into mouse basolateral amygdala has not only an acute anxiolytic effect but also results in accelerated extinction of auditory conditioned fear responses (48), whereas injecting NPS into mouse endopiriform nucleus causes reduction of contextually conditioned fear responses (49). Furthermore, Npsr1 deficient mice display signs of increased anxiety-like behavior (50). Based on these findings, our results, and the similarity of panic attacks and the physiological and behavioral consequences of conditioned fear responses (51), it is tempting to speculate that dysfunction of the NPS-NPSR1 system in the amygdala might play a role in conditioned responses related to PD. To summarize, our genetic evidence from three different populations suggests a role for the NPS-NPSR1 system in predisposition to PD in adults and to tendency for anxiety/depression in children. We further provide preliminary evidence suggesting functional relevance for some of the SNPs showing the strongest predisposing effect in PD. Additionally, we demonstrate altered molecular stress responses in Npsr1 deficient mice. Therefore, further studies of this

recently characterized neuropeptide system in the context of human anxiety disorders are warranted. This work was supported by The Academy of Finland (to RH and JK); The Academy of Finland NEURO Research program and academy research fellow funding (to IH); Sigrid Jusélius Foundation (to JK and IH); Yrjö Jahnsson Foundation (to IH); Yrjö and Tuulikki Ilvonen Foundation (to IH); Biocentrum Helsinki Foundation (to IH); L’Oréal Finland and United Nations Educational, Cultural, and Scientific Organization women in science fellowship (to IH); University of Helsinki (to IH); Nylands Nation vid Helsingfors universitet (to JD); H. Lundbeck A/S (to JD); Finnish Foundation for Psychiatric Research (to JD); The Helsinki Biomedical Graduate School (to JD); Fondo de Investigaciones Sanitarias de la Seguridad Social (P1052565 to RM-S; PI040632, PI040619, and CIBER-CB06/02/0058 to XE); Spanish Ministry of Education and Science (to the Spanish National Genotyping Center, and SAF2005-01005, SAF2007-60827, GEN2003-20651-C06-03 to XE); The Instituto Carlos III (GO3/184 to XE and RM-S); The Fundació la Marató-TV3 (014331 to RM-S); the Department of Universities, Research and Information Society (2005SGR00008; 2005SGR 00322 to XE); Genoma España (to the Spanish National Genotyping Center); Swedish Research Council (to Barn Allergi Miljö Stockholm Epidemiologi [BAMSE] study and JK); Stockholm County Council (to BAMSE study); The Swedish Asthma and Allergy Association (to BAMSE study); The Swedish Foundation for Health Care Science and Allergy Research (to BAMSE study); The Swedish Heart and Lung Foundation (to BAMSE study and JK); Chronic Inflammation - Diagnosis and Therapy/Verket för innovationssystem, Sweden (to JK); Centre for Allergy Research, Karolinska Institutet (to www.sobp.org/journal

482 BIOL PSYCHIATRY 2010;68:474 – 483 BAMSE study); The Bernard Osher Initiative for Research on Severe Asthma at Karolinska Institutet (to LJ); The Swedish Society for Medical Research (to CS); Professor Nanna Svartz Fund (to MD); and The Ruth and Richard Julin Foundation (to MD). We regret deeply the passing away of Leena Peltonen who participated actively in the planning of this study. We thank Sami Heistaro, Juuso Juhila, Laura Kananen, Olli Kiviruusu, Riitta Känkänen, Ville Pulkkinen, Päivi Saavalainen, Tessa Sipilä, and Tuula Vasankari for discussions and help. Dr. Peltonen is a member of the board of Orion Pharma. All other authors reported no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online. 1. Laitinen T, Polvi A, Rydman P, Vendelin J, Pulkkinen V, Salmikangas P, et al. (2004): Characterization of a common susceptibility locus for asthmarelated traits. Science 304:300 –304. 2. Melen E, Bruce S, Doekes G, Kabesch M, Laitinen T, Lauener R, et al. (2005): Haplotypes of G protein-coupled receptor 154 are associated with childhood allergy and asthma. Am J Respir Crit Care Med 171:1089 – 1095. 3. Kormann MS, Carr D, Klopp N, Illig T, Leupold W, Fritzsch C, et al. (2005): G-protein-coupled receptor polymorphisms are associated with asthma in a large German population. Am J Respir Crit Care Med 171: 1358 –1362. 4. Hersh CP, Raby BA, Soto-Quiros ME, Murphy AJ, Avila L, Lasky-Su J, et al. (2007): Comprehensive testing of positionally cloned asthma genes in two populations. Am J Respir Crit Care Med 176:849 – 857. 5. D’Amato M, Bruce S, Bresso F, Zucchelli M, Ezer S, Pulkkinen V, et al. (2007): Neuropeptide S receptor 1 gene polymorphism is associated with susceptibility to inflammatory bowel disease. Gastroenterology 133:808 – 817. 6. Gottlieb DJ, O’Connor GT, Wilk JB (2007): Genome-wide association of sleep and circadian phenotypes. BMC Med Genet 8(suppl 1):S9. 7. Xu YL, Gall CM, Jackson VR, Civelli O, Reinscheid RK (2007): Distribution of neuropeptide S receptor mRNA and neurochemical characteristics of neuropeptide S-expressing neurons in the rat brain. J Comp Neurol 500:84 –102. 8. Xu YL, Reinscheid RK, Huitron-Resendiz S, Clark SD, Wang Z, Lin SH, et al. (2004): Neuropeptide S: A neuropeptide promoting arousal and anxiolytic-like effects. Neuron 43:487– 497. 9. Rizzi A, Vergura R, Marzola G, Ruzza C, Guerrini R, Salvadori S, et al. (2008): Neuropeptide S is a stimulatory anxiolytic agent: A behavioural study in mice. Br J Pharmacol 154:471– 479. 10. Lage R, Dieguez C, Lopez M (2006): Caffeine treatment regulates neuropeptide S system expression in the rat brain. Neurosci Lett 410:47–51. 11. Lage R, Gonzalez CR, Dieguez C, Lopez M (2007): Nicotine treatment regulates neuropeptide S system expression in the rat brain. Neurotoxicology 28:1129 –1135. 12. Smith KL, Patterson M, Dhillo WS, Patel SR, Semjonous NM, Gardiner JV, et al. (2006): Neuropeptide S stimulates the hypothalamo-pituitary-adrenal axis and inhibits food intake. Endocrinology 147:3510 –3518. 13. Badia-Elder NE, Henderson AN, Bertholomey ML, Dodge NC, Stewart RB (2008): The effects of neuropeptide S on ethanol drinking and other related behaviors in alcohol-preferring and -nonpreferring rats. Alcohol Clin Exp Res 32:1380 –1387. 14. Goodwin RD (2003): Asthma and anxiety disorders. Adv Psychosom Med 24:51–71. 15. Goodwin RD, Jacobi F, Thefeld W (2003): Mental disorders and asthma in the community. Arch Gen Psychiatry 60:1125–1130. 16. Goodwin RD, Fergusson DM, Horwood LJ (2004): Asthma and depressive and anxiety disorders among young persons in the community. Psychol Med 34:1465–1474. 17. Roy-Byrne PP, Davidson KW, Kessler RC, Asmundson GJ, Goodwin RD, Kubzansky L, et al. (2008): Anxiety disorders and comorbid medical illness. Gen Hosp Psychiatry 30:208 –225. 18. Lavoie KL, Cartier A, Labrecque M, Bacon SL, Lemiere C, Malo JL, et al. (2005): Are psychiatric disorders associated with worse asthma control and quality of life in asthma patients? Respir Med 99:1249 –1257.

www.sobp.org/journal

J. Donner et al. 19. Richardson LP, Lozano P, Russo J, McCauley E, Bush T, Katon W (2006): Asthma symptom burden: Relationship to asthma severity and anxiety and depression symptoms. Pediatrics 118:1042–1051. 20. McCauley E, Katon W, Russo J, Richardson L, Lozano P (2007): Impact of anxiety and depression on functional impairment in adolescents with asthma. Gen Hosp Psychiatry 29:214 –222. 21. Okamura N, Hashimoto K, Iyo M, Shimizu E, Dempfle A, Friedel S, Reinscheid RK (2007): Gender-specific association of a functional coding polymorphism in the neuropeptide S receptor gene with panic disorder but not with schizophrenia or attention-deficit/hyperactivity disorder. Prog Neuropsychopharmacol Biol Psychiatry 31:1444 –1448. 22. Pirkola SP, Isometsä E, Suvisaari J, Aro H, Joukamaa M, Poikolainen K, et al. (2005): DSM-IV mood-, anxiety- and alcohol use disorders and their comorbidity in the Finnish general population. Soc Psychiatry Psychiatr Epidemiol 40:1–10. 23. Pirkola S, Isometsa E, Aro H, Kestila L, Hamalainen J, Veijola J, et al. (2005): Childhood adversities as risk factors for adult mental disorders: Results from the Health 2000 study. Soc Psychiatry Psychiatr Epidemiol 40:769 –777. 24. Wickman M, Kull I, Pershagen G, Nordvall SL (2002): The BAMSE project: Presentation of a prospective longitudinal birth cohort study. Pediatr Allergy Immunol 13(suppl 15):11–13. 25. The EuroQol Group (1990): EuroQol—a new facility for the measurement of health-related quality of life. Health Policy 16:199 –208. 26. Ostblom E, Lilja G, Pershagen G, van Hage M, Wickman M (2008): Phenotypes of food hypersensitivity and development of allergic diseases during the first 8 years of life. Clin Exp Allergy 38:1325–1332. 27. Donner J, Pirkola S, Silander K, Kananen L, Terwilliger JD, Lönnqvist J, et al. (2008): An association analysis of murine anxiety genes in humans implicates novel candidate genes for anxiety disorders. Biol Psychiatry 64:672– 680. 28. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. (2002): The structure of haplotype blocks in the human genome. Science 296:2225–2229. 29. Dudbridge F (2008): Likelihood-based association analysis for nuclear families and unrelated subjects with missing genotype data. Hum Hered 66:87–98. 30. Tamura K, Dudley J, Nei M, Kumar S (2007): MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596 –1599. 31. Schreiber E, Matthias P, Muller MM, Schaffner W (1989): Rapid detection of octamer binding proteins with “mini-extracts,” prepared from a small number of cells. Nucleic Acids Res 17:6419. 32. Reinscheid RK, Xu YL, Okamura N, Zeng J, Chung S, Pai R, et al. (2005): Pharmacological characterization of human and murine neuropeptide S receptor variants. J Pharmacol Exp Ther 315:1338 –1345. 33. Kwan KY, Lam MM, Krsnik Z, Kawasawa YI, Lefebvre V, Sestan N (2008): SOX5 postmitotically regulates migration, postmigratory differentiation, and projections of subplate and deep-layer neocortical neurons. Proc Natl Acad Sci U S A 105:16021–16026. 34. Benes FM, Lim B, Matzilevich D, Walsh JP, Subburaju S, Minns M (2007): Regulation of the GABA cell phenotype in hippocampus of schizophrenics and bipolars. Proc Natl Acad Sci U S A 104:10164 –10169. 35. Katon W, Lozano P, Russo J, McCauley E, Richardson L, Bush T (2007): The prevalence of DSM-IV anxiety and depressive disorders in youth with asthma compared with controls. J Adolesc Health 41:455– 463. 36. Katon WJ, Richardson L, Lozano P, McCauley E (2004): The relationship of asthma and anxiety disorders. Psychosom Med 66:349 –355. 37. Hasler G, Gergen PJ, Kleinbaum DG, Ajdacic V, Gamma A, Eich D, et al. (2005): Asthma and panic in young adults: A 20-year prospective community study. Am J Respir Crit Care Med 171:1224 –1230. 38. Zaubler TS, Katon W (1996): Panic disorder and medical comorbidity: A review of the medical and psychiatric literature. Bull Menninger Clin 60(2 suppl A):A12–A38. 39. Sourander A, Jensen P, Davies M, Niemela S, Elonheimo H, Ristkari T, et al. (2007): Who is at greatest risk of adverse long-term outcomes? The Finnish From a Boy to a Man study. J Am Acad Child Adolesc Psychiatry 46:1148–1161. 40. Vendelin J, Bruce S, Holopainen P, Pulkkinen V, Rytila P, Pirskanen A, et al. (2006): Downstream target genes of the neuropeptide S-NPSR1 pathway. Hum Mol Genet 15:2923–2935. 41. Faure J, Uys JD, Marais L, Stein DJ, Daniels WM (2006): Early maternal separation followed by later stressors leads to dysregulation of the HPA-axis and increases in hippocampal NGF and NT-3 levels in a rat model. Metab Brain Dis 21:181–188.

J. Donner et al. 42. Dierssen M, Gratacos M, Sahun I, Martin M, Gallego X, Amador-Arjona A, et al. (2006): Transgenic mice overexpressing the full-length neurotrophin receptor TrkC exhibit increased catecholaminergic neuron density in specific brain areas and increased anxiety-like behavior and panic reaction. Neurobiol Dis 24:403– 418. 43. Amador-Arjona A, Delgado-Morales R, Belda X, Gagliano H, Gallego X, Keck ME, et al. (2010): Susceptibility to stress in transgenic mice overexpressing TrkC, a model of panic disorder. J Psychiatr Res 44:157–167. 44. Koo JW, Duman RS (2009): Interleukin-1 receptor null mutant mice show decreased anxiety-like behavior and enhanced fear memory. Neurosci Lett 456:39 – 43. 45. Brambilla F, Bellodi L, Perna G, Bertani A, Panerai A, Sacerdote P (1994): Plasma interleukin-1beta concentrations in panic disorder. Psychiatry Res 54:135–142. 46. Kere J, Laitinen T (2004): Positionally cloned susceptibility genes in allergy and asthma. Curr Opin Immunol 16:689 – 694.

BIOL PSYCHIATRY 2010;68:474 – 483 483 47. Lin PI, Vance JM, Pericak-Vance MA, Martin ER (2007): No gene is an island: The flip-flop phenomenon. Am J Hum Genet 80:531–538. 48. Jungling K, Seidenbecher T, Sosulina L, Lesting J, Sangha S, Clark SD, et al. (2008): Neuropeptide S-mediated control of fear expression and extinction: Role of intercalated GABAergic neurons in the amygdala. Neuron 59:298 –310. 49. Meis S, Bergado-Acosta JR, Yanagawa Y, Obata K, Stork O, Munsch T (2008): Identification of a neuropeptide S responsive circuitry shaping amygdala activity via the endopiriform nucleus. PLoS ONE 3:e2695. 50. Duangdao DM, Clark SD, Okamura N, Reinscheid RK (2009): Behavioral phenotyping of neuropeptide S receptor knockout mice. Behav Brain Res 205:1–9. 51. Gorman JM, Kent JM, Sullivan GM, Coplan JD (2000): Neuroanatomical hypothesis of panic disorder, revised. Am J Psychiatry 157:493– 505.

www.sobp.org/journal