A screen for novel phosphoinositide 3-kinase effector proteins

The cell-surface receptor-regulated, class I phosphoinositide (PI)1 3-kinases play a central role in the intracellular mechanisms that regulate important biological processes including cell division, growth, motility, and metabolism (1, 2). This is emphasized by the association between defects in PI 3-kinase signaling and several disease states (3, 4). Class I PI 3-kinases exert their effects by the unique 3-phosphorylation of their phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2) lipid substrate, to produce phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P3) (5). This lipid acts either to localize and/or modulate the activity of specific effector proteins expressing appropriate PI recognition domains (1, 2, 5, 6). The PtdIns(3,4,5)P3 signal is degraded by both the 3-phosphatase PTEN (phosphatase and tensin homolog deleted on chromosome ten) (7) and by 5-phosphatases including the Src homology 2 domain containing inositol phosphatases 1 and 2 (SHIP1 and SHIP2) (8). However, although these alternative routes both serve to remove PtdIns(3,4,5)P3, they are clearly distinct functionally as impairment of the activity of each has discrete consequences, a feature that may reflect their separate lipid metabolites (9). Thus, whereas PTEN acts simply to reverse PI 3-kinase action by regenerating PtdIns(4,5)P2, 5-phosphatase activity allows the additional synthesis of phosphatidylinositol 3,4-bisphosphate (PtdIns(3,4)P2) and, through subsequent 4-phosphatase activity, of PtdIns3P. The latter lipid is also synthesized directly from phosphatidylinositol by class III PI 3-kinases and many of its functions are well established (10). By contrast, the biological functions of PtdIns(3,4)P2 remain obscure although recent evidence suggests that the specific enzymes responsible for both the synthesis and removal of this lipid fulfill important roles. Hence, the synthesis of PtdIns(3,4)P2 from PtdIns(3,4,5)P3 by the widely distributed 5-phosphatase, Src homology 2 domain containing inositol phosphatase 2 (SHIP2) appears to be tightly regulated (11) and the loss of this enzyme has important consequences (9, 12) that may reflect its influence on cellular PtdIns(3,4)P2 concentrations as well as on those of PtdIns(3,4,5)P3. Similarly, the importance of PtdIns(3,4)P2 catabolism by the types I and II inositol polyphosphate 4-phosphatases (13) is emphasized by the loss of either enzyme. The type I enzyme regulates cell growth downstream of the GATA transcription factor (14) whereas its deficit results in the neuronal loss characteristic of Weeble mice (15) and vulnerability to excitotoxic neuronal death (16). The type II 4-phosphatase may act as a tumor suppressor and recent work implies that this enzyme regulates Akt-dependent cell proliferation (17), a feature consistent with the known ability of PtdIns(3,4)P2 to bind and to regulate the activity of this pivotal serine/threonine kinase (18–22). However, PtdIns(3,4)P2 may also exert effects through other target proteins such as TAPP-1 (tandem pleckstrin homology domain containing protein-1) (23) and lamellipodin (24), which bind this lipid with high selectivity, or DAPP-1 (dual adapter for phosphotyrosine and 3-phosphoinositides) (25) and Irgm-1 (immunity-related GTPase) (26) that interact preferentially with both PtdIns(3,4)P2 and PtdIns(3,4,5)P3. Indeed, it seems likely that it is by acting through these and other similar effectors that PtdIns(3,4)P2 can act independently of, or co-ordinately with, PtdIns(3,4,5)P3 in some circumstances (16, 27). In this context, the activities of PTEN, SHIP2, and the 4-phosphatases can be viewed collectively as a means of tuning the balance of prevailing signal outputs from PI 3-kinase (28). The number of defined molecular targets for PtdIns(3,4)P2 however, remains limited by comparison with the established repertoire of PtdIns(3,4,5)P3 effector proteins. In the current study, we have sought to redress this balance by developing a screen for additional, selective PtdIns(3,4)P2 binding proteins. To do this we have adopted a novel tertiary approach. This combines a primary, selective, cellular recruitment and later recovery of candidates with a secondary, lipid-affinity and other purification of these proteins. This is coupled thirdly, with a ratio-metric, isotope labeling technique that, in conjunction with mass spectrometric analysis, allows quantitative discrimination of background proteins from those authentically responsive to PI 3-kinase/PtdIns(3,4)P2 at the primary step. By introducing multiple tiers of selectivity, including an initial cellular screen that allows the presentation of PtdIns(3,4)P2 in an authentic membrane bilayer to binding partners present in the cellular milieu, this approach aims both to complement and extend earlier, lipid-affinity based efforts to isolate PI binding proteins (29–33). The success of this strategy is demonstrated by our identification of several established 3-PI effectors in addition to several proteins whose potential regulation directly or otherwise by 3-PIs has not been recognized previously. The former demonstrate the feasibility of this approach and lend credibility to the view that the latter include novel, authentic targets whose description here will facilitate a deeper understanding of the subtleties of PI 3-kinase signaling.