Introduction

Major depressive disorder, commonly known as depression, is one of the most prevalent mental disorders worldwide. It affects an estimated 3.8% of the global population, approximately 280 million people1. Depression is characterized by persistent feelings of sadness, hopelessness, and loss of interest and/or pleasure in daily activities, and is often accompanied by changes in appetite, sleep disturbances, fatigue, and suicidal ideation2. The etiology of depression is complex, involving an interplay of biological and environmental factors. Biological factors implicated in the pathophysiology of depression include alterations in monoamine neurotransmission, abnormalities in the hypothalamus‒pituitary‒adrenal axis, reduced neuroplasticity, disruptions in circadian rhythms, inflammation, and dysfunction of neural networks3. Although conventional antidepressants such as selective serotonin (5-HT) reuptake inhibitors (SSRIs), 5-HT and noradrenaline reuptake inhibitors (SNRIs), and tricyclic antidepressants (TCAs) are widely used in the pharmacotherapy of depression, they have significant limitations. These limitations include delayed therapeutic onset, which can take weeks to months, low response rates, with approximately one-third of depressed patients being treatment resistant, and various side effects4. To overcome these limitations, there is an urgent need to develop more rapidly active and effective treatments for depression.

We have shown that the selective histone deacetylase 6 (HDAC6) inhibitor NCT-14b effectively alleviates behavioral despair in mice, as demonstrated by a significant decrease in immobility in the tail suspension test (TST) following the administration of NCT-14b5. HDAC6, a class II HDAC, is a multifunctional cytoplasmic deacetylase that regulates various cellular processes, including microtubule and actin cytoskeleton dynamics6,7, vesicle and mitochondrial transport8,9, protein quality control10,11,12, and stress responses via glucocorticoid receptor maturation13,14,15. The HDAC6 protein is widely distributed in mouse tissues, with the highest levels found in the brain and testis, and is predominantly expressed in dorsal raphe (DR) neurons in the brain. In postmortem human brains, HDAC6 is also expressed in the monoaminergic nervous system, including the raphe nucleus, substantia nigra, ventral tegmental area, and locus coeruleus5. However, the specific mechanism by which HDAC6 inhibition reduces behavioral despair in mice is not well understood. Hdac6 gene knockout (KO) mice show a similar reduction in immobility time to that of wild-type mice treated with NCT-14b in the TST5. Importantly, the immobility of Hdac6 KO mice was further decreased by the prior administration of conventional antidepressants including fluoxetine (an SSRI), imipramine, and desipramine (TCAs), with effects comparable to those in wild-type mice5. This result suggests that HDAC6 depletion alleviates behavioral despair through a mechanism distinct from that of conventional antidepressants. In addition, Hdac6 KO mice also exhibit increased exploratory behavior in novel environments and decreased anxiety-like behavior, which seems to be antithetical to depression5,16. These findings suggest that HDAC6 inhibitors have the potential to improve depression-related behaviors. In the present study, we employed tubastatin A (TubA), a commercially available and potent HDAC6 inhibitor with established in vivo use17,18, instead of NCT-14b. We investigated its effect on mice chronically treated with corticosterone (CORT), a well-established animal model of depression19,20.

Methods

Animals

Five-week-old male and female C57BL/6J mice, obtained from Japan SLC (Shizuoka, Japan), were maintained on a 12 h light/dark cycle (lights on at 7:00 AM) with ad libitum access to food and water, except during the period of chronic CORT treatment. To allow for acclimatization, mice were transferred to a controlled environment (24 ± 1 °C and 50–60% humidity) in the test room at least 30 min before the behavioral experiments. The animal experiments were all performed in accordance with the ARRIVE guidelines 2.0 (Animal Research: Reporting of In vivo Experiments) and approved by the animal experimentation committee of the Institute for Developmental Research at Aichi Developmental Disability Center (Permit number: 2019-003R6). All experimental procedures in this study were performed in accordance with relevant guidelines and regulations.

Chemicals

CORT, DMSO, and PEG300 were purchased from Sigma‒Aldrich (MO, USA). Polyoxyethylene 20 sorbitan monooleate (Tween-80) was obtained from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). TubA was purchased from Selleck Chemicals (TX, USA).

Chronic CORT treatment and drug injection

We dissolved CORT in ethanol and then mixed it with drinking water to a final concentration of 0.1 mg/ml CORT and 1% ethanol. To induce chronic CORT exposure, six-week-old male mice were group-housed three per cage. At seven weeks of age, they were given CORT-containing water in bottles for 21 consecutive days. The control group received drinking water supplemented with 1% ethanol instead of CORT. Individual CORT intake by mice was not directly measured. Instead, the volume of CORT-containing water consumed per cage was monitored every other day, confirming that the consumption was generally consistent across cages. One day before the first female encounter test (day 20 of CORT treatment), the subjects received an intraperitoneal injection of either 10 mg/kg TubA or a DMSO vehicle control (10 ml/kg body weight). We dissolved TubA in 5% DMSO/40% PEG300/5% Tween 80/50% H2O. The dose of TubA was determined based on previous studies of rodent administration17,18. After the first female-encounter test, the mice received regular drinking water.

Experimental groups

For behavioral experiments, mice were divided into three experimental groups (n = 8–12 per group): (1) a chronic CORT-treated group receiving DMSO (DMSO/CORT), (2) a chronic CORT-treated group receiving TubA-treated (TubA/CORT), and (3) an ethanol-treated control group receiving DMSO (DMSO/control). For biochemical analysis, separate cohorts of mice were subjected to the same chronic CORT/ethanol administration and subsequent DMSO/TubA injections as the mice used in the behavioral experiments. These mice were not subjected to any behavioral testing and were sacrificed for tissue collection at two time points post-injection: 24 h and 8 days (n = 4 per group at each time point).

Female-encounter test

We performed the female-encounter test as described by Ago et al.21, with minor modifications. We placed a sexually naïve male mouse (the resident) in the central chamber of an opaque, rectangular Plexiglas box (500 × 420 × 500 mm), which was divided into three interconnected chambers by transparent partitions (300 × 300 mm). Two small wire cages (100 mm diameter × 180 mm height) were positioned at the center of the left and right chambers against the wall (see Supplementary Fig. 1). After a 90-minute habituation period, we introduced unfamiliar, sexually naïve, age-matched male and female mice (the intruders) into the small wire cages. The resident mouse could then interact with them for a 10-minute test period. We recorded the resident mouse’s movements using a video camera mounted above the apparatus. We analyzed the distance traveled, time spent in each chamber, and contact time with the intruders (defined as the time spent within 2 cm of the intruder cages) via EthoVision XT v11.5 video tracking software (Noldus, Netherlands). Female preference was quantified as a percentage. Female preference= (time spent in the female chamber/total time spent in the male and female chambers) × 100. This preference is based on the data from the last 5 min of the test period, as preliminary experiments revealed that the tendency of a 9-week-old male mouse to stay in the female chamber was more pronounced during this time than during the entire 10-minute test period (Supplementary Fig. 1). We considered the total contact time with both intruders during the 10-minute test period as indicative of social interaction, while the distance traveled during the 90-minute habituation period served as an activity measure. The activity during the habituation period of the initial female-encounter test was interpreted as exploratory motivation toward a novel environment. The test was performed during the light cycle, with an illumination of 380 lx in the central chamber.

Forced swim test

Unlike the original protocol22, we conducted the forced swim test (FST) in a single test session without a pretest23. The test was performed a few hours after the second female-encounter test (8 days after drug administration). We filled a transparent cylinder (18.5 cm in diameter and 27 cm in height) with tap water at 24 ± 1 °C to a depth of 14 cm. We placed each test mouse in the cylinder and recorded it for 5 min using a side-positioned video camera. For data analysis, we employed the method described by Detke et al.24. Briefly, we divided the 5-minute test duration into 5-second intervals, and classified each interval as either active (swimming or climbing) or passive (immobility), which were then counted.

Western blot analysis

We euthanized each mouse by cervical dislocation and quickly removed the brain. We dissected coronal Sect. (4 mm thick) between 0.25 and 4.25 mm posterior to Bregma25 using a brain slicer (BS-2000 C, Braintree Scientific, MA, USA), weighed them, and stored them at -80 °C until analysis. This section encompasses the entire hippocampus, as well as portions of the overlying cortex, thalamus, hypothalamus, and midbrain. We homogenized the dissected brain tissue using a Dounce homogenizer in 10 volumes of homogenization buffer (10 mM HEPES, 1 mM EDTA, 320 mM sucrose, 1 mM sodium orthovanadate, 10 mM NaF, and protease inhibitor cocktail [Nacalai Tesque Inc., Kyoto, Japan], pH 7.4), followed by centrifugation at 1,000 × g for 10 min. We centrifuged the resulting supernatants at 34,000 × g for 30 min, and then subjected the pellets (containing the crude synaptosomal fraction) to Western blot analysis followed by densitometric analysis, as described previously16. We normalized the signal to β-actin as a loading control. However, due to the variability in optimal loading amounts required for different target proteins and antibodies, β-actin was not routinely included on every membrane. To ensure accurate quantification, protein loading was optimized for each target protein, and signal intensities were confirmed to be within the linear dynamic range (e.g., Supplementary Fig. 2). Minimal variations in loading and transfer were confirmed by performing multiple Western blots with identical samples and antibodies. These optimization and validation steps confirm the reliability of our protein quantification.

The following antibodies were used in this study: anti-HDAC67, anti-HSP90α/β (F-8, Santa Cruz Biotechnology, CA, USA), anti-α-tubulin (CLT9002, Cedarlane, Ontario Canada), anti-β-actin (AC-15, Abcam, MA, USA), anti-p44/42MAPK (ERK1/2) (137F5, Cell Signaling, MA, USA), anti-p-p44/42MAPK (T202/Y204) (D13.14.4E, Cell Signaling), and anti-TRKB (80E3, Cell Signaling).

Statistical analysis

Statistical analyses were performed using JASP statistical software (version 0.18.1). All data are presented as means ± standard errors of the means (SEMs). Groups were compared using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. A p value less than 0.05 was considered statistically significant. Detailed statistical analyses, including descriptive statistics and ANOVA results for Figs. 2, 3 and 4 and S3, are provided in (Supplementary Table 1).

Results

Rapid and long-lasting antidepressant effects of tubastatin A on depression-related behaviors in mice

To investigate whether HDAC6 inhibition ameliorates depression-related behaviors, we examined the effect of TubA on chronic CORT-treated mice, a mouse model of depression17,18. The time schedule for the experiments performed in this study is illustrated in (Fig. 1). After 21 days of CORT administration, the mice exhibited a decrease in spontaneous movement, impaired coordination, and poor coat conditions compared with the control mice (data not shown). We first assessed anhedonia, a core symptom of depression, via the female-encounter test21. This test evaluates reward-seeking behaviors in male mice, i.e., their preference for an unfamiliar female mouse when presented with both male and female mice simultaneously.

Fig. 1
figure 1

Time schedule for the induction of depression, drug administration, and behavioral tests.

Full size image

As expected, control mice injected with DMSO 24 h before testing clearly preferred females (control, 66.0 ± 4.7%, Fig. 2A). In contrast, chronic CORT-treated mice that were also injected with DMSO 24 h before testing presented a decrease in this preference (DMSO, 43.9 ± 4.8%). This loss of preference suggests the development of anhedonia. However, chronic CORT-treated mice receiving a single TubA injection demonstrated a significant preference for females (TubA, 64.1 ± 5.8%, p < 0.05). These findings suggest that TubA may exert a rapid antidepressant-like effect within 24 h, as evidenced by the restoration of preference for females. During the female-encounter test, we also evaluated exploratory motivation and social interaction. Chronic CORT treatment significantly reduced exploratory behavior (DMSO, 120.9 ± 14.1 m; control, 185.1 ± 5.7 m; p < 0.001 compared with the control; Fig. 2C), whereas a single TubA injection 24 h before the test significantly restored exploratory behavior (TubA, 165.2 ± 6.3 m; p < 0.01 compared with the DMSO group). Interestingly, compared with the control or DMSO, TubA significantly increased the social interaction time by 1.3-fold. Chronic CORT treatment itself did not affect social interaction time (control, 225.4 ± 7.9 s; DMSO, 222.6 ± 23.3 s; TubA, 290.0 ± 22.5 s, p < 0.05; Fig. 2E).

Fig. 2
figure 2

A single systemic injection of tubastatin A induces rapid and sustained antidepressant-like effects in the female encounter test. The effects on anhedonia, exploratory behavior, and social interaction were assessed at 24 h (A,C,E) and 8 days (B,D,F) after a single intraperitoneal injection of tubastatin A (TubA) or DMSO in chronic corticosterone (CORT)-treated mice. (A,B) Effects of TubA on anhedonia, measured as the preference for an unfamiliar female mouse over an unfamiliar male mouse. (C,D) Effects of TubA on locomotor activity during the 90-minute habituation period, considered as a measure of exploratory motivation in (C). (E,F) Effects of TubA on sociability, measured as the total social interaction time with both male and female intruder mice during the 10-minute test session. The data are expressed as the means ± SEMs (DMSO/control, n = 12; DMSO/CORT, n = 8; TubA/CORT, n = 9). Groups were compared using one-way ANOVA with Tukey’s post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001.

Full size image

We then examined the long-lasting effect of TubA. For this purpose, we conducted a second female-encounter test one week after the first test. In the chronic CORT-treated groups, the DMSO-injected mice still showed no preference for females (DMSO, 44.8 ± 6.2%, p < 0.05, compared with the control), whereas the TubA-injected mice again exhibited a clear preference for females (TubA, 68.3 ± 4.1%, p < 0.01, compared with the DMSO), similar to the control group (control, 62.4 ± 4.3%) (Fig. 2B). This result indicates that the TubA-induced recovery from anhedonia was sustained for 1 week. The activity levels and social interaction times did not significantly differ across all the groups (Fig. 2D,F).

To further investigate the effects of a single systemic injection of TubA on depression-related behaviors, we performed an FST a few hours after the second female-encounter test (Fig. 3). In the FST, we assessed active behaviors (climbing and swimming) and passive behavior (immobility). Chronic CORT administration significantly increased immobility (Fig. 3C) and caused a corresponding decrease in swimming time (Fig. 3B). TubA significantly counteracted these changes (Fig. 3A,B).

Fig. 3
figure 3

A single systemic injection of tubastatin A reduces behavioral despair in chronic corticosterone-treated mice in the forced swim test. The effects on climbing (A), swimming (B), and immobility (C) behaviors were evaluated 8 days after a single intraperitoneal injection of tubastatin A (TubA) or DMSO in chronic corticosterone (CORT)-treated mice using the forced swim test. The data are expressed as the means ± SEMs (DMSO/control, n = 12; DMSO/CORT, n = 8; TubA/CORT, n = 9). Groups were compared using one-way ANOVA with Tukey’s post hoc test. *p < 0.05.

Full size image

Effects of chronic CORT treatment and tubastatin a injection on biochemical changes in the mouse brain

Next, we investigated the biochemical changes in the mouse brain following chronic CORT treatment and a single systemic TubA injection. We prepared crude synaptosomal fractions from the brains of chronic CORT-treated mice at 24 h and 8 days after TubA or DMSO injection. We then quantified the protein levels of HDAC6, its substrates, and other proteins potentially associated with antidepressant mechanisms (e.g., extracellular signal-regulated kinase [ERK] and tropomyosin receptor kinase B [TRKB]) in this fraction via Western blotting (Fig. 4).

Chronic CORT treatment significantly decreased the protein level of ERK (Fig. 4A; ERK: DMSO 64%). This reduction was reversed 24 h after TubA injection (ERK: TubA 89%). Additionally, the activation level of ERK, as indicated by the p-ERK/ERK ratio, increased 1.3-fold 24 h after TubA injection compared with that following DMSO injection (p-ERK/ERK: DMSO 96%, TubA 131%). However, these changes in ERK1/2 were not observed 8 days after TubA injection (Fig. 4B). Chronic CORT treatment also decreased the levels of HDAC6 and full-length TRKB140k, and these reductions were not reversed 24 h after TubA injection (Fig. 4A; HDAC6: DMSO 66%, TubA, 61%; TRKB140k: DMSO 74%, TubA 71%). In contrast, the level of the truncated TRKB isoform (TRKB80k) was 1.3-fold greater 24 h after TubA injection than after DMSO injection (TRKB80k: DMSO 88%, TubA 118%). The levels of HDAC6 and TRKB140k,80k were slightly increased 8 days after TubA injection compared with DMSO injection (Fig. 4B; HDAC6: DMSO 93%, TubA, 106%; TRKB140k: DMSO 89%, TubA 102%; TRKB80k: DMSO 72%, TubA 95%). On the other hand, the level of α-tubulin, a major HDAC6 substrate and a component of microtubules, increased 1.6-fold 24 h after TubA injection compared with DMSO injection (Fig. 4A; α-tubulin: DMSO 85%, TubA 132%). At 8 days after drug injection, the α-tubulin levels were comparable between the groups. We observed no significant changes in the levels of HSP90, another HDAC6 substrate. We also assessed the acetylation levels of HDAC6 substrates (α-tubulin and HSP90) in the brain at either 24 h or 8 days after TubA injection, but detected no significant increase (data not shown). In contrast, we observed a modest yet significant increase in α-tubulin acetylation 30 min post-TubA administration in chronic CORT-treated mice (Supplementary Fig. 3). Given the pivotal role of synapses in neuronal communication, we focused our analysis on the crude synaptosomal fraction, which is enriched in synaptic proteins. Notably, we did not observe significant changes in the aforementioned proteins in unfractionated brain extracts (data not shown).

Fig. 4
figure 4

A single systemic injection of tubastatin A affects the protein levels of ERK, TRKB, HDAC6, and α-tubulin in the brains of chronic corticosterone-treated mice. The effects on the protein levels of Hdac6, Hdac6 substrates (α-tubulin and HSP90), ERK, phosphorylated ERK (p-ERK), TRKB isoforms (TRKB140k and TRKB80k) were evaluated at 24 h (A) and 8 days (B) after a single intraperitoneal injection of tubastatin A (TubA). Protein levels were quantified by Western blotting (representative blots on the left), followed by densitometric analysis (graphs on the right). Signal intensities were normalized to β-actin. The data are expressed as the means ± SEMs (DMSO/control, DMSO/CORT, TubA/CORT, n = 4) relative to the DMSO/control group. Groups were compared using one-way ANOVA with Tukey’s post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001. The blots were cropped and full-length blots are presented in (Supplementary Fig. 2).

Full size image

Discussion

In this study, we demonstrated that a single systemic administration of TubA, a specific HDAC6 inhibitor, rapidly alleviated anhedonia in chronic CORT-treated mice within 24 h. These effects were sustained for at least one week. Moreover, TubA not only improved anhedonia but also restored exploratory behavior and reduced behavioral despair, further validating its antidepressant properties. The strong activation of ERK observed 24 h after TubA administration indicates its involvement in TubA’s antidepressant-like effects.

Previous studies have shown that chronic administration of the HDAC6 inhibitor ACY-738 produces antidepressant effects in mice subjected to chronic social defeat stress15,18. In contrast, our study is the first to demonstrate the rapid antidepressant effects of the HDAC6 inhibitor TubA in a chronic CORT-treated mouse model. Considering TubA’s short plasma half-life18 and the lack of enhanced acetylation of HDAC6 substrates 24 h post-injection, our findings suggested that transient HDAC6 inhibition immediately following TubA administration triggered biochemical changes in the brains of chronic CORT-treated mice, leading to antidepressant effects.

The most significant biochemical change observed in the brains of chronic CORT-treated mice after TubA administration was the activation of ERK. This activation was observed 24 h post-administration (Fig. 4A). ERK signaling is crucial in depression. Consistent with this, inhibiting the ERK pathway in the prefrontal cortex or hippocampus, two key brain regions implicated in depression, induces depression-like behavior, whereas various antidepressants exert their behavioral effects partly by normalizing downregulated ERK activity26,27,28. Indeed, ERK is activated downstream of the brain-derived neurotrophic factor (BDNF)-TRKB signaling pathway, which is known to play critical roles in neuronal survival, growth, and synaptic plasticity. These processes are believed to contribute to antidepressant action29. Approximately one-third of total mitogen-activated protein kinases (MAPKs), including ERK, are associated with microtubules in NIH3T3 mouse fibroblasts30. HDAC6, a major deacetylase of α-tubulin at lysine 40 (K40), generally stabilizes microtubules through K40 acetylation31. Given the increase in α-tubulin levels 24 h after TubA administration in the brains of chronic CORT-treated mice (Fig. 4), HDAC6 inhibition may stabilize microtubules and enhance ERK scaffolding, creating a more favorable environment for ERK activation. Notably, one week after drug administration, despite a significant difference in depression-related behaviors between the DMSO- and TubA-injected groups (Figs. 2 and 3), no difference in ERK activation levels was observed. These results suggest that transient ERK activation 24 h after TubA administration is sufficient to initiate antidepressant action, and that sustained high-level ERK activation is not invariably required for the maintenance of normal behaviors such as a preference for females or persistence in escape attempts. Future studies are needed to elucidate the mechanisms underlying the biochemical changes observed 24 h after TubA administration in chronic CORT-treated mice, including the recovery of ERK phosphorylation and the increase in α-tubulin levels.

Inhibition of HDAC6 by TubA potentially impacts both the dopaminergic and serotoninergic systems. The mesolimbic dopamine reward circuit has been implicated in depressive-like behaviors, antidepressant actions, and motivational control16,32,33. In addition, dopamine receptors, particularly D1 and D2, have been found to activate the ERK pathway in various brain regions34,35. We observed that TubA administration restored female preference (Fig. 2A,B) and exploratory behavior (Fig. 2C) in chronic CORT-treated mice, and led to ERK activation. Our findings suggest that the dopamine reward system may be a promising target for TubA. Furthermore, the serotonergic system is also a potential target of TubA. 5-HT is a key neurotransmitter that modulates various physiological and cognitive pathways, including those affecting emotions, memory, sleep, and thermal regulation. Disruptions in 5-HT transmission are commonly implicated in the development of depression3. In fact, susceptibility to anhedonia in mice correlates with the plasticity of 5-HT neurons in the DR36. Notably, HDAC6 is prominently expressed in 5-HT neurons of the DR in humans and mice, and Hdac6 KO mice exhibit reduced behavioral despair and anxiety5. Additionally, 5-HT neuron-selective Hdac6 KO mice are resistant to chronic social defeat stress, with no impairment in social interaction14. Given that social interaction is an index of anxiety-like behavior37, the sociability-enhancing effects of TubA on chronic CORT-treated mice (Fig. 2E) suggest that TubA also exerts anxiolytic effects. Furthermore, TubA improved “swimming” rather than “climbing” in the FST, suggesting enhanced 5-HT rather than noradrenaline neurotransmission (24, Fig. 3). Interestingly, 5-HT also activates the ERK pathway through various receptors, leading to different physiological and behavioral effects38. Taken together, these findings suggest that the antidepressant effects of TubA are likely mediated through both the dopaminergic and serotonergic systems.

The observed rapid and sustained antidepressant effects of TubA, as demonstrated by the female-encounter test, parallel the pharmacological profile of ketamine, an antagonist of the N-methyl-D-aspartate receptor, rather than SSRIs21,24,39,40,41. The mechanistic action of ketamine on striatal GABAergic neurons, which involves the activation of glycogen synthase kinase 3 β (GSK3β) and the inhibition of HDAC642, suggests a potential mechanistic overlap with TubA. Interestingly, chronic CORT administration significantly decreased HDAC6 levels in the mouse brain (Fig. 4). However, whether HDAC6 downregulation is commonly involved in the pathogenesis of depressive disorders remains unclear. Nevertheless, given the antidepressant-like effects of HDAC6 inhibition, HDAC6 downregulation may function as a protective coping mechanism against chronic stress.

In summary, HDAC6 inhibition is likely to influence multiple mechanisms of antidepressant action, possibly involving the activation of ERK signaling. While we cannot completely rule out the possibility of off-target effects of TubA, our findings strongly suggest that the antidepressant effects of TubA are primarily mediated by HDAC6 inhibition, as supported by the parallel between the behavioral phenotypes of Hdac6 KO mice and the effects of TubA in the chronic CORT-treated mouse model. Future studies investigating the effects of various HDAC6 inhibitors in different models of depression will be crucial to further elucidate the specific mechanisms underlying the rapid and long-lasting antidepressant effects of TubA. These studies will also help to definitively exclude the contribution of potential off-target effects.