More Paths to PI3Kγ

In the huge collection of molecules that underpin mammalian biology, only a small number stand out as targets for drug development. PI3Kγ is one such molecule that has received substantial investments to assess whether inhibitors can be developed as potential therapeutics. Although many studies have addressed the structure and regulation of PI3Kγ, our understanding of the enzyme is far from complete. A publication in PLOS Biology from the Wymann group provides new insights into this process by identifying an alternate route to PI3Kγ activation, and these results enable us to understand how PI3Kγ signaling is regulated in mast cells and is thus important in so many cell, tissue, and disease settings.


Roles for PI3Kc in Health and Disease
PI3Kc is expressed strongly in a number of immune cells, including mast cells, neutrophils, and eosinophils. In these cell types, it sits at a key, early, near-receptor stage in proinflammatory, intracellular signaling pathways. This is, in part, why genetic loss or selective inhibition of PI3Kc has little effect on normal mouse development and function but can suppress inflammation in a number of mouse models of disease, including rheumatoid arthritis [1], anaphylaxis [2], atherosclerosis [3], and glomerulonephritis [4]. Interestingly, however, PI3Kc seems to perform a variety of important roles in other cell types/systems where it is often barely detectable-for example, in the heart, where it suppresses cAMP signaling and contractility (and hence may be a useful therapeutic target in certain types of cardiac failure) [5][6][7], and in fat metabolism, where its activity supports fat deposition [8]. Finally, there is evidence that PI3Kc may support tumour growth and progression [9,10].

Class I PI3K Signaling
PI3Kc (phosphoinositide 3-kinase) belongs to the class I PI3K family of signaling enzymes (along with PI3Ks a, b, and d) that 3phosphorylate the membrane phospholipid PtdIns(4,5)P 2 to yield the signaling lipid PtdIns(3,4,5)P 3 . The class I PI3Ks can all be activated by cell surface receptors to drive accumulation of PtdIns(3,4,5)P 3 in the inner leaflet of the plasma membrane. This acts as a signal enabling proteins capable of binding PtdIns(3,4,5)P 3 , typically via PH domains, to concentrate at the cytosolic interface of the plasma membrane. The classic example of a PH domain-containing PtdIns(3,4,5)P 3 effector is PKB (Akt), but there are possibly as many as 60-70 in a mammalian genome and up to at least 20 can be expressed in the same cell. This family of PtdIns(3,4,5)P 3 effectors transduce the lipid signals into many forms, including changes in protein kinase, Rac-family-GEF and Rho-family-GAP activity, and/or distribution.

PI3Kc-The Detail
PI3Kc is made up of a p110c catalytic subunit [11] combined with either a p84 (also called p87 PIKAP ) [12][13][14] or a p101 regulatory subunit [15]. Interestingly, the three proteins are well conserved in eukaryotes from humans to fish; p84 and p101 have even retained their neighbouring genetic location. PI3Kc is thought to exist as a dimer in vivo, although this is only based on the lack of any evidence for the existence of free p110c. It appears that despite immune cells like neutrophils, expressing p101 and p84, mast cells, the subject of the study reported by the Wymann group, only express p84 [16].

The p101 Regulatory Subunit
Our current understanding of PI3Kc activation centers on how the p101 subunit allows the complex to be substantially activated by G-protein bc subunits (Gbcs) liberated from G proteins upon activation of GPCRs. This mechanism is thought to explain why PI3Kc is typically (but not universally) activated by GPCRs. Frustratingly, we do not know the residues/domains in either p101 or p110c involved in the interaction with Gbcs. It appears the interaction between p101 and p110c is very tight and unlikely to have a significant on/off rate under physiological conditions. There is some evidence supporting direct interactions between Gbcs and both p101 and p110c [15,17].

The p84 Regulatory Subunit
But the involvement of the p84 in PI3Kc has been less clear. There is anecdotal evidence in the field that the interaction between p84 and p110c is less tight than that between p101 and p110c. Despite this, once bound, p84/p110c complexes are able to survive gel-filtration chromatography or multiple cycles of washing with traditional detergent-containing lysis buffers through pull-down protocols without significant dissociation, suggesting that their association is not rapidly reversible. It is clear, however, that the p84/p110c heterodimer is substantially less sensitive to Gbcs (at least to a limited subset of Gbcs that have been tested) than the p101/p110c complex [14,18]. activate p110c, p101/p110c, and p84/p110c in vitro or in transfected cells [19]. Furthermore, this interaction is important in vivo, as demonstrated by the effect of knocking-in a Ras-insensitive, but Gbc-sensitive, allele of p110c in mice (p110c DASAA/DASAA ) [20]. There are lines of evidence which suggest that Ras-GTP (and not Gbcs) regulates p84/p110c complexes in transfected cells, while p101/p110c complexes are only regulated by Gbcs [18]. This view suggests that p84/p110c complexes may be less exclusively controlled by GPCRs but rather by any ligand capable of activating Ras, including receptor tyrosine kinase pathways.

Mast Cells
Mast cells are tissue-resident immune cells that express highaffinity receptors for IgE (FceRI). Mast cell FceRI are effectively permanently bound with IgE. These FceRI/IgE complexes are activated by ligands that are specific targets for the variable regions of IgEs, including allergens. Once activated, mast cells can release a huge range of inflammatory mediators, including histamine, LTB4, PAF, and PGD2. The intracellular signals generated by activated FceRIs are primarily based on protein tyrosine kinasemediated mechanisms that would naturally lead to activation of the class IA PI3Ks such as PI3Kd [21,22]; however, PI3Kc has been shown to have an important role in mast cell activation [2]. The role for PI3Kc is through an autocrine/paracrine mechanism, involving among other things released adenosine working back on A3 (A 3 -Ars) GPCRs, that synergizes with the primary effects of FceRI activation.

Insights into Mast Cell Regulation of PI3Kc Activation
A recent study by Walser et al. in PLOS Biology began with several threads, a key one of which was the ability of thapsigargin (a drug that causes a very selective receptor-independent increase in cytosolic Ca 2+ by slowly releasing intracellular ER Ca 2+ stores into the cytosol and activating Ca 2+ influx via the store-operated-Ca 2+ -entry (SOCE) route) to stimulate phosphorylation of PKB in a PI3Kc-dependent manner. That result was very unexpected and a variety of controls suggested it was unlikely to be dependent on release of the mast cell paracrine/autocrine factor adenosine acting back on A3 GPCRs.
Wide-ranging experiments support a model (Figure 1) in which FceRI cross-linking leads to a Ca 2+ signal, involving SOCE, that drives activation of PKCb. Through an interaction between PKCb and the helical domain of p110c, S582-p110c is phosphorylated. This leads to both activation of the kinase activity of PI3Kc (which leads to increased PIP 3 and downstream signaling) and a reduction in the affinity of p110c for p84. The outcome of these events is argued to be a rebalancing in the amount of GPCR-sensitive p84/p110c and free, GPCR-insensitive p110c, thus reducing the sensitivity of PI3Kc to GPCRs. These conclusions represent major conceptual advances for the field. Beyond their major impact on our understanding of mast cell biology, they offer potential molecular explanations for the roles of PI3Kc in a number of important, but ill-understood, physiological and disease settings. Furthermore, the paper introduces beautiful structural data from deuterium-exchange protection assays that reveal where the presence of p84 masks the surface of p110c in its helical domain. Further biochemical experiments demonstrate the potential power of interactions with, or post translational Figure 1. Mast cell environment regulates PI3Kc activation. Coincident with migration and adhesion of mast cells, adaptor protein p84 relays the GPCR signal from GPCR-mediated dissociation of trimeric G proteins to activate the PI3Kc complex. However, when mast cells degranulate, FceRI receptors are clustered via IgE/antigen complexes, and a signaling cascade triggers intracellular Ca 2+ store depletion and the opening of storeoperated Ca 2+ channels. The resulting increase in calcium ion concentration and PLCc-derived diacylglycerol activates PKCb, which binds to p110c and subsequently phosphorylates Ser582. This phosphorylated p110c can no longer interact with p84, and the PI3Kc complex is therefore unresponsive to GPCR inputs. Image credit: Walser et al., 10 modifications of, the helical domain of p110s to regulate class I PI3K function both in terms of changes in catalytic activity and binding of adaptor subunits. This type of mechanism may be a widely important mode of regulation in class I PI3K signaling.
As ever with first steps, many questions remain. The role of Ras in the activation of PI3Kc in mast cells is unclear. Given that RasGrp1 (and RasGrp4) is relatively abundant in mast cells, and activated by a combination of PKC-mediated phoshorylation and diacylaglycerol (DAG) binding, and in lymphocytes is the primary mechanism driving activation of Ras downstream of antibody receptor activation [23,24], it might have been expected to be involved in this pathway, but does not appear to be. This will be clarified by testing the impact of knocking-in a Ras-insensitive version of p110c on mast cell responsiveness. The apparent contradiction between the simple model described above and the dogma that p84 is tightly bound to p110c (that is supported by the data from the deuteriumexchange protection assays) needs to be resolved. It currently leaves it difficult to understand the pattern of events; does PKCb need to compete with p84 to enable significant stoichiometries of phos-phorylation to occur? As always, a high-resolution view of the p84/ p110c complex would be immensely useful.

Concluding Remarks
The PLOS Biology paper from Walser et al. makes many important contributions. It reveals how intracellular signals from mast cell FceRIs, classically thought to engage class IA PI3K signaling, can control p110c directly via the PLC effectors Ca 2+ and DAG/PKC. These effects are mediated by an interaction between PKCb and the helical domain of p110c that enables S582-p110c to be phosphorylated and activate the lipid kinase activity of PI3Kc. The outcome is the emergence of a new mechanism by which class I PI3Ks can be activated. The paper also provides a first insight into the interactions between p110c and p84, hopefully the first step in the exploitation of deuteriumexchange methodologies to reveal more about the interactions of p101, p84, p110c, and Gbcs. Finally, the work may offer molecular explanations for some of the many poorly understood roles that PI3Kc fulfills outside of the immune system.