Interplay between Serotonin 5‐HT _1A and 5‐HT ₇ Receptors in Depressive Disorders

Summary

Serotonin (5‐hydroxytryptamine or 5‐HT) is an important neurotransmitter regulating a wide range of physiological and pathological functions via activation of heterogeneously expressed 5‐HT receptors. Besides the important role of 5‐HT receptors in the pathogenesis of depressive disorders and in their clinical medications, underlying mechanisms are far from being completely understood. This review focuses on possible cross talk between two serotonin receptors, 5‐HT _1A and the 5‐HT ₇. Although these receptors are highly co‐expressed in brain regions implicated in depression, and most agonists developed for the 5‐HT _1A or 5‐HT ₇ receptors have cross‐reactivity, their functional interaction has not been yet established. It has been recently shown that 5‐HT _1A and 5‐HT ₇ receptors form homo‐ and heterodimers both in vitro and in vivo. From the functional point of view, heterodimerization has been shown to play an important role in regulation of receptor‐mediated signaling and internalization, suggesting the implication of heterodimerization in the development and maintenance of depression. Interaction between these receptors is also of clinical interest, because both receptors represent an important pharmacological target for the treatment of depression and anxiety.

Introduction

The brain serotonergic system is known to be involved in the control of a wide range of physiological functions as well as of different kinds of behavior. Such polyfunctionality of serotonin (5‐hydroxytryptamine or 5‐HT) is mediated by the existence of large number of serotonin receptors. Currently, 14 different 5‐HT receptor subtypes expressed in the mammals have been identified. To provide more detailed characterization of 5‐HT receptors, three main criteria are normally used: (1) structural features, including primary amino acid sequence, (2) signal transduction characteristics, including specific signal pathways, and (3) pharmacological profile 1. Based on these criteria, 5‐HT receptors were classified in seven subfamilies, which include both G protein‐coupled and ionotropic receptors 2, 3.

Despite the great clinical interest and large body of investigations dedicated to 5‐HT receptors, the mechanism regulating 5‐HT receptor functions are far from being completely understood. This review focuses on two serotonin receptors: the 5‐HT_1A receptor representing a key player in the brain 5‐HT system 4, and the 5‐HT₇ receptor. The 5‐HT_1A receptor is known to be critically involved in major depression, anxiety, and suicide. More recently, functional studies have also implicated 5‐HT₇ receptors in depression 5, 6, 7. As 5‐HT_1A and 5‐HT₇ receptors are co‐expressed in majority of the brain structures and modulate the same second messenger systems (even though in an opposite way), one intriguing question is the possible functional cross talk between these receptors. Since both receptors represent a pharmacological target for the treatment of depression, receptor–receptor interaction is also of great clinical significance.

5‐HT_1A and 5‐HT₇ Receptors: Genes, Signaling Pathways, and Distribution in the Brain

The 5‐HT_1A receptor is one of the most extensively characterized members of the serotonin receptor family. Increased interest to the 5‐HT_1A receptor is based on several reasons: (1) this receptor plays an important role in the regulation of neuronal development and plasticity 8, 9, (2) presynaptic 5‐HT_1A autoreceptors are involved in the regulation of functions of 5‐HT neurons 2, (3) numerous data demonstrate implication of 5‐HT_1A receptor in the control of various physiological functions 4, and (4) 5‐HT_1A receptors are involved in the mechanisms of depression, anxiety, suicide, and schizophrenia 10, 11, 12, 13, 14.

The 5‐HT_1A receptor gene was cloned in 1987 15. One year later, it has been demonstrated that the protein product of this genomic clone possesses all characteristics of the 5‐HT_1A receptor 16. In following studies, the 5‐HT_1A receptor gene was cloned from rat 17 and mouse 18. It was shown that 5‐HT_1A receptor gene is localized at 5th chromosome in human 15 and at 13th chromosome in mice 19, and does not contain introns in its coding sequence 17. Mouse 5‐HT_1A receptor gene demonstrates 94% homology with rat and 88% homology with human 5‐HT_1A receptor gene (Ensemble Genome Browser, http://www.ensembl.org/index.html). The detailed structure of 5‐HT_1A receptor gene promoter, including position of selective and nonselective enhancer as well as selective silencer, was also determined 20, 21. In addition, several rare single‐nucleotide polymorphisms (SNPs) were described for 5‐HT_1A receptor gene 22. Between those, the SNP C(‐1019)G localized in the repressor region of promoter is of particular interest because of its association with depression, suicidal behavior, and schizophrenia 13 as well as sensitivity to the antidepressant drug treatment 23.

The 5‐HT_1A receptor is known to activate a variety of effectors via G_i/o proteins, and receptor stimulation leads to inhibition of adenylyl cyclase (AC). The 5‐HT_1A receptor is also involved in G_βγ‐mediated activation of a K⁺ current, inhibition of a Ca²⁺ current, and stimulation of the phospholipase C, as well as an activation of the mitogen‐activated protein kinase Erk2 2, 24. One important mechanism regulating 5‐HT_1A receptor function is receptor desensitization and internalization 10. Interestingly, that 5‐HT_1A receptor activation leads to effective internalization of 5‐HT_1A autoreceptors in the nucleus raphe dorsalis, but not of heteroreceptors resided in hippocampal neurons 25, indicating the difference in the regulation of pre‐ and postsynaptic 5‐HT_1A receptors activity.

5‐HT_1A receptors are widely distributed within the brain. In rodents, high receptor density was detected in limbic areas; in hippocampus, septum, and cortical areas; and in the midbrain dorsal and median raphe nuclei 2, 26. Similar distribution was reported in human by postmortem autoradiography 27 as well as in the living human brain using positron emission tomography (PET) 28. Several experimental data suggested correlation between depressive disorders and changes in the distribution of 5‐HT_1A receptor in the brain, although results of such studies are often contradictory. For example, an increase in 5‐HT_1A autoreceptors was reported in postmortem brain from depressed suicide victims 29, 30. In contrast, PET analysis in depressed (submissive) monkeys revealed a reduction in 5‐HT_1A receptor binding throughout the all brain areas 31. At the same time, PET analysis of 5‐HT_1A receptor binding in the brain of depressed patients before and after electroconvulsive therapy demonstrated a strong reduction in 5‐HT_1A receptor binding in regions consistently reported to be altered in major depression 32.

The 5‐HT₇ receptor is one of the most recently described members of the 5‐HT receptor family 33, 34, 35. This receptor is coupled to G_s protein and activates AC. In addition, the 5‐HT₇ receptor is coupled to the G₁₂ protein to activate small GTPases of the Rho family (i.e., Cdc42 and RhoA), leading to enhanced neurite outgrowth, synaptogenesis, and neuronal excitability 36, 37, 38. It has been also demonstrated in cell lines that the 5‐HT₇ receptor can stimulate intracellular calcium release 39.

The gene encoding the 5‐HT₇ receptor is localized at 19th chromosome in mouse and at 7th chromosome in human. Comparison of amino acid sequence of the transmembrane regions of 5‐HT₇ receptor from different species showed homology between 39% and 50%. Interestingly, the homology of transmembrane domains of 5‐HT₇ receptor with those for 5‐HT_1A receptor gene is relatively high (about 40%), which can partly explain the fact that most agonists developed for the 5‐HT_1A and/or 5‐HT₇ receptors have cross‐reactivity between these receptors 40 (see below). The human 5‐HT₇ receptor gene contains three introns within the coding region 41, 42. Detailed analysis of human 5‐HT₇ receptor gene revealed existence of eight SNPs. Noteworthy that two of these polymorphisms (rs3808932 and rs12412496) were shown to be associated with schizophrenia 43. A number of 5‐HT₇ receptor splice variants including 5‐HT_7(a), 5‐HT_7(b) 5‐HT_7(c) 5‐HT_7(e) and 5‐HT_7(a), 5‐HT_7(b) and 5‐HT_7(d), differing in the amino acid sequence of their C‐terminus, but possessing similar pharmacological profile were cloned from rat and human, respectively 41, 42, 44, 45. Noteworthy that the human 5‐HT_7(d) receptor isoform possesses a differential pattern of receptor internalization, which can affect receptor‐mediated signaling 46. The 5‐HT₇ receptors are abundantly expressed in brain limbic structures. Significant density of 5‐HT₇ receptor was also observed in hippocampus and in raphe nuclei area. Moderate or relatively low receptor density was obtained in cortex, caudate putamen, and cerebellum 47.

The 5‐HT₇ receptor is associated with a number of physiological and pathological responses, including serotonin‐induced phase shifting of the circadian rhythm 34 and age‐dependent changes of the circadian timing 48, 49. There are some data implicating 5‐HT₇ receptor in the control of memory 50, emotionality, locomotor, and exploratory activity 51, 52. A large body of evidence indicates involvement of 5‐HT₇ receptor in the anxiety and depression, and recent studies suggest that the 5‐HT₇ receptor can be highly relevant for the treatment of major depressive disorders 53.

Pharmacological Properties of 5‐HT_1A and 5‐HT₇ Receptors

Over the years, several agonists and antagonists for 5‐HT_1A and 5‐HT₇ receptors have been developed and applied to investigate the pharmacology of these receptors. Most frequently used molecules along with their binding affinities at 5‐HT_1A and 5‐HT₇ receptors are summarized in Table 1. The 5‐HT_1A receptor was first identified in 1983 pharmacologically as a distinct binding site for 8‐OH‐DPAT 54, which was considered as the selective 5‐HT_1A receptor reference agonist 55. The affinity of 8‐OH‐DPAT for 5‐HT_1A receptors in rat hippocampal membranes resides in the low nanomolar range (K_i = 1.78 nM) 56. 8‐OH‐DPAT demonstrated the properties of a full 5‐HT_1A receptor agonist in the forskolin‐stimulated cAMP assay for both native and cloned 5‐HT_1A receptors 56. However, when the 5‐HT₇ receptor was cloned in 1993, it became apparent that 8‐OH‐DPAT possesses affinity also for this receptor (K_i = 52 nM) 57 and behaves as a partial agonist (E_max = 62% of 5‐CT response; EC₅₀ = 2345 nM).

Table 1.

Binding affinity values of agonists and antagonists acting on 5‐HT _1A or 5‐HT ₇ receptors

Compound	5‐HT1A receptor			5‐HT7 receptor
	Binding affinity	System	Reference	Binding affinity	System	Reference
8‐OH‐DPAT	Ki = 1.78 nM	Rat hippocampal membranes	52	Ki = 52 nM	Rat cloned receptor	53
WAY‐100635	IC50 = 1.35 nM	Rat hippocampal membranes	54	Ki > 10000 nM	Human cloned receptor	59
WAY‐100135	IC50 = 34 nM	Rat hippocampal membranes	55	Not available
NAN‐190	pKi = 9.2	Rat hippocampal membranes	126	Ki > 1000 nM	Rat cloned receptor	60
NAD‐299	Ki = 0.59 nM	Rat hippocampal membranes	58	Ki > 1000 nM	Rat cloned receptor	58
5‐CT	Ki = 0.53 nM	Human cloned receptor	127	IC50 = 0.83 nM	Rat cloned receptor	31
AS‐19	Ki = 89.7 nM	Human cloned receptor	66	Ki = 0.6 nM	Human cloned receptor	66
E‐55888	Ki = 700 nM	Human cloned receptor	66	Ki = 2.5 nM	Human cloned receptor	66
LP‐211	Ki = 15 nM	Human cloned receptor	67	Ki = 379 nM	Human cloned receptor	67
SB‐269970	pKi < 5	Human cloned receptor	128	pKi = 8.9	Human cloned receptor	128
SB‐258719	pKi < 5.1	Human cloned receptor	129	pKi = 7.5	Human cloned receptor	129
SB‐656104	pKi = 6.25	Human cloned receptor	130	pKi = 8.70	Human cloned receptor	130

Several antagonists of 5‐HT_1A receptor became available in early 1990. One of them, WAY‐100635, is the prototypical silent 5‐HT_1A receptor antagonist, which has been widely used as a pharmacological probe to investigate the distribution and function of 5‐HT_1A receptors 58. WAY‐100635 displaces specific [³H]8‐OH‐DPAT binding to 5‐HT_1A receptors in the rat hippocampus with IC₅₀ = 1.35 nM, and it is more then 100‐fold selective for the 5‐HT_1A site in comparison with a range of other 5‐HT receptors. However, WAY‐100635 also display affinities for 5‐HT_2B (K_i = 24 nM), dopamine D₂ (K_i = 16.4 nM), and adrenergic α _1A (K_i = 19.9 nM) receptors. In addition, WAY‐100135 was initially considered as a silent 5‐HT_1A antagonist 59, but it showed partial agonist properties in a subsequent study 60. Noteworthy that WAY‐100635 together with other 5‐HT_1A receptor antagonists NAN‐190 61 and NAD‐299 62 did not reveal any affinity for 5‐HT₇ receptors 62, 63, 64.

In case of the 5‐HT₇ receptors, 5‐CT has been identified as a high‐affinity receptor agonist (IC₅₀ = 0.83 nM, EC₅₀ = 13 nM, 100% maximal efficacy) that is widely used to study the receptor pharmacology and functions 34. However, due to its high affinity for 5‐HT_1A, 5‐HT_1B, and 5‐HT_1D receptors 65, analysis of 5‐HT₇ receptor functions by 5‐CT requires parallel application of WAY‐100635 (5‐HT_1A receptor antagonist) and GR‐127935 (5‐HT_1B/1D receptor antagonists) 66. Interestingly, 8‐OH‐DPAT has also been used to study the functional effects of 5‐HT₇ receptor in combination with application of the 5‐HT_1A antagonists WAY‐100635, NAN‐190, and NAD‐299 67, 68, 69.

During the last decade, various selective 5‐HT₇ receptor agonists became available in addition to 5‐CT. AS‐19 (K_i of 0.6 nM for the human 5‐HT₇ receptor) was reported to display high selectivity for 5‐HT₇ receptor (more then 100‐fold) in comparison with the other 5‐HT receptor subtypes with exception of 5‐HT_1D receptors (only 11‐fold higher selectivity). AS‐19 was found to behave as a potent (EC₅₀ = 9 nM), but partial 5‐HT₇ receptor agonist with a maximal effect reaching 77% of the cAMP response evoked by the 5‐HT 70. Another selective and potent 5‐HT₇ receptor agonist E‐55888 70 shows high affinity for 5‐HT₇ receptors (K_i = 2.5 nM) and very low affinity for 5‐HT_1A (K_i = 700 nM) with no significant affinity for other 5‐HT receptors. E‐55888 behaves as a full agonist with efficacy and potency (E_max = 99%; EC₅₀ = 16 nM) to induce receptor‐mediated cAMP formation in HEK‐293F cells in a range similar to that obtained for 5‐HT 70. Selective 5‐HT₇ receptor agonist LP‐211 displays high affinity for rat and human 5‐HT₇ receptors (K_i = 0.58 and 15.0 nM, respectively) and moderate to low affinity for other 5‐HT receptors, including the 5‐HT_1A receptor 71. LP‐211 possesses agonistic properties in vivo inducing hypothermia in 5‐HT₇ ^+/+, but not in 5‐HT₇ ^−/− mice.

Recently, several specific 5‐HT₇ receptor antagonists were developed. The prototype of 5‐HT₇ receptor antagonist SB‐269970 demonstrates a high binding affinity for human (pK_i = 8.9) and guinea pig (pK_i = 8.3) native 5‐HT₇ receptors. This compound possesses very high selectivity for 5‐HT₇ receptors (more than 100‐fold) in comparison with other G protein‐coupled receptors (GPCRs), except the human 5‐HT_5A receptor for which selectivity was 50‐fold. SB‐269970 acts as a competitive antagonist of the human 5‐HT₇ receptor in AC assay (pA₂ = 8.5) as well as in the guinea pig native tissue assay (pK_B = 8.3). Interestingly, SB‐269970 produces an inhibition of basal AC activity in the absence of any agonist, suggesting an inverse agonism 72. Other selective 5‐HT₇ receptor antagonists are SB‐258719 73, that is less potent than SB‐269970 (pK_i = 7.5; pA₂ = 7.2), and SB‐656104 74. Although the latter has similar pharmacological properties with SB‐269970 (pK_i = 8.7; pA₂ = 8.4), it demonstrates only 10‐fold selectivity over the 5‐HT_1D receptor.

Functional Cross Talk Between 5‐HT_1A and 5‐HT₇ Receptors

There is a set of studies suggesting functional cross talk between 5‐HT₇ and 5‐HT_1A receptors at the level of thermoregulation. Involvement of 5‐HT_1A receptor in the regulation of the body temperature has been demonstrated more than 25 years ago by the observation that 5‐HT_1A receptor agonist 8‐OH‐DPAT produced considerable hypothermic response 75, 76. This effect has been used as a test for 5‐HT_1A receptor functional activity 75, 76, 77, 78. We have also shown that changes in the expression of 5‐HT_1A receptor gene are linked to natural hibernation and associated hypothermia 79.

Based on the absence of 5‐HT‐induced hypothermic response in 5‐HT₇ ^−/− mice, Hedlund and co‐authors suggested that 5‐HT‐induced hypothermia is mainly mediated via 5‐HT₇ receptor 80. However, in this study, 5‐HT was applied peripherally. Because of the poor ability of 5‐HT to cross the blood–brain barrier 81, 82, the importance of the central 5‐HT₇ receptor in thermoregulation suggested in this study should be therefore considered more carefully.

It was shown that at lower doses (0.3–0.6 mg/kg, i.p.), 8‐OH‐DPAT decreased body temperature in 5‐HT₇ ^+/+ mice but not in 5‐HT₇ ^−/− mice. At a higher dose (1 mg/kg, i.p.), 8‐OH‐DPAT induced hypothermia in both 5‐HT₇ ^−/− and 5‐HT₇ ^+/+ mice. At the same time, it was found that pretreatment of mice with a 5‐HT_1A receptor selective antagonist WAY‐100135 completely blocked the hypothermic response of 8‐OH‐DPAT in 5‐HT₇ ^+/+ mice 83. Moreover, we have shown that intracerebroventricular, but not intraperitoneal administration of selective 5‐HT₇ receptor agonist LP44 produced considerable hypothermic response in mice, indicating involvement of central rather then peripheral 5‐HT₇ receptors in thermoregulation. The comparison of 5‐HT₇ and 5‐HT_1A receptor‐induced hypothermia in eight mouse strains did not reveal any correlation, suggesting the low functional interaction between 5‐HT₇ and 5‐HT_1A receptors in the control of the body temperature. Together with the observation that the 5‐HT₇ receptor selective antagonist did not affect 8‐OH‐DPAT‐induced hypothermia, these data suggest that central 5‐HT₇ and 5‐HT_1A receptors can be independently involved in thermoregulation 84.

Antidepressants effect of selective serotonin reuptake inhibitors (SSRIs) partly depends on 5‐HT_1A autoreceptor functions. Chronic treatment with SSRIs is associated with a progressive recovery of 5‐HT neurons firing rate, correlating with clinical amelioration of depression 10. It is noteworthy that 5‐HT₇ receptors can also be involved in the therapeutic effects of antidepressants. For example, both tricyclic antidepressants as well as SSRIs induced c‐Fos expression in rats in a manner consistent with the 5‐HT₇ receptor activation within the suprachiasmatic nucleus. Moreover, chronic antidepressants treatment led to a decreased 5‐HT₇ receptor binding 85. These data suggest possible interplay between 5‐HT₇ and 5‐HT_1A receptors during antidepressant drug action. One possible mechanism of such cross talk can be selective and regulated heterodimerization (see below).

Oligomerization of 5‐HT_1A and 5‐HT₇ Receptors

Results of multiple biochemical, structural, and functional studies collected during the last decade clearly indicate that GPCRs can exist as oligomeric complexes 86, 87. Moreover, a growing body of evidence points to the functional importance of GPCR oligomers for receptor trafficking, receptor activation, and G protein coupling in native tissues 88. The clinical significance of GPCR oligomerization has also become more evident during recent years, leading to identification of oligomeric complexes as novel therapeutic targets 89, 90.

Oligomerization can occur between identical receptor protomers (homomerization) or between different receptors belonging to the same or different GPCR families (heteromerization). Heteromerization may result in changed receptor pharmacology either by affecting the ligand binding on individual monomers or by the formation of new binding sites 91, 92. More importantly, heteromerization may also influence signaling pathways regulated by a given receptor monomer 93, 94, 95, 96, 97, which can lead to a switch in G protein coupling 98. Thus, heteromerization provides an additional mechanism regulating cellular processes through the fine tuning of receptor‐mediated signaling.

Using a novel Förster resonance energy transfer (FRET) technique based on the spectral analysis, we have recently demonstrated that 5‐HT_1A receptors can form homodimers at the plasma membrane 99, 100. We also showed that FRET efficiency measured for the 5‐HT_1A receptor oligomers significantly decreased in response to agonist stimulation. Combined results of our studies suggest that this decrease was mediated by accumulation of FRET‐negative complexes rather than by dissociation of oligomers to monomers. Formation of 5‐HT_1A homomers (including the higher‐order oligomers) was further confirmed by several recent publications 101, 102. By combination of computational protein–protein docking, site‐directed mutagenesis, and FRET‐based analysis, we have recently demonstrated that transmembrane domains TM4/TM5 can be involved in the formation of 5‐HT_1A receptor dimers and indicated that specific amino acids interactions (e.g., W175^4.64, Y198^5.41, R151^4.40, and R152^4.41) maintain the interaction interface 103.

The 5‐HT₇ receptor has been shown to form homooligomers as well. This was demonstrated in recombinant system by pharmacological studies using the antagonist risperidone, which binds in a pseudo‐irreversible manner to one protomer of a dimer as well as by the Bioluminiscence resonance energy transfer approach 104. Recently, homooligomerization of 5‐HT₇ receptor was confirmed using two different FRET assays 101. Existence of 5‐HT₇ receptor homodimers has been also shown under in vivo like condition in primary cultures of rat cortical astrocytes 105.

Heterooligomerization between 5‐HT_1A and 5‐HT₇ receptors has been demonstrated by combined application of biochemical (co‐immunoprecipitation) and biophysical (FRET) approaches 106. In the same study, it has been also demonstrated that hippocampal neurons express both 5‐HT_1A and 5‐HT₇ receptors and that these receptors are highly co‐localized at the plasma membrane. Moreover, co‐immunoprecipitation studies in mouse brain provided direct evidence that these receptors can form heteromers in neurons in vivo. Functional analysis of oligomerization between 5‐HT_1A and 5‐HT₇ receptors revealed that heterooligomers decreased the 5‐HT_1A receptor‐mediated activation of G_i protein without affecting 5‐HT₇ receptor‐mediated G_s protein activation. In addition to reduced G_i protein coupling of 5‐HT_1A receptor, heterodimerization affected 5‐HT_1A receptor‐mediated activation of G protein‐gated potassium (GIRK) channels, an effect mediated through the G_βγ‐subunits of G_i proteins 107, 108. Noteworthy that inhibitory effect of heterooligomerization on GIRK currents was preserved in hippocampal neurons, suggesting a physiological relevance of heteromerization in vivo 106.

We have also demonstrated that heterodimerization can initiate agonist‐mediated internalization of 5‐HT_1A receptor, which is highly resistant to internalization when expressed alone. Once internalized, 5‐HT_1A receptor can activate G protein‐independent signaling pathways such as a β‐arrestin‐mediated coupling to mitogen‐activated protein kinase (MAPK) 106. Thus, dependent on the relative amount of 5‐HT_1A receptors participating in heterooligomers, the same ligand (serotonin) can activate distinct ERK‐mediated pathways (i.e., G protein dependent or β‐arrestin dependent). This also raises the possibility that conditions selectively promoting or inhibiting heterodimerization could have a significant physiological relevance.

Possible Role of Heterodimerization between 5‐HT_1A and 5‐HT₇ Receptors in Anxiety and Depression

Modulation of receptor signaling by heterodimerization, in particular enhancement of 5‐HT_1A receptor internalization, is also likely to play an important role in pathophysiological processes in CNS. The 5‐HT_1A receptor is expressed as postsynaptic receptor in multiple brain regions including hippocampus and cortex 109, 110. In addition, the 5‐HT_1A receptor is highly expressed as a presynaptic autoreceptor in 5‐HT neurons of raphe nuclei, where it controls serotonin release through feedback inhibition 111, 112. The role of 5‐HT_1A autoreceptor in the regulation of presynaptic serotonin release has led to the hypothesis that its selective and progressive desensitization is a key element responsible for the therapeutic action of the SSRIs. Although widely accepted, this hypothesis still cannot explain the fact that the chronic in vivo SSRI treatment results in preferential functional desensitization of only 5‐HT_1A autoreceptors without affecting the postsynaptic 5‐HT_1A receptors 113, 114. We propose that higher amount of heterodimers produced in presynaptic versus postsynaptic neurons might represent a mechanism responsible for the differential desensitization obtained for the 5‐HT_1A auto‐ and heteroreceptors (Figure 1). Such asymmetric distribution of heterodimers will result in effective co‐internalization of 5‐HT_1A receptor within 5‐HT_1A‐5‐HT₇ heterodimers upon 5‐HT release, which can take place either under physiological conditions or as a consequence of SSRI treatment. It has been demonstrated that the expression level of postsynaptic 5‐HT₇ receptors in hippocampus and forebrain progressively decreased during the postnatal development, while the expression of 5‐HT_1A receptor remained relatively constant 36, 106, 115. This suggests that under physiological conditions, 5‐HT_1A receptor homodimers represent a dominant postsynaptic population in adulthood ([5‐HT_1A‐5‐HT_1A] ≫ [5‐HT_1A‐5‐HT₇]). As 5‐HT_1A receptor homodimers are resistant to the agonist‐mediated internalization, the amount of the postsynaptic 5‐HT_1A receptor will not be affected by 5‐HT. Different relative concentration of 5‐HT_1A‐5‐HT₇ heterodimers between serotonergic presynaptic autoreceptors and target (postsynaptic receptors) neurons can explain not only the differences in desensitization between pre‐ and postsynaptic 5‐HT_1A receptors, but also suggests that the regulated and balanced ratio of homo‐ and heterodimerization on pre‐ and postsynaptic neurons may be critically involved in both the onset as well as the response to treatment of psychiatric diseases such as depression and anxiety (Figure 1). Moreover, differences in relative concentration of 5‐HT_1A‐5‐HT₇ heterodimers in raphe and hippocampus represent an intrigue possibility to explain regional differences in the coupling of 5‐HT_1A receptor to G proteins. It has been shown that in raphe, Gα _i3 is the predominat coupling partner of 5‐HT_1A receptor, while in hippocampus, receptor couples to both Gα _o and Gα _i3 proteins 116. Therefore, future investigations will be needed to more precisely evaluate to which extent the 5‐HT_1A and 5‐HT₇ oligomerization can contribute to selective activation of different G proteins.

Figure 1.

Hypothetical model explaining the functional role of 5‐HT _1A‐5‐HT ₇ heterodimerization. Under physiological conditions (left), the amount of 5‐HT _1A‐5‐HT ₇ heterodimers in presynaptic 5‐HT neurons is higher than in postsynaptic neurons, representing a mechanism responsible for the differential 5‐HT or SSRI‐mediated internalization obtained for the 5‐HT _1A auto‐ versus heteroreceptors. By depression (right), the relationship between 5‐HT _1A‐5‐HT _1A homodimers and 5‐HT _1A‐5‐HT ₇ heterodimers in presynaptic 5‐HT neurons becomes shifted toward 5‐HT _1A‐5‐HT _1A homodimers. This will result in decreased 5‐HT or SSRI‐mediated internalization of 5‐HT _1A autoreceptors, which in turn will lead to 5‐HT _1A receptor‐mediated inhibition of 5‐HT release. On the postsynaptic neurons, higher amount of heterodimers ([5‐HT _1A‐5‐HT ₇] > [5‐HT _1A‐5‐HT _1A]) is expected during depression. Consequently, internalization of postsynaptic 5‐HT _1A receptors will be increased, leading to the increased neuronal excitability.

Abnormalities in 5‐HT_1A receptors have been previously noted in depressed patients and implicated in the pathophysiology of depression. Analysis of the postmortem brains of depressed subjects in comparison with control samples revealed a specific up‐regulation of 5‐HT_1A autoreceptors in the raphe area, with no changes in postsynaptic 5‐HT_1A receptor sites 30. In addition, it has been recently shown that transgenic mice with low level of 5‐HT_1A autoreceptor possess increased resilience to chronic stress and faster response to SSRI treatment as compared to mice with the high level of presynaptic 5‐HT_1A receptor 117. On the other hand, postmortem studies of major depression demonstrated a decrease in postsynaptic 5‐HT_1A receptor expression in hippocampus and prefrontal cortex, consistent with decreased postsynaptic 5‐HT_1A receptor‐mediated signaling observed in depressed suicide tissue 118, 119. Decreased 5‐HT_1A receptor density in the prefrontal cortex was also found in multiple PET studies of human patients with major depression 120, 121, 122. All these results demonstrate that in depression, the number of presynaptic 5‐HT_1A receptors increased and this is accompanied by the concomitant decrease in postsynaptic 5‐HT_1A receptor level.

Although the detailed analysis of 5‐HT₇ receptor distribution in patients with depression is not available yet, recent studies in rats have shown that expression of 5‐HT₇ receptor is up‐regulated in the hippocampus after exposure to stress, and these changes were inhibited by treatment with fluoxetine 123. Similar results were obtained in mice, where chronic fluoxetine treatment led to a down‐regulation of 5‐HT₇ receptor binding in hypothalamus 124. Moreover, analysis of 5‐HT₇ knockout mice revealed that those mice exhibit a behavioral phenotype similar to that of antidepressant‐treated mice 6, 125. These data suggest that in depression, the relationship between 5‐HT_1A homodimers and 5‐HT_1A‐5‐HT₇ heterodimers in presynaptic neurons may be shifted from [5‐HT_1A‐5‐HT₇] > [5‐HT_1A‐5‐HT_1A] to [5‐HT_1A‐5‐HT_1A] > [5‐HT_1A‐5‐HT₇], while in postsynaptic neurons, the opposite effect is expected (i.e., [5‐HT_1A‐5‐HT₇] > [5‐HT_1A‐5‐HT_1A]) (Figure 1). Functionally, an increased concentration of 5‐HT_1A‐5‐HT₇ heterodimers in hippocampus (and consequently inhibition of 5‐HT_1A mediated signaling via its internalization) may cause depression or accentuate the effect of stress. In combination with decreased presynaptic activity due to a relative increase in concentration of 5‐HT_1A homodimers in presynaptic neurons, serotonergic transmission particularly through 5‐HT_1A receptor appears to be compromised in depression.

Conclusion

Serotonin receptors 5‐HT_1A and 5‐HT₇ are co‐expressed in multiple brain regions, where they are critically involved in regulation of various physiological and pathological brain functions. Recent studies demonstrated that these receptors can form heterodimers both in vitro as well as in vivo. From the functional point of view, heterodimerization decreases G_i protein coupling of 5‐HT_1A receptor without substantial changes in the coupling of the 5‐HT₇ receptor to the G_s protein. In addition, heterodimerization facilitates internalization of 5‐HT_1A receptors as well as its ability to activate MAP kinase. Thus, differences in relative concentration of 5‐HT_1A‐5‐HT₇ heterodimers on presynaptic serotonergic neurons and target (post‐synaptic) neurons can explain the differences in desensitization patterns between pre‐ and postsynaptic 5‐HT_1A receptors. More importantly, regulated and balanced ratio of homo‐ and heterodimerization on pre‐ and postsynaptic neurons may be critically involved in both the onset as well as the response to the treatment of psychiatric diseases such as depression and anxiety. As heterodimerization can exist between other 5‐HT receptor subtypes and between 5‐HT receptors and nonserotonergic GPCRs, it would be therefore interesting to analyze in future studies relevance of such receptor complexes for the pathophysiology and treatment of depression.

Conflict of Interest

The authors declare no conflict of interest.