Chemoresistance of Lung and Breast Cancer Cells Growing Under Prolonged Periods of Serum Starvation

Juan Sebastian Yakisich1, Rajkumar Venkatadri1, Neelam Azad1, and Anand Krishnan V. Iyer1

1 Abstract

The efficacy of chemotherapy is hindered by both tumor heterogeneity and acquired or intrinsic multi- drug resistance caused by the contribution of multidrug resistance proteins and stemness-associated prosurvival markers. Therefore, targeting multi-drug resistant cells would be much more effective against cancer. In this study, we characterized the chemoresistance properties of adherent (anchorage- dependent) lung H460 and breast MCF-7 cancer cells growing under prolonged periods of serum starvation (PPSS). We found that under PPSS, both cell lines were highly resistant to Paclitaxel, Colchicine, Hydroxyurea, Obatoclax, Wortmannin and LY294002. Levels of several proteins associated with increased stemness such as Sox2, MDR1, ABCG2 and Bcl-2 were found to be elevated in H460 cells but not in MCF-7 cells. While pharmacological inhibition of either MDR1, ABCG2, Bcl-2 with Verapamil, Sorafenib or Obatoclax respectively decreased the levels of their target proteins under routine culture conditions as expected, such inhibition did not reverse PX resistance in PPSS conditions. Paradoxically, treatment with inhibitors in serum-starved conditions produced an elevation of their respective target proteins. In addition, we found that Digitoxin, an FDA approved drug that decrease the viability of cancer cells growing under PPSS, downregulates the expression of Sox2, MDR1, phospho- AKT, Wnt5a/b and β-catenin. Our data suggests that PPSS-induced chemoresistance is the result of extensive rewiring of intracellular signaling networks and that multi-resistance can be effectively overcome by simultaneously targeting multiple targets of the rewired network. Furthermore, our PPSS model provides a simple and useful tool to screen drugs for their ability to target multiple pathways of cancer resistance. This article is protected by copyright. All rights reserved

Keywords: Lung Cancer; Breast cancer; Paclitaxel, Verapamil, Sorafenib, Bcl-2, Class III tubulin, multidrug-resistance, stemness

2 Introduction

Lung and breast cancers are leading causes of cancer-related death among both men and women. Statistics show that 228,190 new cases were reported, and 159,480 lung cancer deaths occurred in the U.S. in 2013 (Coughlin et al., 2014). Because of the majority of patients being diagnosed at an advanced stage when curative treatment options are limited (Detterbeck et al., 2013), the overall five- year survival rate is only 16 % (Nana-Sinkam, 2013), and the prognosis has remained unchanged for the past three decades. Regarding breast cancer, the number of new cases of invasive breast cancer and breast cancer deaths has been estimated at 232,340 and 39,620 respectively (DeSantis et al., 2014 ) for 2013 in the U.S. The five-year relative survival rate for women diagnosed with localized breast cancer is 98.6% but for those diagnosed with regional and distant-stage breast cancer, the survival rate declines to 84.4% and 24.3%, respectively (DeSantis et al., 2014).
Part of the failure of chemotherapy is due to high intratumoral heterogeneity (Carey et al., 1990; Neelakantan et al., 2015; Rustum et al., 2010) that is responsible for the presence of cancer cells with differential sensitivity to anticancer drugs within a tumor. Like with other cancers, lung cancer and breast cancer are composed of a subpopulation of cells with stem-like properties – cancer stem cells (CSCs) or cancer stem-like cells (CS-LCs) – that are associated with resistance to chemotherapy and tumor relapse (Dawood et al., 2014 ; Lin et al., 2016; Lopez-Ayllon et al., 2014). CSCs/CS-LCs are composed of heterogeneous subpopulations of cells, and it has been demonstrated that non-CSCs and CSCs can interconvert into each other in lung cancer (Akunuru et al., 2012), breast cancer (Gupta et al., 2011 ) as well as in other cancers such as, colon, glioblastoma and melanoma (Cruz et al., 2012; Luo et al., 2014). In theory, following chemotherapeutic intervention, any surviving cell may still be able to generate both non-SCs and CSCs and repopulate the tumor. It is therefore critical to identify and develop drugs or combinations of drugs that can simultaneously target various cellular subtypes so that they may be eliminated concomitantly, thus leading to effective and sustained tumor regression.
We recently found that lung cells growing under Prolonged (7-12 days) Periods of SerumStarvation (PPSS) develop a high level of resistance to conventional anticancer drugs such as Paclitaxel(PX), Colchicine (CX) and Hydroxyurea (HU), but are sensitive to therapeutic concentrations of Digitoxin (DIG), a cardiac glycoside approved by the FDA for the treatment of cardiac diseases(Yakisich, 2016a). We also found a similar pattern of resistance to PX, CX and HU and sensitivity to DIG in floating lung tumorspheres (LTs) generated in serum free media (SFM) without the addition of any mitogenic stimulation (Yakisich et al., 2016b). LTs expressed higher levels of the stemness- associated marker Sox2, suggesting that adherent cells growing under PPSS may also have a similar increase in stemness properties that confers resistance.
Furthermore, while Paclitaxel (PX) is active against a broad range of cancers (Ma and Mumper, 2013), intrinsic or acquired resistance limits its use in lung and breast cancer. Resistance to PX has been attributed to the presence of subpopulations of CSCs/CS-LCs (Kubo et al., 2013), but other factors such as increased expression of Class III β-tubulin (Kubo et al., 2013; Ohashi et al., 2015; Yang et al., 2014) and decreased apoptosis (Paul and Jones, 2014) may play an important role as well. Identifying the factors contributing to PX resistance in cancer cells is important not only for developing effective PX- based therapeutic regimes but also for screening new drugs. Given that we observe similar resistance to PX in both PPSS cells and LTs, it would be interesting to evaluate the pathways that drive resistance so that rate-limiting factors may be identified and targeted.
The aim of this study was to characterize the chemoresistance properties of cancer cells growing under PPSS, to investigate the factor(s) contributing to PX resistance in PPSS cells and explore the signaling pathways in PPSS cells from both lung and breast cancer that are affected by treatment with DIG.

3 Materials & Methods

Drugs

All control treatments were supplemented with equivalent concentrations (0.25%) of DMSO used in drug treatment. Paclitaxel (PX) was purchased from Sigma-Aldrich (St. Louis, MT). Verapamil hydrochloride (VP), Sorafenib (SF), LY294002 (LY) and Wortmannin (WT) were purchased from VWR (Radnor, PA). Obatoclax (GX15-070, OBT) was obtained from Thermo Fisher Scientific (Waltham, MA). PX, SF, LY, WT and OBT were prepared as stock solution in DMSO (1 mM, 10 mM, and 10 mM, 10 mM, 10 mM respectively) and stored in aliquots at -20C. VP (50 mM) was freshly prepared in distilled sterile water prior to use. Final dilutions were freshly prepared in culture media before use.

Cell culture

The human lung epithelial cancer cell line NCI-H460 and breast cancer cell line MCF-7 were obtained from American Type Culture Collection (Manassas, VA). The H460 cell line is considered highly resistant to chemotherapy (Coughlin et al., 2014). NCI-H460 cells were cultured in complete media (CM = RPMI-1640 supplemented with 5% FBS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin) (Medan et al., 2012). MCF-7 cells were cultured in complete media (CM = DMEM/high glucose supplemented with 5% FBS, 4 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin). For serum starvation, cells were cultured in serum-free media (SFM: same as CM but without FBS). All cells were cultured in a 5% CO2 environment at 37C.

Viability assay (MTT assay)

For routine culture conditions (RCCs), cells (~2,000 cells/well) were plated in 96-well cell-culture microplates (Costar, USA) and incubated overnight in CM to allow them to adhere. For cells growing under PPSS, cells (500 cells/well) were plated in 96-well cell-culture microplates and incubated overnight in CM to allow them to adhere and then maintained in SFM for 7-12 days. For drug testing, cells were exposed to the appropriate concentration of drug or vehicle for 72 hs in the corresponding media (CM or SFM) and cell viability was evaluated by the MTT (Sigma-Aldrich, St. Louis, MT) assay. The absorbance of solubilized formazan was read at 570 nm using the Gen 5 2.0 All-In-One microplate reader (Bio-TEK, Instruments Inc.). In all cases, the highest concentration of DMSO was used in the control and this concentration was maintained at or below 0.0025% (v/v). This DMSO concentration did not show any significant antiproliferative effect on the cell line in a short-term assay.

Western blotting

Preparation of cell lysates and Western blotting were performed as described previously (Azad et al., 2010). Antibodies for Sox2, Oct4, Class III β-Tubulin, Nanog, MDR-1, ABCG2, Bcl-2, AKT, phospho- AKT (pAkt), Wnt5a/b, β-catenin and peroxidase-conjugated secondary antibody were purchased from Cell Signaling (Danvers, MA). Antibody for GAPDH was purchased from Santa Cruz Biotechnology (Dallas, TX). The blotting membranes were probed with diluted primary antibody (1:1000) and followed by peroxidase-conjugated secondary antibody (1:4000). Immune complexes were detected by chemiluminescence using SuperSignal™ West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific, Grand Island, NY) and photographed using myECL imager instrument (Thermo Fisher Scientific, Grand Island, NY).

Flow Cytometry

For flow cytometry analysis cells were collected by trypsinization, washed twice with PBS and fixed at least overnight with 70% ethanol at 4°C. Cells were then washed twice with PBS, treated with DNAse- free RNAse (100 µg/ml) and stained with propidium iodide (50 µg/ml). The cells were then analyzed with Acea Novocyte 2060 (Acea Biosciences, San Diego, CA) and NovoExpress (version 1.0.2) cell cycle analysis software. The cell cycle distribution is shown as the percentage of cells containing G0/G1, S and G2/M DNA as identified by propidium iodide staining.

Statistical analysis

All Pairwise Multiple Comparison Procedures (ANOVA, Student-Newman-Keuls Method) were performed with Sigmaplot (V. 11.0) software.

4 Results & Discussion

Cells growing under PPSS display multidrug resistance

We recently found that lung cancer cells growing under PPSS become highly resistant to PX, HU, and colchicine (Yakisich, 2016a). In the current study, by including the widely used MCF-7 breast cancer cell line, we demonstrated that such resistance may not be cell line specific but rather a more general property of all cancer cells. We compared in both cell lines the effect of several anticancer drugs when growing under RCCs versus under PPSS. H460 and MCF-7 cells growing under RCCs or under PPSS were treated with PX (20 nM), HU (1 mM), CX (1 µM), OBT (10 µM), LY (25 µM) or WT (25 µM) for 72 hs and viability was measured using the MTT assay. These concentrations were chosen since they have been shown in pilot studies to decrease the viability of cells by > 50 % in cells growing under RCCs. Figure 1 shows that compared to cells growing under RCCs, cells growing under PPSSs were highly resistant to each of the drugs tested, suggesting that PPSS induces a multi-drug resistant phenotype in cancer cells.

Cells growing under PPSS divide slowly and express different levels of stemness-associated markers in a cell type-dependent manner

While serum starvation arrests non cancer cells at the G0/G1 transition of the cell cycle, this procedure usually fails to synchronize cancer cell lines (Darzynkiewicz et al., 2011). Microscopic observation of cells growing under PPSS revealed that both H460 and MCF-7 cells were able to continue dividing in SFM, and the number of cells increased over time (data not shown). We also performed flow cytometry in H460 cells to demonstrate that cells were not fully arrested. The data revealed a higher number of H460 cells in the G1 and G2 phase of the cell cycle at expense of the S phase (Figure 2 A-B). While tubulin levels vary during the cell cycle (Mujagic et al., 1983), the expression of the class III isoform that has been associated with PX resistance in lung cancers (Ohashi et al., 2015; Yang et al., 2014) as well as with Docetaxel in breast cancer (Li et al., 2014b; Shalli et al., 2005) has not been characterized in cells under PPSS. In order to determine whether changes in the expression of class III β-tubulin may be responsible for PX resistance, we prepared protein lysates from cells growing under both RCCs and PPSS for three and eight days, and analyzed for expression of class III β-tubulin by Western Blot. While PPSS did not significantly alter the expression of class III β-tubulin in H460 cells, a significant decrease in class III β-tubulin was observed in MCF-7 cells, indicating that the expression of this protein is not related to PX resistance in cells growing under PPSS (Figure 2C).
Since we were able to sustain cancer stem cells (CSCs) or cancer stem-like cells (CS-LCs) in SFM conditions (Yakisich et al., 2016b), one possibility is that adherent cells that are serum-starved for prolonged periods may also develop increased stemness properties, a process that is in turn associated with increased chemoresistance. Prolonged culturing of lung (Brower et al., 1986) and breast cancer cells (Calvo et al., 1984) have been performed in serum free media supplemented with specific additives. Since these studies were performed prior to the description of CSCs/CS-LCs in the literature, these studies lacked information on the stemness properties associated with serum starvation including the ability of starved cells to form spheres. Short periods (48 hs) of serum starvation was also shown to increase the side population, a fraction enriched in CSCs/CS-LCs, in three different cancer cell lines (Tavaluc et al., 2007) using flow cytometry. Consequently, we evaluated the expression of sex- determining region Y (SRY)-Box2 (Sox2), Oct4 and Nanog by Western Blotting. Figure 2D shows that PPSS affected the expression of stemness associated markers in a cell type dependent manner. While Nanog levels were slightly increased in H460 cells growing under PPSS, these cells expressed markedly higher levels of Sox2 and decreased levels of Oct4. In MCF-7 cells, PPSS significantly decreased the levels of all three stemness-associated markers. Taken together, our results suggest that increased levels of stemness associated markers per se in PPSS is not a major player in the resistance to PX.
Sox2 is a transcription factor that plays a critical role in embryonic development and in maintaining pluripotency and self-renewal in embryonic stem cells (Santini et al., 2014). In cancer cells, Sox2 antagonizes the Hippo pathway and promote stemness (Basu-Roy et al., 2015 ). Sox2 is a frequently amplified gene in small-cell lung cancer (Rudin et al., 2012) and we previously reported elevated levels of Sox2 in H460 lung tumorspheres (Yakisich et al., 2016b). In our study, the expression of the transcription factor Oct4, also known as POU5F1 (POU domain, class 5), was lower in PPSS cells compared to cells growing under RCCs (Figure 2D). This result could be cell type-dependent and is in agreement with the variation in Oct4 expression found between lung cancer cell lines. For instance, while Oct4 expression increase in SPC-A1 sphere cells compared to SPC-A1 parental cells, an opposite effect is observed in PC9 lung cancer cells. In both cells lines, the expression of Sox2 was higher in spheres compared to their corresponding parental cell line (Zhang et al., 2015b), which is in agreement with our findings.

Prolonged periods of serum starvation induce the expression of MDR1 and ABCG2 in H460 cells, but Verapamil or Sorafenib are unable to reverse PX resistance

Lung CSCs/CS-LCs were found to overexpress multidrug resistance-associated proteins such as MDR1 and ABCG2 (Yang et al., 2015), and its overexpression has been correlated with poor response to PX (Maráz et al., 2011). Overexpression of MDR1 was also reported in PX-resistant MCF-7 cells (Shi et al., 2014). To investigate the contribution of MDR1 and ABCG2 to the PX resistance in H460 cells growing under PPSS, we evaluated the expression levels of these proteins. As shown in Figure 3A, H460 cells growing in SFM for three-eight days showed increased levels of both MDR-1 and ABCG2 compared to cells growing under routine culture conditions. Consistent with results by Han et. al. (Han et al., 2008) in MCF-7 cells, the levels of MDR1 were almost undetectable and its levels were not induced by PPSS. Regarding ABCG2, cells growing under PPSS showed decreased levels suggesting that neither MDR1 nor ABCG2 are responsible for PX resistance in cells growing under PPSS.
To confirm this hypothesis, we evaluated the ability of VP or SF to reverse the resistance to PX. Cells were incubated with VP (50 µM and 100 µM) or SF (2.5 µM and 5 µM) alone or in the presence of 20 nM PX. H460 serum starved cells were resistant to PX (Figure 3B – Control bars), and VP or SF as single agents exerted modest inhibitory effects on the viability of serum-starved H460 cells (Figure 3B, black bars). VP alone at 50 µM concentration exhibited a modest but significant inhibitory effect on cell viability. Serum-starved H460 cells pretreated for 3 h with either VP or SF and then incubated for 72 hs in the presence of 20 nM PX did not show decreased viability compared to untreated cells (Fig. 3B, white bars). We also evaluated the effect of 50 µM VP in combination with 2.5 µM SF, which showed no significant effect either (Figure 3B). Similar experiments were performed in MCF-7 cells, and barring a few differences (e.g. MCF-7 cells were more sensitive to VP and SF at 2.5 µM promoted cell viability) neither VP nor SF alone or in combination reversed PX resistance. In H460 cells, while SF at 2.5 µM and 5 µM modestly potentiated the effect of 20 nM PX, this effect was not observed in MCF-7 cells (Figure 3B), indicating that ABCG2 activity does not have a major role for the multidrug- resistance phenotype in either H460 or MCF-7 cells growing under PPSS.
VP at low micromolar concentration has been shown to sensitize cancer cells to anticancer agents. For instance, i) concentrations as low as 5 µM were able to sufficiently sensitize A549 and H460 lung cancer cells to PX and 17-(Dimethylaminoethylamino)-17-demethoxygeldanamycin (17-DMAG) even in cells with acquired resistance (Kim et al., 2015), and ii) several MDR1-overexpressing human cancer cell lines were sensitized to gemcitabine when co-treated with 25 µM VP (Bergman et al.). On the other hand, SF at a level as low as 2.5 µM was sufficient to potently inhibit ABCG2 (Gongora, 2014; Wei et al., 2012). In our experiments, we used much higher concentrations of VP or SF (up to 100 µM and 5 µM, respectively) as co-treatment drugs with PX, but still could not affect chemoresistance in PPSS cells. Therefore, we conclude that although H460 cells growing under PPSS overexpress MDR1 and ABCG2, pharmacological inhibition of these proteins was not sufficient to reverse PX resistance. This conclusion is further supported by the fact that even for MCF-7 cells growing under PPSS, MDR1 levels was undetectable and ABCG2 levels were lower than cells growing under RCCs.

Prolonged Periods of Serum Starvation induces the expression of the prosurvival protein Bcl-2, but Obatoclax does not reverse PX resistance

Since pharmacological inhibition of the multidrug resistance proteins MDR1 and ABCG2 by VP and SF respectively showed no significant effect on overcoming PX resistance, we investigated the role of the prosurvival protein Bcl-2. It is well established that Bcl-2 overexpression or stabilization confers resistance to apoptosis (Azad et al., 2006; Yang et al., 1997 ) and increases the chemoresistance of lung cancer cells (Nishioka et al., 2014 ). Overexpression of Bcl-2 has also been associated with poor response to PX (Maráz et al., 2011). In vitro, inhibition of Bcl-2 potentiated PX-induced apoptotic cell death (Chatterjee et al., 2015). As shown in Fig. 4A, H460 serum-starved cells showed increases levels of Bcl-2 as compared to cells growing under RCCs. XIAP, another pro-survival caspase inhibitor (Castells et al., 2013) that is a downstream effector of the intrinsic pathway (Lin et al., 2008) remained unchanged. In MCF-7 cells, PPSS did not induced a significant change in the protein levels of Bcl-2 but the protein levels of XIAP were significantly reduced. High concentrations of the Bcl-2 inhibitor OBT (10 µM) reduced cell viability in H460 cells growing under RCCs by more than 95% and led to a 40- 70% decrease in cell viability in cell growing under PPSS; interestingly, OBT (10 µM) had no effect on PX sensitivity (Figure 4B). MCF-7 cells growing under PPSS were also found to be more resistant to 10 µM OBT (Figure 1).
In summary, although H460 cells growing under PPSS overexpress Bcl-2 and become highly resistant to OBT, the pharmacological inhibition of this pro-survival protein has no impact on PX resistance. This result is supported by the above mentioned observation that Bcl-2 levels is MCF-7 does not significantly change under PPSS conditions. This result supports the finding that OBT in combination with docetaxel (another tubulin-targeting drug) had almost no benefit in Phase 1/2 clinical trials for patients with non small-cell lung cancer (Chiappori et al., 2014).

Cells growing under PPSS respond aberrantly to a variety of pharmacological inhibitors, suggesting global network changes in pathways involving key cancer-associated proteins

Given the strong multidrug-resistance phenotype exhibited by cell grown under PPSS (Figure 1), we investigated the effect of various pharmacological inhibitors on their respective target proteins using Western blotting to assess for global changes in protein expression in response to each inhibitor.
Cells growing under RCCs and under PPSS were treated with VP (100 µM), SF (5 µM), OBT (10 µM) or a combination of VP (50 µM) and SF (2.5 µM) for 24 hs, and were assessed for the expression of their respective target proteins. In both H460 and MCF-7 cells growing under RCCs, the tested drugs exerted the expected effects – for instance, SF (5 µM) significantly inhibited the expression of ABCG2 and OBT (10 µM) significantly inhibited the expression of Bcl-2. In both cell lines, the basal expression of MDR1 was extremely low and the effect of VP could not be observed (Figure 5A). In contrast, in cells growing under PPSS, these drugs exerted a paradoxical effect – for instance, VP (100 µM), SF (5 µM) and OBT (10 µM) when used alone significantly increased the expression of MDR1, ABCG2 and Bcl-2 respectively. The combination of VP (50 µM) and SF (2.5 µM) also significantly increased MDR1, ABCG2 and Bcl-2 expression in H460 cells (Figure 5B).
We also observed that SF and OBT were, as expected, relatively highly specific in cells growing under RCCs than in PPSS. In cells growing under RCCs, SF (5 µM) decreased the effect of ABCG2 but had no significant effect on Bcl-2. In a similar manner, OBT (10 µM) significantly decreased the expression of Bcl-2 but did not affect the expression of ABCG in cells growing under RCC (Figure 5A). In contrast, in cells growing under PPSS, VP (100 µM) increased the expression of not only MDR1, but also ABCG2 and Bcl2. In a similar manner, SF and OBT not only increased the expression of Bcl-2 but also increased the expression of MDR1 and ABCG2 in H460 cells (Figure 5B). Overall this data suggests that PPSS may induce a global „rewiring‟ of important intracellular pathways, resulting in negative feedback mechanisms between signaling pathways that regulates the expression of MDR1, ABCG2 and Bcl-2, which may be responsible for their aberrant and sometimes paradoxical response to inhibitors associated with these pathways.

Cells growing under PPSS are resistant to the PI3K inhibitors LY294002 and Wortmannin

Overexpression of Sox2 has been associated with chemoresistance to PX in ovarian (Li et al., 2015), prostate (Li et al., 2014a) and breast (Piva et al., 2014) cancer cells. In ovarian and prostate cancers, the PI3K/Akt pathway has been involved in Sox2 mediated chemoresistance (Li et al., 2014a; Li et al., 2015). Since we found elevated levels of Sox2 in H460 cells growing under PPSS and lower but still significant levels in MCF-7 (Figure 2D), we evaluated the effect of the PI3K inhibitors LY294002 (LY) and Wortmannin (WT) on the viability of cells growing under PPSS as well as their effect on PX resistance. In H460 cells growing under RCCs, both LY and WT (each at 25 µM) were able to decrease viability by ~ 60% (Figure 1). On the other hand, cells growing under PPSS were highly resistant to both LY or WT, and both inhibitors had no effect on PX resistance (Figure 6A). The PI3K/Akt pathway has been shown to antagonize P-glycoprotein-mediated multidrug resistance in leukemic cells (Barancík et al., 2006 ) and further supports our finding that inhibiting MDR-1 with VP in cell growing under PPSS is not sufficient to reverse PX resistance (Figure 3). We also performed Western blotting to evaluate the effect of PI3K inhibitors on their target protein in cells grown under RCCs as well as in cells growing under PPSS. H460 and MCF-7 cells were treated for 24 hs with LY (25 µM) or WT 25 (µM). Both PI3K inhibitors were able to decrease expression of pAKT in H460 and MCF-7 cells regardless of culture conditions (Figure 6B).

Digitoxin decreases the expression of several key markers associated with drug resistance in cancer

We previously demonstrated that therapeutic levels of Digitoxin (20-50 nM) were able to decrease the viability of cells growing under RCCs, PPSS as well as floating lung tumorspheres (LTs) (Yakisich et al., 2016b; Yakisich, 2016a). In LTs, Digitoxin (DIG) reduced the expression of Sox2 (Yakisich et al., 2016b). H460 and MCF-7 cells growing under PPSS or RCCs were treated with DIG (50 nM) for 24 hs and the expression of key proteins associated with increased stemness (Sox2), chemoresistance (MDR-1) and the PI3K/Akt and Wnt signaling pathways (pAKT and AKT; Wnt5a/b and β-catenin) were evaluated. Untreated H460 cells growing under PPSS showed increased levels of Sox2 and MDR1 compared to cells growing under RCCs, and both untreated cells lines showed decreased basal levels of pAKT, and β-catenin. In H460 cells, DIG reduced the levels of Sox2 and MDR1, and in both cell lines DIG reduced the levels of pAKT, Wnt5a/b and β-catenin under PPSS. Remarkably, DIG has a more pronounced effects on the expression of these proteins when cells were cultured under PPSS compared to cells growing under RCCs (Figure 7). The latter seems paradoxical since DIG inhibited the viability of H460 cells more potently in cells growing under RCCs compared to cells growing under PPSS (IC50 = 19.4 ± 1.7 and 88.66 ± 8.12 nM, respectively at 48 h) (Yakisich, 2016a). This apparent paradox may be the result of altered basal activity of signaling networks present in cells growing under RCCs and PPSS that may mediate the effect of DIG on protein pathways in a concentration- and time-dependent manner. Regardless, the data clearly shows that DIG downregulates the expression of key protein (Sox2, MDR1 in H460 cells and, pAKT, Wnt5a/b and β-catenin in both H460 and MCF-7 cell lines) under PPSS conditions. In this context, the Wnt signaling pathway has been recently identified as an important regulator of the stemness of lung cancer cells (Zhang et al., 2015a; Zhang et al., 2015b). CBP-dependent Wnt/β-catenin signaling has been identified as a crucial regulator of MDR1 transcription. Thus, DIG as well as other Wnt inhibitors that exerts potent anticancer effects represents a viable therapeutic strategy by simultaneously targeting important proteins associated with increased stemness and/or multidrug resistance.

Proposed model for chemoresistance in cells growing under PPSS

Our data supports a model in which cells growing under PPSS become multidrug resistant due to rewiring of signaling pathways involved in the regulation of important proteins for chemoresistance (MDR1 and ABCG2) and survival (Bcl-2). Rewiring of signaling networks has been demonstrated during oxidative stress (Garcia et al., 2012 ) as well as an adaptation mechanism to low nutrient and oxygen conditions (Ratnikov et al., 2016). We provide evidence that PPSS may induce or strengthen crosstalk between MDR1, ABCG2 and Bcl-2 that are negligible or non-existent in cells growing under RCC. This could be an important adaptive mechanism for survival under harsh conditions such as serum deprivation or exposure to drugs. The schematic description of this proposed model (Figure 8) recapitulates our finding that while VP, SF and OBT are relatively highly specific and exert an expected effect on their respective target in cells growing under RCCs, these drugs not only affect other targets but also induce a paradoxical response (see figure 4A-B) and 2) in cells growing under PPSS.
There are several other studies showing that cells that are selected for resistance to one specific anticancer drug become resistant to other unrelated anticancer drugs with different targets and overexpress multiple proteins associated with chemoresistance. For instance, MCF-7 cells selected for Docetaxel resistance by intermittent exposure to moderate concentrations of this drug have been found to be highly resistant also for Paclitaxel, Doxorubicin, Methotrexate, and 5-Fluorouracil, and showed elevated level of ABCB1, β-I, and β-III tubulin mRNA (Li et al., 2014b). Similarly, breast cancer cells selected for 18 months with doxorubicin or paclitaxel were found to show cross-resistance to Cisplatin and 5-Fluorouracil (Tegze et al., 2012).
Our model explains the underlying mechanisms by which any given drug (e.g. VP) actually increase the expression of its target (MDR1 in this case) in cell growing under PPSS – while VP targets MDR1 under PPSS, due to the extensive crosstalk and global rewiring of pathways, VP activates ABCG2 and Bcl-2 signaling pathways and/or inhibits regulatory negative feedback mechanisms that mediate MDR1, ABCG2 and Bcl-2 activity. As result, all proteins in this regulatory network are overexpressed. We posit that similar increases in target proteins occur in response to the other inhibitors such as SP and OBT with cells grown under PPSS.
We are aware that our model shown in Figure 8 may be oversimplistic because it only takes into consideration a few key proteins and it is likely that many more proteins may be involved (and vary by cell type). For instance, gene expression studies have identified that in addition to ABCB1 and ABCG2, other proteins such as DNA Topoisomerase II, Major Vault Protein as well as changes in tubulin isoform compositions are likely associated with PX or DOX resistance (Tegze et al., 2012). Regardless, our model explains not only the paradoxical response to VP, SF and OBT found in cells growing under PPSS but it also is the first to propose that multidrug resistance may arise quickly in cancer cells as consequence of lack of external mitogenic stimulation likely by rewiring of signaling pathways and feedback mechanisms.

5 Conclusions

In this study, we demonstrate that H460 and MCF-7 cells growing under PPSS are resistant to conventional anticancer drugs such as Paclitaxel, Colchicine, Hydroxyurea, Obatoclax, and the PI3K inhibitors LY294002 and Wortmannin in a manner similar to CSCs/CS-LCs. Although H460 cells expressed high protein levels of Sox2, MDR1, ABCG2, Bcl-2 and pAKT that are usually associated with increased stemness, they were not associated with increased chemoresistance to PX. This data is supported by the fact 1) that pharmacological inhibition of each of these proteins alone was not enough to reverse the resistance to PX, and 2) in MCF-7 cells growing under PPSS, the basal levels of some of these proteins (e.g. MDR1, ABCG2, Sox2, XIAP) were either unaltered or reduced compared to cells growing under RCCs. Recently, Im et. al. found a strong correlation between Sox2 induction and ROS accumulation in glioblastoma cells growing under serum deprivation (Im et al., 2015 ). However, this study found that treatment with the ROS scavenger N-acetyl-l-cysteine did not inhibit Sox2 expression, suggesting that ROS accumulation is not an essential requirement for induction of Sox2 or may not be the only regulatory factor. Taken together, our data supports the hypothesis that cells growing under PPSS display enhanced chemoresistance likely by multiple mechanisms, and that targeting only one of them is not sufficient to either significantly decrease cell viability or sensitize PPSS cells to other anticancer drugs. Our results may explain why drugs targeting proteins associated with increased resistance usually fail or have modest impact in clinical trials. For instance, i) a Phase III clinical trial demonstrated that SF had no clinical benefit when added to Carboplatin (CB)-PX chemotherapy as first- line treatment for NSCLC (Scagliotti et al., 2010), and ii) a Phase II clinical trial found that although OBT was well tolerated, it did not improve the clinical outcome when added to CB/Etoposide in first- line treatment of ES-SCLC (Langer et al., 2014). Regarding VP, one study enrolling 40 patients showed that only five cases achieved complete response and 29 cases achieved partial response when added in combination with chemotherapy drugs (Huang et al., 2013). These clinical studies provide further support for the utility of the PPSS model as a useful tool to evaluate anti-cancer co-therapy regimens in lung cancer and other cancers in general.
Our finding that DIG at therapeutic concentrations has potent antiproliferative effects on cells growing under RCCs, PPSS (Yakisich, 2016a) and LTSs (Yakisich et al., 2016b) and downregulates expression of several key proteins associated with increased stemness (Figure 6) provides a new alternative treatment approach to eliminating lung and breast cancer cells that are intrinsically resistant to conventional chemotherapy. Finally, cells growing under PPSS provide a simple and complementary model of multiresistance that can be useful as a platform to screen drugs and drugs combinations targeting multiple cancer phenotypes as well as to pursue studies aimed at elucidating underlying mechanisms. An important advantage of this system is the relatively short time that is required to generate a multidrug resistant model in cell lines (less than 2 weeks) compared to traditional methods to generate drug-resistant cell lines, a process that can take up to 18 months (Tegze et al., 2012). This study provides a useful description of the PPSS model and provides valuable insight into the various intracellular protein networks that may drive chemoresistance in these cells.

7 References

Akunuru S, Zhai QJ, and Zheng Y. 2012. Non-small cell lung cancer stem/progenitor cells are enriched in multiple distinct phenotypic subpopulations and exhibit plasticity. Cell Death Dis 3: e352.
Azad N, Iyer AK, Wang L, Lu Y, Medan D, Castranova V, and Rojanasakul Y. 2010. Nitric oxide- mediated bcl-2 stabilization potentiates malignant transformation of human lung epithelial cells. Am J Respir Cell Mol Biol 42: 578-585.
Azad N, Vallyathan V, Wang L, Tantishaiyakul V, Stehlik C, Leonard SS, and Rojanasakul Y. 2006. S- nitrosylation of Bcl-2 inhibits its ubiquitin-proteasomal degradation. A novel antiapoptotic mechanism that suppresses apoptosis. J Biol Chem 281: 34124-34341.
Barancík M, Bohácová V, Sedlák J, Sulová Z, and Breier A. 2006 LY294,002, a specific inhibitor of PI3K/Akt kinase pathway, antagonizes P-glycoprotein-mediated multidrug resistance. Eur J Pharm Sci 29: 426-434.
Basu-Roy U, Bayin NS, Rattanakorn K, Han E, Placantonakis DG, Mansukhani A, and Basilico C. 2015 Sox2 antagonizes the Hippo pathway to maintain stemness in cancer cells. Nat Commun 6: 6411.
Bergman AM, Pinedo HM, Talianidis I, Veerman G, Loves WJ, van der Wilt CL, and Peters GJ. Increased sensitivity to gemcitabine of P-glycoprotein and multidrug resistance-associated protein- overexpressing human cancer cell lines. Br J Cancer 88: 1963-1970.
Brower M, Carney DN, Oie HK, Gazdar AF, and Minna JD. 1986. Growth of cell lines and clinical specimens of human non-small cell lung cancer in a serum-free defined medium. Cancer Res 46: 798- 806.
Calvo F, Brower M, and Carney DN. 1984. Continuous culture and soft agarose cloning of multiple human breast carcinoma cell lines in serum-free medium. Cancer Res 44: 4553-4559.
Carey FA, Lamb D, and Bird CC. 1990. Intratumoral heterogeneity of DNA content in lung cancer.Cancer 65: 2266-2269.
Castells M, Milhas D, Gandy C, Thibault B, Rafii A, Delord JP, and Couderc B. 2013.Microenvironment mesenchymal cells protect ovarian cancer cell lines from apoptosis by inhibiting XIAP inactivation. Cell Death Dis 4: e887.
Chatterjee A, Chattopadhyay D, and Chakrabarti G. 2015. MiR-16 targets Bcl-2 in paclitaxel-resistant lung cancer cells and overexpression of miR-16 along with miR-17 causes unprecedented sensitivity by simultaneously modulating autophagy and apoptosis. Cell Signal 27: 189-203.
Chiappori A, Williams C, Northfelt DW, Adams JW, Malik S, Edelman MJ, Rosen P, Van Echo DA, Berger MS, and Haura EB. 2014. Obatoclax mesylate, a pan-bcl-2 inhibitor, in combination with docetaxel in a phase 1/2 trial in relapsed non-small-cell lung cancer. 9: 121-125.
Coughlin SS, Matthews-Juarez P, Juarez PD, Melton CE, and King M. 2014. Opportunities to address lung cancer disparities among African Americans. Cancer Med 3: 1467-1476.
Cruz MH, Sidén A, Calaf GM, Delwar ZM, and Yakisich JS. 2012. The stemness phenotype model. ISRN Oncol 2012.
Darzynkiewicz Z, Halicka HD, Zhao H, and Podhorecka M. 2011. Cell synchronization by inhibitors of
DNA replication induces replication stress and DNA damage response: analysis by flow cytometry. Methods Mol Biol 761: :85-96.
Dawood S, Austin L, and Cristofanilli M. 2014 Cancer stem cells: implications for cancer therapy. Oncology (Williston Park) 28: 1101-1107, 1110.
DeSantis C, Ma J, Bryan L, and Jemal A. 2014 Breast cancer statistics, 2013. CA Cancer J Clin 64: 52- 62.
DeSantis CE, Lin CC, Mariotto AB, Siegel RL, Stein KD, Kramer JL, Alteri R, Robbins AS, and Jemal A. 2014. Cancer treatment and survivorship statistics, 2014. CA Cancer J Clin 64: 252-271.
Detterbeck FC, Mazzone PJ, Naidich DP, and Bach PB. 2013. Screening for Lung Cancer: Diagnosis and Management of Lung Cancer, 3rd ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 143: e78S–e92S.
Garcia MA, Alvarez MS, Sailem H, Bousgouni V, Sero J, and Bakal C. 2012 Differential RNAi screening provides insights into the rewiring of signalling networks during oxidative stress. Mol Biosyst 8: 2605-2613.
Gongora C. 2014. Sorafenib inhibits ABCG2 and overcomes irinotecan GX15-070 resistance–response. Mol Cancer Ther 13: 764.
Gupta PB, Fillmore CM, Jiang G, ,, Shapira SD, Tao K, Kuperwasser C, and Lander ES. 2011 Stochastic state transitions give rise to phenotypic equilibrium in populations of cancer cells. Cell 146: 633-644. Erratum in: Cell. 2011; 2147(2015):1197. Cell. 2011; 2146(2016):1042. .
Han CY, Cho KB, Choi HS, Han HK, and Kang KW. 2008. Role of FoxO1 activation in MDR1 expression in adriamycin-resistant breast cancer cells. Carcinogenesis 29: 1837-1844.
Huang J, Zhang T, Ma K, Fan P, Liu Y, Weng C, Fan G, Duan Q, and Zhu X. 2013. Clinical evaluation of targeted arterial perfusion of verapamil and chemotherapeutic drugs in interventional therapy of advanced lung cancer. Cancer Chemother Pharmacol 72: 889-896.
Im CN, Yun HH, Yoo HJ, Park MJ, and Lee JH. 2015 Enhancement of SOX-2 expression and ROS accumulation by culture of A172 glioblastoma cells under non-adherent culture conditions. Oncol Rep 34: 920-928.
Kim HJ, Lee KY, Kim YW, Choi YJ, Lee JE, Choi CM, Baek IJ, Rho JK, and Lee JC. 2015. P- glycoprotein confers acquired resistance to 17-DMAG in lung cancers with an ALK rearrangement. BMC Cancer 15: 553.
Kubo T, Takigawa N, Osawa M, Harada D, Ninomiya T, Ochi N, Ichihara E, Yamane H, Tanimoto M, and Kiura K. 2013. Subpopulation of small-cell lung cancer cells expressing CD133 and CD87 show resistance to chemotherapy. Cancer Sci 104: 78-84.
Langer CJ, Albert I, Ross HJ, Kovacs P, Blakely LJ, Pajkos G, Somfay A, Zatloukal P, Kazarnowicz A, Moezi MM, et al. 2014. Randomized phase II study of carboplatin and etoposide with or without obatoclax mesylate in extensive-stage small cell lung cancer. Lung Cancer 85: 420-428.
Li D, Zhao LN, Zheng XL, Lin P, Lin F, Li Y, Zou HF, Cui RJ, Chen H, and Yu XG. 2014a. Sox2 is involved in paclitaxel resistance of the prostate cancer cell line PC-3 via the PI3K/Akt pathway. Mol Med Rep 10: 3169-3176.
Li W, Zhai B, Zhi H, Li Y, Jia L, Ding C, Zhang B, and You W. 2014b. Association of ABCB1, β tubulin I, and III with multidrug resistance of MCF7/DOC subline from breast cancer cell line MCF7.Tumour Biol 35: 8883-8891.
Li Y, Chen K, Li L, Li R, Zhang J, and Ren W. 2015. Overexpression of SOX2 is involved in paclitaxel resistance of ovarian cancer via the PI3K/Akt pathway. Tumour Biol.
Lin CY, Barry-Holson KQ, and Allison KH. 2016. Breast cancer stem cells: are we ready to go from bench to bedside? Histopathology 68: 119-137.
Lin Y, Shi R, Wang X, and Shen HM. 2008. Luteolin, a flavonoid with potential for cancer prevention and therapy. Curr Cancer Drug Targets 8: 634-646.
Lopez-Ayllon BD, Moncho-Amor V, Abarrategi A, de Cáceres II, Castro-Carpeño J, Belda-Iniesta C, Perona R, and Sastre L. 2014. Cancer stem cells and cisplatin-resistant cells isolated from non-small- lung cancer cell lines constitute related cell populations. Cancer Med 3: 1099-1111.
Luo J, Zhou X, and Yakisich JS, . 2014. Stemness and plasticity of lung cancer cells: paving the road for better therapy. Onco Targets Ther 7: 1129-1134.
Ma P, and Mumper RJ. 2013. Paclitaxel Nano-Delivery Systems: A Comprehensive Review. J Nanomed Nanotechnol 4: 1000164.
Maráz A, Furák J, Pálföldi R, Eller J, Szántó E, Kahán Z, Thurzó L, Molnár J, Tiszlavicz L, and Hideghéty K. 2011. Roles of BCL-2 and MDR1 expression in the efficacy of paclitaxel-based lung cancer chemoradiation. Anticancer Res 31: 1431-1436.
Medan D, Luanpitpong S, Azad N, Wang L, Jiang BH, Davis ME, Barnett JB, Guo L, and Rojanasakul Y. 2012. Multifunctional role of Bcl-2 in malignant transformation and tumorigenesis of Cr(VI)transformed lung cells. PLoS One 7: e37045.
Mujagic H, Conger BM, Smith CA, Occhipinti SJ, Schuette WH, and Shackney SE. 1983. Schedule dependence of vincristine lethality in Sarcoma 180 cells following partial synchronization with hydroxyurea. Cancer Res 43: 3598-3603.
Nana-Sinkam SP, Powell, C.A.,. 2013. Molecular Biology of Lung Cancer: Diagnosis and Management of Lung Cancer, 3rd ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 143: e30S–e39S.
Neelakantan D, Drasin DJ, and Ford HL. 2015. Intratumoral heterogeneity: Clonal cooperation in epithelial-to-mesenchymal transition and metastasis. Cell Adh Migr 9: 1-12.
Nishioka T, Luo LY, Shen L, He H, Mariyannis A, Dai W, and Chen C. 2014 Nicotine increases the resistance of lung cancer cells to cisplatin through enhancing Bcl-2 stability. Br J Cancer 110: 1785- 1792.
Ohashi T, Yoshimasu T, Oura S, Kokawa Y, Kawago M, Hirai Y, Miyasaka M, Aoishi Y, Kiyoi M, Nishiguchi H, et al. 2015. Class III Beta-tubulin Expression in Non-small Cell Lung Cancer: A Predictive Factor for Paclitaxel Response. Anticancer Res 35: 2669-2674.
Paul I, and Jones JM. 2014. Apoptosis block as a barrier to effective therapy in non small cell lung cancer. World J Clin Oncol 5: 588-594.
Piva M, Domenici G, Iriondo O, Rábano M, Simões BM, Comaills V, Barredo I, López-Ruiz JA, Zabalza I, Kypta R, et al. 2014. Sox2 promotes tamoxifen resistance in breast cancer cells. EMBO Mol Med 6: 66-79.
Ratnikov BI, Scott DA, Osterman AL, Smith JW, and Ronai ZA. 2016. Metabolic rewiring in melanoma. Oncogene In Press.
Rudin CM, Durinck S, Stawiski EW, Poirier JT, Modrusan Z, Shames DS, Bergbower EA, Guan Y, Shin J, Guillory J, et al. 2012. Comprehensive genomic analysis identifies SOX2 as a frequently amplified gene in small-cell lung cancer. Nat Genet 44: 1111-1116.
Rustum YM, Tóth K, Seshadri M, Sen A, Durrani FA, Stott E, Morrison CD, Cao S, and Bhattacharya A. 2010. Architectural heterogeneity in tumors caused by differentiation alters intratumoral drug distribution and affects therapeutic synergy of antiangiogenic organoselenium compound. J Oncol 2010: 396286.
Santini R, Pietrobono S, Pandolfi S, Montagnani V, D’Amico M, Penachioni JY, Vinci MC, Borgognoni L, and Stecca B. 2014. SOX2 regulates self-renewal and tumorigenicity of human melanoma-initiating cells. Oncogene 33: 4697-4708.
Scagliotti G, Novello S, von Pawel J, Reck M, Pereira JR, Thomas M, Abrão Miziara JE, Balint B, De Marinis F, Keller A, et al. 2010. Phase III study of carboplatin and paclitaxel alone or with sorafenib in advanced non-small-cell lung cancer. J Clin Oncol 28: 1835-1842.
Shalli K, Brown I, Heys SD, and Schofield AC. 2005. Alterations of beta-tubulin isotypes in breast cancer cells resistant to docetaxel. FASEB J 19: 1299-1301.
Shi JF, Yang N, Ding HJ, Zhang JX, Hu ML, Leng Y, Han X, and Sun YJ. 2014. ERα directly activated the MDR1 transcription to increase paclitaxel-resistance of ERα-positive breast cancer cells in vitro and in vivo. Int J Biochem Cell Biol 53: 35-45.
Tavaluc RT, Hart LS, Dicker DT, and El-Deiry WS. 2007. Effects of low confluency, serum starvation and hypoxia on the side population of cancer cell lines. Cell Cycle 6: 2554-2562.
Tegze B, Szállási Z, Haltrich I, Pénzváltó Z, Tóth Z, Likó I, and Gyorffy B. 2012. Parallel evolution under chemotherapy pressure in 29 breast cancer cell lines results in dissimilar mechanisms of resistance. PLoS One 7: e30804.
Wei Y, Ma Y, Zhao Q, Ren Z, Li Y, Hou T, and Peng H. 2012. New use for an old drug: inhibiting ABCG2 with sorafenib. Mol Cancer Ther 11: 1693-1702.
Yakisich JS, Azad N, Venkatadri R, Kulkarni Y, Wright C, Kaushik V, and Iyer AKV. 2016b. Formation of Tumorspheres with Increased Stemness without External Mitogens in a Lung Cancer Model. Stem Cells Int 2016: 5603135.
Yakisich JS, Azad, N., Venkatadri, R., Kulkarni, Y., Wright, C., Kaushik, V., O‟Doherty, G.A., Iyer, A.K.V.,. 2016a. Digitoxin And Its Synthetic Analog MonoD Have Potent Antiproliferative Effects On Lung Cancer Cells And Potentiate The Effects Of Hydroxyurea And Paclitaxel. Oncol Rep 35: 878-886. Yang J, Liu X, Bhalla K, Kim CN, Ibrado AM, Cai J, Peng TI, Jones DP, and Wang X. 1997 Prevention of apoptosis by Bcl-2: release of cytochrome c from mitochondria blocked. Science 275: 1129-1132.
Yang Y, Fan Y, Qi Y, Liu D, Wu K, Wen F, and Zhao S. 2015. Side population cells separated from A549 lung cancer cell line possess cancer stem cell-like properties and inhibition of autophagy potentiates the cytotoxic effect of cisplatin. Oncol Rep 34: 929-935.
Yang YL, Luo XP, and Xian L. 2014. The prognostic role of the class III β-tubulin in non-small cell lung cancer (NSCLC) patients receiving the taxane/vinorebine-based chemotherapy: a meta-analysis. PLoS One 9: e93997.
Zhang X, Lou Y, Wang H, Zheng X, Dong Q, Sun J, and Han B. 2015a. Wnt signaling regulates the stemness of lung cancer stem cells and its inhibitors exert anticancer effect on lung cancer SPC-A1 cells. 32: 95.
Zhang X, Lou Y, Zheng X, Wang H, Sun J, Dong Q, and Han B. 2015b. Wnt blockers inhibit the proliferation of lung cancer stem cells. Drug Des Devel Ther 9: 2399-2407.