The Structure of PI3K

The p85 Adapter Subunit

Class IA PI3Ks are tightly associated with a 85 kDa regulatory subunit called p85.^[2] P85 contains a Src homology 3 (SH3) domain, a breakpoint-cluster region homology (BH) domain between two proline-rich regions, and two C-terminal SH2 domains separated by an inter-SH2 (iSH2) region, which tightly binds p85 to the catalytic subunit.^[3] Since PI3K has multiple protein-interaction domains, p85 is able to interact with several signaling molecules simultaneously, allowing for significant fine tuning of PI3K activity. ^[2]

Src Homology 3 (SH3) Domain

has homologues found in many intracellular signaling proteins. ^[4] It mediates protein-protein interactions by binding to proline-rich motifs in target proteins forming multimeric signaling complexes. ^[5] Of note, the SH3 domain interacts with Src (Src homology 2/α-collagen-related), CDC42GAP (Cdc42 GTPase-activating protein) and the proto-oncogene product Cbl. SH3 binds to proline rich ligands via a network of hydrophobic and hydrogen bond interactions, (3i5r).^[4]

Proline-Rich Regions

The proline rich regions which flank the BH domain are ideal ligands for various SH3 containing non-receptor protein tyrosine kinases like Src, Lyn & Fyn, often with the product of the proteo-oncogene product Cbl as a docking site. ^[6]

Src Homology 2 (SH2) Domains

, an N-terminal (nSH2) domain and a C-terminal (CSH2) domain. ^[1] Both domains recognize similar consensus phosphorylated tyrosine motifs with the pattern: pY-V-X-M in activated receptors and adaptor proteins like PDGF, erbB3, c-Kit and CSF-1 receptors. ^[7] It is upon the interaction of receptor and SH2 domain that the heterodimeric PI3K complex is activated. ^[8] nSH2 was found to interact with the catalytic subunit directly, forming a broad-based scaffold for p110α and coordinates communication between the interacting domains. (Discussed Below). ^[9]

BH Domain

The BH domain specifically interacts with the Rho family proteins, Cdc42 and Rac1. Although no crystal structure of the BH domain has been solved to date, mutagenesis experiments have verified that the conserved residues Arg 151, Lys 187 and Pro 270 play important roles in the interaction with Rac1 and Cdc42. ^[2]

Inter-SH2 (iSH2) Region

The (2v1y), is flanked by the two . The primary purpose of the iSH2 is to (3hhm), effectively holding the PI3K heterodimer together. In fact, disruption of this interaction via antibodies prevents the formation of the PI3K heterodimer completely. ^[2] It is further believed that binding of phosphopetide by the SH2 domains causes conformational strains which is ^[2]

Regulation of Class IA PI3K via p85 Phosphorylation

All PI3K catalytic subunits possess intrinsic protein serine kinase activity. PI3K regulatory subunits can be phophorylated by the catalytic subunit (p110) at specific sites. For example, phophorylation of Ser 608, a residue located in an area of the iSH2 domain that is critical for PIP2 presentation to the catalytic subunit, results in a dramatic reduction in PI3K lipid kinase activity.^[10] Additionally, tyrosines 580 and 607 can be phosphorylated upon stimulation with insulin and growth factor along with . ^[2] Phosphorylation of Tyr 688 in the CSH2 domain by Abl and Lck results in reduced affinity for phosphopeptides and subsequent activation of the catalytic domain. ^[11]

The Catalytic Subunit (P110) of Class 1 PI3Ks

The catalytic subunit, P110 has several isoforms that associate with different classes of PI3Ks. P110α, β, and δ associate with Class IA PI3Ks while p110γ associates with Class 1B PI3ks. ^[2] The an N-terminal adaptor-binding domain (ABD), a Ras binding domain (RBD) a C2 domain that likely binds to the cellular membrane, a helical domain (HD) with unknown function, and the actual catalytic kinase domain. ^[9] The actions of these domains are coordinated by the nSH2 communicating domain in p85.

Communication between nSH2 & The Catalytic Subunit Domains

(residues 340-345) is anchored into Helix α11K of the (residues 1017-1024) nSH2 interacts with the through a network of charge-charge interactions involving two loops on nSH2 (Residues 374-377 & 350-354) and C2 residues 364-371, a strong ^[9]

The , whose function isn’t thoroughly understood, interacts with nSH2 via charge interactions. The HD residue, . These residues are known hotspot mutations which are associated with various types of cancer. ^[9] This loop in which contains the hotspots (residues 542-546) is located precisely where The salt bridge formed between like PDGFR, eliminating nSH2-mediated inhibition of p110α and activating the enzyme to phosphorylate PIP2 into PIP3. The hotspot mutation at Glu 542 accomplishes the same thing by eliminating the salt bridge and uninhibiting p110α. It is the which used to convert PIP2 into PIP3. ^[9]

Model for Catalysis

Although no with bound substate analog has been solved, a model for PIP2 phosphorylation has been developed and is generally supported. ^[12] In this model, the headgroup of PIP2 is between the . This puts the 5-phosphate of PIP2 near Lys 973 and the . The and Lys 973 can bind the 4-Phosphate of PIP2 and help provide the Class I PI3Ks with their specificity for PIP2. Once PIP2 and ATP are bound, it is believed , deprotonating it at the C-3 Hydroxyl position creating a nucleophile. This nucleophile subsequently attacks the gamma phosphate of ATP producing PIP3. ^[12]