HDAC8 Inhibitor WK2-16 Therapeutically Targets Lipopolysaccharide-Induced Mouse Model of Neuroinflammation and Microglial Activation

Glial activation and neuroinflammatory processes play important roles in the pathogenesis of brain abscess and neurodegenerative diseases. Activated glial cells can secrete various proinflammatory cytokines and neurotoxic mediators, which contribute to the exacerbation of neuronal cell death. The inhibition of glial activation has been shown to alleviate neurodegenerative conditions. The present study was to investigate the specific HDAC8 inhibitor WK2-16, especially its effects on the neuroinflammatory responses through glial inactivation. WK2-16 significantly reduced the gelatinolytic activity of MMP-9, and expression of COX-2/iNOS proteins in striatal lipopolysaccharide (LPS)-induced neuroinflammation in C57BL/6 mice. The treatment of WK2-16 markedly improved neurobehavioral deficits. Immunofluorescent staining revealed that WK2-16 reduced LPS-stimulated astrogliosis and microglial activation in situ. Consistently, cellular studies revealed that WK2-16 significantly suppressed LPS-induced mouse microglia BV-2 cell proliferation. WK2-16 was proven to concentration-dependently induce the levels of acetylated SMC3 in microglial BV-2 cells. It also reduced the expression of COX-2/iNOS proteins and TNF-α production in LPS-activated microglial BV-2 cells. The signaling studies demonstrated that WK2-16 markedly inhibited LPS-activated STAT-1/-3 and Akt activation, but not NF-κB or MAPK signaling. In summary, the HDAC8 inhibitor WK2-16 exhibited neuroprotective effects through its anti-neuroinflammation and glial inactivation properties, especially in microglia in vitro and in vivo.


Introduction
Central nervous system (CNS) injuries and neurological diseases, such as head trauma and neurodegenerative diseases, create enormous individual and financial burdens. Much evidence has revealed that neuroinflammation plays a critical role in the development and progression of these neuropathological conditions [1,2]. Microglial activation, astrogliosis, and the release of inflammatory mediators are the major immune components of neuroinflammation. Microglia are the primary Figure 1. WK2-16 suppressed inflammatory mediator production and protected against functional neurological deficits in LPS-induced neuroinflammation in C57BL6 mice. Protein levels of (A) COX-2 and (B) iNOS were assessed eight hours after intrastriatal normal saline or LPS injection (4 µg/3 µL) in the presence or absence of WK2-16 (30 mg/kg) using Western blotting. The results are expressed as fold changes normalized to β-actin. (C) MMP-9 gelatinolytic activity in brain tissue homogenates was evaluated using zymography. (D) Neurological scores according to neurological function changes were recorded eight hours after normal saline or LPS injection in the presence or absence of WK2-16 in C57BL/6 mice. Representative bands shown in Western blots and zymography were classified as the ipsilateral site (LPS-injected brain) and contralateral side (non-operated site). The quantification values are presented as the means ± SD of results from 3-5 animals. I: ipsilateral; C: contralateral. ## p < 0.01, ### p < 0.001 compared to the normal saline-injected group; ** p< 0.01, *** p< 0.001 compared to the LPS-injected group treated with vehicle.

WK2-16 Inhibited Astrogliosis and Microglia Proliferation in LPS-Challenged C57BL6 Mice
The activation of glial cells, such as astrocytes and microglia, plays a major role in the pathogenesis of neuroinflammation via mediation of inflammatory responses and alterations in brain-blood barrier (BBB) permeability [2]. Therefore, we examined whether WK2-16 treatment suppressed astrocyte and microglia activation. Brain sections were immunostained with glial fibrillary acidic protein (GFAP) ( Figure 2B-D) and ionized calcium-binding adaptor molecule 1 (Iba1) ( Figure 3B-D) antibodies, which target astrocytes and microglia, respectively. A significant increase in GFAP-immunoreactive cells in the corpus callosum was found in intrastriatal LPS-injected group compared to the normal saline-injected group ( Figure 2E,F). The morphological changes were characterized by enlarged cell bodies and extended processes in GFAP-expressing astrocytes. The increased number and stainability of GFAP immunoreactivity indicated a dramatic astrogliosis ( Figure 2F). Subsequent WK2-16 (30 mg/kg) treatment significantly reduced GFAP immunoreactivity compared to the LPS-injected group ( Figure 2F,G). The number of cells positive for GFAP was calculated and quantified ( Figure 2H). Intrastriatal LPS-injection markedly increased the number of GFAP-positive cell to 409.3 ± 125.3 as compared with the normal saline-injected group (49.7 ± 6.4, p = 0.003). Treatment with WK2-16 significantly decreased the number of GFAP-expressing cells (222.7 ± 37.9, p = 0.024).

Figure 2.
The inhibition of WK2-16 on LPS-induced astrocyte activation in C57BL6 mice brain. Brain sections (50 μm thick) were prepared eight hours after intrastriatal normal saline or LPS (4 μg/ 3μl) injection in the presence or absence of WK2-16 (30 mg/kg). (A) Schematic illustration of the stereotaxic injection site in the caudate/putamen region (red arrow). (B-D) Representative immunofluorescence staining of GFAP (green) in normal saline-and LPS-injected groups in the presence or absence of WK2-16 (30 mg/kg). (E-G) Higher magnification fluorescence microscopy of the rectangular area in Figure 2A. Figure 2E, F, and G were magnified from Figure 2B, C, and D in the corpus callosum, respectively. GFAP immunostaining was counterstained with DAPI (blue). Scale bar = 75 μm. (H) Quantification of cells positive for GFAP. The data are presented as the means ± SD from three animals. ## p < 0.01 compared to the normal saline-injected group; * p < 0.05 compared to the LPS-injected group treated with vehicle.
We further examined microglia activation using Iba1. The number of Iba-1 signals in the striatum increased eight hours after intrastriatal LPS administration compared to the normal  Figure 2A. Figure 2E,F,G were magnified from Figure 2B, C, and D in the corpus callosum, respectively. GFAP immunostaining was counterstained with DAPI (blue). Scale bar = 75 µm. (H) Quantification of cells positive for GFAP. The data are presented as the means ± SD from three animals. ## p < 0.01 compared to the normal saline-injected group; * p < 0.05 compared to the LPS-injected group treated with vehicle. compared to the normal saline-injected group (183.5 ± 82.9, p < 0.001), and WK2-16 treatment reduced this effect (421.0 ± 92.5, p = 0.046). The anti-proliferative effect of WK2-16 in microglial BV-2 cells was assessed using the sulforhodamine B (SRB) assay in vitro to examine the inhibitory effect on microglia activation. WK2-16 concentration-dependently (0.5, 1, and 2 μM) inhibited the proliferation of BV-2 cells compared to LPS-stimulated BV-2 cells (p = 0.044, p = 0.005, and p < 0.001, respectively) ( Figure 3I). Taken together, WK2-16 treatment markedly suppressed glial activation and proliferation following LPS-induced neuroinflammation.  Figure 3A. Figure 3E,F,G are magnified from Figure 3B, C, and D in the striatum, respectively. Iba-1 immunostaining was counterstained with DAPI (blue). Scale bar = 75 µm. (H) Quantification of cells positive for Iba-1. The data are presented as the means ± SD from 4 animals. (I) BV-2 microglia were pretreated with WK2-16 at varying concentrations followed by LPS (150 ng/mL) incubation for 48 h. The relative cell proliferation was determined using the SRB assay in vitro (n = 3). R: resting; V: vehicle (DMSO). # p< 0.01, ### p< 0.001 compared to the normal saline-injected group in vivo or resting group in vitro; * p < 0.05, ** p < 0.01, *** p < 0.001 compared to the LPS-treated group treated with vehicle in vivo and in vitro.

Discussion
The present study investigated the application of the HDAC8 inhibitor WK2-16 in neuroinflammation-mediated diseases. The results demonstrated that WK2-16 protected against neurological deficits and reduced neuroinflammatory responses in an intrastriatal LPS-induced neuroinflammation model. The neuroprotection was mediated through the suppression of astrogliosis and microglial activation. Our microglial study revealed that WK2-16 primarily affected STAT-1/-3 and Akt activity to suppress endotoxin-mediated inflammation.
Microglia and astrocytes are innate immune cells that develop and maintain the inflammatory response in the CNS, which is characterized by enhanced proliferation and activation [26]. Extended expression of GFAP is a hallmark for CNS pathologies, and it is primarily distributed in astroglia in vivo and in vitro [27]. The activation of astroglia produces neuroprotective effect on the retina by enhancing the expression of cytoprotective factors and restoring neurotransmitter balance. However, proliferative astrogliosis can increase vascular permeability and infiltration of toxic compounds, resulting in the acceleration of neurodegeneration [28]. It was also found that proliferative astroglia could release various inflammatory factors during neurodegeneration, leading to myelin damage and axonal loss, especially in corpus callosum. Application of anti-inflammatory and anti-gliosis agent has been shown to effectively reduce demyelination in the corpus callosum [29,30]. WK2-16 significantly reduced GFAP immunoreactivity in the corpus callosum after intrastriatal LPS stimulation in the present study, which demonstrates its anti-astrogliosis activity and the capacity of CNS repairing. In addition, much evidence demonstrated astrocytes as the major source for MMP-9 production in neuropathies. The overproduction of MMP-9 results in disruption of the BBB and death-inducing ligand release,

Discussion
The present study investigated the application of the HDAC8 inhibitor WK2-16 in neuroinflammation-mediated diseases. The results demonstrated that WK2-16 protected against neurological deficits and reduced neuroinflammatory responses in an intrastriatal LPS-induced neuroinflammation model. The neuroprotection was mediated through the suppression of astrogliosis and microglial activation. Our microglial study revealed that WK2-16 primarily affected STAT-1/-3 and Akt activity to suppress endotoxin-mediated inflammation.
Microglia and astrocytes are innate immune cells that develop and maintain the inflammatory response in the CNS, which is characterized by enhanced proliferation and activation [26]. Extended expression of GFAP is a hallmark for CNS pathologies, and it is primarily distributed in astroglia in vivo and in vitro [27]. The activation of astroglia produces neuroprotective effect on the retina by enhancing the expression of cytoprotective factors and restoring neurotransmitter balance. However, proliferative astrogliosis can increase vascular permeability and infiltration of toxic compounds, resulting in the acceleration of neurodegeneration [28]. It was also found that proliferative astroglia could release various inflammatory factors during neurodegeneration, leading to myelin damage and axonal loss, especially in corpus callosum. Application of anti-inflammatory and anti-gliosis agent has been shown to effectively reduce demyelination in the corpus callosum [29,30]. WK2-16 significantly reduced GFAP immunoreactivity in the corpus callosum after intrastriatal LPS stimulation in the present study, which demonstrates its anti-astrogliosis activity and the capacity of CNS repairing. In addition, much evidence demonstrated astrocytes as the major source for MMP-9 production in neuropathies. The overproduction of MMP-9 results in disruption of the BBB and death-inducing ligand release, propagating the neuroinflammatory response via the recruitment of immune cells. The inhibition of MMP-9 has been shown to protect against neuronal death and improve neurological functions [31,32]. Therefore, MMP-9 may be a therapeutic target in inflammation-mediated neurological diseases. Our study revealed that WK2-16 dramatically suppressed MMP-9 activation, which may positively correlate with the protective effect of neurological functions. On the other hand, over-expression of the marker Iba1 was primarily found in proliferating microglia/macrophages during the active phase of brain disease. Therefore, suppression of microglial proliferation is sufficient to downregulate microglia-mediated inflammatory functions [33,34]. A previous study reported that activation of PI3K/Akt and STAT-3 signaling increased the survival and proliferation of microglia [35], which revealed that Akt and STAT-3 activation was implied in microglia proliferation. Iba1 immunofluorescence in vivo and the microglial SRB assay revealed that WK2-16 significantly attenuated microglia proliferation and reversed morphological changes, including hypertrophic somas and extended processes. The inhibition of proliferation was not involved with cell death because the viability of WK2-16 (2 µM)-treated BV-2 cells was 87.2 ± 4.0% compared to the control cells in the MTT assay (data not shown). Our in vitro study further revealed that the possible inhibitory mechanisms of WK2-16 on microglial proliferation in response to LPS may be probable via Akt and STAT-1/-3 inhibition.
The TLR4 activator LPS activates macrophages/microglia and monocytes to release free radicals generated by iNOS in combination with inflammatory mediators and cytokines, including COX-2 and TNF-α, resulting in neuronal damage [36,37]. A previous study indicated that inhibition of mTOR could reduce microglial proliferation and regulate microglial activation via reduction of iNOS and COX-2 expression in response to inflammatory cytokines [38]. NO-mediated toxicity was involved in dopaminergic neuron loss in a model of LPS-induced Parkinson's disease, and inhibition of iNOS significantly prevented neuronal cell death [39]. In additional, IL-1β-and TNF-α-mediated neuronal death was reported in acute CNS injury and chronic neurodegenerative disease [40]. In the present study, WK2-16 substantially reduced the over-expression of iNOS and COX-2 in intrastriatal LPS-inflamed brains. In vitro studies also revealed that LPS-induced microglial iNOS and COX-2 expression and TNF-α production were suppressed by WK2-16. We examined whether WK2-16 affected NF-κB, STATs, and MAPK signaling pathways, which regulate iNOS and COX-2 expression at the transcriptional level, to further clarify the underlying mechanisms. A previous study demonstrated that sufficient activation of STAT-1 was necessary for the full transcription of iNOS in macrophages, whereas NF-κB was activated [41]. STAT-1 and STAT-3 binding sites are widely located in the microglial genome, and manipulation of the constitutive expression of STAT-1 and STAT-3 mimics LPS stimulation, inducing the transcription of hundreds of genes and results in inflammatory cytokines production in microglia [42]. STAT-1/-3 and p300 are found to recruit iNOS and COX-2 gene promoters in LPS-stimulated microglia, and the downregulation of STAT signaling inhibits iNOS and COX-2 gene expression in vitro and in vitro [43]. In the present study, WK2-16 suppressed the phosphorylation of STAT-1/-3 in LPS-stimulated BV2 microglia, which may positively correlate with the downregulation of iNOS and COX-2 in vitro and in vivo.
Numerous HDAC inhibitors exhibited promising therapeutic effects in neurological conditions. Pan-HDAC inhibitors, such as TSA, SAHA, valproate, and sodium butyrate, reduced neuroinflammatory responses and exhibited neuroprotective effects in rodent models of stroke, neurodegenerative disease, or acute brain injury [16,44]. TSA and SAHA were found to cross the BBB in clinical trials, which highlights their potent protective effects against Parkinson's disease [45]. Notably, most studies focused on the use of pan-HDAC inhibitors, which are nonselective HDAC inhibitors. Therefore, it is difficult to clarify which HDACs regulate the inflammatory response and how specific proteins are modified to reduce neuroinflammation [46]. HDAC inhibitors with higher isoform selectivity may reduce the unfavorable adverse effects [47]. A specific HDAC6 inhibitor ameliorated Alzheimer's disease phenotypes [48], and an HDAC8 inhibition repaired scopolamine-induced learning and memory impairments in animals [49]. The present study addressed whether selective inhibition of Class I HDAC8 by WK2-16 could protect against neurobehavioral deficits and reduce the inflammatory responses in an LPS-induced neuroinflammation mouse model. Some findings revealed that the inhibition of class I/II HDAC polarized microglia from an inflammatory M1-phenotype toward the protective M2-phenotype with decreased expression of iNOS in a model of traumatic brain injury [50]. The specific inhibition of HDAC1/2 also reduced IL-6 and TNF-α production in LPS-activated microglia [46]. In the present study, WK2-16 downregulated microglial proliferation and morphological changes, positively correlating with the suppression of iNOS expression in LPS-inflamed brains. On the other hand, some HDAC inhibitors affect the interaction between NF-κB and IκBα, the DNA-binding ability, and transcriptional activity of p65, leading to a neuroprotective effect in vitro [51,52]. WK2-16 did not reduce p65 phosphorylation in LPS-stimulated microglia, which is consistent with our previous findings in monocytes [20]. The acetylation of histone 3 at lysine 9 by HDACs mediates microglia inflammation in response to LPS. Treatment with the pan-HDAC inhibitor sodium butyrate suppressed proinflammatory (Tnf-α, Nos2, Stat1, Il6) and promoted anti-inflammatory (Il10) genes via the downregulation of HDAC8 and STAT-1/-3 signaling in vivo and in vitro [14,53]. According to these studies, the WK2-16-reduced microglial iNOS, COX-2, and TNF-α production could be mediated by HDAC8 and subsequent STAT-1/-3 inhibition. Blockade of STAT-1/-3 signaling has been proven to be neuroprotective in brain injury via an attenuation of inflammatory responses [54,55].
Taken together, the novel HDAC8 inhibitor WK2-16 exerted the neuroprotective effects via suppressing microglial reactivation, astrogliosis, and neuroinflammation. The therapeutic mechanism of WK2-16 is mediated by HDAC8 inhibition and subsequent blockade of STAT-1/-3 signaling, especially in microglia. Further investigation of its mechanism on the biological activity of acetylation is warranted.

Animals
Male C57BL/6 mice approximately 25-27 g body weight were purchased from BioLASCO Taiwan Co., Ltd. (Taipei, Taiwan). The animals were maintained in an environment with a 12-h light/dark cycle at 25 ± 1 • C with free access to food and water. The Institutional Animal Care and Use Committee of Taipei Medical University approved all animal use protocols (LAC-2017-0182, 25 July 2017), and animals were operated in accordance with guidelines of ARVO statement for ophthalmic and vision research.

Stereotactic Injections and WK2-16 Administration
C57BL/6 mice were processed for intrastriatal LPS injection as previously described [57]. LPS was dissolved in sterilized normal saline and 3 µL (equivalent to 4 µg LPS) was injected. Control animals were received 3 µL of sterilized normal saline. In brief, the normal saline or LPS solution was given stereotaxically into the caudate/putamen region of the right brain with coordinates of 0.3 mm anterior, 2.0 mm lateral, and 3.8 mm ventral to the bregma. Animals were decapitated eight hours after LPS injection, and brains were quickly harvested.
The dose of WK2-16 (30 mg/kg) was used according to a previous study [20]. This compound was dissolved in a co-solvent (ethanol: cremophor: saline = 1: 1: 8). The treatment group received a single dose of WK2-16 intraperitoneally 30 min after intrastriatal LPS microinjection. The controlled group received intraperitoneally injections with the cosolvent.

Neurological Score
A modified neurological score was used to assess neurological function on a scale of 3-18, in accordance with Garcia et al. [21]. The neurological score was composed of motor and sensory tests, including spontaneous activity, symmetry in the movement of four limbs, forepaw outstretching, climbing, body proprioception, and response to vibrissae touch. Mice were scored for clinical symptoms prior to sacrifice. Mice with lower neurological scores exhibited greater neuroinflammation than mice with higher neurological scores.

Western blot Analysis
Homogenates of brain tissue and microglia cell lysates were prepared as described previously [58]. Samples were subjected to SDS-PAGE and transferred onto nitrocellulose membranes. Each specific primary antibody was immersed with the nitrocellulose membrane at 4 • C overnight. The membrane was incubated with homologous HRP-conjugated secondary antibodies, followed by detection of immunoreactive bandsusing ECL. Densitometry was performed using BioLight software V2000.01 (Uhldingen-Mühlhofen, Germany). The optical density of specific protein was normalized to an internal control and expressed as a ratio to control.

Immunofluorescence
Mice were transcardially perfused with 4% paraformaldehyde (PFA), and brains were resected immediately and fixed in 4% PFA at 4 • C overnight. The brain was dehydrated in 30% sucrose, followed by serial sectioning at a thickness of 50 µm using a cryostat (Leica SM2010R; Nussloch, Germany). The following procedure was executed, as described previously [58]. Brain sections were incubated with anti-GFAP and anti-Iba1 antibodies at 4 • C overnight, followed by incubation with the homologous fluorescence-conjugated secondary antibody DyLight 488 for 2 h at room temperature. Samples were counterstained with DAPI. All fluorescence images were captured using an Eclipse 80i fluorescence microscope (Nikon Instruments, Melville, NY, USA). The fluorescence content of GFAP and Iba-1 was quantified using ImageJ software (Bethesda, Rockville, MD, USA).

MTT Assay
Colorimetric MTT assays were used to determine cell viability, as described previously [24]. A total of 3 × 10 5 BV-2 cells were seeded in each well of a 12-well plate and treated with various concentrations of WK2-16 for 22.5 h. MTT was added followed by the further incubation for 1.5 h. Cells were lysed in 1 mL DMSO, and the absorbance values of medium were measured using a microplate reader at 550 nm (Thermo Multiskan GO, Ratastie, Finland).

Sulforhodamine B (SRB) Assay
The cell proliferation of microglia BV-2 cells was determined using the SRB assay, as described previously [60]. BV-2 cells (5 × 10 3 ) were dispensed in a 96-well plate and pretreated with varying concentrations of WK2-16 (0.5, 1, and 2 µM) for 30 min. Cells were subsequently incubated with LPS (150 ng/mL) for an additional 48 h. Cells were fixed in 10% trichloroacetic acid for 10 min and washed with distilled water twice to remove excess trichloroacetic acid. Cells were stained with a 0.4% SRB solution dissolved in 1% acetic acid for 10 min and washed with 1% acetic acid twice to remove excess dye. The protein-bound dye was dissolved in 100 µL Tris-based solution (10 mM), and the absorbance was measured at 515 nm using a microplate reader. Relative cell proliferation was calculated as the absorbance of the treated group/the absorbance of the resting group.

Enzyme-Linked Immunosorbent Assay (ELISA)
TNF-α levels in the conditioned media of BV-2 microglial cells were detected using an ELISA kit according to the manufacturer's instructions (BioLegend, San Diego, CA, USA).

Statistical Analyses
All data are expressed as the means ± SD. Statistical analyses were performed using one-way analysis of variance in Sigma Stat v3.5 software. Statistical differences between experimental groups were determined using the Student-Newman-Keuls test. A p-value < 0.05 was considered statistically significant.