Induction of Krüppel‐like factor 2 reduces K/BxN serum‐induced arthritis

Abstract Krüppel‐like factor 2 (KLF2) critically regulates activation and function of monocyte, which plays important pathogenic role in progressive joint destruction in rheumatoid arthritis (RA). It is yet to be established the molecular basis of KLF2‐mediated regulation of monocytes in RA pathogenesis. Herein, we show that a class of compound, HDAC inhibitors (HDACi) induced KLF2 expression in monocytes both in vitro and in vivo. KLF2 level was also elevated in tissues, such as bone marrow, spleen and thymus in mice after infusion of HDACi. Importantly, HDACi significantly reduced osteoclastic differentiation of monocytes with the up‐regulation of KLF2 and concomitant down‐regulation of matrixmetalloproteinases both in the expression level as well as in the protein level. In addition, HDACi reduced K/BxN serum‐induced arthritic inflammation and joint destruction in mice in a dose‐dependent manner. Finally, co‐immunoprecipitation and overexpression studies confirmed that KLF2 directly interacts with HDAC4 molecule in cells. These findings provide mechanistic evidence of KLF2‐mediated regulation of K/BxN serum‐induced arthritic inflammation.

Krüppel-like transcription factor (KLF)s are involved in regulation of terminal differentiation of several immune cells. Lung KLF (LKLF) or KLF2 plays a significant regulatory role in haematopoietic cell biology including cell quiescence, cell proliferation, differentiation and survival. 9,10 KLF2 also regulates myeloid cell activation and function. 11 During inflammation, KLF2 acts as a potent inhibitor of transcription factors like NF-κB and AP1, as well as of hypoxia-related HIF-1α protein. 12 Importantly, KLF2 is essential for embryonic erythropoiesis and KLF2-deficient mice are embryonic lethal because of the leaky blood vessels and haemorrhages. 13 Krüppel-like factor 2 is a negative regulator of monocyte activation and osteoclast differentiation, directs them towards quiescent state. 11,14 The current study focuses on assessing role of a special class of pharmacological compounds, histone deacetylase inhibitors (HDACi) in modulation of KLF2 in monocytes both in vitro and in vivo. HDACi have been used clinically for the treatment of various cancer malignancies. 15 HDACs are well known for their role in chromatin remodelling, 16 and also regulate large number of non-histone proteins, which controls cell processes associated to various disease conditions. 17 Evidence show that HDACi controls NFκB-driven inflammatory responses through cell cycle arrest. 18 However, the effect of HDACi on inflammatory response varies according to the cell type and stimulus. As cell cycle arrest leads to cell quiescence, and KLF2 is induced upon cell quiescence, we were interested to know whether HDACi can mediate induction of KLF2. Among the common HDACi are suberoylanilide hydroxamic acid (SAHA), entinostat (SNDx-275 or MS-275) and Trichostatin A (TSA) are of consideration. In addition, it is yet to be defined how HDACi regulate myeloid KLF2 and RA pathogenesis.
To investigate the efficacy of HDACi against RA progression and to delineate the mechanistic evidence have driven us to explore this intrigued pathogenesis.

| Isolation and culture of monocytes from murine bone marrow
Bone marrow (BM) cells were isolated from the femurs of 6-to 8-week-old C57BL/6 mice, and cultured in DMEM medium (Sigma) supplemented with 10% foetal bovine serum (ThermoFisher, Grand Island, NY, USA) at 37°C in a 5% CO 2 atmosphere for plastic adherence. Adhered monocytes were either stimulated with suberoylanilide hydroxamic acid (SAHA from Sigma), entinostat (SNDx-275 or MS-275; Cayman Chemical Company, Ann Arbor, MI, USA) and Trichostatin A (TSA from Sigma) for either 24 or 48 hours or cultured without stimulus (DMSO, vehicle) for control.

| In vivo stimulation of mice with HDACi
To investigate whether systemic stimulation with HDACi to the mice have any effect on immune cell-rich tissues, three mice per group were injected intra peritoneal (ip) with MS275 (5 mg/kg BW in 50 μL DMSO) and TSA (10 mg/kg BW in 50 μL DMSO) or DMSO (50 μL, as a control). After 24 hours of HDACi injections, BM, spleen and thymus were harvested. A part of the tissues was homogenized and RNA was extracted, and another part was subjected to protein extraction.

| Isolation of human monocytes
Fresh human peripheral blood (n = 6) was collected with an approved IRB and written consent from donors from The Ohio State University Medical Center, Columbus, OH and processed following earlier described protocol. 11,19 In, brief, peripheral blood mononuclear cells were isolated from freshly collected blood using Ficoll-Paque density centrifugation. CD14 + cells were isolated by using an AutoMACS device, CD14 + antibody and reagents (all from Miltenyi Biotec, San Diego, CA, USA) following earlier established protocol. 11,20 A part of the CD14 + cells was used for RNA extraction, and another part was subjected to protein extraction.

| Real time RT-PCR analysis
Total RNA was isolated from BM-derived monocytes, RAW 264.7 cell (ATCC, Manassas, VA, USA) and human peripheral blood-derived monocytes after 24 hours of HDACi stimulation, or murine tissues after infusion with HDACi to mice using a RNeasy Kit (Qiagen, Ther-moFisher). One microgram of RNA was used for synthesis of cDNA using oligo dT (Invitrogen, ThermoFisher) primer. Real-time RT-PCR was performed with 1 μL of cDNA for the gene-specific primers of KLF2, matrixmetalloproteinase 3 (MMP3), MMP9 and MMP13, keeping GAPDH as an internal control using a standard SYBR green Taqman protocol, and a real-time PCR machine (MX3000P; Stratagene, Santa Clara, CA, USA). Relative fold-expression levels of stated genes were measured considering respective unstimulated cells or tissues (added DMSO) as controls. Experiments were performed in triplicate and were repeated at least three times.

| Histological assessment of arthritis
Arthritis was assessed by histological examination as described, 14 with some modifications. Limbs were fixed in periodate-lysine-paraformaldehyde for overnight and decalcified in 10% EDTA (BDH Chemicals, Victoria, Australia) and 7.5% polyvinylpyrolidone (Sigma) in Tris buffer (pH 6.95) for 7-10 days and processed for paraffin embedding. Tissues were sectioned with 5 μmol/L thickness, placed on aminoalkylsilane-coated slides, and stained for routine histology with haematoxylin and eosin (H&E). Five defined pathological features were graded for severity from 0 (normal) to 5 (severe), according to established protocol, 24 and in a blinded manner. Soft tissue inflammation, assessed in the infrapatellar fat pads, joint capsule, and the area adjacent to the periosteal sheath, was graded according to the extent of cellular infiltration and angiogenesis. Joint space exudate was identified as leucocytes scattered discretely or in aggregates in the joint space. Synovitis (synovial hyperplasia) was defined as hyperplasia of the synovium, but did not include pannus formation. Pannus was defined as hypertrophic synovial tissue forming a tight junction with the articular surface. Evaluation of cartilage and bone damage was based on loss of cartilage matrix, disruption and loss of cartilage surface, and the extent and depth of the subchondral bone erosion. 25 A trained pathologist, Department of Pathology, The Ohio State University Medical Center performed the histomorphometric analyses from multiple H&E sections.

| Statistical analysis
Values were expressed as mean ± SEM and statistical analysis was performed by ANOVA. Student's t test was performed and the results were considered significant when P-values were <0.05.

| Effect of HDAC inhibitors on KLF2 expression in monocytes
To test the effect of HDACi on KLF2 gene expression, three different types of HDACi (MS275, SAHA and TSA) at various concentrations (1-50 nmol/L) were added to the murine monocytic cells (RAW 264.7), and DMSO was added as a vehicle control. After 24 hours of incubation, RNA was isolated from harvested cells, were subjected to KLF2 gene expression using the quantitative RT-PCR method, and simultaneously cell viability was verified by a Trypan blue exclusion test. All HDACi induced KLF2 at various degrees (2-to 17-fold, Figure S1

| Effect of HDAC inhibitor on osteoclastic differentiation of monocytes
We next sought to find whether HDAC inhibitors have any effect on monocyte differentiation to osteoclasts, as they are the important cells in arthritic tissue damages. 26 Figure 4F). The Western blot data was quantified and shown graphically ( Figure S2).

| Effect of HDAC inhibitors in K/BxN seruminduced arthritis in mice
As HDAC inhibitor induced KLF2 in monocytes and reduced osteoclastic differentiation of monocytes, we next tested whether HDAC inhibitor has any effect on arthritic inflammation and severity in joint destruction in mice. Mice were induced arthritis using K/BxN serum and treated with HDAC inhibitors (either SAHA, 10 mg/kg body weight; or SAHA 30 mg/kg body weight; or MS275, 5 mg/kg body weight) during the course of arthritis development for 7 days. Ankles were measured, and found that a significant reduction in inflammation after injection of HDAC inhibitor with SAHA 30 mg/kg body weight; or MS275, 5 mg/kg body weight from day 2 to day 7 ( Figure 5A). When the lower dose of SAHA (10 mg/kg body weight) was used as therapeutics, a marked reduction in ankle inflammation were also observed ( Figure 5A, left panel). These data show that ankle inflammation could be reduced after HDAC inhibitor therapy to K/BxN serum-induced mice.

| Effects of HDAC inhibitors on bone and cartilage damage upon K/BxN serum-induced arthritis
As HDAC inhibitor reduced ankle inflammation in K/BxN seruminduced arthritic mice, we next sought to find whether bone and cartilage damage were also reduced. H&E staining revealed that a reduced bone and cartilage damage after treatment with either of the HDAC inhibitors like SAHA or MS275 ( Figure 5B). The damage of bone and cartilage were significantly lower in animals ( Figure 5C) that were treated with SAHA (30 mg/kg body weight) or MS275 (5 mg/kg body weight).

| DISCUSSION
Rheumatoid arthritis is a progressive autoimmune disease with multifaceted pathobiology involving numerous cells, including monocytes, and various signalling pathways, to drive the inflammation. 28 KLF2, a transcriptional factor, is important in regulating endothelial cell development, homeostasis and activation, 29 and inhibiting proinflammatory activation of monocytes. 11,14 Therefore, delineating the regulatory role of KLF2 is very important in understanding the diverse roles of KLF2 in RA pathogenesis. On the other hand, HDACimediated inhibition of inflammatory gene expression 30 and immune cell activation 31 was reported; however, the exact mechanism is yet to be defined. Therefore, we sought to investigate whether HDACi has any effect on KLF2-mediated regulation of RA pathogenesis.
Among underlying processes in RA, KLF2-mediated regulation of osteoclastic differentiation is critically important. Finding the molecules that regulate KLF2 in monocytes could be useful in managing RA pathogenesis. Herein, we found increased KLF2 expressions both in RNA and protein levels after induction with several HDACi molecules ( Figure 1). We next tested HDACi effects on KLF2 expression in vivo. An increased level of KLF2 expression was found in murine BM, spleen and thymus tissues upon infusion with HDACi ( Figure 2). Interestingly, in vitro differentiation of monocytes to osteoclasts is remarkably reduced upon HDACi addition ( Figure 3). It is possible that the decreased capability of osteoclast differentiation was because of the up-regulated KLF2 level in monocytes. Our finding shows that the up-regulated KLF2 is not only decreased the proliferation of monocytes, but also reduced osteoclastic differentiation.
Out of many inflammatory signals, matrix-proteolytic enzymemediated signalling is highly responsible for destruction of cartilages, bones and articular structures. 32 MMPs can cleave adhesion molecules, cytokines, chemokines, growth factors, and their receptors and binding proteins. 33 Moreover, they also activate other proteases, and thus, mediate a cascade of severe matrix degradation process. 33 Several MMPs are elevated in the synovial fluid and in the serum of RA patients. 34 Herein, we show that reductions in osteoclastic differentiation upon HDACi addition to monocytes are associated with decreased expression of MMPs along with increased level of KLF2 ( Figure 4). Therefore, as metalloproteinase proteolytic axis ensures appropriate initiation and stop signals during the inflammatory reaction. Importantly, we found a significant dose-dependent reduction in ankle thickness upon HDACi with SAHA or MS275 administration to K/BxN serum-induced arthritic mice ( Figure 5). The damage of bone and cartilage was significantly reduced in HDACi-treated F I G U R E 6 HDAC directly interacts with KLF2. (A) Co-immunoprecipitation assays showed that KLF2 directly interacts with HDAC4 molecule. (B) Western blot data revealed that overexpression of KLF2 resulted in increased level of HDAC4 and concomitantly decreased level of MMP9, whereas, levels of HDAC1 and 3 were minimally changed. (C) Quantified Western blot data of stated molecules are presented graphically arthritic mice compared to the untreated control arthritic mice.
These data support HDACi's ability to exert anti-inflammatory effect on K/BxN serum-induced arthritic mice, which is in consistent with the earlier observations with different models. 35 Taken together, these data not only show the efficacy of HDACi against RA but also suggests a potential for future therapeutic modality against RA.
Finally, to determine whether KLF2 and HDAC interact each other in monocytes and mediate an effect on inflammatory transcriptional regulation process. Our data suggests that there is a clear interaction between HDACi and KLF2 in the transcriptional regulation level. We assume that acetylation of histone 3 proteins might play a key role in the HDACi regulated transcription progression, as we found the H3 acetylation is induced by HDACi. Histone acetylation causes alteration in nucleosomal conformation, 36 which can increase the accessibility of transcriptional regulatory proteins to chromatin templates. 37 Our earlier published data suggests that histone acetylation might result in increased transcriptional activity in monocytes by facilitating p65-and c-Fos/c-Jun-dependent transcriptional progression of NFκB and AP1 gene, even in the presence of KLF2. 11,14 However, transcriptional regulation did not lead to inflammatory protein synthesis, rather reduced. Because the level of KLF2 was increased in the presence of HDACi both in vitro as well as in vivo, that mediated inhibition to the inflammatory gene expression and protein synthesis.
In addition, KLF2 directly binds to the HDAC4 in cells ( Figure 6A