HMGN2 regulates non‐tuberculous mycobacteria survival via modulation of M1 macrophage polarization

Abstract Non‐tuberculous mycobacteria (NTM), also known as an environmental and atypical mycobacteria, can cause the chronic pulmonary infectious diseases. Macrophages have been suggested as the main host cell to initiate the innate immune responses to NTM infection. However, the molecular mechanism to regulate the antimicrobial immune responses to NTM is still largely unknown. Current study showed that the NTM clinical groups, Mycobacterium abscessus and Mycobacterium smegmatis, significantly induced the M1 macrophage polarization with the characteristic production of nitric oxide (NO) and marker gene expression of iNOS, IFNγ, TNF‐α, IL1‐β and IL‐6. Interestingly, a non‐histone nuclear protein, HMGN2 (high‐mobility group N2), was found to be spontaneously induced during NTM‐activated M1 macrophage polarization. Functional studies revealed that HMGN2 deficiency in NTM‐infected macrophage promotes the expression of M1 markers and the production of NO via the enhanced activation of NF‐κB and MAPK signalling. Further studies exhibited that HMGN2 knock‐down also enhanced IFNγ‐induced M1 macrophage polarization. Finally, we observed that silencing HMGN2 affected the survival of NTM in macrophage, which might largely relevant to enhanced macrophage polarization into M1 phenotype under the NTM infection. Collectively, current studies thus suggested a novel function of HMGN2 in regulating the anti‐non‐tuberculous mycobacteria innate immunity of macrophage.

non-tuberculous mycobacteria replication. 9 In general, to fight against non-tuberculous mycobacteria infection, macrophage can initiate innate immunity though recognition of pathogen-associated molecular patterns (PAMP) by pattern-recognition receptors (PRRs), 10 such as activation of phagocytic pathway, production of nitric oxide (NO) and antibacterial peptide, [11][12][13][14] release of pro-inflammatory cytokines and chemokines, 12,15,16 which was conceptually introduced as classical activation of macrophage, also termed as M1 polarization. 17 Following the acute inflammatory burst, however, macrophages can remove cell debris and apoptotic inflammatory cells by so-called efferocytosis and polarize into alternatively activated, M2 macrophages. 18 It has been well documented that M1 macrophage polarization is instructed by a variety of environmental stimuli, for instance the acute infection of bacteria and virus, or the stimulation of LPS and IFNγ. The activated macrophages lead to the pro-inflammatory responses and further limit bacterial growth. Currently, it has been widely studied about the immune response against Mycobacterium tuberculosis. For instance, in the initial infection stage, Mycobacterium tuberculosis subcellular components activated M1 macrophage to produce pro-inflammatory cytokines such as tumour necrosis factor alpha (TNF-α), IL-1β, IL-6, IL-12, nitric oxide (NO), reactive oxygen species (ROS), and chemokines via TLR2 signalling. 19 In NTM infection model, it has been reported that NTM-induced different cytokine patterns depend on the strains of NTM which is related to the intracellular NTM growth rate. 20,21 IFNγ is the only subunit of type II interferons. It has been well noted that IFNγ is not only an essential inducer of the M1 macrophage polarization, but also plays the crucial role in anti-Mycobacterium tuberculosis defence. 22 It has been reported that IFNγ promotes NO production to further induce iNOS transcription, which increased the ability of macrophage clearance intracellular pathogens such as the Mycobacterium tuberculosis. 23 The expression of IFNγ could be induced by Mycobacterium tuberculosis in macrophage. In addition, it was reported that IFNγ been up-regulated by released cytokines. For example, IL-12 and TNF-α can induce IFNγ expression. [24][25][26] All in all, tight regulation of cellular cytokine pathways is critical to shape the macrophage polarization state and further impact on the final results of the host anti-bacterial immune responses. Therefore, it is necessary and important to explore the regulatory mechanism of inflammatory cytokine expression and the function of macrophages during mycobacteria infection.
High-mobility group (HMG) is a group of chromosomal proteins which are found in the mammalian nuclei. HMGN serves as a member of the HMG super family, which is involved in the regulation of gene transcription, replication, recombination and DNA repair. 27 HMGN protein family is composed of five subtypes: HMGN1, HMGN2, HMGN3, HMGN4 and HMGN5. Since 2004, we firstly identified HMGN2 functioned as an antimicrobial peptides, 28 and furthermore, we have demonstrated that HMGN2 is a multifunctional protein in immune regulation. For example, we found that HMGN2 promotes LPS-induced β-defensin expression in lung epithelial cells. 29 HMGN2 inhibits the internalization of Klebsiella pneumoniae in lung epithelial cell A549 through decreasing the expression and activity of α5β1 integrin and activated FAK-Src signalling to reduce F-actin polymerization. 30 HMGN2 was found to be used to through Nrf-2 signalling prevents Pseudomonas aeruginosa adhesion and invasion in lung epithelial cells by promoting pyocyanin-induced intracellular ROS clearance. 31 Otherwise, we found HMGN2 participated in MAPK signalling through activating ERK1/2 and P38, and regulating autophagy by AMPK pathway to reduce Uropathogenic Escherichia coli internalization of bladder epithelial cells. 32 However, the role of HMGN2 for regulating the anti-non-tuberculous mycobacteria innate immunity of macrophage is still largely unknown.
In the current study, we firstly investigated the macrophage polarization potential upon the two strains of non-tuberculous mycobacteria infection, and then we show the expression pattern of HMGN2 in macrophages under different infection conditions. Furthermore, we detected HMGN2 functional relevance and its effect on M1-related signalling pathways in M1 macrophage by using the RNA interfering technology. Lastly, we investigated the expression pattern of HMGN2 under IFNγ stimulation and the role of HMGN2 in IFNγ promoted M1 macrophage polarization.
FITC fluorescent-labelled secondary antibody (goat anti-rabbit IgG, green) and tetramethylrhodamine (TRITC)-conjugated secondary antibody (goat anti-rabbit IgG, red) were also purchased from Beyotime Institute of Biotechnology. DAPI and FITC were provided by Sigma-Aldrich. All-in-One cDNA Synthesis SuperMix and 2xSYBR Green RT-qPCR Master Mix were purchased from Biotool. Selective IKK inhibitor BMS-345541 was obtained from Selleck. RPMI 1640 medium was purchased from HyClone, Thermo Scientific. Foetal bovine serum (FBS) was obtained from FuMeng Gene Co., Ltd.. Penicillin-streptomycin was purchased from Beijing Solarbio Science and Technology Co., Ltd.. IFNγ was purchased from immune tool. Other chemical reagents were all analytical grade.

| Cell line and cell culture
Mus musculus macrophage cell lines RAW264.7 and MH-S were purchased from the Cell Bank of the Chinese Academic of Sciences. RAW264.7 cells were cultured in DMEM high glucose medium (HyClone) with 10% foetal bovine serum which be pre-incubated at 56°C for 30 minutes (FBS, FuMeng Gene Co.,Ltd.) and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin). Cells were incubated in humidified air with 5% CO 2 at 37°C. MH-S cells were cultured in RPMI-1640 medium supplemented with 10% foetal bovine serum and 50 µmol/L β-mercaptoethanol and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin incubated in humidified air with 5% CO 2 at 37°C.

| Real-time quantitative polymerase chain reaction (RT-QPCR)
Total RNA from RAW264.7 and MH-S cells was extracted by UNIQ-10 Column total RNA Purification Kit (Sangon Biotech) following the manufacturer's instruction. The purity and concentration of total RNA were measured by Implen NanoPhotometer. The cDNA synthesis was achieved using All-in-One cDNA Synthesis SuperMix (Biotool). The PCR was performed with CFX96 Real-Time PCR, and the PCR products were detected using 2xSYBR Green RT-qPCR Master Mix (Biotool). The PCR primers were synthesized by Qinke, and the primer sequences were as follows: HMGN2, TNF-α, IFNγ, IL-1β, IL-6, IL-10, TGF-β, iNOS, LL37, IFITM1, CXCL1 and CXCL2. The relative mRNA transcripts levels in different groups were evaluated using the 2 −ΔΔCt methods ( Table 1). Equal amounts of 20 µg protein lysate were added into the wells of the SDS-PAGE gels. The separated cells were then transferred onto a PVDF membrane and blocked with 5% blotting milk in PBS buffer for 1 hour, and incubated with the diluted primary antibody in 0.5% PBST buffer overnight at 4°C, then added the secondary antibody (1:1000) for 2 hours at room temperature. The intensity of the band signals were detected by enhanced chemiluminescence (Merck Millipore) and exposed with ChemiDoc™ MP Imager (Bio-Rad).

| NO assay
The production of NO in macrophages was determined by the Total Nitric Oxide Assay Kit (Beyotime) according to the manufacturer's instructions. Briefly, total NO production was measured by Griess reagent to react with NO metabolite nitrite in the medium to form a coloured product and then quantified by a microplate reader at 540 nm.

| Intracellular bacterial assay
According to our previous method, 30

| Statistical analysis
Data were analysed by two-tailed Student's t test to compare the differences in values between experienced group and control group.
P-value < .05 was considered to have a statistical difference. The cytotoxicity of macrophages induced by non-tuberculous mycobacteria strains was tested by CCK-8( Figure S1). Our results showed that NO was induced at 12 hours in RAW264.7 ( Figure 1A) and 24 hours in MH-S cells ( Figure 1B). Then, we measured the ex-

| HMGN2 was induced in non-tuberculous mycobacteria-infected macrophage
Our previous studies showed that HMGN2 was up-regulated in F I G U R E 1 Non-tuberculous mycobacteria infection induces M1 macrophage polarization. RAW264.7 and MH-S cells were exposed to NTM Ma. 24 and Ms 13 separately (MOI = 10:1) at indicated time courses 3, 6, 12 and 24 h. Results shown are representative of at least three independent experiments. A and B, The production of NO was examined by nitrate reduction method. P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between uninfected and infected groups. C and D, The expression levels of iNOS transcription were determined by RT-qPCR P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between uninfected and infected groups. RAW264.7 was exposed to Ma. 24 (E, G and I) and Ms 13 (F, H and J) separately (MOI = 10:1) at indicated time-points 3, 6 and 12 h. Results shown are representative of at least three independent experiments. Pro-inflammatory cytokines TNF-α, IFNγ (E and F), IL-1β and IL-6(G and H), and anti-inflammatory cytokines TGF-β, IL-10 (I and J) mRNA transcription expression were examined by RT-qPCR P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between uninfected and infected groups was higher from 3 to 12 hours (Figure 2A and B). And the protein level of HMGN2 was significantly higher than the uninfected group ( Figure 2C and D). In addition, we examined the cellular distribution of HMGN2 in macrophage. Immunofluorescence assay showed that HMGN2 localized in both cytoplasm and nucleus, and NTM infection does not alter HMGN2 distribution ( Figure S2A).

| HMGN2 knock-down enhanced M1 macrophage polarization with characteristic increase of NO production and the up-regulation of iNOS
iNOS, an inducible enzyme to generate the nitric oxide for killing pathogens, is well described to characterize the M1 polarization. To F I G U R E 2 HMGN2 is up-regulated in NTM-polarized M1 macrophage. RAW264.7 and MH-S cells were exposed to Ma. 24 and Ms 13 separately (MOI = 10:1) at indicated time-points 3, 6 and 12 h. Results shown are representative of three independent experiments. A and B, RT-qPCR examined HMGN2 transcription in infected macrophages. P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between uninfected and infected groups. C and D, Western blot analysis showed the effect of non-tuberculosis mycobacteria infection on HMGN2 expression. Quantification was measured by Image J, and P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between uninfected and infected groups investigate whether HMGN2 can affect NO production in NTM-induced M1 polarization, HMGN2-specific siRNA was transfected into macrophage RAW264.7 and MH-S. As shown in Figure S2B, siRNAmediated HMGN2 knock-down achieved 90% down-regulation of HMGN2 expression in both cell lines compared to negative control.
With using the NO production assay, we measured that the enhanced production of NO in HMGN2 silenced RAW264.7 was at 12 hours after infection, while HMGN2 silenced MH-S displayed the increasing of NO production at 6 hours ( Figure 3A). Furthermore, consistent to the NO production results, we observed that siRNA-mediated knock-down of HMGN2 unregulated the NO synthase iNOS mRNA and protein level in both cell lines from 3 to 12 hours ( Figure 3B and C).

| HMGN2 knock-down enhanced M1-related pro-inflammatory cytokine expression
In addition to the analysis of iNOS gene expression, we also inves-

| HMGN2 knock-down enhanced the activation of NF-κB signalling pathway
NF-κB is regarded as a major immune-regulating transcription factor to regulate the expression of iNOS gene and other pro-inflammatory mediators during M1 macrophage polarization. [33][34][35] At once activation of NF-κB signalling, IκBα was phosphorylated F I G U R E 3 HMGN2 knock-down enhances NTM-induced M1 macrophage polarization. RAW264.7 cells were transfected with siRNA-HMGN2 and NC siRNA (scramble control) for 24 h and then incubated with Ma. 24 (MOI = 10:1) for 3, 6 and 12 h. A, The production of NO was examined by nitrate reduction method. P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between NC-infected and siHMGN2-infected groups. Results shown are representative of at least three independent experiments. B, The expression level of iNOS was determined by RT-qPCR P value was determined by one sample t test. *Indicates significant difference (**P < .01) between NCinfected and siHMGN2-infected groups at the same time-point. C, Western blot analysis showed the protein level of iNOS expression. P value was determined by one sample t test. *Indicates significant difference (*P < .05; **P < .01) between NC-infected and siHMGN2-infected groups at the same time-point and ubiquitination, further be subsequent proteasomal degradation, which leading to P65 phosphorylation and transfer to nuclear.
We examined whether loss the function of HMGN2 could impact on Ma.24-induced IκBα expression. As shown in Figure 5A, knockdown of HMGN2 in macrophage strongly induced IκBα phosphorylation and then we observed a degradation of IκBα. Next, we observed that the phosphorylation of P65, a downstream signal cascade of NF-κB pathway, was significantly enhanced in HMGN2 knock-down cells ( Figure 5A). Immunostaining study showed that HMGN2 knock-down significantly promoted the p65 localization into the nucleus ( Figure 5B). To further confirm HMGN2 regulates NO through NF-κB signalling, we used IKK inhibitor BMS-345541 in the silenced HMGN2 macrophage to inhibit the NF-κB signalling and then measured the production of NO. As shown in Figure 5C, the HMGN2 knock-down induced the production of NO was inhibited by BMS-345541 in NTM-infected macrophage.
Our previous studies showed that HMGN2 can inhibit ERK1/2 and JNK phosphorylation. As MAPK signalling plays an important role in modulation of the expression of pro-inflammatory mediators expression in macrophages. Therefore, we also included analysis of JNK, P38 and ERK1/2 signalling pathways in the current study. As shown in Figure S2C, phosphorylation of JNK, P38 and ERK1/2 was activated by NTM infection from 10 to 15 minutes, which was further enhanced by HMGN2 knock-down in macrophage.

| HMGN2 was involved in IFNγ-induced M1 macrophage polarization
As shown in above results, IFNγ, a well-noted immune regulator to involve both in the induction of M1 macrophage polarization and anti-NTM immune reaction in macrophage, was shown to be regulated by HMGN2. It is supposed that HMGN2 knock-down enhanced M1 polarization were possibly and partially mediated by IFNγ. In order to explore this possibility, we investigated the role of HMGN2 in regulation of IFNγ-induced M1 macrophage polarization. As shown in Figures 6A and B and S4A, the mRNA expression of HMGN2 was induced by IFNγ from 6 to 24 hours, and HMGN2 protein was induced by IFNγ from 6 to 48 hours. Then, we observed that IFNγ-induced TNF-α, IL-1β and IL-6 transcription were further promoted by knockdown of HMGN2 compared with negative control group ( Figure 6C).
Although there is no significant change of iNOS production mediated by HMGN2 knock-down, an obvious increase tendency was showed. In addition, we analysed antiviral gene expression IFITM1, F I G U R E 4 HMGN2 knock-down up-regulates pro-inflammatory cytokines in NTM-induced M1 macrophage polarization. RAW264.7 cells were transfected with siRNA-HMGN2 and NC siRNA (scramble control) for 24 h and then incubated with Ma.24 (MOI = 10:1) for 3, 6 and 12 h. A, The production of IFNγ; IL-1β; TNF-α; and IL-6 was examined by RT-qPCR P value was determined by t test. *Indicates significant difference (*P < .05; **P < .01) between NC-infected and siHMGN2-infected groups. B,The expression levels of IL-10 and TGF-β were determined by RT-qPCR P value was determined by t test. *Indicates significant difference (*P < .05; **P < .01) between NC-infected and siHMGN2-infected groups. Results shown are representative of three independent experiments antibacterial gene LL37 as well as chemokines CXCL1 and CXCL2 ( Figure 6D), and our results exhibited a significant induction of IFITM1 and LL37. Chemokines CXCL1 and CXCL2 were induced by IFNγ treatment, which were also further up-regulated HMGN2-deficient macrophages. Collectively, these results indicated an essential role of HMGN2 in regulating IFNγ-induced macrophage polarization.  24 had no significant change in both macrophages ( Figure 7A). To further confirm these survival results, the FITC-labelled Ma.24 was applied to infect HMGN2 deficient macrophage. As shown in Figure 7D, we observed the higher fluorescent signal from the macrophage with HMGN2 knock-down by using immunofluorescence microscope, which also indicated that HMGN2 knock-down decreased the survival of NTM in macrophage. Collectively, our results proofed that HMGN2-regulated M1 macrophage polarization can be functional relevance to its role in manipulation of NTM survival.

| D ISCUSS I ON
In the present study, we unravel a novel function of HMGN2 in regulation of classic activation of alveolar macrophages induced by NTM.
We showed for the first time that HMGN2 was induced in NTM- In order to investigate the functional relevance of the NTM-induced HMGN2 up-regulation, we firstly analysed the production IFNγ is the key cytokine involved in the protective role to against Mycobacterium.tuberculosis and NTM infection. The involved mechanism includes the induction of autophagy, a number of antimicrobial molecules, and pro-inflammatory cytokines and chemokines. As discussed above, IFNγ was also well known for its role in inducing M1 macrophage polarization. As shown in inhibit Pseudomonas aeruginosa internalization in lung epithelium cells. 31 Our data revealed that HMGN2 also involved in uropathogenic Escherichia coli (UPEC) infection in bladder epithelial cells via regulating autophagy. Therefore, HMGN2 is one of the important regulators of anti-bacterial innate immunity in both epithelial and macrophages.
In conclusion, our study reported that both NTM-and IFNγ-induced M1 alveolar macrophages expressed a high level of HMGN2.
Further functional analysis revealed that HMGN2 can regulate both NTM-and IFNγ-induced M1 polarization, and further impacted on the NTM survival in macrophages. The important M1 polarization signalling pathways NF-κB were found to be relevant to HMGN2 regulated M1 polarization. Collectively, present study suggests an novel role of HMGN2 in anti-NTM innate immunity, and HMGN2specific inhibition or interfering might thus represent a favourable approach for the treatment of NTM-related infection.

ACK N OWLED G EM ENTS
This work was supported by grants from Sichuan University 985 Project-Science and Technology Innovation Platform for Novel Drug Development.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interests.

AUTH O R CO NTR I B UTI O N S
XW, HR, JC, JL, YW and YH conceived the study, acquired the data, and interpreted and analysed the data. XW and SC wrote the manuscript and revised it critically for the important intellectual content.
All authors approved the final version to be published.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated during the study are included in this article.