OSU-03012 sensitizes breast cancers to lapatinib-induced cell killing: a role for Nck1 but not Nck2

The MDA-MB-231 cell line (purchased from American Type Culture Collection, ATCC) and the BT474 cell line (ATCC) were maintained in RPMI (Invitrogen). ATCC published standards are recognized by the American National Standards Institute (ANSI) and are compatible with the requirements of the International Organization for Standardization (ISO). Both cell lines were supplemented with 10% fetal bovine serum (FBS, Invitrogen) and 1% Penicillin / Streptomycin (Invitrogen). All cell lines were maintained in a 95% air / 5% CO2 incubator at 37°C. Cells were passaged once every 3-5 days (~90% confluence), and all experiments were performed during the first 12 passages.

eIF2-α expression plasmids were constructed by Ron et. al. and purchased from Addgene (plasmid numbers #21808 and 21807, [20]). GFP-tagged Nck1 and Nck2 plasmids were a generous gift from Dr. L. Larose [18, 19]. Antibodies to Nck1, phospho-eIF2-α (Serine⁵¹), total eIF2-α, ERK, phospho-ERK, PTEN, phospho-PTEN, PP1, phospho-PP1 and β-actin were purchased from Cell Signaling Technologies. Nck2 antibodies were purchased from Novus Biologicals. siRNA molecules against Nck1 and mutant siRNA molecules were custom manufactured by Dharmacon. The sequence used was previously published by Dr. W. Li and colleagues [21]. A mutant sequence containing 9 mutations was also manufactured as a control to ensure specificity of knockdown. Sequences are as follows (single stranded, sense): siNck1 5’ GGC CTT CAC TCA CTG GAA A 3’; Mutant Nck1 5’ CGC TTC CAC TGC TGA GAG A 3’. Pre-designed and validated siRNA molecules to downregulate eIF2-α and control scrambled siRNA molecules were purchased from Qiagen. siRNAs targeting ATF6 and IRE-1 were generous gifts from the laboratory of Dr. Paul Dent.

Cells were treated as indicated. 24 - 48 hrs later, cells were trypsinized, washed and stained with Annexin V-PE and propidium iodide using the ApoScreen Annexin V Apoptosis Kit (Southern Biotech) according to manufacturer’s instructions. Cells were detected using a BD FACSCanto II and analyzed using the accompanying FACSDIVA software.

Plasmid transfections were accomplished using the Effectene system (Qiagen) according to manufacturer’s instructions. Briefly, plasmid DNA (1 μg) was incubated in the presence of EC buffer and a 150:18 dilution of the Enhancer reagent for 10 minutes followed by the addition of the Effectene reagent (at a 168:20 dilution). Plasmid samples were incubated for a further 10 minutes then diluted to 1 mL with complete medium and added by single drops to the sample. Cells were allowed to accumulate the recombinant proteins for 24-48 hours. All steps excluding the incubation of DNA, EC buffer, Enhancer reagent and Effectene reagent were undertaken in 10% FBS-containing medium.

siRNA transfections were performed using the Dharmafect 1 reagent (Dharmacon) according to manufacturer’s instructions. Briefly, siRNA molecules (25 nM final concentration) were incubated in serum- and antibiotic-free medium. Concurrently, 5 μL Dharmafect 1 reagent was incubated in serum- and antibiotic-free medium. Both tubes were incubated at room temperature for 10 minutes then combined and incubated at room temperature for an additional 20 minutes. siRNA was then added to cells one drop at a time. Cells were incubated for at least 48 hours to achieve downregulation of the target mRNA.

Clonogenic assays were performed as previously described [22]. Briefly, cells were transfected and treated as indicated in the figure legends. Cells were then plated onto 6-well plates at a density of 200-400 cells / well and allowed to form colonies over the next 10-14 days. Colonies were stained using crystal violet stain, and cells that underwent ≥ 50 doublings were counted as a colony.

Cells were plated, cultured and treated as indicated. Cells were washed 2 times in PBS and lysed using CelLytic (Sigma) lysis buffer supplemented 1:100 with protease and phosphatase inhibitors (Cell Signaling) and by sonication. Protein concentration was assessed using Bio-Rad protein assay reagent. Equal amounts (10-20 μg) of protein were subsequently electrophoresed on 10-12% SDS polyacrylamide gels and electrophoretically transferred to PVDF membranes (Bio-Rad). Membranes were blocked in PBS supplemented with 0.1% TWEEN-20 and 5% dry milk and exposed to primary and secondary antibodies as indicated. Membranes were developed using SuperSignal West reagents (Pierce).

Cells were treated as described in figure legends. Cells were then harvested using NP-40 buffer (20 mM Tris-HCl, pH 8.0; 137 mM NaCl; 2 mM EDTA; 2% NP-40; protease and phosphatase inhibitor cocktail (added prior to use)). Lysate was pre-incubated with protein A/G agarose beads (1 h, 4°C with rotation). Concurrently, Protein A/G agarose beads (Pierce) were incubated with antibodies raised against either total eIF2-α or total PP1 (Cell Signaling). Beads were washed 3 times with NP-40 buffer and then added to cell lysates. Lysates + beads were incubated at 4°C for 4 – 16 h with rotation and washed 3 times in NP-40 buffer. Bound proteins were released from the antibody-coated beads using 200 mM glycine, pH 3.0. Electrophoresis and western blotting procedures were then performed as previously described.

Isobologram analyses were performed using the method of Chou and Talalay [23]. Briefly, colony formation assays were performed using stepwise increasing concentrations of OSU-03012 and lapatinib either singly or in combination (1 μM, 2 μM and 3 μM both in single and in combination treatments). Analyses were then performed using the Calcusyn program (Biosoft). Fraction affected (FA) was calculated and the combination index (CI) was then used as a measure of synergy.

All P values refer to paired student’s t-tests; differences with p≤0.05 were considered significant. Analyses were performed using the Sigmaplot software.

PP1 has been found by Larose et al [18, 19] in a complex containing both eIF2 and the protein Nck1. Nck1 (or Nckα), an SH-only adaptor protein, was originally characterized as playing a role in driving cell motility [36], a hallmark of metastatic cancer. Nck1 binds to eIF2-β, preventing the phosphorylation of eIF2-α specifically on Serine⁵¹, and dissociation of Nck1 leads to increased levels of eIF2-α phosphorylation. Thus, we examined the role of Nck1 in the enhanced phosphorylation of eIF2-α on Serine⁵¹. A robust, greater-than-additive decrease in the levels of Nck1 was observed in combination-treated samples (See Figure 5A,B) in contrast to cells treated with a single drug. Nck2 (also known as Nckβ) expression did not follow the same pattern indicating a novel differential role for these two family members in OSU-03012- and lapatinib-induced cell killing.

Next, we examined the role of Nck1 in the cell death and eIF2 Ser⁵¹ phosphorylation induced by the combination of OSU-03012 and lapatinib. The decrease in both clonogenic capacity and eIF2-α phosphorylation in MDA-MB-231 cells after OSU-03012 and lapatinib combination treatment was “rescued” by the ectopic expression of Nck1 (see Figure 5C), but not by ectopically expressing Nck2. Furthermore, Nck1, when co-expressed with wild-type eIF2-α, ablates the increase in cell death induced by OSU-03012 and lapatinib indicating a role in the same pathway for this protein (See Figure 6A and C). In contrast, ectopic co-expression of the Ser⁵¹Ala phospho-deficient mutant of eIF2-α with either Nck1 or Nck2 ablated all cell death induced OSU-03012 and lapatinib in combination (See Figure 6B and D). Co-expression of Nck2 and wild-type eIF2-α did not affect the levels of cell death indicating that this pathway is specific for Nck1.

Finally, in agreement with our hypothesis that decreased Nck1 expression is upstream to the increase in eIF2-α phosphorylation, we showed that downregulation of Nck1 was insufficient to re-sensitize BT474 cells to the ablation of OSU-03012 and lapatinib-induced cell death when the phospho-mutant of eIF2-α is ectopically expressed (Figure 7A). In addition, OSU-03012/lapatinib in combination induces a decrease in the association of eIF2-α with PP1 (Figure 7B). Taken together, these data demonstrate that a major mechanism of cell death via the combination of OSU-03012 and lapatinib is a decrease in Nck1 expression followed by upregulation of eIF2-α phosphorylation, and thus ER stress-related cell death (Figure 7C).

Larose and colleagues [18, 19] found that Nck1 forms a complex with eIF2 and PP1. Dissociation of this complex can lead to eIF2-α phosphorylation at serine⁵¹ and a decrease in protein translation. eIF2-α may also be phosphorylated at serine⁵¹ by the ER resident kinase PERK during ER stress. Since we show in Figure 2 that OSU-03012/lapatinib in combination induces ER stress in part by PERK activation, we performed studies aimed at determining the role of Nck1 in ER stress-induced cell death by OSU-03012 and lapatinib in combination. Our studies showed that ectopic expression of Nck1 abolished the cell death induced by OSU-03012/lapatinib. Furthermore, upregulation of Nck1 “rescues” the cell death induced by wild-type eIF2-α overexpression. Thus, the studies reported here demonstrate that the Nck1/eIF2 complex is a key point at which lapatinib and OSU-03012 act to synergistically kill metastatic breast cancer cells, and generally support Larose’s findings that PP1 is important in this complex.

In contrast to our findings implicating a PP1, Nck1 and eIF2-containing complex in the cytotoxicity/cytostaticity induced by OSU-03012/lapatinib, the Dent laboratory has recently published that lapatinib enhances OSU-03012-induced cell killing in glioblastoma models and that this phenomenon occurs via an ErbB/Akt/PTEN pathway [17]. MDA-MB-231 and BT474 cells as well as GBM6 and GBM12 (used in [17]) cell lines are all PTEN wild-type. Therefore, cancer-type-specific pathways may be responsible for this apparent contradiction. Our data suggest that further experiments may need to take these cancer-specific differences into account when designing therapeutic regimens.

Recently, EGFR-mediated Nck1/Rap1 activation has been shown to upregulate metastasis in a model of metastatic pancreatic carcinoma without affecting primary tumor growth [37]. These findings raise two intriguing possibilities: 1) Nck1 downregulation may be a singularly efficacious inducer of cell death specifically for metastatic breast cancer cells, and 2) eIF2 may play a role in the metastatic process. We observe a small, but insignificant decrease in the viability of BT474 cells (a non-invasive cell line, see Figure 7) after RNAi-mediated inhibition of Nck1, which may be indicative that inhibition of Nck1 alone may induce cell death in more invasive cell lines. In addition, we observe that Nck1 is downregulated only with the combination treatment in MDA-MB-231 (a more invasive cell line) cells even though eIF2-α phosphorylation is upregulated in samples treated with single drugs. eIF-4E, the mRNA cap-binding protein essential for the initiation of translation, has been found to contribute to malignancy by enabling translation of select mRNAs that encode proteins involved in growth, angiogenesis, survival and malignancy [38]. Interestingly, ER stress signaling and eIF2-α phosphorylation have been linked to drug resistance and survival in occult dormant carcinoma cells [39]. However, eIF2-α has never before been characterized specifically as a regulator of metastasis. Therefore, studies aimed at characterizing the involvement of eIF2 in metastasis, both in vivo and in vitro, are a natural continuation of these findings as are studies aimed at examining the potential of Nck1 inhibition as a therapy specific for metastatic breast cancer.