• CRHR1;
  • depression;
  • haplotype;
  • male;
  • SNP;
  • stress;
  • suicidality


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Dysregulation in the stress response of the hypothalamic–pituitary–adrenal axis, involving the corticotrophin-releasing hormone and its main receptor (CRHR1), is considered to play a major role in depression and suicidal behavior. To comprehensively map the genetic variation in CRHR1 in relation to suicidality and depression, as a follow-up to our initial report on SNP rs4792887, we analyzed six new single nucleotide polymorphisms (SNPs), in an extended sample of family trios (= 672) with suicide attempter offspring, by using family-based association tests. The minor T-allele of exonic SNP rs12936511, not previously studied in the context of psychiatric disorders and suicidal behaviors, was significantly transmitted to suicidal males with increased Beck Depression Inventory (BDI) scores (= 347; = 0.0028). We found additional evidence of association and linkage with increased BDI scores among suicidal males with an additional SNP, located proximally to the index SNP rs4792887, as well as with two distal SNPs, which were correlated with index SNP rs4792887. Analysis of haplotypes showed that each of the risk alleles segregated onto three separate haplotypes, whereas a fourth ‘nonrisk’ haplotype (‘CGC’) contained none of the risk alleles and was preferentially transmitted to suicidal males with lowered BDI scores (= 0.0007). The BDI scores among all suicidal males, who carried a homozygous combination of any of the three risk haplotypes (non-CGC/non-CGC; = 160), were significantly increased (= 0.000089) compared with suicidal male CGC carriers (= 181). Thus, while the characteristics of the suicide female attempters remained undetermined, the male suicidal offspring had increased depression intensity related to main genetic effects by exonic SNP rs12936511 and homozygous non-CGC haplotypes.

Suicide is one of the leading causes of death among young and middle-aged men. According to the World Health Organization (WHO), one million people take their own lives each year in the world and at least 10 times as many attempt suicide (Wasserman 2001). Twin, family and adoption studies have shown the involvement of genetic components in suicidality, with heritability estimates in the range of 17–55% (Brent & Mann 2005; Voracek & Loibl 2007). The complexity of suicidality is summarized in a stress-diathesis model, which describes an accumulation of exposures to environmental risk factors as well as genetic predispositions (Mann 2003; Wasserman 2001). Up to date, a number of genetic variants have been studied and implicated as having roles in suicidality, with main focus on the serotonergic system (Bondy et al. 2006; Rujescu et al. 2007).

Dysfunction of the stress-responsive hypothalamic–pituitary–adrenal (HPA) axis is a common feature in anxiety and mood disorders (reviewed in Bale & Vale 2004; Hauger et al. 2006; Nemeroff & Vale 2005; Swaab et al. 2005) as well as in suicidality (Mann & Currier 2007). Furthermore, dysregulation of the HPA axis is the most potent biological marker presently available for predicting suicide among depressed individuals in combination with markers of serotonergic activity (Coryell & Schlesser 2007; Mann et al. 2006). Activation of the HPA axis is controlled and regulated by hypothalamic corticotrophin-releasing hormone (CRH), which activates CRH receptor 1 (CRHR1) in the anterior pituitary, to mediate the production of adrenocorticotrophic hormone, involving, for example transcription factor TBX19 (Lamolet et al. 2001; Liu et al. 2001), in turn promoting the synthesis and release of cortisol from the adrenal gland. The rise of cortisol levels in blood is normally regulated by multipoint feedback loop, at both HPA and also limbic, brainstem and prefrontal brain areas. Furthermore, extrahypothalamic effects are mediated directly through CRHR1 at the central amygdala, affecting, for example the serotonin system. Dysregulations in this system are highly relevant for a variety of psychopathologies (e.g. depression, anxiety and impulsive aggression) commonly found among suicidal individuals.

We previously identified a single nucleotide polymorphism (SNP) in the CRHR1 gene, rs4782887, which showed linkage and association in a subgroup of suicide attempters exposed to low–medium levels of stressful life events (SLEs), among whom most of the males were depressed (Wasserman et al. 2008). Furthermore, a relationship was also shown between neurotic personality traits, suicidality and the genetic variation in the HPA regulatory TBX19 gene (Wasserman et al. 2006a). In the present study, we continued the investigation of the CRHR1 gene with six new SNPs, covering approximately 80% of the genetic variation, in an expanded sample of 672 nuclear family trios with suicide attempter offspring. The objective was to relate the majority of the known genetic variation in the CRHR1 gene with the depression intensity among suicide attempter offspring, taking into account the previously identified subgroup with low–medium levels of SLE exposure (‘SLE1–3’). As discussed previously (Wasserman et al. 2006b, 2008), individuals in such a low-predisposing subgroup of SLE exposure may display increases in genetic effects of certain risk alleles (in contrast to individuals with high levels of SLE exposure in which case the environment may assume dominant effects over certain risk alleles) similar to observations by others (Hansson et al. 2006; Kendler et al. 2005). Because there are differences in suicidal behavior, as well as in the occurrence of depression, between males and females (Wasserman 2001, 2006), the effects were additionally studied in relation to gender.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Experimental subjects

Trios (offspring, with at least one suicide attempt and both parents) were collected as previously described (Wasserman et al. 2005). The collection of research subjects followed the code of ethics of the World Medical Association (Declaration of Helsinki). Before participation in the study, all research subjects were given oral and written information about the project. A written consent was subsequently obtained from them. The study was approved by the Research Ethics Committee at the Karolinska Institute (Dnr 97–188) and by the Ministry of Health in Ukraine. Since our previous report (Wasserman et al. 2008), we had collected an additional sample of 178 trios. The total sample of the present explorative study, consisted of 672 complete nuclear family trios, containing the 178 newly collected trios, as well as 494 (91%) of the previously described trios (Wasserman et al. 2008). Families with Mendelian errors or failed genotypings were not included in the samples. All offspring had performed at least one suicide attempt, and the index attempt was evaluated by using the Medical Damage Rating Scale (Beck et al. 1975), scoring the medical damage of the suicide attempt from 0 (none) to 8 (dead). A medical damage score ≥2 for the offspring was required for inclusion in the study. All research subjects were Caucasian and Ukrainian citizens. The heritage of the offspring in the sample was as follows; 59.0% had at least three grandparents of Ukrainian nationality, 17.1% had at least three grandparents who were of Ukrainian or Russian nationality and 5.3% had two, or more, grandparents of other origin than Ukrainian or Russian, predominantly from other East European regions. The mean (SD) offspring age was 24.1 (7.2) years, with a female/male ratio of 48.4%/51.6%. Previous suicide attempts were admitted by 29.0%. Diagnoses according to the WHO International Classification of Diseases 10 were determined by using the Composite International Diagnostic Interview (Robins et al. 1988), computerized core version 2.1 ( and adjustment disorder diagnosis according to Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria (American Psychiatric Association 2000). Panic disorder, phobias, obsessive compulsive disorder, posttraumatic stress or generalized anxiety disorder were present among 23.5%. Schizophrenia, delusional disorder or acute and transient psychotic disorders were present among 6.8%. Harmful substance use or dependence syndromes were present among 15.5%. Bipolar disorder was diagnosed for 1.0%. Adjustment disorder was observed for 15.5%. Mild depressive episodes were present among 1.8%, moderate depressive episodes among 3.7% and severe depressive episodes were present among 5.5%. Dysthymia was present among 5.0%.

Depression phenotype

The Beck Depression Inventory (BDI) score (Beck et al. 1961) was the main phenotypic depression parameter used for analysis in this study because it had the most complete coverage among the offspring in the sample (98.7%), as well as being positively correlated with the other depression measures, when available. Most the BDI scores had been assessed within the month that followed the suicide attempt (50% within 1 week and 80% within 1 month). The BDI is a measure of current depressive state (during the last 2 weeks) and not of the lifetime history of a depressive episode or chronic depression. Thus, the use of the BDI measure in the present study primarily represented the depressive state of the offspring in connection to the suicide attempt. According to the BDI, 65.6% of the suicidal offspring (61% of the males and 70% of the females, respectively) had symptoms of at least mild mood disturbances (BDI score >9). The median [interquartile range (IQR)] BDI scores among male suicidal offspring was 13 (6–24), which was significantly lower compared with the BDI scores of 17 (8–26) among female suicidal offspring (< 0.01). Similarly, the median (IQR) BDI scores of 6 (2–12) among the male nonsuicidal parents were significantly lower compared with the BDI scores of 10 (5–17) among the female nonsuicidal parents (< 0.0001).

Stressful life events

An SLE inventory was used as previously described (Wasserman et al. 2008). The mean (SD) SLE scores among suicidal offspring were significantly higher compared with the SLEs of the nonsuicidal parents [3.8 (±2.5) vs. 2.8 (±2.2); < 0.0001] (Wasserman et al. 2008) who were used for comparisons in some analyses. The SLEs formed the basis for making median split subgroups, with offspring exposed to low–medium or medium–high levels of SLEs as described previously (Wasserman et al. 2008). Of main importance was the low–medium SLE subgroup (termed SLE1–3, containing all trios with submedian score for offspring, except for offspring with score ‘0’) in which the original finding with SNP rs4792887 was shown (Wasserman et al. 2008). Thus, this subgroup was also used here for comparison. Another SLE >3 subgroup, with offspring having medium–high levels of SLEs (>3), was also analyzed, if indicated.

DNA preparation and genotyping

Venous blood (10 ml) was taken from all research subjects into ethylenediaminetetraacetic acid-containing tubes. DNA isolation was performed as described previously (Geijer et al. 1994). Genotyping of the sample was performed by the custom genetic analysis services using an Illumina BeadStation 500GX (GoldenGate® Assay; Illumina Inc., San Diego, CA, USA) at a quality-assured facility (Molecular Medicine, Department of Medical Sciences, SNP technology platform, Uppsala University, Sweden). Concordance in the output was compared with the two other genotyping platforms used previously (Wasserman et al. 2008), with an overlapping sample of 83 trios, resulting in >99.5% similarity for rs4792887.

Selection of SNPs

A synonymous exonic SNP rs12936511 was selected for being a potential determinant of functional effects at the level of splicing regulation (as determined by using FASTSNP) (Yuan et al. 2006). While exonic SNPs rs12936511 and rs16940665 had not been the focus of any previous studies, the other five SNPs we choose had previously been studied by others in the context of either depression or alcohol intake (Blomeyer et al. 2008; Bradley et al. 2008; Licinio et al. 2004; Liu et al. 2006, 2007; Papiol et al. 2007; Treutlein et al. 2006). Thus, in addition to our previously analyzed index SNP rs4792887 (Wasserman et al. 2008), we chose six additional SNPs for genotyping; rs110402, rs12936511, rs242939, rs242938, rs1876831 and rs16940665 (Fig. 1), which together captured much of the variation in the CRHR1 gene. Five of the SNPs analyzed here are in Hapmap Phase II (rs4792887, rs110402, rs242939, rs1876831 and rs16940665) and are correlated with several of the other CRHR1 SNPs, together representing 80% (35/44) of the genetic variation covered by the Hapmap SNPs in Utah residents with ancestry from northern and western Europe (CEU population; (r2 ≥ 0.80).


Figure 1. The CRHR1 gene and investigated SNPs. Upper panel: a schematic of the CRHR1 gene is shown, with two alternatively spliced variants (as indicated by the Ensembl database), with thin angled lines representing introns, which are spliced away, and vertical lines representing exons. The scaled bar indicates kilobases at chromosome 17 (according to Ensembl). Arrows indicate the positions of the investigated SNPs. Lower panel: a matrice of the pairwise LD values, among the investigated SNPs (numbered 1–7, with the corresponding rs-identifiers indicated), in the total trio sample (= 672), as determined by using Haploview. Numbers in squares represent r2 values between each corresponding SNP (also depicted graphically with shades of grey, from white (r2 = 0) to black (r2 = 1)]. The triangle surrounding the matrice depicts a haplotype block (defined by the confidence interval method) generated by Haploview. Lewontin’s D′ was in the range of 0.89–1.00 between all SNPs.

Download figure to PowerPoint

Statistical analyses

All data were stored and organized for analyses by using FileMaker Pro (FileMaker, Inc., Claris corp., Santa Clara, CA). Analyses of linkage disequilibrium (LD) patterns and transmission disequilibrium test (TDT) with dichotomized phenotypes (i.e. SLE1–3 or ‘BDI > 9’ subgroups) were primarily performed by using Haploview v4.0 (Barrett et al. 2005). Family-based association test (FBAT) was used for analysis of continuous BDI scores, as implemented in fbat v2.0 (Laird et al. 2000) ( Phased offspring and parent haplotypes were obtained by using the best pairs generated by unphased v3.0.12 (Dudbridge 2008) and phase v2.1 (Stephens et al. 2001), respectively. Tests of gene by environment (G × E) interactions were performed by using the algorithms implemented in software pbat v.3.6 (Lange et al. 2004). Other comparisons, for example those being performed as a test of association (confirming the effect(s) among all offspring not restricted to those with heterozygous parents), were performed by using either the Mann–Whitney U test or the 2 × 2 contingency tables (with Fisher’s exact P values shown). Cut offs for high/low comparisons with the BDI variable, were either >9 (bordering to mild mood disturbances) or ≥17 (bordering to clinical depression), used as indicated. Statistical power was calculated using the software TDT Power Calculator (Chen & Deng 2001). The threshold for a significant result was < 0.05. The number of independent SNPs analyzed in the CRHR1 gene was determined by using SNP Spectral Decomposition (Nyholt 2004) ( This established an experiment-wide Bonferroni-corrected level of = 0.0127 for the analysis of the seven CRHR1 SNPs. Permutation testing (105) was also performed when available, as implemented in fbat v2.0 and Haploview v4.0. Uncorrected (nominal) P values are mainly shown, which are referred to as nonsignificant tendencies (> 0.05), nominally significant (uncorrected < 0.05) or significant (corrected < 0.05).


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Novel CRHR1 SNPs associated and linked with increased BDI scores among male suicide attempters, with a main genetic effect by exonic SNP rs12936511

In the present study, we follow up on our initial report (Wasserman et al. 2008) by performing a more complete explorative analysis of the CRHR1 locus, with more novel SNPs, now in a total combined sample of 672 trios, containing 494 (91%) of the previously analyzed trios (Wasserman et al. 2008) as well as 178 new trios (see Materials and methods). Six additional SNPs were selected for the exploratory analysis; rs110402, rs12936511, rs242939, rs242938, rs1876831 and rs16940665 (Fig. 1). These SNPs had never been investigated in relation to suicidality by others.

The SLE1–3 subgroup

Our previous results showed overtransmission of the minor T-allele of SNP rs4792887 in a subgroup of suicide attempter offspring exposed to low–medium (but not high) levels of SLEs (SLE1–3 range) (Wasserman et al. 2008). This low–medium SLE subgroup (SLE1–3) was reasoned to have better potential to reflect the genetic effect in relation to SLEs, as higher levels of SLEs may exert a dominant role over certain genes (>Wasserman et al. 2006b, 2008), similar to observations by others (Hansson et al. 2006; Kendler et al. 2005). Similar approaches of genetic analysis of low-risk subgroups has also been proposed or used by others (Kendler & Eaves 1986; Marenberg et al. 1994; Reynisdottir et al. 2003). Here, novel analysis with the six new SNPs showed that two additional SNPs (rs242939 and rs242938), which were correlated and in LD with index SNP rs4792887 (r2 = 0.75; Fig. 1), had significant overtransmission of their minor alleles in the SLE1–3 subgroup (61 C/A:31 T/G, = 315; = 0.002), as did the index SNP rs4792887 (data not shown) (Wasserman et al. 2008). Interestingly, a new independent result was further observed with exonic SNP rs12936511 (which was not correlated with index SNP rs4792887, r2 = 0.006; Fig. 1), with significant overtransmission of the major C-allele in the same SLE1–3 subgroup (18 T:38 C, = 315; = 0.0075). The other SNPs showed no significant results and none of the SNPs showed transmission distortion in the total sample. Analysis in relation to gender showed that transmission distortion was present in both males and females in the SLE1–3 subgroup (data not shown). We concluded that the CRHR1 locus continued to show transmission distortion for certain SNPs among suicidal individuals in the SLE1–3 subgroup, but not in the total sample, congruent with previous results (Wasserman et al. 2008).

BDI scores

We next wished to further analyze the SNPs also in relation to depression, a parameter which was indicated previously among suicidal male offspring in the SLE1–3 subgroup, using index SNP rs4792887 (Wasserman et al. 2008). We now performed the analysis in relation to BDI scores, a measure of the current depression intensity (see Materials and methods). Individual FBAT tests for each SNP were performed with BDI score as phenotype. Analyzing the total sample (= 663; nine trios had missing BDI scores) showed no significant results (data not shown). Next, the sample was stratified by gender. Table 1 (third column) shows that the minor T-allele of exonic rs12936511 was significantly overtransmitted to males with increased BDI scores (= 0.0028).

Table 1.  Allelic transmission in relation to BDI scores and low–medium SLE exposure among suicidal male offspring
SNPf (MAF)BDI, allBDI in SLE1–3BDI > 9 in SLE1–3
  • SNP, rs# of investigated SNPs, with the previously analyzed SNP are in bold (Wasserman et al. 2008). Minor allele frequencies (MAF) of the corresponding minor alleles (A/C/G/T). Transmission distortion in relation to the BDI score in all males (BDI, all), to BDI score in males exposed to low–medium levels of SLEs (BDI in SLE1–3) or only among males with BDI scores >9 and exposed to low–medium levels of SLEs (BDI > 9 in SLE1–3) are presented with their corresponding uncorrected P value (if < 0.05) or as nonsignificant (NS; ≥ 0.05).

  • FBAT minP test (105 permutations): = 0.0030 (single SNP) and = 0.0127 (all SNPs).

rs47928870.10 (T)105NS50NS1050.0280
rs1104020.48 (G)261NS1270.0300261NS
rs129365110.05 (T)580.002825NS58NS
rs242939/rs2429390.09 (C)/0.09 (A)92NS43NS920.0196/0.0133
rs1876831/rs169406650.14 (T)/0.14 (C)148NS70NS148NS
BDI scores in the SLE1–3 subgroup

Next, we performed joint analysis in relation to BDI scores in the SLE1–3 subgroup (Table 1, fourth column), whereby nominally significant BDI scores were observed with transmission of the major A-allele of rs110402 (= 0.0300). Interestingly, a nonsignificant trend for rs110402 was observed also in the total sample of males (= 0.1041; data not shown). Finally, we also performed FBAT tests with a dichotomized BDI parameter BDI > 9 in the SLE1–3 subgroup (Table 1, fifth column) similar to previous analysis with index SNP rs4792887 (Wasserman et al. 2008). The results showed nominally significant results for the index SNP rs4792887, and the new correlated SNPs rs242939/rs242938, congruent with the previous findings (Wasserman et al. 2008), but no significance for SNPs rs12936511 and rs110402. Females showed no significant results with any SNP in relation to depression (data not shown) as reported previously (Wasserman et al. 2008).

CRHR1 SNP by SLE interactions on BDI outcome

These above described findings with male suicide attempter offspring and depression were also examined for the presence of CRHR1 SNP by SLE score G × E interactions on BDI outcome (BDI scores or BDI > 9, among all males or males in subgroup SLE1–3), showing no significant results (data not shown).

We concluded that this mapping of the CRHR1 locus provided a congruent expansion of the previous findings (Wasserman et al. 2008). In relation to the low–medium SLE exposure (SLE1–3 subgroup), we found overtransmission of minor alleles of previous index SNP rs4792887 and correlated new SNPs rs242939/rs242938 as well as major C-allele of the uncorrelated new exonic SNP rs12936511. In relation to the BDI parameter, we obtained results among suicidal males but not females. In summary, the analyses showed association and linkage to BDI measures among suicide attempter males with three noncorrelated CRHR1 SNPs, with a main effect for novel exonic SNP rs12936511 (minor T-allele, continuous BDI scores) and with lower end gene–environment interplay between the SNPs rs4792887 (minor T-allele, dichotomous BDI > 9)/rs110402 (major A-allele, continuous BDI scores) and the SLE1–3 subgroup.

The relation between T-allele of exonic SNP rs12936511 and increased BDI scores is specific for suicidal males

We wished to further investigate the clinical meaning of the identified association and linkage of the T-allele of SNP rs12936511 in relation to the increases in BDI scores, as well as the suicidality per se, among the males. Thus, comparing the proportions of transmitted T-alleles [= 29; median (IQR) BDI score = 21 (10–28)] vs. C-alleles [= 33; median (IQR) BDI score = 13 (6–19)], in relation to a BDI cutoff (≥17, being the subsample median as well as border for clinical depression), showed that the increase in BDI scores among T-allele receivers occurred at a clinically relevant levels (Fisher’s exact test, = 0.005). The corresponding OR (95% CI) was 5.1 (1.8–15).

Extending the analysis to all male suicidal offspring (in contrast to only the offspring of heterozygous parents, as used in FBAT above), with an allelic association analysis (T carriers vs. non-T carriers), yielded similar results [Table 2; OR (95% CI) = 3.1 (1.4–6.7); = 0.004]. Next, we wished to compare the T-allele and BDI scores in relation to the suicidality among the males. Because all offspring were suicidal in the studied sample, we used the nonsuicidal parents as controls, as was similarly carried out in a previous study (Wasserman et al. 2006a). We compared T carriers vs. non-T carriers (in relation to the BDI cutoff of 17) among suicidal male offspring (= 341) vs. the nonsuicidal male parents (= 648). The results showed a nominally significant increase in proportion of T carriers among the male suicidal offspring with high BDI scores compared with the corresponding male nonsuicidal parents [Table 2; OR (95% CI) = 3.0 (1.1–8.7); = 0.045]. Inversely, suicidal male offspring having low BDI scores showed nonsignificant tendencies for lowered proportion of T carriers compared with the corresponding male parents [Table 2; OR (95% CI) = 0.58 (0.3–1.1); = 0.16). A similar analysis for the female offspring yielded no differences (data not shown). The overall allele frequencies (without consideration to BDI scores) did not differ significantly between any of these groups (range 3.6–5.3%). In summary, the results indicated that the T-allele of exonic SNP rs12936511 constitute a risk factor for increased BDI scores in males in a manner specific for suicidality.

Table 2.  Increased BDI scores, among male T-allele carriers of SNP rs12936511, show specificity for suicidality
SampleNBDIT carriers (n)Non-T carriers (n)
  1. Comparison of the proportions of T- (genotypes TT and CT) and non-T (genotype CC) carriers [% (number of carriers)], among male suicidal offspring vs. male nonsuicidal parents, with a cutoff for the BDI score of 17 (BDI high, score ≥ 17 and BDI low, score < 17). The analyzed cells, which yielded a significant difference related to suicidality vs. nonsuicidality, are in bold.

Suicidal offspring146High14.4% (21)85.6% (125)
Suicidal offspring195Low5.1% (10)94.9% (185)
Nonsuicidal parents76High5.3% (4)94.7% (72)
Nonsuicidal parents572Low8.6% (49)91.4% (523)

A haplotype (‘CGC’) is transmitted to suicidal males with lowered BDI scores

To further investigate the three risk SNPs (rs4792887, rs110402 and rs12936511), which were all associated and linked with the BDI parameter among suicidal males (see above), we further analyzed the transmission of haplotypes in relation to continuous BDI scores. Table 3 shows that the second most frequent haplotype was significantly overtransmitted to suicidal males with low BDI scores (CGC, = 0.32, = 0.0025), an effect that appeared augmented in the SLE1–3 subgroup (= 0.0007). Each of the three other (risk-) haplotypes contained one of the three SNP risk alleles (CAC, major A-allele of rs110402; TGC, minor T-allele of rs4792887 and CGT, minor T-allele of rs12936511) and showed similar results compared with the results of single SNP analysis (cf. Tables 1 and 3 and data not shown). Taken together, the results indicated the CGC haplotype to be a overall indicator of the absence of any SNP risk alleles.

Table 3.  Haplotype CGC is associated and linked with decreased BDI scores among suicidal males
HaplotypefAll malesMales in SLE1–3
  • Analysis of haplotypes [frequencies (f) > 0.01] with three adjacent SNPs (rs4792887, rs110402 and rs12936511) yielded four major haplotypes, as depicted. CAC (major A-allele of rs110402 underlined), CGC (containing no risk alleles), TGC (minor T-allele of rs4792887 underlined) and CGT (minor T-allele of rs12936511 underlined).

  • *

    Hbat permutations: = 0.0021 (χ2 sum) and = 0.0071 (minP).

  • Hbat permutations: = 0.0044 (χ2 sum) and = 0.0017 (minP).


To investigate this further, we analyzed the BDI scores of carriers vs. noncarriers of the CGC haplotype, now extended to all males as an association analysis (‘CGC/CGC’ and ‘CGC/other’ vs. ‘other/other’). The results showed that male homozygous non-CGC individuals, having any combination of the risk allele containing haplotypes (i.e. other/other), had significantly elevated BDI scores, compared with heterozygous or homozygous carriers of the CGC haplotype (i.e. CGC/CGC or CGC/other), in the entire sample (= 0.000089; Table 4) and in the SLE1–3 subgroup (= 0.000597; Table 4), as well as among males in the SLE >3 subgroup (= 0.0142; data not shown).

Table 4.  Suicidal, male ‘non-CGC/non-CGC’ carriers have elevated BDI scores
Haplotype*All malesMales in SLE1–3
nMd (IQR)PnMd (IQR)P
  • *

    ‘CGC’ refers to heterozygous and homozygous CGC carriers (CGC/CGC and CGC/other), whereas ‘non-CGC’ refers to homozygous carriers of any combinations of the other haplotypes CAC, TGC or CGT (i.e. ‘other/other’), each containing either one of the three SNP risk alleles identified (underlined).

  • Median (Md) and IQR of BDI scores.

  • Mann–Whitney U test.

CGC18111 (5–20)0.0000898410 (4–19)0.000597
Non-CGC16018 (8–28)7719 (8–29)

To investigate the relationship with male suicidality per se (as was performed for SNP rs12936511; Table 2), we compared the number of non-CGC and CGC carriers having high BDI scores, with suicidal male offspring vs. nonsuicidal male parents, respectively. The results showed a nominally significant increased proportion of suicidal male offspring non-CGC carriers having increased BDI scores compared with the nonsuicidal male parents [BDI cutoff >9, OR 95% (CI) = 1.6 (1.1–2.4); = 0.013; data not shown].

In summary, we concluded that each of the three risk alleles indicated with increased BDI scores among suicidal males in the present study, that is the minor T-alleles of rs4792287 and rs12936511 as well as the major A-allele of rs110402, were present on three separate haplotypes, respectively. A fourth haplotype, CGC, which did not contain any of the three SNP risk alleles, acted as a compound marker of lowered BDI scores among suicidal males (in a dominant manner). Inversely, homozygous male suicide attempters whom carried any (recessive) combination of the risk alleles (‘non-CGC/non-CGC’) had significantly increased BDI scores.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

The present study is a comprehensive mapping of the genetic variation in the CRHR1 gene, following our initial finding with index SNP rs4792887 in the SLE1–3 subgroup (Wasserman et al. 2008), now further indicating a main genetic effect on depression intensity among suicide attempter males by a new, exonic SNP rs12936511 and homozygous non-CGC haplotypes. The three risk SNPs (rs4792887, rs110402 and rs12936511) indicated here with depression among suicidal males are located within close physical range in the CRHR1 gene, being evenly distributed within a span of 7.4 kb. The SNPs are not well correlated (low r2; Fig. 1), but all three SNP are in LD by Lewontin’s D′ (complete LD, D′ = 1). This is partly because of the greater sensitivity of the r2 measure to differences in allele frequencies and has been similarly observed elsewhere (Long et al. 2004; Tiret et al. 2002). Different changes in the CRHR1 gene may result in similar phenotypic effects (i.e. genetic heterogeneity, as with the prototypic example of the phenylalanine hydroxylase gene) (Scriver 2007). The segregation of the three risk alleles onto separate haplotypes (and their low, pairwise r2 values), but in relationship to a similar phenotype, suggests that there may be more than one causal loci also in CRHR1. Haplotype analysis proved useful for identifying an overall effect of the absence of any of the three risk alleles, risk alleles that each one had less significant results by themself (covering less affected individuals).

Whereas rs4792887 and rs110402 may not likely be functional determinants per se (but are rather in LD with such loci), we speculate that exonic rs12936511 may affect the CRHR1 gene at, for example the level of alternative splicing regulation. While this exonic SNP does not alter the amino acid sequence, our in silico analysis showed that it is located in a putative exonic splicing enhancer and that the presence of the T-allele destroys potential binding of SR protein SRp55, while also altering the binding efficiency of SC35. This is not without relevance here because alternative splicing of CRHR1-messenger RNA is occurring in this region (Fig. 1). Besides affecting splicing, single synonymous SNPs have capacity to have profound effects on a gene by alternative mechanisms (Parmley & Hurst 2007). It would be interesting to determine any possible functional effects linked to the identified SNPs in an experimental setting. Until the causal variants have been identified in full, the use of SNP haplotypes is likely to be the best alternative. The major C-allele of rs12936511 was indicated in the SLE1–3 subgroup and the opposite, minor T-allele with male suicidal depression. We speculate that this dual result may reflect, for example the complexity of splicing regulation, epistasis with other polymorphisms or some form of selective pressure at this SNP. The newly collected trios (= 178) had more than fourfold less families with heterozygous parents (i.e. TDT informative families) compared with the initially investigated sample (Wasserman et al. 2008), which was insufficient for confirmatory replication of the previous findings with (index) SNP rs4792887 (yielding a low statistical power). All TDT analyses with SNP rs4792887 in the new subsample were nonsignificant with neutral tendencies (data not shown).

Congruent with these results presented and discussed here is the observation that many depressed individuals who committed suicide were also nonresponsive to the dexamethasone suppression test, a biological measure of reduced HPA axis feedback sensitivity (Coryell & Schlesser 2007; Mann et al. 2006). Whereas we are presently, to our knowledge, alone on focusing the study of genetic variation in CRHR1 in relation to suicidality and depression in suicidal males, others have performed studies in relation to depression alone (Bradley et al. 2008; Licinio et al. 2004; Liu et al. 2006, 2007; Papiol et al. 2007). Bradley et al. reported results with SNPs in the similar 5′ region of the CRHR1 gene, as was implicated here, with overlap with two of the SNPs reported here (rs4792887 and rs110402) (Bradley et al. 2008). Among other results, it was shown that the T- and G-alleles were linked with increased depression (Bradley et al. 2008), whereas our results show depression being linked with the T- and A-alleles, of SNPs rs4792287 and rs110402, respectively. Our results were among suicidal males with mainly low–medium SLE scores, whereas Bradley et al. did not investigate the suicidality parameter, having their main findings among predominantly females, which may in part explain the differences between us and Bradley et al. concerning the results with the A-allele of rs110402. Papiol et al. showed association with the A-allele of rs110402 with age of onset and seasonal pattern of major depression (Papiol et al. 2007). This was the opposite allele compared with Bradley et al., but the same allele as presented here with suicidal males. The finding of Papiol et al. with the A-allele was made among nonsuicidal females (Papiol et al. 2007), which is congruent with our results of no relation of the A-allele with depression among suicidal females. This may suggest that the A-allele of rs110402 is indicative of increased risk of depression in nonsuicidal females and according to our results related with suicidal depression among the males.

By performing a randomized, placebo and double-blind trial, Licinio et al. observed that those depressed and highly anxious Mexican–Americans, who were homozygous carriers of a GAG haplotype, had an increased response to an 8-week period of antidepressant treatment (Licinio et al. 2004). These results were later replicated in a population of Han Chinese patients (Liu et al. 2007). Interestingly, Licinio et al. reported no genetic association with presence of depression per se (Licinio et al. 2004). Nevertheless, the GAG haplotype, containing the major A-allele of rs242939 (underlined), was indicated with better treatment response (i.e. major allele = nonrisk allele) and is correlated with our index SNP rs4792287 (Fig. 1). The opposite minor alleles of these SNPs were linked with risk for depression by us and Bradley et al. (i.e. minor alleles = risk alleles). Thus, the results of Licinio et al. are congruent with ours and Bradley et al. describing the relationship between minor/major and risk/nonrisk alleles. We speculate that males who carry the minor risk alleles may not only be less susceptible to pharmacological treatment (Licinio et al. 2004) but perhaps also be more at risk for suicidality in relation to depression, as has been observed among inpatients with failed treatment (Mann & Currier 2007). Liu et al. investigated the same three SNPs as Licinio et al. and found that SNP rs242939 (Fig. 1) had a doubling in the frequency of the minor allele (from 0.07 to 0.14), among Han Chinese patients with major depression (Liu et al. 2006). Similarly with the results of Licinio et al., the GAG haplotype was not found to be associated with depression per se (Licinio et al. 2004; Liu et al. 2006). In summary, the results with rs4792887 and rs110402 can be viewed as partly congruent across the studies, whereas the exonic SNP rs12936511, shown here to be linked with suicidal male depression, has not been studied previously by others. Because our findings with rs4792887 were made in relation to a BDI cutoff of 9 in the SLE1–3 subgroup, whereas the finding with exonic SNP rs12936511 was resolved in relation to a cutoff of 17 among all males, the latter SNP may be regarded as being more strongly implicated with clinically relevant increases in depression intensity among the suicidal males.

The apparent sexual dimorphism showed in the analyses here may be caused by either genetic/biological or phenotypic heterogeneities or both. This former is not unlikely because the stress response of the HPA axis is known to act in a sexually dimorphic manner on, for example mood and anxiety, particularly during certain times of the reproductive life cycle (Leibenluft 1999; Rhodes & Rubin 1999). The latter may be a reflection of the fact that the phenotype of suicidal behavior also displays a sexual dimorphism (Wasserman 2006), reflecting that the male attempters are more likely to end up as suicide completers, while the females are more likely to remain as suicide attempters. We speculate that the gender differences observed here may indeed be caused by a combination such as genetic and phenotypic differences. The results and conclusions presented here for suicidal males and depression can thus be viewed in light of the likelihood that a higher proportion of the depressed males may die by suicide, compared with depressed females, among the general population (Angst et al. 2002; Wasserman 2006).

We have identified genetic variants in the CRHR1 gene, which may be of importance in the prediction and treatment of depressed males, at risk of suicidal behavior. Identification of genetic factors affecting the HPA axis, such as reported here, may further help identify the fraction of individuals that are suicidal, from the many nonsuicidal and depressed individuals in the general population (Wasserman 2001, 2006). The results motivate continued exploration of the genetics of HPA axis and suicidal behavior, as it may present new possibilities for improving the needed specificity of biological predictors, as well as to identify novel therapeutic approaches (Mann & Currier 2007).


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  • American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington, DC.
  • Angst, F., Stassenc, H.H., Claytonb, P.J. & Angstc, J. (2002) Mortality of patients with mood disorders: follow-up over 34–38 years. J Affect Disord 68, 167181.
  • Bale, T.L. & Vale, W.W. (2004) CRF and CRF receptors: role in stress responsivity and other behaviors. Annu Rev Pharmacol Toxicol 44, 525557.
  • Barrett, J.C., Fry, B., Maller, J. & Daly, M.J. (2005) Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21, 263265.
  • Beck, A.T., Ward, C.H., Mendelson, M., Mock, J. & Erbaugh, J. (1961) An inventory for measuring depression. Arch Gen Psychiatry 4, 561571.
  • Beck, A.T., Beck, R. & Kovacs, M. (1975) Classification of suicidal behaviors: I. Quantifying intent and medical lethality. Am J Psychiatry 132, 285287.
  • Blomeyer, D., Treutlein, J., Esser, G., Schmidt, M., Schumann, G. & Laucht, M. (2008) Interaction between CRHR1 gene and stressful life events predicts adolescent heavy alcohol use. Biol Psychiatry 63, 146151.
  • Bondy, B., Buettner, A. & Zill, P. (2006) Genetics of suicide. Mol Psychiatry 11, 336351.
  • Bradley, R., Binder, E., Epstein, M., Tang, Y., Nair, H., Liu, W., Gillespie, C., Berg, T., Evces, M., Newport, J., Stowe, Z., Heim, C., Nemeroff, C., Schwartz, A., Cubells, J. & Ressler, K. (2008) Influence of child abuse on adult depression. Arch Gen Psychiatry 65, 190200.
  • Brent, D.A. & Mann, J.J. (2005) Family genetic studies, suicide, and suicidal behavior. Am J Hum Genet C Semin Med Genet 133, 1324.
  • Chen, W.-M. & Deng, H.-W. (2001) A general and accurate approach for computing the statistical power of the transmission disequilibrium test for complex disease genes. Genet Epidemiol 21, 5367.
  • Coryell, W. & Schlesser, M. (2007) Combined biological tests for suicide prediction. Psychiatry Res 150, 187191.
  • Dudbridge, F. (2008) Likelihood-based association analysis for nuclear families and unrelated subjects with missing genotype data. Hum Hered 66, 8798.
  • Geijer, T., Neimann, J., Rydberg, U., Gyllander, A., Jönsson, E., Sedvall, G., Valverius, P. & Terenius, L. (1994) Dopamine D2-receptor gene polymorphisms in Scandinavian chronic alcoholics. Eur Arch Psychiatry Clin Neurosci 244, 2632.
  • Hansson, A.C., Cippitelli, A., Sommer, W.H., Fedeli, A., Bjork, K., Soverchia, L., Terasmaa, A., Massi, M., Heilig, M. & Ciccocioppo, R. (2006) Variation at the rat Crhr1 locus and sensitivity to relapse into alcohol seeking induced by environmental stress. Proc Natl Acad Sci U S A 103, 1523615241.
  • Hauger, R.L., Risbrough, V., Brauns, O. & Dautzenberg, F.M. (2006) Corticotropin releasing factor (CRF) receptor signaling in the central nervous system: new molecular targets. CNS Neurol Disord Drug Targets 5, 453479.
  • Kendler, K.S. & Eaves, L.J. (1986) Models for the joint effect of genotype and environment on liability to psychiatric illness. Am J Psychiatry 143, 279289.
  • Kendler, K.S., Kuhn, J.W., Vittum, J., Prescott, C.A. & Riley, B. (2005) The interaction of stressful life events and a serotonin transporter polymorphism in the prediction of episodes of major depression: a replication. Arch Gen Psychiatry 62, 529535.
  • Laird, N., Horvath, S. & Xu, X. (2000) Implementing a unified approach to family based tests of association. Genet Epidemiol 19 (Suppl. 1), S36S42.
  • Lange, C., DeMeo, D., Silverman, E.K., Weiss, S.T. & Laird, N.M. (2004) PBAT: tools for family-based association studies. Am J Hum Genet 74, 367369.
  • Lamolet, B., Pulichino, A.M., Lamonerie, T., Gauthier, Y., Brue, T., Enjalbert, A. & Drouin, J. (2001) A pituitary cell-restricted T box factor, Tpit, activates POMC transcription in cooperation with Pitx homeoproteins. Cell 104, 849859.
  • Leibenluft, E. (1999) Gender differences in mood and anxiety disorders. In Oldham, J.O. & Riba, M.B. (eds), Review of Psychiatry Series. American Psychiatric Press, Washington, DC. 1177.
  • Licinio, J., O’Kirwan, F., Irizarry, K., Merriman, B., Thakur, S., Jepson, R., Lake, S., Tantisira, K.G., Weiss, S.T. & Wong, M.L. (2004) Association of a corticotropin-releasing hormone receptor 1 haplotype and antidepressant treatment response in Mexican-Americans. Mol Psychiatry 9, 10751082.
  • Liu, J., Lin, C., Gleiberman, A., Ohgi, K.A., Herman, T., Huang, H.P., Tsai, M.J. & Rosenfeld, M.G. (2001) Tbx19, a tissue-selective regulator of POMC gene expression. Proc Natl Acad Sci U S A 98, 86748679.
  • Liu, Z., Zhu, F., Wang, G., Xiao, Z., Wang, H., Tang, J., Wang, X., Qiu, D., Liu, W., Cao, Z. & Li, W. (2006) Association of corticotropin-releasing hormone receptor1 gene SNP and haplotype with major depression. Neurosci Lett 404, 358362.
  • Liu, Z., Zhu, F., Wang, G., Xiao, Z., Tang, J., Liu, W., Wang, H., Liu, H., Wang, X., Wu, Y., Cao, Z. & Li, W. (2007) Association study of corticotropin-releasing hormone receptor1 gene polymorphisms and antidepressant response in major depressive disorders. Neurosci Lett 414, 155158.
  • Long, J.R., Zhao, L.J., Liu, P.Y., Lu, Y., Dvornyk, V., Shen, H., Liu, Y.J., Zhang, Y.Y., Xiong, D.H., Xiao, P. & Deng, H.W. (2004) Patterns of linkage disequilibrium and haplotype distribution in disease candidate genes. BMC Genet 5, 11.
  • Mann, J. (2003) The neurobiology of suicidal behavior. Nat Rev Genet 4, 819828.
  • Mann, J.J. & Currier, D. (2007) A review of prospective studies of biological predictors of suicidal behavior in mood disorders. Arch Suicide Res 11, 316.
  • Mann, J.J., Currier, D., Stanley, B., Oquendo, M.A., Amsel, L.V. & Ellis, S.P. (2006) Can biological tests assist prediction of suicide in mood disorders? Int J Neuropsychopharmacol 9, 465474.
  • Marenberg, M.E., Risch, N., Berkman, L.F., Floderus, B. & De Faire, U. (1994) Genetic susceptibility to death from coronary heart disease in a study of twins. N Engl J Med 330, 10411046.
  • Nemeroff, C.B. & Vale, W.W. (2005) The neurobiology of depression: inroads to treatment and new drug discovery. J Clin Psychiatry 66, 513.
  • Nyholt, D. (2004) A simple correction for multiple testing for SNPs in linkage disequilibrium with each other. Am J Hum Genet 74, 765769.
  • Papiol, S., Arias, B., Gasto, C., Gutierrez, B., Catalan, R. & Fananas, L. (2007) Genetic variability at HPA axis in major depression and clinical response to antidepressant treatment. J Affect Disord 104, 8390.
  • Parmley, J.L. & Hurst, L.D. (2007) How do synonymous mutations affect fitness? Bioessays 29, 515519.
  • Reynisdottir, I., Thorleifsson, G., Benediktsson, R. et al. (2003) Localization of a susceptibility gene for type 2 diabetes to chromosome 5q34-q35.2. Am J Hum Genet 73, 323335.
  • Rhodes, M.E. & Rubin, R.T. (1999) Functional sex differences (‘sexual diergism’) of central nervous system cholinergic systems, vasopressin, and hypothalamic-pituitary-adrenal axis activity in mammals: a selective review. Brain Res Brain Res Rev 30, 135152.
  • Robins, L.N., Wing, J., Wittchen, H.U. et al. (1988) The Composite International Diagnostic Interview. An epidemiologic Instrument suitable for use in conjunction with different diagnostic systems and in different cultures. Arch Gen Psychiatry 45, 10691077.
  • Rujescu, D., Thalmeier, A., Moller, H.J., Bronisch, T. & Giegling, I. (2007) Molecular genetic findings in suicidal behavior: what is beyond the serotonergic system? Arch Suicide Res 11, 1740.
  • Scriver, C.R. (2007) The PAH gene, phenylketonuria, and a paradigm shift. Hum Mutat 28, 831845.
  • Stephens, M., Smith, N.J. & Donnelly, P. (2001) A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68, 978989.
  • Swaab, D.F., Bao, A.M. & Lucasson, P.J. (2005) The stress system in the human brain in depression and neurogeneration. Ageing Res Rev 4, 141194.
  • Tiret, L., Poirier, O., Nicaud, V., Barbaux, S., Herrmann, S.M., Perret, C., Raoux, S., Francomme, C., Lebard, G., Trégouët, D. & Cambien, F. (2002) Heterogeneity of linkage disequilibrium in human genes has implications for association studies of common diseases. Hum Mol Genet 11, 419429.
  • Treutlein, J., Kissling, C., Frank, J., Wiemann, S., Dong, L., Depner, M., Saam, C., Lascorz, J., Soyka, M., Preuss, U.W., Rujescu, D., Skowronek, M.H., Rietschel, M., Spanagel, R., Heinz, A., Laucht, M., Mann, K. & Schumann, G. (2006) Genetic association of the human corticotropin releasing hormone receptor 1 (CRHR1) with binge drinking and alcohol intake patterns in two independent samples. Mol Psychiatry 11, 594602.
  • Voracek, M. & Loibl, L.M. (2007) Genetics of suicide: a systematic review of twin studies. Wien Klin Wochenschr 119, 463475.
  • Wasserman, D. (2001) Suicide – An Unnecessary Death. Martin Dunitz Publishers, Taylor & Francis Group, London.
  • Wasserman, D. (2006) Depression: The Facts. Oxford University Press, New York, NY.
  • Wasserman, D., Geijer, T., Rozanov, V. & Wasserman, J. (2005) Suicide attempt and basic mechanisms in neural conduction: relationships to the SCN8A and VAMP4 genes. Am J Med Genet 133B, 116119.
  • Wasserman, D., Geijer, T., Sokolowski, M., Rozanov, V. & Wasserman, J. (2006a) Genetic variation in the hypothalamic pituitary adrenocortical (HPA) axis regulatory factor, T-box19, and the angry/hostility personality trait. Genes Brain Behav 6, 321328.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

The authors wish to thank all interviewers at the Human Ecological Health organization/Odessa, National Mechnikov University, Odessa, Ukraine, for co-ordination of the material collection in Ukraine; Dr Vladymyr Bogatov and laboratory technician Lars Holmberg for technical assistance and Dr Tatyana Reytarova for logistic assistance. The authors wish to thank all those who have given their consent to participate as research subjects in the present study. D.W. (P.I.), J.W. and M.S. contributed to study concept and design, acquisition of the data, analysis and interpretation of the data and critical revision of the manuscript. D.W. obtained funding. M.S. performed the statistical analysis and drafted the manuscript. V.R. contributed to acquisition of the data and administrative support. The study was funded by the Marianne and Marcus Wallenberg Foundation.