The serotonin transporter (SERT, encoded by the SLC6A4 gene) regulates serotonin (5-HT) neurotransmission by transporting the neurotransmitter 5-HT from synaptic cleft to presynaptic neurons, and it is the main target of SSRI (selective serotonin reuptake inhibitors) antidepressants.

The most investigated polymorphisms within this gene were the 5-HTTLPR (a 44 bp insertion/deletion), the single nucleotide polymorphism (SNP) rs25531, that are both located in the promoter, and a 17 bp VNTR (variable number of tandem repeats) within intron 2 (STin2).

The 5-HTTLPR 16-repeat sequence is called long allele (L) and it shows a twice basal SERT expression compared to the 11-repeat allele (short allele or S) [14]. The S allele was associated with several psychiatric disorders with affective symptomatology and personality traits related to anxiety, impulsivity, and stress, and with poorer antidepressant response, especially in patients of Caucasian ancestry treated with SSRIs [32]. The S allele was also hypothesized to be a risk factor for SSRI-induced side effects mainly in Caucasian populations, with the exception of sexual side effects [14].

The rs25531 SNP was reported to lay within the 5-HTTLPR sequence and influence the functional effect of 5-HTTLPR itself. Indeed, the rs25531 G variant in conjunction with the L allele (L_G) may result in a reduced expression of SLC6A4, equivalent to that conferred by the S allele [14]. Consistently, single-photon emission computed tomography imaging suggested that L_A/L_A carriers may have a more dynamic serotonergic system that seems to confer higher probability of response to SSRIs. Anyway, pharmacogenetic studies showed substantially negative results [14]. The L_A allele was preliminarily associated with lower SSRI-induced side effects but higher sexual dysfunction in subjects of Caucasian ancestry [17, 22].

The STin2 VNTR comprises 9, 10, or 12 copies of a 16–17 bp repeat and may influence gene transcription with a synergistic effect with 5-HTTLPR. Indeed, the 12-repeat variant was shown to cause higher gene expression in vitro and in vivo [14]. Pharmacogenetic findings were mainly negative, while positive results suggested better response in long allele carriers in Asian populations and better response in short allele carriers in Caucasians [14, 25].

In the field of antidepressant pharmacogenetics, the 5-HT1A (encoded by the HTR1A gene) and 5-HT2A (encoded by the HTR2A gene) serotonin receptor subtypes have been the most investigated.

5-HT1A receptor is abundant in corticolimbic regions, and it could be expressed both pre- and postsynaptically. At the level of serotonin cell bodies in the midbrain, dorsal raphe nucleus acts as an autoreceptor, inhibiting the firing of serotonin neurons and reducing the release of 5-HT in the prefrontal cortex. Antidepressants desensitize these inhibitory autoreceptors and this may be responsible for the delay in antidepressant action onset [46]. On the other hand, reductions in postsynaptic 5-HT1A receptors in prefrontal and temporal cortical regions were demonstrated both in depressive and anxiety disorders [35].

In the HTR1A gene, the most investigated SNP is the rs6295 (or 1019C/G, in the upstream regulatory region of HTR1A), which G allele results in an upregulation of the gene. Thus, rs6295 G allele is expected to contrast the enhancement of serotoninergic transmission through a higher number of presynaptic inhibitory 5-HT1A receptors, resulting in poorer antidepressant response. Nevertheless, pharmacogenetic findings were often inconsistent, and the available meta-analyses suggested no effect of the polymorphism [28]. Different stratification factors were hypothesized to explain these inconsistent findings, in detail: particular subtypes of MDD, gender, or gene x gene interactions [14].

The 5-HT2A receptor is a G-coupled postsynaptic receptor with widespread distribution throughout the cortex, with high densities in the frontal cortex. rs6311 (or -1438A/G) and rs6313 (or 102C/T) are functional HTR2A SNPs in linkage disequilibrium (LD) that have been particularly studied. The rs6313 SNP (within HTR2A exon 1) per se has no likely effect on antidepressant response [28], but gene x gene interactions were suggested to modulate this phenotype or a selective effect on SSRI response was reported [14]. rs6311 is located in the promoter of the gene and the A allele has been associated with increased promoter function [14]. A weak association between the AA genotype and nonresponse in subjects of Asian ancestry was demonstrated [28], but the estimated OR was of only 1.66, with 95 % confidence interval lower limit very close to 1; thus, the detrimental effect of the AA genotype is not more than negligible alone. Anyway, gene x gene interactions involving this locus should not be excluded [14], and additive or multiplicative effects with SNPs within HTR2A or other genes may play relevant effects. Interestingly, the region from the downstream to the first intron of the gene was found to harbor other possibly relevant polymorphisms in the context of antidepressant efficacy. Indeed, rs7333412, rs7324017, rs1923882 [15], and rs7997012 [29] are located in this region and were associated with antidepressant response in a large sample (the Sequenced Treatment Alternatives to Relieve Depression or STAR*D).

The key enzymes involved in the metabolism of monoamines play a role in the regulation of their balance in the CNS.

In regard to serotonin biosynthesis, the limiting step is catalyzed by tryptophan hydroxylase (TPH), which is codified by two distinct genes, TPH1 and TPH2. TPH1 is ubiquitarious but predominantly expressed in peripheral organs, while TPH2 is more selectively expressed in the brain. Anyway, TPH1 and TPH2 are actually expressed at similar levels in some brain areas (e.g., frontal cortex, thalamus, hippocampus, and amygdala), and TPH1 may be selectively expressed in particular circumstances (e.g., stress) [14]. The available pharmacogenetic data are mainly referred to TPH1 rs1800532 (or A218C), because this SNP is located in a potential GATA transcription factor-binding site. The rarer A allele is associated with decreased 5-HT synthesis, and according to the monoaminergic theory of MDD, it may determine worse antidepressant efficacy. The hypothesis was confirmed by some pharmacogenetic studies investigating SSRI response, but the greatest part of them failed to replicate the result [14, 28]. A selective effect of rs1800532 on antidepressant response in specific MDD subtypes (with psychotic and melancholic features) has been recently hypothesized [2], suggesting the usefulness of investigating the SNP in these subgroups of patients. No support to the association between SSRI- and SNRI-induced side effects and rs1800532 was provided [14], a possible effect on antidepressant-induced body weight gain apart [37].

MAO (monoamine oxidase) and COMT (catechol-O-methyltransferase) code for the main enzymes involved in the catabolism of monoamines.

In humans, two distinct MAO isoforms are expressed: MAOA, which is the most investigated one in psychiatry and mainly breaks serotonin, norepinephrine, and epinephrine, and MAOB, which is mainly investigated concerning Parkinson’s disease and mainly breaks phenethylamine and benzylamine. A 30-bp VNTR, located 1.2 kb upstream the MAOA coding sequence, was reported to influence the transcription rate of the gene, since alleles with 3.5 or 4 copies of the repeat sequence are transcribed 2–10 times more efficiently than those with 3 or 5 copies of the repeat. Carriers of long alleles were reported to show higher amygdala reactivity in response to aversive stimuli and increase functional coupling of a neural pathway between the ventromedial prefrontal cortex and the amygdala, which was associated to higher levels of harm avoidance, a temperamental dimension related to MDD. Long alleles have been associated with both higher risk of MDD and poorer antidepressant efficacy [30], with a possible selective effect in females [14], consistently with the MAOA position on the X chromosome. Anyway, no all the available studies found an effect of the polymorphism on antidepressant efficacy [14].

The COMT gene is also hypothesized to play a role in MDD pathophysiology and antidepressant response. The COMT Val108/158Met (rs4680) polymorphism shows a relevant functional effect, since the Val/Val genotype catabolizes dopamine at up to four times the rate of Met/Met homozygote, resulting in a significant reduction of synaptic dopamine following neurotransmitter release. Available pharmacogenetic studies mainly reported the Met variant as the favorable allele for antidepressant response, with an allele dose effect (better outcome in Met/Met carriers and intermediate outcome in Met/Val carriers), supporting the monoaminergic theory of MDD. Recently, a better covering of COMT variability was provided, confirming the hypothesis of an effect of this gene on antidepressant efficacy [14].

The most promising pharmacogenetic findings were obtained for interleukin 1β (IL-1β) and interleukin 6 (IL-6). They both code for pro-inflammatory cytokines whose peripheral blood levels were found increased in MDD [16] and inversely correlated with antidepressant response [21]. These cytokines are able to act also within the CNS and affect neuronal death and hippocampal volume [16].

The most interesting polymorphisms within the IL-1β gene are rs16944 and rs1143643. Indeed, functional magnetic resonance imaging showed that the number of G alleles in both rs16944 and rs1143643 was associated with reduced responsiveness of the amygdala and anterior cingulate cortex (ACC) to emotional stimulation [3]. The G allele of both polymorphisms was associated with antidepressant nonresponse [14].

The −174 SNP (rs1800795) within the IL-6 gene is particularly interesting since individuals who carry the G allele have higher plasma concentrations of IL-6 [50], and the polymorphism has been studied as a modulator of interferon (INF)-induced depression [11, 42]. Nevertheless, no data are available about the role of this SNP in antidepressant pharmacological treatment, while the rs7801617 provided a suggestive signal for association with antidepressant response in a GWAS [43].

Corticotropin-releasing hormone (CRH) receptors 1 and 2 (coded by CRHR1 and CRHR2 genes) are the mediators of the effect of glucocorticoids in the CNS. CRHR1 polymorphisms were hypothesized to modulate antidepressant response particularly in anxious depression as well as in generalized anxiety disorder, while negative results were provided by some studies that did not consider anxiety levels [14]. High anxiety-related behavior mice show altered expression of the CRHR1 gene in the pituitary and prefrontal cortex [39]. Furthermore, CRHR1 system is implicated in the programming effects of early life stress on eventual anxious-depressive psychopathology [45]. Few and non-replicated findings are available for the CRHR2 gene as well as for the NR3C1 gene (coding for the glucocorticoid receptor or GR) [14].