Introduction

Although major depressive disorder (MDD) is a main public health concern [1], our treatments mostly rely on symptomatic improvement [2]. Recurrences and early age at onset in patients with a positive family history highly suggest the role of genetics in this illness [3]. Most genetic studies have considered a set of functional polymorphisms relevant to monoaminergic neurotransmission [4].

Observations of increased central nervous system (CNS) renin angiotensin aldosterone system (RAS) function in depression suggest the role of genetic variations of RAS components in pathophysiology and response to treatment of depressed patients. Angiotensin-converting enzyme (ACE) is abundantly found in CNS, and earlier reports have indicated that brain angiotensin II (Ang II), the main product of ACE, facilitates norepinephrine release both in vivo and in vitro [5–7]. Furthermore, it has been reported that environmental stressors elevate circulating and brain levels of renin and Ang II [8]. Saab et al. accordingly have also suggested that increased brain RAS activity elevates the risk of depression via increased Ang II receptor type 1 (AT1) responsiveness to Ang II [9]. ACE is responsible for decomposition of some neuropeptides as well including substance P (SP) [10]. Considering the potential antidepressant properties of SP antagonists, the role of SP along with ACE on pathophysiology as well as mitigation of depression has been hypothesized [11]. Mood-elevating effects of ACE inhibitors [12–14] along with the elevated serum ACE levels in MDD sufferers [15, 16] and all presented evidence divulge an important implication of brain RAS in aetiology and treatment of depression [17].

Genetic polymorphisms influencing RAS function can therefore be among the important determinants of response to therapy in MDD patients. Plasma activity of ACE is mainly governed by an insertion/deletion (I/D) polymorphism in its gene [18]. Other known ACE gene variants affecting serum ACE levels are rs4343 and rs4291 [19]. Since D allele of ACE I/D, G allele of rs4343 and T allele of rs4291 polymorphisms are reported to be associated with higher ACE levels [15, 16, 19], inheritance of them may alter therapeutic efficacy of antidepressant medications as is hypothesized here [20, 21].

Selective serotonin reuptake inhibitors (SSRIs) as effective antidepressant drugs with a vast use in psychiatric disorders are believed to be effective in brain RAS function as well which corroborate their non-serotonin dependant mechanism as a therapeutic option for MDD [22–24]. There is not much evidence as to what extent each formulation in this class benefits best in clinical situations, and robust criteria to select an SSRI for clinical use in depressive disorders do not exist. It mainly depends on the physicians’ discretion, perceived adverse effects, ease of use and drug cost to select one among them. Available data suggest there are critical differences in drug pharmacodynamics besides pharmacokinetic diversities in this group which is responsible for the differential clinical effectiveness. Clinical observations are suggestive of significantly different responses and especially greater cognitive improvement and anxiety relief with sertraline comparing with fluoxetine [25, 26], the two widely used SSRIs, in Iranian population. Fluoxetine is the least selective serotonin reuptake inhibitor with norepinephrine and dopamine reuptake inhibition as well whereas sertraline is the second most potent inhibitor of serotonin reuptake and the second most selective blocker of serotonin over noradrenaline uptake. It is also a more potent dopamine uptake inhibitor than other SSRIs [27, 28]. Moreover, such differential effects could also be explained by considering the alternative actions of these drugs, like what has been said about RAS function. Here, we hypothesized that therapeutic response to the two widely used SSRIs, fluoxetine and sertraline, is partly governed by the inheritance of effector RAS genes in patients with MDD. In this study, in line with our previous report on the presence of a strong association between the specific variant of rs4343 ACE gene polymorphism and incidence of MDD [15], we examined the relationship between three genetic variants (ACE I/D, rs4343 and rs4291) and response to sertraline versus fluoxetine in depressed Iranian patients for the first time.

Methods

Subjects

This work was conducted in line with The Code of Ethics of the World Medical Association (Declaration of Helsinki) and Uniform Requirements for manuscripts submitted to biomedical journals. The local committee for ethics of medical experiments on human subjects of Shiraz University of Medical Sciences approved the study. Prior to interview, the written consent was attained from the participants. All patients were Caucasian and from the same geographical region.

This is a randomized controlled clinical trial of 200 patients with newly diagnosed MDD defined as a negative previous diagnosis of depression and no history of antidepressant medications use. Patients referring by the family physicians to a university hospital outpatient psychiatry clinic with suggestive symptoms of depression were screened by a psychiatrist, and based on the DSM-V criteria [29], a diagnosis of MDD was made. Initially screened patients were then evaluated to obtain a complete history of medical illnesses as well as drug history. Exclusion criteria were as follows: a family history of schizophrenia; bipolar disorder in the first-degree relatives; a personal history of schizophrenia, manic or hypomanic episodes; mood incongruent psychotic symptoms; active substance dependence or primary organic disease; current treatment with antipsychotics, mood stabilizers, ACE inhibitors or angiotensin receptor blockers; and cardiovascular diseases such as coronary artery disease (CAD), myocardial infarction (MI) or heart failure (HF). Afterwards, patients were randomly divided into two parallel groups receiving either fluoxetine (FLUOXETINE-ABIDI®) or sertraline (SERTRALINE-ABIDI®). Fixed doses of 50 mg titrated in 1 week for sertraline and 20 mg for fluoxetine were used. After 12 weeks of treatment with either drug, follow-up examination to measure the response rate was made by another psychiatrist. Severity of depression was appraised using the 21-item Hamilton Rating Scale for Depression (HAMD-21) before and after the treatment course. The level of response was calculated based on 50 % reduction in the reported scores. At the time of randomization, 5 cc of whole blood was acquired for later DNA extraction. Allocation was based on the simple randomization method of two sets of envelopes with the name of the drugs prescribed inside. All the envelopes were shuffled, and the first psychiatrist sequentially opened the envelopes to determine the treatment group for each patient. Neither the prescribing psychiatrist nor the patients were blinded to the treatment but the second psychiatrist who measured the response rate and the analyst were blinded.

DNA extraction and genotype determination

Genomic DNAs were extracted using a salting out method [30]. A standard protocol was used for PCR amplification/detection of ACE I/D [31, 32]. PCR amplification of rs4291 and rs4343 was performed as previously described [33]. All samples were genotyped at least twice and reconfirmed.

Statistics

Data were analysed using SPSS® 21.0 for windows® (SPSS Inc., Chicago, IL). Hardy–Weinberg equilibrium (HWE) for distribution of genotypes was calculated by chi-square (χ 2) test. Continuous variables are demonstrated as mean ± S.D. Genotype frequencies are shown in percentage (%). Distribution of all continuous variables was tested for normal distribution with the Kolmogorov–Smirnov test. Associations between categorical variables were determined by Pearson’s chi-square or Fisher’s exact test and for interval data by Student’s t test. Univariate analysis of genotypes was performed by χ 2 test. Additionally, Bonferroni’s post hoc test was used. Odds ratio (OR) and 95 % confidence intervals (CIs) were obtained. P value <0.05 was considered as statistically significant.

Results

Demographic data of the patients are presented in Table 1. In sertraline treated arm, there were 75 women and 25 men with a mean age of 35.4 ± 12.8 (18–65 years); in the fluoxetine treated arm, there were 69 women and 31 men with a mean age of 33.4 ± 11.3 (18–65 years). Body mass indexes were 24.5 ± 12.8 and 24.8 ± 3.7 for the two groups, respectively. The distribution of study genotypes was in agreement with the Hardy–Weinberg equilibrium. Table 2 shows genotype and allele frequencies of treatment-responsive patients. Regarding ACE I/D polymorphism, DD genotype was strongly associated with response to sertraline comparing with fluoxetine (P = 0.006; OR = 3.7; 95 % CI = 1.66–8.29). D allele of ACE I/D was also significantly higher in patients responding to sertraline (P = 0.0006; OR = 3.0; 95 % CI = 1.80–5.08). Concerning rs4291 polymorphism, frequency of TT genotype was higher in patients responding to sertraline (P = 0.256; OR = 2.8; 95 % CI = 1.04–7.68) although not statistically significant. Concerning rs4343 variant, frequency of GG genotype was also higher in patients responding to sertraline (P = 0.162; OR = 2.63; 95 % CI = 1.09–6.37) although not significantly.

Table 1 Demographic characteristics of patients responsive to sertraline and fluoxetine
Full size table
Table 2 Genotype and allele frequencies in patients responsive to sertraline versus fluoxetine
Full size table

Discussion

To the best our knowledge, this is the first study to show that D allele of ACE gene is associated with better antidepressant response to sertraline comparing to fluoxetine. It signifies the role of genetic factors in response to treatment and will potentially serve as a measure to define criteria for personalized medication. Here, we found out that MDD patients with DD genotype responded approximately four times (OR = 3.7, 95 % CI = 1.66–8.29, P = 0.006) better to sertraline compared to fluoxetine. As this was a clinical trial, we can corroborate the role of the D allele for differential response to SSRIs. Randomized allocation of treatment between the two groups minimizes the chance for unrecognized confounders to be effective and at the same time the magnitude of the observed response difference as well as group matching of the two treatment arms as is shown in Table 1, minimize inaccuracies of our inferences. Moreover, enrolling only patients with newly diagnosed MDD to extinguish the effects of parallel medications and excluding those with concomitant psychiatric disorders or other medical conditions contributed to the demonstration of a more precise cause and effect relationship in this study.

According to overwhelming evidence that suggests genetic factors play a pivotal role in susceptibility to depressive disorders, it is also discerned that therapeutic outcomes are influenced as a consequence. The widely used SSRIs show differences in efficacy because of not only their pharmacological differences but also the influence of patients’ genes. Sertraline is the most potent inhibitor of serotonin reuptake and compared to fluoxetine, it has stronger potency for dopamine and norepinephrine reuptake inhibition [27, 28]. Evidence supports the interactions between brain Ang II and catecholamine systems, and this may explain the complex interplay of various effectors in clinical situations like MDD [6, 34, 35]. Stronger interaction of sertraline with neurotransmitters especially dopamine and serotonin may reflect a more effective interaction with RAS as well. Targeting brain RAS, with its documented functions in memory and learning, has been shown to ameliorate stress, anxiety, brain inflammation and ischemia [36]. Taking the CNS abundance of ACE and the diversity of RAS function in depressive disorders, the hypothesis that genetic determinants of RAS function are among the determinants of differential therapeutic response to SSRIs has been emerged in this study.

Carriers of D allele responded better to sertraline by 3.0-fold (OR = 3.0; 95 % CI = 1.80–5.08) comparing to fluoxetine. Similarly, carriers of GG genotype of rs4343 and TT genotype of rs4291 were more responsive to sertraline [P = 0.162, 17 (30.9 %) vs. 10 (14.1 %) and P = 0.256, 13 (23.6 %) vs. 7 (9.9 %), respectively] although not significantly. These variants (DD, GG and TT genotypes of ACE) were associated with higher serum ACE activity [18, 19] and neuropeptide degradation [37] which can explain an overall higher therapeutic efficacy in the carriers. Furthermore, it has been suggested that ACE I/D polymorphism exerts a noticeable impact on hypothalamic–pituitary adrenal axis hyperactivity in depressed patients [38] which in turn suggests another possible mode of action for medications affecting RAS function in D allele carriers. Supporting evidence for association between the D/D genotype and affective disorders is reported from Japanese and German populations [16, 37] and underlines the pathophysiologic distinction of MDD in patients with this genotype which in turn influence drug responsiveness. We have previously reported that GG variant of rs4343 was strongly associated with depression and also higher ACE levels in a group of depressed patients [15]. Our findings of higher frequencies of GG genotype in patients responsive to sertraline may suggest that this variant may also be effective in treatment efficacy. Likewise, higher frequencies of TT genotype in patients responsive to sertraline propose the same idea. Higher ACE activity in these gene variants is probably responsible for better response to sertraline compared to fluoxetine as well. Patients carrying the mentioned variants experienced more advanced disease with higher initial scores, and sertraline could exert more potent therapeutic effects. On the other hand, RAS over activity as a contributor to or a sign for the severity of MDD is a common finding in carriers of these gene variants. The Ang II dependant mechanisms involved in pathophysiology of depression could be best influenced by sertraline in those with genotypes associated with higher ACE activity then. A better response to sertraline observed in this study among these genetic variant carriers might also offer clues to an interplay between brain RAS and serotonin/dopamine-related mechanisms associated with MDD.

It is to note that one limitation of this study is that the patients were aware of the drug allocated to them. This resulted in the inability to guarantee that the response evaluator was fully blinded to the prescribed medication. Although all the necessary commitments were made to eliminate the response evaluator’s preoccupation affecting the final judgement, theoretically, personal expectations could have an impact on the outcome variables. The fixed doses of sertraline and fluoxetine used in this study are the recommended doses in clinical situations, and thus, the observed differences could have only been minimally affected by the prescribed doses. Another limitation of this study was that the majority of enrolees were women, and further study is warranted to approve such differential effects in men as well. However, the proportions of women in both treatment arms were nearly identical which diminishes the effect of sex on observed differences.

We hope that these findings would help in smarter selection of SSRIs for use in MDD and contribute to personalized medication to more properly target specific disease mechanisms that would result in a better chance for complete remission as well as reducing costs.

Conclusion

Genetic factors are among the basic determinants of altered response to therapy, and here, we showed that carrying D allele of ACE gene, as well as DD genotype of ACE I/D polymorphism, yield to a better response to sertraline compared to fluoxetine.