Despite unusually divergent primary sequences, PH domains share a conserved fold made up of a b-barrel composed of two roughly perpendicular, anti-parallel beta-sheets and a C-terminal alpha amphipathic -helix. PH domains bind to their inositol phosphate ligands via a binding surface composed primarily of residues from the beta1/beta2, beta3/beta4, and beta6/beta7 loops. Basic residues in the beta1-beta2 loop are especially important for binding the phosphatidylinositide phosphates, by establishing a positive electric potential on the face of the PH domain. Comparison of different PH domains with distinct phosphoinosite binding specificities has revealed insights into what structural features dictate binding specificity. For example, a tyrosine residue located in the beta-strand is strictly conserved in PH domains that bind PI3- kinase products with high affinity but not in other PH domains. Interestingly, PH domains have a strikingly similar structure to PTB domains, the Enabled / Vasp homology (EVH1/WH1) and the Ran-binding domain (RanBD) of RBP2, despite the absence of any sequence homology. The figure shows the complex between the PH domain of Phospholipase C-d and inositol-(1,4,5)-trisphosphate (red).

Structure Reference:
Ferguson, KM. et al. (1995) Cell. 83(6): 1037-46. PDB: 1MAI.