PTEN regulates phospholipase D and phospholipase C

Abstract

PTEN is an ubiquitously expressed tumor suppressor which plays a prominent role in the pathogenesis of many types of sporadic solid tumors, including breast cancer, as well as hematologic malignancies. Germline PTEN mutations cause 85% of Cowden syndrome (CS), characterized by a high risk of breast and thyroid cancers, and 65% of Bannayan–Riley–Ruvalcaba syndrome (BRRS), characterized by lipomatosis, hemangiomas and speckled penis. Historically, PTEN's role in tumor suppression has been linked to the down-regulation of the PI3K/AKT pathway by PTEN's lipid phosphatase activity. Beyond the AKT pathway, however, there has been minimal examination of PTEN's responsibility in lipid-derived cellular signaling. As phospholipids have been shown to be critical components in signal transduction and cellular proliferation and PTEN controls cellular phospholipid levels, we hypothesized that PTEN functions as a regulator of lipid signaling and homeostasis. Increased PTEN expression in unstimulated MCF-7 breast cancer cells results in a 51% increase in phosphatidic acid, with a decrease in phosphatidylcholine, suggesting that PTEN may regulate phospholipase D (PLD). PTEN overexpression results in a 30% increase in basal PLD activity. As phospholipase C (PLC) is both involved in PLD activation and is regulated by PIP2/3 levels, we investigated the role of PTEN on PLC activation. Our data suggest that PTEN modulates PLC:PLD activation pathways and indicate that the pathogenesis of CS/BRRS has a more complex biochemical basis beyond simply activating the PI3K pathway. This provides alternative routes for PTEN's tumor suppressor action that may be beneficial in the creation of novel targets for cancer therapy and prevention.

INTRODUCTION

PTEN, a tumor suppressor localized to 10q23.3, is mutated in a variety of human cancers (1). Germline mutations are inherited in patients with the hamartoma-neoplasia syndromes, Cowden syndrome (CS), Bannayan–Riley–Ruvalcaba syndrome (BRRS) and Proteus and Proteus-like syndromes (2–5). As a susceptibility gene for CS, an autosomal dominant disorder, germline PTEN mutations have been documented in 85% of patients (4,6). Additionally, of the malignancies reported in CS, there is an increased incidence of breast and other cancers (7,8). BRRS, which is associated with germline PTEN mutations in 65% of patients (6,9), may also be associated with an increased frequency of breast cancer (9). In addition to disorders with mutations in the germline, somatic PTEN alterations have been found in a variety of sporadic cancers. In particular, somatic PTEN mutations in sporadic breast cancers are rare (10); however, multiple groups have observed 30–40% of non-cultured primary breast carcinomas showing loss-of-heterozygosity of markers in the PTEN region in these tumors (11–13).

As a tumor suppressor, PTEN regulates a variety of cellular processes resulting in the inhibition of cell growth and cancer progression. PTEN is a dual-specificity phosphatase, removing a phosphate group from both protein and lipid substrates. PTEN's protein phosphatase activity negatively regulates the mitogen-activated kinase pathway, promotes cell cycle arrest in G1 and induces apoptosis (14–16). PTEN's lipid phosphatase activity antagonizes the PI3K/AKT pathway by decreasing the levels of cellular phosphatidylinositol (3,4,5)triphosphate (PIP3) (17).

Knowledge of PTEN's role as a negative regulator of the AKT pathway has been clearly defined for sometime. Beyond the AKT pathway, there is little evidence to date that demonstrates PTEN's regulation of other lipid-dependent pathways. However, as PIP2 and PIP3 are important regulators of cellular pathways, it stands to reason that PTEN may regulate pathways beyond AKT. With the understanding that PTEN regulates levels of PIP2 and PIP3 in the cell, we decided to examine whether this regulation affects cellular signaling beyond the AKT pathway. Additionally, we investigated the role of PTEN in the regulation of phospholipase D (PLD) and phospholipase C (PLC), enzymes that can be regulated by PIP3 levels as well as PIP2.

RESULTS

PTEN overexpression changes basal phospholipid levels

In order to examine the role of PTEN in lipid signaling, we first determined the effect of PTEN overexpression on various cellular phospholipid levels in unstimulated cells. We analyzed the effect of PTEN on the major cellular membrane phospholipids, phosphatidic acid (PA), phosphatidylcholine (PC), phosphatidylserine (PS), and phosphatidyletha nolamine (PE) under basal conditions (Fig. 1). Upon PTEN overexpression, statistically insignificant changes were observed for PS and PE. However, we saw significant changes with PC, which exhibited the greatest change, a 50% decrease, between control and wild-type (WT) cells (P = 0.04). In addition, we also saw a change in PA with a 51% increase (P = 0.03). PC is the substrate for PLD, whereas PA is the product of the PLD catalyzed reaction. Consequently, these results may suggest that PTEN overexpression could be directly or indirectly affecting PLD.

Figure 1.

PTEN overexpression results in changes in cellular phospholipid levels. (A) MCF-7 cell lines were plated and PTEN expression was induced by the removal of Tet 48 h prior to harvesting. After harvesting, lipids were extracted and individual phospholipids (PC, PA, PS and PE) were isolated with TLC. Individual lipids were quantified by lipid phosphorous assay and were normalized to total amount of protein extracted. Each bar represents the mean ± SD of three individual experiments. (B) Western blot of PTEN after MCF-7 cell lines were plated and grown in the presence or absence (PTEN induction) of Tet 48 h prior to harvesting. Whole cell lysates were separated by SDS–PAGE and subjected to western analysis. This panel is a representative blot where three separate experiments were performed.

PTEN expression increases PLD activity

The above results suggest that PLD may be activated by PTEN; however, the analysis of PA at the mass level must be approached with caution. The levels of PA in the membrane are in a constant state of flux. PA can be converted to diacylglycerol (DG) by PA phosphohydrolase (18), and similarly, DG can be converted to PA through the action of DG kinase (19), thus making the measurement of PLD activity difficult to ascertain. In order to obtain an accurate measure of PLD activation, we utilized the transphosphatidylation reaction of PLD by growing the cells in butanol, a primary alcohol, to generate phosphatidylbutanol (PBut), a stable molecule (20). Using this assay, we found that overexpression of WT-PTEN led to a 30% increase in PLD activity (P = 0.04) compared with the control cell line (Fig. 2). These data, taken together with our initial mass study of PC and PA, strongly suggest that PTEN increases PLD activation in basal, resting cells.

Figure 2.

PTEN activates basal PLD through lipid phosphatase activity. MCF-7 cell lines were plated and PTEN expression was induced by the removal of Tet 48 h prior to harvesting. (A) Butanol (1%) was added to cells 4 h prior to harvesting. After harvesting, lipids were extracted and PBut was isolated using TLC and quantified with liquid scintillation counting. Each bar represents the mean ± SD of three individual experiments. (B) Western blot of PTEN. MCF-7 cell lines (control, PTEN-WT, C124S and G129E) were plated and grown in the absence of Tet in order to stimulate PTEN expression for 48 h prior to harvesting. Whole cell lysates were separated by SDS-PAGE and subjected to western analysis. This panel illustrates a representative blot from three separate experiments.

We then examined whether PTEN's lipid or protein phosphatase activity plays a role in the activation of PLD in unstimulated cells. To study this question, we utilized cells which can overexpress either C124S-PTEN or G129E-PTEN. The C124S and G129E mutations are both missense mutations in exon 5 that have been detected in the CS (13) and allow one to examine the role of the PTEN's lipid and protein phosphatase activities. The C124S mutation renders a protein that is both lipid- and protein-phosphatase inactive. The G129E mutation renders a protein that is lipid-phosphatase inactive but protein-phosphatase active. We found that in both the C124S-PTEN and G129E-PTEN cell lines, both of which overexpress lipid-phosphatase-inactive PTEN, there was a significant 30% decrease in PLD activity, as evidenced by the decrease in PBut formation (P = 0.02, Fig. 2) compared with cells overexpressing lipid-phosphatase-active PTEN. On the basis of these results, we conclude that the lipid phosphatase activity of PTEN may be responsible for regulating PLD under basal unstimulated conditions.

PTEN's regulation of PLD extends to stimulated conditions

Our data to date suggest that PTEN regulates PLD in resting, unstimulated cells; however, we have not examined stimulated cells. We decided to stimulate the cells with B-estradiol for several reasons. First, PTEN expression has been shown to be regulated by estrogen (21). Secondly, PTEN's effects on cellular signaling are abrogated in the presence of estrogen (14). Thirdly, estrogen has been linked to breast cancer (22). Fourth, estrogen activates the phosphatidylinositol pathway (23). And lastly, the MCF-7 cells used in this study are estrogen-receptor positive, thereby making B-estradiol a feasible stimulant. Upon B-estradiol stimulation, we found that, like the unstimulated cells, cells overexpressing WT-PTEN had a 45% increase in the accumulation of PBut, indicative of PLD activation (P = 0.01, Fig. 3). Of note, this increase was significantly higher in the presence of B-estradiol compared with basal cells, with an observed 15% increase in PLD activity (compare WT in Figs 2 and 3). These data may suggest that B-estradiol may act synergistically with WT-PTEN to activate PLD. Interestingly, however, we found that in B-estradiol-stimulated cells, the C124S-PTEN cells stimulated PLD activity by 26%, whereas the G129E-PTEN cells exhibited the greatest amount of PLD activity (a doubling in PLD activity) compared with all other cell lines. This indicated that the protein phosphatase activity of PTEN plays an important role in estrogen-mediated PLD regulation (Fig. 3), as this was the only cell line that contained only protein phosphatase activity.

Figure 3.

PTEN activates stimulated PLD through protein phosphatase activity. MCF-7 cell lines, (control, PTEN-WT, C124S and G129E) were plated and PTEN expression was induced by the removal of Tet 48 h prior to harvesting. Butanol (1%) was added to cells 4 h prior to harvesting. A total of 1 µm of B-estradiol was added directly to the culture medium 30 s prior to harvesting. After harvesting, lipids were extracted and PBut was isolated using TLC and quantified with liquid scintillation counting. Each bar represents the mean ± SD of three individual experiments.

Translocation studies implicate PTEN's involvement in PLC-gamma translocation

The data examining basal and stimulated PLD activity suggest that PTEN regulates PLD in both these conditions. The method of regulation, however, appears to be through different mechanisms and suggests that PTEN may be regulating events upstream of PLD activation. One of the most notable mechanisms for PLD activation is through activation of PLC and generation of the second messengers DG and inositol triphosphate. PLC-gamma has been shown to be activated by B-estradiol (24), making it a target of interest. In order to examine this, we first looked at the effect of WT-PTEN, in the presence of B-estradiol, on the translocation of PLC-gamma from the cytosol to the membrane. We found that PLC-gamma migrated from the cytosol to the membrane where it can be functionally active (Fig. 4) after 30 s of B-estradiol stimulation. Additionally, we found that in stimulated cells, PLC-gamma remains localized in the cytosolic compartment in the C124S-PTEN and G129E-PTEN cell lines where lipid-phosphatase-inactive PTEN is overexpressed (Fig. 5). These preliminary data, taken together with the translocation observed with the WT cells, suggest that PTEN regulates PLC-gamma translocation through its lipid phosphatase activity.

Figure 4.

PTEN induces translocation of PLC-gamma from the cytosol to the membrane. MCF-7 cell lines were plated and PTEN expression was stimulated by the removal of Tet 48 h prior to harvesting. A concentration of 1 µm of B-estradiol was added directly to the culture medium at varying time points before the cells were harvested. After harvesting, membrane and cytosolic fractions were isolated by ultracentrifugation. Cellular proteins were separated by SDS–PAGE and analyzed by western blot analysis for PLC-gamma. This figure illustrates a representative blot from three separate experiments.

Figure 5.

Lipid phosphatase activity of PTEN responsible for PLC-gamma translocation. MCF-7 cell lines were plated and PTEN expression was induced by the removal of Tet 48 h prior to harvesting. A concentration of 1 µm of B-estradiol was added directly to the culture medium 30 s prior to harvesting. After harvesting, the cytosolic fraction was isolated by ultracentrifugation. Cellular proteins were separated by SDS–PAGE and analyzed by western blot analysis for PLC-gamma. This illustrates a representative blot from three separate experiments.

DISCUSSION

Until recently, the role of PTEN in lipid homeostasis was narrowly defined to its role as a negative regulator of the PI3K/AKT pathway. Our studies here show that PTEN has the ability to regulate both PLD and PLC. We found that PTEN overexpression does not result in a significant change in the levels of PE or PS, suggesting that PTEN may not regulate the catabolism or synthesis of each of these lipids. In this study, we did not examine the phosphatidylinositol species, as PTEN's ability to utilize PIP3 as a substrate has been demonstrated (17) and these lipids are harder to analyze by conventional methods. In contrast to PE and PS, we did observe changes in the cellular levels of PC and PA, suggesting that PTEN effects lipid signaling.

Our experiments revealed that PTEN overexpression stimulates basal PLD activity and that this stimulation in basal cells required PTEN's lipid phosphatase activity. This suggests that PTEN's modulation of PIP3 and PIP2 levels at the cell membrane has an effect on PLD activity. This is not surprising, as both PIP3 and PIP2 have been implicated as regulators/modulators of PLD activity (25).

Given the role of B-estradiol in breast cancer as well as in lipid signaling, we decided to perform preliminary investigations into the effect of B-estradiol on PTEN-mediated lipid signaling. We found that in the presence of B-estradiol, PLD activity was stimulated upon WT PTEN overexpression. Interestingly, however, we found that this effect had a PTEN protein phosphatase aspect, suggesting that in stimulated cells, PTEN's protein phosphatase activity regulates PLD activity. It is interesting to hypothesize that this may be due to MAPK regulation, as PLD activation has been linked to MAPK. Another target may be Shc, given that PTEN's phosphatase activity regulates the Shc-Grb-SOS complex (15).

Although our study suggests that PTEN stimulates the activation of PLD, a recent finding by Kim et al. (26). demonstrated that PLD overexpression inhibits the expression of PTEN. This may indicate that there is reciprocal regulation between PLD and PTEN. It is intriguing to hypothesize that PTEN activation of PLD results in the activation of a feedback loop to prevent additional expression of PTEN. Cellular lipid homeostatsis is important to cellular function and proliferation; thus, it is not surprising that these two pathways cross-talk. Such potential regulation reinforces the complex interplay of lipid-signaling pathways and provides a rich and exciting area of future investigation.

In order to obtain a bit more understanding on potential mechanisms involved in PLD activation, we examined the effect of PTEN on PLC-gamma activity. Like PLD, PLC-gamma has a requirement for both PIP2 (as its substrate) and PIP3 (binding through PLC-gamma's PH domain). In addition, PLC activation is a classical activation pathway for PLD (27). Thus, it seemed reasonable to postulate that PTEN may modulate PLC-gamma. We hypothesize that PTEN overexpression results in the accumulation of PIP2 in the cell membrane, as demonstrated by Dixon and co-worker (17). The remaining PIP3 will serve as an anchor for PLC-gamma (28,29). PLC-gamma activity then results in the activation of PLD through the classic pathway of DG and calcium signaling (27). In order to look at the PLC-gamma activation, we examined the PLC-gamma translocation from the cytosolic (inactive) fraction to the membrane (active) fraction, a classic means of ascertaining its activation (24). We found that PTEN expression results in the translocation of PLC-gamma from the cytosol to the membrane (Fig. 4). This translocation event is both quick and transient in nature (Fig. 4), with PLC-gamma returning to the cytosol after 60 s of stimulation. Owing to this rapid translocation effect (both to the membrane and off), it is not unusual to see PTEN present in the cytosolic fraction at 30 s (Fig. 5). These observations suggest that the cross-talk between PTEN and lipid signaling is dynamic in nature.

The translocation event is dependent upon PTEN's lipid phosphatase activity. In the absence of lipid phosphatase activity (in the C124S and G129E PTEN expressing lines), we also see a dramatic increase in cytosolic PLC-gamma levels in response to B-estradiol (Fig. 5). Because we are looking at rapid and dynamic changes that occur in the span of a minute, it is highly unlikely that this increase is due to B-estradiol-stimulated protein production or even degradation. These data may suggest that the lack of PTEN's lipid phosphatase activity inhibits translocation of PLC-gamma to the membrane and also is involved in the release of PLC-gamma from subcellular compartments to the cytosol. These observations give another layer of complexity to PTEN's regulation of PLD and PLC, which will need to be examined.

The C124S-PTEN and G129E-PTEN mutations provide unique insights into the WT mechanism in which PTEN normally regulates PLC and PLD. They suggest that PTEN uses both of its functional phosphatase activities to regulate lipid-signaling pathways beyond the PI3K/AKT pathway. Although the cross-talk is extensive in these pathways, it is becoming increasingly clear that PTEN's role as a dual-specificity phosphatase is crucial in creating the appropriate intracellular balance of signaling molecules. The data in this study also suggest that a mutation in PTEN's functional domain can also significantly enhance, for example, the activity of PLD in the presence of B-estradiol (Fig. 3). This finding is particularly interesting in light of both the recent discovery by Kim et al. (26). that elevated PLD prevents apoptosis and the 2-fold increase in PLD activity in the G129E-PTEN cells (Fig. 3). Taken together, the data suggest that in the presence the G129E mutation and estrogen, PLD is being activated to prevent apoptosis and cell cycle arrest.

Central to our study is the fact that PTEN's conversion of PIP3 to PIP2 has significant effects on downstream lipid signaling. As these two phospholipids are involved in the regulation or activation of numerous enzymes, PTEN has the potential to modulate a variety of cellular signaling pathways, and we have shown here that PTEN can modulate both PLD and PLC. PLC-gamma has been implicated in cell cycle progression and mitogenesis, whereas PLD has been noted to contribute to various hallmarks of tumor progression which include growth signaling (30,31), suppression of p53 expression (32) and inhibition of apoptosis (33,34). Taken together, our data suggest that PTEN may regulate these cellular pathways and functions by modulating lipid signaling. This provides a new avenue for the role of PTEN in CS, BRRS as well as sporadic neoplasias. Elucidating the roles of PTEN in the cross-talk between lipid signaling and cellular homeostasis is thus important, and in this study, we have begun to show that PTEN regulates PLD and PLC. This opens a new area for investigation and may provide novel targets for cancer prevention and therapy.

MATERIALS AND METHODS

Materials

B-estradiol was obtained from Sigma. PLC-gamma antibody was obtained from Cell Signaling (Beverely, MA, USA). DMEM cell culture media were obtained from Gibco-BRL (Carisbad, CA, USA). Charcoal-treated FBS and phenol-free cell culture media were obtained from Hyclone (Logan, UT, USA). All other materials and reagents were from standard commercial sources.

Cell culture

For all experimental analysis, MCF-7 breast cancer cell lines were used. The cell lines were transformed with the Tet-off inducible vector (35). Each cell line contained one of four vectors: an empty expression vector (control cell line), a vector expressing WT PTEN, a vector containing PTEN with the C124S mutation (C124S-PTEN) or a vector containing PTEN with the G129E mutation (G129E-PTEN) (35). The C124S-PTEN mutation renders PTEN lipid and protein phosphatase inactive, whereas the G129E-PTEN mutation results in a PTEN protein that is lipid-phosphatase inactive but protein-phosphatase active (35). Cell lines were maintained in DMEM supplemented with 10% FBS, 200 mg/ml of G418, 1 mg/ml tetracycline (Tet) and 100 U/ml each of penicillin and streptomycin (35). Media were changed and new Tet was added every 2 days. PTEN expression was induced by growing cells in the absence of Tet for 48 h prior to cell harvesting.

Cellular stimulation

Cells were plated at ~60% confluency in 7 ml of Tet-free media. After 24 h, the media were removed, cells were washed with PBS and fresh Tet-free media were added. After 24 h (48 h after plating), cells were harvested by scraping and lipids were extracted. In some experiments, 1 uCi/ml of [³H]-oleate was added for the first 24 h. Butanol, 1%, was added to the cell culture media 4 h prior to harvesting, where indicated. For experiments where cells were stimulated with B-estradiol, cells were trypsinized with phenol-free trypsin and grown in phenol-free DMEM containing 10% charcoal-treated FBS for 3 days, at which point they were stimulated with B-estradiol as indicated.

Lipid extraction and analysis

After cells were harvested, they were centrifuged and the resulting pellet was resuspended in 1 ml of dH₂O, sonicated, centrifuged again to pellet debris and the resulting supernatant was processed for lipid analysis. Lipid extraction was carried out according to the Folch method (36). Four hundred and fifty microliters of sample was brought up to 2 ml with PBS, to which 6 ml of CHCl₃:MeOH (1:1), 60 µl of HAc and 700 µl of dH₂O were added. Samples were vortexed for 1 min and subjected to centrifugation at 800g for 10 min. The upper aqueous phase was removed and 3 ml of MeOH:dH₂O: CHCl₃:HAc (50:49:3:1) was added to the tube, followed by a 30 s vortex and centrifugation at 800g for 10 min. The top layer was removed and this step was repeated. The organic layer, which consisted of the cellular lipids, was removed, concentrated under N_2(g), reconstituted with 100 µl of CHCl₃ and stored at 4°C (36) under N_2(g) until analysis. Individual phospholipids were isolated by thin layer chromatography (TLC). For analysis of PA, PC, PS and PE, lipids were separated in CHCl₃:MeOH:HAc (65:35:8). PBut was resolved from PA using the upper phase of ethyl acetate:2,2,4-trimethylpentane:HAc:dH₂O (65:10:15:50) (37). After TLC, phospholipids were visualized by I_2(g), scraped and analyzed. Phospholipid mass was determined by measuring phosphorous content using the lipid phosphorous assay (38). Radioactive samples were quantified by liquid scintillation counting. For purposes of comparison, the data from each manipulated line were compared with those from the control cell line, operationally defined as normalized to 100%.

Western blot

Cells were harvested into lysis buffer (62.5 mm Tris, pH = 6.8, and 0.5 M EDTA) containing protease and phosphatase inhibitors (0.75 mm PMSF, 2 µm/ml pepstatin, 2 µm/ml aprotinin and 2 µm/ml leupeptin, 0.2 mm NaOV, 10 mm B-glycerol phosphate and 25 mm NaF). Cells were sonicated and centrifuged at a low speed (800g) in order to remove unbroken cells and nuclei. For PTEN analysis, the whole cell lystate was analyzed, and for the analysis of PLC-gamma translocation, the resulting supernatant was subsequently subjected to ultracentrifugation at 207 000g for 1 h. The resulting supernatant was taken as the cytosolic fraction, and the remaining pellet was resuspended in lysis buffer, sonicated and collected as the membrane fraction. Protein concentration was determined using the bicinchoninic method with bovine serum albumin as a standard (39). Twenty milligrams of protein was separated by 7% SDS–PAGE, according to the Laemmeli method (40). Proteins were then transferred electrophoretically to nitrocellulose (41), and equal loading was confirmed by staining with Ponceau S solution. The membranes were blocked with 5% non-fat milk in TBS-T (10 mm Tris, pH = 7.0, 100 mm NaCl and 0.1% Tween-20) for 1 h. Membranes were probed with either anti-PTEN or an anti-PLC-gamma antibody (1:1000 dilution in 3% BSA) for 2 h and then washed for 1 h with six TBS-T rinses. Blots were then incubated with a rabbit horseradish peroxidase-linked secondary antibody (1:2500 dilution in 5% non-fat milk) for 1 h. Blots were rinsed again with six TBS-T washes for a total of 1 h. Proteins were visualized by chemiluminescence, according to the manufacture's recommendations.

ACKNOWLEDGEMENTS

C.A.A.-B. wishes to thank Michelle Sinden and Rosemary Teresi for technical advice and critical discussions related to the project. Part of this work was performed in partial fulfillment for graduation with distinction in the Department of Evolution, Ecology and Organismal Biology, The Ohio State University (to C.A.A.-B.). These data were presented, in part, as a platform session at the 54th Annual Meeting of the American Society of Human Genetics, Toronto, Canada, November 2004. C.A.A.-B. was funded by a Goldwater Scholarship and an AACR-Thomas J. Bardos Undergraduate Scholarship. This work was funded in part by the Susan G. Komen Breast Cancer Research Foundation (BCTR 2000 462 to C.E.), the American Cancer Society (RSG-02-151-01CCE to C.E.) and the National Cancer Institute (R01CA118980 to C.E.). C.E. is a recipient of the Doris Duke Distinguished Clinical Scientist Award. Funding to pay the Open Access publication charges for this article was provided by Cleveland Clinic Genomic Medicine Institute.

Conflict of Interest statement. The authors do not have any conflict of interest to declare in this study.

Author notes

†

The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors.