Cross-Talk between TLR4 and FcγReceptorIII (CD16) Pathways

Pathogen-pattern-recognition by Toll-like receptors (TLRs) and pathogen clearance after immune complex formation via engagement with Fc receptors (FcRs) represent central mechanisms that trigger the immune and inflammatory responses. In the present study, a linkage between TLR4 and FcγR was evaluated in vitro and in vivo. Most strikingly, in vitro activation of phagocytes by IgG immune complexes (IgGIC) resulted in an association of TLR4 with FcγRIII (CD16) based on co-immunoprecipitation analyses. Neutrophils and macrophages from TLR4 mutant (mut) mice were unresponsive to either lipopolysaccharide (LPS) or IgGIC in vitro, as determined by cytokine production. This phenomenon was accompanied by the inability to phosphorylate tyrosine residues within immunoreceptor tyrosine-based activation motifs (ITAMs) of the FcRγ-subunit. To transfer these findings in vivo, two different models of acute lung injury (ALI) induced by intratracheal administration of either LPS or IgGIC were employed. As expected, LPS-induced ALI was abolished in TLR4 mut and TLR4−/− mice. Unexpectedly, TLR4 mut and TLR4−/− mice were also resistant to development of ALI following IgGIC deposition in the lungs. In conclusion, our findings suggest that TLR4 and FcγRIII pathways are structurally and functionally connected at the receptor level and that TLR4 is indispensable for FcγRIII signaling via FcRγ-subunit activation.


Introduction
The immune system is traditionally divided into innate and adaptive entities. Adaptive immunity is organized around T cells and B cells and requires a process of maturation and clonal selection of lymphocytes. In contrast, innate immunity can be immediately activated during the onset of infection in order to control replication of pathogenic microbes and bring about their clearance from tissues or blood. As an important aspect of innate immunity, pattern-recognition receptors (PRRs) collectively recognize lipid, carbohydrate, peptide, and nucleic-acid structures of invading microorganisms [1]. PRRs comprise the toll-like receptor family (TLR), which consists of at least 12 different evolutionarily conserved membrane proteins that trigger innate immune responses [2]. Initially identified in 1997, TLR4 represents the most thoroughly investigated TLR [3]. TLR4 is essential for responses to bacterial lipopolysaccharide (LPS), a well-known pathogen-associated molecular pattern (PAMP) [3,4]. Besides LPS, various endogenous ligands, such as hyaluronan and high mobility group box 1 protein (HMGB1), appear to engage TLR4 [5,6]. After binding of LPS to the TLR4/MD-2/CD14 receptor complex, activation of the intracellular signaling pathway is initiated, ultimately leading to NF-kB activation and its translocation to the nucleus, resulting in subsequent cytokine/chemokine production and release [7].
As part of the adaptive immune system, antibodies of high affinity binding specifically recognize and neutralize intruding pathogens or their products. After antibody binding to antigen, the Fc domain of immunoglobulin (Ig) is recognized by Fc receptors (FcRs) which are predominantly expressed on immune and inflammatory cells and thereby link antibody-mediated (humoral) immune responses to cellular effector functions [8,9]. Specific FcRs exist for all classes of immunoglobulins. Binding of IgGs to FccRs on phagocytes triggers a wide variety of cellular functions including phagocytosis, release of inflammatory mediators, and clearance of immune complexes [8]. FccRs specifically bind IgG and are divided into four subclasses. FccRI (CD64), FccRIII (CD16), and FccRIV are activating receptors, while FccRII (CD32) mediates inhibitory functions. The cellular response is determined by the balance between activating (ITAM, immunoreceptor tyrosine-based activation motif) and inhibitory (ITIM, immunoreceptor tyrosine-based inhibitory motif) signals [10,11,12,13].
Despite extensive research in the past, the highly complex regulation of innate and adaptive immunity and their interactions are still poorly understood. It has been suggested that adaptive immune responses are controlled by innate immune recognition and vice versa [14,15,16]. In particular, TLRs and FccRs are considered to be important regulators of immune responses [13,17]. Recently, evidence has emerged that there is indirect interaction between TLR4 and FccR pathways. TLR4 has been shown to up-regulate FccR expression in experimental immune complex arthritis; inhibition of TLR4 resulted in attenuation of in vivo cytokine release in models of glomerulonephritis and rheumatoid arthritis [18,19,20]. In the present study, we addressed the question as to whether there is a direct link between TLR4 and FccR pathways in vitro and in vivo.

Exclusion of LPS Contamination of Reagents
In the past, the investigation of TLR4 faced the problem of LPS contamination, which imposed considerable restrictions on the interpretation of data [5]. Therefore, the LPS concentration was determined in reagents used for lung injury induction by deposition of IgG immune complexes (IgGIC), such as DPBS, anti-BSA IgG and BSA, although none of these reagents had been prepared using bacterial (E.coli) systems. Using Limulus Amebocyte Lysate Kinetic-QCL assay, LPS levels were not detectable (,5610 23 units/ml) in any of the reagents (data not shown), suggesting that in vitro stimulation by IgGIC is based upon a genuine agonist effect that is not due to LPS contamination. In addition to determination of LPS contamination (see above), DPBS, anti-BSA IgG and BSA were subjected to endotoxin removal by solid-phase polymyxin. Using the polymyxin-treated reagents, immune complexes were generated and then applied in in vitro experiments or the reagents were administered in mice for the formation of immune complexes in vivo. Furthermore, commercially available, preformed peroxidase/anti-peroxidase immune complexes (PAP IgGIC) were used at the same concentration in order to confirm the results using BSA IgGIC or polymyxin-treated BSA IgGIC. The results of both, polymyxintreated BSA IgGIC and PAP IgGIC, are presented in the corresponding figures. In summary, using different in vitro and in vivo approaches, it is highly unlikely that any of the effects following IgGIC stimulation in the present study are based on LPS contamination of the reagents.

Association between TLR4 and FccRIII after IgG Immune Complex Activation
In order to assess whether crosstalk between TLR4 and FccR might occur at the receptor level, neutrophils (PMNs) and macrophages from wild-type (Wt) mice were incubated in vitro with IgGIC, LPS, or the combination of the two. After incubation, cell lysates were immunoprecipitated (IP) with anti-TLR4 and then analyzed for FccRII/III by immunoblotting (IB). As shown in Figure 1A,B, immunoprecipitated TLR4 was associated with FccR after cell exposure to IgGIC. Inversely, LPS incubation did not result in an association of both receptors as indicated by the absence of bands for FccR, whereas the combination of LPS+IgGIC seemed to enhance the signal for FccR coimmunoprecipitated by anti-TLR4 IgG ( Figure 1A,B). The band for FccR under the conditions described above indicated a protein mass of 55 kDa, in accord with the reported molecular weight for FccRIII [21,22]. In contrast, there was no band at the 40 kDa position (data not shown), the molecular weight of FccRII, which is also recognized by the anti-FccR antibody (mAb, clone 2.4G2) used for Western blot analyses [23,24]. In accord with Figure 1A,B, reverse direction immunoprecipitation using FccRIII antibody followed TLR4 Western blots revealed bands at around 90 kDa, consistent with the reported molecular weight of TLR4 ( Figure 1C,D). However, under these conditions bands also occurred after stimulation of phagocytes with LPS ( Figure 1C,D), which may suggest that FccRIII and TLR4 heterodimerize upon LPS stimulation, although to a lesser extent as compared to IgGIC treated cells. When PMNs and macrophages from FccRIII 2/2 mice were exposed to the same in vitro conditions (IgGIC, LPS, LPS+IgGIC), the band for FccRIII failed to appear, confirming its specificity ( Figure 1E,F). In order to examine whether the interaction between TLR4 and FccRIII was specific for these two receptors or whether there also might be multimerization with other TLRs or Fc receptors, lysates from Wt phagocytic cells under various conditions (see above) were subjected to immunoprecipitation with anti-TLR6 or anti-CD23 (anti-FceRII), followed by Western blots for FccRIII or TLR4, respectively ( Figure 1G-J). In both combinations, specific bands for either FccRIII (after immunoprecipitation with anti-TLR6; Figure 1G,H) or TLR4 (immunoprecipitation of cell lysates with anti-TLR6; Figure 1I,J) failed to appear, whereas the strong bands in the lower panels (loading controls) demonstrate that immunoprecipitation of the samples worked properly. In addition, macrophages from Wt mice were incubated with polymyxin-treated BSA IgGIC and PAP IgGIC, followed by immunoprecipitation with anti-TLR4 and Western blotting with anti-FccRIII. As shown in Figure 1K, receptor heterodimerization occurred under these conditions as well, confirming the results shown in Figure 1A,B.
In summary, these findings indicate that association of TLR4 and FccRIII occurs following activation of phagocytes with IgGIC and/or LPS and that this receptor association is a specific phenomenon for FccRIII and TLR4.
Attenuated In Vitro Cytokine Production by TLR4 Mutant PMNs and Macrophages Following IgGIC or LPS Exposure Elicited peritoneal neutrophils (PMNs) and macrophages were obtained from Wt and TLR4 mut mice. The cells were incubated in vitro with IgGIC or LPS. Subsequently, supernatant fluids were collected and evaluated by ELISA for intereukin-6 (IL-6) and tumour necrosis factor alpha (TNFa) levels ( Figure 2). PMNs from Wt mice showed significant release of IL-6 and TNFa after exposure to either IgGIC or LPS. In the case of TLR4 mut PMNs, cytokine responses to IgGIC or LPS were lost (Figure 2A-D).

Author Summary
The immune system is traditionally divided into innate and adaptive entities. Pattern-recognition receptors (PRRs) collectively recognize molecular structures of invading microorganisms, followed by initiation of immune responses. PRRs comprise the toll-like receptor (TLR) family, including TLR4, which is essential for responses to bacterial lipopolysaccharide (LPS). As part of the adaptive immune system, Fc receptors (FcRs) on immune cells recognize antigen-antibody complexes and link antibody-mediated immune responses to cellular effector functions. Here, we describe cross-talk between the pathogen-recognitionreceptor toll-like receptor 4 (TLR4) and receptors for IgG immune complexes (IgGIC), Fcc receptors (FccRs). We found that TLR4 is involved in FccRIII (CD16) signaling and that heterodimerization of TLR4 and FccRIII occurs in the presence of IgGIC but not LPS. Consequently, dysfunctional TLR4 signaling results in unresponsiveness of immune cells in vitro to both LPS and IgGIC, resulting in absence of acute lung injury after intratracheal administration of LPS or intrapulmonary immune complex deposition. In summary, we describe that TLR4 and FccRIII pathways are structurally and functionally connected. These findings provide new insights of the interplay between innate and adaptive immunity, which closely interact with each other at the receptor level and post receptor signaling pathways. When peritoneal macrophages were employed in the same protocol, similar results were found ( Figure 2E,F). There was a 4-fold increase in IL-6 after exposure of Wt macrophages to LPS, and a 3-fold increase in IL-6 after IgGIC exposure ( Figure 2E). Likewise, there was a robust release of TNFa by Wt macrophages into supernatant fluids after stimulation with IgGIC or LPS. When TLR4 mut macrophages were used under the same conditions, IL-6 and TNFa responses to IgGIC or LPS were greatly abolished ( Figure 2E,F). Similar results were found when macrophages were incubated with polymyxin-treated BSA IgGIC or PAP IgGIC indicating that the results are reproducible and not based on LPS contamination of the reagents ( Figure 2E,F). Thus, the lack of a functional TLR4 is associated with the in vitro inability of PMNs and macrophages to respond to LPS or IgGIC.
In order to assess if the impaired response of TLR4 mut cells observed in vitro might be due to a general impairment of the inflammatory response, peritoneal PMNs and macrophages from Wt and TLR4 mut mice were exposed to opsonized zymosan particles as well as to Pam3Cys, which is a specific ligand for TLR2 [25,26,27] . As displayed in Figure S1, Wt cells showed a significant increase of IL-6 ( Figure S1A,C,E,G) and TNFa ( Figure S1B,D,F,H) release when incubated in vitro with Pam3Cys or opsonized zymosan particles. In contrast to the findings described above (incubation with LPS or IgGIC), PMNs (Figure S1A-D) and macrophages ( Figure S1E-H) from TLR4 mut mice showed full responses for IL-6 and TNFa when incubated with opsonized zymosan particles or Pam3Cys. These data indicate that the ability to produce cytokines in response to non-TLR4 agonists is intact in TLR4 mut cells and that the impairment of the inflammatory response to LPS and IgGIC is specific for the non-functional TLR4 protein.
In another set of experiments, cells from FccRIII-deficient mice were tested for responsiveness to LPS. Peritoneal PMNs and macrophages from Wt and FccRIII 2/2 were incubated with LPS and opsonized zymosan (as a positive control) under the same conditions described above and supernatant fluids were analyzed for IL-6 and TNFa levels by ELISA. As shown in Figure 3, phagocytes from FccRIII +/+ and FccRIII 2/2 mice robustly produced cytokines when incubated with LPS, opsonized zymosan or IgGIC. There was no difference in cytokine secretion between the FccRIII +/+ and FccRIII 2/2 cells, except for LPS-induced TNFa release by FccRIII 2/2 PMNs, which was lower as compared to FccRIII +/+ PMNs, but significantly elevated above baseline levels. As expected, FccRIII +/+ macrophages robustly released IL-6 and TNFa into supernatant fluids when stimulated with IgGIC, whereas macrophages from FccRIII 2/2 mice were unresponsive to IgGIC ( Figure 3C,D).
These results suggest that FccRIII-deficient phagocytes can respond to LPS and that FccRIII is not required for direct TLR4 signaling, while FccRIII is essential for the mediation of IgGICinduced responses. Phosphorylation of FcR c-Subunit Requires the Integrity of TLR4 After binding of LPS, TLR4 engages intracellular signaling pathways via the adaptor molecules MyD88 and TRIF [27]. In the case of FccR-immune-complex interaction, intracellular pathways are activated by tyrosine phosphorylation of the FcRcsubunit ITAM region [8,28]. This subunit is known to be the common adaptor of FccRI, FccRIII and FceRI [29,30], the first two being essential for development of IgGIC induced acute lung injury [31]. In order to evaluate the mechanism behind the impaired response of TLR4 mut cells to IgGIC, tyrosine phosphorylation of the FcRc-subunit was investigated in vitro. When peritoneal PMNs ( Figure 4A) or macrophages ( Figure 4B) from Wt mice were exposed to IgGIC, rapid tyrosine phosphorylation (PY) of the FcRc-subunit occurred over the first 30 min, as indicated by robust bands in the Western blots. In striking contrast, phosphorylation of the FcRc-subunit failed to occur when TLR4 mut cells were used. Here, the intensity of the bands was comparable to those in non-stimulated cells ( Figure 4A,B). When LPS was used as a stimulus ( Figure 4C,D), slight phosphorylation of the FcRc-subunit occurred in Wt cells (but not in TLR4 mut cells), indicating that TLR4 has little ability to activate the FcRc-subunit as an intracellular signaling event ( Figure 4C,D). Furthermore, the above mentioned results were confirmed in macrophages by using polymyxin-treated BSA IgGIC for stimulation under the same conditions in order to exclude LPS contamination of the reagents ( Figure 4E). Collectively, these data suggest that the integrity of TLR4 seems to be required for a proper function of FccR activation via phosphorylation of the FcRc-subunit, further suggesting communication between the TLR4 and FccR signaling pathways.
Acute Lung Injury in Wt, TLR4 Mutant, and TLR4 2/2 Mice Using the LPS and IgGIC models of ALI, Wt, TLR4 mut, TLR4 +/+ and TLR4 2/2 mice were evaluated for responses following lung deposition of IgGIC or LPS. While FccRs play a key role in the IgG immune complex (IgGIC) model of ALI [31,32], TLR4 is critical for the development of lung injury in the LPS model [33,34,35]. As indicated in Figure 5A, LPS-induced lung injury, as defined by the permeability index (leak of plasma albumin into the extravascular lung compartment), showed a 4-fold increase in Wt mice (compared to controls, ctrl) and remained at the control level in LPS-challenged TLR4 mut mice. In the case of IgGIC (Figure 5B), the permeability index rose 5-fold above control (basal) levels in Wt mice. However, TLR4 mut mice unexpectedly showed no evidence of injury after deposition of IgGIC ( Figure 5B). TLR4 2/2 mice behaved similar to TLR4 mut mice in terms of lung injury, with virtually no lung injury in response to deposition of either LPS or IgGIC ( Figure 5A,B). When IL-6 levels were measured in bronchoalveolar lavage (BAL) fluids, LPS and IgGIC induced high levels of IL-6 in Wt mice and very low levels in TLR4 mut mice ( Figure 5C). Similar patterns were found for TNFa levels ( Figure 5D).
Similarly, induction of ALI by intrapulmonary deposition of polymyxin-treated BSA IgGIC in Wt and TLR4 mut mice ( Figure 5E) revealed no difference to the results displayed in Figure 5B; when polymyxin-treated reagents were administered for intrapulmonary IgGIC formation lung permeability rose 3.5 fold in Wt mice whereas mice TLR4 mut mice did not show a significant increase. Thus, these findings support the conclusion that lung injury induction by IgGICs is not linked to contamination of the reagents with endotoxin. In addition, reagents that were used for the formation of IgGIC were administered separately in vivo at the same concentration as they were used in combination for intrapulmonary IgGIC deposition ( Figure 5F). When BSA was injected intravenously, followed by intratracheal PBS injection lung permeability was not different from control mice. Similarly, intratracheal injection of anti-BSA and subsequent intravenous DPBS injection (containing a trace amount of I 125 -labelled BSA) did not result in increased lung permeability. In striking contrast, the combination of anti-BSA (i.t.) and by BSA (i.v.) injection lead to the development of acute lung injury, as also shown in Figure 5B and 5E. These data indicate that the development of lung injury in the IgG model is dependent on the in vivo formation of immune complexes and may not be explained by putative LPS contamination of the reagents since their separate, independent administration failed to increase lung permeability. Finally, IgGIC lung injury was induced in FcR c-subunit-deficient mice, which do not express FccRI and FccRIII on the surface of PMNs and macrophages [36]. In contrast to Wt mice (FcR c-subunit +/+ ), FcR c-subunit 2/2 mice did not develop acute lung injury after intrapulmonary IgGIC deposition, as determined by lung permeability ( Figure 5G). These findings suggest that the IgGICinduced lung injury using anti-BSA and BSA is strictly dependent on the FccR-mediated signalling, and not on LPS-induced activation of TLR4. However, the caveat remains that there is always a concern about LPS contamination in the context of sensitive assays and in vivo responses. In particular, the possibility that LPS was present at concentrations below the detection limit of the available assays, which would not result in any in vivo (and in vitro) responses alone, but would be responsible for putative synergistic effects and an augmentation of IgGIC-induced inflammatory responses cannot be entirely excluded. Expression of FccRIII, FcRc-Subunit, and C5aR in Wt and TLR4 Mutant Mice It is well established that engagement of FccRIII with IgGIC as well as activation of the complement system with generation of C5a and its interaction with C5aR play crucial roles in the pathogenesis of IgGIC-induced ALI [31,37,38]. Therefore, elicited peritoneal PMNs were evaluated by flow cytometry for surface expression of FccRII/III and C5aR protein. As shown in Figure 6A,F, the levels of each receptor on the surface of PMNs were the same in Wt versus TLR4 mut cells. The original flow cytometry data of FccRII/III expression on Wt and TLR4 mut PMNs are displayed in Figure 6B,C. In addition, the total content of FccRIII and FcRc-subunit in cell lysates from Wt and TLR4 mut PMN ( Figure 6D) and macrophages ( Figure 6E) were analyzed by Western blotting. In accordance with the flow cytometry results ( Figure 6A,B), unstimulated phagocytes from both mouse strains expressed the same levels of FccRII/III and FcRc-subunit. The analysis for the house keeping protein GAPDH (lower bands) indicates equal loading of the cell lysates. Thus, the inability of TLR4 mut mice to respond to IgGIC or LPS is not associated with reduced surface content of FccR protein on PMNs, consistent with the findings that there is cross-talk between FccR and TLR4 signaling pathways such that downstream production of IL-6 and TNFa upon IgGIC stimulation requires participation of both pathways. Collectively, these data indicate that TLR4 is required for proper FccRIII functions.

Discussion
The mechanisms by which the recognition of pathogens leads to host responses are inadequately understood. The modulation of immune responses is inter alia mediated by cell surface receptors that are associated with signaling molecules that contain ITAMs (immunoreceptor tyrosine-based activation motifs), TREMs (triggering receptors expressed on myeloid cells) and OSCARs (human osteoclast-associated receptors) [1]. Intracellular signaling after TLR4 activation is mediated through the adaptor proteins, MyD88 and TRIF, whereas FccRI and FccRIII both contain the FcRc-subunit, which is phosphorylated at tyrosine residues by Src and Syk kinases upon FccR activation [28,30,39,40]. Interestingly, ligation of FcRc-subunit containing FcRs results in inhibition of IL-12 production by monocytes in response to TLR ligands [41]. The specificity of IL-12 downregulation appears to be based on inhibition at the transcription level [41]. Moreover, TLRs are considered to control activation of acquired immunity [14], supporting the hypothesis for an instructive role of innate immunity in adaptive immune responses [15]. In the present study, we describe that TLR4 and FccRIII associate, possibly by heterodimerization, following stimulation with IgGIC in vitro (Figure 1). Binding of IgGICs to the extracellular domain of FccRs causes clustering of these receptors, followed by phosphorylation of tyrosine residues within the ITAM region, and subsequent activation of intracellular signaling cascades [28,30,40]. TLR signaling is initiated by dimerization of TLRs, which can form homo-or heterodimers [42]. Previously, it has been suggested that TLR4 co-associates with FccRIII after activation of human monocytes [43]. Based on our findings, it is possible that TLR4 and FccRIII multimerize into clusters following stimulation by LPS or IgGIC, a mechanism known as capping [44], which is required for engagement of intracellular signaling pathways. TLR4 may represent the central component for such signaling or ''docking platforms'' [45] and interconnect intracellular signaling pathways via association to adaptor proteins. As demonstrated in the present study, dysfunction of TLR4 results in impaired signaling in FccRIII pathways (Figure 4).
The mutation that is responsible for the endotoxin tolerance of C3H/HeJ mice has recently been demonstrated to cause suppressed tyrosine phosphorylation by Src tyrosine kinases (Lyn) in the toll-IL-1 resistance (TIR) domain of TLR4, resulting in signaling-incompetence [45]. Altered or suppressed TLR4 tyrosine phosphorylation correlated with impaired MyD88 association and suppressed IRAK-1 activation [45]. In addition, our data suggest that this mutation in the TLR gene not only hinders phosphorylation of its own TIR domain but also blocks the tyrosine phosphorylation of the ITAM-containing FcRc-subunit, the consequence of which ultimately leads to impaired signaling after engagement of FccRIII.
In the LPS model of acute lung injury, TLR4 mut or TLR4 2/2 mice were, as expected, highly protected from the development of tissue damage in the LPS-induced model of acute lung injury ( Figure 5). It is well established that mice with mutation in the TLR4 gene or genetic deficiency of TLR4 are non-responsive to LPS [4], including LPS-mediated lung injury [33,34,35]. In the present study, TLR4 mut and TLR4 2/2 mouse strains unexpectedly also showed greatly attenuated susceptibility to IgGICinduced lung injury ( Figure 5). For this model, it is known that, besides complement activation, FccRs are critical for initiation and development of IgGIC alveolitis [31,32], particularly through engagement and activation of ITAM-containing FccRs (FccRI and FccRIII) [31]. In accordance, mice with targeted disruption of the FcRc-subunit showed an impaired inflammatory response in the reverse passive Arthus reaction [46]. In our study, TLR4 mut mice not only were resistant to lung injury, but also failed to locally release cytokines in vivo after intrapulmonary IgGIC deposition, as indicated by baseline levels of IL-6 and TNFa in BAL fluids ( Figure 5). In companion experiments, in vitro exposure of TLR4 mut phagocytes to IgGIC resulted in complete suppression of proinflammatory cytokines (TNFa, IL-6) in comparison to phagocytes from Wt mice ( Figure 2). Furthermore, TLR4 mut cells showed impaired tyrosine phosphorylation of the FcRcsubunit when exposed to IgGIC, in striking contrast to Wt cells (Figure 4). The fact that TLR4 mut PMNs and macrophages responded with cytokine release when incubated with opsonized zymosan particles or with Pam3Cys ( Figure 3) indicates that 1.) the mutation in the TLR4 gene does not lead to a global impairment of the cellular inflammatory/immune response and 2.) the intracellular signaling pathways are intact since other TLRs (such as TLR2 and TLR6), which share common pathways, could be activated in vitro. On the other hand, phagocytes from FccRIIIdeficient mice are fully responsive to LPS (Figure 3), suggesting that TLR4 signaling does not depend on the functional integrity of FccRIII, whereas TLR4 is required for FccRIII signaling.
Especially in the field of immunology, there is an increasing number of reports describing effects of receptor interactions. Examples include a previous study suggesting cross-talk between IFN-gamma and IFN-alpha receptors with signaling pathways [47]. In brief, signalling by IFN-gamma was shown to depend on the IFN-alpha/beta receptor components. A more recent publication describes that signalling triggered by NKG2D and DAP10 is coupled to the interleukin 15 receptor signalling pathway, suggesting that coupling of activating receptors to other receptor systems may regulate cell type-specific signaling events [48]. In the case of innate immunity, it has been proposed several times that there is a link between TLR4 and the complement system, especially to the C5a signalling pathway, which can negatively regulate TLR4-induced responses [49,50]. Under physiological conditions, receptor interactions and cross-talk between signalling pathways might represent important regulatory mechanisms of the immune system to provide distinct but fine-tuned responses. In the case of TLR4 and FccRIII, cross-talk may provide an optimal and rapid response against invading microorganisms by mediating an interplay between adaptive and innate immunity. However, in certain conditions, such as systemic inflammation (sepsis) or autoimmune diseases that are characterized by a loss of inhibitory action or uncontrolled activation of signalling pathways, a loss of control over otherwise carefully orchestrated receptor interactions can become instruments of harm.
Taken together, the present findings strongly suggest that (i) there is a direct link between TLR4 and FccR pathways, (ii) phosphorylation of tyrosine residues in the ITAM-containing FcRc-subunit requires the presence and integrity of TLR4 during cellular activation after binding of IgGICs to FccRs, and (iii) presence of IgGICs results in an association between TLR4 and FccRIII (CD16) on phagocytic cells. These data imply that innate and adaptive immunity are closely connected at the receptor level and post receptor signaling pathways, which might have ramifications for a variety of inflammatory conditions, such as IgGIC-mediated autoimmune diseases (rheumatoid arthritis or glomerulonephritis), ischemia/perfusion injury, trauma or systemic inflammation (sepsis), etc.

Ethics Statement
All studies were performed in accordance with the University of Michigan Committee on Use and Care of Animals.

Immunoprecipitation and Western Blotting
After incubation of peritoneal PMNs or macrophages with either IgG immune complexes (100 mg/ml; prepared as described elsewhere [52] or LPS (20 ng/ml) for 5 to 30 min, supernatant fluids were removed and pellets were lysed with 1X RIPA buffer containing Vanedate and protease inhibitors (Roche Diagnostics). Protein concentrations were determined in cell lysates using BCA protein assay (Pierce). Equal protein amounts of supernatants were then incubated overnight with preblocked protein A and G beads (Santa Cruz) in the presence of anti-FcRc-subunit IgG (Upstate) or anti-TLR4 IgG(Santa Cruz), respectively. Reverse direction immunoprecipitation included anti-FccRIII IgG (Santa Cruz).

ELISA for Mouse IL-6, TNFa
For measurement of IL-6 and TNFa in BAL fluids and supernatant fluids after in vitro incubation of mouse PMNs and macrophages, commercially available ELISA-kits (''Duo set'', R&D Systems) were used according to the manufacturer's protocol.

Immune Complex Lung Injury
To induce IgGIC lung injury, tracheae of mice were surgically exposed and 125 mg rabbit anti-BSA IgG (MP Biomedicals) was administered using a 30 gauge needle (volume of 42 ml/mouse) followed by intravenous injection of BSA (500 mg; Sigma). For determination of the permeability index as a quantitative marker for vascular leakage, 125 I-labelled bovine serum albumin (1 mCi 125 I-BSA/mouse) was injected intravenously. After the development of acute lung injury, the pulmonary vasculature was flushed with 2.0 ml PBS. The amount of lung radioactivity was then measured as a ratio of radioactivity present in 100 ml blood recovered from the inferior vena cava at the time of animal euthanasia and that in lung. For bronchoalveolar lavage retrieval, lung injury was performed as described above, but without the intravenous injection of 125 I-BSA. The airways were flushed with 0.8 ml ice cold PBS using a blunt 20 gauge needle and BAL fluids were recovered for further studies.

LPS Lung Injury
50 mg LPS from E.coli (serotype O111:B4; Sigma) were given intratracheally (volume of 42 ml/mouse). When lung permeability was measured, a trace amount of 125 I-BSA was injected intravenously, as described above. The permeability index was determined and BAL fluids were collected as described for the IgGIC model.

Detection of Possible LPS Contamination
Reagents other than LPS, such as DPBS, BSA, anti-BSA IgG that were used for the in vivo and in vitro experiments were tested for LPScontamination. For quantification of LPS content, samples were conducted in Limulus Amebocyte Lysate Kinetic-QCL assay (Cambrex) according to the manufacturer's protocol and as described elsewhere [53]. In addition, reagents used for immune complex formation (DPBS, BSA, anti-BSA IgG) were subjected to endotoxin removal (Pierce) prior to induction of lung injury or preparation of immune complexes used stimulation of phagocytes in vitro.

Analysis of FccR and C5aR on PMNs
Flow cytometric analysis was conducted after whole blood collection of untreated wild-type and TLR4 mut mice in a citratecontaining syringe. Rabbit anti-mouse C5aR serum (1:10 dilution; Lampire) was incubated with mouse whole blood. Non-specific rabbit serum (Jackson Immunoresearch) was added to control samples in equal amounts. For detection of FccR on PMNs, mouse whole blood was either incubated with 1 mg monoclonal anti-FccRII/III IgG (clone 2.4G2; BD Pharmingen) or with the appropriate isotype IgG control (Jackson Immunoresearch). After washing, cells were suspended in Phycoerythrin (PE)-labeled anti-rabbit IgG (Invitrogen) diluted 1:200 in staining buffer and incubated at room temperature for 45 min. Erythrocytes were lysed by addition of 16 FACS lysing solution (BD Pharmingen) for 10 min. After washing, the leukocytes were resuspended in a 1%-paraformaldehyde fixing solution and analyzed on a flow cytometer (BD Pharmingen).

Statistical Analysis
All values were expressed as mean6SEM. Data sets were analyzed by one-way analysis of variance (ANOVA); differences in mean values among experimental groups were then compared using Tukey multiple comparison test. Results were considered statistically significant when P,0.05. Figure S1 Cytokine response of PMNs and macrophages to Zymosan and Pam3Cys. In vitro cytokine responses to non-TLR4 agonists of elicited peritoneal phagocytes from Wt or TLR4 mut mice. PMNs (A-D) and macrophages (E-H) (3610 6 cells/ml) were incubated (for 4 hr) with serum-opsonized zymosan particles (300 mg/ml) or Pam3Cys (1 mg/ml). Ctrl = control levels of nonstimulated cells. For each condition n$3. Differences between controls and stimulated cells were found to be statistically significant (p,0.05). Found at: doi:10.1371/journal.ppat.1000464.s001 (0.50 MB EPS)