PI3K signalling in B- and T-lymphocytes: new developments and therapeutic advances

Abstract

Activation of PI3K (phosphoinositide 3-kinase) is a shared response to engagement of diverse types of transmembrane receptors. Depending on the cell type and stimulus, PI3K activation can promote different fates including proliferation, survival, migration and differentiation. The diverse roles of PI3K signalling are well illustrated by studies of lymphocytes, the cells that mediate adaptive immunity. Genetic and pharmacological experiments have shown that PI3K activation regulates many steps in the development, activation and differentiation of both B- and T-cells. These findings have prompted the development of PI3K inhibitors for the treatment of autoimmunity and inflammatory diseases. PI3K activation, however, has both positive and negative roles in immune system activation. Consequently, although PI3K suppression can attenuate immune responses it can also enhance inflammation, disrupt peripheral tolerance and promote autoimmunity. An exciting discovery is that a selective inhibitor of the p110δ catalytic isoform of PI3K, CAL-101, achieves impressive clinical efficacy in certain B-cell malignancies. A model is emerging in which p110δ inhibition disrupts signals from the lymphoid microenvironment, leading to release of leukaemia and lymphoma cells from their protective niche. These encouraging findings have given further momentum to PI3K drug development efforts in both cancer and immune diseases.

Figure 1. Domain structure of class IA and class IB PI3K isoforms

Black arrows indicate constitutive interactions of the heterodimers. Dashed red arrows indicate inhibitory interactions between the regulatory and catalytic isoforms of class IA PI3K that maintain low basal activity of the enzyme. Phosphorylation of tyrosine (Y) residues on the conserved pY-X-X-M motifs on various receptors or adaptor proteins (CD19/BCAP for B cells, CD28/ICOS for T cells) recruits the regulatory isoforms via the two SH2 domains and this binding releases the inhibitory interactions. Other protein:protein interactions also contribute to class IA and IB recruitment and activation.

Figure 2. PI3K engagement in B lymphocytes and the rheostat concept

BCR engagement triggers tyrosine (Y) phosphorylation on CD19 and BCAP to recruit PI3K dimers mainly consisting of p85α and p110δ. PI3K activation promotes signalosome assembly for Ca2+ mobilization and diacylglycerol (DAG) production, and increases activity of AKT. BCR dependent Ca²⁺ flux, AKT activation and proliferation is mainly dependent on PIP₃ pools generated by dimers of the p110δ catalytic isoform with the p85α regulatory isoform. Cytokine (BAFF and IL-4) dependent survival signals require p110α as well, which might generate distinct pools of PIP₃ as shown. Two parallel membranes drawn in light brown represent a 3-dimensional cell surface rather than distinct membranes. In vivo, overall PI3K activity serves as a ‘rheostat’ (gray circle) whose signal output strength determines the nature of the response. In B cells, high PI3K activity opposes class switch recombination and promotes plasma cell differentiation. Low PI3K activity promotes class switch recombination, with p110δ inhibition selectively augmenting IgE production in mice. A similar rheostat concept applies in CD4 T cells to generate the variety of different subsets required depending on the immune context.

Figure 3. Role of PI3K and downstream effectors in the differentiation of activated CD4 T lymphocytes

Antigen encounter triggers clonal expansion of naïve CD4 T cells that is mainly dependent on p110δ. Depending on the immune microenvironment (cytokines, etc.), primed CD4 T cells differentiate into different CD4 T cell subsets (Th1, Th2, Th17, Tfh, and iTregs). TORC2-dependent AKT activation promotes Th1 differentiation through TORC1 and NFκB, and TORC2-dependent PKCθ activation promotes Th2 differentiation. PDK-1 also triggers NFκB activation through a complex with PKCθ. AKT increases Th17 differentiation through TORC1 and HIF-1α activation. Already established CCR6+ human memory Th17 cell function (IL-17/IL-22 secretion) also depends on PI3K/AKT activity via modulation of FOXO and its target genes including KLF2. ICOS stimulation triggers Tfh differentiation through a p110δ-dependent pathway, with a speculative role for p110α binding to ICOS. The total PI3K output acts as a rheostat (gray circle) with decreased PI3K/AKT activity favoring iTreg induction. Isoforms other than p110δ including p110α probably suppress iTreg induction in this context. PI3K/AKT blockade results in the induction of FOXO-mediated FoxP3 expression and low TORC1 activity leads to decreased metabolic activity via S6K and HIF-1α. Established FoxP3+ Tregs also depend on p110δ activity for proper suppressive function such as IL-10 secretion. Due to space constraints, the nucleus is not depicted where all of the genes are regulated.

Figure 4. A model for the efficacy of a p110δ inhibitor in CLL

The two illustrations show a lymph node containing CLL tumor cells before treatment (Left panel) or after treatment with the p110δ inhibitor, CAL-101 (Right panel). Chemokines drive the entry of CLL cells and other lymphocytes into the lymph node from blood vessels via the high endothelial venules (HEV). Cells exit the lymph node via the efferent lymph. The legend at the lower right explains the arrows and symbols. The scale at the upper right depicts the differential p110δ enzyme activity in cells shown in the figure. NLC = nurse-like cells. Th = T helper. Other stromal cells also interact with CLL cells in the tumor cell niche.