Exp Neurobiol 2018; 27(5): 437-452
Published online October 31, 2018
© The Korean Society for Brain and Neural Sciences
Selien Sanchez1,†, Stefanie Lemmens1,†, Paulien Baeten1, Daniela Sommer1, Dearbhaile Dooley2, Sven Hendrix1,‡*, and Myriam Gou Fabregas1,‡
1Department of Morphology, Biomedical Research Institute, Hasselt University, Diepenbeek BE3590, Belgium.
2Health Science Centre, School of Medicine, University College Dublin, Dublin D04 V1W8, Ireland.
Correspondence to: *To whom correspondence should be addressed.
TEL: 32 11 26 92 46
†Equally contributing first authors.
‡Equally contributing senior authors.
After spinal cord injury (SCI), monocyte derived macrophages play a detrimental role. Histone deacetylases (HDACs) are central epigenetic regulators of macrophage-polarization. We hypothesized that HDAC3 inhibition suppresses the pro-inflammatory macrophage phenotype (M1), promotes the anti-inflammatory phenotype (M2) and improves functional recovery after SCI. Therefore, two inhibitors of HDAC3 were selected, namely scriptaid and RGFP966. The impact on macrophage polarization was studied by investigating the effect on gene and protein expression of selected M1 and M2 markers. We show that scriptaid differentially influences M1 and M2 markers. It increases CD86 and iNOS gene expression and decreases GPR18, CD38, FPR2 and Arg-1 gene expression as well as the production of IL-6 and NO. RGFP966 primarily increased the expression of the M2 markers Arg-1 and Ym1 and reduced the production of IL-6 (M1). RGFP966 and scriptaid reduced the formation of foamy macrophages. Finally, to investigate the impact of HDAC3 inhibition on functional recovery after SCI, we studied the effects of RGFP966 and scriptaid in an
Spinal cord injury (SCI) is a complex disorder, during which, excessive inflammation and glial scar formation aggravate the tissue damage and have a negative impact on neuronal regeneration [1,2,3,4,5,6,7]. Major players in this neuro-inflammatory response are macrophages. These cells react rapidly after injury by producing several cytokines, neurotrophic factors and by playing a role in scar tissue remodelling . Macrophages can be polarized to different subpopulations including
Histone deacetylases (HDACs) have recently been identified as important therapeutic targets in many human central nervous system (CNS) disorders . HDACs remove acetyl groups from histones or other protein substrates. Protein acetylation modulates the function of many proteins and can therefore affect numerous cellular processes, such as gene transcription, microtubule dynamics, metabolism and aging. Valproic acid (VPA), a broad-acting HDAC inhibitor, has been shown to affect the macrophage phenotype and to improve functional recovery after SCI [17,18]. Scriptaid, which is a more specific HDAC inhibitor that targets HDAC1, HDAC3 and HDAC8, prevents white matter injury and improves motor functions after traumatic brain injury (TBI). In addition, Wang et al. showed that scriptaid shifted microglia/macrophage polarization towards the protective M2 phenotype [19,20]. However, the underlying mechanisms and the specific HDAC responsible for these effects, remain unclear. Since, HDAC3 is a key regulator of macrophage polarization, it represents a promising therapeutic target .
In this study, we examined the effects of scriptaid and RGFP966, a specific HDAC3 inhibitor, on macrophage polarization and SCI recovery. We hypothesized that HDAC3 inhibition induces the protective macrophage M2 phenotype leading to improved functional recovery after SCI. The influence of both inhibitors on macrophage polarization has been investigated by determining the expression of selected M1 and M2 markers on the gene and protein level. Secondly, the effects of RGFP966 and scriptaid on functional recovery after SCI have been evaluated by assessing the Basso Mouse Scale (BMS) in the T-cut hemisection SCI mouse model. Our results indicate that the specific HDAC3 inhibitor RGFP966 promotes alternative activation of macrophages, but does not lead to better functional recovery after SCI.
Bone marrow derived macrophages (BMDMs) were obtained by isolating primary bone marrow cells from femurs and tibias of female Balb/c mice (Envigo, Cambridgeshire, U.K.) as previously described . These femoral and tibial cell suspensions were flushed out of the bones with sterile ice-cold 1x phosphate-buffered saline (PBS). Cells were cultured and differentiated in RPMI 1640 medium (Lonza, Basel, Switzerland) for 10 days, supplemented with 10% heat-inactivated fetal calf serum (hiFCS; Gibco, Waltham, U.S.A.), 1% penicillin/streptomycin (Sigma-Aldrich, Saint Louis, U.S.A.) and 15% L929 conditioned medium (LCM), at 37℃ and 5% CO2. After 10 days, the differentiated BMDMs were plated out for in vitro experiments.
For qPCR (1×106 cells/condition), western blot (0.5×106 cells/condition), the arginase activity assay (1×106 cells/condition), the ELISA (0.5×106 cells/condition) and the Griess assay (0.5×106 cells/condition), the BMDMs were first pre-stimulated for 1 hour either with LPS (200 ng/ml; EMD Millipore, Billerica, U.S.A.) to obtain a polarization towards M1 macrophages or with IL-4 (33 ng/ml; Peprotech, Rocky Hill, U.S.A.) or IL-13 (33 ng/ml; Peprotech) to obtain a polarization towards M2 macrophages (Figs. 1, 2, 3, 4, 5, 6). Thereafter, the BMDMs were stimulated for 24 hours with RGFP966 (5 µM or 10 µM, Cayman Chemicals, Ann Arbor, U.S.A) or scriptaid (0.2 µM or 1 µM; Sigma-Aldrich). These doses displayed no toxicity. RGFP966 was chosen because it has an IC50 of 0.08 µM for HDAC3 and showed no effective binding for any other HDAC at concentrations up to 15 µM . A condition with dimethylsulfoxide (DMSO; Sigma-Aldrich) has been used as a vehicle control. This experimental set-up is shown in Fig. 3A. Pictures were taken of the BMDMs in all these conditions after 24 hours using a Nikon eclipse TS100 microscope, the results are shown in Fig. 2.
For the Oil red O (ORO; 0.75×106 cells/condition) assay, BMDMs were stimulated for 1 hour with either LPS or IL-4 to obtain polarization towards an M1 or M2 macrophage phenotype. Next, these macrophages were treated with RGFP966 (5 µM, 10 µM) or scriptaid (0.2 µM, 1 µM) for 24 hours. After stimulation with the inhibitors, the cells were treated for 48 hours with spinal cord debris (50 µg), fixated and stained with ORO.
For the quantitative polymerase chain reaction (qPCR) cell lysis was performed with β-mercaptoethanol in RLT-buffer (1:100; RNeasy Mini Kit, Qiagen, Venlo, The Netherlands). mRNA was isolated using RNeasy Mini Kit (Qiagen) following manufacturer's instructions and the mRNA concentration was measured by Nanodrop (Thermo Fisher Scientific, Merelbeke, Belgium). Furthermore, the reverse transcription to cDNA and the qPCR reaction was performed as previously described, the primer sequences that were used are displayed in Table 1 . Relative quantification of gene expression was accomplished with the comparative Ct method. Data were normalized to YWHAZ and HMBS, which were selected as the most stable reference genes using geNorm.
Stimulated BMDMs were lysed in SDS lysis buffer (2% (w/v) in 125 mM Tris) to collect total protein lysates. Using Pierce BCA protein assay kit (Thermo Fisher Scientific) according to manufacturer's instructions and iMARK Microplate Reader (Bio-Rad Laboratories, Hercules, U.S.A.), protein concentrations were measured. Protein samples (10 µg) were separated on 12% or on 7.5% (for iNOS) SDS gels for 45 minutes at 200 V. The western blot was performed as previously described . The primary antibodies used were: rabbit anti-mouse acetylated histone 3 (1/2000; Cell signalling, Leiden, The Netherlands), goat anti-mouse arginase-1 (Arg-1; 1/1000; Santa Cruz Technologies, Dallas, U.S.A.), mouse anti-mouse inducible nitric oxide synthase (iNOS; 1/500; Sigma-Aldrich), mouse anti-mouse β-actin (1/5000, Santa Cruz). The measured values were normalized to the level of β-actin or total histone 3 and 4, as internal controls. Selected blots (Fig. 1B, 5D and 6B), which derive each from the same membrane have been cut and reordered to increase readability.
Cell pellets from BMDMs stimulated with IL-4 or IL-13 were collected to measure the arginase activity, which is a characteristic of M2 polarisation. The assay was performed using the ‘Arginase activity assay kit’ (Sigma-Aldrich) according to the manufacturer's instructions.
IL-6 production, which is a characteristic of M1 polarisation, was measured in medium of LPS stimulated cells. The assay was performed using the ‘eBioscience Mouse IL-6 ELISA Ready-SET-Go Kit’ (Thermo Fisher Scientific) according to the manufacturer's protocol.
To measure nitrite (NO2− concentration), which is a measure of NO production and a characteristic of M1 polarisation, a Griess assay was performed on LPS stimulated cell media. This assay was performed using the ‘Griess reagent system’ kit (Promega, Leiden, The Netherlands) according to the manufacturer's instructions.
After culturing and stimulation (Treatment of BMDMs) as previously described, BMDMs were fixed using 4% paraformaldehyde (Sigma-Aldrich) and stained with ORO (Sigma-Aldrich). The staining was extracted using isopropanol and its absorbance was measured at 490 nm using an iMARK Microplate Reader (Bio-Rad Laboratories).
Experiments were performed using 10-week old female Balb/c mice (Envigo). The animals were housed in groups under regular conditions (temperature- and humidity-controlled, 12-hour light/dark cycle and food and water
T-cut spinal cord hemisection injury was performed as previously described [4,5,6,7,24]. Mice were anesthetized with 3% isoflurane (IsofFlo, Abbot Animal Health, Waver, Belgium) and received a subcutaneous injection of the analgesic buprenorphine Temgesic (0.1 mg/kg bodyweight; Val d'Hony Verdifarm, Beringen, Belgium). A partial laminectomy was performed at thoracic level T8 to expose the spinal cord. Iridectomy scissors were used to transect left and right dorsal funiculus, the dorsal horns and the ventral funiculus. This “T-cut” procedure results in a specific and clean transection of the dorsomedial and ventral corticospinal tract. The muscles were sutured and the back skin was closed with wound clips (Autoclip, Clay-Adams Co., Inc. Franklin Lakes, US). Glucose solution (20%) was administered intraperitoneally (i.p.) post-op to compensate for any blood loss during surgery. Bladders were manually emptied daily until autonomic control was restored. Mice were treated i.p. for 3 consecutive days, starting 2 hours after SCI with either RGFP966 (10 mg/kg, dissolved in 7.7% DMSO in NaCl), scriptaid (3.5 mg/kg, dissolved in 0.9% DMSO in NaCl) or vehicle containing 0.9% DMSO in NaCl. The treatment protocol was adapted from .
Functional recovery in mice was measured using the BMS starting one day after injury until day 27 by a trained investigator blinded to the experimental groups . In the first week, mice were scored daily and thereafter every other day until the end of the observation period, 27 days post injury (dpi).
At 28 dpi, animals were overdosed with dolethal (Vetiquinol NV, Aartselaar, Belgium) and transcardially perfused with ringer-heparin and 4% paraformaldehyde in 0.1 M PBS. Immunofluorescence analysis was performed as previously described [4,6]. To determine lesion size demyelination, gliosis and inflammatory infiltrate, cryosections were blocked using 10% protein block (Dako Agilent Technologies, Diegem, Belgium) in PBS for 30 minutes at room temperature. Next, the following primary antibodies were incubated overnight at 4℃: mouse anti-GFAP (1/500, Sigma-Aldrich), rat anti-MBP (1/250, EMD Millipore), rabbit anti-Iba-1 (1/350, Wako Chemicals GmbH, Neuss, Germany), rat anti-CD4 (1/25, BD Biosciences, San Jose, U.S.A.). To identify classically or alternatively activated macrophages, sections were permeabilized using 0.1% Triton X-100 for 30 minutes and blocked with 20% donkey serum or 10% goat serum in Tris-buffered saline (TBS, pH 7.5) for 1 hour. Incubation with rat anti-MHC-II (1/200, Santa Cruz Technologies) and goat anti-Arg1 (1/50, Santa Cruz Technologies) primary antibodies was performed overnight at 4℃. Next, the cryosections were incubated for 1 hour at room temperature with the corresponding secondary antibodies: goat anti-rat IgG Alexa Fluor 568 and 488 (1/250, Invitrogen), goat anti-mouse Alexa Fluor 568 (1/250, Invitrogen), goat anti-rabbit Alexa Fluor 488 (1/250, Invitrogen), donkey anti-goat IgG Alexa Fluor 555 and 488 (1/400, Invitrogen, Carlsbad, U.S.A.), rabbit anti-rat biotin (1/400, Dako) and streptavidin 488 (1/2000, Invitrogen). DAPI (1/1000, Sigma-Aldrich) counterstaining was performed for 10 minutes. Omission of the primary antibody was used to confirm the specificity of the secondary antibodies. All slides were mounted with fluorescent mounting medium (Dako). Fluorescent images were taken using a Leica DM4000 BLED fluorescence microscope with a Leica DFC450C (Leica microsystems, Diegem, Belgium). Quantitative image analysis was performed using unmodified pictures in ImageJ open source software. GFAP staining was used to evaluate the lesion size by delineating the GFAP- area, in the same way the MBP- area was delineated to evaluate the demyelinated area. GFAP and Iba-1 expression was quantified by intensity analysis within rectangular areas of 100 µm×100 µm extending 600 µm cranial to 600 µm caudal from the lesion epicentre. CD4+ cells were counted in total spinal cord sections, Arg-1+ cells and MHCII+ cells were counted at the lesion area. The analyses were done on 4~7 spinal cord sections per mouse representing the lesion area, as previously described .
The analyses were performed using GraphPad Prism 5.0 (Graph-Pad Software, La Jolla, U.S.A.). D'Agistino and Pearson omnibus normality tests were used to test normal distribution. To compare two groups, a non-parametric Mann-Whitney test was used. When comparing multiple groups, a Kruskal-Wallis with Dunn's multiple comparison was used if normality was not achieved. For functional recovery
First, the potency of the HDAC inhibitors scriptaid and RGFP966 has been confirmed by testing their effects on histone acetylation. The expression of histone 3 and 4 acetylation was analyzed by means of a western blot on total protein lysates from BMDMs treated with these inhibitors. Both RGFP966 and scriptaid significantly increased histone 3 and 4 acetylation (Fig. 1).
Secondly, the effects of scriptaid and RGFP966 were examined on macrophage morphology. Macrophages were left either unstimulated (control) or were stimulated with LPS, IL-4 or IL-13 to obtain M1 or M2 macrophages respectively. As expected, there are clear morphological differences between M0, M1 and M2 macrophages (Fig. 2A~T). M1 macrophages display a flattened, rounded shape while M2 macrophages are more elongated and spindle shaped . Macrophages treated with scriptaid (0.2 and 1 µM) have a similar morphology compared to control (Fig. 2A~C). Treatment with RGFP966 results in elongated cells, similar to an M2 phenotype morphology especially in the 10 µM condition (Fig. 2A, D, E). M1 macrophage morphology after LPS treatment is shown in Fig. 2F. Treatment of LPS-pre-stimulated macrophages with both HDAC3 inhibitors leads to elongated cells (Fig. 2G~J), which is most prominent in the higher concentrations (1 µM scriptaid and 10 µM RGFP966, Fig. 2H and 2J). In Fig. 2K and 2P the typical M2 morphology is shown after IL-4 and IL-13 treatment respectively. Treatment of these cells with scriptaid make them resemble the control cells (Fig. 2L~M, and Q~R vs. Fig. 2A). Treatment with RGFP966 of the IL-4 and IL-13 pre-stimulated cells does not change the morphology (Fig. 2N~O and S~T vs. Fig. 2K and P).
In a next step, the influence of scriptaid and RGFP966 on the gene expression of selected M1 (CD86, iNOS, IL-6, IL-1β GPR18, CD38 and FPR2) and M2 (Arg-1, Ym1 and Fizz1) markers was examined. When M1 macrophages were treated with scriptaid, the gene expression of CD86 and iNOS was increased (Fig. 3B, C), while the gene expression levels of GPR18, CD38 and FPR2 were decreased (Fig. 3F~H). There was no effect on the gene expression of IL-6 and IL-1β (Fig. 3D, E). When M2 macrophages were treated with scriptaid, the gene expression levels of Arg-1 and Ym-1 were not affected (Fig. 4A, B) while Fizz-1 expression was reduced (Fig. 4C). Treating M1 macrophages with RGFP966 had no effect on the gene expression of iNOS, IL-6, IL-1β, GPR18, CD38 and FPR2 (Fig. 3B~G), but increased the gene expression of CD86 (<0.5-fold change vs control+LPS; Fig. 3B). IL-4-stimulated M2 macrophages showed an increased expression of Arg-1 and of Ym-1 after RGFP966 treatment (Fig. 4A, B). The gene expression level of Fizz-1 was unchanged (Fig. 4C). When the macrophages were treated with IL-13, there was an increase in Arg-1 gene expression and a decrease in Fizz-1 gene expression after RGFP966 treatment (Fig. 4D~F). There was no effect on the gene expression of Ym-1 (Fig. 4E). Briefly, RGFP966 promoted the gene expression of selected M2 markers (Arg-1 and Ym-1) after IL-4 or IL-13 stimulation and did not affect the expression of the investigated M1 markers (iNOS, IL-6, IL-1β, GPR18, CD38, FPR2), except for CD86.
Additionally, we investigated the effect of scriptaid and RGFP966 on the protein expression of the M1 markers iNOS, NO and IL-6 . Macrophages were stimulated with LPS to obtain an M1 phenotype followed by treatment with the inhibitors. Both, IL-6 and NO production were significantly downregulated after scriptaid treatment (Fig. 5A, B), while iNOS expression was left unchanged (Fig. 5C). On the protein level, RGFP966 significantly reduced the levels of IL-6 but had no effect on NO production (Fig. 5A, B) and iNOS expression (Fig. 5C).
To further investigate the effects of scriptaid and RGFP966 on macrophage polarization on the protein level, we studied the impact of HDAC3i treatment on the expression of Arg-1, a well-established M2 marker [28,29]. The macrophages were skewed towards an M2 phenotype using IL-4 or IL-13 followed by scriptaid or RGFP966 treatment. IL-4-induced Arg-1 upregulation is fully eliminated by scriptaid, while IL-13-induced Arg-1 upregulation was only partially reduced (Fig. 6A~D). RGFP966 significantly increased the Arg-1 expression by M2 macrophages (Fig. 6A~D) and had no effect on arginase activity (Fig. 6E, F).
After SCI, foamy macrophages are persistent and become proinflammatory due to lipid overload . Since HDAC3 deficiency in macrophages improves their lipid handling , we investigated the effects of HDAC3 inhibition on the formation of foamy macrophages. M0, M1 and M2 macrophages were treated with scriptaid and RGFP966 after which they were incubated with spinal cord debris to become foamy, as visualized by ORO staining. Scriptaid did not affect the foaminess of the M0 or M2 macrophages (Fig. 7A, B) but reduced the foaminess of M1 macrophages (Fig. 7C). Analysis of ORO staining's showed that RGFP966 significantly reduced the formation of foamy macrophages regardless of the macrophage phenotype (Fig. 7A~C).
Finally, we investigated the effects of RGFP966 and scriptaid on functional recovery after SCI using the BMS and examined whether HDAC3 inhibition affected lesion size, demyelinated area and immune cell infiltration after SCI. Functional and histological analyses after SCI showed no significant differences between the mice treated with scriptaid and the vehicle control group (Figs. 8, 9, 10), except for the level of astrogliosis which were significantly increased in the treatment group compared to the control group (Fig. 8B). Assessing the BMS for 27 days after applying RGFP966 showed no significant difference in functional recovery between the treatment group and the vehicle group (Fig. 8A). On the histological level, there were no differences in lesion size, astrogliosis, demyelinated area and in the numbers of Iba-1, CD4+, MHCII+ and Arg-1+ infiltrating cells between the vehicle group and the group treated with RGFP966 (Figs. 8, 9, 10).
In this study we investigated whether HDAC inhibition promotes functional recovery by polarizing the perilesional environment towards an anti-inflammatory milieu. Our findings indicate that the non-specific inhibitor scriptaid, reduces the expression of the anti-inflammatory macrophage marker Arg-1
Our results demonstrate that RGFP966 has no substantial effect on the phenotype of pro-inflammatory macrophages and promotes the alternative activation of macrophages. Similarly, Mullican et al.  showed that HDAC3 restricts many IL-4 targeted genes in macrophages (e.g. Arg-1, Chi3I3 and Clec7a). They demonstrated that HDAC3 deficiency prevents the reduction in IL-4 targeted genes and promotes the switch towards an IL-4 induced alternatively activated macrophage.
Phagocytosis also plays a major role in macrophage polarization. After SCI, macrophages become foamy via phagocytosis of spinal cord debris, thereby converting them to an M1 phenotype . In atherosclerosis, it has been shown that HDAC3 deletion reduces lipid accumulation . Similarly, we wanted to investigate whether HDAC3 inhibition with HDAC inhibitors could decrease the formation of foamy macrophages
Scriptaid, which inhibits HDAC1 and HDAC8 in addition to HDAC3, displayed differential effects on macrophage phenotype. Unexpectedly, in our hands, scriptaid reduced the protein expression of the M2 marker Arg-1, whereas others have shown that loss of HDAC3 promotes M2 polarization . These contrasting results may be due to the inhibition of HDAC1 besides HDAC3 by scriptaid. HDAC1 has been shown to suppress LPS-induced genes, characteristic for an M1 phenotype, hence inhibition of HDAC1 may lead to M1 macrophage polarization . In addition, previous studies have also shown contradictory results when using HDAC inhibitors depending on the cell type or even within the same cell type. A potential explanation is that HDAC inhibition affects transcription factors that activate diverse sets of genes with contradicting effects as a result . Therefore, future studies should elucidate the signalling pathways modulated by HDACs to gain better understanding of the underlying mechanisms when using HDAC inhibitors as a treatment for CNS disorders.
Both inhibitors were administered individually in the T-cut hemisection mouse model and we monitored the functional recovery using the BMS. Surprisingly, neither scriptaid nor RGFP966 affected functional recovery. Consistently, the histological analyses indicated no differences between the treatment groups and the control groups.
In a parallel study, it was shown that RGFP966 significantly improved functional recovery after a
In conclusion, HDAC3 inhibition by RGFP966 promotes the alternative activation of macrophages and reduces the formation of foamy pro-inflammatory macrophages
|Gene||Forward primer||Reverse primer|
Arg-1, arginase-1; CD38, cluster of differentiation 38; CD86, cluster of differentiation 86; Fizz, found in inflammatory zone; FPR2, formyl peptide receptor 2; GPR 18, G protein-coupled receptor 18; HMBS, hydroxymethylbilane synthase; IL-6, interleukin 6; iNOS, inducible nitric oxide synthase; Ym1, chitinase-like 3; YWHAZ, 14-3-3 protein zeta/delta.