A novel anti-mycobacterial function of mitogen-activated protein kinase phosphatase-1

Background Mycobacterium tuberculosis (MTB) is a major cause of morbidity and mortality in the world. To combat against this pathogen, immune cells release cytokines including tumor necrosis factor-α (TNF-α), which is pivotal in the development of protective granulomas. Our previous results showed that Bacillus Calmette Guerin (BCG), a mycobacterium used as a model to investigate the immune response against MTB, stimulates the induction of TNF-α via mitogen-activated protein kinase (MAPK) in human blood monocytes. Since MAPK phosphatase-1 (MKP-1) is known to regulate MAPK activities, we examined whether MKP-1 plays a role in BCG-induced MAPK activation and cytokine expression. Results Primary human blood monocytes were treated with BCG and assayed for MKP-1 expression. Our results demonstrated that following exposure to BCG, there was an increase in the expression of MKP-1. Additionally, the induction of MKP-1 was regulated by p38 MAPK and extracellular signal-regulated kinase 1 and 2 (ERK1/2). Surprisingly, when MKP-1 expression was blocked by its specific siRNA, there was a significant decrease in the levels of phospho-MAPK (p38 MAPK and ERK1/2) and TNF-α inducible by BCG. Conclusions Since TNF-α is pivotal in granuloma formation, the results indicated an unexpected positive function of MKP-1 against mycobacterial infection as opposed to its usual phosphatase activity.


Background
Tuberculosis (TB) remains a major cause of morbidity and mortality in the world, especially in the developing countries [1]. The disease is caused by Mycobacterium tuberculosis (MTB) and approximately one third of the world's population has been infected by this pathogen. In a recent report, World Health Organization (WHO) estimated that there are 9.2 million new TB cases around the world in 2006 [1].
In response to MTB infection, induction of cytokines by immune cells is an important defense mechanism. The infected macrophages secrete intercellular signaling factors, proinflammatory cytokines, to mediate the inflammatory response leading to the formation of granuloma and induction of T-cell mediated immunity [2]. In order to understand TB pathogenesis, signaling pathways induced by mycobacteria have long been a subject of interest. Mitogen activated protein kinases (MAPKs) including extracellular signal-regulated kinase 1 and 2 (ERK1/2), p38 MAPK, and c-Jun N-terminal kinase (JNK) have been implicated as important cellular signaling molecules activated by mycobacteria [3]. Previous reports have shown that p38 MAPK and ERK1/2 are required in the induction of TNF-α expression in human monocytes infected with M. tuberculosis H37Rv [4]. We have further revealed the significant role of MAPKs in the signal transduction events of mycobacterial activation of primary human blood monocytes (PBMo) leading to cytokine expressions via the interaction with PKR [5]. However, the subsequent events as to how MAPK is regulated and how such regulation affects cytokine production in response to mycobacteria remain to be elucidated.
Since MAPKs are activated by phosphorylation, dephosphorylation of MAPKs seems to be an efficient process to inactivate their activities. It can be achieved by specific protein kinase phosphatases which can remove the phosphate group from MAPKs. Examples of these phosphatases include tyrosine phosphatases, serine/threonine phosphatases, and dual-specificity phosphatases (DUSPs). Some DUSPs are also known as MAPK phosphatases (MKPs) [6][7][8]. Currently, there are at least 10 MKPs identified, while MKP-1 is the most studied member of the family. The regulatory role of MKP-1 on cytokine induction is best demonstrated by MKP-1 knockout (KO) macrophages in response to lipopolysaccharide (LPS), a cell wall component of Gram-negative bacteria. MKP-1 KO macrophages showed prolonged phosphorylation of p38 MAPK and JNK as well as increased production of TNF-α in response to LPS treatment [9]. Consistent with these results, another group further revealed that LPS-treated MKP-1 KO bone marrow-derived macrophages show increased AP-1 DNA-binding activity [10]. Also, they showed that LPS-induced MKP-1 expression is dependent on myeloid differentiation factor 88 (MyD88) and TIR domain-containing adaptor inducing IFN-β (TRIF) [10], thus demonstrating the role of MKP-1 in signal transduction.
Not only LPS, other TLR inducers including CpG, peptidoglycan, poly IC, and Pam 3 Cys can regulate cytokine expressions including TNF-α, IL-10 via MKP-1 activities [10,11]. In these processes, MKP-1 serves to mitigate the undesirable effects of septic shock and maintain organ functions by restraining the inflammatory responses following bacterial infection. Another example of MKP-1 function is the immune response to Staphylococcus aureus (S. aureus), a Gram positive bacteria. There are higher levels of cytokine production including TNF-α, IL-6, and MIP-1α in MKP-1 KO mice infected with S. aureus [12]. Also, the mice would have a rapid development of multiorgan dysfunction as well as faster mortality rate upon challenge with heat-killed S. aureus [12]. Taken together, these results suggest that MKP-1 protects the host from overactivation of the immune system in response to Gram negative or Gram positive bacteria.
In the past, it was believed that different MKP/DUSP family members have overlapping functions. However, the emergence of DUSP2 turned the concept up side down [13]. It was shown that DUSP2 behaves differently and is opposite to the function as stated above. In DUSP2 KO cells, they produced less inflammatory mediators, implying that DUSP2 may play a role in mediating instead of limiting inflammation. For instances, when DUSP2 KO macrophages were treated with LPS, there were less TNF, IL-6, nitric oxide, IL-12-producing cells when compared to those of the wild type counterparts [13]. When the DUSP2 KO bone marrow-derived mast cells were first sensitized with immunoglobulin E (IgE) receptor (FcεRI) and then stimulated with dinitrophenol-heat stable antigen, they produced lower TNF mRNA levels, diminished IL-6 production, less phosphorylation of ERK1/2, p38 MAPK, and less transcriptional activities by Elk1 and NFAT-AP-1 [13].
These unexpected positive regulations of immune cell functions by DUSP2 have been hypothesized to be due to crosstalks between MAPKs [13]. Stimulation of KO mast cells and macrophages showed increases in phosphorylation of JNK. Moreover, inhibition of JNK by small molecule inhibitors showed increases in phosphorylation of ERK [13]. The authors also showed that there were physical interactions of DUSP2 with ERK2, DUSP2 with JNK2, as well as DUSP2 and p38 MAPK after stimulation of the cells with dinitrophenol-heat stable antigen. Nevertheless, the details of the crosstalks between MAPKs and phosphatases need further investigation. Thus, the MKP family plays a critical role in the regulation of immune responses.
Innate immune response protects the host from MTB infection by secretion of cytokines including TNF-α in immune cells. Meanwhile, MAPK is one of the critical proteins in the regulation of immunity and cytokine expression. Since MAPK is regulated by MKP-1 in response to LPS and the activation of MAPK is important in BCGinduced cytokine expression, we hypothesize that MKP-1 plays a critical role in the immune regulation of BCG in human monocytes. We examined the involvement of MKP-1 in BCG-induced MAPK activation and its consequent cytokine expression. Here, we present evidences that MKP-1 plays an unexpected role in the regulation of cytokine induction by BCG through its control of MAPK phosphorylation.

Different inducers show differential induction of MKP-1
It has been reported that many inducers including growth factors, LPS, peptidoglycan, and dexamethasone can stimulate the expression of MKP-1 in human macrophages, microglia, mast cells or fibroblasts [6]. To investigate the role of different TLR inducers in MKP-1 induction process in human blood monocytes, the level of MKP-1 mRNA was measured by quantitative polymerase chain reaction (QPCR) method. PBMo were isolated from primary human blood mononuclear cells and stimulated with Pam 3 Cys (TLR2 agonist), poly IC (TLR3 agonist), or LPS (TLR4 agonist) for 1 and 3 hours. Following exposure to Pam 3 Cys or LPS, there were significant inductions of MKP-1 mRNA levels within 1 hour of treatment ( Figure  1A). These effects on MKP-1 induction continued for 3 hours post-treatment with Pam 3 Cys ( Figure 1A). In contrast, poly IC did not induce MKP-1 ( Figure 1A). The results indicate that different inducers showed differential up-regulation of MKP-1 expression.

LPS induces MKP-1 transcription
LPS has been extensively used to demonstrate the role of MKP-1 in immune response both in vivo and in vitro [9,12]. To establish a foundation for interpretation of subsequent experimental results, LPS was used as a positive control for the induction of MKP-1 expression. To determine the levels of MKP-1 in response to LPS, kinetics of MKP-1 transcription were determined by QPCR. There was a significant induction of MKP-1 mRNA, which peaked as early as 1 hour upon LPS stimulation, and the levels gradually decreased over a course of 6 hours. These results showed that LPS induced MKP-1 expression (Figure 1B).

BCG induces MKP-1 expression
Next, to demonstrate the induction of specific phosphatases by BCG, kinetics of MKP-1 expression in PBMo was studied by using QPCR during BCG treatment. Similar to the results produced by LPS, upon the addition of BCG (MOI = 1 CFU/cell), there was a significant induction of MKP-1 mRNA within 1 hour of BCG treatment as determined by Taqman probe specific for MKP-1 ( Figure  2A). The effects lasted for at least 6 hours ( Figure 2A).
To examine whether the changes of protein production were in parallel to that of the mRNA levels, the protein levels of MKP-1 were measured by Western blotting. In response to BCG, PBMo produced the MKP-1 protein as early as 30 minutes after treatment. The protein levels were maintained for 2 hours and dropped to basal levels at 3 hours ( Figure 2B). The results demonstrated that there was MKP-1 induction in response to BCG activation in human monocytes.

BCG-induced MKP-1 expression is dependent on p38 MAPK and ERK1/2
It has been shown that inhibition of p38 MAPK either by specific inhibitor or siRNA reduced the expression of MKP-1 in LPS-or peptidoglycan-treated macrophages [14]. To determine the mechanisms involved in the BCGinduced MKP-1 expression, PBMo were pretreated with several inhibitors including PD98059 (inhibitor for MAP kinase kinase [MEK] or ERK1/2), SB203580 (inhibitor for p38 MAPK), SP600125 (inhibitor for JNK), and CAPE (inhibitor for NF-κB) for 1 hour. A range of concentrations of each inhibitor was used to test their optimal concentrations and effects on cell viability and kinase inhibitions. BCG was added afterwards and total RNA was harvested. The results demonstrated that, with the inhibition of ERK1/2 and p38 MAPK activities by their corresponding relatively specific inhibitors, MKP-1 expressions were significantly reduced ( Figure 3). In addition, using higher dose of SB203580, we showed that the inhibition is increased further (data not shown). On the contrary, pretreatment of the cells with CAPE and SP600125 did not affect the induction of MKP-1 by BCG ( Figure 3). These results suggest that BCG-induced MKP-1 expression is dependent on both p38 MAPK and ERK1/2.

Transfection of MKP-1 siRNA into PBMo can reduce BCGinduced MKP-1 production
Throughout the above experiments, the primary goal was to examine the induction of MKP-1 by BCG in human monocytes. Thus, to further examine the role of MKP-1 in BCG-induced signaling, transfection of siRNA into PBMo was used to knockdown the activity of MKP-1. To demonstrate that the MKP-1 siRNA can indeed knockdown the target gene, PBMo were first transfected with control or MKP-1 siRNA and then treated with BCG for 3 hours. Levels of MKP-1 mRNA were measured by RT-PCR method.
In Figure 4A, BCG stimulated MKP-1 expression (lanes 1 and 2). In MKP-1 siRNA transfected monocytes, induction of MKP-1 by BCG was significantly decreased (lanes 2 and 4). The results showed that the siRNA does abrogate the levels of MKP-1 mRNA.
To further determine whether MKP-1 siRNA affects BCGinduced MKP-1 at protein levels, PBMo were treated as above and MKP-1 proteins were measured by Western blotting. The results showed that BCG could induce MKP-1 proteins as usual for cells transfected with control siRNA ( Figure 4B, lanes 1-3). However, the levels of BCGinduced MKP-1 protein expression were reduced in cells transfected with MKP-1 siRNA ( Figure 4B, lanes 4-6). Together, the results suggest that MKP-1 siRNA not only reduced the MKP-1 mRNA in BCG treatment but also abrogated the BCG-induced MKP-1 protein.

LPS-induced cytokine expression is increased by MKP-1 siRNA
As stated in the literature [9], MKP-1 KO mice showed increased TNF-α production in response to LPS. On the basis of the above MKP-1 siRNA results, LPS was then used as a control to demonstrate the effects of this MKP-1 siRNA system. cytokine expression induced by LPS in MKP-1 siRNA transfected cells suggest that the siRNA system is effective in knocking down the MKP-1 expression and MKP-1 acts as a negative regulator in LPS-induced TNF-α expression.

BCG-induced cytokine expression is decreased by MKP-1 siRNA
To investigate the effect of MKP-1 siRNA on BCG-induced cytokine expression, the levels of TNF-α, IL-6 and IL-10 mRNA were measured by QPCR method. PBMo were transfected with either control or MKP-1 siRNA. Following exposure to BCG with control siRNA, there were significant inductions of TNF-α, IL-6 and IL-10 mRNA levels for 3 hours after treatment as previously reported ( [5] and data not shown). Next, the effects of MKP-1 siRNA were examined on the cytokine expression induced by BCG. Surprisingly, there was a significant abrogation of BCGinduced TNF-α expression by MKP-1 siRNA ( Figure 4D). With the knockdown of MKP-1, the level of BCG-induced TNF-α was only 60% compared to that of the control cells, while BCG-induced IL-6 and IL-10 were unchanged in MKP-1 siRNA transfected cells. The results revealed that MKP-1 plays a role in the induction of TNF-α expression upon BCG stimulation, which may be different from that of its conventional functions in which MKP-1 acts as a negative regulator in LPS-induced signaling pathways [7].

BCG-induced MAPK phosphorylation is decreased by MKP-1 siRNA
The unexpected observations in cytokine expression lead to the investigation on the effects of MKP-1 siRNA on BCG-induced MAPK activation. MKP-1 was found to have a preferential substrate binding to p38 MAPK and JNK than ERK1/2 [7]. The phosphorylation status of MAPKs was assessed in control or MKP-1 siRNA transfected PBMo. Western blotting results demonstrated that BCGinduced both p38 MAPK and ERK1/2 phosphorylation in 15 minutes (data not shown) and peaked at 30 minutes, and then returned to basal levels in cells treated with the control siRNA ( Figure 5). Similar to the results of cytokine expression, phosphorylation of both p38 MAPK and ERK1/2 in response to BCG was decreased in monocytes transfected with MKP-1 siRNA instead of the expected increase in phosphorylation ( Figure 5). The results suggest that MKP-1 knockdown would result in reduced MAPK phosphorylation by BCG, implying that the reduced level of TNF-α production in BCG stimulated monocytes is due to reduced phosphorylation of MAPKs by MKP-1 siRNA.

Discussion
This report presented evidences that a novel function of MKP-1 is uncovered in cytokine regulation in response to mycobacterial infection. BCG induces MKP-1 as a rapid response (Figure 2). The induction mechanism of MKP-1 by BCG is dependent on both ERK1/2 and p38 MAPK ( Figure 3). Using siRNA approach, the functions of MKP-1 can be examined in primary human monocytes. The results showed that the BCG-induced MAPKs activation as well as cytokine expression are downstream of MKP-1 ( Figures 4D and 5). Thus, MKP-1 is a critical signaling molecule that is involved in BCG-induced cytokine expression.
Previous reports have shown that MKP-1 induced by LPS or peptidoglycan is dependent on p38 MAPK [14]. Accordingly, BCG-induced MKP-1 can be inhibited by both p38 MAPK and ERK1/2 inhibitors. Interestingly, it has been shown that degradation of MKP-1 is reduced after ERK1/2 phosphorylation [15]. It can be hypothesized that BCG-induced MKP-1 proteins can be stabilized by ERK1/2 and the detailed mechanisms involved require more exploration. Also, since the inhibition of MKP-1 expression by both inhibitors (for p38 MAPK and ERK1/ 2) was not complete, it is believed that other proteins may be involved in the BCG-induced MKP-1 expression.
On the basis of the literature results on LPS effects ( Figure  6), the original expectation for this project is that MKP-1 acts as a negative regulator. LPS-stimulated MKP-1 KO peritoneal macrophages showed prolonged phosphorylation of p38 MAPK and JNK as well as increased production of TNF-α [9]. In doing so, LPS-induced MKP-1 could BCG-induced MAPK phosphorylation is decreased by MKP-1 siRNA  prevent prolonged TNF-α production as in sepsis which may lead to severe damage to the host. It was expected that BCG induces MKP-1 and its induction would correlate with the dephosphorylation of MAPKs including p38 MAPK. By blocking the MKP-1 using siRNA, it was expected to have increased p38 MAPK phosphorylation and prolonged TNF-α production in response to BCG. Nevertheless, our results shown here are diametrically opposite. One possibility for the unexpected results may be due to non-specific effects of transfection or siRNA. However, this was not the case since there was a prolonged and increased TNF-α expression after the MKP-1 siRNA-transfected monocytes were treated with LPS (Figure 4C).
There is now a new hypothesis to explain such paradoxical effects of MKP-1 in TNF-α regulation in which the phosphatase plays a role in positive regulation of TNF-α production in response to BCG as in the case of DUSP2 [13]. The structures of MKP-1 and DUSP2 are similar, with which they both contain a MAPK-interacting domain and a phosphatase catalytic site. By contrast, other DUSP may have extra domains, e.g., PEST [6]. Here, we postulate that the function of MKP-1 in BCG-induced signaling is similar to that of the DUSP2/PAC1.
Actually, the discovery of DUSP2 has initially created some paradoxical questions. As described, DUSP2 behaves differently from other MKP family members [13]. In DUSP2 KO macrophages treated with LPS, they produced less inflammatory mediators including less TNF, IL-6, nitric oxide, and IL-12-producing cells, when compared to that of the wild type counterparts [13]. Indeed, the results of these published studies on DUSP2 studies are quite similar to that of our reported results here.
It is plausible that these unexpected positive regulations of immune cell functions by DUSP2 were due to crosstalks between MAPKs [13]. It was shown that there are interactions between JNK and ERK1/2 pathways [16].
Here, we showed that the sustained activation of JNK blocks ERK activation ( Figure 6). In the DUSP2 situation, stimulation of KO mast cells and macrophages shows increased phosphorylation of JNK, and inhibition of JNK by its own specific inhibitor restores phosphorylation of ERK1/2 [13].
In the BCG-MKP-1 situation, there is an early phosphorylation of p38 MAPK and ERK1/2. Therefore, it is possible that JNK may play a role in the crosstalk interaction of MAPK. However, our preliminary data suggest that the level of phosphorylated JNK was not increased in PBMo MKP-1 plays a critical role in the regulation of cytokine expression upon mycobacterial infection Figure 6 MKP-1 plays a critical role in the regulation of cytokine expression upon mycobacterial infection. LPS model was provided according to literature findings (Left). In this scenario, LPS activates MKP-1, which in turn dephosphorylates and deactivates phospho-p38 MAPK, resulting in less TNF-α induction. However, the situation in DHP-HSA activation of DUSP2 is more complicated (Middle), since the phosphatase activity causes subsequent inhibition of phospho-JNK which leads to the derepression of phospho-p38 MAPK. Consequently, the combined effects of this cascade results in more TNF-α expression. The unexpected antimycobacterial role of MKP-1 (Right) may be explained by events similar to the DUSP2 effects. In this case (Right), there was an inhibition of unknown pathways or kinases downstream of MKP-1, and the unknown factor in turn inhibits MAPKs activation leading to more TNF-α induction. The details and kinase targets are yet to be identified. transfected with MKP-1 siRNA (data not shown). Thus, the details of the crosstalk between MAPKs need further investigation. Here, we present a model to summarize the results and to hypothesize the existence of an as yet unidentified intermediary factor or factors in the pathways downstream of MKP-1 effects in the BCG-induced signaling cascade. The unexpected antimycobacterial role of MKP-1 ( Figure 6) may be explained by events similar to the DUSP2 effects. In this case, BCG induces MKP-1 expression while also activates MAPKs including p38 MAPK and ERK1/2. Downstream of MKP-1, there is an inhibition of unknown pathways or kinases. The unknown factor in turn inhibits MAPKs activation, which ultimately leads to more TNF-α induction ( Figure 6).

Conclusions
In summary, MKP-1 plays a critical role in the regulation of cytokine expression upon mycobacterial infection. Inhibition of unknown pathways or kinases downstream of MKP-1, which in turn inhibits MAPKs activation, may be used to explain the novel function of MKP-1 in enhancing MAPK activity and consequent TNF-α expression following BCG treatment ( Figure 6). Taken together, the role of MAPK crosstalks need further exploration.  (3) TNF-α, 30 cycles (TM = 56°C), upstream, 5'-GGCTCCAGGCGGTGCTTGTTC-3', downstream, 5'-AGACGGCGATGCGGCTGATG-3'. PCR products were analyzed on a 1% agarose gel with ethidium bromide and visualized under ultraviolet light. In order to check the size of the PCR products, 1 kb Plus DNA Lad-der™ (Invitrogen, USA) was run along with the PCR products.

Quantitative polymerase chain reaction (QPCR)
To perform QPCR, the levels of MKP-1, and TNF-α mRNA as well as the reference gene GAPDH (as internal control) were assayed by the gene-specific Assays-on-Demand reagent kits (Applied Biosystems, USA). All samples were run in duplicates or triplicates and with no template controls on an ABI Prism 7700 Sequence Detector. The analysis method of QPCR was the comparative cycle number to threshold (C T ) method as described in user bulletin no. 2 of the ABI Prism 7700 Sequence Detection System. The number of C T of the targeted genes was normalized to that of GAPDH in each sample (ΔC T ). The C T value of the treated cells was compared with that of the untreated or mock-treated cells (ΔΔCT). The relative gene expression of the targeted genes (fold induction) was calculated as 2 -ΔΔCT .

Western blot analysis
Western blot was done as described [20]. Equal amounts of protein were separated by 10% SDS-PAGE, electroblotted onto nitrocellulose membranes (Schleicher & Schuell), and followed by probing with specific antibod-ies for Actin, MKP-1 (Santa Cruz Biotech., USA), phospho-p38 MAPK, phospho-ERK1/2 (Cell Signaling, USA). After three washes, the membranes were incubated with the corresponding secondary antibodies. The bands were detected using the Enhanced Chemiluminescence System (Amersham Pharmacia Biotech) as per the manufacturer's instructions.

Statistical Method
Statistical analysis was performed by Student's t test. Differences were considered statistically significant when p values were less than 0.05.