MGCD0103, a selective histone deacetylase inhibitor, coameliorates oligomeric Aβ25-35 -induced anxiety and cognitive deficits in a mouse model.

AIMS: Recently, histone deacetylase (HDAC) inhibitors are considered a possible therapeutic strategy in Alzheimer's disease (AD). However, HDACi treatments exhibit diverse functions with unfavorable effects in AD. Thus, the development of selective HDACi without side effects is urgently needed.
METHODS: HDACi, namely, BML210, MGCD0103, PXD101, and Droxinostat, were screened in mouse hippocampal primary cultures incubated with oligomeric Aβ25-35 (50 μmol/L). MGCD0103 was chosen for in vivo tests and was intraperitoneally injected into C57BL/6J mice (0.5 mg/kg, once per day) for 4 weeks following an intrahippocampal CA1 injection of oligomeric Aβ25-35 . Brain samples were collected for pathological analyses after the behavioral analyses including open- field test (OFT), elevated plus maze (EPM), Y-maze, and Morris water maze (MWM).
RESULTS: Among the HDACi, MGCD0103 exhibited significant neuroprotection against the Aβ toxicity in primary cultures. MGCD0103 coattenuated cognitive deficits and anxiety against Aβ damage in mice. MGCD0103 further ameliorated pathological features such as the levels of acetylated histone 3 at Lys 9 site (H3K9) and α-tubulin, synaptophysin, Aβ, tau protein phosphorylation, and serotonergic neuron loss against Aβ toxicity. Furthermore, chronic MGCD0103 treatment did not show liver or kidney toxicity in mice.
CONCLUSIONS: These results reveal MGCD0103 could be a potential therapeutic agent against AD.

INTRODUCTION

Alzheimer's disease (AD) is the most common neurodegenerative disease and the major cause of dementia in the aging population. The major pathological features of AD are extracellular senile plaques of aggregated Aβ and hyperphosphorylated tau protein‐enriched intraneuronal neurofibrillary tangles (NFT) in brain tissue, including in the temporal lobe, limbic system, and neocortex.1, 2 However, the development of therapeutic AD drugs, including anti‐Aβ, anti–tau protein phosphorylation, antioxidants, and anti‐inflammatory agents, has been disappointing for the multifactorial elements of AD.3, 4, 5, 6 Therefore, ongoing investigations and promising new ways of understanding the puzzle pieces of AD are urgently needed.

Environmental factors such as stress, anxiety, and epigenetic alterations have been reported to play a significant role in the pathogenesis of sporadic or late‐onset AD (LOAD).7, 8, 9, 10 Epigenetic mechanisms regulate gene expression through histone acetylation and DNA methylation of normal nucleotide sequences of a gene. There are four families of histone deacetylase (HDAC) enzymes encoded from both the human and rodent genomes: class I (including HDAC 1‐3 and 8), class IIa (HDAC 4, 5, 7, and 9), class IIb (HDAC 6 and 10), and class IV (HDAC11).11 Evidence further suggests that the development of the epigenetic drugs may provide a new direction in AD.12, 13 Previous experimental treatments with pan‐HDAC inhibitors (nonselective HDAC inhibitors, HDACi), such as sodium butyrate, trichostatin A, suberoylanilide hydroxamic acid, or sodium phenylbutyrate, have shown promising therapeutic effects in animal models of AD and neuropsychiatric disorders.14, 15, 16 However, pan‐HDACi should be used with caution because of the unfavorable side effects that they can cause in the brain.17 Therefore, the development of selective HDACi with the reduction in unwanted side effects needs to be addressed in AD treatments.

In this study, we first identified a lead compound selected from four HDACi that would serve as an appropriate chemical probe to test its therapeutic potential in AD treatment. In mouse primary hippocampal neuronal cultures, we found that MGCD0103 showed neuroprotective effects superior to those of the three other HDACi. MGCD0103, an isotype‐selective HDACi, has been clinically evaluated for the treatment of tumors18 and has also shown a promising pharmacological profile as an anticancer agent, including reliability, tolerance, and safety.19, 20 Therefore, the potential therapeutic effects and mechanisms of MGCD0103 were characterized in mice received with an acute bilateral intrahippocampal CA1 injection of oligomeric Aβ25‐35. Our data show that MGCD0103 can effectively and safely attenuate the anxiety and cognitive deficits associated with an increase in acetylated histone and nonhistone (α‐tubulin) proteins and can ameliorate the pathological features of AD, including Aβ deposition, tau protein phosphorylation, gliosis, and loss of serotonergic neurons and synapse‐related proteins. Based on these results, we suggest that MGCD0103 could be developed as a novel therapeutic approach for AD.

METHODS

Animals and ethics statement

All experimental procedures using mice received specific approval from the Institutional Animal Care and Use Committee (IACUC) of the National Taiwan Normal University, Taipei, Taiwan (Permit Number: 104010). C57BL/6J mice were purchased from the National Breeding Centre for Laboratory Animals. In vivo animal experiments were carried out with mature (12‐week‐old) males, and pregnant females were utilized for primary hippocampal cultures. Mice were maintained on a 12 hours light/dark cycle with a constant ambient temperature (20‐25°C) and relative humidity (60%), with food and water available ad libitum. Every effort was made to minimize any suffering to the animals.

Preparation of Aβ25‐35 oligomer and HDAC inhibitors

Aβ25‐35 peptides were purchased from Sigma (SI‐A4559, Saint Louis, MO 63103, USA), solubilized in sterile distilled water (3 mg/mL), and stored at −20°C until use. For in vitro use, Aβ25‐35 peptides were diluted to 50 μmol/L with physiological saline and aggregated by incubation at 37°C for 4 days before use, as previously described.21 For in vivo use, Aβ25‐35 peptides were diluted to 10 nmol/L with physiological saline and aggregated by incubation at 37°C for 7 days before bilateral injection into the hippocampal CA1 subregion (i.c.). The IC50 values of four HDACi (BML210; MGCD0103; PXD101; Droxinostat) were determined by an MTT assay (see Supporting Information) using a SH‐SY5Y neuroblastoma cell line (Table S1). HDACi were diluted with DMSO and applied to the primary culture at three different concentrations (Table S1; low, medium, and high).

Administration of MGCD0103

The elimination half‐life in plasma of MGCD0103 is 7‐11 hours.22, 23, 24 Chronic intrathecal delivery (4 weeks Alzet minipumps) of MGCD0103 (30 or 60 nmol/d) was administrated into the spinal cord of nerve injury rats.25 In addition, four HDACi (BML210, MGCD0103, PXD101, Droxinostat) in three different doses (high, medium, and low dose based on IC50) were screened using mouse primary hippocampal neuronal culture (Figure S1). We found that the neuroprotective effect of MGCD0103 was superior to the other three HDACi, especially in low and high doses (Figure S1). Furthermore, the performance in neuroprotection of MGCD0103 in high dose was better than low dose (Figure S2). Therefore, the in vivo dose of MGCD0103 (0.5 mg/kg; Biovision) was determined based on the results of the primary hippocampal neuronal cultures (high dose) and evidence of its ability to penetrate the BBB.23, 24 MGCD0103 was first dissolved in DMSO (40 mg/mL) and then diluted with a mixed buffer (saline:Kolliphor, 4:1; Kolliphor (C5135; Sigma‐Aldrich)) to 0.5 mg/kg before intraperitoneal injection (i.p.) in mice once a day for 28 days following stereotaxic surgery. The toxicity of chronic MGCD0103 injections in liver and kidney was assessed by hematoxylin and eosin (HE) staining (Table S2).

Experimental design

Six days after adaption in an animal room, mice were anesthetized on day 7 by an intraperitoneal injection of avertin (0.4 g/kg body weight; Sigma, Saint Louis, MO 63103, USA) and mounted in a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA). 3 μL of either saline or oligomeric Aβ25‐35 (10 nmol/L) was injected bilaterally into the hippocampal CA1 with a 10.0 μL Hamilton syringe using the following coordinates: anteroposterior −0.23 mm from bregma, mediolateral ±0.2 mm from midline, depth −0.15 mm from skull. After surgery recovery, mice were administered MGCD0103 (0.5 mg/kg, i.p.) for 28 days (days 8‐36). Mice were divided into four groups: (a) saline (i.c.)/saline (i.p.); (b) saline (i.c.)/MGCD0103 (i.p.); (c) Aβ25‐35 (i.c.)/saline (i.p.); or (d) Aβ25‐35 (i.c.)/MGCD0103 (i.p.). During MGCD0103 treatment, mice underwent a series of behavioral tests: an open‐field test (OFT) (day 24), an elevated plus maze (EPM) test (day 26), a Y‐maze test (day 28), and a Morris water maze (MWM) test (days 30‐36). After behavioral testing, mice were sacrificed for pathological analyses on day 37.

Evaluation of behavioral tasks

During the MGCD0103 treatment period, mice were evaluated in OFT, EPM, Y‐maze, and MWM tests. Before the behavioral tests, mice were acclimatized to the behavioral testing room for 30 minutes. All experiments were completed between 7:00 and 17:00, and the behavioral performance of the mice was recorded and analyzed by an automatic tracking system (EthoVision‐XT; Noldus, Wageningen, Netherlands). The test box was sequentially cleaned with 70% and 30% ethanol before the next mouse was introduced to the box.

Open‐field test

The OFT was conducted as previously described.7 Mice were individually placed in the central zone (15 × 15 cm) of a white acrylic box (30 × 30 × 30 cm) and then recorded for 10 minutes with a video camera. Anxiety‐like phenotypes were evaluated as a decrease in the time spent in the center of the open field arena.

Elevated plus maze

The maze consisted of four arms (30 × 5 cm) extending out from a central square (10 × 10 cm), where two arms had 15‐cm high walls and were 50 cm above floor level. Mice were placed in the central square facing one of the open arms, and their activity was video tracked for 5 minutes. The total time that mice spent in the open arms was determined and used as an index of anxiety.

Y‐maze

The Y‐maze apparatus consisted of 3 identical symmetric arms (46 × 3 × 17 cm) with white acrylic walls. The Y‐maze task was conducted as described.26 Each mouse was placed in the central space and allowed to explore freely for 8 minutes, while their rates of spontaneous alternations were assessed as an indicator of short‐term memory. Arm entry was defined as the entry of all four paws into one arm. Spontaneous alternation percentage was calculated by the equation: [successive entries/(total arm entries − 2) × 100].

Morris water maze

The spatial learning and memory abilities of the mice were assessed through the MWM task as previously described.7 Before the training phase, the swimming ability of the mice was tested on day 30. Mice that floated in the pool were not enrolled in the task. In the training phase, each mouse underwent four 1‐minute trials to locate the hidden platform on 4 consecutive days (days 31‐34). The time to climb onto the hidden platform was recorded as the escape latency. On day 35, three testing trials were performed as an index of spatial learning acquisition. Twenty‐four hours after the last testing trial (day 36), a probe test was performed. The platform was removed from the pool, and each mouse was allowed to swim freely for 60 seconds. The amount of time spent in the quadrant of the original platform location was calculated as an index of long‐term spatial memory.

Immunohistochemistry (IHC)

Immunohistochemistry staining was performed as previously described.7 Mice (n = 3‐5 per group) were deeply anesthetized with avertin (0.4 g/kg body weight) and then transcardially perfused with normal saline followed by 4% paraformaldehyde (PFA) in 0.1 mol/L phosphate buffer. Harvested brain tissues were postfixed in the same solution at 4°C for 4 hours and then dehydrated in a graded series of sucrose solutions until they were fully permeated. Frozen brain sections were cut 30‐μm thick in a coronal plane using a cryostat microtome (CMS3050S, Leica Microsystems, Nussloch, Germany), and IHC staining was performed on 3‐4 sections per mouse. Slices were incubated with the primary antibodies as Table 1 overnight at room temperature and the appropriate biotinylated secondary antibodies (1:200, Vector Laboratories, Burlingame, CA, USA) for 1 hour at room temperature, followed by incubation with an avidin‐biotin‐peroxidase complex (ABC kit; Vector Laboratories, Burlingame, CA, USA) for 1 hour at room temperature. After rinsing with PBS (10 minutes, three changes), the sections were incubated with a diaminobenzidine (DAB) kit (Vector Laboratories, Burlingame, CA, USA). Slices were stained and observed using a light microscope (Leica, Wetzlar, Germany). Images were obtained using Image‐Pro Plus 5.1 (Image‐Pro Plus Media Cybernetics, Washington, MD, USA), and the threshold intensity was manually set at a constant value for all images. Pixel counts were derived from the average of three adjacent sections per animal.

Table 1

List of primary antibodies in immunohistochemistry and Western blot

Antibody	Species	Supplier	WB dilution	IHC dilution
Iba‐1	Rabbit	Wako, Osaka, Japan		1:1000
GFAP	Mouse	Millipore, Temecula, CA, USA		1:1000
5‐HT	Rat	Millipore, Temecula, CA, USA		1:200
ChAT	Rabbit	Millipore, Temecula, CA, USA		1:1000
TH	Rabbit	Millipore, Temecula, CA, USA		1:1000
6E10	Mouse	Covance, Princeton, NJ, USA		1:1000
pS202Tau	Rabbit	AnaSpec, San Jose, CA, USA		1:1000
pS9GSK3β	Rabbit	Cell Signaling, Danvers, MA, USA	1:1000
GSK3β	Rabbit	Cell Signaling, Danvers, MA, USA	1:1000
pT216GSK3β	Mouse	Millipore, Temecula, CA, USA	1:1000
pT181Tau	Rabbit	Millipore, Temecula, CA, USA	1:1000
pS396Tau	Rabbit	Invitrogen, Rockford, USA	1:1000
pT205Tau	Rabbit	Invitrogen, Rockford, USA	1:1000
Total tau (HT7)	Mouse	Thermo, Rockford, IL, USA	1:500
BACE1	Rabbit	Cell Signaling, Danvers, MA, USA	1:1000
Synaptophysin	Mouse	Abcam, Cambridge, UK	1:1000
PSD95	Goat	Santa Cruz, Dallas, TX, USA	1:1000
ERK	Rabbit	Cell Signaling, Danvers, MA, USA	1:1000
pERK	Rabbit	Cell Signaling, Danvers, MA, USA	1:1000
CDK5	Mouse	Millipore, Temecula, CA, USA	1:1000
pCDK5	Mouse	Santa Cruz, Dallas, TX, USA	1:1000
IDE	Rabbit	Abcam, Cambridge, UK	1:1000
NEP	Mouse	Santa Cruz, Dallas, TX, USA	1:1000
AcH3K9	Rabbit	Cell signaling, Danvers, MA, USA	1:1000
Histone 3	Rabbit	Millipore, Temecula, CA, USA	1:1000
Acα‐tubulin	Mouse	Santa Cruz, Dallas, TX, USA	1:1000
α‐tubulin	Mouse	Sigma, Bath, UK	1:5000
β‐actin	Mouse	Millipore, Temecula, CA, USA	1:2000

Western blot

Three to five mouse hippocampi from each group were collected and homogenized in RIPA lysis buffer (1:2, w/v). An aliquot of 20 μg protein was used for Western blot analysis as previously described.7 After blocking, membranes were incubated with the primary antibodies listed in Table 1 at 4°C for 16 hours. Next, membranes were incubated with the secondary antibodies corresponding to the type of primary antibody (1:10 000, Amersham Pharmacia Biotech, Piscataway, NJ, USA) for 2 hours at room temperature and developed using an ECL kit (ECL, Amersham, MA, USA). β‐actin was used as a loading control. Blots were scanned by a LAS‐4000 chemiluminescence detection system (Fujifilm, Tokyo, Japan), and the quantification of band density was performed using Gel‑Pro Analyzer software (GelPro32, version 4.0; Media Cybernetics, Inc., Rockville, MD, USA).

Statistical analysis

Data are shown as the mean ± SEM. Data were analyzed by SPSS 20.0 statistics software (Armonk, New York, USA) using one‐way ANOVA or Student's t tests. P < 0.05 was considered to be statistically significant.

RESULTS

MGCD0103 attenuated the anxiety and cognitive deficits induced by oligomeric Aβ25‐35

To investigate whether MGCD0103 attenuated the anxiety and cognitive deficits induced by oligomeric Aβ25‐35, we performed OFT, EPM, Y‐maze, and MWM experiments. The total distance travelled in the OFT was not different between the four groups (Figure 1A), which indicates that mice from all groups had normal motor activity. Despite this normal motor activity, the oligomeric Aβ25‐35 significantly lowered the cumulative time spent in the center zone compared to the intrahippocampal injection of saline (P < 0.01; Figure 1B). However, the administration of MGCD0103 significantly increased the cumulative time spent in the center zone compared to the saline treatment of mice who received oligomeric Aβ25‐35 injections (P < 0.001; Figure 1B). In the EPM task, the oligomeric Aβ25‐35‐treated group showed significantly reduced total time spent in the open arms compared to the saline‐treated group (P < 0.001; Figure 1C). MGCD0103 treatment induced a significant increase in the cumulative time spent in the open arms compared to saline treatment for mice who received oligomeric Aβ25‐35 injections (P < 0.05; Figure 1C). These results suggest that the treatment of MGCD0103 ameliorated the anxiety caused by oligomeric Aβ25‐35 injections.

Figure 1

Effects of MGCD0103 on Anxiety and Cognitive Deficits in Oligomeric Aβ25‐35‐Treated Mice. A, Motor activity in an open‐field test (OFT). There was no difference among the four groups. B‐C, Time spent in the central zone in the OFT and overall time spent in the open arms of the EPM, which were evaluated as indexes of anxiety. MGCD0103 reduced anxiety in oligomeric Aβ25‐35‐treated mice. D, The spontaneous alternation rate of mice in the Y‐maze was measured as an index of short‐term memory. MGCD0103 reduced the short‐term memory deficits induced by oligomeric Aβ25‐35. E, Swimming velocity was measured on the pretraining day of the MWM task. No differences were identified among the four groups of mice. F, The learning profiles of the four groups of mice during the 4 training days. MGCD0103 improved learning abilities in the oligomeric Aβ25‐35 group. G, Latency to identify the platform during the testing phase. MGCD0103 treatment reduced the time for the mice to climb onto the platform in the oligomeric Aβ25‐35 group but increased spatial learning acquisition impairments in the saline group. H, Results of long‐term memory retrieval testing. Oligomeric Aβ25‐35 impaired long‐term memory retrieval, and MGCD0103 attenuated this deficit. Data are presented as the mean ± SEM for each group. n = 12‐15. *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (#, P < 0.05; **, ##, P < 0.01; ***, ###, P < 0.001)

In addition, the oligomeric Aβ25‐35‐treated group exhibited significantly reduced short‐term memory compared with the saline‐treated group in the Y‐maze test (P < 0.001; Figure 1D). However, the recognition memory performance inhibited by oligomeric Aβ25‐35 was effectively restored by the i.p. administration of MGCD0103 (P < 0.001; Figure 1D). The treatment with MGCD0103 recovered the short‐term memory damage induced by oligomeric Aβ25‐35.

In the MWM task, the time for the mice to reach the hidden platform was compared among the four groups (Figure 1E‐H). First, no significant difference in swimming speeds was found among the four groups of mice (Figure 1E). In the 4‐day positioning navigation test, the escape latency of the mice was not significantly decreased by oligomeric Aβ25‐35 treatment following training days 1‐4 (F 3,59 = 0.68, P > 0.05; Figure 1F). Compared with the saline‐treated group, the group administered Aβ25‐35 had a significant longer escape latency on all 4 days (P < 0.001; Figure 1F), indicating that the administration of Aβ25‐35 induced a learning deficit. This deficit was ameliorated by the administration of MGCD0103 on days 2‐4 (P < 0.05‐0.01; Figure 1F).

On the 5th day of the MWM (testing phase), mice from the Aβ25‐35‐treated group spent more time locating the platform than did saline‐treated mice (P < 0.001; Figure 1G), indicating deficits in memory acquisition. The administration of MGCD0103 significantly improved the memory acquisition damage induced by oligomeric Aβ25‐35 (P < 0.05; Figure 1G). In the saline‐treated group, MGCD0103 treatment significantly impaired memory acquisition (P < 0.01; Figure 1G), which suggests that MGCD0103 should not be used before the onset of disease. On the 6th day of the MWM (probe phase), deficits in memory retention were observed in the oligomeric Aβ25‐35‐treated group (P < 0.001; Figure 1H), and the administration of MGCD0103 improved memory retention (P < 0.001; Figure 1H). The results obtained from these different behavioral tasks suggest that the administration of MGCD0103 attenuated the anxiety and cognitive deficits in oligomeric Aβ25‐35‐treated mice.

MGCD0103 decreased Aβ levels via increasing the expression levels of IDE in oligomeric Aβ25‐35‐treated mice

Immunohistochemistry staining revealed the extensive Aβ deposition in the hippocampus of oligomeric Aβ25‐35‐treated mice compared to saline‐treated mice (P < 0.01; Figure 2A,B). Furthermore, the deposition of Aβ was significantly decreased in the hippocampus of the MGCD0103‐treated group under oligomeric Aβ25‐35 conditions (P < 0.01, Figure 2A,B). In addition, BACE1, IDE, and NEP were also evaluated to elucidate the effects of MGCD0103 on the metabolism of Aβ. The levels of BACE1 in the oligomeric Aβ25‐35‐treated group were significantly higher than those of the saline‐treated group (P < 0.01; Figure 2C,D); however, there was no change following the administration of MGCD0103 (Figure 2C,D). In addition, MGCD0103 markedly enhanced the reduced expression of IDE induced by oligomeric Aβ25‐35 (P < 0.01; Figure 2C,E). There was no difference in the levels of NEP among the four groups of mice (data not shown). These results indicated that MGCD0103 administration may mitigate Aβ deposition by increasing the levels of IDE.

Figure 2

Effects of MGCD0103 on Aβ Levels in Oligomeric Aβ25‐35 Mice. (A) Representative immunohistochemical staining of amyloid deposits in the mouse hippocampus using 6E10 antibody. Scale bar = 100 μm. The arrowheads indicate positive staining (n = 3‐4 per group). (B) Quantitative results of immunohistochemical staining. Representative Western blots (C) and quantitative densitometry results for BACE1 (D) and IDE (E) expression levels. β‐actin was used as an internal control. The quantitative data are presented as the mean ± SEM for each group (n = 3‐5 per group). *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (**, ##, P < 0.01; ***, P < 0.001)

MGCD0103 attenuates the increased expression of pTau seen in oligomeric Aβ25‐35‐treated mice

Cyclin‐dependent kinase 5 (CDK5) and GSK3β are major kinases that participate in the pathological hyperphosphorylation of tau in both 3xTg‐AD mice and the human brain.27, 28 Therefore, we used IHC staining and Western blot analyses to evaluate the expression levels of pTau, CDK5, and GSK3β in the hippocampus and basolateral amygdala (BLA) of mice. Injections of oligomeric Aβ25‐35 significantly reduced the levels of pS9GSK3β (inactive form of GSK3β; P < 0.05; Figure 3A,B) and increased the levels of pTau at the T205 (P < 0.05; Figure 3A,C) and S202 sites (P < 0.001; Figure 3D,E) in the hippocampus and the S202 site in the BLA (P < 0.001; Figure 3D,F) compared to injections of saline. Moreover, the administration of MGCD0103 significantly reversed these changes in pTau‐related proteins induced by the Aβ25‐35 oligomer in the hippocampal and BLA regions (P < 0.05‐0.001; Figure 3). However, the levels of CDK5, a second critical tau kinase, were not modulated by MGCD0103 treatment (data not shown). Therefore, MGCD0103 treatment increased the levels of pS9GSK3β expression that accompanied the reduction in tau protein phosphorylation at the T205 and S202 sites in oligomeric Aβ25‐35‐treated mice.

Figure 3

Effects of MGCD0103 on Tau Phosphorylation‐Related Proteins in Oligomeric Aβ25‐35 Mice. (A) Representative Western blots and quantitative densitometry results for the ratios of pS9GSK3β/GSK3β (B) and pT205Tau/HT7 (total tau) (C). β‐actin was used as an internal control. (D) Representative immunohistochemical staining of pS202Tau in the CA1 subregion of the hippocampus and the BLA. Scale bar = 100 μm. The arrowheads indicate positive staining of pS202Tau (n = 3‐4 per group). Quantitative results for pS202Tau in the CA1 subregion of the hippocampus (E) and the BLA (F). *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (*, #, P < 0.05; ##, P < 0.01; ***, ###, P < 0.001)

MGCD0103 reduced the gliosis induced by oligomeric Aβ25‐35

Gliosis is a common pathological process in AD29; thus, the modifications in gliosis were evaluated. First, the expression of GFAP (a marker for astrocytes) and Iba1 (a marker for microglia) proteins was measured by IIHC staining. Staining results in the hippocampal region revealed that Aβ25‐35‐inoculated mice had significantly increased levels of GFAP compared to saline‐inoculated mice (P < 0.01; Figure 4A,B). These effects were ameliorated by systemic treatment with MGCD0103 (P < 0.05; Figure 4A,B). In addition, the number of activated microglia in the hippocampal area showed increased levels of activated microglia in the oligomeric Aβ25‐35‐treated group compared to the levels in the saline‐treated group (P < 0.001; Figure 4A,C); this effect was counteracted by systemic treatment with MGCD0103 (P < 0.001; Figure 4A,C). Therefore, we suggest that the administration of MGCD0103 reduced the gliosis that was elevated by the toxicity of oligomeric Aβ25‐35.

Figure 4

Effects of MGCD0103 on Gliosis in Oligomeric Aβ25‐35 Mice. (A) Representative immunohistochemical staining of astrocytes and microglia using anti‐GFAP and anti‐Iba1 antibodies in mouse hippocampal tissue. Scale bar = 100 μm. The arrowheads indicate positive staining of activated microglia and astrocytes, which are magnified in the insets of each figure (n = 3‐4 per group). Quantitative results for astrocytes (B) and activated microglia (C). *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (#, P < 0.05; **, P < 0.01; ***, ###, P < 0.001)

MGCD0103 ameliorated the loss of serotonergic neurons in oligomeric Aβ25‐35‐treated mice

The progression of cognitive deficits is closely reflected by neuronal loss.30 The staining results showed that the numbers of noradrenergic in locus coeruleus (LC) and serotonergic neurons in raphe nucleus but not cholinergic neurons in medial septal nucleus/diagonal band (MS/DB) were significantly decreased in Aβ25‐35‐treated mice compared to the numbers in saline‐treated mice (P < 0.001; Figure 5A‐D). The number of serotonergic neurons was protected by systemic MGCD0103 treatment under oligomeric Aβ25‐35 conditions (P < 0.05; Figure 5A,D).

Figure 5

Effects of MGCD0103 on Neuronal Loss in Oligomeric Aβ25‐35 Mice. (A) Representative immunohistochemical staining of cholinergic neurons in the MS/DB, noradrenergic neurons in the LC, and serotonergic neurons in the raphe nucleus. Scale bar = 200 μm for cholinergic and serotonergic neurons and 100 μm for noradrenergic neurons. The arrowheads indicate positive staining (n = 3‐4 per group). Quantitative results of cholinergic neurons in the MS/DB (B), noradrenergic neurons in the LC (C), and serotonergic neurons in the raphe nucleus (D). *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (#, P < 0.05; ***, P < 0.001)

MGCD0103 increased the levels of acetylated histone 3 and α‐tubulin and the expression of synaptophysin

In this study, we found that the levels of acetylated H3 at the Lys 9 site (AcH3K9) and α‐tubulin were not different in the groups treated with oligomeric Aβ25‐35 or saline (Figure 6A‐C). However, MGCD0103 significantly increased the acetylation of H3K9 in saline‐ and oligomeric Aβ25‐35‐treated mice (P < 0.05, Figure 6A,B) and α‐tubulin (nonhistone) in oligomeric Aβ25‐35‐treated mice (P < 0.05, Figure 6A,C). Evidence also shows that synaptic disruptions, such as decreases in the levels of the presynaptic protein synaptophysin and the postsynaptic protein PSD‐95, have been observed in the AD brain.31, 32 We thus investigated whether synaptic proteins were altered. MGCD0103 significantly increased the expression levels of synaptophysin in oligomeric Aβ25‐35‐treated mice (P < 0.05, Figure 6D,E). However, no changes in PSD‐95 were detected following MGCD0103 treatment (Figure 6D,F).

Figure 6

Effects of MGCD0103 on the Levels of Acetylated Histone 3, α‐Tubulin, and Synaptophysin Expression in Oligomeric Aβ25‐35 Mice. Representative Western blots (A) and quantitative densitometry results for the ratios of AcK9H3/H3 (B) and Acα‐tubulin/α‐tubulin (C). Representative Western blots (D) and quantitative densitometry results for the expression levels of synaptophysin (E) and PSD95 (F). β‐actin was used as an internal control. The quantitative data are presented as the mean ± SEM for each group (n = 3‐5 per group). *, compared to the saline (i.c.)/saline (i.p.) group; #, compared to the oligomeric Aβ25‐35 (i.c.)/saline (i.p.) group (*, #, P < 0.05)

DISCUSSION

In the present study, we demonstrated that the systemic chronic administration of MGCD0103 attenuated anxiety and cognitive deficits and AD‐associated neuropathological markers in oligomeric Aβ25‐35‐treated mice. In addition, chronic treatment of MGCD0103 did not show toxicity in liver and kidney of mice. These results indicate that MGCD0103 may be a potential therapeutic agent against AD.

In this study, bilateral intrahippocampal injection of oligomeric Aβ25‐35 increased anxiety, cognitive dysfunction, Aβ deposition, IDE expression level, tau protein phosphorylation, neuroinflammation, and loss of noradrenergic and serotonergic neurons and decreased the synaptophysin expression level in mice. AD is a complex multifactorial disease. Mounting evidence has demonstrated that the accumulation of soluble Aβ oligomers played an important role in AD pathogenesis.33, 34 The accumulation of Aβ in the brain initiates a cascade of events, including neuroinflammatory response, oxidative injury, and kinases/phosphatase activities, which lead patients to present with the symptoms of dementia.35 Intrahippocampal injection of oligomeric Aβ via stereotaxic is widely used to generate an animal model for AD study.36, 37, 38, 39, 40 Evidence also demonstrated that the platform of intrahippocampal Aβ25‐35 injection can be used to evaluate the cognitive effects of drug in animals.38 In addition, it is reported that Aβ1‐40, the major component of Aβ deposition, which is soluble and nontoxic and is easily converted by brain proteases to truncated toxic fragments, Aβ25‐35/40. This may account for the lag between Aβ deposition and neurodegeneration in AD.41 Therefore, mice received an acute bilateral intrahippocampal Aβ25‐35 injection were established as an AD animal model in the study.

The administration of MGCD0103 attenuated the anxiety and cognitive deficits that accompanied a decrease in serotonergic neuron loss in the raphe nucleus of oligomeric Aβ25‐35‐treated mice. The most common manageable side effects of MGCD0103 therapy at doses 70‐110 mg (orally 3 times a week in 28‐day cycles) in tumor therapy are fatigue, nausea, and vomiting.22, 23 The dose (0.5 mg/kg once daily) we applied to the mice in this study was much lower than the doses in the treatment of tumors. The dose could be even lowered when it is applied to patients with AD according to the body surface area (BSA) normalization method.42 In addition, there is no difference in motor activity in the groups of mice with or without the treatment of MGCD0103 during this study. Furthermore, we did not observe significant lesion in pathological examination of livers and kidneys of the sacrificed mice at the end of treatment. These results reveal the low toxicity of MGCD0103 for the AD model. Psychological symptoms of dementia are an integral part of AD.43, 44 Early evidence shows marked serotonergic neuron loss, predominantly in the raphe nucleus, in patients with AD.45 Evidence also suggests that a dysregulation of the serotonergic system induces anxiety‐like behaviors.46 Previous immunohistochemical evidence also suggests that serotonergic fibers have dense terminals in the basolateral complex of the amygdala.47 In addition, HDACi such as valproic acid (VPA) and trichostatin A (TSA) modulate the activity of serotonergic neurons.48, 49 Growing evidences also suggest that HDACi ameliorated deficits in synaptic plasticity, cognition, and stress‐related behaviors in a wide range of neurologic and psychiatric disorders including AD, anxiety, and mood disorders.16, 50, 51 Furthermore, the serotonergic neurons in the raphe nucleus also innervate the medial prefrontal cortex (mPFC)52; thus, it is reasonable to suggest that the loss of serotonergic neurons in the raphe nucleus can affect cognitive behavior. Evidence also suggests that anxiety traits exacerbate the cognitive impairments and hippocampal vulnerability in patients with AD.53 Therefore, we suggest that the administration of MGCD0103 attenuated the anxiety and cognitive deficits induced by oligomeric Aβ25‐35 via the raphe nucleus and limbic‐related regions. The anxiolytic and attenuated cognitive deficit effects of MGCD0103 could provide an important benefit for both patients with AD and caregivers.

The systemic administration of MGCD0103 showed protective effects on the molecular alterations, including histone and nonhistone (α‐tubulin) acetylation, Aβ deposition, tau phosphorylation, gliosis, and synapse‐related protein changes, induced by oligomeric Aβ25‐35. Interestingly, we found that there was no significant difference in the acetylation of H3K9 and α‐tubulin in the hippocampus of oligomeric Aβ25‐35‐ and saline‐treated mice. However, chronic MGCD0103 treatment induced an increase in the acetylation of H3K9 and α‐tubulin in the hippocampus of oligomeric Aβ25‐35‐treated mice. These results are consistent with a previous study on APP/PS1 mice54 and suggest that the memory deficits seen in acutely oligomeric Aβ25‐35‐treated mice are not maintained by abnormalities in histone or nonhistone acetylation. However, the chronic administration of MGCD0103 caused elevations in histone and nonhistone acetylation in oligomeric Aβ25‐35‐treated mice. Increased levels of H3K9 acetylation has also been observed in sodium butyrate–treated animals.55 MGCD0103 also protected microtubule stability–related protein activity, including α‐tubulin acetylation, S9GSK3β phosphorylation, and tau protein phosphorylation at the T205 and S202 sites, against the effects of oligomeric Aβ25‐35 treatment. Evidence has shown that acetylated tubulin plays a critical role in microtubule stability at different ages or experimental conditions.56 Previous studies also suggest that tau protein phosphorylation at both sites (S202 and T205) not only enhances polymerization but also increases filament formation sensitivity.57 Therefore, treatment with MGCD0103 may protect the microtubule structure from the damage induced by oligomeric Aβ25‐35 via increasing the levels of α‐tubulin acetylation and pS9GSK3β expression and decreasing the levels of phosphorylated tau at the T205 and S202 sites. Next, we further found that MGCD0103 treatment increased only the expression levels of synaptophysin and did not increase those of PSD95. These results are consistent with a previous study that assessed treatment with another HDACi, 4‐phenylbutyrate.54 In this study, oligomeric Aβ25‐35 increased Aβ deposition and BACE1 expression and decreased IDE expression. MGCD0103 treatment increased only the levels of IDE, not NEP, under oligomeric Aβ25‐35 conditions. Furthermore, evidence shows that the reduction in gliosis was not consistent in HDACi treatment.58 However, treatment with MGCD0103 attenuated the gliosis induced by oligomeric Aβ25‐35.

Taken together, these results reveal MGCD0103 could be a potential therapeutic agent against AD.

CONFLICT OF INTEREST

The authors have no conflict of interests to declare.

Click here for additional data file.