Background
Germ line mutations in the
PTEN gene are present in 80% of Cowden Syndrome (CS), and in 60% of Bannayan-Riley-Ruvalcaba syndrome (BBRS) [
1]. CS patients are predisposed to breast, thyroid, skin, and endometrial malignancies [
2].
PTEN mutations are found scattered along exon 1 to 8 in both CS and BBRS patients. Significant numbers of mutations (>30%) are found in exon 5 affecting codon 123-131 within the core catalytic domain [
3]. However, germ-line mutation in the
PTEN gene is uncommon in early onset (<40 year old) breast cancer patients with wild-type
BRCA1 allele [
4]. Although the loss of heterozygosity at chromosome 10q23 region has been reported in 17 out of 42 (40%) of invasive breast carcinomas,
PTEN mutations in sporadic breast carcinomas are rare [
5‐
7]. However, PTEN may still play a role in sporadic breast cancer due to its reduced expression [
8].
The
PTEN gene encodes a 54 kDa lipid phosphatase with specificity towards phosphatidylinositol (3,4,5) triphosphate (PIP
3) [
9]. The 185-amino acid (aa) catalytic domain in the amino(N)-terminus is followed by a phospholipid-binding C2 domain between aa186-353. PTEN C2 domain resembles that of the Ca
2+-independent membrane recruitment motif found in Protein Kinase C isotypes [
10]. While the membrane-facing side of C2 domain is characterized by the polybasic CBRIII loop between aa260-269, the cytosolic-facing portion features an unstructured region of 33-aa between codon 282-314, referred to as the C2 loop [
11]. Interestingly, codon Lys289 (K289) within the C2 loop is a target site for mono-ubiquitination and is implicated in nuclear import [
12]. Additional ubiquitination sites have also been identified in the N-terminus but these sites are modified by polyubiquitination and play a role in the stability of PTEN through proteosome-mediated degradation [
13]. These ubiquitination events have been shown to be catalyzed by NEDD4-1, a ring domain containing E3 ligase [
13].
A mutation in the PTEN C2 loop has been identified in a case of CS at codon K289 [
12]. The K289E mutation disrupts monoubiquitination and impedes PTEN nuclear import. The resulting loss of growth suppression may explain the numerous intestinal polyps found in this patient. In this manuscript, we report the characterization of another C2 loop mutant uncovered during a screen for aberrant PTEN protein expression in a panel of human breast cancer cell lines.
Methods
Cell cultures
All cell lines were obtained from the cell bank of Dr. Stuart Aaronson (Mount Sinai School of Medicine) and maintained in DMEM supplemented with 10% FBS. Pten
-/- MEFS were a gift from Dr. Hong Wu (UCLA). For transfection of 293T cells, 1 × 106 cells were transfected in 100-mm plates with 10 μg of DNA using 12 μl of Lipofectamine2000 (Invitrogen) in 6 ml of serum- and antibiotic-free DMEM for 4 h. Cells were lysed after 48 h of incubation. For gene transfer in PC3 cells, 1 × 105 cells per well in a 6-well plate were transfected with 2-4 μg of DNA using 2 μl of Lipofectamine2000 (Invitrogen) in 1 ml of serum- and antibiotic-free DMEM for 4 h. Cells were analyzed after 48 h incubation. For retroviral gene transfer, 10 μg of pBabe-puro expression vectors was co-transfected with either 10 μg of pCL-ECO or pCL-AMPHO helper viral DNA for Pten
-/- MEFS and U87MG, respectively. Cultures were infected with predetermined volume of viral supernatants to generate similar ectopic gene expression. After 6 h of incubation, cells were selected in 1.0-1.5 μg/ml of puromycin and marker-selected mass cultures were used in subsequent analysis. For the electroporation of MCF7 cells, approximately 1.5 × 106 cells were electroporated with 2 μg of expression plasmids using the nucleofection kit according to the manufacturer's instructions (Amaxa). For lipofection of MCF7, approximately 3 × 105 cells in 6-well plates were transfected with 2 μg of plasmid DNA and 5 μl of lipofectamine 2000 (Invitrogen) for 6 h. Cells were analyzed after 24 to 48 h.
Sequencing analysis
cDNAs were generated from total RNA of MDA-MB-453 using SuperScript® III First-Strand Synthesis System (Invitrogen). PCR fragments were generated with the following primer pairs covering the entire coding region of PTEN: Exon1 and 2: P1 - 5' atgacagccatcatcaaagagatc 3', P8 - 5' aatattgttcctgtatacgccttc 3' Exon3 and 4: P16 - 5' agacttgaaggcgtatacagg 3', P76 - 5' gtcatcttcacttagccattggt 3' Exon5: P26 - 5' cccttttgtgaagatcttgac 3', P41 -5' cagtgccactggtctataatccag 3' Exon6 and 7: P3 - 5' ctggattatagaccagtggcactg 3', P70 - 5' ctgtttgtggaagaactctac 3' Exon8: P71 - 5' gtagagttcttccacaaacag 3', P24 - 5' cagcttcaccttaaaatttgg 3' Exon9: P34 - 5' gccaaccgatacttttctccaaat 3', P2a - 5' tcagacttttgtaatttgtgtatg 3' Amplified fragments were resolved on agarose gels and PCR products were excised and purified by a gel extraction kit (Qiagen). Amplified products were subjected to sequencing analysis using the primers listed above.
Expression plasmids
All expression plasmids have been reported previously [
14‐
16]. The
PTEN E307K mutant plasmid was constructed using PCR-based site-directed mutagenesis and cDNA was subcloned in the
BamHI and
EcoRI sites of both pBabe-puro and pCEFL-KZ-AU5 expression vectors.
Reagents
LY294002 and rapamycin were purchased from Sigma. Cisplatinum was obtained from Sicor, Inc.
Antibodies
The anti-PTEN mouse monoclonal antibody (A2B1, Santa Cruz), anti-PTEN rabbit polyclonal antibody (#9552, Cell Signaling), anti-AU5 monoclonal antibody (Covance), anti-Akt, anti-p-S473-Akt, anti-p38 antibodies (Cell Signaling), anti-EGFR, anti-SP1, anti-tubulin, and anti-HRP-actin (Santa Cruz) antibodies were purchased from the indicated commercial sources.
Saponin subcellular fractionation
Cells were solubilized in 10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.01% Saponin, 5 mM EDTA, 2 mM EGTA, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1 mM Na3VO4, 1 mM Na4P2O7, and 10 mM NaF for 20 min on ice. The lysates were centrifuged for 30 min at 18,000 × g. The supernatant was collected as the cytosolic fraction. The pellet was resuspended in 50 mM HEPES, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1.5 mM MgCl2, 1 mM EGTA, 10% glycerol, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1 mM Na3VO4, 1 mM Na4P2O7, 10 mM NaF, and centrifuged for 30 min at 18,000 × g. The supernatant was collected as the membrane fraction.
Nuclear-cytosol fractionation
Nuclear and cytoplasmic fractions were prepared using the NE-PER fractionation kit (Pierce) according to a modified manufacturer's protocol.
Purification of recombinant PTEN from insect cells
Sf9 cells were infected with baculoviruses harboring His-tagged PTEN WT and E307K expression plasmids for 5-7 days. Monomeric PTEN was purified as described previously [
16].
Phosphatase activity assay
Recombinant PTEN was incubated with 50 μM synthetic diC8-PI(3,4,5)P3 (Echelon) with or without 50 μM synthetic diC8-PI(4,5)P2 (Echelon), both in soluble form for 10 min at 37°C in a final volume of 25 μl. The reaction buffer was in 10 mM DTT and 25 mM Tris-HCl, pH 8.0. The reaction was stopped with 100 μl of malachite green reagent (Echelon). The amount of phosphate released was measured by reading the absorbance at 620 nm. For measuring PTEN activity in vitro, 5 × 106 293T cells were transfected with 15 μg of expression plasmids using 9 μl of lipofectamine 2000 (Invitrogen). After 48 h, cells were solubilized in 600 μl of a lysis buffer containing 25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Triton X-100, 2 mM DTT, 1 mM EDTA, 1 mM PMSF, 10 μg/ml aprotinin, and 10 μg/ml leupeptin. Approximately 3 mg of total cell extracts were immunoprecipitated with 5 μg of an anti-AU5 antibody for 4 h and immunocomplexes were affinity absorbed onto 30 μl of GammaBind G sepharose beads. PTEN phosphatase activity was then assayed as described above.
Discussion
In this report, we have characterized a unique PTEN C2 loop mutant present in a widely used human metastatic carcinoma cell line, MDA-MB-453. This tumor was originated from a 42 year old female diagnosed with breast cancer in 1970 at MD Anderson Cancer Center [
26]. A right mastectomy was performed followed by radiation and chemotherapy. Tumor relapse was recorded in 1974 and after further treatments, the patient died from the disease in 1976. MDA-MB-453 represents a molecular apocrine breast cancer, which is estrogen receptor-negative but androgen receptor-positive, and is commonly detected in patients with CS [
27,
28]. However, we do not have evidence that the E307K mutation represents a germ line mutation similar to those found in CS patients. The lack of access to germ line tissues hampers our efforts in addressing this possibility.
Mutations in PTEN C2 loop are not without precedent. As mentioned above, a K298E mutation was uncovered in a CS patient and the nuclear PTEN levels in the dysplastic intestinal polyps were reduced [
12]. More interestingly, PTEN C2 loop is highly acidic in nature with 10 out of 32 (~31%) amino acids are either glutamate or aspartate residues. The substitution with a basic lysine residue at E307 is expected to introduce a considerable change to the charge surfaces. Indeed, a search of the Sanger Tumor Gene Mutation Database has identified similar E to K substitutions at codons E284, E288, and E291. Coincidentally, these mutations were either from endometrial or vulva cancers [
29‐
31]. This cluster of E to K mutations highlights the likelihood that C2 loop mutations play a role in either the initiation or progression of certain human malignancies. It is also note worthy that there was no evidence of a loss of heterozygosity in the MDA-MB-453 tumor as the WT protein species is expressed to a similar level. Whether the E307K mutant can act as a dominant acting oncogene or has dominant inhibitory effects on the WT counterpart is currently unknown. Alternatively, haploinsuffiency may play a role in the progression of the MDA-MB-453 tumor as has been demonstrated for the
PTEN tumor suppressor gene [
32].
The E307K mutant displays similar lipid phosphatase activity as the WT protein. This is not surprising since eliminating the 24-aa C2 loop between aa286-309 did not significantly altered PTEN enzymatic activity and membrane binding affinity [
11]. The finding of a higher level of the E307K mutant in the saponin soluble membrane fraction is intriguing. One possibility is that the substituted lysine may serve as a
de novo binding interface with the negatively charged membrane. Previously, our group and others have identified phosphorylation sites in the PTEN C-terminal tail that may serve to mask C2 domain from the plasma membrane [
16,
33,
34]. One speculation is that E307K may serve as a decoy binding site for C-terminal tail and thereby unmasking the C2 domain for membrane targeting. Demonstrating this would need biophysical evidence but will pose considerable challenge due to the unstructured nature of the C2 loop [
11].
The greater membrane localization of the E307K mutant is consistent with the observation of greater suppression of p-Akt levels in both
Pten
-/- MEFS and MCF7 cells. This finding is, however, inconsistent with the fact that E307K shows similar growth suppressing activity as WT. A plausible explanation is that Akt activation does not correlate with cell proliferation in
Pten
-/- MEFS. Alternatively, the threshold of repressed p-Akt level sufficient for growth suppression is rather low for differences to be registered. Also, it is counterintuitive that breast cancer cells would acquire a mutation in PTEN that has greater ability in suppressing the PI3-K pathway. At present, we do not have a solid explanation to this paradox. One speculation is that since Akt1 has been shown to block breast cancer cell migration [
35,
36], the presence of a E307K mutant may repress Akt1 activation and promote metastatic growth. An alternative explanation is that the persistent suppression of PI3-K > Akt pathway in cells harboring the E307K mutant may promote secondary mutations that can subvert this repression. Indeed, MDA-MD-453 harbors a H1047R oncogenic mutation in the
PIK3CA gene [
18]. This is consistent with the fact that the PI3-K inhibitor, LY294002, could completely abolish p-Akt in this tumor line.
The heightened polyubiquitination observed with the E307K mutant reaffirms the role of C2 loop in the post-translational modification of PTEN. However, it is not clear which sites are being modified. It is tempting to speculate that the aberrant polyubiquitination of E307K can result in PTEN nuclear exclusion since monoubiquitination appears to be critical for nuclear import [
12]. In addition, the increase in polyubiquitination does not result in greater degradation of the PTEN E307K protein since its level is almost identical to that of WT.
Several cellular functions have been ascribed to nuclear PTEN including p53-dependent oxidative stress response [
22], protection against DNA double strand breaks [
23], and growth suppression through binding to the APC-CDH1 complex [
37]. Currently, it is unclear if PTEN E307K mutant binds p53 or APC-CDH1 with altered affinity. The inability of PTEN E307K mutant in sensitizing cells to cisplatinum would argue for a negative role of nuclear PTEN in promoting DNA damage response. Given the pro-survival function of Akt, our data is inconsistent with the fact that PTEN K307E mutant possesses a greater ability in suppressing Akt activation in
Pten
-/- MEFS. To resolve these issues, clearly defining the Akt isoforms being regulated would be important. To clearly define the role of E307K mutation, it may be necessary to eliminate this mutant
PTEN allele by either gene silencing or somatic gene knockout in MDA-MB-453. In summary, this study provides an initial characterization of a PTEN C2 loop mutant present in a breast cancer cell line. To our knowledge, MDA-MB-453 is the only known cell line that harbors a PTEN C2 loop mutation. Thus, this study highlights the availability of a commonly used breast cancer cell line for future research into the role of PTEN C2 loop in nuclear/cytosol partitioning.
Acknowledgements
The authors thank Dr. Xuejun Jiang (Memorial Sloan Kettering, NY) for providing the NEDD4-1 proteins, and both Drs. Zhen-Qiang Pan and Kenneth Wu (Mount Sinai School of Medicine, NY) for the purified E1 and E2 enzymes. The authors also thank Asoka Banno for technical assistance.
Financial Support: G. Singh was supported by an NCI pre-doctoral training grant, T32CA078207. L. Odriozola, was supported by a MECD-Fulbright Fellowship. A. Chan was supported by NIH CA095063, CA133669, Army Breast Cancer Research Program IDEA award, DAMD17-03-1-0682, Midwest Athletes Against Childhood Cancer (MACC) Fund, Advancing a Healthier Wisconsin Award, and Wisconsin Breast Cancer Showhouse Research Award.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
All authors read and approved the final manuscript. GS made the observations of the nuclear exclusion and membrane localization of the E307K mutant in MDA-MB-453. LO performed the purification and lipid phosphatase assays. HG, and CRK performed both growth assays and Western blotting analysis. AMC analyzed the data and wrote the manuscript.