Histone deacetylase 3 facilitates TNFα-mediated NF-κB activation through suppressing CTSB induced RIP1 degradation and is required for host defense against bacterial infection

As important enzymes regulating acetylation, histone deacetylases (HDACs) participate in a series of cell physiological process. However, the mechanisms responsible for individual HDAC family members in regulating innate immunity remained to be elucidated. Here we sought to reveal the mechanism of HDAC3 in regulating the inflammatory response of macrophages. RNAseq was done to detect the transcriptional influence of HDAC3 on macrophages. Kyoto Encyclopedia of Genes and Genomes was used to reveal the change of signaling pathways after HDAC3 knockout. CHIPseq was done to detect the deacetylation modification of HDAC3 on chromosome. Western blot, immunofluorescence, and real-time quantitative PCR were used to measure the change of genes and proteins’ levels. Mice were intratracheal instillation with lipopolysaccharide or Pseudomonas aeruginosa to determine the influence of HDAC3 on inflammatory response in vivo. HDAC3-deficient macrophages had increased expression of cathepsins resulting from elevated histone acetylation. Over-expressed cathepsins such as cathepsin B (CTSB) caused remarkable degradation of receptor (TNFRSF)-interacting serine-threonine kinase 1 (RIP1), which reduced TNFα mediated NF-κB activation and inflammatory response. Consistently, mice with macrophage specific knockout of HDAC3 were impaired in inflammatory response and thereby susceptible to Pseudomonas aeruginosa infection. HDAC3 was required for protecting RIP1 from degrading by CTSB in macrophages. Decreased RIP1 in HDAC3 knockout macrophages impaired TNFα mediated NF-κB activation. Our studies uncovered important roles of HDAC3 in the regulation of cathepsin-mediated lysosomal degradation and RIP1-mediated inflammatory response in macrophages as well as in host defense against bacterial infection.


Introduction
As important deacetylases, the function of HDACs has been well studied [1][2][3]. Through interacting with histones and non-histones, they exert various effects on chromatin remodeling, transcriptional regulation and protein stability [4], with important contributions to physiological or pathological processes. Eighteen HDAC family members of eukaryotes are divided into four classes. Class I (HDAC1, HDAC2, HDAC3, HDAC8), Class II (HDAC4, HDAC5, HDAC6, HDAC7, HDAC9, HDAC10) and Class IV (HDAC11) HDACs are Zn ++ -dependent enzymes, while Class III (Sirt1-Sirt7) HDACs are NAD + -dependent enzymes [5]. The importance of HDACs promotes the development of HDAC inhibitors, five of which have been approved by US Food and Drug Administration (FDA) or China FDA (CFDA) for the treatment of hematological tumors [6]. The influence of HDACs on macrophages has also been taken notice recently by studying both repression and activation of gene transcription in mouse bone-marrowderived macrophages (BMDM) with myeloid-specific loss of HDAC3 [7]. However, the mechanisms responsible for individual HDAC family members in regulating inflammatory response in macrophages and in host defense against bacterial infections remained to be elucidated.
Lysosomal degradation system is the major intracellular degradation system. Both extracellular materials and cellular components can be delivered to the lysosome for degradation [8]. Cathepsins are a family of proteases [9], acting as important executors of lysosomal degradation system through digesting of internalizes wasted cell proteins and peptides. Most of the cathepsin members belong to cysteine proteases, and a few are aspartic protease (CTSD, CTSE) and serine protease (CTSA, CTSG). Abnormal expression of cathepsins can lead to the disorder of lysosomal degradation system's activity, inducing the unbalance of protein degradation and cell homeostasis. A series of articles have been reported the links between cathepsins and diseases including cancer [10], mental illness [11], etc.
RIP1 is an essential molecule in TNFα or Myd88independent Toll-like receptor (TLR) activation induced NF-κB activity. Reduced NF-κB activity and inflammatory response was observed in cells with RIP1 genetic interference [12,13] or drug inhibition [14]. In accordance with these, in vivo application with Nec-1s, a RIP1 inhibitor showed decreased inflammatory response after injection with LPS [15]. However, impaired inflammatory response can sensitize host to infection. RIP1 deficient mice showed more death after infection with Yersinia [16,17] or Candida albicans [18], indicating the key role of RIP1in host defense against bacterial and fungus infections.
Here we reported that HDAC3 was a critical transcriptional repressor of cathepsins. Elevated expression of cathepsins in HDAC3 deficient macrophages results in remarkable lysosomal degradation of RIP1, leading to reduced TNFα mediated NF-kB activation and the inflammatory response. More importantly, macrophage specific HDAC3 knockout mice were impaired in host defense against bacterial infection. These results indicated a novel function of HDAC3 in host defense against bacterial infection by suppressing cathepsin-mediated RIP1 degradation and also raised a potential concern of opportunistic bacterial infections during clinical application of HDAC inhibitors.

Elevated expression of cathepsins in HDAC3 deficient macrophages
In order to explore the role of HDAC3 in the regulation of the cellular activity of macrophages, we constructed HDAC3 knockout murine monocytic cell line RAW264.7 cells through the CRISPR/Cas9 gene editing system (Fig. 1A). RNAseq was done to detect the change after HDAC3 knockout. Results showed 1333 genes and 1136 genes were significantly upregulated or downregulated after HDAC3 knockout, respectively (Fig. 1B). KEGG analysis of genes significantly changed after HDAC3 knockout showed that HDAC3 deficiency in macrophages significantly upregulated genes involved in the lysosomal pathway (Fig. 1C, D). We surprisingly found that the mRNA levels of multiple cathepsins which are important executors for protein degradation of lysosomal degradation system mentioned above, including Ctsa, Ctsb, Ctsd, Ctsl, Ctss and Ctsz, were significantly elevated response. Consistently, mice with macrophage specific knockout of HDAC3 were impaired in inflammatory response and thereby susceptible to Pseudomonas aeruginosa infection.
Conclusion: HDAC3 was required for protecting RIP1 from degrading by CTSB in macrophages. Decreased RIP1 in HDAC3 knockout macrophages impaired TNFα mediated NF-κB activation. Our studies uncovered important roles of HDAC3 in the regulation of cathepsin-mediated lysosomal degradation and RIP1-mediated inflammatory response in macrophages as well as in host defense against bacterial infection.
CTSB, a critical protein in the lysosomal system, was among the cathepsins with the highest expression in macrophages (Fig. 1F). In consistent with mRNA expression, the protein levels of CTSB detected by western blot and confocal analysis were also significantly higher in Hdac3 −/− as compared to Vector RAW264.7 cells ( Fig. 2A-D).
We have further generated BMDMs from Lysm Cre and Lysm Cre Hdac3 f/f mice. The successful loss of HDAC3 in Lysm Cre Hdac3 f/f BMDMs was confirmed by PCR (Additional file 1: Fig. S1A) and western blot analysis (Additional file 1: Fig. S1B). And we found that HDAC3-deficient BMDMs had higher levels of CTSB mRNA and protein than Lysm Cre BMDMs ( Fig. 2E-G). We also explored the influence of other members of Class I HDACs on CTSB with HDAC1, HDAC2, HDAC8 knockout RAW264.7 cells generated by CRISPR/Cas9. Besides HDAC3, which had the most obvious effect on CTSB, HDAC2 and HDAC8 knockout cells also showed slightly elevated CTSB expression, while HDAC1 had no influence on the expression of CTSB ( Fig. 2H-J). We also observed among Class I HDACs, CTSD was expressed at the highest levels in Hdac3 −/− RAW264.7 (Additional file 1: Fig. S2A-C). Taken together, these data indicated the important role of HDAC3 in suppressing the expression of CTSB.

HDAC3-dependent histone deacetylation on the promoters of cathepsins
As histone deacetylation has negative effects on transcription and HDAC3 is an important deacetylase, we speculated that HDAC3 knockout might promote the histone acetylation on the promoter of these cathepsins to increase their expression. Confocal analysis revealed that HDAC3 was mainly located in the nucleus (Additional file 1: Fig. S3). Chromatin immunoprecipitation (ChIP) was then performed against H3K27ac. Correlation matrix between ChIP-seq experiments demonstrated good consistency between samples (Fig. 3A). The CHIP peak-calling program Model Based Analysis for CHIP-Seq data (MACS2) revealed that H3K27ac located near the transcriptional start sites (TSS) (Fig. 3B, C) and its binding was enriched on the promoter of the genes (Fig. 3D), indicating the regulation of acetylation on the transcription. More importantly, we observed significantly increased acetylation on the promoter of Ctsa, Ctsb, Ctsd, Ctsl, Ctss, Ctsz in HDAC3-deficient macrophages as compared to the corresponding Vector macrophages ( Fig. 3E-J). These data indicated that HDAC3 may negatively regulated the expression of these cathepsins through deacetylase the histones on their promoters.

Reduced protein level of RIP1 in HDAC3 deficient macrophage
CTSB has been reported to be involved in the cleavage of RIP1 [19], which is a critical molecule on the signaling pathway for TNFα mediated NF-κB activation [20,21]. We therefore explored the influence of the elevated CTSB expression in HDAC3 knockout macrophages on RIP1 expression. Decreased expression level of RIP1 protein was observed in Neuro-2a cells transfected with CTSB ( Fig. 4A-C) and also in HDAC3 deficient RAW264.7 and BMDMs as compared with the corresponding control cells ( Fig. 4D-G). Furthermore, application of either Leupeptin, an inhibitor of the lysosomal system or CA-074, an inhibitor of CTSB, restored the protein level of RIP1 in HDAC3 deficient RAW264.7 cells ( Fig. 4H-K). The same phenomenon was also observed in HDAC3 deficient BMDMs ( Fig. 4N-Q).
As CTSB degrades proteins mainly in lysosomes, we also observed the co-localization of RIP1 and lysosomes in Neuro-2a cells. Cells treated with CA-074 had increased the resident of RIP1 in lysosomes as compared to control cells, indicating CTSB mediated RIP1 degradation mainly occurred in lysosomes (Additional file 1: Fig. S4). Lysosomal system and ubiquitin-proteasome are two main ways involved in protein degradation. We also applied cells with MG132, an inhibitor of the proteasome mediated degradation system and found it failed to restore the protein level of RIP1 in HDAC3 knockout macrophages (Fig. 4L, M). The above results indicated that the decreased protein level of RIP1 was caused by the over-reactive lysosomal degradation after HDAC3 knockout. We also explored the protein level of RIP1 in Vector, Hdac1 −/− , Hdac2 −/− , Hdac3 −/− , Hdac8 −/− RAW264.7 cells and found HDAC3 deficiency in macrophages resulted in the most significantly decreased protein level of RIP1, which was consistent with the increase of CTSB protein levels ( Fig. 4R, S).

The role of HDAC3 in host defense against pseudomonas aeruginosa infection
To further validate the effect of HDAC3 on inflammatory response of macrophages in vivo, Lysm Cre and Lysm Cre Hdac3 f/f mice were intratracheally instilled with endotoxin (LPS). The BALF was collected for the detection of the inflammatory cytokine secretion. Decreased secretion of IL-6 and TNFα was found in Lysm Cre Hdac3 f/f mice (Fig. 6A, B), indicating the impaired inflammatory response in vivo. Insufficient inflammatory response lowers the body's resistance to infection. We further conducted lung infection experiments in Lysm Cre and Lysm Cre Hdac3 f/f mice by intratracheal drip of pseudomonas aeruginosa. Consistent with our expectations, the survival rate of Lysm Cre Hdac3 f/f was significantly lower than Lysm-Cre mice (Fig. 6C). The basic status of Lysm Cre Hdac3 f/f mice were also significantly worse, manifested by a remarkable reduced activity, lower body temperature (Fig. 6D) and difficulty in breathing. In addition, higher numbers of bacteria were accumulated in Pseudomonas aeruginosa infected Lysm Cre Hdac3 f/f than Lysm Cre mice (Fig. 6E).
Further pulmonary histopathological examination showed that Pseudomonas aeruginosa infection led to more severe lung injury, manifested as pulmonary edema, hemorrhage, inflammatory cell infiltration, and cellulose exudation, in Lysm Cre Hdac3 f/f than Lysm Cre mice (Fig. 6F, G). Increased hemorrhage was also showed as the color of the BALF of Lysm Cre Hdac3 f/f was redder than Lysm Cre mice after Pseudomonas aeruginosa infection (Fig. 6H). HMGB1 is a conserved nuclear protein in cells. It is loosely bound to chromosomes and can be passively secreted through the broken cell membrane in necrotic cells. Thus, it can be an indicator of lung injury. We observed that the levels of HMGB1 in Lysm Cre Hdac3 f/f mice were higher than those in Lysm Cre mice after infection (Fig. 6I). In conclusion, HDAC3 was required for inhibiting the expression of Ctsb through possible deacetylating the histone on its promoter. Degradation of RIP1 resulted from elevated expression of Ctsb in HDAC3 deficient macrophages impaired TNFα mediated NF-κB activation and eventually impaired the inflammatory response and anti-infection immunity of host (Fig. 6J).

Discussion
As the major protein post-translational modification, acetylation which depends on the balance between HAT and HDAC determines the normal development, structure and function of cells. Abnormal expression or mutations of HDAC family members are associated with numerous diseases including cancers and autoimmune diseases. As more and more HDAC inhibitors have been and are being developed, potential unknown targets and side effects of these HDAC inhibitors should be considered given the broad physiological and pathology functions of this family members. Our studies here have provided evidence demonstrating that HDAC3 is an important regulator in suppressing molecule lysosomal degradation system, and facilitating inflammatory response. HDAC3 knockout resulted in overexpressed CTSB through promoting acetylation of the histones on its promoters. The increased CTSB acted as a scissor for the degradation of RIP1, leading to the impaired activation of TNFα-induced NF-κB signaling pathway. More importantly, mice with macrophage specific knockout of HDAC3 showed more susceptible than control mice to  HDACs may affect the lysosomal degradation system through multiple steps, including the influence on the lysosomal pH, the expression of lysosome related protein, autophagosome-lysosome fusion, lysosomal exocytosis and the acetylation modification of targeted protein.
Reduced lysosomal pH was observed in neck squamous cell carcinoma cell line applied with HDAC inhibitor trichostatin A, leading to increased cathepsin activity and cell death [23]. Another HDAC inhibitor SAHA was able to increase the transcriptional activity of TFEB [24] and the expression of LAMP1, LAMP2, CTSB [25]. HDAC10 was required for depressing LAMP2A expression [26]. The deacetylation of syntaxin 17 (STX17), a key autophagosomal SNARE protein mediating autophagosome maturation by HDAC2 [27] and Cortactin by HDAC6 [28] promotes autophagosome-lysosome fusion. As for lysosomal exocytosis, HDAC inhibitor ITF2357 was able to induce hyperacetylation of tubulin, preventing the exocytosis of IL-1β-containing secretory lysosomes [29]. In addition, the acetylation modification of targeted protein such as PKM [30] and Tau [31] promoted themselves degraded by lysosomal system. However, these previous studies are mainly based on the effects of HDAC inhibitors on particular targets, the role of specific HDAC in overall lysosomal degradation system is not clear. Here we first reported that HDAC3 was a key regulator in suppressing the expression of multiple cathepsin genes through deacetylating histones on their promoter regions. Moreover, our RNAseq analysis between control and HDAC3 knockout macrophages have identified many more HDAC3 suppressed genes in the lysosomal degradation system. Further studies especially proteomic studies are needed to understanding the broad impact of HDAC3 in controlling lysosomal degradation system and potential side effects of using HDAC inhibitors.
Inflammatory response as an important part of immune system involves regulation of numerous proinflammatory genes such as TNF-α, IL-1β and IL-6 and anti-inflammatory genes such as IL-10, IL-4 and TGF-β. Defective inflammatory response may lead to uncontrolled infectious diseases and cancers whereas over reactive inflammatory response may be associated with various inflammatory diseases. Although many different signal transduction pathways are involved in regulating inflammatory response, the NF-kB pathways are known to play critical roles. Previous studies have reported that the CTSB inhibitor CA-074 decreased NF-κB activity and inflammatory factors' expression [32] and CTSB knockdown inhibited doxorubicin-induced NF-κB activation [33], suggesting an important role of CTSB for inflammation. However, the molecular mechanisms responsible for regulation of inflammatory response by HDAC3 and cathepsins are not clear. We have reported that HDAC3deficient macrophages have elevated levels of cathepsins and CTSB promoted RIP1 degradation through the lysosomal degradation pathway. We have also shown that HDAC3-deficient macrophages are impaired in NF-kB activation and downregulating numerous inflammatory genes in response to TNFa stimulations. More importantly, we have demonstrated that mice with macrophage specific HDAC3 knockout are more susceptible to infection with bacteria such as pseudomonas aeruginosa. Currently, US FDA has approved four HDAC inhibitors (Vorinostat, Romidepsin, Belinostat, and Panobinostat) and Chinese FDA has approved two HDAC inhibitors (Chidamide and HBI-800) for various cancer treatments, while many more HDAC inhibitors are at different stages of clinical trials. In addition, specific inhibitors for individual HDAC family members including proteolysis-targeting chimera (PROTAC) compounds that can mediate degradation of specific HDAC family members are being developed. Our studies uncover the role of HDAC3 in facilitating NF-κB activation and inflammatory responses by preventing cathepsin-mediated RIP1 degradation as well as the requirement of its expression in macrophages for host defense against bacterial infection, which indicates the importance of future work in considering the risk factors of infections in the clinical applications of HDAC inhibitors.

Quantitative real-time PCR (RT-qPCR)
RNA of 2 × 10 6 cells seeded on six-well plates was extracted by NucleoZol (MNG). Briefly, 600 μl Nucleo-Zol was added with 240 μl ddH 2 O and then centrifugated at 12,000g, 4 ℃ for 15 min, and 700 μl supernatant was added with 700 μl isopropanol. After centrifugation, the RNA precipitation was washed by 75% ethanol for two times. PrimeScript ™ II 1st Strand cDNA Synthesis Kit (Takara, #6210A) was used for Reverse transcription of RNA into complementary DNA (cDNA). And SYBR Green (Takara) was used for qPCR. All the steps are done according to the manufacturer's instructions. The sequences for primer pairs are listed in Additional file 1: Table S2. RNAseq was done by the company (Novogene) for database construction and sequencing.

Western blot
Cells were lysed for 30 min on ice by lysis buffer (Beyotime Biotechnology, #P0013) which contained phosphatase inhibitor cocktail (Roche, #04906845001) and complete protease inhibitor cocktail (Roche, #04693132001). After centrifugation at 12,000g, 4 ℃ for 15 min and boiled with 4X Protein SDS PAGE Loading (Takara, #9173) for 10 min, the proteins were separated through SDS-PAGE gel and transferred onto PVDF membrane. The membrane was blocked with 5% skimmed milk for 2 h, and then incubated with interested antibodies at 4 ℃ overnight. After stained with HRP-coupled secondary antibodies for 2 h, the proteins on membrane were then detected by Chemiluminescence system (Bio-Rad ChemiDoc MP).

Chromatin immunoprecipitation (ChIP)
Chromatin IP Kit (CST, #9005) was used for Chromatin Immunoprecipitation. All process was done according to manufacturer's protocol. In brief, RAW264.7 cells in 20 ml DMEM complete medium were added with 540 μl
Neuro-2a cells seeded on Glass Bottom Dish were transfected with 1.5 μg PCMV-C-EGFP-RIP1 for 24 h and CA-074 was added at 12 h after transfection. Then the cell culture medium was removed and lyso-tracker (Solarbio, #L8010, 1:15,000) staining solution was added for 1 h at 37 ℃. Cells were washed by PBS for 3 times and The nuclei were stained with Hoechst (Beyotime Biotechnology, #C1027, 1:100).
Cells were imaged on a confocal microscope (Nikon Tl-S).

Mouse acute lung injury (ALI) model
Hdac3 f/f mice on a C57BL/6 J background were bought from the Jackson Lab (Stock number: 024119). The Lysm-Cre mice were bought from the Jackson Laboratory (Stock number: 004781). All mice were bred in Animal center of Suzhou Institute of Systematic Medicine. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Suzhou Institute of Systems Medicine (ISM-IACUC-0019-R). Age-matched male Lysm Cre and Lysm Cre Hdac3 f/f littermates were used for all experiments. In LPS-mediated acute lung injury model, 5 mg/kg LPS was injected into the airway of mice. After 12 h, the mice were executed for BALF extraction. In pseudomonas aeruginosa-mediated acute lung injury model, the mice were intratracheal instillation of a dose of pseudomonas aeruginosa (2 × 10 7 CFU/per mouse). The survival rate was observed every 12 h up to 120 h. The temperature of mice was monitered until death appears. BALF extraction and lung tissue collection were done in 36 h after pseudomonas aeruginosa infection.

Bronchoalveolar lavage fluid (BALF) extraction and analysis
An incision in the bronchus of mice was cut with scissors after anesthesia. Then the syringe was used for injecting 1 ml cold PBS into the airway and about 70-80% PBS was withdrawn. After twice the same steps, about 1.5 ml BALF was obstained. Centrifugation was performed to separate the cells and supernatants for inflammatory factors assay.

Histological analysis
The scissors were used to open the thoracic cavity of mice and the syringe was used to collect the blood from heart to avoid blood contamination after anesthesia. Then the lung was resected and fixed in 4% paraformaldehyde (PFA). The samples were then embedded with paraffin and sliced into 4 um sections for haematoxylin and eosin (H&E) staining. The injury score was calculated according to previous research [34].