Analysis of Antidepressant-like Effects and Action Mechanisms of GSB-106, a Small Molecule, Affecting the TrkB Signaling

Induction of BDNF-TrkB signaling is associated with the action mechanisms of conventional and fast-acting antidepressants. GSB-106, developed as a small dimeric dipeptide mimetic of BDNF, was previously shown to produce antidepressant-like effects in the mouse Porsolt test, tail suspension test, Nomura water wheel test, in the chronic social defeat stress model and in the inflammation-induced model of depression. In the present study, we evaluated the effect of chronic per os administration of GSB-106 to Balb/c mice under unpredictable chronic mild stress (UCMS). It was observed for the first time that long term GSB-106 treatment (1 mg/kg, 26 days) during ongoing UCMS procedure ameliorated the depressive-like behaviors in mice as indicated by the Porsolt test. In addition, chronic per os administration of GSB-106 resulted in an increase in BDNF levels, which were found to be decreased in the prefrontal cortex and hippocampus of mice after UCMS. Furthermore, prolonged GSB-106 treatment was accompanied by an increase in the content of pTrkB706/707 in the prefrontal cortex and by a pronounced increase in the level of pTrkB816 in both studied brain structures of mice subjected to UCMS procedure. In summary, the present data show that chronic GSB-106 treatment produces an antidepressant-like effect in the unpredictable chronic mild stress model, which is likely to be associated with the regulation of the BDNF-TrkB signaling.


Introduction
Major depressive disorder, commonly referred to as clinical depression, is a severe mental chronic disease that has a high prevalence in almost all developed countries, with more than 280 million people affected worldwide (approximately 4.4% of the population), and is expected to become the leading cause of disease burden by 2030 [1]. Complicated etiology of depression involves large number of factors including stressful life events, genetic risk, and type of personality response to the adverse life experiences [2]. Multiple basic and clinical studies proposed that combination of those factors causes disturbances within mesocorticolimbic circuits (nucleus accumbens, prefrontal cortex and hippocampus), thus contributing to the pathophysiology and the core symptoms of depression [3]. Indeed, magnetic resonance imaging (MRI) revealed smaller volumes of hippocampus, the inferior anterior cingulate, and the orbital prefrontal cortex in patients suffering from depression that are consistent with the post-mortem studies and the data on animal models of depression-like behavior [4]. Lower volume of hippocampal and prefrontal cortex found in the subjects with depression are believed to be related to the neuronal atrophy through an increase in the apoptosis-associated DNA fragmentation and necrotic neuron death, as well as with alterations of synaptic reorganization proteins and, predominantly, with the reactive astrogliosis in the distinct hippocampal regions [5].
Although the neurobiology of depression is still not clearly understood, the monoaminebased, the glucocorticoid-dependent, the glutamate-driven, the neurotrophin signaling deficiency, the neuroinflammatory, the impaired brain circuit and networks hypotheses have been proposed [6][7][8][9][10]. To date, clinically active antidepressants target primarily the monoamine systems in the brain, inhibiting the reuptake of monoamines (selective serotonin or/and noradrenaline and dopamine reuptake inhibitors, tricyclic antidepressants), decreasing their degradation rate (inhibitors of monoamine oxidase), or acting on receptors (noradrenergic antagonist-specific serotonin antagonist, serotonin modulating antidepressants) [11]. However, the long onset time for antidepressants effects (several weeks to months), a significant number of patients that do not respond to typical antidepressants and a broad range of side effects significantly limit the efficacy of treatment [12]. Experimental studies have reported that the delayed onset of antidepressants action is associated with a slowly developed increase and rearrangements of synaptic strength in the depressionrelated circuits, requiring changes in the gene expression and protein translation that are dependent, to some extent, on the engagement of BDNF/TrkB pathways [13,14]. Numerous basic research findings confirm that typical antidepressants increase the BDNF mRNA and protein expression as well as the TrkB activation in the hippocampus and the cortex [3,7,15], thus promoting the structural plastic changes associated with the antidepressant action mechanisms [8,16,17]. Moreover, conventional antidepressants have been shown to rapidly increase the phosphorylation of Tyr 706/707 at the TrkB catalytic domain and the phospholipase Cγ-binding site (Tyr 816 ), while leaving the phosphorylation of the Shc-interaction site Tyr 515 unaffected [18][19][20]. A rapidly developing and sustained antidepressant effect of ketamine which depends on the NMDA receptors inhibition, leads to the activation of mTORC1 complex followed by an increase in the translation and release of BDNF, an enhancement of BDNF/TrkB signaling, and a subsequent activation of the related cellular plasticity cascades [21]. Casarotto and colleagues have recently reported that structurally unrelated antidepressants (fluoxetine, imipramine, moclobemide, ketamine, esketamine, R,R-HNK) directly bind to transmembrane region of dimerized TrkB and potentiate the receptor activation by BDNF [22]. Overall, a number of previous observations as well as some more recent evidence emphasize the significance of TrkB as a valuable target for antidepressants [23,24] and reaffirm the critical role of BDNF/TrkB signaling in the pathophysiology of depression and in antidepressants action [7,25,26].
The given data indicate the importance of BDNF/TrkB pathway for depression, leading to several proposed approaches aiming to restore or enhance neurotrophin signaling, such as: a local delivery of neurotrophin by engineered cells or viral vectors; an increase of BDNF levels and the neurotrophin effects; a regulation of TrkB receptor synthesis and dimerization; an indirect Trk receptors activation via transactivation mechanisms; a modulation of individual steps in the desired signal transduction pathway; an interaction with TrkB receptor (e.g., small peptide and non-peptide compounds for TrkB); a functional mimicry [23,[27][28][29]. Although the intracerebral administration of BDNF produced an antidepressant-like effect in animal models of depression [30,31], clinical use of neurotrophin is limited, mainly due to its suboptimal pharmacokinetic properties and side effects. One of the ways to overcome these factors is to augment BDNF/TrkB pathway with small molecules that could function as TrkB ligands or enhance neurotrophin signaling [27]. Several low-molecular-weight peptide and non-peptide compounds with favorable pharmacokinetic profiles acting on the BDNF/TrkB pathway and exerting antidepressant properties in various depression-like animal behavioral models have been developed to date. The best characterized and well-studied examples are the 7,8-dihydroxyflavone [32,33] and its synthetic derivative [34], the low-molecular weight TrkB antagonist with anxiolytic and antidepressant activity ANA-12 [35], and the low molecular weight BDNF mimetic GSB-106 [36].
A crystallographic analysis, on par with mutagenesis studies and gain-of-function approaches based on chimeric recombination of BDNF protein, allowed to propose the amino acid residues, implicated in neurotrophin interaction with TrkB. In particular, residues 91-97 of the 4th loop of BDNF (Thr91-Met92-Asp93-Ser94-Lys95-Lys96-Arg97) have been recognized as one of the sites for neurotrophin/TrkB complex formation (reviewed in [37,38]). Orally available homodimeric dipeptide GSB-106 (bis-(N-monosuccinyl-L-seryl-L-lysine) hexamethylenediamide; Figure 1), which resembles the part of BDNF loop 4, has been recently developed at the Zakusov Institute of Pharmacology [36]. Namely, GSB-106 has been designed based on the BDNF loop 4 β-turn amino acid sequence -Asp93-Ser94-Lys95-Lys96-, where the putative TrkB-binding residues -Ser94-Lys95-were retained, the upstream residue -Asp93-was replaced by its bioisostere, a succinic acid residue, and -Lys96-was substituted by the amide group; C-terminal dimerization was performed using oligomethylenediamine spacer [36,39]. The compound has been shown to exert a trophic effect, up-regulate the TrkB phosphorylation and downstream PI3K/Akt, MAPK/Erk and PLCγ signaling pathways in vitro [40,41]. Furthermore, GSB-106 has been found to act as a partial TrkB receptor agonist, promoting protection of serum-deprived neuronal-like cells by counteracting cell apoptosis through the activation of TrkB-dependent pro-survival mechanisms, including inactivation of pro-apoptotic BAD protein and suppression of caspases 9 and 3/7 [42]. Behavioral studies revealed the substantial antidepressant-like activity of GSB-106. Particularly, GSB-106, administered to rodents either orally or intraperitoneally, under acute, subchronical or chronical conditions demonstrated an antidepressant-like activity in the Porsolt test, a tail suspension test and the Nomura water wheel test, as well as in the chronic social defeat stress model and in the inflammation-induced model of depression [36,[43][44][45][46]. ligands or enhance neurotrophin signaling [27]. Several low-molecular-weight peptide and non-peptide compounds with favorable pharmacokinetic profiles acting on the BDNF/TrkB pathway and exerting antidepressant properties in various depression-like animal behavioral models have been developed to date. The best characterized and well-studied examples are the 7,8-dihydroxyflavone [32,33] and its synthetic derivative [34], the low-molecular weight TrkB antagonist with anxiolytic and antidepressant activity ANA-12 [35], and the low molecular weight BDNF mimetic GSB-106 [36].
А crystallographic analysis, on par with mutagenesis studies and gain-of-function approaches based on chimeric recombination of BDNF protein, allowed to propose the amino acid residues, implicated in neurotrophin interaction with TrkB. In particular, residues 91-97 of the 4th loop of BDNF (Thr91-Met92-Asp93-Ser94-Lys95-Lys96-Arg97) have been recognized as one of the sites for neurotrophin/TrkB complex formation (reviewed in [37,38]). Orally available homodimeric dipeptide GSB-106 (bis-(N-monosuccinyl-L-seryl-L-lysine) hexamethylenediamide; Figure 1), which resembles the part of BDNF loop 4, has been recently developed at the Zakusov Institute of Pharmacology [36]. Namely, GSB-106 has been designed based on the BDNF loop 4 β-turn amino acid sequence -Asp93-Ser94-Lys95-Lys96-, where the putative TrkB-binding residues -Ser94-Lys95-were retained, the upstream residue -Asp93-was replaced by its bioisostere, a succinic acid residue, and -Lys96-was substituted by the amide group; C-terminal dimerization was performed using oligomethylenediamine spacer [36,39]. The compound has been shown to exert a trophic effect, up-regulate the TrkB phosphorylation and downstream PI3K/Akt, MAPK/Erk and PLCγ signaling pathways in vitro [40,41]. Furthermore, GSB-106 has been found to act as a partial TrkB receptor agonist, promoting protection of serum-deprived neuronal-like cells by counteracting cell apoptosis through the activation of TrkB-dependent pro-survival mechanisms, including inactivation of pro-apoptotic BAD protein and suppression of caspases 9 and 3/7 [42]. Behavioral studies revealed the substantial antidepressant-like activity of GSB-106. Particularly, GSB-106, administered to rodents either orally or intraperitoneally, under acute, subchronical or chronical conditions demonstrated an antidepressant-like activity in the Porsolt test, а tail suspension test and the Nomura water wheel test, as well as in the chronic social defeat stress model and in the inflammation-induced model of depression [36,[43][44][45][46]. Generally, a panel of rodent tests and models is used to evaluate the potential antidepressant effect of the compounds. Based on its close resemblance to certain human depressive symptoms, unpredictable chronic mild stress (UCMS) is considered as one of the most translationally-relevant and valuable models for studying the pathophysiology of depression and the efficacy of antidepressants in animals [47]. It has been observed that rats and mice subjected to UCMS demonstrate anhedonia in a sucrose preference test Generally, a panel of rodent tests and models is used to evaluate the potential antidepressant effect of the compounds. Based on its close resemblance to certain human depressive symptoms, unpredictable chronic mild stress (UCMS) is considered as one of the most translationally-relevant and valuable models for studying the pathophysiology of depression and the efficacy of antidepressants in animals [47]. It has been observed that rats and mice subjected to UCMS demonstrate anhedonia in a sucrose preference test and depression-like scores in the tests of learned helplessness such as the Porsolt test [48][49][50], while antidepressants eliminate those manifestations [51,52]. Most studies usually raise the question of drug efficacy in the settings of the developed behavioral abnormalities, and the course of potential antidepressants is initiated after several weeks of ongoing stress stimuli [53]. It has been widely shown that chronic stress-induced behavioral changes are associated with neuronal atrophy and reduced levels of BDNF and phosphorylated/activated TrkB in the prefrontal cortex and the hippocampus in rodents, while almost all antidepressants can restore BDNF level and activate TrkB receptors in rodent chronic stress models (reviewed in [7]).
The purpose of this study is to examine the potential antidepressant-like effects of GSB-106 as demonstrated by an UCMS model of depression in Balb/c mice and to evaluate whether this effect is mediated through amelioration of BDNF/TrkB signaling.

Effect of Chronic GSB-106 Administration on Depressive-like Responses in the Porsolt Test
The GSB-106 compound administered to mice for 14 days in parallel with ongoing UCMS (Figure 2a) dose-dependently contributed to a decrease in the immobilization time in the behavioral test compared to the untreated animals ( Figure 3). and depression-like scores in the tests of learned helplessness such as the Porsolt test [48][49][50], while antidepressants eliminate those manifestations [51,52]. Most studies usually raise the question of drug efficacy in the settings of the developed behavioral abnormalities, and the course of potential antidepressants is initiated after several weeks of ongoing stress stimuli [53]. It has been widely shown that chronic stress-induced behavioral changes are associated with neuronal atrophy and reduced levels of BDNF and phosphorylated/activated TrkB in the prefrontal cortex and the hippocampus in rodents, while almost all antidepressants can restore BDNF level and activate TrkB receptors in rodent chronic stress models (reviewed in [7]).
The purpose of this study is to examine the potential antidepressant-like effects of GSB-106 as demonstrated by an UCMS model of depression in Balb/c mice and to evaluate whether this effect is mediated through amelioration of BDNF/TrkB signaling.

Effect of Chronic GSB-106 Administration on Depressive-like Responses in the Porsolt Test
The GSB-106 compound administered to mice for 14 days in parallel with ongoing UCMS (Figure 2a) dose-dependently contributed to a decrease in the immobilization time in the behavioral test compared to the untreated animals ( Figure 3).   As shown in Figure 3, stressed mice exhibited a 1.4-fold increase (F = 12.83; p = 0.001) in immobilization time compared to non-stressed animals in the Porsolt test. A higher dose (1 mg/kg, per os) of GSB-106 attenuated depressive-like responses in mice by 1.2 (F = 12.83; p = 0.05) times compared to stressed mice treated with vehicle. Amitriptyline (10 mg/kg, per os) reduced the dysphoric behavior of stressed animals by 1.5 times (F = 12.83; p = 0.0001) compared to the control stressed mice. Thus, the preliminary stage of assessing the behavioral effects of substances in UCMS model indicated a dose-dependent anti-dysphoric effect of GSB-106.
After the effective dose (1 mg/kg) was determined, antidepressant-like activity and underlying mechanisms of GSB-106 were evaluated upon the prolonged use of the compound (26 administrations versus 14) (Figure 2b). As follows from the results of the behavioral assessment GSB-106 effects in Porsolt test (Figure 4), GSB-106 caused a 1.3-fold decrease in the immobilization periods compared to the vehicle (F = 34.58; p = 0.0499) in the absence of UCMS. In the vehicle-treated mice group subjected to UCMS a 1.2 folds increase in immobilization duration was registered (F = 34.58; p = 0.0423) compared to non-stressed animals treated with vehicle. In mice exposed to UCMS, chronic GSB-106 treatments decreased the immobilization periods by 2.9 folds (F = 34.58; p = 0.0001), compared to the stressed mice treated with vehicle. Thus, the Porsolt test data indicate the anti-dysphoric effect of GSB-106 after 26 days of oral administration in Balb/c mice. At the same time, animals subjected to the stress procedure exhibited a deeper antidepressant effect of the substance compared to the corresponding control.

Effects of Chronic GSB-106 Administration on BDNF Content and TrkB Site-Specific Phosphorylation in the Prefrontal Cortex and the Hippocampus of Mice
To examine the protein level of BDNF and TrkB site-specific phosphorylation (pTyr 706/707 , pTyr 515 , pTyr 816 ) in the different brain regions following UCMS and GSB-106 treatment a Western blotting was performed. Figure 5a shows that BDNF levels significantly decreased in the prefrontal cortex and the hippocampus (F = 18.5; p = 0.001 and F = 27.06; p = 0.001 respectively) of mice exposed to UCMS, compared to the control non-stressed animals, which is consistent with the previous data [54,55]. A chronic treatment with GSB-106 (1 mg/kg, per os, 26 days) during ongoing UCMS procedure restored UCMS-induced BDNF decrease in both brain structures (Fcortex = 18.5; p = 0.001 and Fhip = 27.06; p = 0.001) compared to that of the vehicle-treated UCMS mice. However, the administration of GSB-106 to animals that were not exposed to UCMS had no effect on BDNF content in these brain regions (Fcortex = 18.5; p = 0.98 and Fhip = 27.06; p = 0.07).

Effects of Chronic GSB-106 Administration on BDNF Content and TrkB Site-Specific Phosphorylation in the Prefrontal Cortex and the Hippocampus of Mice
To examine the protein level of BDNF and TrkB site-specific phosphorylation (pTyr 706/707 , pTyr 515 , pTyr 816 ) in the different brain regions following UCMS and GSB-106 treatment a Western blotting was performed. Figure 5a shows that BDNF levels significantly decreased in the prefrontal cortex and the hippocampus (F = 18.5; p = 0.001 and F = 27.06; p = 0.001 respectively) of mice exposed to UCMS, compared to the control nonstressed animals, which is consistent with the previous data [54,55]. A chronic treatment with GSB-106 (1 mg/kg, per os, 26 days) during ongoing UCMS procedure restored UCMSinduced BDNF decrease in both brain structures (F cortex = 18.5; p = 0.001 and F hip = 27.06; p = 0.001) compared to that of the vehicle-treated UCMS mice. However, the administration of GSB-106 to animals that were not exposed to UCMS had no effect on BDNF content in these brain regions (F cortex = 18.5; p = 0.98 and F hip = 27.06; p = 0.07). Abundant evidence indicates that action mechanisms of pharmacologically diverse antidepressants implicate a phosphorylation and an activation/transactivation of the TrkB receptors in the brain regions is associated with the depression-like behavior [24]. Our next step was to perform immunoblotting with antibodies to specific phospho tyrosines of TrkB was performed in the prefrontal cortex and the hippocampus of chronically stressed mice treated with GSB-106. Figure 5b  As follows from the data presented in Figure 5c, a decrease in the pTrkB 816 level was observed in the prefrontal cortex (F = 15.31; p = 0.0001) and in the hippocampus (F = 12.65; p = 0.0001) of UCMS mice compared to unstressed animals. Notably, data on TrkB 816 phosphorylation in the prefrontal cortex of stressed animals shows bimodal distribution.  (Tyr 515 ) proteins expression in mice prefrontal cortex and hippocampus. After the completion of UCMS procedure and the compound treatment, mice were decapitated and brain structures were collected. Protein extracts were subjected to polyacrylamide gel electrophoresis and transferred for Western blotting. Blots were probed with anti-BDNF, anti-phosphorylated TrkB antibodies and then reprobed with anti-TrkB antibody, anti-a-tubulin antibodies. For every picture, representative images for immunoblots are shown on the top panels, and quantitative data are shown on the bottom panels (red dots on the graphs correspond to mice 4, 8, 10 and 15; dark dots are all mice samples in the group). All data are expressed as mean ± SEM (n = 3; the significance of intergroup differences was estimated by one-way analysis of the variance (ANOVA), followed by Tukey's multiple comparisons test).
Abundant evidence indicates that action mechanisms of pharmacologically diverse antidepressants implicate a phosphorylation and an activation/transactivation of the TrkB receptors in the brain regions is associated with the depression-like behavior [24]. Our next step was to perform immunoblotting with antibodies to specific phospho tyrosines of TrkB was performed in the prefrontal cortex and the hippocampus of chronically stressed mice treated with GSB-106. Figure 5b shows that 4-weeks of UCMS exposure did not alter Tyr 706/707 phosphorylation in both structures (F cortex = 8.26; p = 0.98; F hip = 8.37; p = 0.89). Prolonged GSB-106 administration to UCMS subjected mice caused an increase in the pTrkB 706/707 content in the prefrontal cortex (F = 8.26; p = 0.05) compared to UCMS animals, while the effect of GSB-106 on pTrkB 706/707 level in both brain structures of nonstressed animals appeared to be statistically insignificant (F cortex = 8.26; p = 0.65; F hip = 8.37; p = 0.15).
As follows from the data presented in Figure 5c, a decrease in the pTrkB 816 level was observed in the prefrontal cortex (F = 15.31; p = 0.0001) and in the hippocampus (F = 12.65; p = 0.0001) of UCMS mice compared to unstressed animals. Notably, data on TrkB 816 phosphorylation in the prefrontal cortex of stressed animals shows bimodal distribution. However, none of the available data sets, other than those indicated, appear to be mathematically bimodal. At the same time, all our data pass the normality test (D'Agostino & Pearson normality test, Shapiro-Wilk normality test). Conceivably, the bimodal distribution of pTrkB 816 values might reflect the individual variability in response to GSB-106 treatment in the cohort of participating animals.
Administration of GSB-106 for 26 days along with a continued stress procedure promoted a rise of TrkB phosphorylation at Tyr 816 site in the prefrontal cortex (F = 15.31; p =0.0001) and in the hippocampus (F = 12.65; p = 0.0001) compared with control UCMS mice. Notably, in the absence of stress, GSB-106 caused a reduction of pTrkB 816 in the prefrontal cortex (F = 15.31; p = 0.05), whereas in the hippocampus there was a nonsignificant trend for decreased TrkB phosphorylation at that tyrosine residue (F = 12.65; p = 0.09). We also found that UCMS produced a decrease in TrkB phosphorylation at Tyr 515 site in the prefrontal cortex and the hippocampus (F = 8.92; p = 0.0006 and F = 4.4; p = 0.05, respectively) relative to control (Figure 5d). Further, we observed a decline of pTrkB 515 in both structures (F cortex = 8.92; p = 0.0015 and F hip = 4.4; p = 0.045) of non-stressed mice after chronic GSB-106 treatment. However, no changes in pTrkB 515 values were established in the prefrontal cortex and in the hippocampus of UCMS mice chronically treated with GSB-106 (F cortex = 8.92; p = 0.059 and F hip = 4.4; p = 0.11).

Discussion
The presented results for the first time show that prolonged per os GSB-106 administration (1 mg/kg, 26 days) produced an antidepressant-like behavioral response during an ongoing stress exposure in UCMS-treated Balb/c mice, as evaluated by the Porsolt test. Furthermore, GSB-106 restored a decrease in BDNF level induced by the UCMS procedure in the prefrontal cortex and in the hippocampus. Additionally, chronic administration of GSB-106 resulted in a TrkB activation (as estimated by phosphorylation of neurotrophin receptor at Tyr Y706/707 site) in the prefrontal cortex and induced a pronounced increase of TrkB Y816 phosphorylation in the prefrontal cortex and the hippocampus of UCMS subjected mice. Overall, these results indicate that GSB-106 exhibits an obvious antidepressant-like effect in a mouse model of UCMS-induced depression that was mediated, at least partially, by an increase in BDNF level and site-selective phosphorylation of TrkB Y706/707 and TrkB Y816 .
Two experimental design modifications were used to evaluate behavioral effects of GSB-106 on chronic stress. Initially, mice were exposed to stressful context for 42 days and 14 administrations of GSB-106 were carried out; then mice were subjected to a more prolonged stress procedure (54 days) and more GSB-106 treatments (26) were delivered. However, both experiments had one fundamental point in common, namely, they initially formed dysphoria-like chronic stress-induced behavior followed by treatment. Most studies utilizing this model of chronic stress have a similar design, where once the long-lasting stressful conditions are discontinued, recovery processes are likely to develop. In the present study, a 1 mg/kg dose of GSB-106 attenuated the depression-like responses of mice by 2.9 times following a 26-days treatment and by 1.2 times following a 14-days course, compared to the corresponding controls.
It is well known that long-term exposure to stress, as well as other conditions that contribute to the depressive-like behavior in animals, is accompanied by a BDNF expression decrease in the prefrontal cortex and in the hippocampus, while an increase in the level of neurotrophin has been observed in the nucleus accumbens [56,57]. In particular, numerous studies reported a decrease in the BDNF mRNA or protein levels, as well as an attenuation of TrkB receptor activity (detected by the level of phosphorylation of tyrosine residues at Tyr 706/707 ) and a decrease of Tyr 515 and Tyr 816 phosphorylation in the prefrontal cortex and hippocampus of rodents subjected to UCMS, that is consistent with the depression-like behavior [23,55,58]. In our study, a decrease of BDNF protein level both in the prefrontal cortex and the hippocampus in UCMS-induced depressed mice was found. Furthermore, a reduction of pTrkB 515 and pTrkB 816 content in the prefrontal cortex and the hippocampus of UCMS-treated mice was also revealed, while no changes in TrkB 706/707 phosphorylation were observed in both brain structures of UCMS mice, generally indicating the alleviation of TrkB-dependent signaling in vulnerable brain regions of stressed animals. Overall, our behavioral and western-blot findings have demonstrated that UCMS produces depressionlike symptoms (dysphoria), which are accompanied by the alterations of BDNF content and TrkB phosphorylation at specific tyrosine residues in the limbic brain structures, which are similar to those evidenced elsewhere. Importantly, the present study revealed that continuous UCMS procedure concomitantly with long-term per os GSB-106 treatment resulted in a BDNF content increase in the prefrontal cortex and the hippocampus and induction of site-specific phosphorylation of the TrkB receptor-at Tyr 706/707 and Tyr 816 sites, but not at Tyr 515 site, in this way mimicking the pattern attributed for most of the conventional antidepressants. We consider that an increase in TrkB 706/707 phosphorylation in the prefrontal cortex of stressed animals upon GSB-106 treatment might be a consequence of increased BDNF content, evoked by GSB-106, although it is quite difficult to determine the exact cause-and-effect relationships under the conditions of in vivo experiment with the multiple crosstalk and feedback loop mechanisms all of which suggest that this point requires additional investigations. Obtained data indicate the ability of GSB-106 to reverse depressive-like behavior. Simultaneously, GSB-106 normalizes the BDNF content, reduced due to a prolonged UCMS in animals, as well as the site-specific phosphorylation of TrkB receptor.
It has been experimentally established that the development of a depressive phenotype in UCMS is based on the elevated functionality of the hypothalamus-pituitary-adrenal (HPA) axis and its effects on the feedback systems of the limbic structures [59,60].
When a stressful event occurs, an activation of the HPA axis leads to the release of glucocorticoids and triggers a feedback mechanism, the main contributors of which are the hippocampus and the prefrontal cortex. As a result, the effects of stress are limited in time and the mechanism for restoring homeostasis and behavior is triggered. Prolonged activation of the HPA axis under chronic stress causes prolonged release of glucocorticoids which are toxic both for the hippocampus and the frontal cortex which leads to a decrease in the BDNF content [61,62], depletion of synaptic connections [63] and their morphological decrease develop [63,64].
In contrast to the hippocampus and the prefrontal cortex, the amygdala provides positive feedback to the HPA axis. Prolonged UCMS with the loss of hippocampal inhibition over the HPA axis increases activity in the amygdala. UCMS has been found to cause an increased coherence of activity in the amygdala [65], along with an increase in the dendrites' length, the density of dendritic spines [66][67][68], and synaptic proteins [66]. The chronic stress activation causes a decrease in the activity of the mesolimbic dopamine system, which may lead to the development of anhedonia as a result of stress which could be a target of antidepressant effect [69].
It is widely believed that chronic antidepressant treatment causes an increase in the synthesis or release of BDNF, activation of TrkB receptors and related signal transduction pathways, triggers up-regulation of genes associated with BDNF-induced plasticity [7,24,70,71]. Duman and co-workers were the first, who have demonstrated the stimulating effect of chronic (21 days), but not acute tranylcypromine, sertraline, desipramine, mianserin treatment on the BDNF and TrkB mRNA expression in rat hippocampus [15]. Then, multiple studies supported the positive effect of prolonged antidepressant courses on the BDNF mRNA expression and their ability to prevent/abolish a decrease in BDNF protein level caused by a chronic stress paradigm in vivo [72][73][74][75][76][77].
Saarlainen et al. [18], Rantamaki et al. [19,20] have demonstrated that, unlike BDNF, which requires prolonged administration of antidepressants to get a level increase, activation of the TrkB receptor in the cortex and hippocampus can occur upon even acute drug treatment. Particularly, antidepressant-induced TrkB 706/707 phosphorylation takes place within 30 min after a single dose, whereas no changes were detected at TrkB 515 residue (Shc binding site) either under acute or chronic administration. Furthermore, an increase in Tyr 706/707 phosphorylation was observed after sustained (21 days) antidepressant treatment as well [18]. However, phosphorylation at Tyr 816 residue in the mice hippocampus was detected 1 h after acute administration of fluoxetine, riboxetine, citalopram, imipramine; a similar effect on the phosphorylation of Tyr 816 was also established upon long-term administration (21 days) of fluoxetine [19]. Altogether, multiple complementary findings obtained to date, indicate that both acute and chronic courses of most clinically used antidepressants could induce the kinase activity of the receptor (increased phosphorylation at Tyr 706/707 ), followed by an increase in Tyr 816 (binding site for PLC-γ1). Notably, through preferential activation of TrkB-PLC-γ1-CREB-mediated signaling, antidepressants have been found to raise the level of synaptic plasticity in critical neural circuits to the degree required for relieving depressive symptoms, pointing out the significance of PLC-γ1/IP3/Ca 2+ pathway for antidepressants response [78][79][80]. Our results on the tyrosine-specific phosphorylation of TrkB (Tyr 706/707 and Tyr 816 sites, but not at Tyr 515 ) in the brain areas implicated in depression, caused by prolonged GSB-106 treatment in UCMS-exposed mice, are in line with the data observed on typical antidepressants, which provide TrkB-Tyr 816 /PLC-γ1 specific effects [18,19]. As was already mentioned, detailed animal studies revealed the substantial antidepressant-like activity of GSB-106 both in various tests (Porsolt, tail suspension, Nomura water wheel tests) and in animal depression models (chronic social defeat stress, LPS-induced anhedonia) [43][44][45][46]. Remarkably, it was clearly demonstrated that antidepressant-like properties of GSB-106 (acute, i.p.) depend on TrkB receptor activation and the PLC-γ1 signaling pathway in the Porsolt test in mice, as it was evident from the experiments with TrkB and PLC-γ1 inhibitors. GSB-106 (21 days, per os) has been shown to restore the reduced levels of synaptophysin and CREB in the hippocampus of C57BL/6 mice subjected to the 28-days chronic social defeat stress [45].
To our knowledge, this is the first report showing a site-selective phosphorylation of TrkB, elicited by small peptide or non-peptide BDNF analogs with antidepressant effects in the robust behavioral model of depression. Currently, several compounds with antidepressant properties were identified as putative ligands of the TrkB receptor, such as 7,8-dihydroxyflavone (TrkB agonist), ANA-12 (TrkB antagonist) and tetraterpenoid deoxygedunin. Zhang et al. demonstrated that chronic 7,8-dihydroxyflavone (28 days, i.p.) treatment induce antidepressant-like effect in mouse chronic mild stress model, accompanied by a restoration of BDNF level, TrkB phosphorylation at Tyr 706/707 residue, a recovery of the synaptic proteins PSD95 and synaptophysin levels in the prefrontal cortex in CMS-subjected animals [33]. The antidepressant properties of 7,8-dihydroxyflavone and ANA-12 as well as their ability to increase BDNF and pTrkB-Tyr 706/707 levels have also been demonstrated in a social defeat stress model and LPS-stimulated depressive-like behavior [81,82]. Another interesting fact is that 7,8-dihydroxyflavone treatment triggered phosphorylation of TrkB Y816 but not TrkB Y515 with subsequent activation of the PLC-γ1 pathway in primary striatal neurons and selectively reverted the decrease of pTyr 816 but not pTyr 515 in striatum of the chronically treated R6/1 transgenic mice [83].
Although the comprehensive evidence on the engagement of BDNF/TrkB signaling in antidepressant responses has been acquired, detailed mechanisms of this interaction remained unresolved until recently. Casarotto and colleagues reported the direct binding of both typical and rapid-acting antidepressants (in micromolar affinity) to the transmembrane region of TrkB dimers that promotes translocation and the synaptic localization of TrkB and facilitation of BDNF binding to TrkB and its activation. It is important to note that cholesterol has been shown to potentiate the effects of antidepressants and BDNF on TrkB signaling through the presumed impact on the conformation of TrkB dimers and the binding site of antidepressants. Furthermore, it was revealed that fluoxetine and ketamine increased Tyr 816 phosphorylation in primary cortical neurons, whereas fluoxetine, imipramine, ketamine, R,R-HNK enhanced TrkB interaction with PLC-γ1 [22]. To date, ability to directly interact with the extracellular domain of TrkB and promote receptor dimerization was reported only for 7,8-dihydroxyflavone amongst the low-molecular weight putative ligands of TrkB with antidepressant properties [32]. Whilst GSB-106 has been shown to mediate its neurotrophic [40,84], neuroprotective [85,86] and antidepressantlike [87] effects through the TrkB-dependent pathways, the mechanisms, underlying this interplay remained rather elusive, since we have not yet explored the direct association of GSB-106 with TrkB. What is important, GSB-106 could produce an indirect receptor activation, implicated Src-dependent TrkB transactivation in SH-SY5Y serum-deprived cells [42], and elucidation of this issue in animal models will provide novel knowledge about the mechanisms of small TrkB ligands antidepressant action.

Animals
The experiments were conducted on 128 male Balb/c mice weighing 22-25 g (Research Center of Biomedical Technology, Federal Medical and Biological Agency, Russia). The animals were group-housed under standard conditions (9-12 mice per home cage), with a 12-h dark-light cycle at a temperature of 22 ± 2 • C and ad libitum access to granulated chow (MEST, Russia) and water. All experimental procedures were in compliance with the 322 the EC Directive 86/609/EEC for animal experiments and were approved by the bioethics committee of the FSBI "Research Zakusov Institute of Pharmacology" (protocol No. 1 of 31 January 2021).

Substance
GSB-106 (MW 746.85; Figure 5) was synthesized as described previously [36] at Medicinal Chemistry Department of Zakusov Institute of Pharmacology.

Experimental Design
The UCMS model was used to induce chronic stress conditions [88]. While being maintained in a vivarium, animals were exposed to various stress factors, applied in a quasi-random manner (wet bedding, dirty boxes, water deprivation, reduction in daylight hours, etc.) within four weeks. Starting from day 29 stress factors application was accompanied by GSB-106, vehicle or amitriptyline treatment. The total duration of exposure to moderate chronic stress was either 42 or 54 days (Figure 1). The experimental design assumed the initiation of the treatment with the study drug after the manifestation of depressive-like responses in animals and its constant administration during the ongoing stress, which is usually applicable in similar experimental conditions and corresponds to the clinical situation. The in vivo experimental design was developed in compliance with the 3R principles. All compounds (GSB-106, amitriptyline) were dissolved in water immediately before administration. Compounds or vehicle were administered at a volume of 0.1 mL/10 g body weight.
The second experimental series consisted of two parallel stages-a behavioral assay and BDNF and pTrkB protein contents assessment (Figure 2b). Both stages began with the UCMS simulation for 28 days; on the 29th day of stress, the administration of GSB-106 or vehicle started (26 sequential per os treatments, 1 time per day) while stressing procedure was continued. 60 min after the 26th injection, 37 mice were subjected to Western-blot analysis. The following groups were formed: (1) vehicle, no stress (n = 9); (2) GSB-106 (1 mg/kg), no stress (n = 10); (3) vehicle, stress (n = 9); (4) GSB-106 (1 mg/kg), stress (n = 9). Behavioral tests were carried out with the animals of the second experimental cohort under a similar schedule for the control and the experimental groups. The Porsolt test was performed 60 min after 26th administration of the vehicle or GSB-106.

Porsolt Test
The antidepressant activity was studied in the Porsolt test in mice [89]. Cylindrical transparent Plexiglas tanks (30 cm height × 10 cm diameter) were filled with water (25 ± 1 • C) up to 20 cm from the bottom. On the test day mice were put individually in the cylinders for a 6-min swim session, and from the 4th min it was video-taped marking the duration of immobility periods. Immobility was defined as a lack of activity other than that required from the animal to keep its head above water: tail movements and limited limb movements.

Samples Preparation
All mice were decapitated by cervical dislocation and the brain was rapidly removed. The hippocampus and prefrontal cortex were dissected on wet ice covered with filter paper dampened in 0.32 M sucrose solution at a temperature of 0-4 • C. Each brain structure was frozen in liquid nitrogen (−196 • C), weighed, and stored at −80 • C.

Statistical Analysis
Analysis was performed using the GraphPad Prizm 7.0 software (GraphPad, La Jolla, CA, USA, www.graphpad.com). The normal distribution of the data was evaluated using the Shapiro-Wilk test. The results are presented as means ± SEM. The significance of intergroup differences was estimated by one-way analysis of the variance (ANOVA), followed by Tukey's multiple comparisons test.

Conclusions
Altogether, irrespective of the mode of GSB-106 interaction with the TrkB, prolonged per os treatment with dimeric dipeptide ameliorates depressive-like behavior in UCMSstressed mice. Remarkably, our findings strongly suggest that an increase in BDNF in the prefrontal cortex and hippocampus, as well as the restoration of phosphorylation at the TrkB Y706/707 (prefrontal cortex) and at the PLC-γ1 interaction site TrkB Y816 (prefrontal cortex and hippocampus) of UCMS mice might contribute to antidepressant-like mechanisms of GSB-106.

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets generated during the current study are available from the corresponding authors on reasonable request. Acknowledgments: Authors highly appreciate Tagir Kabirov for the scientifically-qualified language editing.

Conflicts of Interest:
The authors declare that they have no conflict of interests.