Nucleotide-Binding Oligomerization Domain 2 Contributes to Limiting Growth of Mycobacterium abscessus in the Lung of Mice by Regulating Cytokines and Nitric Oxide Production

Mycobacterium abscessus is a prominent cause of pulmonary infection in immunosuppressed patients and those with cystic fibrosis. Nucleotide-binding oligomerization domain (NOD) 2 is a cytosolic receptor which senses a bacterial peptidoglycan component, muramyl dipeptide (MDP). Although nucleotide-binding oligomerization domain 2 (NOD2) contributes to protect host against various microbial infections, it is still unclear whether NOD2 is essential to regulate host immune responses against M. abscessus infection. In this study, we sought to clarify the role of NOD2 and the underlying mechanism in host defense against M. abscessus infection. Mice were infected intranasally with M. abscessus and sacrificed at indicated time points. Bacterial survival, cytokines production, and pathology in the lungs were determined. Bone marrow-derived macrophages were used to clarify cellular mechanism of NOD2-mediated immune response. Bacterial clearance was impaired, and pathology was more severe in the lungs of NOD2-deficient mice compared with the wild-type mice. In macrophages, NOD2-mediated activation of p38 and JNK were required for production of proinflammatory cytokines and nitric oxide (NO) and expression of iNOS in response to M. abscessus. NO was critical for limiting intracellular growth of the pathogen. Intranasal administration of MDP reduced in vivo bacterial replication and thus improved lung pathology in M. abscessus-infected mice. This study offers important new insights into the potential roles of the NOD2 in initiating and potentiating innate immune response against M. abscessus pulmonary infection.

to most classical anti-tuberculous drugs and antibiotics, which thereby limits therapeutic options and often leads to a poor prognosis (6). Nucleotide-binding oligomerization domain 2 (NOD2) is a member of cytosolic nucleotide-binding oligomerization domain (NOD)-like receptors family (7). Upon detecting muramyl dipeptide (MDP), a bacterial peptidoglycan component, NOD2 recruits receptor-interacting protein serine/threonine protein kinase 2 (RIP2/RICK/RIPK2) through the CARD-CARD interaction (8). Subsequently, RIP2 mediates recruitment and activation of TGFβ-activated kinase 1, which is required for activation of the IKK complex and mitogen-activated protein kinases (MAPKs) (9). Finally, this activation cascade leads to the expression of proinflammatory cytokines (9).
Previous studies have shown that NOD2 pathway contributes to the host innate immune response against pulmonary bacterial infections and that its role depends upon the species of bacterial pathogens (10)(11)(12)(13). NOD2 is necessary for optimal phagocytosis of Streptococcus pneumoniae by alveolar macrophages and neutrophils in vitro (13). In addition, NOD2 cooperates with TLR2 to reduce the bacterial replication in the upper respiratory tract, although deficiency of NOD2 alone does not influence bacterial replication, dissemination, or lung pathology in mice infected with S. pneumoniae (12,13). Moreover, NOD1/NOD2-RIP2 signaling pathways improved host immune responses against Chlamydophila pneumoniae infection, resulting in enhanced bacterial clearance in the lungs and longer survival time in the infected mice (10). In contrast, in a pulmonary infection model of Legionella pneumophila, NOD2 was dispensable for bacterial clearance in the lungs, although it was involved in the recognition of heat killed L. pneumophila (11).
The role of NOD2 in mycobacterial infections has been primarily studied in the Mycobacterium tuberculosis infection model, wherein NOD2 activation by MDP limited intracellular growth of M. tuberculosis within alveolar macrophages (14). NOD2 deficiency resulted in both increased bacterial loads in lung tissue and enhanced mortality during the chronic phase of infection, although the lung pathology during the early stage of infection was reduced (15). A recent study revealed that TLR2 is critical for initiating the protective Th1-type immune response against M. abscessus, but the exact role of NOD2 in initiating and propagating the host innate immune response against M. abscessus still remains unknown (16). To address these limitations in the present study, both the in vitro and in vivo roles of NOD2 were assessed in the context of host innate immunity against M. abscessus infection.

Mice
Wild-type (WT), NOD2, and iNOS-deficient mice on a C57BL/6 background were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). The animal studies were conducted under protocols approved by the Institutional Animal Care and the Use Committee of Chonnam National University (Approval No. CNU IACUC-YB-2015-32).

Bacterial culture and infection experiments
Culture and single-cell suspensions of the isogenic rough variant of M. abscessus ATCC19977 T (Manassas, VA, USA) were performed as previously described (16). Briefly, the bacterium was cultivated in 7H9 broth supplemented with 0.5% glycerol, 10% oleic acid, albumin, dextrose, and catalase (OADC; BD Biosciences, San Jose, CA, USA) for 7-10 days. The seed lots were kept in small aliquots at −80°C until use. Mice were intranasally (i.n.) inoculated with 1.5 × 10 7 CFU of bacteria. At 5, 10, and 20 days post infection (dpi), mice were sacrificed and the lung tissues were collected aseptically. For the in vivo MDP treatment, mice were i.n. dosed with N-glycolyl-MDP (150 μg/animal; Invivogen, San Diego, CA, USA), which exists in Mycobacteria (17).

lung Tissue Preparation and Bacterial count
The right lobes of the lungs were weighed and homogenized with phosphate buffered saline (PBS). Half of the lysate was diluted with sterile PBS, plated onto 7H10 agar plates, and incubated at 37°C for 4 days. The colonies were counted and the number of bacteria (measured as CFU/g lung tissue) was calculated. The other half of the lysate was centrifuged and the supernatant was stored in a deep freezer for cytokine measurement.

histopathological examination
The left lobes of the lungs were fixed in 10% neutral formalin for 24 h, followed by tissue processing and paraffin embedding. The paraffin blocks were sectioned at 2 µm, stained with hematoxylin and eosin, and examined under microscopy. The lung pathology was evaluated blindly by an expert in a field of laboratory animal pathology. with an arbitrary scoring index ranging 0-5, which was determined by the degree of inflammatory cell infiltration and the extent of the lesion area (0, normal; 1, mild; 2, mild to moderate; 3, moderate; 4, moderate to severe; 5, severe).

nitrite assay
Bone marrow-derived macrophages were infected with M. abscessus at a MOI 1:10 along with mouse IFN-γ (BioLegend, San Diego, CA, USA) in the absence or presence of 1 mM of N G -nitro-l-arginine methyl ester (l-NAME; Sigma Aldrich, St. Louis, MO, USA) for indicated times. Nitrite formation in culture supernatants was determined via the Griess reaction as described previously (19).

cytokine Measurement
The concentrations of IL-6, TNF-α, IL-1β, and IFN-γ in supernatants from BMDM cultures and lung tissue lysates were determined using a commercial ELISA kits (R&D Systems, Minneapolis, MN, USA).

Western Blotting
Cellular lysates (25 μg/well) were separated using 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes. Membranes were then incubated with primary antibodies to regular or phosphorylated IκB-α, p38, ERK, JNK (Cell signaling Technology, Beverly, MA, USA), and iNOS (BD Biosciences, San Jose, CA, USA). A primary antibody to β-actin (Santa Cruz Biotechnology, CA, USA) was used to verify equal loading of the protein samples. Following incubation with the relevant secondary antibodies, proteins were detected using an enhanced chemiluminescence reagent (Intron Biotechnology, Seongnam, Korea).

statistical analysis
The statistical significance of differences between groups was determined by a two-tailed Student's t-test or one-way ANOVA followed by the Bonferroni post hoc test for multigroup comparisons (GraphPad Prism 5; GraphPad Software Inc., La Jolla, CA, USA). Values of P that were less than 0.05 were considered statistically significant.

The absence of nOD2 Fails Bacterial clearance and exacerbates Pathology in the lungs of M. abscessus-infected Mice
To determine whether NOD2 controls in vivo growth of M. abscessus, WT and NOD2-deficient mice were i.n. infected and the bacterial loads in the lungs were determined at 5, 10, and 20 dpi. No mortality of mice was observed in any group during the experiment. The bacterial loads in the lungs were significantly higher in NOD2-deficient mice as compared with those of the WT mice at all time points measured ( Figure 1A). Upon histopathological examination, infiltration of inflammatory cells, mostly composed of monocytes and lymphocytes, were observed in the peribronchioles, and thickened alveolar septa were also seen in the lungs of M. abscessus-infected mice ( Figure 1B). NOD2-deficient mice exhibited more severe lung pathology at 10 and 20 dpi (Figures 1B,C). Cellular composition in bronchoalveolar lavage fluids was not different between WT and NOD2-deficient mice groups at least until day 10 (data not shown). These results indicate that NOD2 may play an important role in host defense against M. abscessus pulmonary infection.

nOD2 is essential for cytokines Production in the lungs of Mice infected with M. abscessus during the early Phase of infection
We further investigated whether NOD2 deficiency influences cytokines production in the lungs of M. abscessus-infected mice. Levels of IL-6, TNF-α, and IL-1β were significantly lower in the lungs of NOD2-deficient mice as compared with those of WT mice at 5 dpi (Figures 2A-C). At 10 dpi, NOD2-deficient mice still produced less IL-6 compared to the WT mice, whereas TNF-α levels were comparable and IL-1β was increased in the lung homogenates of NOD2-deficient mice (Figures 2A-C). The levels of IL-6 and TNF-α were significantly higher in the lung homogenates of NOD2-deficient mice than those of the WT mice at 20 dpi (Figures 2A,B). When we measured levels of IFN-γ, the principal Th1 effector cytokine, NOD2-deficient mice exhibited less IFN-γ production at 10 and 20 dpi relative to that of the WT mice ( Figure 2D). These results indicate that NOD2 is essential for the production of cytokines during the early stages of M. abscessus infection and may be required for optimal induction of the Th1-type immune response.
Production of cytokines and nO and expression of inOs are impaired in nOD2-Deficient Macrophages during M. abscessus infection Next, we sought to determine the role of NOD2 in a cell-specific immune response using macrophages, one of the central cells of the innate immune system. Production of IL-6, TNF-α, and IL-1β increased in M. abscessus-infected BMDMs in a dose-dependent manner but which levels were significantly lower in NOD2deficient cells (Figures 3A-C). We further compared iNOS  Figure 1a and that of uninfected group is five. The results are from one representative experiment of two independent experiments (*P < 0.05, **P < 0.01, ***P < 0.001).  detectable levels of NO production in macrophages ( Figure 3F).
In the presence of IFN-γ, M. abscessus-induced substantial levels of NO production in macrophages, which was significantly less in NOD2-deficient cells (Figure 3F). To confirm the phenotype of NOD-deficient BMDMs, cytokines production in response to LPS was examined in the absence or presence of MDP. As expected, both WT and NOD2-deficient BMDMs produced comparable level of IL-6 and TNF-α in response to LPS (Figures 3G,H). LPS synergies with MDP to produce both cytokines in WT BMDMs, but not in NOD2-deficient cells (Figures 3G,H).

P38 and JnK MaPKs are involved in nOD2-Mediated Production of cytokines and nO in M. abscessus-infected Macrophages
We further investigated whether NOD2 is required for activations of NF-κB and MAPKs in BMDMs, the two key molecular factors regulating cytokine production in macrophages (20). Similar levels of IκB-α degradation and phosphorylation were detected at 15 and 30 min after M. abscessus infection between WT and NOD2-deficient cells (Figure 4A). At 15 min after the infection, phosphorylation of p38 and JNK was highly induced only in the WT BMDMs compared with those of NOD2-deficient cells ( Figure 4A). Comparable levels of ERK phosphorylation were also detected in the WT and NOD2-deficient BMDMs at the 15 and 30 min time points, although the signal was undetectable in NOD2-deficient cells at 60 min ( Figure 4A). An inhibitor assay revealed that both the p38 and JNK signaling pathways are involved in NO production in M. abscessus-infected BMDMs ( Figure 4B). Moreover, M. abscessus-induced production of IL-6, TNF-α, and IL-1β was reduced in a dose-dependent manner by both inhibitors (Figures 4C-E). These findings suggest that M. abscessus activates p38 and JNK MAPKs via NOD2-mediated signaling pathway, which contributes to production of cytokines and NO in macrophages.  00 0.22 0.76 0.76 1.30 0.20 0.69 0.85   0.03 0.90 0.22 0.17 0.03 1.06 0.19 0.02   0.02 0.89 0.74 0.03 0.03 0.39 0.06 0.02   0.16 1.30 1.65 1.31 0.36 0.60 0.37 0.14   0.07 0.07 1.14 0.54 0.06 0.07 0.91 0.63

A B
C D E

nOD2-Mediated nO Production is essential for restricting intracellular growth of M. abscessus within Macrophages
Although there is no direct evidence showing the role of NO in controlling the growth of M. abscessus, it is well known that NO plays an important role in the host defense against microbial infections (21). As M. abscessus-induced NO production was impaired in NOD2-deficient macrophages (Figure 3F), we sought to determine whether NO contributes to restriction of the bacterial growth within macrophages. In the absence of IFN-γ treatment which is essential for NO production by M. abscessus, as shown in Figure 3F, there was no significant difference in intracellular growth of M. abscessus between WT and NODdeficient BMDMs (Figure 5A). However, NOD2-deficient cells exhibited significantly higher CFU densities than did WT cells in the presence of IFN-γ at 24, 48, and 72 h after M. abscessus infection ( Figure 5B). Moreover, treatment with l-NAME, a direct inhibitor of NOS, abolished M. abscessus-induced NO production ( Figure 5C) and the difference in the intracellular bacterial CFUs between the WT and NOD2-deficient BMDMs ( Figure 5D). Consistently, the ability to restrict the bacterial growth was impaired in iNOS-deficient BMDMs ( Figure 5E). Taken together, NOD2 seems to be involved in the inhibition of intracellular survival and replication of M. abscessus by mediating iNOS expression and NO production in macrophages.

M. abscessus
To determine whether NOD2 activation by MDP has a beneficial effect on the host defense against M. abscessus infection, MDP was given i.n. before and after M. abscessus infection as shown in Figure 6A. The bacterial loads in the lungs were significantly reduced in mice with MDP treatment at both 5 and 10 dpi versus the PBS-treated mice (Figure 6B). Paralleling the microbiological outcomes, the lung pathology was also improved in MDP-treated mice (Figures 6C,D). The MDP-treated mice also exhibited significantly lower levels of IL-6, TNF-α, and IL-1β in the lung homogenates than those of PBS-treated mice at day 10, although only marginal differences in the cytokine levels were observed between the two groups at day 5 (Figures 6E-G). These results

DiscUssiOn
In the present study, we revealed that NOD2 is essential for the optimal clearance of M. abscessus and reduced pathology in murine lungs. The bacterial load in the lungs was higher in NOD2-deficient mice even at 5 dpi, suggesting that NOD2 may contribute to inducing the innate immune response against M. abscessus during the early phase of infection. In line with these data, IL-6, TNF-α, and IL-1β levels in the lungs were significantly lower in NOD2-deficient mice at 5 dpi. Moreover, the in vitro study showed that the production of IL-6, TNF-α, and IL-1β in response to M. abscessus was impaired in NOD2deficient macrophages. Such cytokines have been shown to be important in the host defense against M. tuberculosis infection (22)(23)(24)(25). Besides pulmonary infections, TNF-α was also essential for bacterial clearance in the spleen and liver of mice infected by intravenous injection of M. abscessus, reducing the lethality of the infection (26). Taken together, NOD2 seems to inhibit M. abscessus infection by initiating cytokine production during the early phase of infection. However, during the late phase of infection, it is likely that NOD2 controls the growth of M. abscessus through different mechanisms, because despite levels of IL-6, TNF-α, and IL-1β being significantly higher in NOD2-deficient mice at 10 or 20 dpi, the bacterial burden was still higher in those mice. It is, therefore, assumed that increased bacterial densities induced high-level production of cytokines through a NOD2independent pathway, which resulted in severe lung pathology in those M. abscessus-infected mice. In addition, levels of IFN-γ were lower in the lungs of NOD2-deficient mice at 10 and 20 dpi. In previous publications, it has been reported that bacterial clearance in organs was impaired in IFN-γ-or RAG2-deficient mice infected with M. abscessus (26,27). Moreover, activation of the antigen-specific T cell response was reduced in NOD2-deficient mice during mycobacterial infections (15). Although a more detailed role of NOD2 in Th1 immunity should be investigated in the future, our data indicate that NOD2 may be involved in the control of M. abscessus growth during the late phase of infection by regulating T cell response and IFN-γ-mediated signaling.
Macrophages are the representative innate immune cells and play an important role in host immunity against mycobacterial infections by producing cytokines and NO. In the present study, we demonstrated that NOD2 is essential for M. abscessus-induced cytokine production in murine macrophages. Although N-acetyl MDP exists in the peptidoglycan layer of most bacteria, mycobacteria and related actinomycetes possess N-acetyl muramic acid hydroxylase, which converts their MDP to an N-glycolylated form. N-glycolyl MDP is more potent than N-acetyl MDP in producing proinflammatory cytokines and activating NF-κB and JNK (17). Accordingly, it is assumed that NOD2 plays a much more important role in the host immunity against mycobacterial   (20,(28)(29)(30). In contrast, production of cytokines in response to various Mycobacterium spp. was impaired in NOD2-deficient macrophages (31)(32)(33)(34).
Mycobacterium abscessus could also activate NF-κB and all MAPKs (p38, ERK, and JNK) in murine BMDMs, but only p38 and JNK activation were influenced by NOD2 deficiency (Figure 4A). The inhibitor assay confirmed that both p38 and JNK signaling is necessary for optimal cytokine production in macrophages in response to M. abscessus. However, significance of p38 on M. abscessus-induced production of cytokines remains controversial. Shin et al. reported that M. abscessus induces p38 and ERK phosphorylation in murine macrophages (35). The inhibitor assay revealed that ERK, but not p38, is involved in TNFα, IL-6, and IL-12p40 production by macrophages in response to M. abscessus infection (35). In contrast, Sampaio et al. reported that induction of TNF-α production in response to M. abscessus is abolished by the p38 inhibitor SB203580 in human monocytes (36). It is well known that pyridinyl imidazole, inhibitors of p38 MAPKs such as SB203580, inhibits autophosphorylation of RIP2 (37,38), which is an adaptor protein for NOD1 and NOD2 (39). It is likely that SB203580 inhibits cytokines and NO production in M. abscessus-infected BMDMs by suppressing RIP2-mediated signaling. It should be further elucidated whether p38 directly regulates M. abscessus-induced production of cytokines and NO in macrophages using other experimental systems such as gene silencing or p38 knockout cells.
Nitric oxide plays an important role in the innate immune response and in the host defense against mycobacterial infection NOD2 in M. abscessus Infection Frontiers in Immunology | www.frontiersin.org November 2017 | Volume 8 | Article 1477 in mice, although there is still controversy about the role of NO in the control of M. tuberculosis in humans (21). iNOSdeficient mice displayed decreased survival times and impaired bacterial clearance in response to M. tuberculosis infection (40). NOD2 is required for iNOS expression and NO production in M. tuberculosis-infected macrophages (31,41). Consistently, in the present study, transcriptional and translational expression of iNOS induced by M. abscessus was decreased in NOD2deficient macrophages. Moreover, in the presence of IFN-γ, NOD2 deficiency led to impairment of NO production. NOD2 was also required for control of the bacterial growth in IFN-γtreated cells, and l-NAME increased intracellular survival of the bacteria, suggesting that NOD2-mediated NO production may be crucial for restriction of M. abscessus growth within murine macrophages. In line with the NO production data, the bacterial growth control was impaired in iNOS-deficient macrophages. These findings were unique compared to previous reports studying M. tuberculosis infection. Although M. tuberculosis-induced NO production was impaired in NOD2-deficient macrophages, the bacterial replication was not different between WT and NOD2-deficient cells (31). It should be clarified whether or not NO-mediated restriction of bacterial replication in macrophages is specific to M. abscessus among the various Mycobacterium spp. In addition, in human macrophages, the role of NO in controlling the intracellular growth of M. abscessus may differ in that NOS activity is lower than in murine macrophages (42). Nevertheless, a recent study showed that compassionate inhaled NO treatment is very effective in bacterial clearance in cystic fibrosis patients with persistent M. abscessus infection (43). This implies that NO-based therapies can be effectively applied in patients with M. abscessus infection and support our results. It has been known that MDP activation of NOD2 has protective effect against several diseases, such as colitis, atherosclerosis, and influenza infection, in mice (44)(45)(46). As NOD2 deficiency impairs bacterial clearance and pathology in the lungs of M. abscessus-infected mice, we sought to determine the effect of MDP-induced activation of NOD2 on host defense against the bacteria. Our results revealed that intranasal administration of MDP improves the bacterial clearance and pathology in the lungs of mice. As well as NOD2 activation, MDP can enhance TLRs-mediated immune responses in murine macrophages (20,47). In addition, TLR2 contributes to early protective Th1 immune response against M. abscessus (16). Accordingly, NOD2independent signaling such as TLR2 should not be overlooked in the protective effect of MDP, although MDP is well known as a NOD2-specific ligand (39). To clarify this, experiments using NOD2-or TLR2-deficient mice should be performed.
In summary, the present results suggest that (i) NOD2 defi ciency leads to impaired bacterial clearance and thus more severe lung pathology in M. abscessus-infected mice, (ii) NOD2 contributes to the production of cytokines by regulating the activation of p38 and JNK in the infected macrophages, (iii) NOD2 is also involved in iNOS expression and NO production in M. abscessus-infected macrophages, which restricts intracellular replication of the bacteria, and (iv) in vivo administration of MDP enhances the host's ability to control bacterial growth and reduces the severity of lung pathology in mice. In terms of that M. abscessus infection is dominant in patients with cystic fibrosis or immunosuppression, it seems to be important to know whether and how these underlying diseases modulate NOD2-mediated immune response in host. And if the results of this study are proved similar in humans, NOD2-mediated signaling can be a therapeutic target for M. abscessus infection.