1. Introduction
It has been long known that mental illness can affect different behaviors resulting in social dysfunctions [
1]. Psychopathologies such as schizophrenia, autism spectrum disorder and depression, are all good examples of how social relationships can be severely impacted [
2,
3,
4]. According to current scientific evidence, environmental factors and multiple risk genes are the key features for the development of psychopathologies, causing subtle changes in the brain neurotransmission that leads ultimately to behavioral symptoms and emotional instability. One brain structure that is reportedly involved in many mental disorders is the serotoninergic median raphe nucleus.
The median raphe region (MRR, also known as superior central nucleus, B5+B8 [
5]) is located in the midline of the brainstem, and it is constituted by the median and the paramedian raphe regions [
6]. It is also part of the mesolimbic serotonergic pathway that projects to the septum and to the hippocampus [
7,
8]. This pathway has been the subject of a large number of studies, which demonstrate the involvement of MRR in anxiety [
9,
10,
11,
12,
13] , social behavior [
14,
15,
16,
17], depression [
18] and in the control of the circadian rhythms [
19] by regulating the theta rhythms of the hippocampus [
20,
21].
Although the MRR is widely known as a serotoninergic nucleus, recent studies have shown that serotoninergic neurons are only a minority of the MRR (8.5% of all MRR neurons) and instead, the majority of MRR neurons are GABAergic (61%) or can be characterized by the presence of glutamatergic transporters (vesicular glutamate transporter 2 (VGluT2) and 3 (VGluT3)) [
6,
22]. The presence [
23,
24,
25] and production [
26,
27,
28,
29] of dopamine (DA) in the MRR was repeatedly confirmed in rats, and its glial uptake – suggesting functional relevance- was also demonstrated [
30]. It was even suggested that the hippocampus might get DAergic innervation from the MRR [
31]. The presence of DA was also found in the MRR of deermice, Peromyscus maniculatus [
32] and its production was shown also in chicken [
33]. Nevertheless, most of the studies used tyrosine hydroxylase (TH) immunohistochemistry, which is the rate limiting enzyme of catecholamine synthesis including DA, noradrenaline and adrenaline. In rats considerable amount of noradrenaline was also detected in the MRR [
23,
24,
25], therefore, confirmation of dopamine-β-hydroxylase (DBH, the key enzyme for noradrenaline synthesis) negativity of the TH positive cells seemed to be important [
35]. Moreover, one study concentrating on the dorsal raphe (DR) did not find TH immunopositivity in the MRR of rats [
34].
DA is present both in the central and peripheral nervous system, and it has been described as a key player in the regulation of a wide range of behaviors [
36]. Imbalances in DA can contribute to the emergence of neurological and psychiatric disorders with disturbances in mood, locomotion and cognitive functions [
37,
38] such as Parkinson’s disease, schizophrenia, addiction and attention hyperactivity disorders [
39,
40,
41,
42]. Despite each condition having different etiology, most of these disorders are characterized by abnormal social behavior. Moreover, in some cases (especially in Parkinson’s disease and schizophrenia) the major target of presently available therapies is the dopaminergic system [
44,
45,
46].
Our aim was (i) to confirm the presence of dopaminergic cells in the MRR of mice (by immunohistochemistry and RT-PCR) as well as in humans (by RT-PCR); (ii) investigate their role in social behavior by chemogenetics (virally introduced designer receptors exclusively active by designer drugs (DREADDs) activated by the synthetic ligand Clozapine-N-oxide (CNO) [
47,
48] in dopamine transporter-Cre (DAT-Cre) animals (
Figure 1); (iii) summarize additional information on locomotion, anxiety and memory gained during the behavioral testing, as alterations of these behavior might influence the outcome of social tests.
3. Discussion
Our results suggest that the MRR contains dopaminergic neurons (both in mice and in humans) and that these cells do not influence locomotion, anxiety and memory, however, their stimulation in mice decreased social behavior during the social discrimination test, whereas their inhibition increased the friendly social behavior in the social interaction test.
The presence of DAT in the human pontin raphe nucleus confirmed the translational value of our results. Despite the adequate RNA content (see
Supplementary Table S1.) only very faint expression was observed in sample 6. It is possible that this person was in agony for a longer time than others (see
Supplementary Table S2.) and his oxygen supply may have been permanently reduced due to previous pneumonia as well, which may negatively affect RNA integrity. However, it cannot be ruled out that the sampling does not always succeed with the same accuracy, which is also why there may be differences in the DAT expression of the samples. The absence of DAT mRNA in cortical areas was in line with a previous rat study [
52] and can be explained by its expression in the cell bodies rather than on axon terminals.
Table 1 summarizes all the observed changes during behavioral experiments performed on DAT-Cre mice in this study.
In the tests used to measure locomotor activity of the mice (OF, closed arm entries in EPM and total arm entries in Y-maze) no difference was found between excitatory, inhibitory and control groups. According to the literature [
53,
54,
55], DA is directly responsible for the locomotor activity. However, mainly the nigrostriatal pathway was implicated in this behavior. Additionally, a previous study has shown that injecting different drugs (GABA agonists, opioid agonists) into the MRR may lead to hyperlocomotion [
56]. Not all effect was antagonized by haloperidol (D2 antagonist) injection, suggesting a DA-dependent as well as an independent MRR-related influence on locomotion [
56,
57]. We have to add, due to its vast projection [
58], it is not easy to determine which pathway the DAT-MRR is involved. Nevertheless, in light of the current findings, we might conclude that dopaminergic cells from the MRR might have a different role than the nigrostriatal pathway.
During the habituation phase of the sociability test, there was no sign of fear from the object with any group tested. Although DA is implicated in fear, it is more connected to the extinction than the realization of the fear response [
59,
60] and the ventral tegmental area (VTA) is suggested as a main source [
59]. However, D1 receptors of the prefrontal cortex has a role in the acquisition of contextual fear conditioning [
61]. We cannot entirely close out that dopaminergic cells of the MRR may project to this area, as DA may colocalize with VGluT3 [
62], and previous studies showed VGluT3 innervation from the MRR to the prefrontal region [
63]. Additionally, our recent study suggested that MRR VGluT2 neurons regulate the acquisition of negative experience in mice [
22]. However, it is still not yet known if there is a DA-VGluT2 interaction in the MRR. Further studies are required to address the detailed role of DA cells of the MRR in the context of fear.
In the second phase of the sociability test, the social interested was investigated, but no changes were observed among the groups. All animals displayed more interest toward the conspecific stimulus mice rather than the empty cage. In contrast to our results, Bariselli, et al. [
64] described that inhibition of DA neurons decreased the sociability among conspecifics, however, they investigated the DA neurons of the VTA. In our hands during the social discrimination phase, excitation of the DAT-MRR cells resulted in a
decrease in social interest toward both conspecific stimulus mice. This result corroborates with the finding of the social interaction test, where inhibition of the DAT-MRR cells
increased the friendly social behavior. In contrast, Liu, et al. [
65] found that the dendritic cell fact 1 (Dcf1) knockout mice displayed social interaction deficit and it could have been reverted by DA or D1 receptor agonist, suggesting that lower rather than enhanced dopaminergic tone will lead to reduced social interest. This difference might be due to the divergent role of DA on different brain areas and also the usage of different genetic techniques. Nevertheless, our results suggest that the dopaminergic neurons of the MRR might be also involved in the regulation of social behavior, an important manifestation of normal as well as pathological behavior.
In the resident intruder test, no statistically significant difference was observed between the studied groups. This might be due to the fact that social behavior and ethological aggression are regulated by slightly different mechanisms [
66]. Another possible explanation is that the later experiment was carried out in dark, thus, was less distressing for the nocturnal animal. Indeed, previously it was shown that the outcome of the test is highly dependent on the anxiogenic nature of the environment. Although both tests (SIT and RIT) represent a social challenge and induce similar behaviors, but involve different contexts, thereby different levels of anxiety. For example, Haller, et al. [
67] showed that the cannabinoid receptor 1 (CB1) KO mice was more aggressive during RIT, while less aggressive during SIT compared to wild-type, which was explained by the different stressfulness of the tests (e.g. familiar vs non-familiar environment; light vs dark during the test).
Although we were not able to find any anxiety-related effect in the EPM test, this test was done also in dark, where the CB1 KO animals also behaved normally [
67]. However, our EPM results were in contrast to the findings of Bahi and Dreyer [
68], presenting a decreased anxiety both in the OF and EPM after silencing the DAT in the nucleus accumbens. Once again, this variation is probably due to the different brain region targeted, and the environmental conditions may also influence the outcome (the mentioned study was done during the light, inactive phase of the animals). Nonetheless, given the successful history of evaluating the anxiety-like behavior in mice with the EPM test [
69,
70,
71], our results might indicate that manipulation of DAT-MRR cells does not have a strong effect on anxiety control.
Previous studies have shown that acquisition and consolidation of memory involve dopaminergic activity [
72,
73]. In addition, a recent study found that a decrease in dopaminergic cells might be a neurocognitive signature of Alzheimer’s disease [
74]. All these studies suggest a close relationship between DA and memory. We assessed working memory by Y-Maze and social memory by SD without treatment effects. In the SD test, the test mice acted as if they have never met the familiar mice. This result corroborates with the notion that social memory only lasts a few hours under laboratory conditions [
75,
76], although it has been reported that vasopressin release as well as group housing could prolong this effect [
76,
77,
78]. These results might indicate that DAT-MRR is not involved in either working or social memory. However, it might have a role in spatial memory. Indeed, injection of a D1/D5 antagonist into the CA1 region of the hippocampus decreased the SD abilities at 24h [
79]. As VGluT3-MRR cells projecting to the hippocampus [
63] may be co-localized with DA, they might provide the source of DA in the hippocampus. Although, these assumptions need further confirmation.
Our experiments have certain limitations that need to be addressed in future studies. (i) We tried to use the most optimal tests to measure each behavior, but in order to have a general view we used a long test battery and repeated CNO injections. However, our previous optogenetic manipulation showed that a single 5 min stimulation of the MRR may induce long lasting, plastic changes in the animal’s behavior [
80]. (ii) Although one might expect that the excitatory and inhibitory groups would behave in opposite ways, our results did not confirm this assumption. Nevertheless, it is not that surprising as the two types of DREADD sequences affect different signaling pathways (Gq and Gi). Moreover, stimulation seems to be a more active process, while inhibition mostly diminishes the effect of other stimulatory signals.
Our results further extend our understanding regarding the role of dopaminergic cells in our brain and raise some questions that could be addressed in future research. This is utmost important as many people worldwide take drugs influencing their dopaminergic system [
81] such as: L-DOPA for treatment of Parkinson’s disease, [
82], Aripiprazole for the treatment of schizophrenia and bipolar disorder [
83,
84], Tetrabenazine for the treatment of Huntington’s Disease [
85] and Pramipexole for the treatment of Restless legs syndrome [
86] and essential tremor [
87], which might have unwanted behavioral side effects leading to discontinuation of the pharmacotherapy. We focused on social behavior as it is a fundamental property of every day interactions and serves as the basis for survival and reproduction [
88]. Understanding the mechanisms behind this behavior could help scientists to provide a better treatment for those who suffer from psychological and psychiatric disorders.