Activity of Suberoylanilide Hydroxamic Acid Against Human Breast Cancer Cells with Amplification of Her-2

Approximately 25% of breast cancers have amplification and overexpression of her-2/neu oncogene (1), which encodes for Her-2, a member of the family of epidermal growth factor receptor tyrosine kinases (2, 3). Active dimers of Her-2 with the other members of the family, e.g., Her-3 and Her-1, result in pro-growth and pro-survival signaling through the phosphoinositide-3-kinase (PI3K)/AKT and/or Ras/Raf/Erk pathways (2, 3). Preclinical studies have shown that Her-2-mediated signaling also confers resistance to anti–breast cancer agents, including tubulin polymerizing agents, e.g., paclitaxel and docetaxel, and the antiestrogen tamoxifen (4, 5). Although Her-2 overexpression has been associated with poor prognosis in breast cancer, it may also be associated with improved clinical outcome following chemotherapy, including tubulin polymerizing agents (6, 7). However, overall, Her-2 has been regarded as an excellent therapeutic target in breast cancer (7–9). Indeed, treatment with the recombinant, humanized, monoclonal anti-Her-2 antibody, trastuzumab (Herceptin), which dephosphorylates and down-regulates Her-2, has exhibited significant clinical efficacy against breast cancer and has been shown to sensitize breast cancer cells to chemotherapeutic agents, especially tubulin polymerizing agents and radiation therapy (10–15). Unfortunately, resistance to trastuzumab, administered alone or in combination with chemotherapeutic agents is common, mediated by mechanisms that often involve the activation of PI3K/AKT and/or Erk1/2 signaling or depletion of p27^KIP1 (referred to as p27; refs. 16–20). This highlights the need to develop new strategies that would target these mechanisms of resistance to trastuzumab, thereby increasing the sensitivity of breast cancer cells to trastuzumab and chemotherapeutic agents.

The reciprocal activities of histone deacetylases and histone acetyltransferases catalyze the deacetylation and acetylation, respectively, of specific lysine residues in the core nucleosomal histones and several transcription factors (21, 22). Treatment with histone deacetylase inhibitors (HDI), including the hydroxamic acid analogues (HA), e.g., trichostatin A, suberoylanilide hydroxamic acid (SAHA), and LAQ824, which inhibit class I and II histone deacetylases, has been shown to induce hyperacetylation of the histones, which transcriptionally up-regulates p21^WAF1 (referred to as p21) and increase p27 expression (21–24). This is associated with cell cycle arrest and apoptosis of cancer cells, and is shown to induce the in vivo regression of tumors (21–25). Recently, treatment with the HA-HDIs was also shown to induce acetylation of the heat shock protein 90 (hsp90), which inhibited its ATP binding and chaperone association with its client proteins, e.g., Her-2, mutant p53, AKT, and c-Raf (24, 26). This targets the client proteins for polyubiquitylation and degradation by the proteasome (27, 28). In cancer cells, these effects of HA-HDI treatment were similar to those observed following treatment with the hsp90 inhibitor geldanamycin and its analogue, 17-AAG (27, 28). 17-AAG mediated depletion of c-Raf-1 and inhibition of Erk1/2 phosphorylation, as well as attenuation of p-AKT, leads to disruption of the Raf-Mek-Erk1/2 and PI3K/AKT signaling, resulting in cytostasis and apoptosis of cancer cells (27–31). Treatment with 17-AAG has also been shown to down-regulate Her-2 and its downstream signaling, as well as sensitize breast cancer cells to chemotherapeutic agents (27–29, 32, 33). Taken together, these reports suggest that inhibition of hsp90 by an HA-HDI or 17-AAG exerts anti–breast cancer effects through depletion of Her-2 and its signaling, as well as by attenuation of AKT and c-Raf-1 activity that may be independent of down-regulation of Her-2 signaling (24, 31). SAHA is an orally bioavailable prototype of the HA-HDIs, which has been shown to be relatively safe and active as an anticancer agent (21, 34). In present studies, we determined the effects of SAHA on the levels of Her-2 and the downstream pro-growth and pro-survival signaling proteins. Treatment with SAHA induced acetylation of hsp90, which was associated with decreased binding of Her-2 to hsp90 and increased polyubiquitylation and depletion of Her-2. SAHA treatment also depleted the levels of p-AKT, c-Raf, XIAP, survivin, Bcl-2, and Bcl-x_L but induced p27 and Bim. Importantly, cotreatment with trastuzumab or docetaxel with SAHA induced synergistic cytotoxic effects against human breast cancer cells with amplification of Her-2.

Reagents and antibodies. SAHA was kindly provided by Aton Pharma, now wholly owned subsidiary of Merck & Co., Inc. (Flemington, NJ). Docetaxel was a gift from Aventis Pharmaceuticals (Bridgewater, NJ). Trastuzumab was a gift from Genentech Inc., (South San Francisco, CA). Docetaxel was made at a concentration of 10 mmol/L in DMSO and stored at −20°C. Antibodies for the immunoblot analyses to detect the levels of Her-2, c-Raf-1, AKT, p-AKT, ERK1/2, p-ERK1/2, acetylated lysine, p21, p27, acetylated histone H3 and H4 proteins, Bak, Bax, poly-ADP ribose polymerase (PARP), Bim, XIAP, survivin, Bcl-2, Bcl-x_L, and β-actin were obtained, as previously described (24, 30, 35–37).

Cell culture. The human breast cancer cell lines BT-474, SKBR-3, MCF-7 and MB-468 were obtained from American Type Culture Collection (Manassas, VA) and maintained in culture, as previously described (24).

Flow cytometric analysis of cell cycle status. The flow cytometric evaluation of the cell cycle status was done according to a previously described method (35, 36). The percentage of cells in the G₁, S phase, and G₂-M phases were calculated.

Apoptosis assessment by Annexin V staining. After drug treatments, cells were resuspended in 100 μL staining solution (containing Annexin V fluorescein and propidium iodide in a HEPES buffer, Annexin V-FITC, BD PharMingen, San Diego, CA). Following incubation at room temperature for 15 minutes, cells were analyzed by flow cytometry (35–37).

Morphology of apoptotic cells. After drug treatment, cells were washed with PBS and resuspended in the same buffer. Cytospin preparation of 50 × 10³ cells were fixed and stained with Wright-Giemsa stain (Biochemical Sciences Inc., Swedesboro, NJ). Cell morphology was determined by light microscopy. The percentage of apoptotic cells was calculated for each experiment, as described previously (35–37).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltertrazolium bromide assay. Cytotoxic effect of the drugs was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltertrazolium bromide (MTT) assay, as previously described (37). Cells (1-5 × 10³ in 100 μL per well) were grown overnight in 96-well plates. Fresh medium (100 μL) containing the designated drug was added for specified exposure intervals. At the completion of incubation, media were replaced with fresh complete media (100 μL). Three hours prior to the end of the incubation period, 20 μL of PBS containing MTT (5 mg/mL) was added to each well. Following this, the plates were centrifuged at 200 × g for 5 minutes and the medium was removed. The precipitate was then resuspended in 100 μL of DMSO. The absorbance was measured on a plate reader at 540 nm. Each experiment was done in triplicate. Analysis of synergism between SAHA and trastuzumab or docetaxel in inducing cytotoxic effects in BT-474 and SKBR-3 cells was done by median dose-effect analysis using the commercially available software (Calcusyn, Biosoft, Ferguson, MO; ref. 38).

Western blot analyses. Western analyses of Her-2, c-Raf-1, AKT, p-AKT, p21, p27, acetylated histone H3 and H4 proteins, and β-actin were done using specific antisera or monoclonal antibodies (see above), as described previously (35–37). Horizontal scanning densitometry was done on Western blots by using acquisition into Adobe PhotoShop (Adobe Systems Inc., San Jose, CA) and analysis by the NIH Image Program (U.S. NIH, Bethesda, MD). The expression of β-actin was used as a control.

Immunoprecipitation of conformationally changed Bax. Cells are lysed in CHAPS lysis buffer [150 mmol/L NaCl, 10 mmol/L HEPES (pH 7.4), 1% CHAPS] containing protease inhibitors. Immunoprecipitation is done in lysis buffer by using 500 μg of total cell lysate and 2.5 μg of anti-Bax 6A7 monoclonal antibody (Sigma Chemicals Co., St. Louis, MO). The resulting immune complexes as well as the supernatants are subjected to immunoblotting analysis with anti-Bax polyclonal rabbit antiserum, as previously described (36).

Acetylation of Hsp90. Untreated or SAHA-treated cells were lysed in total cellular proteins were quantified using the bicinchoninic acid protein assay. Hsp90 was immunoprecipitated from 200 μg of total protein using mouse hsp90 antibody, and immunoprecipitates were immunoblotted with anti–acetyl lysine or a monoclonal anti-hsp90 antibody, as previously described (24, 35).

Binding of Her-2 to hsp90 and polyubiquitylation of Her-2. Following drug treatment, cells were lysed in the lysis buffer [20 mmol/L Tris (pH 8), 150 nmol/L sodium chloride, 1% NP40, 0.1 mol/L sodium fluoride, 10% glycerol, 1 mmol/L phenymethylsulfonyl fluoride, 1 mmol/L sodium orthovanadate, 2.5 μg/ml leupeptin, 5μg/mL aprotinin], for 30 minutes on ice, and the nuclear and cellular debris cleared by centrifugation. Cell lysates (200 μg) were incubated with the Her-2 specific polyclonal antibody for 1 hour at 4 °C. To this, washed protein A agarose beads were added and incubated overnight at 4 °C. The immunoprecipitates were washed thrice in the lysis buffer and proteins were eluted with the SDS sample-loading buffer prior to the immunoblot analyses with specific antibodies against hsp90, polyubiquitin, or Her-2 (38).

Preparation of detergent-soluble and insoluble fractions. After the designated drug treatments, cells were lysed with TNSEV buffer [50 mmol/L Tris-HCl (pH 7.5), 2 mmol/L EDTA, 100 mmol/L NaCl, 1 mmol/L sodium orthovanadate, 1% NP40 containing 20 μg/mL aprotonin, 20 μg/mL leupeptin, 1 mmol/L phenylmethylsulfonyl fluoride, 25 mmol/L NAF, and 5 mmol/L N-ethylmaleimide; refs. 24, 38]. The insoluble fraction (pellet) were solubilized with SDS buffer [80 mmol/L Tris (pH 6.8), 2% SDS, 100 mmol/L DTT, and 10% glycerol]. Fifty micrograms of proteins from the NP40 soluble and insoluble fractions were separated on 7.5% SDS-polyacrylamide gel and analyzed by Western blotting (24, 38).

Statistical analyses. Data were expressed as mean ± SE. Comparisons used Student's t test or ANOVA, as appropriate. P values of < 0.05 were assigned significance.

SAHA increases p21 and p27 levels, as well as induces apoptosis of breast cancer cells. We first determined the effect of SAHA on histone acetylation, as well as on the level of p21 and p27 in breast cancer cells. Exposure to 2.0 or 5.0 μmol/L SAHA for 24 hours increased the acetylation of histone H3 and H4 (data not shown) in BT-474, SKBR-3 and MCF-7 cells (data not shown; Fig. 1A and B). This was associated with induction of p21 in a dose-dependent manner. SAHA treatment increased p27 levels in a dose-dependent manner in BT-474 but not in SKBR-3 or MCF-7 cells. Previous reports have indicated that unlike p21, increase in p27 level following treatment with SAHA is not due to increase in the mRNA transcript levels of p27, but may be due to a posttranscriptional mechanism (23). SAHA treatment also induced the accumulation of SKBR-3 and BT-474 cells in the G₂-M phase of the cell cycle in a dose-dependent manner (Fig. 1C). Treatment with SAHA also resulted in a dose-dependent increase in apoptosis of SKBR-3 and BT-474 cells, as detected by positive staining for Annexin V (Fig. 2A). This was associated with an increase in Bax conformation change, as determined by immunoblotting with the 6A7 antibody, as well as with the caspase-dependent cleavage of PARP (Fig. 2B). We next determined the effect of SAHA on the levels of pro-death and pro-survival proteins that regulate apoptosis. Figure 2C shows that treatment of BT-474 cells with SAHA increased the levels of all three isoforms of Bim, a pro-death BH3 domain-only containing protein, as well as of the multi-BH domain pro-death family member Bak but not of Bax in a dose-dependent manner. Conversely, treatment with SAHA in a dose-dependent manner depleted the levels of the pro-survival Bcl-2 and Bcl-x_L, as well as attenuated the levels of the IAP protein family members XIAP and survivin. Similar findings were also observed in SKBR-3 cells (data not shown). These findings suggest that treatment with SAHA modulates the levels of some of the key regulators of the apoptotic threshold in a manner that could sensitize breast cancer cells with Her-2 amplification to apoptotic stimuli.

SAHA induces hsp90 acetylation and depletes Her-2, p-AKT, and ERK1/2 levels. Next, we investigated the effect of SAHA on the levels of pro-growth and pro-survival Her-2, AKT, and c-Raf in the breast cancer cells. Treatment with 2.0 or 5.0 μmol/L SAHA for 24 hours depleted Her-2, p-AKT, AKT, and c-Raf levels in BT-474 and SKBR-3 cells (Fig. 3A). SAHA also mediated the depletion of c-Raf, p-ERK1/2, p-AKT, and AKT in the MCF-7 cells, which lack Her-2 amplification but express ERα and are responsive to estrogen (Fig. 3B). SAHA-mediated depletion of c-Raf in these cell types was associated with decrease in the levels of p-ERK1/2 but not ERK1/2 levels. Additionally, exposure to SAHA also induced p21 but depleted c-Raf, p-AKT, and AKT levels in MB-468 cells that lack both Her-2 amplification and Erα expression (Fig. 3B). Previous reports have shown that the HA-HDIs, e.g., LAQ824, induce acetylation of hsp90, which inhibits the ATP binding and chaperone association of hsp90 with its client proteins, including Her-2, AKT, and c-Raf, leading to their polyubiquitylation and degradation by the 20S proteasome (24, 26). Consistent with these reports, exposure intervals to 2.0 or 5.0 μmol/L of SAHA as short as 8 hours also induced acetylation of hsp90 in the SKBR-3 and MB-468 cells (Fig. 4A), which depleted the binding of Her-2 to hsp90 in SKBR-3 cells (Fig. 4B). Similar to LAQ824, SAHA also induced the acetylation of α-tubulin in SKBR-3 (Fig. 4A) and MB-468 cells (data not shown). This may be due to inhibition of histone deacetylase-6, a previously documented α-tubulin deacetylase that is also known to be associated with the microtubules (Fig. 4A; refs. 39–41). Treatment with SAHA increased polyubiquitylation and accumulation of Her-2 in the detergent-insoluble fraction of the cell lysates, directing Her-2 to proteasomal degradation (Fig. 4C and D). As shown in Fig. 4D, cotreatment with the proteasomal inhibitor ALLnL augmented SAHA-mediated accumulation of Her-2 in the detergent-insoluble fraction of the cells. Overall, these effects of SAHA caused the depletion of Her-2 levels in SKBR-3 and BT-474 cells (Fig. 3A). Although not investigated in the present studies, treatment with SAHA similar to the other HA-HDIs may also deplete the mRNA transcript levels of Her-2, which may contribute to the attenuation of Her-2 in the breast cancer cells (24, 42).

Cotreatment with SAHA and trastuzumab and docetaxel exerts synergistic cytotoxic effects in breast cancer cells. Activated Her-2 dimers trigger pro-growth and antiapoptotic signaling through the activation of PI3K/AKT, which mediates chemoresistance of breast cancer cells (2, 3, 16). Conversely, depletion of Her-2 and AKT activity induces chemosensitization in breast cancer cells (13, 43, 44). Consistent with this, as shown in Fig. 5A, cotreatment with 2.0 or 5.0 μmol/L of SAHA, which attenuates Her-2 and AKT signaling, and trastuzumab induced significantly more apoptosis of BT-474 cells than either agent alone (P < 0.01). Similar findings were also observed in SKBR-3 cells (Fig. 5B). The combinations of SAHA and trastuzumab at various dose levels induced synergistic cytotoxic effects against BT-474 cells, as determined by median dose-effect isobologram analysis (Fig. 5C). Additionally, combined treatment with 1.0 or 5.0 μmol/L of SAHA and 10 or 20 μg/mL trastuzumab caused more depletion of Her-2 and p-AKT levels, but induced p27 and Bim_EL than treatment with either agent alone (Fig. 5D). Combination of SAHA and trastuzumab also induced more processing of PARP consistent with more apoptosis of BT-474 and SKBR-3 cells (data not shown). Figure 6A and B show that, similar to trastuzumab, cotreatment of SAHA and docetaxel induced significantly more apoptosis than treatment with either agent alone (P < 0.05). The combination also induced synergistic cytotoxic effects against SKBR-3 cells, as determined by median dose-effect isobologram analysis (Fig. 6C). A similar effect was observed against BT-474 cells (data not shown). Exposure of SKBR-3 cells to docetaxel increased p-AKT levels, which were abrogated when treatment with docetaxel was combined with SAHA (1.0 or 5.0 μmol/L; Fig. 6D). Similar abrogation of p-AKT levels was observed in BT-474 cells, following combined treatment with SAHA and docetaxel (concentration ranging between 5.0 and 50 nmol/L; data not shown). The combination of SAHA and docetaxel also induced more processing of PARP than either agent alone, consistent with more apoptosis of SKBR-3 cells (Fig. 6D).

Preclinical in vitro and in vivo studies have shown that treatment with SAHA induces p21 and cell cycle growth arrest, associated with differentiation and apoptosis of breast cancer cells (23, 45, 46). A phase I study of SAHA has indicated that it is a relatively safe agent with clinical antitumor activity (34). This creates the rationale to further evaluate the activity of SAHA as a single agent and in combination with anti–breast cancer agents. Present studies focused on the in vitro activity of SAHA and/or trastuzumab or docetaxel against human breast cancer cells with amplification of Her-2. The results highlight several mechanisms that may contribute to SAHA-mediated cell cycle effects and apoptosis of breast cancer cells. First, SAHA treatment is shown to induce hsp90 acetylation, which inhibits the chaperone association of Her-2 with hsp90, directing Her-2 to polyubiquitylation and degradation by the 20S proteasome. These findings are consistent with our previous reports demonstrating that treatment with other HA-HDIs induce acetylation of hsp90, which promotes polyubiquitylation and accumulation of the misfolded client proteins, including Her-2, AKT, c-Raf, Bcr-Abl and mutant FLT-3, in the detergent-insoluble fraction of the cells leading to proteasomal degradation of the client proteins (24, 35, 47). The chaperone-binding E3 ubiquitin ligase for Her-2 is carboxyl-terminal hsc70-interacting protein (CHIP), which ubiquitylates Her-2 prior to its degradation by the 20S proteasome (48, 49). Similar to hsp90 inhibition by 17-AAG, SAHA-induced acetylation of hsp90 may also promote CHIP-mediated polyubiquitylation and degradation of Her-2, although this was not ascertained in the present studies. Treatment with SAHA also induced α-tubulin acetylation, suggesting that the cytosolic histone deacetylase-6 may be involved in the deacetylation of hsp90, which is inhibited by SAHA (40–42). Which lysine residue(s) on hsp90 is acetylated following SAHA treatment, leading to the inhibition of ATP binding and chaperone association of hsp90 with the client proteins, remains to be determined. Additionally, whether SAHA, following treatment with 17-AAG, also promotes CHIP-mediated polyubiquitylation of Her-2, has yet to be determined.

SAHA-mediated inhibition of the chaperone function of hsp90 also led to proteasomal degradation and depletion AKT and c-Raf, two of the most prominent pro-growth and pro-survival proteins often dysregulated in cancer cells (24, 27, 30). Whereas reduced levels of p-ERK1/2 may be due to decreased c-Raf levels, the depletion of p-AKT has been shown to be due to restoration of the activity of a previously described AKT phosphatase PP1, following depletion of Her-2 due to SAHA (30, 50). Present findings also show that treatment with SAHA attenuated Bcl-2, Bcl-xL, XIAP, and survivin, whereas simultaneously increasing Bak and Bim expression. Collectively, these modulations are likely to lower the threshold for apoptosis in breast cancer cells. Although the precise mechanism(s) responsible for all of these modulations have not been elucidated, depletion of AKT and/or its activity is known to inhibit p27 phosphorylation and increase its nuclear localization (51, 52). Reduced AKT activity also attenuates the phosphorylation of Foxo3A and increases its nuclear localization, which results in transactivation and induction of p27 and the prodeath proteins Bim and TRAIL (53, 54). Taken together, these findings suggest that treatment with SAHA, and other HA-HDIs, inhibits Her-2 and downstream AKT and Ras/Raf/ERK signaling pathways, as well as modulating apoptosis regulators in a manner which could lower the apoptotic threshold to anti–breast cancer agents (24, 43, 55, 56). Consistent with this, SAHA and docetaxel combination induced synergistic cytotoxic effects against breast cancer cells with amplification of Her-2. Combination with other HA-HDIs has also been shown to enhance the activity of anticancer agents with diverse mechanisms of action (24, 57). Molecular perturbations induced by SAHA could also explain why cotreatment with SAHA and trastuzumab exerted synergistic cytotoxic effects in Her-2-amplified breast cancer cells. A similar outcome was observed when the multiple cyclin-dependent kinase inhibitor, flavopiridol, which also induces molecular perturbations that lower the apoptosis threshold, was combined with trastuzumab against breast cancer cells (58). Although not directly tested here, SAHA-mediated depletion of p-AKT and p-ERK1/2, as well as down-modulation of the apoptotic threshold, may also sensitize breast cancer cells to other targeted agents against Her-2 heterodimers, e.g., pertuzumab (24, 29, 59, 60).

Several reports have highlighted a number of potential mechanisms of resistance to trastuzumab in breast cancer cells. In a majority of patients with breast cancers with amplification of Her-2, this manifests both as primary and acquired resistance to trastuzumab, most likely due to mechanisms that involve loss of PTEN or activation of insulin-like growth factor receptor 1 signaling (17, 19). This ultimately results in deregulated pro-growth and pro-survival activity of PI3K/AKT, mediated through depletion of p27 levels and Forkhead transcription factor activity, as well as through increased mTOR/4E-BP1 phosphorylation leading to increased Cap-dependent mRNA translation (16, 20, 61, 62). Thus, strategies, such as cotreatment with SAHA, which attenuates PI3K/AKT signaling, would help override the primary and acquired mechanisms of resistance to trastuzumab (63, 64). At least some of the in vivo anti–breast cancer activity of trastuzumab may depend on its IgG₁ Fc region engaging the Fc-γ receptors on immune effector cells (65). Treatment with HA-HDIs, including SAHA, has been shown to enhance the expression of TRAIL and its death receptors, DR4 and DR5, which may augment TRAIL-dependent antitumor activity of immune effector cells (66–68). This is a particularly attractive hypothesis, because cotreatment with trastuzumab has been shown to enhance the activity of TRAIL (69). Collectively, these reports, taken together with our findings, suggest that cotreatment with SAHA may also help override immune function–based mechanism of resistance to trastuzumab.

In summary, as shown here, SAHA mediated attenuation of Her-2, pERK1/2, and p-AKT, as well as SAHA-induced potentiation of trastuzumab and docetaxel activity against Her-2-amplified breast cancer cells represent novel findings that can be therapeutically exploited. Combination of docetaxel and trastuzumab is active and commonly employed against breast cancer with amplification of Her-2 (70). Therefore, our findings also support the in vivo testing of SAHA with this combination. To this end, a phase I/II study of SAHA and trastuzumab has been designed and will soon be implemented by a member institution (Eastern Cooperative Oncology Group).