American Journal of Medical Genetics Part B (Neuropsychiatric Genetics) 137B:45 –50 (2005) Association Study of CREB1 and Childhood-Onset Mood Disorders I. Burcescu,1 K. Wigg,1 N. King,2 Á. Vetró,3 E. Kiss,3 L. Katay,3 J.L. Kennedy,2 M. Kovacs,4 C.L. Barr,1,5* and The International Consortium for Childhood-Onset Mood Disorders 1 Toronto Western Research Institute, University Health Network, Toronto, Ontario, Canada Neurogenetics Section, Centre for Addiction and Mental Health, Toronto, Ontario, Canada 3 Department for Child and Adolescent Psychiatry, Szeged University, Szeged, Hungary 4 University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 5 The Hospital for Sick Children, Toronto, Ontario, Canada 2 Several lines of evidence suggest that the cellular pathways involved in synaptic plasticity contribute to the risk of depression. These findings include the evidence that chronic antidepressant treatment upregulates the cAMP signal transduction cascade resulting in increased expression and function of the cAMP responsive element binding protein (CREB), a transcription factor that increases the expression of key growth factors involved in synaptogenesis and neurogenesis. Recently, linkage to CREB1 was reported for early-onset depression in families recruited from the Pittsburgh area. This finding was significant only in female sibling pairs from those families. Two specific DNA variants, 656G/A and a C insertion/deletion in intron 8, were identified in CREB1 that co-segregated with depression in two of the families. We sought to investigate the relationship of CREB1 to childhood-onset mood disorders (COMD) using a sample of 195 nuclear families (225 affected children) collected in Hungary. We genotyped the two CREB1 DNA variants previously identified as linked to depression as well as three additional polymorphisms spanning the gene. In addition, we genotyped the 656G/A DNA change and the intron 8 polymorphism in a Members of the International Consortium for Childhood-Onset Mood Disorders: István Benák, Viola Kothencné Osváth, Edit Dombovári, Emı́lia Kaczvinszky and Andor Kanka, Szeged University Medical Faculty, Department of Child and Adolescent Psychiatry, Szeged. Júlia Gádoros, Ildikó Baji, Zsuzsa Tamás, Márta Besnyo, and Judit Székely, Vadaskert Hospital, Budapest. László Mayer, Margit Hospital, Csorna. Heads of the Units and Departments where the patients were collected: Rózsa Hasuly, MD, Szent Rókus Hospital, Outpatient Unit of Child Psychiatry, Budapest; Ilona Riegler, MD, Márta Fohn, MD, Heim Pál Hospital Outpatient Unit of Child Psychiatry, Budapest; Katalin Benko, MD, Outpatient Unit of Child Psychiatry; Szeged. Mária Mojzes, MD, Outpatient Unit of Child Psychiatry, Baja; Róza Oláh, MD, Kenézy Gyula Hospital, Department of Child and Adolescent Psychiatry, Debrecen; Mária Károlyfalvi, MD, Gyöngyi Farkas, MD, Zsuzsa Bánk, MD, Outpatient Unit of Child Psychiatry, Debrecen. Ferenc Dicso, MD, Dénes Kövendy, MD, Jósa András Hospital Outpatient Unit of Child Psychiatry, Nyı́regyháza; Mária Gyurcsó, MD, Petz Aladár Hospital, Outpatient Unit of Child Psychiatry, Gyor; Zsuzsanna Fekete, MD, Mariann Vados, MD, Szent György Hospital, Outpatient Unit of Child Psychiatry, Székesfehérvár; Zsuzsa Takács, MD, Szekszárd Hospital, Outpatient Unit of Child Psychiatry, Szekszárd; Eszter Gyenge, MD, Mária Palaczky, MD, Ágnes Horváth, MD, Outpatient Department of Child Psychiatry, Pécs; Ilona Mógor, MD. Péter ß 2005 Wiley-Liss, Inc. sample of 112 probands with mood disorders collected in the Pittsburgh area and matched controls, and examined the distribution of alleles. The 656A allele was not observed in our samples and there was no evidence for association of the intron 8 polymorphism in either the sample from Pittsburgh (x2 ¼ 0.061, 1 d.f., P ¼ 0.803) or Hungary (x2 ¼ 0.040, 1 d.f., P ¼ 0.842). We found no evidence for an association with the other three polymorphisms or with the haplotypes of these markers. Further, we found no sex-specific relationship. Our results, therefore, do not support the previous evidence for this gene as a major factor contributing to depression. ß 2005 Wiley-Liss, Inc. KEY WORDS: CREB1; depression; mood disorders; gene genetics; INTRODUCTION Mood disorders that onset in childhood have a significant impact on subsequent adjustment, and present challenges to conventional treatments. In young patients, an episode of major depressive disorder (MDD) has a recurrence rate of approximately 70%, indicating that many young patients Steiner, MD, Csolnoky Ferenc Hospital, Outpatient Unit of Child Psychiatry, Veszprém. Eniko Juhász Szolnok Hospital Outpatient Unit of Child Psychiatry, Szolnok; Mária Révhelyi, MD, Petz Aladár Hospital Department of Child and Adolescent Psychiatry, Gyor; Éva Gyulai, MD, Baja Hospital Outpatient Unit of Child Psychiatry, Baja; Katalin Bense, MD, Edina Farkas, MD, Kecskemét Hospital Outpatient Unit of Child Psychiatry, Kecskemét; Sörfozo Zsuzsanna, MD, Kaposi Mór Hospital, Outpatient Unit of Child Psychiatry, Kaposvár. Interviewers: Judit Tömöri, Tünde Horváth M., Eszter Szamosi. Regional research office members: Cserép Melinda, Edit Sitkei, Noémi Takács. Central research office members: Zoltán Széll, László Tóth-Soma Ph.D, Márta Dobó, and Éva Lehoczky. Grant sponsor: National Institute of Mental Health Program Project; Grant number: MH 56193; Grant sponsor: National Alliance for Research on Schizophrenia and Depression. *Correspondence to: C.L. Barr, 399 Bathurst St. Toronto Western Research Institute, University Health Network, Toronto, Ontario, Canada. E-mail: CBarr@uhnres.utoronto.ca Received 22 October 2004; Accepted 21 April 2005 DOI 10.1002/ajmg.b.30201 46 Burcescu et al. remain at significant risk for depressive disorders [Kovacs and Devlin, 1998]. Mood disorders and depressive symptoms are estimated to have a heritable component (31–42%) [Sullivan et al., 2000] and probands with early-onset, including childhood-onset, carry a higher familial risk [Orvaschel, 1990; Kovacs et al., 1997], and have higher heritiability estimates (70–82%) [Thapar and McGuffin, 1994]. Extensive research has established that population base rates of depression are higher in women than in men [Weissman and Klerman, 1977; Nolen-Hoeksema, 1987; Kessler et al., 1993]. However, prior to around 14 years of age, depression appears to affect boys and girls equally [Wade et al., 2002]. Further, in psychiatrically referred depressed children, there is no evidence for significant sex differences in clinical presentation in terms of age of onset, severity, risk of recurrence, episode length, or comorbid disorders [Kovacs, 2001]. The sex differences in depression become apparent following puberty: almost twice as many adolescent girls as boys present with MDD [Lewinsohn et al., 1993; Angold et al., 1998]. Some twin studies have suggested a greater heritability of depression in women than in men [Bierut et al., 1999; Kendler et al., 2001], with evidence that women and men, for the most part, share common genetic influences for depression, but with some gender specific genes contributing to the risk [Kendler et al., 2001]. Evidence has been reported for linkage of the 2q33–35 chromosomal region to mood disorders among women, but not men, in families with recurrent (2 episodes) early onset (onset age 25) major depressive disorders (MDD) [Zubenko et al., 2003b]. A particularly strong candidate for depression located in this region, the gene for cAMP responsive element binding protein, CREB1, was investigated in an additional study that implicated CREB1 as a potential sex-limited susceptibility gene for MDD [Zubenko et al., 2003a]. In that study, two sequence variants in CREB1 were identified, one located in the promoter region (656A/G) and one in intron 8 (cytosine insertion/deletion located 16 bp 50 to exon 9) that yielded significant linkage scores (P ¼ 0.002 and 0.01, respectively) for mood disorders, or their absence, among the women in 2 of a total of 81 families in the study. The CREB1 gene encodes a 43 kDa, 341 amino acids nuclear phosphoprotein, cAMP-responsive element (CRE)binding protein (CREB), which is a member of the basic leucine zipper family of DNA binding proteins [Mayr and Montminy, 2001]. Alternate splicing of the gene’s exons (spanning 68.9 kb) results in two transcript variants encoding different isoforms with distinct properties in different tissues and during development [Walker et al., 1996; Rudolph et al., 1998]. Phosphorylation at Ser133 by several protein kinases in response to, amongst others, cAMP, activates CREB. Once activated, the protein binds as a homodimer to the CRE site (an octameric palindrome) of a number of gene promoters, thereby inducing transcription. Several reports on the synergistic interaction of CREB with nuclear estrogen receptors have suggested mechanisms by which CREB facilitates sex-specific patterns of gene expression [Lazennec et al., 2001; McEwen, 2001]. This gene is a strong candidate for depression based on the hypothesis that cellular pathways involved in synaptic plasticity contribute to the risk of depression [Duman et al., 1999; Vaidya and Duman, 2001]. Recurrent major depression has been associated with significant hippocampal damage, and chronic (but not acute) antidepressant treatment upregulates the cAMP signal transduction cascade involved in synaptic plasticity. Upregulation of cAMP results in increased expression and function of CREB1, increasing the expression of target genes including brain-derived neurotropic factor (BDNF), leading to neuronal sprouting and neurogenesis. Chronic stress has been demonstrated to decrease hippocampal neurogenesis and to induce atrophy and death of hippocampal and cortical neurons. Genetic variability in the components of this system may result in an inherent decrease in the ability to make adaptive cognitive changes to stressful events and decreased resilience to negative thoughts. Directly relevant is our previous genetic studies of BDNF that provides evidence for this gene as a susceptibility gene in COMD in the samples used in this study, specifically supporting the role of the cAMP pathway in the genetic risk for COMD [Strauss et al., 2004]. These converging lines of evidence and the study by Zubenko et al. [2003a] prompted our group to investigate whether the CREB1 gene was associated with COMD, and whether this association was related to sex. To test for such an association, we genotyped the two CREB1 DNA variants described above [Zubenko et al., 2003a] as well as three additional polymorphisms in a sample of 195 small nuclear families with 225 affected children that had been collected in Hungary. The transmission disequilibrium test (TDT) was then used to test for biased transmission of alleles and of haplotypes from heterozygous parents to their affected offspring. Using an independent sample of 112 probands with early-onset mood disorders collected in Pittsburgh and matched controls, we also genotyped the two DNA variants identified in the Zubenko et al. [2003a] study and then examined the distribution of the alleles. MATERIALS AND METHODS Subjects The subjects enrolled in this study were part of a multidisciplinary Program Project to study risk factors in childhoodonset mood disorders (COMD) and have been previously described in detail [Adams et al., 2005]. We used two samples for this study: (i) family-based association study: The sample consisted of 195 families with 225 affected children recruited from 21 mental health facilities across Hungary. The probands were diagnosed using the Interview Schedule for Children and Adolescents-Diagnostic Version (ISCA-D), an extension and modification of the ISCA [Sherrill and Kovacs, 2000]. The probands met DSM-IV criteria for a mood disorder (unipolar major depression or bipolar disorder), with the onset of the first episode by age 14. Probands were interviewed twice, approximately 1 month apart. Consensus diagnosis of two independent child psychiatrists was used as final diagnosis. (ii) Case-control association study: An extensive description of the subjects enrolled in this study was published previously [Adams et al., 2005]. Briefly, the 112 proband sample consisted of young adults from the Pittsburgh, PA area that met DSM-III or DSMIV criteria for depression (major depressive and/or dysthymic disorder) by age 14 or bipolar spectrum disorder (bipolar I, II or cyclothymic disorder) by age 17. Psychiatric assessments were conducted and verified by experienced and trained clinical evaluators and independent best-estimate psychiatrists. Each proband was individually matched to a control with no history of mental illness, of the same sex, and of the same ethnic background. The sample consisted of 69 (62%) females and 43 (38%) males. For both probands and controls, ethnicity matching was based on detailed self-reported description of ancestry. Proband and control groups consist of 83% EuropeanNorth American, 16% African-North American, and 1% mixed ethnicity. Previously, genomic control testing [Devlin and Roeder, 1999] was performed with 17 biallelic markers distributed randomly across the genome. The distribution of the w2 values for the COMD probands versus controls showed no deviation from that expected by chance indicating that the controls were of similar ethnic background as the COMD probands and no correction for stratification was required for this sample [Strauss et al., 2004]. 0.602 0.273 0.842 0.040 0.442 0.591 0.478 b 30 UTR 208432020 rs6785 Creb1-9224 Intron 8 208425646 Creb1-int8del Intron 4 208395853 rs2551920 Creb1-9089 Intron 1 (untranslated region) 208376555 rs2709357 Creb1-7093 a 71 63 73 64 12 13 69 63 1.000 0.000 0.763 0.237 0.762 0.238 0.967 0.033 0.768 0.232 1 2 1 2 1 2 1 2 1 2 G A A G C G Ins C Del C G A Promoter Based on genomic contig NT005403, S53722 (and using NCBI information regarding location on chromosome 2 for the SNPs with reference sequence number). Intron structure based on NCBI Build 33, the April 2003 UCSC human reference sequence. 63 71 64 73 13 12 63 69 Non-transmissions Transmissions Allele frequency Allele DNA variant Locationb Nucleotide numbera Reference sequence number 208358054 For this study we genotyped five markers (Table I), two of which had been reported previously [Zubenko et al., 2003a]. Creb1-pro RESULTS Polymorphism Statistical Analysis The degree of linkage disequilibrium (LD) between marker alleles in this study was evaluated using the ldmax program (available in the GOLD software package [Abecasis and Cookson, 2000] at http://www.sph.umich.edu/csg/abecasis/ GOLD). LD was expressed as the coefficient of LD D0 [Devlin and Risch, 1995] and the square of the standardized measure, r2 [Hill and Weir, 1994]. Hardy–Weinberg equilibrium was tested using the program PedStats (http://www.sph.umich. edu/csg/abecasis/PedStats/index.html). All markers were found to be in Hardy–Weinberg equilibrium. The extended TDT (ETDT) program [Sham and Curtis, 1995] was used to test for biased transmission of alleles in the family-based association study (Hungarian samples). The transmission of haplotypes was analyzed using the TRANSMIT program [Clayton, 1999] with the robust estimator option, and where only haplotypes with frequencies greater than 0.10 were used in the analysis. For the case-control association study, chi-square analyses were performed to compare the distribution of alleles. TABLE I. CREB1 Polymorphisms, Allele Frequencies in the Hungarian Sample Parental Chromosomes, and TDT Results Blood, or in some cases, buccal samples, were collected from subjects and genomic DNA was extracted using a standard high salt method or NaOH and Proteinase K (QIAmp DNA Mini Kit, Qiagen, Inc., Mississauga, ON, Canada), respectively. Genotypes for the two Zubenko et al. [2003a] DNA variants, a G to A change in the promoter region (marker CREB1-pro) and a C insertion/deletion in the CREB1 intron 8 (marker CREB1-int8del), were determined by PCR amplification of 60–100 ng DNA, followed by restriction enzyme digestion of the PCR products, as described previously [Zubenko et al., 2003a]. The promoter region polymorphism was genotyped with the following primers: F: GTA GGA TGG GGC ATA TTT CCA G and R: CGG GTT TTC CTT TCC GAG ACT CC amplified at 608C and digested with MspI. The intron 8 insertion/deletion polymorphism was initially genotyped with restriction enzymes as previously described but was then converted into a TaqMan1 50 nuclease assay for allelic discrimination assay (Applied Biosystems, Foster City, CA) for more efficient genotyping. The primers for amplification (608C) were F: TAT GTG GAA ATC ATT TGC AT and R: TTT TCA AGC ACT GCC ACT CTG T with probes VIC: TCC CGT CTC TTT TG and FAM: CCG TCC TCT TTT G. The remaining three single nucleotide polymorphisms (SNPs; CREB1-7093, CREB1-9089, CREB1-9224) were genotyped using the TaqMan1 50 nuclease assay with primers and probes available commercially (Applied Biosystems). The PCR primers and allelic discrimination probes were designed (CREB1-int8del) and/or obtained from ABI (assays-ondemand1) based on published reference sequence numbers. Five microliters PCR reactions contained 30 ng of genomic DNA, 5 mmol of TaqMan1 Universal PCR Master Mix (Applied Biosystems), and 0.125 ml of allelic discrimination mix, which contained 36 mM of each primer and 8 mM of each probe. The thermal cycling conditions for the Assays-on-Demand1 obtained from Applied Biosystems were 958C for 10 min, then 50 cycles of 928C for 15 sec and 1 min at 608C. For CREB1int8del, an initial 2 min step of 508C was added. Negative controls were included with each 96-well plate. Plates were read on the ABI 7900-HT Sequence Detection System using the allelic discrimination end-point analysis mode of the software package version 2.0 (Applied Biosystems). w2 Genotyping 0.489 P-value Association of CREB1 and Childhood-Onset Mood Disorders 47 48 Burcescu et al. The first of these, a G to A transition is predicted to eliminate a consensus-binding motif for AP-2, a transcriptional activator, and a potential CpG methylation site in the CREB1 promoter (Creb1-pro). The second is a C insertion/deletion in intron 8 of the gene (Creb1-int8del). We selected the remaining polymorphisms from the NCBI SNP database (http://www. ncbi.nlm.nih.gov/SNP/index.html), located either in introns, or in the 30 untranslated region (30 UTR) of the gene, and thus we were able to genotype polymorphic markers across the gene from intron 1 to the 30 UTR. The location of the polymorphisms and allele frequencies in the Hungarian parental chromosomes are listed in Table I. We did not observe the A allele of the promoter DNA variant in any of 376 probands (n ¼ 124) affected siblings (n ¼ 21) or parents (n ¼ 231) screened in the Hungarian family sample, nor in the 112 COMD probands collected in Pittsburgh. To test for association of the CREB1 gene to COMD, we used the TDT to analyze the inheritance of four polymorphisms, both individually and as haplotypes, in the families from Hungary. We found no evidence of biased transmission of individual marker alleles (Table I). For the intron 8 C insertion/ deletion polymorphism, the number of informative transmissions was low (25) therefore not to be considered definitive; however, there was no evidence of even a trend for transmission of either allele. We also genotyped this marker in the sample of COMD probands collected in Pittsburgh as this sample would be of similar ethnic background to the families previously identified with evidence for linkage of this marker. The frequency of the intron 8DC allele in the probands (0.040) collected in Pittsburgh did not differ significantly from the frequency observed in the control sample (0.036) and therefore we find no evidence for association to COMD (w2 ¼ 0.06, 1 df, Pvalue ¼ 0.803). Further, the frequency did not differ significantly from the frequency observed in the Hungarian parental (0.033) or probands’ chromosomes (0.039). We also examined the transmission of haplotypes in the Hungarian sample, since haplotypes may be more informative than single markers alone. The four-marker haplotype frequencies were determined from the parental chromosomes, and as Table II shows, only two haplotypes were found at haplotypes frequencies greater than 10% and therefore at a high enough frequency to be used meaningfully in TDT analysis. We found no evidence of biased transmission of haplotypes in our nuclear family sample. Since Zubenko et al. [2003a] identified sex differences in the risk conferred by the intron 8DC insertion/deletion and promoter variants, we analyzed our female and male probands separately in an attempt to determine whether this would indicate biased allele transmission. As Table III shows, even when sex of the subjects was considered separately, we found no biased mode of transmission in our Hungarian family sample for any of the markers. The previous report by Zubenko and colleagues of linkage for the intron 8 insertion/deletion polymorphism indicated that the DC allele was a protective factor in women but not men. It would then be predicted that there would be evidence for biased non-transmission of this allele in the sample of female subjects, but this was not supported in our sample. There was also no significant difference in the allele frequencies for the intron 8 marker between males (0.046) and females (0.036) in the Pittsburgh probands (w2 ¼ 0.146, 1 d.f., P ¼ 0.703). D0 [Devlin and Risch, 1995] and r2 [Hill and Weir, 1994] are commonly used measures for describing the degree of LD between alleles at two markers. Strong LD (>0.90 between all markers) as measured by D0 is evident across the gene in the Hungarian sample. r2 can be used as an indication of the power to detect an association with r2 values greater than 0.80 considered a stringent threshold, able to detect 80% of the existing haplotypes [Carlson et al., 2004]. With the exception of the relatively uninformative intron 8DC insertion/deletion marker, the remaining three markers have r2 values greater than 0.93 indicating that we would have high power to detect an association using these three markers if this gene is a major susceptibility factor contributing to depression as indicated from the previous linkage studies of this region. DISCUSSION In this study, we failed to find support for the CREB1 gene recently reported as a candidate gene for susceptibility to early-onset mood disorders [Zubenko et al., 2003a]. The evidence that this gene is related to mood disorders was based on an initial linkage study that yielded a peak multipoint LOD score of 3.77 at D2S2208 for female relative pairs affected by recurrent, early-onset, unipolar mood disorders [Zubenko et al., 2002]. Subsequent studies [Zubenko et al., 2003a] identified two DNA changes in the CREB1 gene that cosegregated with mood disorders (or their absence). The first of these variants was a possible functional G to A transition at position 656 in the promoter region (Creb1-pro); heterozygosity for the CREB1 promoter variant was linked with mood disorders among women (w2 ¼ 9.76, df ¼ 1, P ¼ 0.002) in one early-onset MDD family (family A) that previously revealed a positive LOD score with the 2q33–35 locus. This change was also observed to segregate in a nuclear family with four affected children, two males and two females (family B). The second DNA change was a C deletion in intron 8 (Creb1-int8del). Unlike the promoter variant, heterozygosity for the intron 8 TABLE II. CREB1 Common Haplotype Frequencies and TRANSMIT Analyses CREB1 marker alleles Haplotype number 1 2 3 4 5 6 7 8 2 Transmission 7093 9089 Int8Del 9224 Haplotype frequency 1 2 1 1 1 1 1 2 1 1 2 1 2 1 2 2 1 1 1 2 2 1 1 1 1 1 1 1 1 2 2 2 0.72 0.01 0.00 0.03 0.00 0.00 0.00 0.23 2 Obsa Expb w2 P-value (two-sided) 324.00 317.37 1.16 0.28 100.00 103.47 0.37 0.55 Global w ¼ 7.350, 7 d.f., P ¼ 0.3934; common haplotypes (those with frequencies >10%) w ¼ 1.778, 2d.f., P ¼ 0.411. Only haplotypes with frequencies >0.01 are listed. Others were found with frequencies <0.01. a Test statistic representing the observed number of transmissions. b Expected value of the test statistic under the null hypothesis of no linkage or association. 0.502 0.450 1.000 0.000 0.502 0.450 0.317 1.000 36 45 37 43 7 7 37 43 1.000 0.000 0.763 0.091 0.158 0.691 45 36 43 37 7 7 43 37 0.890 0.019 27 26 27 30 6 5 26 26 26 27 30 27 5 6 26 26 One degree of freedom. a Creb1-9224 Creb1-int8del Creb1-9089 1 2 1 2 1 2 1 2 Creb1-7093 Non-transmissions Polymorphism Allele Transmissions Non-transmissions w2 P-valuea Transmissions Female subjects Male subjects TABLE III. TDT Results of CREB1 Markers in the Sample of Hungarian Families According to Sex w2 P-value Association of CREB1 and Childhood-Onset Mood Disorders 49 polymorphism yielded a protective effect against mood disorders in women but not men in family A (P ¼ 0.01). In our sample of Hungarian families, the promoter polymorphism was not observed and we did not find significant evidence for biased transmission of the alleles of the intron 8 polymorphism, nor any significant sex-biased results. Because ethnic variation could result in different alleles contributing to the phenotype in different populations, we genotyped the promoter DNA variant and the DC DNA changes in a sample of probands drawn from the Pittsburgh population. We did not observe the promoter DNA change in the Pittsburgh sample, therefore, although possibly functional due to the predicted change in the consensus binding site for the transcription factor AP-2, this DNA variant is too rare to be a major susceptibility allele contributing to the COMD phenotype. We also did not observe any difference in allele frequencies for the intron 8DC allele in the Hungarian and Pittsburgh samples and there was no evidence for an association. The Zubenko and colleagues study found that these two CREB1 sequence variants contributed to the susceptibility of mood disorders in only 2 of 31 early-onset MDD families (or 2.5% of a total of 81 recurrent early-onset unipolar mood disorders families in their collection). Further analysis indicated that the 79 remaining families generated a statistically significant LOD score of 4.24 that occurred in the 451 kb region between adjacent markers D2S2321 and D2S2208 that contained CREB1 indicating that the remaining families are linked to this region. This allows for the possibility that as yet unidentified DNA variants in CREB1 contribute to the disorder. Results in our sample did not support this as a possible explanation as we did not find any evidence for an association using markers spanning the gene. There is strong LD across this gene, therefore in our sample, we should be able to detect an association if this gene is a major genetic factor contributing to depression as suggested by the results of the linkage studies. Different factors influence power in linkage versus association studies thus it is difficult to compare power between these different methods of analysis. The high LOD scores for this region indicate that the locus in this region is a gene of major effect contributing to depression. Therefore, using an association approach which is much more powerful, and multiple markers spanning the gene with high LD across the gene, our sample of 195 nuclear families should have sufficient power to detect an association. We can estimate power for the DC allele specifically, assuming complete LD (this allele is the causative change) and allele frequency of 0.03. Based on these assumptions, the Hungarian sample will have 80% power at the alpha 0.05 level to detect an association if this allele carries a relative risk of 2.6 or greater (Genetic Power Calculator, http:// statgen.iop.kcl.ac.uk/gpc/qtlassoc.html). Linkage analysis based on sharing of alleles within families generally results in positive LOD scores over large chromosomal regions. Association studies, however, rely on LD and therefore association can generally only be detected over small chromosomal regions generally under 1 Mb. The genome scan results provided the highest evidence for support spanned a region of 451 kb with significant multipoint LOD scores spanning a much larger region of 15 cM. Based on the evidence supporting genes involved in synaptic plasticity in mood disorders, particularly CREB1 [Duman et al., 1999, 2000], this gene was by far the strongest candidate in the 2q33–35 region. However, linkage to another gene in this region co-segregating with the CREB1 DC and the promoter variant in the families previously linked to CREB1 cannot be ruled out on the basis of the linkage data, and other genes in the region should be considered. Lastly, differences in the definition of the phenotype between the previous study of CREB1 and ours may be a factor in the findings. For example, different mechanisms may 50 Burcescu et al. contribute to childhood-onset depression, defined here as onset before 14 compared to early-onset defined as onset prior to 25 years of age defined in the Zubenko and colleagues study. Further, the sample from Pittsburgh used here consists of both early-onset unipolar as well as bipolar cases whereas the previous study used only unipolar cases. Phenotypic, or clinical features, as contributing factors to the differential findings cannot be ruled out at this time. REFERENCES Abecasis GR, Cookson WO. 2000. GOLD—Graphical overview of linkage disequilibrium. Bioinformatics 16(2):182–183. Adams JH, Wigg KG, King N, Burcescu I, Vetro A, Kiss E, Baji I, George CJ, Kennedy JL, Kovacs M, Barr CL. 2005. Association study of neurotrophic tyrosine kinase receptor type 2 (NTRK2) and childhood-onset mood disorders. Am J Med Genet B Neuropsychiatr Genet 132(1):90–95. Angold A, Costello EJ, Worthman CM. 1998. Puberty and depression: The roles of age, pubertal status and pubertal timing. Psychol Med 28(1): 51–61. Bierut LJ, Heath AC, Bucholz KK, Dinwiddie SH, Madden PA, Statham DJ, Dunne MP, Martin NG. 1999. Major depressive disorder in a community-based twin sample: Are there different genetic and environmental contributions for men and women? Arch Gen Psychiatry 56(6): 557–563. Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, Nickerson DA. 2004. Selecting a maximally informative set of single-nucleotide polymorphisms for association analyses using linkage disequilibrium. Am J Hum Genet 74(1):106–120. Clayton D. 1999. A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission. Am J Hum Genet 65(4):1170– 1177. Devlin B, Risch N. 1995. A comparison of linkage disequilibrium measures for fine-scale mapping. Genomics 29(2):311–322. Devlin B, Roeder K. 1999. Genomic control for association studies. Biometrics 55(4):997–1004. Duman RS, Malberg J, Thome J. 1999. Neural plasticity to stress and antidepressant treatment. Biol Psychiatry 46(9):1181–1191. Duman RS, Malberg J, Nakagawa S, D’Sa C. 2000. Neuronal plasticity and survival in mood disorders. Biol Psychiatry 48(8):732–739. Hill WG, Weir BS. 1994. Maximum-likelihood estimation of gene location by linkage disequilibrium. Am J Hum Genet 54(4):705–714. Kendler KS, Gardner CO, Neale MC, Prescott CA. 2001. Genetic risk factors for major depression in men and women: Similar or different heritabilities and same or partly distinct genes? Psychol Med 31(4):605– 616. Lazennec G, Thomas JA, Katzenellenbogen BS. 2001. Involvement of cyclic AMP response element binding protein (CREB) and estrogen receptor phosphorylation in the synergistic activation of the estrogen receptor by estradiol and protein kinase activators. J Steroid Biochem Mol Biol 77(4–5):193–203. Lewinsohn PM, Hops H, Roberts RE, Seeley JR, Andrews JA. 1993. Adolescent psychopathology: I. Prevalence and incidence of depression and other DSM-III-R disorders in high school students. J Abnorm Psychol 102(1):133–144. Mayr B, Montminy M. 2001. Transcriptional regulation by the phosphorylation-dependent factor CREB. Nat Rev Mol Cell Biol 2(8):599–609. McEwen BS. 2001. Invited review: Estrogens effects on the brain: Multiple sites and molecular mechanisms. J Appl Physiol 91(6):2785–2801. Nolen-Hoeksema S. 1987. Sex differences in unipolar depression: Evidence and theory. Psychol Bull 101(2):259–282. Orvaschel H. 1990. Early onset psychiatric disorder in high risk children and increased familial morbidity. J Am Acad Child Adolesc Psychiatry 29(2): 184–188. Rudolph D, Tafuri A, Gass P, Hammerling GJ, Arnold B, Schutz G. 1998. Impaired fetal T cell development and perinatal lethality in mice lacking the cAMP response element binding protein. Proc Natl Acad Sci USA 95(8):4481–4486. Sham PC, Curtis D. 1995. An extended transmission/disequilibrium test (TDT) for multi-allele marker loci. Ann Hum Genet 59(Pt 3):323–336. Sherrill JT, Kovacs M. 2000. Interview schedule for children and adolescents (ISCA). J Am Acad Child Adolesc Psychiatry 39(1):67–75. Strauss J, Barr CL, George CJ, King N, Shaikh S, Devlin B, Kovacs M, Kennedy JL. 2004. Association study of brain-derived neurotrophic factor in adults with a history of childhood onset mood disorder. Am J Med Genet B Neuropsychiatr Genet 131(1):16–19. Sullivan PF, Neale MC, Kendler KS. 2000. Genetic epidemiology of major depression: Review and meta-analysis. Am J Psychiatry 157(10):1552– 1562. Thapar A, McGuffin P. 1994. A twin study of depressive symptoms in childhood. Br J Psychiatry 165(2):259–265. Vaidya VA, Duman RS. 2001. Depresssion—Emerging insights from neurobiology. Br Med Bull 57:61–79. Wade TJ, Cairney J, Pevalin DJ. 2002. Emergence of gender differences in depression during adolescence: National panel results from three countries. J Am Acad Child Adolesc Psychiatry 41(2):190–198. Walker WH, Girardet C, Habener JF. 1996. Alternative exon splicing controls a translational switch from activator to repressor isoforms of transcription factor CREB during spermatogenesis. J Biol Chem 271(33):20145–21050. Weissman MM, Klerman GL. 1977. Sex differences and the epidemiology of depression. Arch Gen Psychiatry 34(1):98–111. Kessler RC, McGonagle KA, Swartz M, Blazer DG, Nelson CB. 1993. Sex and depression in the National Comorbidity Survey. I: Lifetime prevalence, chronicity and recurrence. J Affect Disord 29(2–3):85–96. Zubenko GS, Hughes HB 3rd, Maher BS, Stiffler JS, Zubenko WN, Marazita ML. 2002. Genetic linkage of region containing the CREB1 gene to depressive disorders in women from families with recurrent, early-onset, major depression. Am J Med Genet 114(8):980–987. Kovacs M. 2001. Gender and the course of major depressive disorder through adolescence in clinically referred youngsters. J Am Acad Child Adolesc Psychiatry 40(9):1079–1085. Zubenko GS, Hughes HB 3rd, Stiffler JS, Brechbiel A, Zubenko WN, Maher BS, Marazita ML. 2003a. Sequence variations in CREB1 cosegregate with depressive disorders in women. Mol Psychiatry 8(6):611–618. Kovacs M, Devlin B. 1998. Internalizing disorders in childhood. J Child Psychol Psychiatry 39(1):47–63. Zubenko GS, Maher B, Hughes HB 3rd, Zubenko WN, Stiffler JS, Kaplan BB, Marazita ML. 2003b. Genome-wide linkage survey for genetic loci that influence the development of depressive disorders in families with recurrent, early-onset, major depression. Am J Med Genet B Neuropsychiatr Genet 123(1):1–18. Kovacs M, Devlin B, Pollock M, Richards C, Mukerji P. 1997. A controlled family history study of childhood-onset depressive disorder. Arch Gen Psychiatry 54(7):613–623.