NEDD4‐induced degradative ubiquitination of phosphatidylinositol 4‐phosphate 5‐kinase α and its implication in breast cancer cell proliferation

Abstract Phosphatidylinositol 4‐phosphate 5‐kinase (PIP5K) family members generate phosphatidylinositol 4,5‐bisphosphate (PIP2), a critical lipid regulator of diverse physiological processes. The PIP5K‐dependent PIP2 generation can also act upstream of the oncogenic phosphatidylinositol 3‐kinase (PI3K)/Akt pathway. Many studies have demonstrated various mechanisms of spatiotemporal regulation of PIP5K catalytic activity. However, there are few studies on regulation of PIP5K protein stability. Here, we examined potential regulation of PIP5Kα, a PIP5K isoform, via ubiquitin‐proteasome system, and its implication for breast cancer. Our results showed that the ubiquitin ligase NEDD4 (neural precursor cell expressed, developmentally down‐regulated gene 4) mediated ubiquitination and proteasomal degradation of PIP5Kα, consequently reducing plasma membrane PIP2 level. NEDD4 interacted with the C‐terminal region and ubiquitinated the N‐terminal lysine 88 in PIP5Kα. In addition, PIP5Kα gene disruption inhibited epidermal growth factor (EGF)‐induced Akt activation and caused significant proliferation defect in breast cancer cells. Notably, PIP5Kα K88R mutant that was resistant to NEDD4‐mediated ubiquitination and degradation showed more potentiating effects on Akt activation by EGF and cell proliferation than wild‐type PIP5Kα. Collectively, these results suggest that PIP5Kα is a novel degradative substrate of NEDD4 and that the PIP5Kα‐dependent PIP2 pool contributing to breast cancer cell proliferation through PI3K/Akt activation is negatively controlled by NEDD4.

pathways and functional roles of phosphatidylinositol 4,5-bisphosphate (PIP2), one of the phosphoinositides, have been extensively studied because of its significant impacts on membrane signalling, trafficking and dynamics. 1,2 PIP2 metabolism is also implicated in various human diseases including cancer. 3,4 Phosphatidylinositol 3kinase (PI3K)-catalysed phosphatidylinositol 3,4,5-trisphosphate (PIP3) formation from PIP2 is essential for Akt phosphorylation and activation, promoting cancer cell growth and proliferation. 5 The type I phosphatidylinositol 4-phosphate 5-kinase (PIP5K) family members comprising PIP5Ka, PIP5Kb and PIP5Kc primarily account for PIP2 production at the plasma membrane by catalysing the phosphorylation of phosphatidylinositol 4-phosphate. 1,3 Accumulating evidence has indicated involvement of PIP5Kdependent PIP2 production in cancer progression. For example, the focal adhesion-specific PIP5Kc90 (a 90 kD splice variant of PIP5Kc) enhances migration, invasion and proliferation of breast cancer cells. [6][7][8] In addition, PIP5Ka is required for invadopodia formation and promotes survival of breast cancer cells. 9,10 A previous report also showed that PIP5Ka inhibition using a small compound reduced invasion and caused apoptotic cell death in prostate cancer cells. 11 It is proposed that the oncogenic property of PIP5Ka is closely linked to the activation of PI3K/Akt signalling pathway. 10,11 There have been many studies elucidating the underlying mechanisms, by which PIP5K activation is spatiotemporally regulated. [12][13][14] For instance, PIP5K catalytic activities are modulated by upstream regulators such as Rho family small GTPases, binding partner proteins and post-translational modifications such as phosphorylation. [12][13][14][15] However, relatively little is known about the potential regulation of PIP5K via protein stability. To our knowledge, there is actually no evidence of protein degradation of PIP5Ka. In this regard, we have been intrigued by the possibility that PIP5Ka is a target of the ubiquitinproteasome system (UPS), one of main degradative pathways, for its regulation, which may also play a role in cancer development.
Among a number of E3 ubiquitin ligases, we especially paid attention to neural precursor cell expressed, developmentally downregulated gene 4 (NEDD4) as a candidate for PIP5Ka ubiquitination in that it regulates a wide range of membrane proteins including receptors, ion channels, adaptors and signalling players. 16,17 NEDD4 (also called NEDD4-1) is a prototype of the NEDD4 family of E3 ubiquitin ligases and is known to regulate various proteins that are related to cancer. 18,19 NEDD4 is composed of an N-terminal C2 domain that binds to calcium or acidic membrane lipids such as phosphoinositides, four central WW domains that recognize the PY motifs (PPXY or LPXY) in target substrates through direct interaction and a C-terminal catalytic HECT (homologous to the E6-AP carboxyl terminus) domain that mediates ubiquitination. 17,20 In this study, we attempted to determine whether NEDD4 could regulate PIP5Ka protein stability. Here, we showed for the first time that NEDD4 interacted with PIP5Ka and mediated its degradation via the UPS, resulting in reduction of PIP2 levels. Our results from PIP5Ka-deficient and -reconstituted breast cancer cells supported that PIP5Ka positively mediates breast cancer cell proliferation and epidermal growth factor (EGF)-induced Akt activation. Furthermore, a PIP5Ka mutant (K88R) that was resistant to degradation by NEDD4 had greater effects on them compared to wild-type (WT) PIP5Ka. Overall, these results suggest that proteasomal degradation of PIP5Ka by NEDD4 can be an alternative way for controlling the protein and PIP2 levels leading to suppression of PI3K/Akt-associated cancer progression.

| Reagents and antibodies
Most chemicals, including MG132, cycloheximide, DMEM, EGF, anti-FLAG M2 affinity gels, and antibodies to a-tubulin, FLAG and

| Western blotting and immunoprecipitation
Cell lysate preparation, protein quantification, SDS-PAGE, Western blotting and anti-FLAG immunoprecipitation (IP) were performed as described previously. 21

| Cell imaging and immunostaining
Fluorescent images were captured with a Zeiss LSM 710 confocal microscope (Carl Zeiss GmbH, Jena, Germany) as previously described. 21,22 In brief, cells were fixed with 4% paraformaldehyde for 20 minutes. At indicated, cells were immunostained with mouse monoclonal anti-HA or anti-FLAG antibody, followed by staining with Alexa Fluor 594-conjugated secondary antibodies.

| PIP5Ka knockout
Cas9-mediated gene editing was performed by the lentiviral infection of a single guide RNA (sgRNA) and CRISPR/Cas9 system using the lentiCRISPRv2 vector (a gift of Prof. Daesik Lim, KAIST, Daejeon, Korea). 23,24 The guide RNA sequences used for this study were as follows (Bioneer): upper, 5 0 -caccgCGCCCTGCCGGGCTTACCTG-3 0 , and bottom, 5 0 -aaacCAGGTAAGCCCGGCAGGGCGc-3 0 for human PIP5Ka; upper, caccgATCGTTTCCGCTTAACGGCG, and bottom, 5 0 -aaacCGCCGTTAAGCGGAAACGATc-3 0 for a non-target control. The oligo annealing and subcloning, the lentiviral production and the cell transduction were carried out according to the instructions. Cells were infected with recombinant lentiviruses for 2 days and then cultured with fresh complete media containing puromycin (3.0 lg/mL) for 2 weeks. Puromycin-resistant clones were isolated and screened for the PIP5Ka gene knockout using Western blot analysis and genomic DNA sequencing.

| Colony formation assay
PIP5Ka sgRNA-or non-targeting sgRNA-expressing cells were seeded in 6-well plates at a density of 500-1000 cells/well. For PIP5Ka complementation experiments, FLAG-PIP5Ka plasmids were transiently transfected into PIP5Ka knockout cells using Lipofectamine 2000 before cell seeding. After 7-10 days, cells were fixed in an acetic acid:methanol mixture (1:7, v/v) for 1 minutes at room temperature and stained with 0.5% crystal violet, and then the number of cell colonies was counted.

| Statistical analysis
All experiments were performed independently at least three times with similar results. Band intensities of Western blots were measured using NIH ImageJ software (National Institutes of Health, Bethesda, MD, USA). Data shown in the graphs are presented as the mean AE SEM. The statistical significance of the data was determined using a one-way analysis of variance with Tukey's multiple comparison tests using GraphPad Prism software (La Jolla, CA, USA).

| NEDD4 induces the proteasomal degradation of PIP5Ka
As a first step to examine the protein stability of PIP5Ka, we tested the possibility of its proteasomal or lysosomal degradation. Changes  Conversely, treatment with cycloheximide, an inhibitor of protein synthesis, gradually decreased PIP5Ka protein levels in the same time ranges ( Figure 1C).
The HECT domain-containing E3 ubiquitin ligase NEDD4 plays important roles in regulating various proteins present in the plasma membrane. 16,17 Here, we examined NEDD4 for its potential effect on PIP5Ka protein levels. Transfected HA-tagged NEDD4 induced degradation of endogenous PIP5Ka in a dose-dependent manner ( Figure 1D). In contrast, when other NEDD4 family members NEDD4L and ITCH 20 were similarly tested, NEDD4L and ITCH did not show a degrading effect on PIP5Ka ( Figure S1). As expected, PIP5Ka mRNA levels were not significantly changed under the same conditions ( Figure S2), indicating that the decreased PIP5Ka protein levels by NEDD4 are a post-translational event. NEDD4 gene knockdown with two different siRNAs resulted in relatively high levels of PIP5Ka compared to the control siRNA ( Figure 1E). We also examined the effects of NEDD4 on transiently expressed PIP5Ka with cells cotransfected with HA-NEDD4 and GFP-tagged PIP5Ka. The dose-dependently expressed HA-NEDD4 reciprocally altered GFP-PIP5Ka levels ( Figure 1F). The catalytically inactive C867A mutant of HA-NEDD4 was less effective in degrading FLAG-PIP5Ka compared to the WT NEDD4, as expected ( Figure 1G). In addition, pre-treatment with MG132 blocked the degradation of endogenous PIP5Ka by HA-NEDD4 ( Figure 1H). All these results indicate that NEDD4 induces the proteasomal degradation of PIP5Ka. However, endogenous and transfected PIP5Kc90 was not degraded by NEDD4 ( Figure S3), supporting the specificity of NEDD4-mediated PIP5Ka degradation.

| NEDD4 mediates ubiquitination of PIP5Ka
We examined whether PIP5Ka is ubiquitinated by NEDD4.  Figure 2B). These observations support that PIP5Ka is a novel NEDD4 substrate for protein ubiquitination.
Protein substrates to be degraded by the UPS are polyubiquitinated through K48-linked ubiquitin chains. 26 We tested the possibility that

| PIP5Ka interacts with NEDD4 through its Cterminal region
The ubiquitination of target proteins by NEDD4 requires their physical interaction with NEDD4. We tested a possible interaction between NEDD4 and PIP5Ka with cells transfected with FLAG-PIP5Ka and/or HA-NEDD4. HA-NEDD4 coprecipitated with FLAG-PIP5Ka only when both proteins were coexpressed ( Figure 3A). We carried out PIP5Ka or NEDD4 IP in intact HEK293 cells. Anti-NEDD4 immunoblotting of PIP5Ka immunoprecipitates showed that endogenous NEDD4 coprecipitated with PIP5Ka, and likewise, the coprecipitation of PIP5Ka and NEDD4 was observed in NEDD4 IP samples ( Figure 3B). No immunoreactivity with PIP5Ka and NEDD4 in IP samples with control IgG confirmed the specificity of this experiment. Alternatively, we generated a GST-fusion protein of fulllength PIP5Ka. A control GST and GST-PIP5Ka protein conjugated to glutathione-Sepharose beads were used in the pull-down assay with HA-NEDD4-expressing cell lysates. HA immunoblotting showed that HA-NEDD4 was pulled down by GST-PIP5Ka but not by GST alone ( Figure 3C).
The PPXY or LPXY motifs in NEDD4 substrates mediate the NEDD4 binding 27 and PIP5Ka contains a single LPGY motif (193-196 aa) in the catalytic domain ( Figure S4A). To examine whether this LPGY motif is responsible for NEDD4 binding, we introduced mutations (LPGY?AAGF) into the FLAG-PIP5Ka. Somewhat unexpectedly, however, the AAGF mutant retained almost the same binding activity as the WT for HA-NEDD4 ( Figure S4B) and was still degradable by NEDD4 at levels comparable to those of the WT (  Based on these results, we tested the colocalization between these two proteins. GFP-PIP5Ka seemed to localize to the plasma membrane and form membranous puncta and aggregates, and mCherry-NEDD4 was also, at least partially, found on the same membrane sites although it also showed a cytoplasmic distribution ( Figure 3F).

PIP2 levels
We examined the possibility that the NEDD4-mediated degradation of PIP5Ka could influence PIP2 levels using the GFP-tagged PH domain of PLCd (GFP-PLCd-PH) that has been used as a PIP2 reporter. 28 The fluorescent protein having a binding affinity for PIP2 undergoes translocation between the plasma membrane and cytoplasmic space depending on the lipid levels. A prominent GFP fluorescence of the PIP2 probe was found at the cell surface when it was solely transfected ( Figure 4A). Upon cotransfection with mCherry-NEDD4, GFP-PLCd-PH diffused into the cytoplasm, indicative of a reduction in PIP2 levels ( Figure 4A). Similarly, we used another fluorescent construct of the membrane-bound transcription factor Tubby for monitoring PIP2 levels. Tubby selectively binds to plasma membrane PIP2 through its Tubby domain, and the mutated (R332H) Tubby more sensitively reflects PIP2 changes. 29,30 Transfected Tubby-R332H-YFP alone localized to the plasma membrane as expected, but it became concentrated in the perinuclear sites by cotransfection with mCherry-NEDD4 ( Figure 4B). We compared the effects of WT and catalytically inactive NEDD4 on PIP2 levels.
Tubby-R332H-YFP still localized to the cell surface in the cells expressing HA-NEDD4 C867A, whereas it was relatively soluble in HA-NEDD4 WT-expressing cells ( Figure 4C). The NEDD4-induced membrane-to-cytosol translocation of Tubby-R332H-YFP further supports that PIP5Ka degradation by NEDD4 leads to a decrease in PIP2 levels.

| PIP5Ka is ubiquitinated by NEDD4 at the lysine 88 residue
As an attempt to determine the specific lysine residue mediating PIP5Ka degradation by NEDD4, we utilized several ubiquitination site prediction methods. The lysine in the STKPER motif is shown to have a relatively high probability among multiple candidates, and the equivalent lysine 97 of human PIP5Kc90 was reported to be ubiquitinated by HECTD1, a HECT domain-containing E3 ubiquitin ligase. 7 Thus, we substituted K88, the corresponding residue in mouse HA immunofluorescence in magenta colour and YFP fluorescence were visualized by confocal microscopy. Scale bars, 10 lm 4 hours after the addition of cycloheximide, but the K88R mutant remained at relatively high levels ( Figure 5A). The PIP5Ka K88R mutant remained stable irrespective of the presence of HA-NEDD4 in contrast to WT PIP5Ka ( Figure 5B) and was resistant to NEDD4mediated ubiquitination unlike the PIP5Ka WT ( Figure 5C).
We compared binding affinities for HA-NEDD4 between the WT and K88R FLAG-PIP5Ka. The K88R mutant maintained a binding affinity for NEDD4 as strong as the WT ( Figure 5D). Moreover, we repeated the pull-down assay with full-length GST-fusion proteins of PIP5Ka WT and K88R. Consistently, the NEDD4-binding abilities of the GST-PIP5Ka K88R were similar to those of the GST-PIP5Ka WT ( Figure 5E), suggesting that the non-degradable PIP5Ka K88R mutant against NEDD4 is likely due to a defect in ubiquitination rather than in binding. WT or K88R FLAG-PIP5Ka-cotransfected cells with Tubby-R332H-YFP were examined for PIP2 levels by monitoring YFP localization. Confocal images showed that the plasma membrane targeting of the PIP2-specific fluorescent probe was more robust in the K88R mutant-expressing cells than in the WT-expressing cells (Figure 5F), indicating a relatively high level of PIP2 in the K88R mutant. Collectively, these data support that K88 is one of the main degradative ubiquitination sites of PIP5Ka by NEDD4.

| PIP5Ka K88R potentiates Akt activation and breast cancer cell proliferation
It has been reported that PIP5Ka plays a role in cancer progression. [9][10][11] In this regard, we examined whether NEDD4-induced PIP5Ka degradation might be implicated in cancer. To test this idea, we first developed PIP5Ka knockout in breast cancer cell lines Cas9-mediated sgRNA targeting PIP5Ka (sgPIP5Ka) or a non-targeting sgRNA (sgControl) as a negative control. We confirmed that PIP5Ka protein expression levels in the sgPIP5Ka-expressing cells were clearly reduced compared to those in the sgControl-expressing cells ( Figure 6A). We observed comparable expression levels of endogenous NEDD4 protein in the three breast cancer cells (Figure 6A). Interestingly, when we performed the colony formation assay, we observed that PIP5Ka depletion by sgPIP5Ka caused a significant decrease in the number of surviving colonies, suggesting that PIP5Ka is required for cell proliferation (Figure 6B and C). Consistent with this result, inhibition of PIP5Ka with a small compound ISA-2011B 11 led to defective proliferation in SKBR3 cells ( Figure S5A).
The activation of the PI3K/Akt signalling pathway plays a central in growth factor-induced cell growth and proliferation. 31 We thus examined the effect of PIP5Ka knockout on Akt activation by EGF.
For this, we introduced the lentiviral sgControl and sgPIP5Ka expression into a gastric cancer cell line, NCI-N87, as described in Figure 6A and then measured EGF-induced Akt phosphorylation at the S473 residue, a marker of PI3K/Akt activation. EGF stimulation rapidly increased phospho-Akt to the maximal level within 5 min, followed by dephosphorylation to the basal level during 60 min in sgControl cells ( Figure 6D). In contrast, Akt phosphorylation was blunted in PIP5Ka knockout cells during the time course of EGF stimulation ( Figure 6D). PIP5Ka knockout also decreased colony formation in NCI-N87 cells ( Figure S5B). The PIP5Ka knockout NCI-N87 cells were reconstituted with WT or K88R FLAG-PIP5Ka by transient transfection. We found that reconstitution with the PIP5Ka K88R mutant could more significantly potentiate Akt phosphorylation by EGF than WT PIP5Ka reconstitution ( Figure 6E). When we tested the same PIP5Ka reconstitution experiments with PIP5Ka knockout SKBR3 cells, similar results were obtained ( Figure 6F). As expected, the NEDD4-mediated PIP5Ka degradation was detectable in SKBR3 cells as evidenced by immunoblot analysis (Figure 6G), which was consistent with the result in Figure 1D. Furthermore, when HA-NEDD4 was coexpressed in PIP5Ka-reconstituted SKBR3 cells, EGF-induced Akt phosphorylation was also much higher in the K88R-reconstituted cells than in the WT-reconstituted cells ( Figure 6H).
We tested the potential effects of reconstituted WT and K88R FLAG-PIP5Ka on cell proliferation by performing the colony formation assay with PIP5Ka knockout SKBR3 cells. Defects in colony formation by PIP5Ka knockout were recovered by the WT, and the K88R mutant had a greater colony-forming ability ( Figure 6I and J).
We assessed transcriptional changes in E2F1, CDK1 and CCND1 whose expression levels correlate with increased cell proliferation and are, in general, elevated in various cancers including breast cancer cells. [32][33][34][35] As shown in Figure 6K, the mRNA expression levels of E2F1, CDK1 and CCND1 were reduced, but conversely, the mRNA levels of the tumour suppressor FOXO3 36,37 were increased in the PIP5Ka knockout SKBR3 cells compared to those in non-targeting control cells. Moreover, the transient re-expression of WT PIP5Ka in the PIP5Ka knockout cells at least partially reversed those changes, and similar rescuing effects of the PIP5Ka K88R mutant on the expression of E2F1, CDK1 and CCND1 were relatively high compared to the WT to some extent ( Figure 6K). Reconstitution of the NEDD4 binding-deficient PIP5Ka ΔCT that harbours the catalytic domain ( Figure 3D and E) also increased colony formation as much as the WT ( Figure 6L). Notably, HA-NEDD4 coexpression suppressed colony formation by PIP5Ka WT, but, in contrast, did not show such inhibitory effect on the PIP5Ka ΔCT ( Figure 6L). The relevant expressions of transfected HA-NEDD4 and FLAG-PIP5Ka proteins were confirmed by Western blot analysis ( Figure 6M). Taken together, all these results suggest that PIP5Ka acts upstream of PI3K/Akt signalling by supplying the PI3K substrate PIP2, leading to increased cell proliferation and that NEDD4 serves as a negative regulator of this signalling by inducing proteasomal degradation of PIP5Ka (Figure 7).

| DISCUSSION
In this present study, we showed that PIP5Ka is a ubiquitinated protein that undergoes proteasomal degradation under steady-state conditions, suggesting the dynamic control of PIP5Ka protein levels through the ubiquitin-proteasome pathway. Our data revealed that NEDD4 is a reliable candidate to mediate the UPS-mediated proteolysis of PIP5Ka. As a result, NEDD4 could reduce PIP5Ka-dependent plasma membrane PIP2 pool. We also demonstrated that K88 of PIP5Ka is the likely residue ubiquitinated by NEDD4. Accordingly, the PIP5Ka K88R mutant enhanced the protein stability of PIP5Ka.
The mutants of PIP5Ka K88R, as well as the catalytically inactive NEDD4, led to relatively high levels of plasma membrane PIP2 compared to their WT controls. Considering the fact that many studies of PIP5Ks have focused on the identification of the underlying regulatory mechanisms of their catalytic activities, [12][13][14][15] this study presents another perspective on the regulation of PIP5K-dependent PIP2 generation through the NEDD4-dependent control of PIP5K protein stability.
Our results indicated that NEDD4 interacts with the C-terminal region of PIP5Ka that does not harbour the NEDD4-binding P(L)PXY motifs. Similarly, a previous study showed that a non-P(L)PXY motif of activated fibroblast growth factor receptor 1 was responsible for NEDD4 binding in the receptor's ubiquitination and endocytosis by NEDD4. 38 In addition, the interaction of PIP5Kci5, a splice variant of PIP5Kc, with NEDD4 was mediated through its C-terminal region that contains no P(L)PXY motifs, and this interaction suppressed NEDD4-mediated ubiquitination and proteasomal degradation of the tumour suppressor, mitogen-inducible gene 6, without PIP5Kci5 degradation. 39 Apparently, NEDD4 did not mediate PIP5Kc90 degradation, suggesting a differential regulation of PIP5Ka and PIP5Kc90 by NEDD4.
Our results showed that PIP5Ka is necessary for breast cancer cell proliferation, as evidenced by the increases in the cell proliferation markers E2F1, CDK1 and CCND1 and the decrease in the TRAN ET AL. and mRNA expression levels (K) were quantified relative to those in the respective sgControl (C, K), vector control (J), or FLAG-PIP5Ka WT-reconstituted (L) cells. Values in the graphs are presented as the mean AE SEM. *P < .05, **P < .01 CCND1 transcription than the WT PIP5Ka. It is well known that the aberrant activation of PI3K/Akt pathway is one of the primary causes of diverse cancers. 5 The activation of PI3K/Akt signalling through PIP5Ka has been identified in breast cancer and prostate cancer models. 10,11 Therefore, a significant implication of this study is that the regulation of PIP5Ka by NEDD4 can play a role in limiting activation of the PI3K/Akt pathway linked to breast cancer progression ( Figure 7).
Previous studies showed that mRNA and protein expression levels of PIP5Ka were highly elevated in breast cancer cells 9 and prostate cancer cells, 11 respectively. Similarly, a previous study using human breast cancer database revealed that only PIP5Ka showed increased gene expression among PIP5K family members. 40 We also observed more abundant PIP5Ka protein in the breast cancer cells Overall, it will be interesting to investigate those possibilities with efforts to clarify as yet undefined molecular mechanisms.
On the other hand, it has been demonstrated that NEDD4 mediates ubiquitination and degradation of various cancer-related proteins. For example, NEDD4 degrades the tumour suppressors, such as phosphatase and tensin homologue (PTEN), a lipid phosphatase that hydrolyses PIP3 41 and Beclin 1. 42 However, NEDD4 is shown to be dispensable for the PTEN regulation. 43 In addition, the oncoproteins such as Ras 44 and Myc 45 are targeted for degradation by NEDD4. It seems that NEDD4 can distinctly regulate degradative ubiquitination of those different types of protein substrates in various cancer models, which leads to promotion or suppression of tumorigenesis. Based on this study, our results suggest that NEDD4 can act as a suppressor of breast cancer by negatively regulating PIP5Ka.
PIP2 is a critical regulator of a wide range of membrane signalling, vesicle trafficking and actin dynamics, and such pleiotropic roles played by PIP2 are mediated through its binding to and regulatory effect on related cytosolic proteins. 1-3,12,13 NEDD4 also plays an important role in regulating various membrane proteins of receptors, channels and adaptors at the cell surface. 16,17 In addition, the N-terminal C2 domain expressed in NEDD4 is known as a recognition module for calcium and phosphoinositides including PIP2, 46 and Smurf, a NEDD4 family member, was proposed to be activated upon binding of its C2 domain to phosphoinositides. 47 Our results indicated the physical interaction between NEDD4 and PIP5Ka. These results raise the possibility that NEDD4 may be functionally implicated in the physiological roles of PIP5Ka-dependent PIP2.
In conclusion, our present results demonstrate that PIP5Ka undergoes proteasomal degradation through interaction with and ubiquitination by the ubiquitin ligase NEDD4, which is accompanied by a reduction in plasma membrane PIP2 levels. Thus, this study proposes an alternative way of controlling PIP2 through NEDD4. Our results indicate that PIP5Ka-dependent PIP2 production contributes to the activation of PI3K/Akt signalling and breast cancer cell proliferation, which can be impeded by NEDD4, at least partially, through F I G U R E 7 A model for regulation of cancer cell proliferation by PIP5Ka that is a target of NEDD4-dependent proteasomal degradation. PIP5Ka that produces PIP2 from phosphatidylinositol 4-phosphate (PI(4)P) contributes to cancer cell survival and proliferation by supplying PIP2, a substrate of PI3K for PIP3 formation, which activates its downstream Akt (top). The PIP2 levels are reduced by NEDD4-mediated PIP5Ka degradation via UPS, leading to downregulation of the oncogenic PI3K/Akt signalling pathway (bottom)