Role of PKCtheta in macrophage-mediated immune response to Salmonella typhimurium infection in mice

Background The serine/threonine protein kinase C (PKC) theta has been firmly implicated in T cell-mediated immunity. Because its role in macrophages has remained undefined, we employed PKCtheta-deficient (PKCtheta−/−) mice in order to investigate if PKCtheta plays a role in macrophage-mediated immune responses during bacterial infections. Results Our results demonstrate that PKCtheta plays an important role in host defense against the Gram-negative, intracellular bacterium Salmonella typhimurium, as reflected both by markedly decreased survival and a significantly enhanced number of bacteria in spleen and liver of PKCtheta−/− mice, when compared to wild-type mice. Of note, albeit macrophages do not express detectable PKCtheta, PKCtheta mRNA expression was found to be profoundly upregulated during the first hours of lipopolysaccharide (LPS)/interferon-gamma (IFNgamma)-, but not IL-4-mediated cell polarization conditions in vitro. Mechanistically, despite expressing normal levels of classically activated macrophage (CAM) markers, PKCtheta-deficient CAMs expressed significantly higher levels of the anti-inflammatory cytokine IL-10 in vivo and in vitro when challenged with S. typhimurium or LPS/IFNgamma. Neutralization of IL-10 recovered immune control to S. typhimurium infection in PKCtheta-deficient macrophages. Conclusions Taken together, our data provide genetic evidence that PKCtheta promotes a potent pro-inflammatory CAM phenotype that is instrumental to mounting protective anti-bacterial immunity. Mechanistically, PKCtheta exerts a host-protective role against S. typhimurium infection, and acts as an essential link between TLR4/IFNgammaR signaling and selective suppression of the anti-inflammatory cytokine IL-10 at the onset of CAM differentiation in the course of a bacterial infection.


Background
Macrophages (Mφ) play important roles in inflammatory and infectious diseases. Mφ polarization phenotypes, however, are heterogeneous and profoundly affected by local factors within the microenvironment. Microbial stimuli such as lipopolysaccharide (LPS), interleukin-1beta (IL-1beta), and cytokines secreted by Th1 lymphocytes, such as interferon-gamma (IFNgamma) induce a "classic" Mφ phenotype (CAM). Activated CAMs are pro-inflammatory Mφ and produce high levels of pro-inflammatory cytokines like tumor necrosis factor-alpha (TNF-alpha), IL-6, IL-1beta, IL-23 and IL-12. CAMs are essential components of anti-microbial host defense and further characterized by production of reactive oxygen species and reactive nitrogen species during the promotion of a Th1-type driven response [1][2][3]. Because these pro-inflammatory CAMs may cause extensive tissue damage, their activation is tightly controlled. Indeed, CAMs are established to be involved in autoimmune diseases such as systemic lupus erythematosus and rheumatoid arthritis [4]. In contrast, Th2 lymphocyte cytokines such as IL-4 and IL-13 promote the alternatively activated Mφ phenotype (AAM) that dampens the inflammatory state by producing anti-inflammatory mediators such as IL-10. Signature genes for AAMs are Arginase-1, chitinase-like molecules YM1 and YM2, and resistin-like molecule Fizz-1 (Relm-alpha, Retnla) [5,6]. AAMs, in particular, produce high amounts of the anti-inflammatory cytokines IL-10 and Tumor growth factor beta (TGFβ), thereby inhibiting pro-inflammatory immune responses. Thus, AAMs are involved in tissue remodelling [7] and tumor progression [8].
Protein kinase C (PKC) isotypes are members of the serine/threonine protein kinase subfamily, and play an important role in the regulation of a variety of cell functions. PKCtheta (PKCtheta), a member of the nPKC subfamily, is predominantly expressed in T cells. Most of our current knowledge about PKCtheta therefore arises from studies performed in T cells. We and others could show that PKCtheta plays a critical role in the NF-kappaB, AP-1, and Ca 2+ /NFAT pathways to activate e.g. the interleukin-2 cytokine promoter [9,10]. Madaro et al. demonstrated that PKCtheta has an immune cell intrinsic role in muscle tissue repair during muscular dystrophy inflammation [11]. More recently, Ma et al.
reported a role for PKCtheta in cholesterol metabolism in human Mφ [12]. Because different PKCs, namely PKCalpha, PKCbeta, PKCdelta and PKCzeta, have been shown to have a critical role in Mφ antimicrobial immune responses [13][14][15][16], we investigated the functional role of PKCtheta in Mφ biology. Here we demonstrate that the genetic ablation of PKCtheta expression in mice leads to exacerbation of disease progression and early death in Salmonella (S.) typhimurium infection.

Results
PKCtheta mRNA expression is selectively induced in LPS/ IFNgamma polarizing conditions PKCtheta has been reported to regulate the expression of various genes in T cells [17]. To examine whether PKCtheta modulates Mφ polarization in pro-or antiinflammatory milieu conditions, we analysed mRNA expression patterns of PKCtheta in bone marrow-derived macrophages (BMDM) polarized with lipopolysaccharide (LPS)/interferon-gamma (IFNgamma) or with IL-4, respectively ( Fig. 1a). For proper polarization, marker gene expression of iNOS (for CAM) and Fizz-1 (for AAM) was monitored (Fig. 1b + c). As a remarkable result, PKCtheta expression was strongly induced in LPS/  Fig. 1 PKCtheta mRNA levels are transiently induced in an LPS/IFNgamma-dependent manner in both mouse and human macrophages: a BMDMs were differentiated and polarized to CAMs (LPS/IFNgamma) and AAMs (IL-4). PKCtheta gene expression was determined by qRT-PCR after different time points and was found to be selectively expressed under CAM polarization. Data shown are derived from at least three independent experiments and qRT-PCR data were normalized to Gapdh. Error bars denote the mean ± s.e.m. *p < 0.05; ** p < 0.01; Expression of unstimulated macrophages was arbitrarily set to 1. b + c In order to control proper polarization of macrophages, gene expression of prominent markers (iNOS for CAM, Fizz-1 for AAM) were validated. d CD14+ human monocytes from peripheral blood were differentiated and polarized with LPS/IFNgamma. After predetermined time periods, gene expression of PKCtheta was analysed with qRT-PCR, revealing an increase in PKCtheta mRNA levels upon stimulation in the human setting, consistent with results from mouse experiments. Data shown are relative to expression of unstimulated macrophages, which was arbitrarily set to 1. All data were normalized to Gapdh IFNgamma-polarized but not in IL-4-polarized cells. Of note, PKCtheta mRNA up-regulation in CAMs peaked at 4 h. These results suggest a subset-selective role for PKCtheta as "signaling intermediate" of pro-inflammatory Mφ. Reminiscent to mouse cells, PKCtheta is inducibly expressed in human CD14+ monocytes polarized with LPS/IFNgamma, indicating a possible role of PKCtheta also in fine-tuning of the cellular phenotype of human cells (Fig. 1d).
PKCtheta is causally involved in the protection from peritoneal S. typhimurium infection Based on significant PKCtheta expression level in CAMs, we investigated its potential role in bacterial killing, employing infection assays with intracellular Gramnegative S. typhimurium bacteria in vivo and ex vivo. Evaluation of wild-type and PKCtheta −/− mice during systemic infection showed that all of the PKCtheta −/− mice died within 16 days whereas a minimum of 50 % wild-type mice survived, suggesting a critical role for PKCtheta in the host defense against S. typhimurium (Fig. 2a). As PKCtheta −/− mice started to die around day 4-5, we monitored the bacterial load in spleen and liver at day 4 which revealed a significant increase in the number of colonyforming units (cfu) of S. typhimurium in PKCtheta −/− mice (Fig. 2b). These data clearly establish an essential and non-redundant positive role of PKCtheta in protection from S. typhimurium. during S. typhimurium infection, we depleted CD4 + and CD8 + T cells with antibodies before injection of S. typhimurium (Fig. 2c). Depletion antibodies were re-injected every 3 days during the course of the experiment. Isotype-matched antibodies were injected as depletion control and effective depletion was controlled by FACS analysis (Fig. 2d). The results showed no significant difference in survival of mice injected either with CD4 and CD8 depletion antibodies or isotype control, indicating that T cells are not primarily responsible for the reduced survival rate of PKCtheta −/− mice. These data suggest that PKCtheta −/− mice fail to mount the appropriate innate immune responses after infection with S. typhimurium.

PKCtheta plays a crucial role in repression of IL-10 production in vitro and in vivo
In T cells, PKCtheta is known to affect a wide range of transcription factors such as AP-1, NF-kappaB and NFAT, regulating i.e. IL-2 activation responses. Additionally, PKCtheta influences the transcription of miRNA clusters [18] and, in turn, controls expression of many other genes. To investigate a PKCtheta signaling function upstream of cytokine responses in Mφ, LPS/ IFNgamma-polarized BMDMs from wild-type and PKCtheta −/− mice were monitored for pro-and antiinflammatory cytokine production at different time points. Although we could not detect any differences in the production of IL-6, IL-12p40, TNF-alpha as well as iNOS expression (data not shown) in polarized PKCtheta −/− CAMs, we found a significantly increased production of the anti-inflammatory cytokine IL-10 in vitro (Fig. 3a). Of, note, the anti-inflammatory cytokine TGFβ response of polarized PKCtheta −/− CAMs, remained unchanged both at the protein and the mRNA levels ( Fig. 3b and data not shown), indicating a selectivity of PKCtheta signaling function. Mechanistically and albeit our results do not rule out effects of PKCtheta deficiency on TLR4 signaling responses in general, PKCtheta does not regulate TLR4 signaling leading to IRF3 activation response (Fig. 3c). In agreement with this observation, the significantly lowered survival rates of PKCtheta −/− mice during peritoneal S. typhimurium infection correlated with significantly higher IL-10 serum levels (Fig. 3d). Additionally, after injecting LPS intraperitoneally, the anti-inflammatory IL-10 serum response in PKCtheta-deficient mice was profoundly increased after 2 to 24 h (Fig. 3e). This further confirms the increased IL-10 production by myeloid cells in another in vivo model. Thus, although Mφ from either PKCtheta −/− or wild-type mice were polarized into CAMs, only wild-type Mφ but not PKCtheta-deficient Mφ mounted a functional anti-bacterial activity.

PKCtheta-mediated IL-10 repression affects bacterial killing of macrophages in vitro
Because containment and killing of S. typhimurium after invasion of Mφ is a key event in host protection, we monitored intracellular bacterial loads of S. typhimurium-infected Mφ of both genotypes, reproducibly demonstrating increased bacterial numbers in PKCtheta −/− CAMs as compared to wild-type macrophages (Fig. 4a). In order to investigate a direct role of IL-10 hyper-production for the reduced killing capacity of Salmonella-infected PKCtheta −/− Mφ, IL-10 was neutralized with IL-10 blocking-or isotype control antibodies. We observed that IL-10 blocking antibody treatment rescued the killing capacity of PKCthetadeficient cells back to the wild-type level (Fig. 4a). Similar to genetic PKCtheta depletion, pharmacological PKCtheta inhibition by the PKC-selective inhibitor AEB071 (Sotrastaurin) [19] significantly reduced the killing capacity of S. typhimurium-infected wild-type Mφ (Fig. 4b). Taken together, these results suggest that PKCtheta is critically involved in bacterial killing of S. typhimurium-infected Mφ by suppressing IL-10 production.

Discussion
S. typhimurium is a Gram-negative, motile, facultative intracellular bacterium, which invades and multiplies within mononuclear phagocytic cells in liver, spleen, lymph nodes, and Peyer's plaques [20,21]. S. typhimurium causes severe gastrointestinal disorders in humans and typhoid fever with systemic infections in mice. Mφ functions have been strongly implicated in infectious diseases. Mφ, as one of the first barriers of the innate immune system, rapidly control S. typhimurium. Importantly, however, these bacteria may evade immune control and even multiply within Mφ by mechanisms that are insufficiently understood so far [22,23]. The immune response to bacterial infection is regulated by the counter-play between different Mφ subsets, namely CAMs and AAMs. Changes in the balance of CAM-and AAM-subsets and their effector cytokines result in alterations of disease progression. Recognition of Gram-negative bacteria in Mφ involves the binding of LPS to TLR4, leading to secretion of key cytokines [24,25]. Furthermore, Mφ produce ROS by the phagocyte NADPH oxidase (NOX2; phox) and RNS by the inducible nitric oxide synthase (iNOS; NOS2) to combat bacteria. To counteract chronic pro-inflammatory immune effector mechanisms, the production of antiinflammatory proteins IL-10 and TGFbeta by Mφ is induced [26,27]. NO, generated by IFNgamma-induced iNOS, is shut down with a shift to Arg1, and YM1 and YM2 are strongly induced, subsequently leading to the induction of a Th2 immune response [28].
IL-10 plays an essential part in dampening inflammation [29]. Consequently, dysregulation of IL-10 is linked with susceptibility and an impaired clinical course of numerous infections in mouse models and in humans [30][31][32]. Mechanistically, distinct signaling and epigenetic chromatin remodelling of the IL10 locus control the production of IL-10. Furthermore, post-transcriptional events through e.g. microRNA represent a critical step for IL-10 transcript stability.
In response to LPS, TLR4 mediated signaling utilizes IRF3 to induce an IL-10 transcription response. Nevertheless, a number of additional transcription factors including signal transducers and activators of transcription (STAT), activator protein (AP), cAMP response element binding protein (CREB), CCATT enhancer/binding protein (C/EBP), c-musculoaponeurotic fibrosarcoma factor (c-MAF), and nuclear factor kappa B (NF-kappaB), have been characterized as essential or critical in myeloid cell-type-specific IL-10 gene regulation. Thus a comprehensive understanding of the detailed cellular processes of PKCtheta that contribute in a CAM cellspecific manner to IL-10 transcriptional regulation remains elusive. Albeit we cannot define the exact signaling function of PKCtheta upstream of Il10 transcriptional regulation, the results are in a line with our observation of increased IL-10 formation in PKCtheta-deficient BMDMs and that neutralization of IL-10 results in an improved control of bacterial proliferation that is indistinguishable between wild-type and PKCtheta-deficient macrophages.

Conclusions
We demonstrate that PKCtheta contributes to a hostprotective immune response against peritoneal S. typhimurium infection by its selective involvement during early (up to 24 h) pro-inflammatory CAM polarization. This time frame is in agreement with the fact that in these cells PKCtheta is induced immediately and transiently after LPS/IFNgamma engagement. PKCtheta deficiency, as defined in this study, leads to an altered CAM phenotype, based on an incomplete suppression of the macrophage deactivating cytokine IL-10. The detailed molecular basis of PKCtheta-dependent regulation of IL-10 production under CAM-promoting conditions, and thus the phenotypic instability of PKCtheta −/− CAMs, however, remains unknown. Nevertheless, the data define a previously unknown regulatory role of PKCtheta in macrophage driven innate defense mechanisms against bacterial infections. As an innovative paradigm, CAM-intrinsic PKCtheta may represent a transiently inducible signaling intermediate leading to IL-10 suppression, and subsequently, to an effective anti-bacterial immunity against intracellular bacteria such as S. typhimurium. This hypothesis is based on the increased IL-10 expression observed during the PKCtheta −/− CAM-priming phase. Of note, this deregulated IL-10 expression level observed in PKCtheta −/− CAMs is directly responsible for their bacterial killing defects after infection with S. typhimurium.
In summary, the findings of this study provide genetic evidence that PKCtheta is an early factor involved in developing a stable CAM-mediated immune response in vitro and in vivo. Our results provide first experimental evidence for an involvement of PKCtheta in Mφ biology, a finding relevant for the understanding of PKCtheta in innate antimicrobial immune effector function.

Gene expression analysis
Total RNA was isolated using the RNeasy® Mini Kit (Qiagen) according to the manufacturer's instructions. cDNA was synthesized using the Qiagen Omniscript RT Kit and oligo(dT) primers (Promega). Expression analysis was performed with real-time PCR using TaqMan technology (assays from Applied Biosystem). Reactions were run with Applied Biosystems 7500 Fast Sequence

In vivo T cell depletion
Mice were injected with 500 μg of anti-mouse CD4 (clone GK1.5; BE003-1) and anti-mouse CD8 antibody (clone YTS 169.4; BE0117), or the corresponding IgG2b (clone LTF-2; BE0090) control (all from BioXCell, USA) 1 day prior to infection with S. typhimurium. Facial vein blood was taken after 24 h to control T cell depletion by FACS analysis using antiCD4-FITC, antiCD8-APC, and antiCD3-PE (all from eBioscience). Thereafter, 5,000 to 30,000 cfu S. typhimurium were injected intraperitoneally. 300 μg of depletion antibodies or isotype control were administered every third day throughout the entire experiment.

Cytokine measurements
IL-10 and TGFbeta cytokine levels from from serum or cell culture supernatants were analyzed with Bio-Plex multianalyte technology (BioRad).

Endotoxin-induced infection model
WT and PKCtheta −/− mice were treated with intraperitoneal injection of 15 mg LPS (lipopolysaccharide from S. typhimurium L6511; Sigma-Aldrich) per kg of body weight. Serum was collected 4 h after LPS challenge for analysis of IL-10.

Salmonella infection in vitro
Macrophages were isolated as described above and incubated in complete DMEM without antibiotics. Wild-type strain S. typhimurium was cultured in LB broth to latelogarithmic phase. Mφ were infected with S. typhimurium at a multiplicity of infection (MOI) of 10 for 1 h. Thereafter, cells were washed with PBS and incubated in complete DMEM containing gentamycin (Gibco). For killing experiments, prior to infection, cells were treated with a monoclonal rat anti-mouse IL-10 antibody (10 μg/ml; clone JES5-2A5; 504903; BioLegend) or the appropriate isotype control (10 μg/ml; clone RTK2071; 400413; BioLegend) for 23 h. AEB071 (sotrastaurin) was used as PKC inhibitor in killing experiments and cells were treated with 1 μM for 23 h. After infection, intracellular bacterial loads were harvested with 0.5 % sodium deoxycholic acid (D6750; Sigma-Aldrich) as described previously [33].