Adult male Sprague-Dawley (SD) rats (8–10 weeks old; Charles River Laboratories, Wilmington, MA, USA) weighing 250 to 280 g were pair-housed and maintained on a 12-h light-dark cycle with access to food and water ad libitum. All procedures were in strict accordance with Institutional Animal Care and Use Committee (IACUC) guidelines, and the use of laboratory animals was approved by the Hanyang University Animal Care and Use Committee. All animals were randomly assigned to one of the experimental groups.

3-(2-Carboxypiperazin-4-yl)propyl-1-phosphonic acid (CPP, Sigma) was dissolved in 0.9% saline at a concentration of 0.5 mg/kg and was injected intraperitoneally. In the controls, 0.9% saline was injected intraperitoneally.

To simulate CUS, animals were subjected to the exact sequence of 12 stressors (2 per day for 28 days) described in Koo et al. [15]: light on, light off, cage tilt, cage rotation, wet bedding, restraint, strobe, odor, food deprivation, overcrowding, cold (4℃), and isolation.

The FST is one of the most common behavioral experiments used to assess depression-like behavior in rodents. In this test, rats are forced to swim for 10 min in a cylinder from which there is no escape, after which the immobility time is measured. Immobility (passive) behavior was defined as floating without any motion, while swimming across the cylinder or climbing the wall of the cylinder was considered active behavior. All FSTs were recorded with a camcorder for subsequent video analysis.

The locomotor activity of rats was evaluated by an automatic video tracking system (SmarTrack®, Smartech, Madison, WI). Briefly, rats were placed in the marginal part of the open-field arena (76.5 cm×76.5 cm×40 cm) and allowed to explore the arena for 10 min. The total distance travelled in 10 min was measured by the video tracking system.

Animals were food deprived for 12 h and on the test day were placed in an open field (76.5 cm×76.5 cm×40 cm) with eight pellets of food in the center. The animals were given 8 min to approach the food and eat. The test was stopped as soon as the animal took the first bite. The latency to eat was recorded in seconds. Home cage food intake was also measured as a control.

Rat hippocampal tissues were fixed with 1% formaldehyde in PBS. Lysates were sonicated and extracted DNA was cut to lengths between 200-500 bps. The sonicated samples were rotated to preclear with protein A agarose (Roche Applied Sciences) for 1 h at 4℃ and anti-HDAC5 antibody (Abcam) was added and incubated overnight at 4℃ followed by incubation with fresh Protein A agarose for 2 h at 4℃. Precipitated chromatin complexes were eluted from the beads. Finally, the protein-DNA cross-links were incubated with protease K (Roche Applied Sciences) for 2 h at 42℃ and immunoprecipitated DNA was purified. Purified DNA samples were normalized and subjected to PCR analysis.

Immunoprecipitated DNA samples were resuspended in H₂O and fractions used for real-time PCR (iCycler, Bio-Rad, Hercules, CA, USA). Input (non-immunoprecipitated DNA) and immunoprecipitated DNA were PCR amplified in triplicate in the presence of SYBR Green (Bioline, London, UK). Ct values for each sample were obtained using Sequence Detector 1.1 software. QPCR data were analyzed by ΔΔCt method.

Equal amounts of protein extracts (20–25 µg) were mixed with SDS sample buffer and denatured at 95℃. The proteins were then electrophoresed on 7–12% polyacrylamide gels and transferred onto nitrocellulose membrane filters (Amersham Pharmacia Biotech, Amersham, UK). Blots were blocked with 1-5% non-fat milk in 1X TTBS (0.1% Tween-20 in 1X TBS) for 1 h at room temperature and incubated overnight at 4℃ with primary antibodies.

Virus was generated using a triple-transfection, helper-free method and purified according to a published protocol (Invitrogen). Briefly, HEK 293T cells (ATCC, Manassas, VA, USA) were cultured in five 150×25 mm cell culture dishes and transfected with the ViraPower lentiviral expression system (Invitrogen) using Lipofectamine (Invitrogen). Cells were collected, pelleted, and resuspended in buffer (0.15 M NaCl and 50 mM Tris, pH 8.0) at 66–70 h after transfection. After two freeze-thaw cycles to lyse the cells, benzonase was added (final concentration, 50 U/ml), and the mixture was incubated at 37℃ for 30 min. The lysate was added to PEG-it solution and processed according to the manufacturer's protocol (System Biosciences, Inc., Mountain View, CA, USA). The final purified virus was stored at −80℃.

Rats were anesthetized with xylazine hydrochloride (25 mg/kg) and tiletamine hydrochloride/zolazepam hydrochloride (50 mg/kg). Purified virus was injected bilaterally into the dentate gyrus (DG) at coordinates −4.1 mm (anterior/posterior), ±2.4 mm (lateral), and −4.6 mm (dorsal/ventral) relative to bregma. A total of 6 µl of purified virus was delivered at a rate of 0.1 µl/min, followed by 10 min of rest. The needle was removed, and the scalp incision was closed with wound clips. Rats in the same cage were randomly assigned to different experimental groups for the behavioral study; the order of testing was distributed across groups.

After animals were perfused with PBS, their brains were fixed overnight in 4% paraformaldehyde and then transferred to 30% sucrose. Brain sections (30 µm thickness) were cut using a microtome for visualization of EGFP. Gene transfer efficiency was assessed by immunofluorescence staining 4 weeks after surgery. EGFP staining was observed predominantly in dentate granule cells surrounding the infusion site in each hemisphere. Anti-GFP monoclonal antibodies were used (1:300, Roche). Slices were incubated in secondary antibodies conjugated to Alexa Fluor 488 (1:300, Invitrogen) and then mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA) to preserve fluorescence. Images were captured using a fluorescence microscope (Nikon, Tokyo, Japan).

All data were obtained from at least three independent experiments. Data are presented as mean±S.E.M. Statistical significance was calculated by unpaired Student's t-test. ^*p<0.05, ^**p<0.01, ^***p<0.001. For the behavioral experiment, the statistical significance of differences between the four groups was determined by analysis of variance (ANOVA; StatView 5, SAS Software, Cary, NC, USA) followed by LSD post hoc analysis.

We first investigated whether CPP induces HDAC5 phosphorylation and determined when CPP-mediated HDAC5 phosphorylation reached its peak level. To this end, 8-week-old rats were treated intraperitoneally for various lengths of time with 0.5 mg/kg CPP. This concentration was chosen because antidepressant-like behaviors were previously observed in the FST at 0.5 mg/kg [13]. Since HDAC5 phosphorylation at Ser259 and Ser498 has been shown to result in nuclear export of HDAC5 by 14-3-3 proteins [16], the levels of HDAC5 phosphorylation at Ser259 and at Ser498 were measured by immunoblotting. Intraperitoneal administration of CPP to the rats induced phosphorylation of HDAC5 at Ser259 and Ser498 in a time-dependent manner, with peak phosphorylation at approximately 0.5 h (Figs. 1A and B).

We next evaluated whether cytoplasmic localization of HDAC5 is triggered by CPP at 1 h. Rat hippocampal tissue extracts were separated into cytoplasmic and nuclear fractions, and the levels of HDAC5 phosphorylation at Ser259 and Ser498 were measured in the cytoplasmic fractions. The results showed that CPP rapidly increased phosphorylation of HDAC5 at Ser259 and Ser498 in the rat hippocampal cytoplasm by 1 h post-treatment (Figs. 1C and D). Therefore, these results indicate that CPP rapidly induces HDAC5 phosphorylation, and that the phosphorylated HDAC5 translocates to the cytoplasm. These observations suggest that phosphorylated HDAC5 initiates the transcription of genes associated with antidepressant-like effects.

Phosphorylation of HDAC5 has been reported to be regulated by CaMKII activity in neurons [1718] and by protein kinase D (PKD) in other cell types [1920]. To determine whether HDAC5 phosphorylation in response to CPP is associated with PKD and CaMKII in the rat hippocampus, the levels of phosphorylated PKD and CaMKII induced by CPP were investigated by immunoblotting. Western blotting revealed that CPP stimulated phosphorylation of PKD and CaMKII in hippocampus at 1 h post-treatment (Figs. 2A and B). The phosphorylation of PKD and CaMKII were increased in a time-dependent manner with a peak level at approximately 10-30 min after injection, a time point that is earlier than that of HDAC5 phosphorylation (Figs. 2C and D). These indicate that phosphorylation of PKD and CaMKII was increased before the induction of HDAC5 phosphorylation and suggest that CPP-mediated phosphorylation of HDAC5 might be regulated by a PKD- and CaMKII-dependent pathway.

BDNF plays a key role in neuroprotective and neurotrophic processes, which are relevant to stress and mood disorders [2122]. CPP rapidly increased the protein level of BDNF at 1 h (Figs. 2A and E). To determine whether CPP regulates BDNF gene expression through HDAC5 in hippocampal neurons, the DNA binding activity of HDAC5 to the Bdnf gene promoter was measured by ChIPs in hippocampus. HDAC5 was less enriched at the Bdnf promoter in hippocampal neurons treated with CPP than in those cells treated with saline (Fig. 2F). Given that association with HDAC5 in the specific promoter regions is linked to their repression, these results were consistent with the induction of BDNF proteins by CPP. These results suggest that CPP increases HDAC5 phosphorylation through a PKD- and CaMKII-dependent pathway, thereby inducing neuroprotective and neurotrophic processes that ameliorate depression-like behavior and CPP modulates BDNF transcriptional activity through HDAC5.

Exposure of rats to CUS has consistently been shown to induce cognitive impairment and to reduce the rate of body weight gain over the course of CUS exposure. These effects are similar to the components of major depression and anxiety disorders [23]. To investigate whether CPP has antidepressant-like effects on behavior, 8-week-old rats were exposed to a 4-week CUS regimen or a control period. The FST and NSFT were then conducted 1 h after intraperitoneal CPP injection (0.5 mg/kg) (Fig. 3A).

Consistent with a previous study [23], the mean body weight of the CUS group was significantly lower than that of the non-stressed group (Fig. 3B), indicating that the CUS-exposed rats were affected by the various stressors. In the FST, which uses decreased immobility time as an index of antidepressant efficacy, the immobility time was reduced by CPP in the nonstressed group and in the CUS group in the early stage (5 min) (Fig. 3C). However, at the late stage of the FST (10 min), the immobility time was only decreased by CPP in the CUS group (Fig. 3D). In the NSFT, which uses the latency period to feeding as an index of anxiety and antidepressant-like response, CPP reduced latency to feeding in the CUS group (Fig. 3E), consistent with a previous report on anxiolytic behavior in animals injected with ketamine [11]. There was no difference in the home cage food intake between groups (Fig. 3F). In the locomotor activity test (LMA), which uses total distance moved as an index of motor function, there was no difference between groups (Fig. 3G). When combined with the results demonstrating an effect of CPP on antidepressant-like behaviors in the FST, these behavioral results indicate that CPP rapidly induces an antidepressant response that is more effective under stress conditions.

To further address the role of HDAC5 in the behavioral response to chronic stress and CPP, a lentivirus expressing shRNA targeted against rat HDAC5 (lenti-shHDAC5) was infused into the DG, a hippocampal subregion whose volume is reduced in major depressive disorder (MDD) [24]. Next, rats were given a CUS regimen for 4 weeks (Fig. 4A). Given that HDAC5 is expressed in the granule cell layer of the hippocampus [25], lenti-GFP was infused into DG granule cells for rats in the control groups (Fig. 4B). Lenti-shHDAC5 infusions decreased the level of HDAC5 protein as assessed by immunoblotting (Fig. 4C) and also reduced the Hdac5 mRNA level in DGs as measured by quantitative PCR (Fig. 4D). To investigate whether lentivirus infusion affected rat motor function, a locomotor activity was tested. In the LMA, there was no difference between the lenti-GFP and lenti-shHDAC5 groups (Fig. 4E). This finding indicates that lenti-GFP- and lenti-shHDAC5-infused rats had identical motor function.

Consistent with the results shown in Fig. 3C, the immobility time during the initial 5 min period was reduced by CPP in animals infused with the lenti-GFP control virus (Fig. 4F). Rats injected with lenti-shHDAC5 showed behaviors similar to those on antidepressants. Compared to its effects in lenti-GFP-infused rats, CPP had no effects in lenti-shHDAC5-expressing rats. Since immobility in the FST was not affected (Figs. 4F and G), the antidepressant effects of CPP are likely blocked by lenti-shHDAC5 infusion. Furthermore, we found that viral-mediated hippocampal knockdown of HDAC5 induced Bdnf mRNA and protein expression, and blocked the enhancing effects of CPP on BDNF expression in stressed rats (Figs. 4H and I). Taken together, these results indicate that HDAC5 knockdown produces antidepressant effects and prevents the actions of CPP in chronically stressed animals.