The Complex and Intriguing Lives of PIP2 with Ion Channels and Transporters

See allHide authors and affiliations

Science's STKE  04 Dec 2001:
Vol. 2001, Issue 111, pp. re19
DOI: 10.1126/stke.2001.111.re19


Phosphatidylinositol-4,5-bisphosphate (PIP2), the precursor of several signaling molecules in eukayotic cells, is itself also used by cells to signal to membrane-associated proteins. PIP2 anchors numerous signaling molecules and cytoskeleton at the cell membrane, and the metabolism of PIP2 is closely connected to membrane trafficking. Recently, ion transporters and channels have been discovered to be regulated by PIP2. Systems reported to be activated by PIP2 include (i) plasmalemmal calcium pumps (PMCA), (ii) cardiac sodium-calcium exchangers (NCX1), (iii) sodium-proton exchangers (NHE1-4), (iv) a sodium-magnesium exchanger of unknown identity, (v) all inward rectifier potassium channels (KATP, IRK, GIRK, and ROMK channels), (vi) epithelial sodium channels (ENaC), and (vii) ryanodine-sensitive calcium release channels (RyR). Systems reported to be inhibited by PIP2 include (i) cyclic nucleotide-gated channels of the rod (CNG), (ii) transient receptor potential-like (TRPL) Drosophila phototransduction channels, (iii) capsaicin-activated transient receptor potential (TRP) channels (VR1), and (iv) IP3-gated calcium release channels (IP3R). Systems that appear to be completely insensitive to PIP2 include (i) voltage-gated sodium channels, (ii) most voltage-gated potassium channels, (iii) sodium-potassium pumps, (iv) several neurotransmitter transporters, and (v) cystic fibrosis transmembrane receptor (CFTR)-type chloride channels. Presumably, local changes of the concentration of PIP2 in the plasma membrane represent cell signals to those mechanisms sensitive to PIP2 changes. Unfortunately, our understanding of how local PIP2 concentrations are regulated remains very limited. One important complexity is the probable existence of phospholipid microdomains, or lipid rafts. Such domains may serve to localize PIP2 and thereby PIP2 signaling, as well as to organize PIP2 binding partners into signaling complexes. A related biological role of PIP2 may be to control the activity of ion transporters and channels during biosynthesis or vesicle trafficking. Low PIP2 concentrations in the secretory pathway would inactivate all of the systems that are stimulated by PIP2. How, in detail, is PIP2 used by cells to control ion channel and transporter activities? Further progress requires an improved understanding of lipid kinases and phosphatases, how they are regulated, where they are localized in cells, and with which ion channels and transporters they might localize.


Most of us were introduced to cell signaling by adenosine 3′,5′-monophosphate (cAMP), a simple and elegant second messenger that transmits extracellular signals into the cell by activating protein kinases (1). Next came inositol 1,4,5-trisphosphate (IP3), which transmits another set of hormonal messages into the cell by Ca2+ release from intracellular stores (2). IP3 also seemed simple and elegant, but it was slightly disconcerting that a complex membrane phospholipid, phosphatidylinositol-4,5-bisphosphate (PIP2), would be used to generate a water-soluble messenger. Why should phospholipids be involved? The answer sacrificed simplicity, but it maintained elegance. The phospholipid product of PIP2 hydroysis by phospholipase C (PLC), diacylglycerol (DAG) (3), turned out to be a membrane-associated messenger that activates a family of protein kinases and other signaling molecules (4). Then came phosphatidylinositol 3,4,5-trisphosphate (PIP3), generated by phosphorylation of PIP2 at the 3′ position (5), and with it came surprising new insights into the regulation of cell proliferation (6, 7), cell growth (8), and apoptosis (9), as well as some forms of membrane insertion (that is, vesicle fusion) (10) and cell movement (11).

With simplicity forever lost, elegance now also seems to be on hold. PIP2 itself has emerged as a major signaling molecule. At minimum, PIP2 is a cofactor, an anchor, and a modulator of numerous membrane-associated signaling molecules and the cytoskeleton (12). Perhaps PIP2 was one of the first eukaryotic signaling molecules, closely tied to evolution of the secretory pathway (13). In any case, numerous groups are providing glimpses of PIP2 playing diverse roles in membrane trafficking (14-18), regulation of the cytoskeleton (19-21), and regulation of ion transporters and channels (22). This Review focuses on the regulation of ion channels and transporters, first with an update of the various systems found to be PIP2-sensitive, and then with a discussion of the possible roles that PIP2 is playing.

PIP2: Gregarious But Not Promiscuous

Reconstitution experiments revealed long ago that some membrane transporters, notably plasmalemmal Ca2+ (PMCA) pumps (23) and Na+/Ca2+ exchangers (24), are strongly activated by anionic phospholipids. When purified transporter proteins were incorporated into vesicles composed of purified phospholipids, transporter activity was enhanced by, or even required, the presence of anionic phospholipids with long acyl chains, preferably with double bonds (25). This seemed at first like a "constitutive" effect, but some phospholipids were clearly more effective than others were. Penniston proposed that PIP2 could be a physiological activator of the PMCA pumps (26), and the suggestion has been supported recently for PMCA pumps in plants (27). Luciani and colleagues made a similar proposal for Na+/Ca2+ exchangers (28).

In experiments with excised membrane patches, the activation of Na+/Ca2+ exchangers (29) and some potassium channels (30) by cytoplasmic adenosine triphosphate (ATP) was related to the generation of anionic lipids, and the underlying mechanism turned out to be the phosphorylation of phosphatidylinositol (PI) by lipid kinases to generate PIP2 (30). Because these results were obtained with native cell membranes, and the interpretation did not rely on effects of exogenous phospholipids, the results suggested that PIP2 is a physiologically important ligand of both membrane transporters and ion channels.

For each of the three transport systems just mentioned, Ca2+ pumps of the PMCA type (31), Na+/Ca2+ exchangers of the NCX type (32), and inward rectifier potassium channels (IRK) (33, 34), cationic domains have been identified on the cytoplasmic side of the protein that bind PIP2 and probably mediate its functional effects. The sites in Na+/Ca2+ exchangers called "XIP" domains have homology to calmodulin-binding domains of myosin light chain kinase, but they do not bind calmodulin. All of the cationic domain sites identified to date that bind PIP2 contain multiple cationic residues with interspersed hydrophobic residues. They have no obvious homology to the PIP2-binding pleckstrin homology (PH) domains (35), and they are usually less selective than PH domains, because other anionic phospholipids at high concentrations can substitute for PIP2. Why do they bind PIP2 preferentially? On the basis of electrostatics alone, nonspecific cationic binding sites at the protein-membrane interface of a native membrane would probably bind PIP2 preferentially. PIP2 is a trivalent ion at pH7. On the basis of electrostatic interactions alone, it will bind to polyvalent cationic sites with about 100-times higher affinity than monovalent phospholipids will bind to the same sites. PIP2 constitutes about 1% of anionic phospholipids in most cells, whereas the abundance of PIP3 probably never exceeds a few percent of the level of PIP2 (36). Thus, modest evolutionary improvements of nonspecific cationic binding sites might have achieved fairly specific binding sites for PIP2 in a physiological phospholipid milieu. Of course, each case must be analyzed carefully, and, as outlined below, certain ion channels are activated highly specifically by PIP2.

Over the last few years, several groups have probed a wide range of transporters and channels for PIP2 sensitivity (Fig. 1). Tests that can determine the functional dependence of these molecules on changes of PIP2 include the following: (i) a PLC can deplete membranes of PI and thereby remove substrate for the generation of PIP and PIP2 (30); (ii) PIP2-specific PLCs can specifically deplete PIP2 from the membrane (30, 37); (iii) PIP2 and most phospholipids can be incorporated into excised patches, because they bind to the pipette and diffuse into the cytoplasmic membrane face (30, 33, 38); (iv) membrane-targeted PIP2-specific phosphatases can be expressed in cells to deplete PIP2 (39); and (iv) PIP2-specific ligands, including PH domains and PIP2 antibodies, can be applied to functionally deplete or sequester PIP2 (33).

Fig. 1.

The major membrane transporters and channels that have been found to be PIP2-sensitive. Membrane proteins in the diagram to the left of PIP2 are stimulated: PMCA, the plasmalemmal calcium pump; NCX, the cardiac Na+/Ca2+ exchanger; NHE, a Na+/H+ exchanger; a Na+/Mg2+ exchanger of unknown identity; ROMK, epithelial potassium channels; IRK, classical inward rectifier potassium channel; KATP, ATP-inhibited potassium channels; GIRK, G protein-activated IRK; ENaC, epithelial Na+ channels; and RyR, ryanodine-sensitive calcium release channels. Proteins depicted to the right of PIP2 are inhibited: CNG, cyclic nucleotide-gated channels of the rod; TRPL, Drosophila phototransduction channels; VR1, capsaicin-activated transient receptor potential (TRP) channels; and IP3R, IP3-gated calcium release channels. ATP, adenosine triphosphate; ADP, adenosine diphosphate; Pi, inorganic phosphate.

Ion Transporters

Na+/H+ exchangers, like Na+/Ca2+ exchangers, are activated by ATP-dependent processes. Grinstein and colleagues have shown that PIP2 synthesis underlies the ATP-dependence of Na+/H+ exchangers, at least in part, and that PIP2 depletion by multiple means strongly inhibits Na+/H+ exchangers (39). As with the Na+/Ca2+ exchangers, cationic domains on the cytoplasmic side of Na+/H+ exchangers are implicated in PIP2 binding and mediating exchanger activation. Na+/Mg2+ exchangers of still-unknown physical identity mediate Mg2+ transport in epithelia and control the cytoplasmic free magnesium concentration in many cell types. In dialysis experiments using giant barnacle muscle fibers, Rasgado-Flores and colleagues have shown that PIP2 is a strong activator of Na+/Mg2+ exchange (40).

Ion-Selective Channels

All of the background potassium channels, called "inward rectifier potassium" (that is, the IRK channels), are apparently bound and activated by PIP2 (33). Despite this first impression of "ubiquity," there is need for caution in too readily accepting this impression (41). For the ATP-inhibited potassium (KATP) channels, there are hints that phosphatidic acid (42) and PIP3 (43), rather than PIP2, may be important physiologically. In the Xenopus oocyte expression system, PIP2 appears to be a requirement for opening of several IRK channels, but the requirement may not be absolute in native membranes (44). It is probable that PIP2 binds to the COOH-terminal cytoplasmic domain, in close proximity to domains where ATP binds in KATP channels, and where G protein βγ-subunits bind in G protein-activated IRK (GIRK) channels (45, 46). The Kir2.1 channels of the inward rectifier family do not bind ATP or heterotrimeric guanosine triphosphate (GTP)-binding proteins (G proteins); instead, the equivalent regulatory domain is replaced by a second PIP2-binding domain (46).

Of several voltage-gated K+ channels tested in our lab, the only PIP2-sensitive channels that we have identified are the Kv2.1 channels. They are activated by PIP2, and PIP2 can completely prevent their otherwise rapid run-down in excised patches (114). Interestingly, these channels may localize to so-called "lipid rafts" (47), putative phospholipid microdomains that, as discussed subsequently, are thought to be enriched in PIP2. The Na+-selective ENaC channels of epithelial cells are reported to be activated by PIP2 in the presence of a G protein α subunit (Gα), and PIP2 binding evidently occurs on the β subunit of the channel (48).

Sensory Ion Channels

A number of ion channels involved in sensation are PIP2 sensitive. Olfactory receptor neurons in the lobster express a nonselective, Na+-activated cation channel that carries a substantial part of the depolarizing receptor current and is PIP2 activated (49). The cyclic nucleotide-gated channels of the rod (CNG) photoreceptor, but not the related olfactory ion channels, are strongly inhibited by PIP2 when they are expressed in Xenopus oocytes (50), and the effects of PIP2 are not mimicked by any other anionic phospholipid. Indeed, a highly specific PIP2 site would be required to mediate PIP2 effects in the rod outer segment, because PIP2 levels are substantially lower in rod outer segments than in other cell types (36). Some members of the transient receptor potential (TRP) channel family appear to be regulated by PIP2. For example, the Drosophila TRP-like (TRPL) channels, which carry much of the phototransduction current in the Drosophila complex eye, are inhibited by PIP2 when they are expressed heterologously (51). Another type of TRP channel, the capsaicin-activated VR1 channel, also appears to be strongly inhibited by PIP2, in that channel activity can be strongly increased by treatment of patches with PLCs or PIP2 antibodies (37).

Intracellular Calcium Release Channels

Both ryanodine receptors and IP3 receptors are reported to be PIP2 sensitive. The activity of ryanodine receptors, when incorporated into membrane bilayers, can be markedly increased by inclusion of PIP2 in the bilayer (52). In fragmented sarcoplasmic reticulum vesicles, as well as skinned skeletal muscle fibers, PIP2 at submicromolar concentrations can cause calcium release (53,54), enhance ryanodine binding (53), and enhance caffeine-induced calcium release (55). When IP3 receptors from brain microsomes are incorporated into membrane bilayers, their activity and sensitivity to IP3 are strongly enhanced by a PIP2 antibody, and their activity is strongly inhibited by incorporating PIP2 into the bilayers (56, 57). Presumably, PIP2 is bound by the channels with very high affinity, because it can remain bound during channel isolation and incorporation into phospholipid bilayers. IP3 receptors from the nuclear envelope of Xenopus oocytes can be studied using patch clamp techniques (58), and they display a very high activity and IP3 sensitivity that is reminiscent of channels in bilayers treated with PIP2 antibody. Possibly, then, PIP2 is depleted by physiological mechanisms occurring in the oocyte. How can one demonstrate that PIP2-dependent inhibition of IP3 channels is indeed physiologically important? Circumstances must be identified in which Ca2+ release is regulated primarily by a change of the concentration of PIP2, rather than IP3.

Obviously, the range of mechanisms that are sensitive to changes of PIP2 is very broad, from both a structural and a physiological perspective. In some cases, more rigorous tests for sensitivity to PIP2 are still required. In the case of calcium release channels, the results described do not provide any evidence that PIP2 is a physiological channel modulator (59). Certainly, many transporters and channels, probably the majority, are completely PIP2 insensitive. From unpublished results of this lab, PIP2-insensitive mechanisms include Na+/K+ pumps, most voltage-gated K+ channels, voltage-gated Na+ channels, and probably voltage-gated Ca2+ channels, the γ-aminobutyric acid (GABA) transporter GAT1, and the glutamate transporter EAAT. The remainder of this Review examines the clues that are currently available to suggest how PIP2-sensitivity might be used to physiologically regulate ion channel and transporter activities.

The Excitement and Pain of PIP2

Regulation of the transcription factor Tubby (60) by PIP2 can be viewed as a model for localized PIP2-dependent regulation of ion channels. Tubby specifically binds PIP2 and is thereby anchored to the surface membrane. Tubby is released from the membrane, to enter the nucleus, by the activation of a PLC-β. To achieve signaling specificity, one may speculate that a ternary complex forms between Tubby, PIP2, and the PLC, within which PIP2 can be cleaved. That hypothesis remains to be proved. Ternary complexes formed between several PIP2-binding proteins, PIP2, and a PIP2 antibody can be seen as a precedent for such interactions. The PIP2-binding protein, PIP2, and PIP2 antibody complexes remain intact during sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot analysis (61, 62).

How localized can PIP2 signaling at ion channels be? Total PIP2 in cells amounts to 5 to 25 μm/liter cell water (36, 63), and it seems likely that the majority is bound and does not diffuse freely in the membrane (64). One speculation then is that PIP2-sensitive ion channels and transporters interact with other PIP2-binding proteins, possibly in signaling complexes that include PLCs, lipid kinases, or lipid phosphatases, or any combination of the three. In the extreme case, PIP2 and its metabolites could function as membrane-delimited cofactors for channel signaling, analogous to GTP and guanosine diphosphate (GDP) in G protein signaling, with other PIP2-binding proteins playing the roles of PIP2 exchange proteins.

One of the most tantalizing cases in which localized PIP2 signaling may regulate ion channels is the phototransduction process of the invertebrate complex eye (63), a signaling cascade that takes place within signaling complexes organized by a PDZ domain protein called INAD in the rhabdomere (65). The cascade starts with rhodopsin absorbing a photon, which is followed by activation of a G protein, a PLC, and then ion channels that depolarize the surface membrane. The depolarizing response in Drosophila is mediated by two ion channels, TRP and TRPL, which are founding members of the greater TRP channel family (66). Although PLC activation initiates the cascade, the photoresponse remains normal after targeted deletion of the Drosophila IP3 receptor (10). Thus, IP3 is not expected to play a second-messenger role. One alternative mechanism (to which this discussion will be limited) is that phototransduction channels are both inhibited by PIP2 and activated by DAG. The attraction of this functional antagonism is that PLC activation would then simultaneously remove an inhibitor and supply an activator to the channel.

Unfortunately, it has been difficult to test this hypothesis. The architecture of the rhabdomere does not allow effective dialysis of the cytoplasm through a patch pipette, nor does it allow excision of patches with phototransduction channels. In addition, the Drosophila TRP channels do not express well in cell cultures. The native phototransduction channels are activated by metabolic inhibition, which could correspond to PIP2 depletion (67). As already mentioned, the heterologously expressed TRPL channels can be inhibited by PIP2, and they can be activated to some extent by DAG analogs (51). DAG and DAG analogs activate a number of TRP-like channels (68, 69). The native phototransduction channels are also strongly activated by unsaturated fatty acids (70) [such as arachidonic acid (AA) and linoleic acid], but the relationship of those effects to the phototransduction response is not clear. For TRPL channels expressed in SF9 cells, activation by fatty acids is reportedly blocked by PLC inhibitors (51).

Recent studies of Hardie and colleagues demonstrate that profound PIP2 changes can indeed occur in the rhabdomere during photoresponses (71). Briefly, a functional assay for PIP2 changes was developed by generating transgenic flies in which PIP2-activated K+ channels (Kir2.1) were genetically targeted to rhabdomeres. Currents generated by the expressed K+ channels then track, at a first approximation, changes of PIP2 concentration within the rhabdomere. The results indicate that PIP2 can be largely depleted from the entire rhabdomere within 1 s during the response to high-intensity light, and that Ca2+ influx is a critical signal to resynthesize PIP2. Furthermore, the second class of phototransduction channels, the TRP channels, inactivate over the time course of PIP2 depletion, and PIP2 depletion appears to control their inactivation. Because the Kir2.1 K+ channels probably do not localize directly to the PLC-containing (INAD) complexes, PIP2 depletion in the immediate vicinity of the phototransduction channels must be taking place faster than represented by the K+ current changes. Thus, the results are entirely consistent with PIP2 depletion being a major factor in the activation of TRPL channels, presumably together with activation by DAG. Clearly, firm conclusions require new transgenic models, which can be expected in the very near future.

A vertebrate sensory system in which a role of PIP2 is implicated is the sensation of pain and heat. Vanilloid receptors (VR) are ion channels of the greater TRP family that appear to be involved in multiple aspects of pain sensation (72). The channels designated VR1, which can be activated with capsaicin, have been shown in knockout experiments to mediate, at least in part, the sensation of both pain and heat in mice (73). These channels open in a temperature-sensitive manner, over the range of 40 to 50°C. They also open in response to extracellular acidification. Recent work by Julius and colleagues now suggests that the activation of these channels is augmented when PIP2 is cleaved, and that this mechanism may explain why the experience of pain is potentiated during tissue injury (37). Briefly, multiple hormones that may be released during tissue injury sensitize the VR1 channels so that they open with smaller increases of temperature or extracellular proton concentrations. In excised patches, both cleavage of PIP2 with an exogenous PLC and sequestration of PIP2 with a PIP2 antibody sensitize the channels so that they open more readily (37).

On the basis of coexpression experiments in cultured cells, this mechanism can explain sensitizing effects of bradykinin and nerve growth factor (NGF) on VR1 channels (37). When hormone receptors are expressed with the channels, these two hormones can sensitize the VR1 channels to open more readily, and these effects are not blocked by inhibitors of PKCs. Further experiments suggest the following model: When NGF binds to its receptor, PLC-γ is recruited to the tyrosine kinase receptor TrkA, where the VR1 channel already resides as part of a complex. Within this signaling complex, or in its immediate vicinity, PIP2 is cleaved and depleted, thereby sensitizing the VR1 channels. Given the medical importance of pain, this model will be scrutinized closely. First and foremost, verification of the model in a sensory neuron will be essential. Tyrosine kinases are typically connected to multiple signaling pathways, including both PKC-dependent and PIP3-dependent pathways, which could still be important for the physiological responses in native neurons. Direct effects of DAG on VR1 channels also appear possible, because both phorbol esters and a PI-specific PLC, which generates DAG from PI, are noted to potentiate VR1 activity (37). As with TRPL channels, a functional antagonism between PIP2 and DAG might be at play, and the possibility of "on-site" PIP2 cleavage deserves further attention.

Just Setting the Mood for Signaling?

PIP2 never seems to be alone. It is always there with a signaling partner, sometimes in a synergistic and sometimes in an antagonistic relationship. In the following cases, PIP2 interacts with a partner on the same molecule. For the cardiac Na+/Ca2+ exchanger, PIP2 increases the apparent affinity of a regulatory binding site for Ca2+, which activates the exchanger (22). An increase of PIP2, and therefore exchange activity, would be expected to increase both the Ca2+ influx and extrusion functions of the exchanger; upon depolarization, the exchanger brings Ca2+ into heart cells, in parallel with Ca2+ channels, and it extrudes Ca2+ during relaxation, in competition with intracellular Ca2+ pumps. For KATP channels, PIP2 strongly decreases the apparent affinity of channels for ATP, which inhibits the channel (38, 74). An increase of PIP2 would cause channels to open more quickly during metabolic inhibition and potentially would activate significant numbers of KATP channels without metabolic inhibition. For GIRK (specifically, Kir3) channels, PIP2 synergizes with the βγ-subunits of G proteins in channel activation (33). Thus, in the intact heart, an increase of PIP2 in pacemaking cells would be expected to potentiate the decrease of heart rate that occurs when acetylcholine is released and activates GIRK channels. AA is another ligand at some Kir channels, and it can be an activator (68) or an inhibitor (75). AA inhibits GIRK channels specifically by suppressing long channel openings that are induced by ATP-dependent stimulation of the channels (75), which presumably reflects PIP2 generation (33, 75). A mutagenesis study suggests that the functional antagonism between PIP2 and AA (or one of its metabolites) actually reflects a physical overlap of binding sites for AA and PIP2 on the channel (76). One implication is that the characteristic prolonged openings of highly activated cardiac GIRK channels may reflect the kinetics of PIP2 dissociation from the channels.

In vertebrate phototransduction, two synergistic effects of PIP2 have been identified, an inhibition of guanosine 3′,5′-monophosphate (cGMP)-gated ion channels and a strengthening of phosphodiesterase activation (50). As a reminder, vertebrate phototransduction involves the activation of phosphodiesterases to deplete cGMP and thereby close cGMP-activated ion channels, which results in membrane hyperpolarization. Thus, these two effects would presumably strengthen the photoresponse when PIP2 is increased, and weaken the response when PIP2 is depleted.

How are these modulatory effects put to use in cells? This question really cannot be answered until we know much more about the regulation of PIP2 in the surface membrane of the various cells of interest. At minimum, this requires knowing how lipid kinases, lipid phosphatases, and lipid transfer proteins, in addition to PLCs, are regulated. Growth factors (77) and small GTPases (guanosine triphosphatases) (78), cell attachment and cell-cell interactions (79), changes in cell volume (80, 81), cell differentiation state (82), and cell stress (83, 84) affect PIP2 levels. In general, many mechanisms are thought to involve the regulation of lipid kinases. Clear cause-and-effect relations have not been established with respect to ion channel and transport activity in any of these cases. A first "hit" may be the activation of Na+/H+ exchange (85) and Na+/Ca2+ exchange (86) that occurs with cell shrinkage. The activation of Na+/H+ exchangers may be important for regulating cell volume, and the response does not involve classical shrinkage-activated signaling pathways (85). Large and rapid increases in the concentrations of PIP2 and PIP occur in a wide range of cell types (87), including plants (74). The underlying mechanisms, which remain to be clarified, could therefore be key to understanding cell volume regulation.

To date, most suggestions as to how changes of PIP2 might regulate ion channels and transporters are related to the activation of PLCs. For example, a desensitization or "fade" in GIRK channel currents is observed during prolonged muscarinic receptor activation in isolated sinoatrial node cells (88) and atrial myocytes (89), and the fade has been ascribed to PIP2 depletion as a result of PLC activation through Gq-coupled muscarinic receptors (89, 90). Although desensitization can occur by additional mechanisms (91, 92), there is agreement that PIP2 depletion during Gq-coupled PLC activation can occur and is inhibitory. Pharmacological studies suggest that inhibition of GIRK channel currents by α-adrenergic receptor stimulation may also be mediated by PIP2 deletion (93), and in rat ventricle, α-adrenergic receptor stimulation can indeed cause PIP2 depletion (94). In Cos cells, the activation of overexpressed M1 receptors can evidently inhibit KATP channels through PIP2 depletion (95).

Whether or not total cellular PIP2 decreases significantly during Gq activation depends strongly on cell type and cell state. In tracheal muscle, for example, contraction is activated and maintained by the release of Ca2+ from intracellular stores. For a maintained contraction, the opening of IP3-gated channels must be maintained, and phospholipases are presumably more active than other cell types. Thus, PIP2 depletion can be substantial during agonist exposure in tracheal smooth muscle (96), in spite of the fact that PIP2 synthesis is actually enhanced as part of the agonist response (97, 98). Interestingly, the PIP2 depletion is maintained with prolonged agonist exposure, whereas IP3 levels decline; thus, it is possible that the depletion of PIP2 helps to maintain the IP3 receptors in an activated state by releasing them from PIP2-dependent inhibition.

In intact cardiac tissue, PIP2 levels are 5 to 10 times higher than in isolated myocytes (36, 99), possibly due to a loss of extracellular matrix-dependent or cell-cell contact-dependent signaling of lipid kinases in the myocyte preparations (7, 99). In fact, PIP2 in intact atrial tissue may not decrease at all during prolonged activation of muscarinic receptors (100). In intact heart, the response to repeated vagal nerve stimulation fades because of depletion of acetylcholine in nerve terminals, not depletion of PIP2 (101). As expected, if PIP2 does not decrease in intact tissue during continued muscarinic stimulation, then action potential shortening in superfused atria (94) and decrease of frequency in arterially perfused atria (101) are entirely stable tissue responses. Whether or not "fade" is an artifact of the isolated myocyte, it seems clear that PIP2 levels in the heart are less dependent on PLC activity than in tracheal muscle or the rhabdomere of the Drosophila complex eye. Obviously, there is a great need to understand how and whether lipid kinases and phosphatases are regulated in cardiac cells, because they may be the primary regulators of PIP2.

Controlling Migrants from the Secretory Pathway

Lipid phosphorylation was first demonstrated to change in connection with hormone-stimulated secretion and thereby membrane insertion during vesicle or granule fusion (102). The regulation of membrane trafficking and related cytoskeletal changes by phosphatidylinositides is the now the subject of intense investigation, and multiple recent reviews of the subject are available (17, 103-106). On one hand, the generation of PIP2 is closely connected with membrane insertion at the cell surface, and on the other hand, recruitment of PIP kinases and PIP2-binding proteins to the surface membrane initiates membrane retrieval (107); whether PIP2 is dephosphorylated before or after retrieval is not certain. Although PIP2 can be synthesized on internal membranes, including the Golgi, its concentration there appears to be low compared to that on the surface membrane. Thus, ion channels and transporters that are activated by PIP2 will probably be inactive during their processing and passage through the secretory pathway. If organelles of the secretory pathway require a membrane potential, regulate their internal ion concentrations, or both, in addition to maintaining an acidic interior (108, 109), then constitutively active potassium channels (such as IRK channels) and sodium channels (such as ENaC channels) would probably be toxic. Conversely, mechanisms that maintain a low internal pH in the secretory pathway, namely, V-type ATPases (adenosine triphosphatases) and "ClC"-type chloride channels (109), must be presumed to appear in the surface membrane during constitutive membrane cycling. The high PIP2 concentration they experience in the surface membrane might serve as an inactivation signal.

PIP2: A Membrane Organizer

What other roles might the PIP2 dependence of channels and transporters play? (i) Some data suggest that PIP2 is generated and maintained mostly in so-called "lipid rafts," which are membrane domains high in cholesterol and sphingomyelin (110-113). If these structures really exist in intact cells, then ion channels and transporters with PIP2 binding sites might tend to cluster and be activated in such domains. (ii) If ternary complexes between PIP2 and two PIP2-binding proteins are common, then PIP2 might serve to adhere ion channels with PIP2-binding signaling molecules or PIP2-binding cytoskeletal proteins. (iii) Hydrophobic signaling molecules and drugs may act by modifying PIP2-channel or PIP2-transporter interactions. In our experience, a wide range of cationic, hydrophobic molecules strongly decrease activation of Na+/Ca2+ exchange and KATP channels by PIP2, perhaps by sequestering PIP2 away from the transporters into a different membrane microdomain. Sphingosine and lidocaine are biologically and pharmacologically interesting examples, inhibiting Na+/Ca2+ exchange in submicromolar concentrations from the cytoplasmic side (114). Fatty acids are converted to acylcarnitines for transport into mitochondria, and inhibition of PIP2-sensitive IRK channels by acylcarnitines (116), as well as Na+/Ca2+ exchangers, might contribute to their pathological effects during ischemia. Some anesthetics and alcohols are also suggested to disrupt PIP2 interactions with potassium channels, in particular the GIRK channels (116).

The Challenge to Settle Down

In summary, PIP2 binds to and modulates the function of numerous ion channels and transporters. PIP2 may be a membrane-restricted second messenger that is used in sensory signaling cascades to regulate the activity of ion channels or transporters. It may also connect ion channels and transporters to important signaling mechanisms besides those immediately involved in hormone responses, in particular, cell signaling changes induced by cell proliferation, cell-cell contact, and mechanical perturbation. PIP2 may help to localize signaling and regulate the activity of ion transporters and channels in a location-dependent fashion. In terms of cellular homeostasis, differences in the levels of PIP2 in internal compared to surface membranes may help to control the activity of proteins as they progress through the secretory pathway to the surface membrane. For the moment, we do not know how, in detail, PIP2 is used to regulate ion channel or ion transport activity. Real progress requires new experimental models that can provide insights into the regulation of lipid kinases and phosphatases, as well as the lateral distribution and movements of PIP2 in the surface membranes of cells.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.
  99. 99.
  100. 100.
  101. 101.
  102. 102.
  103. 103.
  104. 104.
  105. 105.
  106. 106.
  107. 107.
  108. 108.
  109. 109.
  110. 110.
  111. 111.
  112. 112.
  113. 113.
  114. 114.
  115. 115.
  116. 116.
  117. 117.
View Abstract

Navigate This Article