MicroRNA-325-3p Facilitates Immune Escape of Mycobacterium tuberculosis through Targeting LNX1 via NEK6 Accumulation to Promote Anti-Apoptotic STAT3 Signaling

Intracellular survival of Mycobacterium tuberculosis results in bacterial proliferation and the spread of infection in lungs, consequently deteriorating the conditions of tuberculosis (TB) patients. This research discovers a new immune escape pathway of M. tuberculosis by modulating host miR-325-3p expression, thus leading to the intracellular survival of M. tuberculosis. These findings make a contribution to the understanding of the immune escape of M. tuberculosis, and they provide a theoretical basis for the development of therapeutic approaches for drug-resistant TB.

length) that transcriptionally/posttranscriptionally regulate gene expression (6). miRNA-mRNA networks are involved in many immune signaling pathways (7), especially in the immune escape of M. tuberculosis: microRNA let-7 targets A20 to help M. tuberculosis suppress innate immune responses in macrophages (8); microRNA-27a targets endoplasmic reticulum (ER)-located Ca 2ϩ transporter CACNA2D3 to inhibit autophagosome formation and to promote the intracellular survival of M. tuberculosis (9). In addition, the results from microRNA sequencing (miRNA-seq) showed that a large amount of miRNAs were up-or downregulated during M. tuberculosis infection (10). These findings suggested that miRNAs are likely to influence the outcome of the encounter between macrophages and M. tuberculosis.
Ubiquitination is a critical posttranslational modification in several cellular responses (11). The ubiquitin system is composed of activating enzyme (E1), conjugating enzyme (E2), and ligase (E3). The high efficiency and exquisite selectivity of ubiquitination reactions are attributable to the function of E3. To be more specific, E3 recognizes substrates based on the presence of a specific ubiquitination signal and catalyzes the formation of an isopeptide bond between a substrate lysine residue and the C terminus of ubiquitin. Ubiquitin-tagged substrates are recognized by proteasomes, which can induce protein degradation (12). The ubiquitin-mediated degradation of pivotal proteins is essential to the precise control of the immune system (13). For example, as a potent inhibitor of interferon-dependent antiviral responses, E3 ubiquitin ligase RNF2 directly binds to STAT1 after interferon stimulation and increases K33-linked polyubiquitination of STAT1 at position K379. In this way, RNF2 promotes the disassociation of STAT1/STAT2 from DNA and consequently suppresses transcription of interferonstimulated genes (ISG) (14). The immune escape pathway of M. tuberculosis associated with ubiquitination has also been reported. For example, PtpA, a secreted tyrosine phosphatase essential for tuberculosis pathogenicity, can suppress innate immunity by competitively binding to ubiquitin-interacting domain of the host adaptor TAB3 (15). These findings revealed a close link between ubiquitination and the host immune system.
In previous studies, we found that microRNA-325-3p (miR-325-3p) was upregulated in a susceptible mouse substrain after aerosol infection with M. tuberculosis. Also, relatively higher levels of miR-325-3p were found in latent TB patients than in healthy controls (data not published), indicating an important role for miR-325-3p in M. tuberculosis intracellular survival. In this study, we demonstrated for the first time that Mir325-deficient mice were resistant to M. tuberculosis infection. More importantly, we found that the ligand of numb-protein X 1 (LNX1), a direct target of miR-325-3p, is an E3 ubiquitin ligase of NIMA-related expressed kinase 6 (NEK6). During M. tuberculosis infection, LNX1 reduction promotes the abnormal accumulation of NEK6, resulting in an inhibition of apoptosis and promotion of the intracellular survival of M. tuberculosis. All these results not only made an important contribution to the understanding of the immune escape pathways of M. tuberculosis but also provided a theoretical basis for the development of therapeutic approaches for drug-resistant TB.

RESULTS
miR-325-3p is associated with M. tuberculosis survival in macrophages. To determine the role of miR-325 during M. tuberculosis infection, expression levels of primary (pri-), precursor (pre-), and mature miR-325 transcripts were analyzed in RAW 264.7 cells infected with M. tuberculosis. M. tuberculosis infection increased the expression levels of pri-miR-325, pre-miR-325, and miR-325-3p instead of miR-325-5p (Fig. 1A). The upregulation of miR-325-3p was also observed in mouse bone marrow-derived macrophages (BMDMs) at 24 h postinfection (Fig. 1B). Furthermore, the results from Northern blotting showed elevated miR-325-3p in macrophages after M. tuberculosis infection (Fig. 1C). Since gamma-irradiated M. tuberculosis had an ability to upregulate miR-325-3p similar to that of live M. tuberculosis (Fig. 1D), we used gamma-irradiated M. tuberculosis instead of live M. tuberculosis in some of the following studies. Mycobacterium bovis BCG, which lacks the RD1 genomic region, was unable to upregulate miR-325-3p (Fig. 1D). RD1 encodes two major antigens, ESAT6 and CFP10, which are associated with the virulence of M. tuberculosis strains (16). The loss of ESAT6 or CFP10 affected the upward adjustment of M. tuberculosis to miR-325-3p (see Fig. S1A in the supplemental material). In order to examine whether miR-325-3p plays a role in M. tuberculosis survival, RAW 264.7 cells were transfected with an miR-325-3p mimic or inhibitor (Fig. S1B). After infection with M. tuberculosis, bacterial CFU were scored every 24 h and converted to M. tuberculosis growth rates. The results showed that M. tuberculosis survival was augmented in the presence of the miR-325-3p mimic (Fig. 1E). Conversely, M. tuberculosis survival was compromised in macrophages treated with the miR-325-3p inhibitor (Fig. 1F). Furthermore, the function of miR-325-3p in M. tuberculosis pathogenesis was studied using wild-type (WT) and Mir325-deficient (Mir325 Ϫ/Ϫ ) mice. The median survival time of Mir325 Ϫ/Ϫ mice aerosol infected with M. tuberculosis, compared to that of the WT mice, was significantly lengthened (Fig. 1G). At 21 days postinfection (dpi), the lungs, spleens, and livers of Mir325 Ϫ/Ϫ mice infected with M. tuberculosis had lower bacterial loads than those of the WT mice (Fig. 1H). After M. tuberculosis stimulation, the macrophages from Mir325 Ϫ/Ϫ mice showed a higher apoptosis rate (Fig. 1I). Pathology analysis at 21 dpi showed that increased necrotic lesions in lungs of WT mice, while only a trail of inflammation was left in those of Mir325 Ϫ/Ϫ mice ( Fig. 1J and Table S1). This phenomenon indicated that Mir325 Ϫ/Ϫ mice were more resistant to M. tuberculosis infection than WT mice. Finally, an in vivo transfection method was used to confirm the functional effect of miR-325-3p. The miR-325-3p mimic or control mimic was transfected into Mir325 Ϫ/Ϫ mice through the tail intravenously (Fig. S1C), and the survival rates, organ bacterial burdens, and macrophage death mechanisms were determined afterwards to identify pathological damage. The results showed that Mir325 Ϫ/Ϫ mice transfected with the miR-325-3p mimic had restored susceptibility to M. tuberculosis (Fig. 1K to M). Likewise, pathology analysis of lungs supported increased necrotic lung lesions in mice with a higher miR-325-3p level ( Fig. 1N and Table S1). Taken together, these data suggested that miR-325-3p is upregulated after M. tuberculosis infection and is unfavorable for host resistance to M. tuberculosis.
LNX1 is a direct target of miR-325-3p. We next investigated whether miR-325-3p targets and regulates specific genes during M. tuberculosis infection. The targets of miR-325-3p were predicted by four microRNA databases (TargetScan Mouse 7.1, miR-TarBase, miRDB, and miRcode) and GenBank to obtain reference sequences and functional annotations. A series of transcripts were found as potential targets of miR-325-3p, including Gtf2a2, Cox7a1, Prss1, Tuba1c, Dgat1, and Lnx1; Lnx1 was selected for further study. Lnx1 had a specific binding site for miR-325-3p in its 3= untranslated region (3= UTR) ( Fig. 2A and Fig. S2). The 3= UTR of Lnx1 and its mutant were cloned into luciferase reporter vectors. Reporters were transfected into HEK293T cells along with the miR-325-3p mimic, and luciferase activities were measured. Luciferase activity was downregulated in the 3= UTR of WT Lnx1 but not in that of the mutant (Fig. 2B). When different concentrations of the miR-325-3p mimic were cotransfected with the WT Lnx1 3= UTR into cells, this caused a downregulation of the luciferase activity in a dose- Mouse survival data were plotted as Kaplan-Meier curves and compared by log rank (Mantel-Cox) test. Bacterial loads were analyzed using the Mann-Whitney U test. For other data, statistical significance between groups was determined by two-tailed Student's t test or one-way ANOVA followed by Bonferroni post hoc test. All data are presented as the mean Ϯ SDs and were derived from three independent experiments. *, P Ͻ 0.05; **, P Ͻ 0.01. n.s., not significant.

tuberculosis-infected Mir325
Ϫ/Ϫ mice transfected with miR-325-3p mimic in vivo at 21 dpi. Statistical significance between groups was determined by two-tailed Student's t test or one-way ANOVA followed by Bonferroni post hoc test. All data are presented as the means Ϯ SDs and were derived from three independent experiments. All blots are representative of three independent experiments. **, P Ͻ 0.01. dependent manner (Fig. 2C). Consistent with luciferase activity results, there was a reverse correlation between miR-325-3p and Lnx1 mRNA levels (Fig. 2D). In accordance with the upregulation of miR-325-3p during infection with gamma-irradiated M. tuberculosis, dual-luciferase assay and real-time quantitative PCR (qRT-PCR) were used to detect the transcription of Lnx1 in response to stimulation with gamma-irradiated M. tuberculosis. Luciferase activity was downregulated in the presence of WT Lnx1 3= UTR at 24 h postinfection (Fig. 2E). The Lnx1 mRNA level was decreased with the prolongation of M. tuberculosis infection (Fig. 2F). RNA-induced silencing complex (RISC) uses miRNA as a template to recognize the complementary mRNA sequence and to activate Argonaute, and Argonaute 2 (Ago2) has been identified as the catalytic center of RISC in mice (17). Therefore, we immunoprecipitated Myc-tagged Ago2 in the presence of the miR-325-3p mimic in RAW 264.7 cells and found that there was a significant enrichment with Lnx1 mRNA, suggesting that Lnx1 is a direct target of miR-325-3p (Fig. 2G). In order to further confirm the role of miR-325-3p in regulating LNX1, LNX1 expression levels were evaluated by Western blotting. The results showed that LNX1 was suppressed in the presence of the miR-325-3p mimic but upregulated by the miR-325-3p inhibitor (Fig. 2H). To further confirm the results, BMDMs from WT mice and RAW 264.7 cells were infected with gamma-irradiated M. tuberculosis and the expression levels of LNX1 were measured. Infection with gamma-irradiated M. tuberculosis prevented the expression of LNX1 in both types of macrophages (Fig. 2I). This phenomenon did not appear in the BMDMs from Mir325 Ϫ/Ϫ mice (Fig. 2J). However, the downregulation of LNX1 was restored in the Mir325 Ϫ/Ϫ mice after transfection with the miR-325-3p mimic (Fig. 2K). In addition, the expression of LNX1 in organs also showed that LNX1 was upregulated in Mir325 Ϫ/Ϫ mice but not in Mir325 Ϫ/Ϫ mice transfected in vivo with the miR-325-3p mimic ( Fig. 2L and M). These observations indicated that miR-325-3p targets LNX1 to suppress its expression during M. tuberculosis infection.
LNX1 is the E3 ubiquitin ligase of NEK6. To assess the functional effects of LNX1 during M. tuberculosis infection, we used the yeast two-hybrid system to screen the interacting proteins of LNX1. Since LNX1 has been reported as an E3 ubiquitin ligase (18), UbiBrowser and Ubiqsite were also used to predict potential targets and to search for the substrate of LNX1. Afterwards, potential candidates were further screened by Western blotting, and NEK6 was selected due to its low expression in M. tuberculosisresistant macrophages after infection. NEK6 is a serine/threonine protein kinase which is ubiquitously expressed in several tissues (19)(20)(21)(22), while the antituberculosis mechanism has not been reported. In order to determine the expression patterns of NEK6 and LNX1 in macrophages, hemagglutinin (HA)-tagged LNX1 or Lnx1 small interfering RNA (siRNA) was transfected into RAW 264.7 cells. The results showed that the changes of NEK6 were negatively correlated with LNX1 (Fig. 3A). Transfection of miR-325-3p or infection with gamma-irradiated M. tuberculosis can also downregulate LNX1, resulting in the accumulation of NEK6 in macrophages ( Fig. S3A and B). The CRISPR-Cas9 method was used to create LNX1-deficient L929 cells (LNX1 KO-1 and LNX KO-2). In the absence of LNX1, a higher expression level of NEK6 was observed. After transfection of an LNX1-overexpressing vector into LNX1-deficient cells, NEK6 was suppressed (Fig. 3B). Furthermore, we established myeloid cell-specific LNX1-deficient mice (Lnx1 fl/fl Lyz2-Cre) and Lnx1 fl/fl littermates. BMDMs from Lnx1 fl/fl Lyz2-Cre and Lnx1 fl/fl mice were infected with gamma-irradiated M. tuberculosis, and NEK6 levels were tested at the desired time points. The expression level of NEK6 showed no significant difference during infection in LNX1-deficient macrophages, while an upregulation of NEK6 was identified in cells from Lnx1 fl/fl mice (Fig. 3C). These data suggested that LNX1 is a negative regulator of NEK6.
Ubiquitination is an important enzymatic posttranslational modification; it can induce proteasome-mediated protein degradations (11). Due to the reverse correlation between LNX1 and NEK6, the polyubiquitination of NEK6 in the presence or absence of LNX1 was studied. To better observe the polyubiquitination of NEK6, MG132, a 26S proteasome inhibitor, was used to inhibit the degradation of NEK6 by proteasomes. The results showed that LNX1 did not change the expressions of E1 and E2 (Fig. S3C), while the polyubiquitination of NEK6 was significantly promoted in the presence of LNX1 compared to that in the absence of LNX1 (Fig. 3J and K and Fig. S3D and E). Consistently, the in vitro ubiquitination assay confirmed that LNX1 directly targeted NEK6 (Fig. S3F). Moreover, gamma-irradiated-M. tuberculosis-induced downregulation of LNX1 with a decreasing polyubiquitination of NEK6 was observed in BMDMs from Lnx1 fl/fl mice, whereas this phenomenon did not occur in macrophages from Lnx1 fl/ flLyz2-Cre mice (Fig. 3L). In addition, we confirmed that Nek6 mRNA level was not associated with LNX1 (Fig. S3G). Taken together, these data indicated that LNX1 is an E3 ubiquitin ligase that directly binds to NEK6 and then promotes the polyubiquitination and proteasome-mediated degradation of NEK6.
LNX1 promotes K48-linked polyubiquitination of NEK6 at the K174 site. Next, we constructed domain truncations of LNX1 to investigate the LNX1-NEK6 interaction (Fig. 4A). In LNX1-deficient L929 cells, coimmunoprecipitation showed that the deletion of RING or PDZ3 domain in LNX1 impaired LNX1-mediated polyubiquitination of NEK6 ( Fig. 4B and Fig. S4A). Transfection of RING and PDZ3 domains (RϩP3) into LNX1deficient cells restored the polyubiquitination of NEK6 (Fig. 4C and Fig. S4B). Further, lysine mutants (K85A, K92A, K98A, K135A, K174A, and K187A) of NEK6 were constructed to identify the recognition site of ubiquitin (Fig. 4D). Replacement of lysine with alanine at position 174 (K174A) abrogated LNX1-mediated polyubiquitination of NEK6 ( Fig. 4E and Fig. S4C). The ubiquitination and proteasomal degradation pathway of NEK6 was impeded because of the downregulation of LNX1 in macrophages stimulated with gamma-irradiated M. tuberculosis (Fig. 4F); this was consistent with our previous findings ( Fig. 3J and K). In contrast, the NEK6 K174A mutant could not be ubiquitinated during infection (Fig. 4F). Considering that different types of polyubiquitin link directly to distinct functions, we cotransfected NEK6 and LNX1 together with a series of ubiquitin mutants (K6O, K11O, K27O, K29O, K33O, K48O, and K63O) (23) into macrophages, each of which contained only one lysine available for polylinkage (Fig. 4G). The results showed that only the K48-linked polyubiquitination of NEK6 can be upregulated by LNX1. Meanwhile, the K48R ubiquitin mutant could not mediate polyubiquitination of NEK6 completely ( Fig. 4H and Fig. S4D). Collectively, these results demonstrated that LNX1 promotes K48-linked polyubiquitination of NEK6 at the K174 site. The RING and PDZ3 domains of LNX1 are necessary for the polyubiquitination of NEK6.
NEK6 inhibits apoptosis through activation of STAT3. We further investigated the specific role of NEK6 in the intracellular survival of M. tuberculosis. Since NEK6 is a protein kinase targeting STAT3 in mouse skin epidermal cells (24), we first focused on the regulation of the STAT pathway during M. tuberculosis infection. The phosphorylation of STAT3 was upregulated at 24 h postinfection (Fig. 5A). This result was consistent with the observed phosphorylation level of STAT3 with NEK6 overexpression or knockdown (Fig. 5B and Fig. S5A). Nek6 Ϫ/Ϫ mice showed a relatively lower level of phosphorylation of STAT3 both in BMDMs (Fig. 5C) and in organs (Fig. 5D) than the controls. STAT3 is a transcription activator that mediates the expression of a variety of genes in response to cell stimuli and thus plays a key role in many cellular processes, such as apoptosis and immune responses (25,26). Therefore, we tested the downstream genes that could be directly regulated by STAT3 (Fig. S5B). Notably, BCL-2 and its family genes (BCL-X L , BCL-W, and MCL-1) were upregulated during infection with gamma-irradiated M. tuberculosis infection in BMDMs from WT mice but not in those from NEK6-deficient mice (Fig. 5E and G). On the other hand, proapoptotic genes (such as Bax, Bik, Bad, and Bak) were not affected in NEK6-deficient macrophages (Fig. S5C). To further confirm our findings, interleukin 6 (IL-6) was used to activate the STAT3 pathway in NEK6-deficient BMDMs (27). As a result, the expression levels of the BCL-2 family were restored (Fig. 5F and H). Due to the anti-apoptotic function of the BCL-2 family, NEK6-sufficient macrophages showed an inhibition of apoptosis, which was (I) BMDMs from WT and Nek6 Ϫ/Ϫ mice were stimulated with 10 g/ml of gamma-irradiated M. tuberculosis for 24 h, and the apoptosis rates were detected by flow cytometry. (J) BMDMs from Nek6 Ϫ/Ϫ mice were pretreated with 20 ng/ml of IL-6 for 4 h, and then cells were stimulated with 10 g/ml of gamma-irradiated M. tuberculosis for 24 h. The apoptosis rates were detected by flow cytometry. Statistical significance between groups was determined by two-tailed Student's t test. All data are presented as the means Ϯ SDs and were derived from three independent experiments. All blots are representative of three independent experiments. **, P Ͻ 0.01. beneficial to the intracellular survival of M. tuberculosis (Fig. 5I and J). BCL-2 family genes regulate cell apoptosis by controlling mitochondrial membrane permeability (28). Based on this, we also found that NEK6-deficient BMDMs released cytochrome c from mitochondria to the cytoplasm and showed a high production of reactive oxygen species (ROS) (Fig. S5D to F). Taken together, these data indicated that NEK6 regulates innate immune responses through phosphorylation of STAT3 and participates in the regulation of apoptosis during M. tuberculosis infection.
Suppression of the NEK6/STAT3 pathway contributes to the defense against TB. To better understand the role of the NEK6-mediated immune response to M. tuberculosis infection, we measured the survival rates of Nek6 Ϫ/Ϫ mice infected with M. tuberculosis. Nek6 Ϫ/Ϫ mice showed an improved median survival time compared to that of WT mice (Fig. 6A). Bacterial burdens were lower in the lungs, spleens, and livers of Nek6 Ϫ/Ϫ mice (Fig. 6B). In the BMDMs from Nek6 Ϫ/Ϫ mice, M. tuberculosis multiplied more slowly (Fig. 6C). Through pathology scoring, less pathological changes were observed in the lungs of Nek6 Ϫ/Ϫ mice, whereas WT mice showed larger necrotic lesions ( Fig. 6D and Table S1). The fact that Nek6 Ϫ/Ϫ mice were more resistant to M. tuberculosis infection suggested that NEK6 contributed to the intracellular survival of M. tuberculosis. Moreover, to assess the effects of STAT3 on NEK6-mediated immune responses, a model of myeloid cell-specific LNX1-deficient mice (Lnx1 fl/fl Lyz2-Cre) that has a higher NEK6 expression in BMDMs was constructed. LNX1-deficient mice were treated with oral administration of BP-1-102 (a selective STAT3 inhibitor that is orally bioavailable) to suppress the activation of STAT3 (29). BP-1-102 had no direct toxic effect on intracellular M. tuberculosis (Fig. S6). LNX1-deficient mice were treated with BP-1-102, and their littermates were infected with M. tuberculosis. Figure 6E to I and Table S1 show that LNX1-deficient mice were extremely susceptible to M. tuberculosis, with increased necrotic lesions in the lungs (Fig. 6I). However, LNX1-deficient mice treated with BP-1-102 showed suppression of STAT3 accompanying a better resistance to M. tuberculosis. All these findings suggested that NEK6 is a negative regulator in the antituberculosis immune responses through the excessive activation of STAT3.
Collectively, our results indicate that we discovered a new immune escape pathway of M. tuberculosis by modulating host miR-325-3p expression. miR-325-3p targets LNX1 and results in the abnormal accumulation of NEK6, thus leading the inhibition of cell apoptosis and the intracellular survival of M. tuberculosis (Fig. 7).

DISCUSSION
miRNAs play significant roles in regulatory mechanisms of various biological processes, including host-pathogen interactions. More than 30 miRNAs have been shown to participate in immune responses during M. tuberculosis infection (10). miR-155 is upregulated in TB patients and in experimental settings upon M. tuberculosis infection, while the regulatory effect of miR-155 on M. tuberculosis clearance is controversial (30)(31)(32). A lot of miRNAs are involved in regulating innate immune signaling pathways through targeting genes. Let-7 targets A20 to suppress innate immune responses (8), miRNA-27a targets ER-located Ca 2ϩ transporter CACNA2D3 to inhibit autophagy (9), and miR-30a targets MyD88 to inhibit Toll-like receptor (TLR)/MyD88 activation and cytokine expression (33). In the previous study, we found that miR-325-3p was upregulated in a susceptible mouse substrain after aerosol infection with M. tuberculosis, yet there had been no interpretation of this phenomenon (data not published). In this study, we established mouse models and surprisingly, the survival time of M. tuberculosis-infected Mir325 Ϫ/Ϫ mice was significantly improved. This indicated that miR-325 might worsen the illness in M. tuberculosis infection.
The gene encoding LNX1, an E3 ubiquitin ligase, is one of the target genes of miR-325-3p. Human c-Src kinase is an important substrate of LNX1. The ubiquitination of c-Src by LNX1 has multiple effects on cell growth and cell differentiation (18,34). In this study, we proved that LNX1 and its substrate, NEK6, were downstream of miR-325-3p during M. tuberculosis infection. miR-325-3p directly targets LNX1, resulting in the accumulation of NEK6. The LNX1-NEK6 axis demonstrates the role of posttranslational modifications in host-M. tuberculosis interaction. LNX1 contains a RING domain and four PDZ domains, which are necessary for the recognition of substrates. For instance, the ubiquitination of PTPRF interacting protein alpha 1 is dependent on a PDZ binding motif, which contains a carboxyl-terminal cysteine that binds to LNX1 PDZ2 (18). In this study, we illustrated that the RING and PDZ3 domains of LNX1 are necessary Mouse survival data were plotted as Kaplan-Meier curves and compared by log rank (Mantel-Cox) test. Bacterial loads were analyzed using the Mann-Whitney U test. For other data, statistical significance between groups was determined by two-tailed Student's t test. All data are presented as the means Ϯ SDs and were derived from three independent experiments. *, P Ͻ 0.05; **, P Ͻ 0.01.

Fu et al.
® for the formation of the LNX1-NEK6 complex. Moreover, we showed that LNX1 catalyzes K48-linked polyubiquitination of NEK6 at the K174 site.
The NEK protein kinase subfamily can be identified in most eukaryotes, and its functions have been partially reported. For instance, NLRP3 inflammasome activation requires the interaction of NLRP3 with centrosomal kinase NEK7 (35). NEK6 is a protein kinase of CDK7, HSP70, and STAT3 (24,36,37). The accumulation of NEK6 certainly phosphorylated STAT3. In the classic JAK/STAT3 pathway, Tyr705 is a readout of JAK/STAT signaling (38). NEK6 may compete with JAK in STAT3 signaling. Future studies are needed to clarify the mechanisms by which NEK6 accumulation leads to the phosphorylation of STAT3 at Tyr705. As a transcriptional regulator, STAT3 has key roles in vertebrate development and mature tissue function, including the control of inflammation and immunity (39). STAT3 signaling is a major intrinsic pathway for inflammation. The continuous activation of STAT3 in tumor cells increases the expression of cytokines (transforming growth factor ␤ [TGF-␤] and IL-10) and then promotes tumor development and metastasis in a positive-feedback manner (40). Also, STAT3-deficient macrophages, neutrophils, and dendritic cells (DCs) produce elevated amounts of proinflammatory cytokines upon TLR4 activation, including tumor necrosis factor alpha (TNF-␣), IL-6, IL-12, and gamma interferon (IFN-␥) (41,42). However, in our research, the activation of STAT3 by NEK6 inhibited cell apoptosis through regulation of the transcription of anti-apoptotic BCL-2 family genes. Apoptotic cell death is one of the most important manners for the macrophages to eliminate the intracellular M. tuberculosis. The inhibition of apoptosis gives M. tuberculosis a chance to develop latent infection (43). Notably, BP-1-102, a selective inhibitor of STAT3, was proved to relieve pulmonary nodules in M. tuberculosis-infected mice through oral administration. This result provides a potential therapeutic approach to TB patients. Interestingly, we found that not only proinflammatory factor IL-6 but also anti-inflammatory factor IL-10 was raised by the phosphorylated STAT3 during M. tuberculosis infection. Accordingly, the involvement of inflammation with the NEK6-mediated immune escape of M. tuberculosis requires further investigation.
Together, our findings showed that miR-325-3p participates in the regulation of apoptosis during M. tuberculosis infection through targeting LNX1. According to the previous study, miR-325-3p directly targets RIPK3 to program necrosis in mice (44). Silencing of RIPK3 increases the viability of cardiomyocytes under hypoxic conditions expression levels. Data were analyzed using the threshold cycle (2 ϪΔΔCT ) method. All primer sequences used for qRT-PCR are listed in Table S3.
Luciferase assays. A dual-luciferase reporter (DLR) assay system (Promega) was used to perform luciferase assays as previously described (48). In brief, cells were cotransfected with luciferase reporter plasmid and internal control plasmid pRL-SV40. Cells were lysed for DLR assays 24 h after treatment. Data were collected with a VICTOR X5 multilabel plate reader (PerkinElmer, Waltham, MA), and relative luciferase activities were measured by firefly luciferase luminescence divided by Renilla luciferase luminescence.
Flow cytometry. Annexin V staining, paired with propidium iodide (PI), was used to identify apoptotic cells with an annexin V-fluorescein isothiocyanate (FITC) apoptosis detection kit (BD Biosciences). Flow cytometry analysis was performed according to standard procedures.
Statistical analysis. All data were derived from at least three independent experiments and are presented as the means Ϯ standard deviations (SD) unless otherwise indicated. Two-tailed Student's t test was used to compare the means between two groups, and one-way analysis of variance (ANOVA) followed by Bonferroni post hoc test was used for multiple comparisons. Mouse survival data were plotted as Kaplan-Meier curves and compared by log rank (Mantel-Cox) test. Bacterial loads were analyzed using the Mann-Whitney U test. All blots are representative of three independent experiments. A P value of Ͻ0.05 was considered significant.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.