The angiotensin-converting enzyme (ACE) catalyses the conversion of angiotensin I to physiologically active angiotensin II. The angiotensin system is implicated in the control of blood pressure as well as cardiovascular and respiratory functions, and exerts central effects within the brain.1 Recently, a functional role of polymorphisms in the ACE gene was reported in unipolar depression.2 The implication of ACE in anxiety disorders is unclear so far, but its close relation to cardiovascular and respiratory control may suggest a role in the etiology of panic attacks. Three previous genetic association studies focused on the insertion/deletion polymorphism in the ACE gene in patients with panic disorder. A Japanese study failed to detect an effect, whereas two European studies reported a significant association of the less active I allele with panic disorder mainly in the male subgroup of panic patients.3, 4, 5
To further elucidate the role of the ACE in the genetic vulnerability for panic attacks, we first genotyped the 24 single nucleotide polymorphisms (SNPs) covering the ACE gene region (±10 kb) reported in Baghai et al. (2006), in 175 patients with syndromal panic attacks (including panic disorder with and without agoraphobia 87.7%, social and special phobia 13.3%), who were recruited from our Anxiety Disorders Outpatient Clinic, versus 445 healthy controls matched for ethnicity, age and sex (sample 1, Max Planck Institute for Psychiatry (MPI)). Patients and controls underwent a thorough clinical and laboratory examination. The controls were screened for the presence of anxiety and affective disorders using the Composite International Diagnostic Screener6 and excluded in the case of any indication for one of these disorders. The diagnosis in patients was ascertained by trained psychiatrists applying Diagnostic and Statistical Manual of Mental Disorders Version IV (DSM-IV) criteria and the Structured Clinical Interview for DSM-IV (severe combined immuno-deficiency I and II).6 The replication of significant SNPs was performed in a second German sample from the University of Münster, consisting of patients with panic disorder with and without agoraphobia (n=169) and blood donors as controls (n=334) matched for gender and age (sample 2, Münster). Genotyping was performed as described in Baghai et al.2
In the discovery sample, we found nominal associations in five intronic SNPs: rs4311 (P=0.001), rs4329 (P=0.02), rs4333 (P=0.01), rs4344 (P=0.01) and rs4353 (P=0.04). rs4291, the promoter polymorphism found to be associated with unipolar depression in two independent samples, did not show a significant association with panic disorder (P=0.08). All five SNPs were in linkage disequilibrium (LD) and located in the 3′ region of the second LD block within the ACE gene, identified as described in Baghai and co-workers, which spans the promoter region as well as all coding regions for ACE isoforms 1 and 2.2, 7 Only for rs4311, the P-value was still significant after Bonferroni correction for multiple testing (Table 1). All five SNPs entered the replication and significant P-values after Bonferroni correction were found for rs4311 and rs4333 (Table 1); rs4329 (P=0.04), rs4344 (P=0.01), rs4353 (P=0.04) were nominally associated in sample 2.
Although the association replicated in both samples, the direction of this association was different. The T allele of rs4311 and 4333 was the risk allele in sample 1, while we observed a heterozygous disadvantage for both SNPs in sample 2. The reason for true but diverse, in extreme reversed associations, known as the flip-flop phenomenon, are discussed in a recent publication by Lin et al.8 Flip-flop associations may indicate heterogeneous effects of the same true variant due to differences in the LD structure between the genotyped SNPs and the disease susceptibility locus in two investigated samples.8 Following this idea, the reported SNPs could be non-causal but in LD with a genuine causal variant and/or multilocus effects that we are not able to account for, and this may lead to different associations. Interestingly, these SNPs are located in the same LD block as the promoter variant associated with unipolar depression, but are clustered at the 3′ end of this block. While the variant rs4291 associated with unipolar depression was not significantly associated with panic disorders, all of the five SNPs associated with panic disorder were also nominally associated with unipolar depression.2 The ACE gene may thus be an example for a genetic or locus overlap between anxiety and depression.
Taken together, we have identified an association of two SNPs (rs4311 and rs4333) in the ACE gene with syndromal panic attacks, which could be replicated in an independent case/control sample, although in a different direction. Further studies have to be performed to elucidate the functional correlates of this association, especially regarding ACE activity in vivo in patients with panic attacks.
About this article
American Journal of Medical Genetics Part B: Neuropsychiatric Genetics (2012)