Sterile-α- and Armadillo Motif-Containing Protein Inhibits the TRIF-Dependent Downregulation of Signal Regulatory Protein α To Interfere with Intracellular Bacterial Elimination in Burkholderia pseudomallei-Infected Mouse Macrophages

Abstract

Burkholderia pseudomallei, the causative agent of melioidosis, evades macrophage killing by suppressing the TRIF-dependent pathway, leading to inhibition of inducible nitric oxide synthase (iNOS) expression. We previously demonstrated that virulent wild-type B. pseudomallei inhibits the TRIF-dependent pathway by upregulating sterile-α- and armadillo motif-containing protein (SARM) and by inhibiting downregulation of signal regulatory protein α (SIRPα); both molecules are negative regulators of Toll-like receptor signaling. In contrast, the less virulent lipopolysaccharide (LPS) mutant of B. pseudomallei is unable to exhibit these features and is susceptible to macrophage killing. However, the functional relationship of these two negative regulators in the evasion of macrophage defense has not been elucidated. We demonstrated here that SIRPα downregulation was observed after inhibition of SARM expression by small interfering RNA in wild-type-infected macrophages, indicating that SIRPα downregulation is regulated by SARM. Furthermore, this downregulation requires activation of the TRIF signaling pathway, as we observed abrogation of SIRPα downregulation as well as restricted bacterial growth in LPS mutant-infected TRIF-depleted macrophages. Although inhibition of SARM expression is correlated to SIRPα downregulation and iNOS upregulation in gamma interferon-activated wild-type-infected macrophages, these phenomena appear to bypass the TRIF-dependent pathway. Similar to live bacteria, the wild-type LPS is able to upregulate SARM and to prevent SIRPα downregulation, implying that the LPS of B. pseudomallei may play a crucial role in regulating the expression of these two negative regulators. Altogether, our findings show a previously unrecognized role of B. pseudomallei-induced SARM in inhibiting SIRPα downregulation-mediated iNOS upregulation, facilitating the ability of the bacterium to multiply in macrophages.

INTRODUCTION

Infection with Burkholderia pseudomallei, a Gram-negative intracellular pathogen and the causative agent of melioidosis, may lead to several outcomes ranging from asymptomatic seroconversion to life-threatening clinically apparent manifestations such as pneumonia and multiple abscesses (1). In recent years, there has been a worldwide increase in research focusing on the pathogenesis of this bacterium in order to gain an insight into the mechanisms used by this bacterium to persist in the host. The bacterium is notoriously known to possess an array of virulence factors by which it can easily evade and/or resist the host defense mechanisms (2, 3). A prominent feature of B. pseudomallei pathogenesis is its ability to invade, survive, and intracellularly multiply in a variety of phagocytic and nonphagocytic cells (4). After cellular uptake, this bacterium can escape from the endocytic vacuole into the cytoplasm (5, 6). The bacteria can then replicate in the cytoplasm and induce the formation of actin-associated membrane protrusions and multinucleated giant cells (MNGC) (5, 7).

The initial immune response after the recognition of B. pseudomallei by the immune cells is considered to be crucial for determining the pathogenesis outcome (8, 9). Macrophages act as the front line of defense against invading B. pseudomallei by playing a critical role for the early control of B. pseudomallei infection (10). However, these immune cells become the target of immune modulation by B. pseudomallei to make a favorable environment for its intracellular replication and spread. Previously, it was demonstrated that virulent wild-type B. pseudomallei (strain 1026b) could multiply in the mouse macrophage cell line and primary human macrophages over an extended period of time (11, 12). In contrast, the less virulent lipopolysaccharide (LPS) mutant of B. pseudomallei (lacking the O-antigenic polysaccharide moiety of LPS) was found to be susceptible to the macrophage antibacterial response, leading to restricted intracellular bacterial growth (11–13). These studies suggest that the virulent B. pseudomallei strain has the strategy to manipulate macrophage intracellular killing mechanisms for its multiplication.

Macrophages recognize and respond to B. pseudomallei by the use of Toll-like receptors (TLRs), which are important for the production of proinflammatory cytokines and antimicrobial effectors (1, 14). TLRs consist of two signaling pathways, the MyD88-dependent and TRIF-dependent (also known as MyD88-independent) pathways (14). To date, more than 10 TLRs have been identified in mice (14). The importance of these pathways in macrophage killing of invading pathogens has been demonstrated in various studies (15–18). Although TLR4 signaling can take place through both pathways, the majority of genes expressed in LPS-activated macrophages appear to use a TRIF-dependent pathway, implying that the TRIF-dependent pathway may serve as a pivotal regulator of host defense against Gram-negative pathogens (19). The involvement of the TRIF signaling pathway in controlling the B. pseudomallei multiplication in mouse macrophages has been previously demonstrated by our group (18). The virulent wild-type B. pseudomallei strain was demonstrated to be able to suppress the TRIF signaling pathway (18). Because of such a feature, the wild-type-infected macrophages fail to upregulate the expression of genes downstream of this pathway such as inoS and ifn-β (18, 20). This mechanism contributes to the intracellular multiplication of wild-type bacteria in mouse macrophages. In contrast, the less virulent LPS mutant B. pseudomallei was found to be able to activate the TRIF signaling pathway, resulting in the induction of inoS and ifn-β gene expression, and thus shows restricted intracellular growth (18). The underlying mechanisms responsible for the differential nature of activation of the TRIF signaling pathway by these two strains are not fully understood. However, our recent findings suggest the involvement of negative regulators of TLRs, signal regulatory protein α (SIRPα), and sterile-α- and armadillo motif-containing protein (SARM) in the inhibition of the TRIF signaling pathway (21, 22).

SARM is a TLR adaptor protein that can interfere with the TRIF-dependent gene expression by direct interaction with TRIF (23), whereas SIRPα is an immunoreceptor tyrosine-based inhibition motif (ITIM)-containing transmembrane receptor and can recruit SH2 domain-containing protein tyrosine phosphatase 1 (SHP-1) and SHP-2 through its ITIM motif (24). Both of these negative regulators have been demonstrated to inhibit TLR signaling pathways, particularly the TRIF-dependent pathway, in response to TLR ligands (25–27). We previously found that wild-type B. pseudomallei can induce the SARM expression and also can prevent the downregulation of constitutively expressed SIRPα in mouse macrophages, resulting in the inhibition of inducible nitric oxide synthase (iNOS) expression and an increase in intracellular survival (21, 22). In contrast, the LPS mutant can strongly suppress the expression of these two negative regulators over a prolonged time period of infection (21, 22). However, the functional relationship of these two negative regulators in the evasion of the macrophage defense mechanism by B. pseudomallei has not been understood. Although a distinct role of TLR4 signaling in activating the LPS-induced SIRPα downregulation in mouse macrophages has been demonstrated, a possible involvement of such a mechanism that may cause the differential SIRPα expression patterns among wild-type- and LPS mutant-infected macrophages has not been investigated (27). Therefore, for the present report, we extended our study to determine such a possibility and examined the role of SARM in the regulation of SIRPα expression upon infection by these two B. pseudomallei strains.

MATERIALS AND METHODS

Cell line and culture conditions.

Mouse macrophages (RAW 264.7 cells) (ATCC) were cultured in complete Dulbecco's modified Eagle's medium (DMEM) (HyClone, Logan, UT), defined as DMEM supplemented with 10% fetal bovine serum (HyClone) and 1% l-glutamine (Gibco Laboratories, Grand Island, NY), at 37°C in a 5% CO₂ atmosphere.

Bacterial strains and growth conditions.

Wild-type B. pseudomallei (strain 1026b) as used in this study was previously described (13). The B. pseudomallei LPS mutant (mutant strain of parental 1026b strain) that lacks the O-antigenic polysaccharide moiety was used for comparison (13). Before use in the experiments, frozen bacteria were subcultured onto Luria-Bertani (LB) agar. A single bacterial colony from an LB agar plate was always selected to inoculate the bacteria into LB broth, which was then cultured at 37°C with agitation at 150 rpm. Overnight cultures were washed twice in phosphate-buffered saline (PBS) and adjusted to a concentration of 10⁸ CFU/ml in complete DMEM by measuring the optical density at 650 nm. CFU was calculated from the precalibrated standard curve.

Infection of mouse macrophages.

An overnight culture of mouse macrophages (1 × 10⁶ cells) in a 6-well plate was infected with either wild-type or LPS mutant B. pseudomallei at a multiplicity of infection (MOI) of 2 for 1 h. To remove uninternalized bacteria, cells were washed twice with 1 ml of PBS and residual bacteria were killed by incubating the cells in complete DMEM containing 250 μg/ml of kanamycin (Gibco) for 2 h. Thereafter, the infection was allowed to continue in the medium containing 20 μg/ml of kanamycin until the experiment was terminated (7). Tetracycline (Sigma-Aldrich, St. Louis, MO) at a concentration of 50 μg/ml was supplemented in the medium when the LPS mutant was cocultured with the macrophages in order to avoid a possible reversion to the wild type. In the case of gamma interferon (IFN-γ) prestimulation, macrophages were pretreated overnight with recombinant IFN-γ (CytoLab, Rehovot, Israel) (10 U/ml), and the cells were washed twice with PBS before infection. Subsequent steps were the same as those described above. Host cell viability was determined using 0.04% trypan blue staining to ensure that more than 90% of the cells remained alive at the final time point (unless otherwise indicated).

Preparation of LPS.

LPS was extracted from pathogenic B. pseudomallei (strain 1026b) by a modified phenol-chloroform-petroleum-ether method (28) and characterized by SDS-PAGE and immunoblotting as described previously (29). Escherichia coli LPS (Sigma-Aldrich) and Porphyromonas gingivalis LPS (Sigma-Aldrich) were used for comparisons. Each of these LPS at a concentration of 100 ng/ml was added to macrophage cultures, and they were incubated together until the indicated time points.

Depletion of SARM and TRIF in mouse macrophages.

Mouse macrophages (RAW 264.7 cells) were transfected with SARM or TRIF small interfering RNA (siRNA) according to the manufacturer's protocol. In brief, macrophages (1.5 × 10⁵ cells/well) were seeded overnight in a 6-well plate. Cells were then transfected with 60 nM (each) negative-control siRNA (catalogue no. 12935-200; Invitrogen, Hilden, Germany), SARM siRNA (catalogue no. 1320001, oligonucleotide identification [ID] no. MSS239396; Invitrogen), or TRIF siRNA (catalogue no. 1320001, oligonucleotide ID no. MSS200702; Invitrogen) using Lipofectamine 2000 (Invitrogen). After 24 h of incubation, the expression of SARM and TRIF protein was determined by immunoblotting. Viability of the SARM-depleted or the TRIF-depleted macrophages was determined to be more than 80% for all samples by the trypan blue staining method.

Quantification of intracellular bacteria.

To determine the intracellular number of viable bacteria, a standard antibiotic protection assay was performed as described previously (7). In brief, at the times indicated, the infected cells were washed twice with 1 ml PBS and the intracellular bacteria were released by lysing the macrophages with 0.1% Triton X-100. The released bacteria were serially diluted with tryptic soy broth, and appropriate dilutions were plated on tryptic soy agar. The number of intracellular bacteria, expressed as CFU, was determined by bacterial colony counting after incubating at 37°C for 48 h.

Immunoblotting.

Cells were lysed in lysis buffer containing 20 mM Tris, 100 mM NaCl, and 1% NP-40. The lysates were separated in 8% SDS-PAGE. Proteins were transferred onto a nitrocellulose membrane (Pall Corporation, Penscola, FL). The nonspecific binding sites on the membrane were blocked with 10% blocking solution (Roche Diagnostics, Mannheim, Germany) for 1 h before proteins were allowed to react with specific primary antibodies against SIRPα, iNOS, or SARM (Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C overnight. Polyclonal rabbit primary antibodies of TRIF (Abcam, Cambridge, MA, United Kingdom) were used for the detection of TRIF protein expression levels. The membrane was washed three times with 0.1% PBS–Tween 20 (PBST) and incubated with corresponding horseradish peroxidase-conjugated secondary antibodies (Pierce, Rockford, IL) for 1 h at room temperature. Thereafter, the membrane was washed four times with 0.1% PBST before a chemiluminescence substrate (Roche Diagnostics) was added and proteins were detected by the enhanced chemiluminescence method.

RT-PCR.

Total RNA was extracted from uninfected and infected cells according to the instructions of the manufacturer (GE Healthcare, Freiburg, Germany) and used for cDNA synthesis with the avian myeloblastosis virus (AMV) reverse transcription (RT) enzyme (Promega, Madison, WI). PCR was then performed using primer pairs specific for the detection of SIRPα, iNOS, IFN-β, tumor necrosis factor alpha (TNF-α), TRIF, and β-actin mRNAs. The primer sequences are shown in Table 1. The amplified products were electrophoresed using a 1.5% or 2.5% agarose gel and stained with ethidium bromide before visualization under an UV lamp.

Table 1.

Nucleotide primer sequences used for PCR amplification

Primer	Sense primer sequence	Antisense primer sequence
Mouse SIRPα	GCA AGG ACC CAG CTC TGT AG	TGG AGC TTT ACC AAC CAA CC
Mouse SARM	GGA GGC TCA GTG CAT AGG AG	CAG GTC TGG ACC TCA GCT TC
Mouse TNF-α	GTA GCC CAC GTC GTA GCA AA	CCC TTC TCC AGC TGG GAG AC
Mouse TRIF	ATG GAT AAC CCA GGG CCT TCG C	AGAA ATG GAG TGG CTG GAA ACC A
Mouse IFN-β	TCC AAG AAA GGA CGA ACA TTC G	TGA GGA CAT CTC CCA CGT CAA
Mouse iNOS	GGT ATC ATT CGG AGC AA	ACA GAG GGA GGG TGG AAT CT
Mouse β-actin	CCA GAG CAA GAG AGG TAT CC	CTG TGG TGG TGA AGC TGT AG

Statistical analysis.

Unless otherwise indicated, all experiments were conducted at least three times. Experimental values were expressed as means ± standard deviations. The statistical significances were analyzed by Student's unpaired t test using Prism 4 software (GraphPad). P values of less than 0.05 were considered significant.

RESULTS

Inhibition of SARM expression leads to decreased SIRPα expression and thereby results in iNOS upregulation and restriction of bacterial multiplication in B. pseudomallei-infected mouse macrophages.

We previously demonstrated that B. pseudomallei could rapidly induce SARM expression in mouse macrophages, leading to the inhibition of the TRIF signaling pathway activation (22). Therefore, it is of interest to know whether the SARM upregulation has any effect on the SIRPα expression in B. pseudomallei-infected mouse macrophages. To determine this, macrophages were infected with wild-type B. pseudomallei at an MOI of 2 after transfecting the cells with SARM siRNAs and the level of SIRPα protein in these cells was then analyzed by immunoblotting. The level of SARM was substantially decreased in SARM-depleted macrophages compared to that of control siRNA-transfected cells (Fig. 1A). The results indicate that the inhibition of SARM expression in B. pseudomallei-infected macrophages was associated with decreased expression of SIRPα protein and a marked upregulation of the iNOS level (Fig. 1A). These results suggest that the upregulation of SARM prevents the decrease of SIRPα expression in B. pseudomallei-infected macrophages, resulting in the inhibition of iNOS expression.

Fig 1.

Inhibition of SARM expression leads to decreased SIRPα expression and thereby results in iNOS upregulation and restriction of bacterial multiplication in B. pseudomallei-infected mouse macrophages. (A) Macrophages (1.5 × 10⁵ cells/well) were transfected with siRNAs against SARM prior to infection with wild-type B. pseudomallei at an MOI of 2. At 6 h of infection, the infected cells were lysed to determine the SARM, SIRPα, and iNOS protein levels by immunoblotting. β-Actin protein was used as an internal loading control. Data are representative of three independent experiments. (B) Macrophages (1.5 × 10⁵ cells/well) were transfected with siRNAs against SARM prior to infection with wild-type B. pseudomallei at an MOI of 2. The SARM siRNA-treated cells and control cells were pretreated without or with 1 mM AG for 1 h before infection, and the internalization of bacteria (at 2 h of infection) and intracellular survival of bacteria (at 6 h of infection) were determined by an antibiotic protection assay. Data indicate the means and standard deviations of the results of three separate experiments, each carried out in duplicate (*, P < 0.05; ns [not significant], P > 0.05).

In order to demonstrate the involvement of iNOS-mediated nitric oxide (NO) production in the control of B. pseudomallei multiplication in SARM-depleted macrophages, the intracellular survival of bacteria was determined in the SARM siRNA-treated cells with or without treatment with aminoguanidine (AG), a known iNOS inhibitor that can significantly suppress NO production by activated macrophages (30). The results show an abrogation of a significant decrease of intracellular survival of the wild-type strain in SARM-depleted cells after AG treatment (Fig. 1B), suggesting that the NO production after iNOS elevation is involved in limiting the multiplication of wild-type bacteria in SARM-depleted cells. It should be noted that SARM depletion with or without AG treatment could not interfere with the internalization of bacteria as judged at 2 h of infection (Fig. 1B), indicating that the observed intracellular survival of the wild-type strain in SARM-depleted and/or AG-treated cells was not due to altered cellular uptake of bacteria by these cells.

Activation of the TRIF signaling pathway is essential for SIRPα downregulation.

Since it has been shown that SARM interacts with TRIF and inhibits the gene expression downstream of TRIF (25), it is tempting to determine further whether the activation of TRIF can contribute to SIRPα downregulation. In order to test this possibility, macrophages were infected with either wild-type or LPS mutant B. pseudomallei at an MOI of 2 after depletion of TRIF by TRIF siRNAs. After 4 and 6 h of infection, cells were analyzed for SIRPα mRNA and protein expression levels by reverse transcription-PCR (RT-PCR) and immunoblotting, respectively. The levels of TRIF mRNA and protein were substantially decreased in TRIF siRNA-treated macrophages compared to those in control siRNA-treated cells (Fig. 2). The results show that the TRIF-depleted LPS mutant-infected macrophages failed to downregulate SIRPα mRNA and protein expression, while equal levels of SIRPα downregulation were observed in untreated and control siRNA-treated infected macrophages (Fig. 2). The inability of LPS mutant to downregulate SIRPα expression in TRIF-depleted macrophages was correlated with a decreased expression of genes downstream of the TRIF signaling pathway but not of those of the MyD88-dependent signaling pathway, such as tumor necrosis factor α (Fig. 2A). However, there was no alteration in the levels of SIRPα mRNA and protein in wild-type B. pseudomallei-infected TRIF-depleted cells, which is consistent with its inability to induce the expression downstream of the TRIF signaling pathway of genes such as inoS and ifn-β (Fig. 2). Altogether, these results suggest that the activation of the TRIF signaling pathway can lead to SIRPα downregulation, resulting in the enhanced expression of genes downstream of this pathway.

Fig 2.

Activation of the TRIF signaling pathway is essential for SIRPα downregulation. (A) Macrophages (1.5 × 10⁵ cells/well) were transfected with siRNAs against TRIF prior to infection with either wild-type or LPS mutant B. pseudomallei at an MOI of 2. The levels of mRNA expression of triF, sirPα, the TRIF-dependent genes (inoS and ifn-β), and the MyD88-dependent gene (tnf-α) were determined by RT-PCR after 4 h of infection. (B) The levels of TRIF, SIRPα, and iNOS protein were analyzed by immunoblotting after 6 h of infection. Data are representative of three independent experiments.

TRIF-mediated SIRPα downregulation is involved in restricting the intracellular multiplication of B. pseudomallei.

It was previously demonstrated that the activation of the TRIF signaling pathway is essential for the control of intracellular bacterial replication, including the replication of B. pseudomallei, in the mouse system (17, 18, 31, 32). To investigate the consequent effect of SIRPα expression following TRIF depletion on the intracellular survival of B. pseudomallei, TRIF-depleted macrophages were infected with either the wild-type or LPS mutant strain at an MOI of 2. After 4, 6, and 8 h of infection, the infected cells were lysed and the numbers of viable intracellular bacteria were determined by an antibiotic protection assay. As shown in Fig. 3A, similar numbers of wild-type bacteria were observed in TRIF-depleted, untreated control, and control siRNA-treated macrophages. In contrast, depletion of TRIF resulted in an increase in the numbers of viable intracellular LPS mutant B. pseudomallei at 6 and 8 h of infection compared to those of untreated control and control siRNA-treated cells (Fig. 3B). The increase in the number of intracellular LPS mutant bacteria was not due to the effect of increased bacterial uptake by TRIF-depleted macrophages, as the similar numbers of intracellular bacteria among untreated control, control siRNA-treated, and TRIF siRNA-treated macrophages were observed at 2 h of infection (Fig. 3B). These results imply that the failure of SIRPα downregulation caused by inhibition of the TRIF signaling pathway is associated with increased bacterial survival in B. pseudomallei-infected macrophages.

Fig 3.

Intracellular survival of B. pseudomallei in TRIF-depleted macrophages. Macrophages (1.5 × 10⁵ cells/well) were transfected with siRNAs against TRIF prior to infection with either wild-type or LPS mutant B. pseudomallei at an MOI of 2. Internalization of bacteria (at 2 h of infection) and intracellular survival of bacteria (at 4, 6, and 8 h of infection) for wild-type (A) or LPS mutant (B) B. pseudomallei in TRIF-depleted macrophages were determined by an antibiotic protection assay. Data indicate the means and standard deviations of the results of three separate experiments, each carried out in duplicate (*, P < 0.05).

IFN-γ-mediated SIRPα downregulation in B. pseudomallei-infected mouse macrophages involves the inhibition of SARM expression but does not require the TRIF signaling pathway.

We recently demonstrated that IFN-γ influences the ability of wild-type B. pseudomallei to downregulate SIRPα expression, leading to the induction of iNOS expression and the suppression of bacterial multiplication in macrophages (21). However, the underlying mechanism behind IFN-γ-mediated SIRPα downregulation has not been determined. As we observed the SIRPα downregulation after inhibition of SARM expression in B. pseudomallei-infected macrophages (Fig. 1A), it is of interest to determine whether IFN-γ-mediated SIRPα downregulation may also involve this possibility. As shown in Fig. 4, IFN-γ stimulation enabled the B. pseudomallei-infected cells to inhibit the SARM expression. This result is consistent with the SIRPα downregulation and the iNOS upregulation observed in IFN-γ-treated infected cells (Fig. 4, lanes 7, 8, and 9). To examine the possible involvement of TRIF signaling in the downregulation of SIRPα within IFN-γ-treated infected cells, IFN-γ was added to macrophage cultures after 8 h of transfection with TRIF siRNAs, and the cultures were incubated together overnight before infection with wild-type B. pseudomallei at an MOI of 2. The results show that, in the absence of IFN-γ, B. pseudomallei alone can induce SARM and inhibit SIRPα downregulation equally in untreated control, TRIF-depleted, and control siRNA-treated macrophages (Fig. 4, lanes 4, 5, and 6). It should be mentioned that the IFN-γ treatment alone did not alter SARM and SIRPα expression either with TRIF depletion or without TRIF depletion (Fig. 4, lanes 1, 2, and 3). However, in the presence of IFN-γ, B. pseudomallei failed to upregulate SARM and to prevent SIRPα downregulation in the TRIF-depleted as well as control siRNA-treated macrophages (Fig. 4, lanes 7, 8, and 9). These results correlate with the elevated iNOS levels observed in all of the IFN-γ-pretreated infected cells (Fig. 4, lanes 7, 8, and 9). Altogether, these data suggest that the IFN-γ can inhibit SARM as well as SIRPα expression, leading to iNOS upregulation occurring independently of the TRIF signaling pathway in B. pseudomallei-infected macrophages.

Fig 4.

IFN-γ-mediated SIRPα downregulation in B. pseudomallei-infected mouse macrophages involves the inhibition of SARM expression but does not require the TRIF signaling pathway. Macrophages (1.5 × 10⁵ cells/well) were transfected with siRNAs against TRIF. After 8 h after transfection, IFN-γ (10 U/ml) was added to the macrophage cultures and left overnight. Cells were then infected with wild-type B. pseudomallei at an MOI of 2. After 6 h of infection, the levels of expression of SIRPα, SARM, and iNOS protein were determined by immunoblotting. Data are representative of three independent experiments.

Purified B. pseudomallei LPS induces SARM expression and fails to downregulate constitutively expressed SIRPα in mouse macrophages.

We previously observed SARM upregulation and inhibition of SIRPα downregulation in heat-killed wild-type B. pseudomallei-treated macrophages (21, 22). Since LPS is a major heat-stable component on the surface of Gram-negative bacteria, including B. pseudomallei, we tested whether the activation of the TRIF signaling pathway might also accompany SIRPα downregulation when the purified B. pseudomallei LPS, typical LPS (from E. coli), and atypical LPS (from P. gingivalis) were used. Macrophages were treated with the LPS from B. pseudomallei, P. gingivalis, or E. coli at a concentration of 100 ng/ml for 8 and 10 h, and the levels of SARM and SIRPα mRNA as well as protein were analyzed. The results show that, in a manner consistent with the use of live bacteria, B. pseudomallei LPS upregulates SARM expression and fails to downregulate constitutively expressed SIRPα (Fig. 5). On the other hand, E. coli LPS is able to inhibit both SARM and SIRPα expression (Fig. 5). Similar to B. pseudomallei LPS, P. gingivalis LPS, which is considered to be less effective in activating TLR4 (33), was also able to upregulate SARM and failed to downregulate SIRPα (Fig. 5). Failure to downregulate SIRPα by B. pseudomallei LPS and P. gingivalis LPS is consistent with their inability to activate the TRIF signaling pathway, as judged by the expression levels of iNOS and genes expressed downstream of TRIF (Fig. 5). In contrast, the observed inhibition of SARM and SIRPα expression by E. coli LPS correlated with its ability to induce the genes that are expressed downstream of the TRIF signaling pathway, such as inoS and ifn-β (Fig. 5A). Altogether, our results suggest that the features of SARM upregulation and SIRPα downregulation failure observed in B. pseudomallei-infected mouse macrophages may be due to the atypical-like LPS of the bacterium, which can be involved in inhibition of the TRIF signaling pathway during infection.

Fig 5.

Purified B. pseudomallei LPS induces SARM expression and fails to downregulate constitutively expressed SIRPα in mouse macrophages. Macrophages (1 × 10⁶ cells/well) were treated with B. pseudomallei LPS, P. gingivalis LPS, or E. coli LPS at a concentration of 100 ng/ml for 8 and 10 h. The levels of expression of SIRPα, SARM, and iNOS mRNAs (A) and proteins (B) were determined as described above. Data are representative of three independent experiments.

DISCUSSION

The importance of TLR-mediated control of B. pseudomallei infection has been demonstrated by several studies (18, 21, 34–36). However, how this pathogen regulates the TLR signaling response for its own survival in the host is largely unknown. A study by Wiersinga et al. demonstrated the importance of the MyD88-dependent pathway in the early control of B. pseudomallei infection in mice by showing the involvement of this pathway in the recruitment and activation of neutrophils at the primary site of infection (34). The essential role of the TRIF signaling pathway in controlling the intracellular multiplication of B. pseudomallei in mouse macrophages has been demonstrated by our previous studies (18, 21). We demonstrated that the virulent B. pseudomallei has the ability to suppress the TRIF signaling pathway by modulating the infected cells to upregulate SARM and to inhibit rapid SIRPα downregulation (21, 22). In contrast, the less virulent LPS mutant is unable to modulate the macrophages to express these two negative regulators, so that the signals after bacterial recognition by TLR4 can be transduced to downstream signaling molecules, leading to increased expression of an iNOS enzyme (21, 22). This enzyme contributes to the bacterial elimination by stimulating the infected macrophages for the production of microbicidal NO (37). The essential role of iNOS in the elimination of invading bacteria has been observed not only for B. pseudomallei but also for other intracellular pathogens, including Mycobacterium tuberculosis (38), Toxoplasma gondii (39), Leishmania donovani (40), and Salmonella enterica serovar Typhimurium (41). Therefore, the mechanisms involved in the regulation of iNOS expression level are important for understanding the evasive strategies used by intracellular pathogens, including B. pseudomallei, for intracellular survival in macrophages.

Although we had previously demonstrated distinct differences in regulation of SARM and SIRPα expression in wild-type- and LPS mutant-infected mouse macrophages, the underlying mechanisms by which these two isogenic strains are able to differentially regulate the levels of expression of these two proteins have not been identified (21, 22). The results presented here demonstrate that the inability of the wild-type bacterium to downregulate constitutively expressed SIRPα is due to its ability to induce SARM expression in infected macrophages. As shown in Fig. 1A, the SARM-depleted macrophages infected with wild-type bacteria were observed with a marked decrease in SIRPα expression and consequent upregulation of iNOS levels compared to those observed in untreated control and control siRNA-treated cells (Fig. 1A). Moreover, the abrogation of restricted multiplication of wild-type bacteria after AG treatment in the SARM-depleted wild-type-infected cells suggests the involvement of NO production in the control of bacterial multiplication after SARM depletion (Fig. 1B). It should be mentioned that the NO production level by SARM siRNA-treated wild-type-infected cells at up to 8 h of infection was observed to be below the limits of detection by a Griess assay (data not shown), suggesting that the lesser NO production in these SARM siRNA-treated cells could play a bacteriostatic role to restrict the multiplication of intracellular B. pseudomallei. Altogether, these results indicate the importance of SARM in the regulation of SIRPα and iNOS expression during B. pseudomallei infection.

Since it has been demonstrated that expression of SARM inhibits TRIF signaling pathway in LPS-stimulated human monocytes (42) or in B. pseudomallei-infected mouse macrophages (22), we further investigated the role of this pathway in SIRPα downregulation during B. pseudomallei infection. As expected, downregulation of SIRPα was abrogated during LPS mutant infection when the activation of this pathway was prevented in TRIF-depleted macrophages and this was accompanied by the disappearance of iNOS expression (Fig. 2). These results imply that the regulation of SIRPα expression in macrophages is associated with the TRIF signaling pathway. Consistent with our previous findings, SIRPα downregulation was observed and found to be associated with the activation of this pathway, as judged by the increased expression of signature genes of this pathway and the iNOS upregulation, among untreated control and control siRNA-treated macrophages that were infected with the LPS mutant (Fig. 2). The reduced level of iNOS expression observed in the LPS mutant-infected TRIF-depleted macrophages correlated with the significant increase in the intracellular survival of LPS mutant bacteria compared to those observed in the nondepleted macrophages (Fig. 3B). These findings regarding the SIRPα involvement in impairing the elimination of LPS mutant B. pseudomallei in TRIF-depleted cells is consistent with our previous study in which wild-type bacteria were used in SIRPα-depleted macrophages (21).

Although a previous study pointed out the essential role of TLR4 signaling in SIRPα downregulation required for macrophage activation in response to LPS stimulation, the present study further determined the direct involvement of TLR4-TRIF signaling pathway to be responsible for SIRPα downregulation (27). Moreover, the inability of wild-type bacterium with respect to the activation of TRIF signaling was associated with its failure in downregulating SIRPα expression and with its ability to survive and multiply inside macrophages (Fig. 2 and 3A). The observed absence of a role of the TRIF signaling pathway in control of wild-type B. pseudomallei replication is consistent with a previous in vivo study that demonstrated similar patterns of animal survival and bacterial outgrowth in wild-type and TRIF mutant mice following intranasal B. pseudomallei (strain 1026b) infection (34), suggesting the possible involvement of these two negative regulators in attenuation of this pathway in mice with experimentally induced melioidosis. Overall, our results suggest an essential role of the SARM in the inhibition of TRIF-dependent SIRPα downregulation during B. pseudomallei infection and that this mechanism may contribute to this bacterium for the evasion of macrophage killing.

It has long been known that IFN-γ plays a pivotal role in the control of B. pseudomallei infection in murine models as well as in in vitro systems (43, 44). However, the underlying molecular mechanism by which IFN-γ can provide protection to the host against this bacterium is largely unknown. In this regard, our previous studies showed that IFN-γ-mediated rapid SIRPα downregulation could contribute to the induction of iNOS expression, leading to an restricted intracellular survival of the bacteria, in B. pseudomallei-infected IFN-γ-activated macrophages via an unknown mechanism (21). In the present study, we found a distinct correlation of SIRPα downregulation with the inhibition of SARM expression, suggesting that the attenuation of SARM expression may serve as a mechanism used by B. pseudomallei-infected IFN-γ-activated macrophages to downregulate SIRPα required for iNOS expression. It should be noted that IFN-γ in the absence of bacteria did not alter the expression level of either SIRPα and SARM (Fig. 4), indicating that the signal generated after bacterial recognition is essential for the IFN-γ-mediated inhibition of the expression of these two negative regulators. Unexpectedly, IFN-γ-mediated SIRPα downregulation was not abrogated when TRIF was depleted prior to infection with B. pseudomallei, suggesting that IFN-γ could inhibit both SARM and SIRPα expression even in the absence of TRIF (Fig. 4). Moreover, the observed IFN-γ-mediated TRIF-independent downregulation of SIRPα and upregulation of iNOS are consistent with the results showing reduced intracellular survival of the bacteria in TRIF-depleted IFN-γ-activated macrophages (data not shown). These findings regarding the involvement of the TRIF-independent pathway in the expression of mediators downstream of TRIF, such as iNOS, in IFN-γ-stimulated macrophages are similar to those of a previous study that demonstrated increased expression of iNOS after depletion of TRAF family member-associated NF-κB activator (TANK) binding kinase 1 (TBK-1) protein (a signaling molecule downstream of TRIF) in IFN-γ-stimulated macrophages that were infected with B. pseudomallei (18). It is possible that intracellularly replicating B. pseudomallei or its components could be involved in the induction of iNOS expression via a TLR-independent mechanism in IFN-γ-stimulated macrophages (45).

Previous studies demonstrated that upregulation of TRIF-dependent genes in LPS- and polyriboinosinic:polyribocytidylic acid [(poly(I·C)]-treated mouse macrophages involves the downregulation of constitutively expressed SIRPα (26, 27). Therefore, we attempted to examine the involvement of the TRIF signaling pathway in SIRPα downregulation using LPS isolated from virulent B. pseudomallei. The results show that B. pseudomallei LPS is able to upregulate SARM expression and inhibits SIRPα downregulation, resulting in the inhibition of gene expression downstream of the TRIF signaling pathway (Fig. 5). In contrast, E. coli LPS-treated macrophages exhibited a noticeable inhibition of SARM expression and downregulation of SIRPα; hence, these cells showed an upregulated gene expression profile of the TRIF signaling pathway (Fig. 5). However, P. gingivalis LPS, an atypical type of LPS that is known to activate macrophages via TLR2 (46), demonstrated results similar to those seen with B. pseudomallei LPS with regard to the ability to induce SARM and prevent SIRPα downregulation (Fig. 5), implying that the ability of B. pseudomallei to regulate these two negative regulators for the evasion of macrophage killing may be at least dependent on the atypical nature of its LPS. It should be noted that typical and atypical types of LPS share a general structure of LPS; however, the atypical LPS differs from the typical type by bearing the substituents in the phosphate group and the variable and long fatty-acid chain in lipid A (47). Due to such features, it has been previously demonstrated that the atypical LPS is recognized and responded to by the TLRs in a manner different from that seen with the typical LPS (46). It is still uncertain how atypical P. gingivalis LPS as well as B. pseudomallei LPS is able to induce SARM expression in the macrophages. It is possible that these two LPS may be preferentially recognized by TLR2 or another receptor(s) that might lead to the induction of SARM expression. Such possibilities are under investigation by our group.

In summary, it can be concluded that virulent B. pseudomallei-induced SARM expression prevents the TRIF-dependent downregulation of constitutively expressed SIRPα in mouse macrophages, resulting in the lack of iNOS expression by B. pseudomallei-infected macrophages. This mechanism may contribute to the ability of this pathogen to persist and multiply inside the macrophage.