Many cancers, including breast cancer, harbor loss-of-function mutations in the catalytic domain of phosphatase and tensin homologue deleted on chromosome 10 (PTEN) or have reduced PTEN expression through loss of heterozygosity and/or epigenetic silencing mechanisms. However, specific phenotypic effects of PTEN inactivation in human cancer cells remain poorly defined without a direct causal connection between the loss of PTEN function and the development or progression of cancer. To evaluate the biological and clinical relevance of reduced or deleted PTEN expression, a novel in vitro model system was generated using human somatic cell knockout technologies. Targeted homologous recombination allowed for a single and double allelic deletion, which resulted in reduced and deleted PTEN expression, respectively. We determined that heterozygous loss of PTEN in the nontumorigenic human mammary epithelial cell line MCF-10A was sufficient for activation of the phosphoinositide 3-kinase/AKT and mitogen-activated protein kinase pathways, whereas the homozygous absence of PTEN expression led to a further increased activation of both pathways. The deletion of PTEN was able to confer growth factor–independent proliferation, which was confirmed by the resistance of the PTEN−/− MCF-10A cells to small-molecule inhibitors of the epidermal growth factor receptor. However, neither heterozygous nor homozygous loss of PTEN expression was sufficient to promote anchorage-independent growth, but the loss of PTEN did confer apoptotic resistance to cell rounding and matrix detachment. Finally, MCF-10A cells with the reduction or loss of PTEN showed increased susceptibility to the chemotherapeutic drug doxorubicin but not paclitaxel. [Cancer Res 2009;69(21):8275–83]

Phosphatase and tensin homologue deleted on chromosome 10 (PTEN) is a tumor suppressor gene that dephosphorylates phosphatidylinositol 3,4,5-trisphosphate, the product of the lipid kinase phosphoinositide 3-kinase (PI3K). PTEN antagonizes activated PI3K to maintain normal cell growth or arrest, survival, or apoptosis. PTEN and PI3K exist in a tight regulatory loop, and a reduction or deletion of PTEN or the acquisition of an activating mutation in PI3K leads to abnormal activation of the PI3K pathway.

PTEN is the second most frequently mutated gene in human cancers following TP53. Immunohistochemistry studies of tumors from breast, pancreatic, and ovarian cancers have shown a loss of PTEN protein in 30% to 50% of samples (15). In breast cancer, this loss correlates highly with lymph node metastasis (5). Additionally, germ-line mutations in the PTEN gene are associated with multineoplastic, autosomally dominant human syndromes such as Cowden's disease and Bannayan-Zonana syndrome. Each syndrome features a predisposition to the formation of different malignancies, including breast cancer in 20% to 50% of the affected females (6, 7). The high frequency of reduction or loss of PTEN in breast cancer suggests its potential role in initiation and/or progression of human breast cancer.

Previously, the effects of PTEN loss have primarily been measured in tumor cell lines, which harbor numerous other transforming and oncogenic mutations. This has made it difficult to determine which phenotypes are directly conferred by the loss of PTEN and to define the stages of tumorigenesis that are specifically altered in cells with PTEN loss. To elucidate the connection between PTEN loss and the initiation and/or progression of human breast carcinomas, somatic cell gene targeting technology was used to more closely mimic the physiologic reduction or loss of PTEN in epithelial cells. This in vitro model system involved targeted homologous recombination to disrupt each allele of PTEN in the nontumorigenic mammary epithelial cell (MEC) line, MCF-10A. Using these isogenic somatic cell PTEN knockout lines, we determined that PTEN loss induces not only the activation of the PI3K pathway but also the mitogen-activated protein kinase (MAPK) pathway. This increase in pathway activation led to growth factor–independent proliferation that was suppressed with either PI3K or MAPK inhibitors. PTEN loss in MCF-10A cells also increased anchorage-independent survival but was insufficient to confer anchorage-independent growth. Also, of potential clinical importance, PTEN loss confers susceptibility to the chemotherapeutic drug, doxorubicin, but not paclitaxel, two agents commonly used for breast cancer therapy.

Cell lines and cell culture

MCF-10A cells were purchased from the American Type Culture Collection and maintained as described previously (8) supplemented with 0.1 μg/mL cholera toxin. Minimal assay medium was composed of DMEM/F-12 without phenol red, 1% charcoal stripped dextran–treated fetal bovine serum (Hyclone), and 100 units/mL penicillin-streptomycin without exogenous growth factors. MCF-10A.Bcl2 cells were created by stable transfection with the pBP/Bcl2 expression vector (9) and maintained in MCF-10A growth medium supplemented with 2.5 μg/mL puromycin. Cells were maintained in a 37°C incubator with 5% CO2.

PTEN+/− and PTEN−/− cell line generation

Heterozygote clones were created as described previously (10). A second construct was generated to delete exon 2 of PTEN on the second allele. Briefly, sequences with exact homology to intronic regions flanking exon 2 were cloned into the pAAV-MCS (Stratagene) via the pSept vector (11, 12). The adeno-associated virus was generated using the AAV Helper-free system from Stratagene following the manufacturer's instructions. Individual G418-resistant clones were tested via PCR for the presence of a homozygote PTEN knockout. Positive clones were treated with a Cre recombinase virus to excise the IRES-neoR gene. Cells were maintained in MCF-10A growth medium as described above.

Proliferation assays

Cell were seeded in quadruplicate at 2.0 × 103 per well in 96-well plates in minimal assay medium. The next day, the appropriate medium with or without drug was added. Cell viability was quantified using MTT (Sigma). After MTT treatment, the medium was removed, the converted dye was solubilized in 0.01 mol/L glycine in DMSO, and the absorbance (450 nm) was measured. For growth in the absence of drugs, a set of cells were exposed to MTT on day 0 to accurately assess starting cell number. Growth was calculated as a percent above cell number on day 0. For the erlotinib and gefitinib (LC Laboratories) studies, untreated wells of each cell line were used as the control of calculation percent viability.

Western blot analysis

Cell lysates were prepared in radioimmunoprecipitation assay lysis buffer [0.5 mol/L Tris-HCl (pH 7.4), 1.5 mol/L NaCl, 2.5% deoxycholic acid, 10% NP-40, 10 mmol/L EDTA] supplemented with protease inhibitor cocktail EDTA-free (Roche) and phosphatase inhibitor cocktail II (Calbiochem). Western blotting was done using NuPage gels (Invitrogen). Primary antibodies for PTEN, pAKT (S473), AKT, phosphorylated extracellular signal-regulated kinase 1/2 (pERK1/2), ERK1/2, and poly(ADP-ribose) polymerase (PARP; Cell Signaling) and glyceraldehyde-3-phosphate dehydrogenase (Abcam) were used at the manufacturers' recommended dilutions.

Flow cytometry

For sub-G1 analysis, cells were ethanol-fixed and treated with RNase A (1 mg/mL) and propidium iodide (20 μg/mL). Cells were analyzed by a Becton Dickinson LSR-II at the Flow Cytometry Core Laboratory, CVD Immunology Group at the University of Maryland.

Survival assays

Cell were seeded in quadruplicate at 1.5 × 104 per well in 96-well plates in minimal assay medium. The next day, the appropriate concentrations of doxorubicin (Calbiochem) or paclitaxel (Invitrogen) were added to the cells. On day 5, cell viability was quantified using MTT. Untreated wells of each cell line were used as the control of calculation percent viability.

PTEN heterozygous and homozygous loss promotes activation of both PI3K and MAPK pathways in MECs

Reduction of PTEN expression or complete loss is observed in ∼40% of human breast cancers (2, 5, 13). To determine the oncogenic phenotype of reduced or absent PTEN expression in human breast epithelial cells, we generated isogenic human MECs using MCF-10A cells. MCF-10A breast epithelial cells were chosen for targeted PTEN knockout because these cells are human, mostly diploid, nontumorigenic cell line and karyotype analysis of late-passage cells show genetic stability (data not shown). The use of nontumorigenic cell lines allows us to assess any oncogenic effects directly resulting from loss of PTEN. Several independently derived heterozygous knockout clones (PTEN+/−) were identified containing one active PTEN allele. Subsequent targeting of the second allele was accomplished to yield homozygous PTEN (PTEN−/−) knockout clones.

Somatic cell gene knockout was accomplished via homologous recombination between the genomic locus and the targeting vector to delete exon II of PTEN and replace it with a promoterless, IRES-neoR gene flanked by LoxP sites (Fig. 1A). For each round of targeting, positive clones were verified by PCR. Subsequent removal of the IRES-neoR cassette was accomplished by treatment with Cre recombinase. At least three single, independent isogenic PTEN+/− and PTEN−/− clones from separate infections were isolated and used to account for any clonal variations.

Figure 1.

Single and biallelic deletion of the PTEN gene. A, to remove each copy of PTEN, our targeting construct contained a neomycin resistance cassette flanked by LoxP sites and homologous sequences to the intronic regions 5′ and 3′ of exon 2. B, MCF-10A cells (lane 1) and PTEN+/− (lanes 2-4) and PTEN−/− (lanes 5-7) clones were harvested at subconfluence during exponential growth. PTEN expression is reduced in PTEN+/− clones and absent in PTEN−/− clones. With a reduction and deletion of PTEN, activated AKT (pAKT) is increased (top bands), whereas total AKT levels remain equivalent. Activated ERK (pERK1/2) levels are also increased in the clones. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loading control.

Figure 1.

Single and biallelic deletion of the PTEN gene. A, to remove each copy of PTEN, our targeting construct contained a neomycin resistance cassette flanked by LoxP sites and homologous sequences to the intronic regions 5′ and 3′ of exon 2. B, MCF-10A cells (lane 1) and PTEN+/− (lanes 2-4) and PTEN−/− (lanes 5-7) clones were harvested at subconfluence during exponential growth. PTEN expression is reduced in PTEN+/− clones and absent in PTEN−/− clones. With a reduction and deletion of PTEN, activated AKT (pAKT) is increased (top bands), whereas total AKT levels remain equivalent. Activated ERK (pERK1/2) levels are also increased in the clones. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loading control.

Close modal

MCF-10A PTEN+/− clones showed a decrease in PTEN levels (Fig. 1B,, lanes 2-4) and PTEN−/− clones completely lacked PTEN expression (Fig. 1B,, lanes 5-7). The absence of PTEN mRNA in PTEN−/− clones was verified by real-time PCR (data not shown). All PTEN−/− clones maintained increased pAKT levels over their PTEN+/− and parental counterparts. Interestingly, activated ERK (pERK1/2) levels were also increased in PTEN−/− clones over the MCF-10A parental cells. Although PTEN+/− clones showed an increase in activated ERK levels from the parental cells, the increase was less dramatic or consistent among PTEN+/− clones (Fig. 1B). However, in conjunction with an increase in pERK1/2 in PTEN−/− clones, decreased total ERK levels were consistently observed.

PTEN loss confers growth factor–independent proliferation

To determine whether the activation of the PI3K and MAPK pathways altered proliferation rates, MCF-10A, PTEN+/−, and PTEN−/− clones were analyzed over a period of 9 days. Interestingly, at early passage, PTEN+/− and PTEN−/− clones with reduced or deleted PTEN grew significantly slower than parental MCF-10A cells (Fig. 2A; P < 0.05), whereas, at later passage, MCF-10A cells maintain a similar growth rate to that of their earlier-passage counterparts; the proliferation rates of PTEN+/− and PTEN−/− clones increase over time (Fig. 2A). The parental MCF-10A, PTEN+/−, and PTEN−/− clones do not undergo any significant cell death over the first 7 days due to the absence of a sub-G1 population (Supplementary Fig. S1). However, once the cells achieve contact inhibition by day 7, all cells begin to die as shown by a drop in viability and the presence of a sub-G1 peak. The increase proliferation rate is likely due to a variety of mechanisms downstream of AKT activation, such as increased cyclin D1 expression, inhibition of forkhead transcription factors, or reduction of p27Kip1, all of which positively regulate G1-S cell cycle progression (reviewed in ref. 14).

Figure 2.

PTEN loss confers growth in minimal medium. A, early- and late-passage growth of MCF-10A cells and PTEN+/− and PTEN−/− clones in normal MCF-10A growth medium. Points, mean of two independent experiments done in quadruplicate; bars, SD. B, cell growth of MCF-10A cells and PTEN+/− and PTEN−/− clones after exposure to increasing doses of the EGF receptor antagonists gefitinib and erlotinib. Points, mean of three independent experiments done in quadruplicate; bars, SD. C, cell growth in minimal medium. Points, mean of two independent experiments done in quadruplicate; bars, SD. On day 6, MCF-10A, PTEN+/−, and PTEN−/− cells were fixed and stained with a solution of 10% PBS-buffered formalin and 0.25% crystal violet. Representative wells from MCF-10A, PTEN+/−, and PTEN−/− cells are shown. D, MCF-10A, two PTEN+/−, and two PTEN−/− clones were plated in either normal growth medium (NG) or minimal assay medium and harvested at the indicated times by direct addition of radioimmunoprecipitation assay lysis buffer.

Figure 2.

PTEN loss confers growth in minimal medium. A, early- and late-passage growth of MCF-10A cells and PTEN+/− and PTEN−/− clones in normal MCF-10A growth medium. Points, mean of two independent experiments done in quadruplicate; bars, SD. B, cell growth of MCF-10A cells and PTEN+/− and PTEN−/− clones after exposure to increasing doses of the EGF receptor antagonists gefitinib and erlotinib. Points, mean of three independent experiments done in quadruplicate; bars, SD. C, cell growth in minimal medium. Points, mean of two independent experiments done in quadruplicate; bars, SD. On day 6, MCF-10A, PTEN+/−, and PTEN−/− cells were fixed and stained with a solution of 10% PBS-buffered formalin and 0.25% crystal violet. Representative wells from MCF-10A, PTEN+/−, and PTEN−/− cells are shown. D, MCF-10A, two PTEN+/−, and two PTEN−/− clones were plated in either normal growth medium (NG) or minimal assay medium and harvested at the indicated times by direct addition of radioimmunoprecipitation assay lysis buffer.

Close modal

A well-known characteristic of MCF-10A MECs is their epidermal growth factor (EGF) requirement for cellular proliferation. Growth factor–independent proliferation is a common hallmark in cancer cells containing oncogenic phenotypes and aberrantly activated signaling (15). Because PTEN+/− and PTEN−/− clones have an increase in activated PI3K and MAPK pathways, we examined whether the increased activation of these pathways was sufficient to confer EGF-independent growth by treating the cells with increasing concentrations of the clinically administered EGF receptor small-molecule inhibitors gefitinib and erlotinib (Fig. 2B). Compared with the parental and PTEN+/− clones, PTEN−/− clones were significantly more resistant to growth inhibition via the EGF receptor inhibitors, indicating a decreased requirement of EGF for proliferation. To confirm this observation, MCF-10A cells and PTEN+/− and PTEN−/− clones were maintained in minimal assay medium devoid of exogenous growth factors for 9 days. As observed previously using compounds to disrupt EGF signaling, MCF-10A cells and PTEN+/− clones showed reduced growth (Fig. 2C). However, PTEN−/− cells survived and continued to slowly proliferate although considerably slower than in medium supplemented with EGF. At later passage, PTEN−/− cell proliferation in the absence of mitogens became more robust (Supplementary Fig. S2).

Immunoblots confirmed the increased activity of each pathway in the presence and absence of mitogens (Fig. 2D). After 24 h in growth medium, the levels of pAKT are highest in PTEN−/− clones and slightly more elevated in PTEN+/− clones compared with the MCF-10A parental cells. Assessment of the same set of cells grown in medium without mitogens showed reduced levels of pAKT. Although there is a minimal level of detectable pAKT in the parental and PTEN+/− clones, PTEN−/− clones continue to maintain elevated pAKT levels at 24, 48, and 72 h.

pERK levels were also slightly elevated in all PTEN+/− and PTEN−/− clones compared with the MCF-10A parental cells when grown in normal culture medium (Fig. 2D). The levels of activated ERK were similar between PTEN+/− and PTEN−/− clones. This result differed from the earlier immunoblot results (Fig. 1B) likely due to different harvest times after replating. Cells in Fig. 1B were harvested during exponential growth 3 days after replating, whereas cells in Fig. 2D were harvested only 24 h after replating when they are not yet in exponential growth. However, after 3 days in medium devoid of growth factors, PTEN+/− cells have increased pERK levels over the MCF-10A parental cells, and PTEN−/− cells have pERK levels higher than PTEN+/− clones, similar to the pattern of ERK phosphorylation seen after 3 days in culture medium (Fig. 1B). This pattern of increasing levels of activated ERK, from MCF-10A parental cells to their PTEN−/− isogenic counterparts, indicates that, even under growth factor–reduced conditions, loss of PTEN allows for activation of the ERK signaling pathway.

To determine if the improved cell growth in PTEN−/− cells resulted from differences in cell cycling or reduced apoptosis, we performed flow cytometry measurements of DNA content throughout the growth in minimal medium (Fig. 3). These results showed that both MCF-10A and PTEN+/− cells undergo apoptotic DNA cleavage beginning on day 3 that increases to 87% and 81%, respectively, by day 9, whereas PTEN−/− cells show no signs of apoptosis. The increased viability of PTEN−/− cells is primarily from resistance to apoptosis rather than a difference in cell cycling. The elevation of apoptosis in MCF-10A and PTEN+/− cells during exposure to minimal medium is clearly sufficient to offset any cell growth and keep the cell population from increasing.

Figure 3.

MCF-10A and PTEN+/− cells undergo apoptosis, whereas PTEN−/− cells survive in minimal medium. A, flow cytometry analysis of propidium iodide–stained MCF-10A and PTEN+/− cells reveals massive cell death beginning at day 3 and continuing through day 9 until only a sub-G1 population is left. However, PTEN−/− cells continue to survive and slowly cycle due to the presence of a minimal G2 peak present until day 9 when the cells become growth arrested at confluence. B, graphical representation of the cell death in minimal medium.

Figure 3.

MCF-10A and PTEN+/− cells undergo apoptosis, whereas PTEN−/− cells survive in minimal medium. A, flow cytometry analysis of propidium iodide–stained MCF-10A and PTEN+/− cells reveals massive cell death beginning at day 3 and continuing through day 9 until only a sub-G1 population is left. However, PTEN−/− cells continue to survive and slowly cycle due to the presence of a minimal G2 peak present until day 9 when the cells become growth arrested at confluence. B, graphical representation of the cell death in minimal medium.

Close modal

Growth factor–independent proliferation due to PTEN loss can be inhibited by pharmacologic blockade of PI3K and MAPK pathways

To confirm the requirement of active PI3K and MAPK pathways for continued cell proliferation in the absence of growth factors, PTEN−/− cells were treated with inhibitors of each pathway and growth factor–independent proliferation was assessed. Only PTEN−/− cells were used in this experiment because the parental and heterozygote clones do not grow under these conditions (Fig. 2C). Increasing concentrations of the PI3K inhibitor, LY294002, in minimal assay medium was added to the cells, and after 5 days, a dose-dependent inhibition of growth was observed. The addition of 10 μmol/L LY294002 led to nearly complete inhibition of proliferation of all PTEN−/− clones (Fig. 4A). Similarly, PTEN−/− clones were grown in the presence of the MEK1/2 inhibitor, U0126 (Fig. 4B). There was also a dose-dependent growth inhibition of PTEN−/− cells following exposure to the MEK inhibitor. In the presence of 1 μmol/L U0126, growth factor–independent proliferation was inhibited by >50% and almost completely inhibited with 2.5 μmol/L U0126.

Figure 4.

Inhibition of the PI3K and MAPK pathways suppresses proliferation in minimal medium of PTEN−/− MECs. A, percent growth of three PTEN−/− clones in minimal assay medium with increasing concentrations of LY294002. B, percent growth of three PTEN−/− clones in minimal assay medium with increasing concentrations of U0126. C, three PTEN−/− clones were plated in minimal assay medium and allowed to recover overnight. The next day, fresh minimal assay medium was added with or without 10 μmol/L LY294002 (L) or 10 μmol/L U0126 (U) for 1 h. Western blot analysis showing pAKT reduction in the cells treated with the PI3K inhibitor LY294002 and pERK reduction in the cells treated with the MEK inhibitor U0126.

Figure 4.

Inhibition of the PI3K and MAPK pathways suppresses proliferation in minimal medium of PTEN−/− MECs. A, percent growth of three PTEN−/− clones in minimal assay medium with increasing concentrations of LY294002. B, percent growth of three PTEN−/− clones in minimal assay medium with increasing concentrations of U0126. C, three PTEN−/− clones were plated in minimal assay medium and allowed to recover overnight. The next day, fresh minimal assay medium was added with or without 10 μmol/L LY294002 (L) or 10 μmol/L U0126 (U) for 1 h. Western blot analysis showing pAKT reduction in the cells treated with the PI3K inhibitor LY294002 and pERK reduction in the cells treated with the MEK inhibitor U0126.

Close modal

To verify inhibition of the PI3K and MAPK pathways by LY294002 and U0126, respectively, immunoblots were done. In the absence of the PI3K or MEK1/2 inhibitors, PTEN−/− clones displayed high levels of pAKT and pERK (Fig. 4A). After LY294002 treatment, all PTEN−/− clones showed a significant drop in pAKT levels. Following U0126 treatment, the levels of pERK dropped to almost undetectable levels.

Anchorage-independent survival and growth

Because PTEN loss highly correlates with increased breast cancer lymph node metastasis (1618), it was next determined if PTEN loss alone would lead to the transformation of nontumorigenic breast epithelial cells. Anchorage-independent growth in soft agar is a property of transformed cells that best correlates with in vivo tumorigenicity (19). MCF-10A nontumorigenic parental cells and PTEN−/− clones were plated in soft agar and incubated for 21 days. MCF-7 breast cancer line was used as a positive control for colony growth and only incubated for 14 days due to the formation of multiple, large colonies. As expected, MCF-10A cells did not form colonies. Likewise, PTEN−/− cells were unable to form colonies in soft agar and unable to form tumors in severe combined immunodeficient mice (n = 5) after 24 weeks (data not shown). However, colony formation in soft agar and in vivo tumor growth rely on anchorage-independent proliferation but are not a sufficient test for increased resistance to anoikis or apoptosis after matrix detachment. Because previous data have shown that activation of the PI3K pathway contributes to cell survival after detachment (20), MCF-10A, PTEN+/−, and PTEN−/− cells were next tested for apoptotic resistance during cell rounding and detachment. Included in these studies, as a control for resistance to cell rounding and anoikis, were MCF-10A cells overexpressing the antiapoptotic gene Bcl2 (8). PARP cleavage, an indicator of apoptosis, was examined after the cells were treated with latrunculin-A to induce cell rounding (Fig. 5A). Latrunculin-A is a specific inhibitor of actin polymerization that has been used to induce rapid rounding of MCF-10A cells while allowing the cells to maintain attachment to the tissue culture dishes (8). MCF-10A, PTEN+/−, and PTEN−/− cells treated with vehicle control in minimal assay medium showed similar, low levels of PARP cleavage. MCF-10A.Bcl2 cells, with verified resistance to apoptosis, had undetectable PARP cleavage. On addition of latrunculin-A, the MCF-10A parent line and PTEN+/− clones undergo significant PARP cleavage, whereas PTEN−/− cells maintain high levels of full-length PARP, similar to that of MCF-10A.Bcl2 cells. To examine whether PTEN−/− cells exhibited a general resistance to apoptosis, all cells were treated with the death receptor ligand, tumor necrosis factor–related apoptosis-inducing ligand (TRAIL). Binding of TRAIL to transmembrane death receptors stimulates apoptosis via the extrinsic pathway, which is independent of AKT and the mitochondria. Within 2 h of TRAIL-related apoptosis inducing ligand treatment, all MCF-10A cells and variants began to undergo apoptosis confirmed by the increase of cleaved PARP levels. The apoptotic resistance of PTEN−/− cells therefore seems restricted to the intrinsic apoptosis pathway, because it cannot prevent apoptosis that occurs downstream of mitochondrial apoptosis signaling. Additionally, to determine anoikis resistance, the cells were plated over low-attachment plates. Without tissue culture–treated plastic, the cells remained in suspension. After detachment for 24 h, MCF-10A parental cells and PTEN+/− clones undergo massive cell death as shown by high levels of cleaved PARP and the presence of a sub-G1 peak (Fig. 5B and C). PTEN−/− cells maintained high levels of full-length PARP and a lower percentage of cells in sub-G1, indicating their resistance to anoikis. pERK levels are only very slightly elevated in PTEN+/− clones; these cells maintained similar levels of PARP cleavage to the parental cells. Although the first PTEN−/− clone revealed an increase in pERK levels compared with the other PTEN−/− clone, no differences in the levels of cleaved PARP were observed between clones. Therefore, whereas MEK activation was required for continued cell growth (Fig. 4), the levels of ERK activation were relatively independent from apoptosis during detachment. It is more likely that AKT activation was responsible for the resistance to anoikis, because high levels of pAKT were maintained in the suspended PTEN−/− cells.

Figure 5.

PTEN loss promotes resistance to apoptosis on cell rounding and anoikis. A, MCF-10A, a PTEN+/− clone, a PTEN−/− clone, and MCF-10A.Bcl2 cells were plated in minimal assay medium. The next day, the medium was changed to fresh minimal medium with or without 5 μmol/L latrunculin-A (LA) to induce cell rounding for 24 h or 1 μg/mL TRAIL for 2 h and harvested at the indicated times by direct addition of radioimmunoprecipitation assay lysis buffer. B, MCF-10A, two representative PTEN+/− clones, two representative PTEN−/− clones, and MCF-10A.Bcl2 cells were incubated in suspension in normal growth medium for 24 h. C, flow cytometry analysis of MCF-10A, PTEN+/− clone 1, and PTEN−/− clone 1 incubated in suspension in DMEM/F-12 for 24 h. A larger percentage of MCF-10A (42.8%) and PTEN+/− cells (49.4%) undergo apoptosis than PTEN−/− cells (29.6%).

Figure 5.

PTEN loss promotes resistance to apoptosis on cell rounding and anoikis. A, MCF-10A, a PTEN+/− clone, a PTEN−/− clone, and MCF-10A.Bcl2 cells were plated in minimal assay medium. The next day, the medium was changed to fresh minimal medium with or without 5 μmol/L latrunculin-A (LA) to induce cell rounding for 24 h or 1 μg/mL TRAIL for 2 h and harvested at the indicated times by direct addition of radioimmunoprecipitation assay lysis buffer. B, MCF-10A, two representative PTEN+/− clones, two representative PTEN−/− clones, and MCF-10A.Bcl2 cells were incubated in suspension in normal growth medium for 24 h. C, flow cytometry analysis of MCF-10A, PTEN+/− clone 1, and PTEN−/− clone 1 incubated in suspension in DMEM/F-12 for 24 h. A larger percentage of MCF-10A (42.8%) and PTEN+/− cells (49.4%) undergo apoptosis than PTEN−/− cells (29.6%).

Close modal

PTEN loss sensitizes cells to the chemotherapeutic drug doxorubicin but not paclitaxel

Studies suggest that loss of PTEN expression correlate with poor prognosis as well as resistance to chemotherapies (21, 22). To determine whether PTEN loss mediates chemotherapeutic resistance and increases cell survival, the PTEN isogenic MCF-10A cells were exposed to increasing concentrations of doxorubicin and paclitaxel. Primary normal breast epithelial cells are alive and metabolizing but not actively proliferating. To mimic healthy, growth-arrested epithelial cells, MCF-10A cells were plated at high density (1.5 × 104) per well in a 96-well plate in minimal assay medium. After 24 h, the cells attached as ∼90% confluent monolayer. Although PTEN−/− cells have the ability to grow in minimal medium, they remain contact-inhibited and growth-arrested at confluence; therefore, the plating densities used matched those of the MCF-10A parental cells. Drugs were added after growth arrest to determine cell survival. Interestingly, PTEN+/− and PTEN−/− cells responded differently to the drugs. Doxorubicin similarly reduced the percentage of surviving cells in both PTEN+/− and PTEN−/− cells (Fig. 6A). Even at the low concentration of 5 nmol/L, doxorubicin reduced PTEN+/− and PTEN−/− clones by 10% and 24%, respectively. No difference in susceptibility was observed between the isogenic cells following exposure to paclitaxel (Fig. 6B).

Figure 6.

PTEN loss sensitizes MECs to the commonly used breast cancer chemotherapeutic doxorubicin but not to paclitaxel. Cell growth of MCF-10A cells and PTEN+/− and PTEN−/− clones after exposure to increasing doses of doxorubicin (A) and paclitaxel (B). Points, mean of three independent experiments done in quadruplicate; bars, SD.

Figure 6.

PTEN loss sensitizes MECs to the commonly used breast cancer chemotherapeutic doxorubicin but not to paclitaxel. Cell growth of MCF-10A cells and PTEN+/− and PTEN−/− clones after exposure to increasing doses of doxorubicin (A) and paclitaxel (B). Points, mean of three independent experiments done in quadruplicate; bars, SD.

Close modal

The loss of PTEN expression or the acquisition of activating PI3K mutations (PIK3CA) occurs in ∼50% to 75% of breast cancers, illustrating the importance of the PI3K pathway in breast cancer. Notably, loss of PTEN expression and PIK3CA mutation are mutually exclusive events (23) likely because PTEN and PI3K exist in a tight, regulatory loop, strictly controlling phosphatidylinositol levels. The loss of PTEN or the acquisition of an activating mutation in PIK3CA are reciprocal alterations, either of which would result in increased levels of phosphatidylinositol 3,4,5-trisphosphate and remove the selective pressure to convert phosphatidylinositol 3,4,5-trisphosphate to its phosphatidylinositol 4,5-bisphosphate counterpart. Our findings support the role of PTEN loss in breast cancer based on the ability of PTEN−/− cells to proliferate in the absence of growth factors and their resistance to anoikis. However, PTEN loss is insufficient to promote active tumorigenesis of the MCF-10A cells, suggesting a need for other oncogenic events (24). This result is contradictory to recent data in which the overexpression of two clinically relevant PI3K mutations (H1047R and E545K) conferred anchorage-independent growth of MCF-10A cells in soft agar (25). However, in these studies, the expression of the PIK3CA mutant cDNA is under control of a cytomegalovirus promoter that may increase PIK3CA expression to levels not observed in primary tumors or derived cell lines. In support of this hypothesis, knockin of the same PIK3CA activating mutations, H1074R and E545K, did not cause anchorage-independent growth of MCF-10A cells (26). Similar to MCF-10A PTEN−/− cells, knockin mutant PIK3CA cells were not tumorigenic, did not form colonies in soft agar, and did not alter acinar growth in three-dimensional Matrigel culture.

Although metastasis is the cause of 90% of human cancer deaths (27), the metastatic process presents numerous challenges to tumor cells, including apoptosis that results from detachment (anoikis) or cell shape change (amorphosis; ref. 28). Resistance to apoptosis allows tumor cells to survive these challenges (29) but does not promote immediate tumor outgrowth at the secondary site, yielding a period of tumor dormancy (30). There is currently tremendous clinical interest in such dormant tumor cells, because their presence in the bloodstream strongly predicts poor patient outcome in breast cancer (31, 32). The importance of defining the mechanisms that promote tumor dormancy is also emphasized by the observation that breast tumor patients who are diagnosed early with no detectable regional metastases have >30% chance of recurrence when followed for 10 to 15 years (33, 34). Our results indicate that PTEN loss induces a dormant tumor cell phenotype by promoting resistance to apoptosis without inducing complete anchorage-independent growth. Recent evidence shows that MECs that have not fully transformed to anchorage-independent growth are still fully capable of metastasizing to the lung in a dormant state and then recurring once growth-initiating oncogenes are activated (35, 36). Systems based on fibroblasts or exogenous overexpression of PI3KCA display active tumor growth, whereas our system based on homologous knockout of PTEN in MCF-10A MECs more effectively models the dormant phenotype of carcinoma cells. However, such dormant tumor cells are typically difficult to treat with traditional chemotherapies, because they persist without active cell division. Defining which types of chemotherapy are able to effectively target tumor cells in such a dormant state will be critical to treating metastatic recurrence.

A variety of chemotherapeutic agents converge on a common final pathway leading to apoptotic cell death. Certain studies have shown that activation of the PI3K pathway enhances the survival of cancer cells in response to such agents and contribute to chemotherapy resistance. However, these previous studies employed an overexpressed, constituently active AKT1 (37, 38), which may not recapitulate the physiologically active AKT levels due to PTEN loss or PIK3CA mutations, whereas other studies overexpressed an oncogene such as constitutively active Ras (39) or HER-2 (40) in the MCF-7 breast adenocarcinoma line, which already contains a PIK3CA mutation (E545K; ref. 41). Here, to more closely recapitulate physiologic levels of active AKT, we used the MCF-10A isogenic PTEN knockout clones to determine chemotherapeutic response to doxorubicin and paclitaxel. Surprisingly, PTEN−/− clones were more susceptible to doxorubicin than their parental PTEN-expressing counterparts. However, no difference in survival was observed between the isogenic clones when treated with paclitaxel. The susceptibility of PTEN+/− and PTEN−/− cells to doxorubicin and not paclitaxel may be explained by the different mechanisms of action of each drug. Paclitaxel is a microtubule-stabilizing compound that interferes with the normal breakdown of this cytoskeletal component. This drug immediately and adversely affects cell function as microtubule-inherent dynamic instability is necessary for their function to transport other cellular components. Doxorubicin is known to intercalate within the DNA and inhibition of topoisomerase II progression, eliciting DNA damage. On DNA damage, normal cells undergo growth arrest to either repair the damage or undergo apoptosis if the damage is substantial. However, constitutively active PI3K and pathway components have been shown to override DNA damage–induced cell arrest (4244). Haploinsufficiency and deletion of PTEN may allow for cell cycle progression and death due to massive DNA damage. Further work to elucidate the mechanisms by which PTEN expression loss may contribute to chemotherapy susceptibility is warranted.

There is a significant clinical relevance for the creation and characterization of this PTEN isogenic model system. Using this model, we have shown that PTEN expression loss in MECs results in constitutive AKT activation and induces multiple phenotypic alterations characteristic of breast tumor cells, including growth factor–independent proliferation and protection from anoikis. PTEN expression loss also confers increased susceptibility to doxorubicin but not paclitaxel. Together, these data support the notion that the cancer-associated PTEN expression loss may significantly contribute to breast cancer cell survival and tumor dormancy.

No potential conflicts of interest were disclosed.

Grant support: Maryland Cigarette Restitution Fund (K.E. Bachman), Susan G. Komen Breast Cancer Foundation grant PDF104506 (M.I. Vitolo), National Cancer Institute grants T32-DK067872 (M.B. Weiss), R01-CA115699 (T. Waldman), and R01-CA124704 (S.S. Martin), Breast Cancer Research Foundation (B.H. Park), and National Cancer Institute and General Medicine/NIH grants R01-CA107331 and R01 GM58888 (D.J. Weber).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Agnes Cheung for developing the MCF-10A.Bcl2 cell line and Regina Harley for flow cytometry experiments.

1
Bruni
P
,
Boccia
A
,
Baldassarre
G
, et al
. 
PTEN expression is reduced in a subset of sporadic thyroid carcinomas: evidence that PTEN-growth suppressing activity in thyroid cancer cells mediated by p27kip1
.
Oncogene
2000
;
19
:
3146
55
.
2
Depowski
PL
,
Rosenthal
SI
,
Ross
JS
. 
Loss of expression of the PTEN gene protein product is associated with poor outcome in breast cancer
.
Mod Pathol
2001
;
14
:
672
6
.
3
Ebert
MP
,
Fei
G
,
Schandl
L
, et al
. 
Reduced PTEN expression in the pancreas overexpressing transforming growth factor-β1
.
Br J Cancer
2002
;
86
:
257
62
.
4
Kurose
K
,
Zhou
XP
,
Araki
T
,
Cannistra
SA
,
Maher
ER
,
Eng
C
. 
Frequent loss of PTEN expression is linked to elevated phosphorylated Akt levels, but not associated with p27 and cyclin D1 expression, in primary epithelial ovarian carcinomas
.
Am J Pathol
2001
;
158
:
2097
106
.
5
Perren
A
,
Weng
LP
,
Boag
AH
, et al
. 
Immunohistochemical evidence of loss of PTEN expression in primary ductal adenocarcinomas of the breast
.
Am J Pathol
1999
;
155
:
1253
60
.
6
Liaw
D
,
Marsh
DJ
,
Li
J
, et al
. 
Germline mutations of the PTEN gene in Cowden disease, an inherited breast and thyroid cancer syndrome
.
Nat Genet
1997
;
16
:
64
7
.
7
Marsh
DJ
,
Zheng
Z
,
Zedenius
J
, et al
. 
Differential loss of heterozygosity in the region of the Cowden locus within 10q22-23 in follicular thyroid adenomas and carcinomas
.
Cancer Res
1997
;
57
:
500
3
.
8
Martin
SS
,
Leder
P
. 
Human MCF10A mammary epithelial cells undergo apoptosis following actin depolymerization that is independent of attachment and rescued by Bcl-2
.
Mol Cell Biol
2001
;
21
:
6529
36
.
9
Pinkas
J
,
Martin
SS
,
Leder
P
. 
Bcl-2-mediated cell survival promotes metastasis of EpH4 βMEKDD mammary epithelial cells
.
Mol Cancer Res
2004
;
2
:
551
6
.
10
Lee
C
,
Kim
JS
,
Waldman
T
. 
PTEN gene targeting reveals a radiation-induced size checkpoint in human cancer cells
.
Cancer Res
2004
;
64
:
6906
14
.
11
Topaloglu
O
,
Hurley
PJ
,
Yildirim
O
,
Civin
CI
,
Bunz
F
. 
Improved methods for the generation of human gene knockout and knockin cell lines
.
Nucleic Acids Res
2005
;
33
:
e158
.
12
Rago
C
,
Vogelstein
B
,
Bunz
F
. 
Genetic knockouts and knockins in human somatic cells
.
Nat Protoc
2007
;
2
:
2734
46
.
13
Garcia
JM
,
Silva
J
,
Pena
C
, et al
. 
Promoter methylation of the PTEN gene is a common molecular change in breast cancer
.
Genes Chromosomes Cancer
2004
;
41
:
117
24
.
14
Liang
J
,
Slingerland
JM
. 
Multiple roles of the PI3K/PKB (Akt) pathway in cell cycle progression
.
Cell Cycle
2003
;
2
:
339
45
.
15
Hanahan
D
,
Weinberg
RA
. 
The hallmarks of cancer
.
Cell
2000
;
100
:
57
70
.
16
Vranic
S
,
Bilalovic
N
,
Lee
LM
,
Kruslin
B
,
Lilleberg
SL
,
Gatalica
Z
. 
PIK3CA and PTEN mutations in adenoid cystic carcinoma of the breast metastatic to kidney
.
Hum Pathol
2007
;
38
:
1425
31
.
17
Janssen
EA
,
Soiland
H
,
Skaland
I
, et al
. 
Comparing the prognostic value of PTEN and Akt expression with the mitotic activity index in adjuvant chemotherapy-treated node-negative breast cancer patients aged <55 years
.
Cell Oncol
2007
;
29
:
25
35
.
18
Schmitz
M
,
Grignard
G
,
Margue
C
, et al
. 
Complete loss of PTEN expression as a possible early prognostic marker for prostate cancer metastasis
.
Int J Cancer
2007
;
120
:
1284
92
.
19
Shin
SI
,
Freedman
VH
,
Risser
R
,
Pollack
R
. 
Tumorigenicity of virus-transformed cells in nude mice is correlated specifically with anchorage independent growth in vitro
.
Proc Natl Acad Sci U S A
1975
;
72
:
4435
9
.
20
Khwaja
A
,
Rodriguez-Viciana
P
,
Wennstrom
S
,
Warne
PH
,
Downward
J
. 
Matrix adhesion and Ras transformation both activate a phosphoinositide 3-OH kinase and protein kinase B/Akt cellular survival pathway
.
EMBO J
1997
;
16
:
2783
93
.
21
Campbell
RA
,
Bhat-Nakshatri
P
,
Patel
NM
,
Constantinidou
D
,
Ali
S
,
Nakshatri
H
. 
Phosphatidylinositol 3-kinase/AKT-mediated activation of estrogen receptor α: a new model for anti-estrogen resistance
.
J Biol Chem
2001
;
276
:
9817
24
.
22
Garcia
JM
,
Silva
JM
,
Dominguez
G
, et al
. 
Allelic loss of the PTEN region (10q23) in breast carcinomas of poor pathophenotype
.
Breast Cancer Res Treat
1999
;
57
:
237
43
.
23
Saal
LH
,
Holm
K
,
Maurer
M
, et al
. 
PIK3CA mutations correlate with hormone receptors, node metastasis, and ERBB2, and are mutually exclusive with PTEN loss in human breast carcinoma
.
Cancer Res
2005
;
65
:
2554
9
.
24
Ciardiello
F
,
Gottardis
M
,
Basolo
F
, et al
. 
Additive effects of c-erbB-2, c-Ha-ras, and transforming growth factor-α genes on in vitro transformation of human mammary epithelial cells
.
Mol Carcinog
1992
;
6
:
43
52
.
25
Isakoff
SJ
,
Engelman
JA
,
Irie
HY
, et al
. 
Breast cancer-associated PIK3CA mutations are oncogenic in mammary epithelial cells
.
Cancer Res
2005
;
65
:
10992
1000
.
26
Gustin
JP
,
Karakas
B
,
Weiss
MB
, et al
. 
Knockin of mutant PIK3CA activates multiple oncogenic pathways
.
Proc Natl Acad Sci U S A
2009
;
106
:
2835
40
.
27
Weigelt
B
,
Peterse
JL
,
van't Veer
LJ
. 
Breast cancer metastasis: markers and models
.
Nat Rev Cancer
2005
;
5
:
591
602
.
28
Martin
SS
,
Vuori
K
. 
Regulation of Bcl-2 proteins during anoikis and amorphosis
.
Biochim Biophys Acta
2004
;
1692
:
145
57
.
29
Mehlen
P
,
Puisieux
A
. 
Metastasis: a question of life or death
.
Nat Rev Cancer
2006
;
6
:
449
58
.
30
Martin
SS
,
Ridgeway
AG
,
Pinkas
J
, et al
. 
A cytoskeleton-based functional genetic screen identifies Bcl-xL as an enhancer of metastasis, but not primary tumor growth
.
Oncogene
2004
;
23
:
4641
5
.
31
Riethdorf
S
,
Wikman
H
,
Pantel
K
. 
Review: biological relevance of disseminated tumor cells in cancer patients
.
Int J Cancer
2008
;
123
:
1991
2006
.
32
Riethmuller
G
,
Klein
CA
. 
Early cancer cell dissemination and late metastatic relapse: clinical reflections and biological approaches to the dormancy problem in patients
.
Semin Cancer Biol
2001
;
11
:
307
11
.
33
Fisher
B
,
Jeong
JH
,
Dignam
J
, et al
. 
Findings from recent National Surgical Adjuvant Breast and Bowel Project adjuvant studies in stage I breast cancer
.
J Natl Cancer Inst Monogr
2001
;
62
6
.
34
Wallgren
A
,
Bonetti
M
,
Gelber
RD
, et al
. 
Risk factors for locoregional recurrence among breast cancer patients: results from International Breast Cancer Study Group Trials I through VII
.
J Clin Oncol
2003
;
21
:
1205
13
.
35
Podsypanina
K
,
Du
YC
,
Jechlinger
M
,
Beverly
LJ
,
Hambardzumyan
D
,
Varmus
H
. 
Seeding and propagation of untransformed mouse mammary cells in the lung
.
Science
2008
;
321
:
1841
4
.
36
Weinberg
RA
. 
Leaving home early: reexamination of the canonical models of tumor progression
.
Cancer Cell
2008
;
14
:
283
4
.
37
Schmidt
M
,
Hovelmann
S
,
Beckers
TL
. 
A novel form of constitutively active farnesylated Akt1 prevents mammary epithelial cells from anoikis and suppresses chemotherapy-induced apoptosis
.
Br J Cancer
2002
;
87
:
924
32
.
38
VanderWeele
DJ
,
Zhou
R
,
Rudin
CM
. 
Akt up-regulation increases resistance to microtubule-directed chemotherapeutic agents through mammalian target of rapamycin
.
Mol Cancer Ther
2004
;
3
:
1605
13
.
39
Jin
W
,
Wu
L
,
Liang
K
,
Liu
B
,
Lu
Y
,
Fan
Z
. 
Roles of the PI-3K and MEK pathways in Ras-mediated chemoresistance in breast cancer cells
.
Br J Cancer
2003
;
89
:
185
91
.
40
Knuefermann
C
,
Lu
Y
,
Liu
B
, et al
. 
HER2/PI-3K/Akt activation leads to a multidrug resistance in human breast adenocarcinoma cells
.
Oncogene
2003
;
22
:
3205
12
.
41
Bachman
KE
,
Argani
P
,
Samuels
Y
, et al
. 
The PIK3CA gene is mutated with high frequency in human breast cancers
.
Cancer Biol Ther
2004
;
3
:
772
5
.
42
Kandel
ES
,
Skeen
J
,
Majewski
N
, et al
. 
Activation of Akt/protein kinase B overcomes a G(2)/m cell cycle checkpoint induced by DNA damage
.
Mol Cell Biol
2002
;
22
:
7831
41
.
43
Zhan
Q
,
Antinore
MJ
,
Wang
XW
, et al
. 
Association with Cdc2 and inhibition of Cdc2/Cyclin B1 kinase activity by the p53-regulated protein Gadd45
.
Oncogene
1999
;
18
:
2892
900
.
44
Shtivelman
E
,
Sussman
J
,
Stokoe
D
. 
A role for PI 3-kinase and PKB activity in the G2/M phase of the cell cycle
.
Curr Biol
2002
;
12
:
919
24
.

Supplementary data