Open Access Kent State (OAKS)

Phosphatidylinositol polyphosphate lipids (phosphoinositides) form only a minor pool of membrane phospholipids but are involved in many intracellular signaling processes, including membrane trafficking, cytoskeletal remodeling, and receptor signal transduction. Phosphoinositide properties are largely determined by the characteristics of their headgroup, which at physiological pH is highly charged but also capable of forming hydrogen bonds. Many proteins have developed special binding domains that facilitate specific binding to particular phosphoinositides, while other proteins interact with phosphoinositides via nonspecific electrostatic interactions. Despite its importance, only limited information is available about the ionization properties of phosphoinositides. We have investigated the pH-dependent ionization behavior of all three naturally occurring phosphatidylinositol bisphosphates as well as of phosphatidylinositol 3,4,5-trisphosphate in mixed phosphoinositide/phosphatidylcholine vesicles using magic angle spinning ³¹P NMR spectroscopy. For phosphatidylinositol 3,5-bisphosphate, where the two phosphomonoester groups are separated by a hydroxyl group at the 4-position, the pH-dependent chemical shift variation can be fitted with a Henderson−Hasselbalch-type formalism, yielding pK_a2 values of 6.96 ± 0.04 and 6.58 ± 0.04 for the 3- and 5-phosphates, respectively. In contrast, phosphatidylinositol 3,4-bisphosphate [PI(3,4)P₂] as well as phosphatidylinositol 4,5-bisphosphate [PI(4,5)P₂] show a biphasic pH-dependent ionization behavior that cannot be explained by a Henderson−Hasselbalch-type formalism. This biphasic behavior can be attributed to the sharing of the last remaining proton between the vicinal phosphomonoester groups. At pH 7.0, the overall charge (including the phosphodiester group charge) is found to be −3.96 ± 0.10 for PI(3,4)P₂ and −3.99 ± 0.10 for PI(4,5)P₂. While for PI(3,5)P₂ and PI(4,5)P₂ the charges of the individual phosphate groups in the molecule differ, they are equal for PI(3,4)P₂. Differences in the charges of the phosphomonoester groups can be rationalized on the basis of the ability of the respective phosphomonoester group to form intramolecular hydrogen bonds with adjacent hydroxyl groups. Phosphatidylinositol 3,4,5-trisphosphate shows an extraordinary complex ionization behavior. While at pH 4 the ³¹P NMR peak of the 4-phosphate is found downfield from the other two phosphomonoester group peaks, an increase in pH leads to a crossover of the 4-phosphate, which positions this peak eventually upfield from the other two peaks. As a result, the 4-phosphate group shows a significantly lower charge at pH values between 7 and 9.5 than the other two phosphomonoester groups. The charge of the respective phosphomonoester group in PI(3,4,5)P₃ is lower than the corresponding charge of the phosphatidylinositol bisphosphate phosphomonoester groups, leading to an overall charge of PI(3,4,5)P₃ of −5.05 ± 0.15 at pH 7.0. The charge of all investigated phosphoinositides at pH 7.0 is equal or higher than the corresponding charge of soluble inositol polyphosphate headgroup analogues, which is the opposite of what is expected on the basis of simple electrostatic considerations. This higher than expected headgroup charge can be rationalized with mutual intermolecular hydrogen bond formation. Measurements using different concentrations of PI(4,5)P₂ in the lipid vesicles (1, 5, and 20 mol %) did not reveal any significant concentration-dependent shift of the two phosphomonoester peaks, suggesting that PI(4,5)P₂ is clustered even at 1 mol %.