Phosphoinositide 3-Kinases γ and δ, Linkers of Coordinate C5a Receptor-Fcγ Receptor Activation and Immune Complex-induced Inflammation*

Fcγ receptors (FcγR) and the C5a receptor (C5aR) are key effectors of the acute inflammatory response to IgG immune complexes (IC). Their coordinated activation is critical in IC-induced diseases, although the significance of combined signaling by these two different receptor classes in tissue injury is unclear. Here we used the mouse model of the passive reverse lung Arthus reaction to define their requirements for distinct phosphoinositide 3-kinase (PI3K) activities in vivo. We show that genetic deletion of class IB PI3Kγ abrogates C5aR signaling that is crucial for FcγR-mediated activation of lung macrophages. Thus, in PI3Kγ-/- mice, IgG IC-induced FcγR regulation, cytokine release, and neutrophil recruitment were blunted. Notably, however, C5a production occurred normally in PI3Kγ-/- mice but was impaired in PI3Kδ-/- mice. Consequently, class IA PI3Kδ deficiency caused resistance to acute IC lung injury. These results demonstrate that PI3Kγ and PI3Kδ coordinate the inflammatory effects of C5aR and FcγR and define PI3Kδ as a novel and essential element of FcγR signaling in the generation of C5a in IC disease.

Activated complement component C5a is a pleiotropic molecule that regulates the activity of many cell types, with a broad range of biological functions in the immune system (1). C5a binds to at least two seven-transmembrane domain receptors, C5aR 4 (CD88) and C5L2, expressed on a variety of immune cells, including circulating leukocytes, mast cells, basophils, macrophages, and many others. C5aR-dependent activation of these cells by C5a results in inflammatory mediator release and granule secretion, which in turn alters vascular permeability, induces smooth muscle contraction, and promotes cell migration (2). It is well established that this C5a-triggered cascade of events contributes to the pathogenesis of various diseases in humans, including myocardial ischemia/reperfusion injury and respiratory distress syndrome (3)(4)(5). In addition, genetic deletion of C5aR is very effective in preventing inflammation in animal models of type III hypersensitivity (as modeled by the passive reverse Arthus reaction) and arthritis, as well as antibody-dependent type II autoimmunity (6 -8).
Complement activation occurs through multiple pathways (classical, alternative, and lectin binding) in the circulation, each of which produces C5a. Interestingly, C5a is also formed within the extravascular tissue compartments through activation of resident innate immune effector cells, such as tissue macrophages, and requires the presence of receptors for the Fc portion of IgG, Fc␥R (reviewed in Ref. 9). Fc␥R exert their function through paired expression of activating (Fc␥RI, Fc␥RIII, and Fc␥RIV) and inhibitory (Fc␥RIIB) receptors (reviewed in Refs. 10 and 11). Compelling evidence suggests that the ratio of the opposing signaling Fc␥R is critical in setting the cellular thresholds for the pathogenic activity of autoantibodies (reviewed in Refs. 10 -12) and that C5a regulates this ratio, thus amplifying the Fc␥R-mediated inflammatory response in autoimmunity (8,(13)(14)(15).
Although C5a/C5aR and Fc␥R likely cooperate in the context of immunological diseases, the precise molecular mechanisms of their combined signaling remain to be elucidated. C5a/C5aR has recently been shown to induce and suppress transcription of the two Fc␥RIII and Fc␥RII genes on macrophages in a PI3Kdependent manner (16). In addition, PI3K-mediated signal transduction is also known for Fc␥R-induced cell activation (17). Distinct members of the PI3K family of signaling molecules may thus participate in the regulatory cross-talk between individual Fc␥R and C5aR, which was previously proposed to be a key event in the initiation of the inflammatory cascade in vivo (8,9,13,14).
The PI3K family can be divided into three subfamilies (I, II, and III) on the basis of their structural characteristics, activation mechanisms, and substrate specificity. Class IA PI3K members are heterodimers consisting of a regulatory subunit (seven distinct isoforms including splice variants are known, which are derived from five genes: p85␣, p85␤, p55␣, p55␤, and p55␥) and a catalytic subunit (which is one of three types: p110␣, p110␤, and p110␦) (reviewed in Refs. 18 -21). Whereas class IA PI3K␣ and PI3K␤ are widely expressed, PI3K␦ is found primarily in cells of the immune system, where it is activated by receptors involved in protein-tyrosine kinase signaling, such as cytokine receptors, antigen receptors present on T and B cells, and the mast cell IgE Fc receptor, Fc⑀RI (reviewed in Refs. [22][23][24]. The only class IB PI3K enzyme, PI3K␥, also consists of a catalytic (i.e. p110␥) and a regulatory (i.e. p101 or p87) subunit and is also preferentially expressed in immune cells (reviewed in Ref. 25). In contrast to PI3K␦, however, class IB PI3K␥ is predominantly activated by G-protein-coupled receptors, such as C5aR.
Recent studies in mice lacking PI3K␥ and PI3K␦ (26 -29) have revealed a key role of class I PI3K members in many cellular immune responses at the interface of innate and adaptive immunity. For instance, PI3K␥ deficiency leads to defects in chemokine-induced migration of neutrophils, macrophages, and dendritic cells to sites of infection and tissue injury (27)(28)(29)(30). Moreover, G-protein-coupled receptor-induced activation of mast cells and platelets is also affected in PI3K␥ Ϫ/Ϫ mice (31,32), whereas PI3K␦ Ϫ/Ϫ mice show a severely impaired B cell response of antibody production to T cell-dependent and -independent antigens (26,(33)(34)(35)(36).
In this study, we analyzed PI3K␥ Ϫ/Ϫ mice in the experimental model of the pulmonary Arthus reaction and found defective C5aR-mediated Fc␥R regulation and macrophage activation, leading to impaired induction of the inflammatory response. Notably, however, C5a production was not affected in PI3K␥ Ϫ/Ϫ mice, but was impaired in the genetic absence of PI3K␦. Conse-FIGURE 1. Attenuation of IgG IC inflammation in mice receiving the general PI3K inhibitor wortmannin. Induction of the inflammatory response in the lung was performed by intratracheal application of 150 g of purified anti-OVA IgG Ab, followed by systemic 20 mg/kg OVA Ag in WT B6 mice (IC) or WT mice treated with wortmannin (IC ϩ wortm.). Mice receiving only anti-OVA Ab served as controls (Ab). After 4 h, lungs were lavaged. BALF was assayed for PMN infiltration (A); pulmonary hemorrhage (B); chemotactic activity (C); and production of CXC chemokines (MIP-2 and KC) (D), MIP-1␣ (E), and TNF␣ (F). Data are expressed as the means Ϯ S.E. (n ϭ five to eight mice for each group). Differences in IC compared with IC ϩ wortmannin treatment groups are significant or highly significant for all parameters (*, p Ͻ 0.05; **, p Ͻ 0.001). RBC, red blood cells.
quently, PI3K␦ Ϫ/Ϫ mice failed also to develop IC-induced inflammation, indicating that both PI3K␥ and PI3K␦ are critical linkers of coordinate C5aR-Fc␥R activation in IC disease.

Mice
The generation of B6 PI3K␥-and PI3K␦-deficient mice and their phenotypic characterization have been described previously in detail (26,27). WT B6 mice were used for all comparisons. All mice were maintained under dry barrier conditions at the animal facilities of the Hanover Medical School and Heinrich-Heine University.

Passive Reverse Lung Arthus Reaction
Mice were anesthetized with ketamine and xylazine; the trachea was cannulated; and 150 g of protein G-purified anti-OVA IgG Ab (ICN) was applied. Immediately thereafter, 20 mg/kg OVA Ag was given intravenously. Ab control animals received phosphate-buffered saline instead of OVA Ag. In PI3K-blocking experiments, 2 g of wortmannin (Calbiochem) was given intratracheally together with the application of anti-OVA IgG. Mice were killed at 2-4 h after initiation of lung inflammation. BAL was performed five times with 1 ml of 0.9% NaCl at 4°C. The total cell count of BALF was assessed with a hemocytometer. The amount of red blood cells represented the degree of hemorrhage. For quantitation of PMN accumulation, differential cell counts were performed on Cytospins (10 min at 55 ϫ g) stained with May-Grünwald/Giemsa using 300 -500 l of BALF. The concentrations of MIP-1␣, MIP-2, KC, and TNF␣ in BALF were measured in duplicate in appropriately diluted samples with enzyme-linked immunosorbent assay kits (R&D Systems) according to the manufacturer's instructions.

Fc␥R Expression Analysis in Vitro and in Vivo
Luciferase Reporter Gene Assays-The Fc␥R promoter-reporter plasmids Fc␥RII p(Ϫ729/ϩ501)Luc and Fc␥RIII p(Ϫ808/ϩ18)Luc (0.7 g) were cotransfected with 0.3 g of the reference plasmid pRL-CMV into 5 ϫ 10 5 macrophage MH-S cells in 12-well plates using Lipofectamine TM as

PI3K-mediated Activation of C5aR and Fc␥R
described (16). Cells were recovered after 24 h, cultured for 24 h in 1% fetal calf serum-containing RPMI 1640 medium, and further treated with 50 ng of recombinant human C5a for 4 h. In some experiments, cells were pretreated for 1 h with an anti-C5aR Ab to block C5aR activity. In PI3K inhibition experiments, wortmannin (20 M) and a PI3K␥-specific inhibitor (5-quinoxalin-6-ylmethylenethiazolidine-2,4-dione, 10 nM; Calbiochem) were used. The Src kinase inhibitor PP2, which does not inhibit C5aR signaling (16), served as a negative control. Cells were then lysed and measured for luciferase activities using the Dual-Luciferase reporter assay system (Promega). Firefly luciferase activity was normalized by Renilla luciferase activity to yield the relative promoter activity.
Real-time RT-PCR-Total RNA was prepared from BAL-AM cells of the indicated mice at 0 and 2 h after C5a treatment (intratracheal application of 200 ng of recombinant human C5a) using TRIzol reagent and analyzed for Fc␥RII, Fc␥RIII, and FcR␥ transcript levels normalized to tubulin by real-time RT-PCR using published primers (8,13).

C5/C5a Analysis
Real-time RT-PCR-Total RNA was prepared from BAL-AM cells of mice at 2 h after Ab and IC treatments and analyzed for C5 transcription by real-time RT-PCR as described (8).
Detection of C5a-dependent Chemotactic Activity in Vivo-Bone marrow cells (containing 64 -68% PMN) from C5aR-deficient mice or B6 controls were suspended at 7.5 ϫ 10 6 cells/ml of RPMI 1640 medium and 0.1% bovine serum albumin (fatty acid-free). 100 l of the bone marrow cell suspension was placed into the insert of a Transwell chemotaxis chamber, and the bottom well was filled with 600 l of RPMI 1640 medium and 0.1% bovine serum albumin (negative control) or BALF diluted 1:2 in RPMI 1640 medium and 0.1% bovine serum albumin. BALF was obtained from PI3K␥ Ϫ/Ϫ and PI3K␦ Ϫ/Ϫ mice, wortmannin-treated B6 mice, or WT controls at 2-4 h after OVA/anti-OVA IC inflammation. BALF from mice receiving Ab but not OVA Ag served as controls. Inserts were combined in the lower chambers and incubated at 37°C and 6% CO 2 for 2 h. After the incubation, 50 l of 70 mM EDTA was added to the lower chambers to release adherent cells from the lower surface of the membrane and from the bottom of the well. Plates were further incubated for 30 min at 4°C; inserts were removed; and the transmigrated neutrophils were vigorously suspended and counted with a FACSCalibur for 1 min at 12 l/min with gating on forward and side scatter. Migration of PMN from the insert to the bottom well was quantitated as the percentage of total PMN loaded into the upper chamber.

Statistical Analysis
Data for comparison of mean values among samples were analyzed by a two-sided unpaired Student's t test.

In Vivo Effects of General PI3K Inhibition in IgG IC-induced
Inflammation-To determine whether inhibition of PI3K signaling has a protective effect on IgG IC inflammation, we conducted lung Arthus experiments in mice receiving wortmannin, a general PI3K inhibitor. Assessment of lung inflammation at 4 h revealed that local intratracheal application of 2 g of wortmannin suppressed alveolar PMN influx, hemorrhage, and chemotactic activity in IgG IC-challenged B6 mice (Fig. 1, A-C). In addition, PI3K inhibition resulted in strong reduction of IC-induced mediator production of CXC chemokines (MIP-2 and KC), MIP-1␣, and TNF␣ (p Ͻ 0.001) (Fig. 1,  D-F). These data indicate that PI3K activity is required for the generation of crucial inflammatory mediators in the initiation of acute IC inflammation, thus confirming previous observations on the key role of PI3K in other experimental models of immunological disease (reviewed in Refs. 38 -41). (13), thereby tuning Fc␥R-mediated release of PMN recruitment factors in lung IC inflammation (42). Previous studies also suggested abrogation of C5aR signaling and impaired Fc␥RIII-mediated functions in mice lacking G␣ i2 , a selective upstream regulator of PI3K␥ in macrophages (14), implicating a specific involvement of PI3K␥ in Fc␥R regulation by these cells. We have used three distinct experimental strategies to determine the potential role of the class IB PI3K␥ isoform in C5a-and IgG IC-induced Fc␥R regulation.

PI3K␥-dependent Regulation of Macrophage Fc␥RII and Fc␥RIII in Vitro and in Vivo-C5a is a major regulator of Fc␥R with induction of activating Fc␥RIII and suppression of inhibitory Fc␥RII
In the first approach, we performed in vitro reporter gene assays using the AM cell line MH-S and the recently characterized C5a-responsive Fc␥RII (Ϫ729 to ϩ501) and Fc␥RIII (Ϫ808 to ϩ18) gene promoters (16). Enhanced versus suppressed Fc␥RIII and Fc␥RII promoter activities in transfected MH-S cells were obtained by stimulating them with C5a for 4 h (Fig.  2A). The simultaneous positive and negative Fc␥RIII and Fc␥RII regulation was equally sensitive to inhibition with wortmannin or a PI3K␥-selective inhibitor (5-quinoxalin-6-ylmethylenethiazolidine-2,4-dione) or an anti-C5aR mAb. As a negative control, the effect of PP2, an Src kinase inhibitor that does not interfere with C5aR signaling (16), was analyzed, and no inhibition was observed. In the second approach, we directly instilled 200 ng of C5a into the tracheas of WT B6 and PI3K␥ Ϫ/Ϫ mice and tested for in vivo changes of Fc␥R transcription in BAL-AM cells. As shown in Fig. 2B, both Fc␥RIII␣ and Fc␥RIII␥ mRNA levels were up-regulated, whereas Fc␥RII transcription was suppressed within 2 h after application of C5a. This inverse regulation of Fc␥RII and Fc␥RIII was impaired in the absence of PI3K␥ (Fig. 2B). IC-induced Fc␥R mRNA (data not shown) and protein (Fig. 2C) regulation was also abrogated in PI3K␥ deficiency. In the third approach, AM cells were collected at 0 and 4 h after IgG IC challenge and analyzed by flow cytometry as described (13). Fluorescenceactivated cell sorter analysis revealed IC-induced changes in increased staining of Fc␥RIII (0-h versus 4-h groups, 297 Ϯ 71 versus 912 Ϯ 33 mean fluorescence intensity Ϯ S.E.; p ϭ 0.039) and reduced staining of Fc␥RII (0-h versus 4-h groups, 424 Ϯ 37 versus 191 Ϯ 23 mean fluorescence intensity Ϯ S.E.; p ϭ 0.049) in WT B6 mice, indicating that the observed changes in Fc␥R mRNA (13,14) correlate with modulated Fc␥RII/III surface membrane expression. However, regulation of Fc␥R, which requires C5aR (13), was absent in PI3K␥ Ϫ/Ϫ mice (Fig. 2C).

. Different effects of general PI3K inhibition and PI3K␥ deficiency on IC-induced C5a production. Pulmonary 4-h IC inflammation was induced in the indicated mice (WT and PI3K␥
Ϫ/Ϫ ; A) (WT treated or not with wortmannin (wortm.); B), and BALF was assayed for C5a bioactivity (IC and IC ϩ wortm.). Controls received anti-OVA Ab without OVA Ag (Ab). Chemotactic activity was determined by Transwell migration assays of neutrophils (PMN isolated from bone marrow of C57BL/6 and C5aR Ϫ/Ϫ mice) elicited with 300 l of BALF from the Ab, IC, and IC ϩ wortmannin treatment groups. Results are expressed as the percentage of PMN number loaded into the upper chamber that had migrated to the bottom well (mean Ϯ S.E., n ϭ four to nine mice for each group). Differences in the BALF-induced migration of C57BL/6 and C5aR Ϫ/Ϫ PMN represent bioactive C5a. Significant differences were determined by Student's t test (**, p Ͻ 0.001).
Collectively, these data indicate that PI3K␥ regulates Fc␥R expression in response to C5aR activation in vitro and in vivo. Impaired IC Inflammation but Normal C5a in PI3K␥-deficient Mice-Because PI3K␥ is an apparent critical mediator of both in vivo responses to C5a-and IC-induced Fc␥R modulation, we then asked whether PI3K␥ not only affects the expression level of Fc␥R but also regulates the subsequent Fc␥R-mediated production of factors that dictate PMN migration in lung disease. Whereas TNF␣ induces expression of adhesion molecules, thus promoting PMN migration more indirectly, MIP-2 and MIP-1␣ represent direct recruitment factors for PMN (42,43). Both PMN influx and hemorrhage were blunted in PI3K␥ deficiency (Fig. 3, A and B), and IC-induced release of all mediators tested (MIP-2, MIP-1␣, and TNF␣) and KC (data not shown) was decreased to background levels (Fig. 3, D-F). Curiously, however, the chemotactic activity of BALF from IC-treated PI3K␥ Ϫ/Ϫ mice showed only partial reduction (Fig.  3C), indicating that most, but not all, acute changes associated with the lung Arthus reaction were dependent on PI3K␥.
We have previously shown that IC activation of lung macrophages by Fc␥R triggers C5a production both in vitro and in vivo (14). To test the possibility that cell-derived C5a might be responsible for the remaining chemotactic activity seen in PI3K␥ Ϫ/Ϫ mice, BALF from both PI3K␥ Ϫ/Ϫ and wortmannintreated mice was analyzed for C5a activity by Transwell chemotaxis assays using bone marrow PMN cells from C5aR Ϫ/Ϫ mice.
The difference in the percentage of C5aR Ϫ/Ϫ and C5aR ϩ/ϩ PMN indicated the presence of bioactive C5a. In contrast to BALF from PI3K␥-deficient animals (Fig. 4A), the C5a-depend-  A and B), and C5a-mediated C5aRtriggered chemotactic activity of BALF was determined by Transwell migration assays of neutrophils (PMN isolated from bone marrow of C57BL/6 and C5aR Ϫ/Ϫ B6 mice) (C). Results are shown as the means Ϯ S.E. (n ϭ four to seven mice for each group). Significant differences were determined by Student's t test (*, p Ͻ 0.05; **, p Ͻ 0.001). AU, arbitrary units; n.d., not determined. FIGURE 6. Impaired IC inflammation in PI3K␦-deficient B6 mice. WT B6 and PI3K␦ Ϫ/Ϫ mice received 150 g of anti-OVA Ab intratracheally and 20 mg/kg OVA Ag intravenously, and the inflammatory response in the lung was allowed to proceed for 4 h (IC). Mice not receiving OVA Ag served as Ab controls (Ab). Lungs were lavaged, and BALF was assayed for production of MIP-2 (A), TNF␣ (B), KC (C), and MIP-1␣ (D); PMN infiltration (E); and pulmonary hemorrhage (F). Data are expressed as the means Ϯ S.E. (n ϭ four to seven mice for each group). Differences for the IC treatment groups of WT mice compared with PI3K␦ Ϫ/Ϫ mice are highly significant for all parameters (**, p Ͻ 0.001). RBC, red blood cells.
FIGURE 7. Summarized model of the alternative roles of the two class I PI3K members (PI3K␥ and PI3K␦) in IC inflammation. As discussed under "Results," the initial contact between IgG IC and AM cells causes induction of C5 and cleavage into C5a that is PI3K␦-dependent. Engagement of C5aR by such PI3K␦-mediated C5a triggers PI3K␥ activation, resulting in regulated Fc␥R expression, induced cytokine/chemokine production, and, in turn, neutrophil inflammation. ent chemotactic activity detected in BALF of 4-h IC-treated WT mice was markedly suppressed upon PI3K inhibition (Fig.  4B), suggesting that generation of C5a critically involves a PI3K activity, which, however, is not related to PI3K␥.
PI3K␦-dependent C5/C5a in IC Inflammation-Because macrophages normally coexpress class IA PI3K␦ and class IB PI3K␥ (25) (data not shown), we analyzed PI3K␦ Ϫ/Ϫ mice to determine the possible contribution of PI3K␦ to the production of C5/C5a in IC inflammation. At 2 h after IgG IC challenge, BAL-AM cells displayed increased C5 mRNA synthesis in WT mice (p Ͻ 0.001 compared with Ab controls) (Fig. 5A). In contrast to both WT and PI3K␥ Ϫ/Ϫ mice (Fig. 5A), C5 mRNA levels remained largely unchanged in AM cells of IC-treated PI3K␦ Ϫ/Ϫ mice (Fig. 5B). To examine whether the changes in PI3K␦-dependent C5 gene induction correlate with release of C5a in acute lung injury, BALF from WT and PI3K␦ Ϫ/Ϫ mice was assayed for the appearance of C5a. In contrast to BALF from WT littermates, C5a-dependent chemotactic activity detected in BALF from 2-and 4-h IC-treated PI3K␦ Ϫ/Ϫ mice was strongly decreased (Fig. 5C), indicating that PI3K␦ mediates generation of C5 and C5a by macrophages in lung IC inflammation.
Impaired IC Inflammation in PI3K␦ Ϫ/Ϫ Mice-In C5-defective Hc Ϫ/Ϫ mice, it has been shown that IC-mediated cytokine production depends on C5a (43), so we finally assessed the biological relevance of PI3K␦-mediated C5a to the release of cytokines and chemokines as well as hemorrhage and neutrophil infiltration. The impaired production of C5a in PI3K␦ Ϫ/Ϫ mice (Fig. 5) seemed to correlate with reduced TNF␣ and MIP-2 mRNA levels in AM cells at 2 h of IC challenge compared with WT mice (data not shown), decreased appearance of these mediators in BALF (at 4 h) (Fig. 6, A and B), reduced synthesis of KC and MIP-1␣ (Fig. 6, C and D), and significantly lower levels of red blood cells (Fig. 6E) and PMN (Fig. 6F) in alveoli. Taken together, our results support a model of IC inflammation in which the cellular communication of Fc␥R and C5aR is coordinated by distinct class I PI3K members in the initiation of the inflammatory cascade (Fig. 7).

DISCUSSION
In this study, we have analyzed the role of class IA and IB PI3K signaling molecules in the Arthus reaction, the classical animal model of IC disease. Genetic deletion of PI3K␥ completely abolishes the inflammatory cascade with respect to C5aR-mediated Fc␥R regulation and subsequent cytokine/chemokine production, hemorrhage, and alveolar PMN transmigration, indicating that PI3K␥ is the pivotal downstream mediator of C5aR through which the C5aR-Fc␥R axis (13,14) controls early neutrophil accumulation in the bronchoalveolar compartment. These data significantly extend previous findings that provided evidence for a role of PI3K␥ in PMN movement in chemokine-induced lung neutrophilia in vivo (44). Moreover, they support the concept that, in the complex situation of an inflammatory disease, a major function of the G-protein-regulated PI3K␥ is to transmit early C5aR-triggered signals in macrophages, thereby promoting a shift in the receptor balance between activating Fc␥RIII and inhibiting Fc␥RII that is essentially required for the recruitment of neutrophils via Fc␥R-mediated synthesis of various secondary mediators (including MIP-2, KC, MIP-1␣, and TNF␣). In addition, our analysis of general PI3K inhibition versus selective PI3K␥ deficiency also implicates the participation of an additional PI3K in the generation of a full chemotactic gradient that requires C5a.
PI3K␦ is the most prevalent class IA PI3K isoform in leukocytes and controls a wide range of distinct immunological responses, such as antigen presentation, cytokine and chemokine secretion by natural killer/T cells, Th2-mediated inflammation, and many others (45)(46)(47)(48)(49)(50). An important new mechanistic aspect of this study is the finding that PI3K␦ plays a non-redundant role in the initial phase of the lung Arthus reaction with a specific requirement of PI3K␦ for C5a production upstream of C5aR and PI3K␥. As a consequence of impaired C5a, PI3K␦ Ϫ/Ϫ mice exhibit reduced cytokine production and PMN migration, suggesting that macrophage PI3K␦ is important for neutrophil recruitment in the initiation of the inflammatory response. A recent study in C5-defective Hc Ϫ/Ϫ mice has concluded that activation of C5 is central to the inflammatory reaction in IC-mediated disease (43). We found a similar protective phenotype in PI3K␦ deficiency, thus indicating that cellular activation of PI3K␦ is the relevant C5/C5a-generating mechanism in lung pathology. It is important to note, however, that it is at present not clear whether, in addition to macrophages, other cell types are involved in the PI3K␦-mediated process of C5a production.
In summary, PI3K␥ Ϫ/Ϫ and PI3K␦ Ϫ/Ϫ mice allowed dissection of the signaling mechanisms of two major cooperating receptor systems, C5aR and Fc␥R, in immune inflammation. The results suggest that the different receptors have distinct PI3K␥and PI3K␦-specific requirements. PI3K␥ is the central player in C5aR signaling that sets the threshold for Fc␥R activation. Conversely, both Fc␥R (8,14) and PI3K␦ are critical for local C5 and C5a production. Our observations also indicate that the two class IA and IB PI3K molecules might be essential components of the immune response to pathogenic IC and, as recently discussed for rheumatoid arthritis (41,51), may represent potential molecular targets for pharmacological intervention in inflammatory and rheumatic autoimmune diseases.