Drug-resistance in doxorubicin-resistant FL5.12 hematopoietic cells: elevated MDR1, drug efflux and side-population positive and decreased BCL2-family member expression

Chemotherapeutic drug treatment can result in the emergence of drug-resistant cells. By culturing an interleukin-3 (IL-3)-dependent cell line, FL5.12 cells in the presence of the chemotherapeutic drug doxorubicin, we isolated FL/Doxo cells which are multi-drug resistant. Increased levels of drug efflux were detected in FL/Doxo cells which could be inhibited by the MDR1 inhibitor verapamil but not by the MRP1 inhibitor MK571. The effects of TP53 and MEK1 were examined by infection of FL/Doxo cells with retroviruses encoding either a dominant negative TP-53 gene (FL/Doxo+ TP53 (DN) or a constitutively-activated MEK-1 gene (FL/Doxo + MEK1 (CA). Elevated MDR1 but not MRP1 mRNA transcripts were detected by quantitative RT-PCR in the drug-resistant cells while transcripts encoding anti-apoptotic genes such as: BCL2, BCLXL and MCL1 were observed at higher levels in the drug-sensitive FL5.12 cells. The percentage of cells that were side-population positive was increased in the drug-resistant cells compared to the parental line. Drug-resistance and side-positive population cells have been associated with cancer stem cells (CSC). Our studies suggest mechanisms which could allow the targeting of these molecules to prevent drug-resistance.

The RAS/RAF/MEK/ERK pathway also consists of proteins which are considered as oncogenes, e.g., RAS, RAF and other proteins which may have tumor suppressor activities e.g., protein phosphatase 2A (PP2A), dual specificity protein phosphatase 1 (DUSP1) and others. The altered expression of components of this pathway can result in resistance to targeted therapeutics [33].
Proteins that are frequently referred to as "drugtransporters" are often upregulated in drug-resistant cells [62][63][64]. The "drug-transporting" functions of these proteins are often considered a "moon-lighting" activity of these proteins as they often have physiological roles in various organs (e.g., gut, colon) to transport xenobiotics [65][66][67].
Increased expression of MDR1 and/or MRP1 has been observed in certain drug-resistant hematopoietic cells such as HL-60 [68]. The MDR1 (a.k.a., Pgp, ABCB1) has been shown to efflux various chemotherapeutic drugs such as doxorubicin, daunorubicin, paclitaxel and others [69]. MRP1 (a.k.a. ABCC1) will transport various molecules including: daunorubicin, doxorubicin, edatrexate, etoposide, epirubicin, idarubicin, irinotecan, methotrexate, paclitaxel, vinblastine, vincristine and others [70]. Drug transporters may in some cases be regulated by miRs which in turn will change the sensitivity of certain tumors to chemotherapeutic drugs [71].
FL5.12 is an interleukin-3 (IL-3)-dependent hematopoietic cell line derived from the fetal liver of inbred mice [72]. They absolutely require IL-3 for their survival and after 18-24 hours of IL3-deprivation they commence to undergo apoptosis [73]. The IL3-dependency of these cells can be relieved after introduction of various activated oncogenes either by themselves (v-ABL, BCR-ABL) [74] or in combination (AKT and RAF) [75]. In addition, activated cytokine genes such as IL-3 genes with mutations in the 3'untranslated region in the sequence motifs associated with mRNA (in)stability will relieve the dependency of these cells on the addition of exogenous IL-3 [76,77]. FL5.12 cells do not normally cause tumors upon injection into nude immunocompromised mice, however, once their dependency on addition of exogenous IL-3 is abrogated, they usually will form tumors in immunocompromised mice [77,78].
We derived a panel of doxorubicin-resistant FL5.12 cells (FL/Doxo) by culturing them in the presence of doxorubicin for prolonged time periods [79]. These cells have been determined to have altered TP53 and ERK signaling. Furthermore, these cells show an altered response to certain small molecule inhibitors which target various molecules such as the mTORC1 complex [80]. We have previously shown that the doxorubicin-resistant FL/ Doxo cells are cross resistant to daunorubicin, paclitaxel but not 5-fluorouracil (5FU) [79]. We demonstrated recently that increased levels of mRNA transcripts encoding MDR1 but not MRP1 were detected by q-RT-PCR in FL/Doxo in comparison to FL5.12 cells [80]. The drug-resistance of the FL/Doxo cells could be increased further by infection of the cells with retroviruses which encode either dominant negative (DN) TP53 (FL5.12 cells have wild-type TP53) or constitutively-active (CA) MEK1 [79].
In the following studies, the drug efflux abilities of doxorubicin-resistant FL/Doxo and doxorubicinsensitive FL5.12 cells were determined, as well as, the presence of side-population positive cells in both types of cells. We document the effects of TP53(DN) and MEK1(CA) oncoproteins on drug-resistance and sidepopulation positive cells. These studies provide insights into mechanisms responsible for drug-resistance and sidewww.impactjournals.com/oncotarget positive population cells and may indicate approaches to target these cells which are often responsible for cancer development and progression.

IL-3 suppresses effects of doxorubicin in parental FL5.12 cells
The effects of doxorubicin were examined on the cytokine-dependent FL5.12 cells. When these cells were cultured in the presence of IL-3, the IC 50 for doxorubicin was approximately 20 nM (Figure 1, Panel A). In contrast, when the cells were cultured in the absence of IL-3, the level of growth decreased and the IC 50 for doxorubicin was approximately 3 nM ( Figure 1, Panel A) which was 6.7-fold lower than when the cells were cultured in IL-3.

Doxorubicin-resistance of FL/Doxo and derivative cells
The doxorubicin-sensitivities of the FL5.12, FL/ Doxo, FL/Doxo + TP53(DN) and FL/Doxo + MEK1(CA) cells are presented in Figure 1, Panel B. In this experiment the doxorubicin IC 50 s for: FL5.12, FL/Doxo, FL/Doxo + TP53(DN) and FL/Doxo + MEK1(CA) cells were approximately 12, 30, 100 and 55 nM respectively. These IC 50 s were all measured on the same day. There is some variation in IC 50 s measured on different days that is why it is better to make comparisons on results obtained on the same day.

Drug-resistant FL/Doxo cells are frequently larger cells than drug-sensitive FL5.12 cells
The morphologies of the FL5.12 and FL/Doxo cells were examined by phase contrast microscopsy ( Figure 2). FL/Doxo cells (Panel B) were larger than parental FL5.12 cells (Panel A). Often there were "giant" cells present in the FL/Doxo cultures [81]. Large multi-nucleate cells have been observed with drug-resistant prostate cancer cells [82].  cells to different concentrations doxorubicin and then examined by the 2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay as described [79,80]. The MTT assay is a colorimetric assay that measures the reduction of MTT by mitochondrial succinate dehydrogenase as described [80]. Arrows pointing to the X-axis indicate where the IC 50 can be estimated. Statistical analysis (unpaired t test results) indicated that the two-tailed P value: for (Panel A) FL5.12 in the presence of IL-3 and FL5.12 in the absence of IL-3 is less than 0.0001 which is considered to be extremely statistically significant.

Drug-resistant FL/Doxo cells display altered drug efflux activity
The abilities of the FL5.12 and FL/Doxo cells to efflux the dye rhodamine 123 (Table 1) were determined by drug efflux assays in the presence and absence of the MDR1 (Pgp, ABCB1) pump inhibitor verapamil or the MRP1 inhibitor MK571. Verapamil is a substrate of ABCB1 (MDR1) [83]. A FACS assay was used to quantitate drug efflux [83].
The background flouresence in FL/Doxo cells (25.9) was higher than that detected in FL5.12 cells (4.9). When the FL/Doxo cells were resuspended in medium containing verapamil, a value of 24.4 was observed while when FL5.12 cells were set up under similar conditions, a value of 6.8 was observed. Higher background levels were observed when the cells were incubated with the MRP1 inhibitor, MK571 as a value of 43.3 was observed with the FL/Doxo cells and 29 was observed with the FL5.12 cells. These values decreased to 34.8 and 21.3 in the FL/Doxo and FL5.12 cells respectively at the T2 hr. time points.
The fold increase in MDR1 expression in the cells cultured presence of IL-3 ranged from 3.8-fold in FL/Doxo to 19.2-fold in FL/Doxo+MEK1(CA) cells. In contrast, the levels of MRP1 mRNA transcripts were similar in FL5.12 and FL/Doxo, when the cells were cultured in the presence of IL-3. In contrast, lower levels (2-and 1.8-fold respectively) of MRP1 mRNA transcripts were detected in FL/Doxo+TP53(DN) and FL/Doxo+MEK1(CA) cells respectively than in FL5. 12

Effects of signal transduction inhibitors on sensitivity to doxorubicin in FL5.12 and drug resistant FL/Doxo derivative cells
The effects of titrating doxorubicin in the presence of the multi-kinase inhibitor sorafenib, which can inhibit RAF and other kinases, the MEK1 inhibitor (PD0325901), and the BCL2 inhibitor ABT737 were examined in Figures 5-7. In all of these graphs, the effects of suboptimal doses of the inhibitors and doxorubicin titrations were determined the same day. They are graphed in this manuscript to illustrate the differences between doxorubicin in the absence of inhibitor and doxorubicin in the presence of the inhibitor in the individual cell lines. In a previous manuscript, the effects of some of these and other inhibitors on the four different FL5.12-derived cell lines were plotted in the same graphs together [80].
FL5.12 cells were cultured in the presence of doxorubicin in the absence and presence of 500 nM of When FL/Doxo + TP53(DN) cells were cultured in the absence of doxorubicin, an IC 50 of 120 nM was observed. Upon culture of FL/Doxo +TP53(DN) in the presence of doxorubicin and 500 nM sorafenib, the IC 50 for doxorubicin remained the essentially the same (Panel C).
When FL/Doxo + MEK1(CA) cells were cultured in the absence of doxorubicin, an IC 50 of 120 nM was observed. Upon culture of FL/Doxo +TP53(DN) in the presence of doxorubicin and 500 nM sorafenib, the IC 50 for doxorubicin remained the essentially the same (Panel D).
The effects of the MEK1 inhibitor PD0329501 on the doxorubicin IC 50 s were also examined. When FL5.12 cells were cultured in the presence of 500 nM PD0329501 MEK1 inhibitor, the IC 50 did not change from when they were cultured in the absence of the MEK1 inhibitor ( Figure 6, Panel A). Likewise, when FL/Doxo and FL/ Doxo + TP53(DN) cells were cultured in the presence and absence of 500 nM PD0325901 MEK inhibitor, the doxorubicin IC 50 s did not change ( Figure 6, Panels B and C). In contrast, when FL/Doxo + MEK1(CA) cells were cultured in the presence of 500 nM PD0329501, the doxorubicin IC 50 was approximately 100 nM while when the cells were cultured in the absence of the MEK1 inhibitor, the doxorubicin IC 50 was approximately 150 nM ( Figure 6, Panel D). Thus, the MEK1 inhibitor treatment reduced the IC 50 for doxorubicin approximately 1.5-fold in the FL/Doxo + MEK1(CA) cells.
The effects of the BCL2 inhibitor ABT737 on the doxorubicin IC 50 s were also examined ( Figure 7). When FL5.12 cells were cultured in the presence of 500 nM ABT 737 inhibitor, the level of growth decreased and the IC 50 for doxorubicin declined from 20 nM to 8 nM, a 2.5-fold decrease (Figure 7, Panel A). Likewise, when FL/Doxo,

Effects of combination of signal transduction and BCL2 inhibitors on FL5.12 and FL/Doxo derivative cells
The effects of combining sorafenib with the BCL2 inhibitor ABT737 were examined in Figure 8. When FL5.12 cells were cultured in the presence of the multikinase sorafenib inhibitor, an IC 50 of 5,000 nM was observed (Figure 8, Panel A). In contrast, when they were incubated with sorafenib in the presence of 500 nM ABT-737 inhibitor, the level of growth decreased and an IC 50 of approximately 65 nM (77-fold decrease) was observed.
When FL/Doxo cells were cultured in the presence of sorafenib, an IC 50 was not obtained even with concentrations up to 5,000 nM ( Figure 8, Panel B). In contrast, the level of growth decreased when FL/Doxo cells were cultured with sorafenib in the presence of 500 nM ABT-737 inhibitor and an IC 50 of approximately 3 nM (approximately a 1,666-fold decrease) with respect to the sorafenib inhibitor was observed.
When FL/Doxo + TP53 (DN) cells were cultured in the presence of sorafenib, the IC 50 was not obtained even with doses of 5,000 nM doxorubicin (Figure 8, Panel C). In contrast, the level of growth decreased when FL/Doxo + TP53(DN) cells were cultured with sorafenib in the presence of the BCL2 inhibitor and an IC 50 of approximately 2,000 nM sorafenib was observed. When FL/Doxo + MEK1(CA) cells were cultured in the presence of sorafenib, an IC 50 was not observed even with concentrations of 5,000 nM sorafenib (Figure 8, Panel D). Addition of the BCL2 inhibitor decreased growth and an IC 50 of approximately 4,000 nM sorafenib was observed. In the FL/Doxo + MEK1(CA) cells, the decrease was not observed until approximately 100 nM sorafenib, demonstrating that the FL/Doxo + MEK1(CA) cells were more resistant to the BCL2 inhibitor than the other cells.
Essentially the reciprocal experiments were performed. Namely, the effects of titrating cells with the BCL2 inhibitor in the presence and absence of 500 nM sorafenib (Figure 9). Addition of 500 nM sorafenib did not increase the effect of the BCL2 inhibitor on FL5.12 cells (Figure 9, Panel A). When FL/Doxo and FL/Doxo + TP53(DN) cells were treated with 500 nM sorafenib and titrated with the BCL2 inhibitor, the IC 50 s decreased from 40 to 10 nM (4-fold) in FL/Doxo (Panel B) and from 170 to 130 nM (1.3-fold) in FL/Doxo + TP53(DN) cells respectively (Panels B and C). Treatment of the FL/Doxo + MEK1(CA) cells with up to 5,000 nM BCL2 inhibitor did not reach an IC 50 . Combined treatment with different doses of the BCL2 inhibitor and 500 nM sorafenib reduced cell growth but did not reduce the IC 50 for the BCL2 inhibitor. The experiments mirror the results presented in Figure 8 and document that the FL/Doxo + MEK1(CA) cells were more resistant to the BCL2 inhibitor than the other FL5.12 and FL/Doxo-derived cells.
When FL5.12 cells were cultured in the presence of the PD0329501 MEK1 inhibitor, an IC 50 was not obtained even with concentrations up to 5,000 nM ( Figure 10, Panel A). In contrast, the level of growth decreased when they were cultured with the MEK1 inhibitor in the presence of 500 nM ABT-737 inhibitor and an IC 50 of approximately 100 nM with respect to the MEK1 inhibitor was observed.
When FL/Doxo cells were cultured in the presence of the PD0329501 MEK1 inhibitor, an IC 50 was not obtained even with concentrations up to 5,000 nM ( Figure  10, Panel B). In contrast, the level of growth decreased when FL/Doxo cells were cultured with the MEK1 inhibitor in the presence of 500 nM ABT-737 inhibitor and an IC 50 of approximately 3 nM with respect to the MEK1 inhibitor was observed.
When FL/Doxo + TP53 (DN) cells were cultured in the presence of the PD0329501 MEK1 inhibitor, an IC 50 of approximately 3,000 nM was observed ( Figure 10, Panel C). The level of growth decreased slightly when FL/Doxo + TP53(DN) cells were cultured with the MEK1 inhibitor in the presence of 500 nM ABT-737 inhibitor, but the IC 50 for the MEK1 inhibitor was not obtained even with doses of 5,000 nM PD0329501.
When FL/Doxo + MEK1(CA) were cultured in the presence of the PD0329501, an IC 50 of approximately 400 nM MEK1 inhibitor was observed ( Figure 10, Panel D). In contrast, the addition of 500 nM BCL2 inhibitor did not decrease the concentration of the MEK1 inhibitor required to reach the IC 50 , if anything, it antagonized slightly with the MEK1 inhibitor in FL/Doxo + MEK1(CA) cells.
Essentially the reciprocal experiments were performed. Namely, the effects of titrating cells with the BCL2 inhibitor in the presence and absence of the 500 nM MEK1 inhibitor were determined ( Figure 11). Addition of 500 nM MEK inhibitor did not reduce the effect of the BCL2 inhibitor on FL5.12 cells (Figure 11  more sensitive to the effects of the BCL2 inhibitor than the MEK1 inhibitor.

Drug-resistant FL/Doxo cells displayed increased levels of side-population positive cells
The presence of side-population positive cells in the FL5.12 and different FL/Doxo cells was examined. Sidepopulation positive cells is often associated with drugresistance and CSCs [86]. MDR1 expression has been shown to be important in glioma CSCs [87]. Verapamil has been shown to inhibit tumor progression by targeting the side population of certain types of cancer cells [88].
Side-population positive cells were detected as we and others have previously described [88][89][90]. Statistically significant higher levels of side-population positive cells were detected by FACSs analysis in FL/Doxo than in FL5.12 cells (Figure 12). Approximately 1.9-fold more side population positive cells were detected in FL/Doxo cells than in FL5.12 cells. The detection of side population positive cells was prevented with the FL5. 12

DISCUSSION
The cytokine IL-3 was protective in FL5.12 cells against the effects of doxorubicin. When FL5.12 cells were cultured in the absence of IL-3, the IC 50 was approximately 6.7-fold less than when they were grown in the presence of IL-3. FL5.12 cells will undergo apoptosis when they are cultured in the absence of IL-3 [75,79,81]. Upon their incubation in the absence of IL-3 in the presence of doxorubicin, their growth is even more inhibited than when they are cultured in the presence of IL-3 and doxorubicin. Doxorubicin-resistant FL/Doxo cells often were larger than doxorubicin-sensitive FL/ Doxo cells and acridine orange staining has indicated that they are often multi-nucleate [81].
Doxorubicin induced more apoptosis in the doxorubicin-sensitive cells at lower doxorubicin concentrations than in the doxorubicin-resistant cells [79]. Similar results were observed with daunorubicin and paclitaxel with the FL5.12 and FL/Doxo cells [79]. Previously we demonstrated that there was decreased caspase activation (caspases 3, 8 and 10) in the doxorubicin-resistant FL/Doxo cells than in the doxorubicin-sensitive FL5.12 cells after doxorubicin treatment [79]. In this current manuscript, we have focused on drug efflux, the effects of the MDR1 and the MRP1 inhibitors on drug efflux, drug transporter gene expression, and the effects of signal transduction and BCL2 inhibitors on growth, and the presence of side-population positive cells in drug-sensitive and drug-resistant cells.
Both FL5.12 and FL/Doxo cells could readily efflux rhodamine 123. In both FL5.12 and FL/Doxo cells, the level of fluorescence detected after rhodamine 123 staining decreased after 2 hrs. on ice. The efflux of rhodamine 123 was inhibited in both FL5.12 and FL/Doxo cells after verapamil treatment. In contrast, after treatment with the MK571 MRP1 inhibitor, the efflux of rhodamine 123 was not as inhibited. The FL/Doxo cells appeared to efflux approximately 2.9-fold more rhodamine 123 than the FL5.12 cells when they were treated with the MK571 MRP1 inhibitor. The levels of mRNAs of MDR1 and MRP1 have been quantitated by qRT-PCR [80]. In this current manuscript we discuss the fold differences in mRNA expression. The cell lines with the highest degree of drug-resistance, FL/Doxo+TP53(DN) and FL/ Doxo+MEK1(CA) expressed approximately 10-100-fold more MDR1 mRNA transcripts than in the drug-sensitive FL5.12 cells. FL/Doxo cells expressed approximately 4-fold more MDR1 transcripts than FL5.12 cells. The fold differences in MDR1 expression between the doxorubicinsensitive FL5.12 cells and doxorubicin-resistant FL/Doxo and derivative cells increased when the cells were cultured in the presence of doxorubicin as the levels of MDR1 mRNAs detected in the FL5.12 cells decreased.
In contrast, 2-to 4-fold less BAX, BIM, BCL2 and BIM mRNAs and between 1.3 to 2-fold less MCL1 mRNA transcripts were detected in the cells with the highest-degree of drug-resistance, FL/Doxo+TP53(DN) and FL/Doxo+MEK1(CA) than in the drug-sensitive FL5.12 cells [80]. The FL/Doxo cells expressed 2-4-fold less BCL2 mRNA transcripts than the FL5.12 cells. The FL/Doxo cells were very sensitive to the BCL2/BCLXL inhibitor ABT-737 [80]. Our results with the expression of the anti-apoptotic BCL2 and BCLXL mRNA expression were interesting as the cells that were more doxorubicinsensitive expressed more of these transcripts as detected by q-RT-PCR. Thus, in our system, doxorubicin-resistance is associated with lower expression of BCL2 and BCLXL RNA expression. Previously we determined that ectopic expression of BCL2 could enhance the ability of the RAF and MEK1 oncoproteins to abrogate the cytokinedependence of IL3-dependent cells [91,92].
Our results indicate the MEK1(CA) and the TP53(DN) genes can have effects on both the expression of MDR1 and the presence of side-population positive cells. These results are important as these characteristics are often linked with CSC. CSCs have been detected in almost every cancer type. Novel methods to target CSCs are under intense investigation. Other studies have shown the side-population positive cells includes CSCs [93]. Chemotherapeutic drug-resistance is often associated with drug transporters, side-population positive and CSCs [86]. Other investigators have documented the sensitivity of other cancer cells to verapamil treatment and their enhanced abilities to form colony in vitro and tumors in animals [94]. In the above study, the presence of side-population positive AsPC1 pancreatic cancer cells was not eliminated completely after verapamil treatment. These and our results indicate that there may be other drug transporters not inhibited by verapamil that are present in some of the cells which contribute to their side-population positive cells and drug resistance. Alternatively, the concentration of verapamil used may not have been sufficient to inhibit all the MDR1 molecules expressed in the drug-resistant cells.
An effect of sorafenib on the sensitivity to doxorubicin was observed in FL/Doxo but not in the other FL/Doxo derivative or parental FL5.12 cells. In most experiments, the FL/Doxo are more sensitive to doxorubicin, than either FL/Doxo + TP53(DN) or FL/ Doxo + MEK1(CA) cells. In these studies, addition of 500 nM PD0325901 MEK inhibitor only affected the doxorubicin IC 50 in FL/Doxo + MEK1(CA) cells. In contrast, addition of 500 nM ABT-737 BCL2 inhibitor reduced both the level of cell growth and the doxorubicin IC 50 in FL5.12, FL/Doxo and FL/Doxo +TP53(DN) and FL/Doxo + MEK1(CA) cells. Thus, the ABT-737 BCL2 inhibitor could reduce the IC 50 for doxorubicin in all cell lines. In contrast, the FL/Doxo+ MEK1(CA) cells were relatively resistant to the BCL2 inhibitor by itself or in combination with certain signal transduction inhibitors.
The PI3K/PTEN/AKT/mTORC1/GSK-3 pathway regulates the expression of various drug transporter molecules [95][96][97]. The PI3K/PTEN/AKT/mTORC1/ Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (no verapamil) and FL/Doxo (no verapamil) is 0.0195 which is considered to be statistically significant. Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (plus verapamil) and FL/Doxo (plus verapamil) is 0.2879 which is not statistically significant. Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (no verapamil) and FL/Doxo + TP53(DN) (no verapamil) is 0.0005 which is considered to be extremely statistically significant. Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (plus verapamil) and FL/Doxo + TP53 (DN) (plus verapamil) is 0.0001 which is considered extremely statistically significant. Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (no verapamil) and FL/Doxo + MEK1(CA) (no verapamil) is 0.0013 which is considered to be very statistically significant. Statistical analysis (unpaired t test results) indicated that the two-tailed P value for the FL5.12 (plus verapamil) and FL/Doxo + MEK1 (CA) (plus verapamil) is 0.0002 which is considered extremely statistically significant. GSK-3 pathway can induce multidrug-resistance in AML cells that were co-cultured with stromal cells [98].
Various signaling pathways have been implicated in drug resistance, both resistance to classical chemotherapeutic drugs, as well as, targeted therapeutics. Often resistance is associated with CSCs and CSCs frequently display altered activity of proteins which transport drugs. Our studies have documented the importance of activated MEK1 in drug-resistance and side-population positive cells in hematopoietic cells.
Parental FL5.12 and FL/Doxo + TP53(DN) cells were also sensitive to the BCL2 inhibitor, but the FL/Doxo + MEK1(CA) cells were relatively resistant to the BCL2 inhibitor ABT-737. It is possible that the activated MEK1 present in the FL/Doxo + MEK1(CA) cells, activated ERK1,2 which phosphorylated a BCL2 family member which prevented the effects of the BCL2 inhibitor. Importantly, the doxorubicin-sensitivity of all the cell lines examined could be reduced upon treatment with the BCL2 inhibitor.
Various BCL2 family members have been observed to be phosphorylated by RAF/MEK/ERK signaling which could have various effects on apoptosis, cellular proliferation, prevention of apoptosis and drug-resistance [99][100][101][102][103][104][105][106]. Interactions between RAF/MEK/ERK and BCL2 family members have been shown to abrogate the cytokine-dependency of certain hematopoietic cells [91,92]. BCL2 inhibitors increased the sensitivity of the cells to doxorubicin in the current study. Inhibition of RAF/ MEK/ERK signaling by sorafenib also had some effects on the doxorubicin-sensitivity of FL/Doxo cells.
Mutation of TP53 may also contribute to drugresistance and CSCs. We have demonstrated that introduction of the DN-TP53 gene into FL/Doxo cells resulted in more drug-resistance and increased sidepopulation positive cells. Inheritance of DN-TP53 also resulted in decreased anti-apoptotic and pro-apoptotic BCL2 family expression. In contrast, inheritance of DN-TP53 resulted in increased levels of beta2-microglobulin mRNA transcripts [80]. Interestingly, FL/Doxo + TP53 (DN) cells were very sensitive to the BCL-2 inhibitor, ABT-737. These cells were relatively resistant to MEK1 inhibitor PD0329501. Addition of 500 nM MEK inhibitor to cells titrated with the BCL2 inhibitor decreased the IC 50 required for the BCL2 inhibitor less than 2-fold.
Similar experiments performed with the combination of the MEK1 inhibitor and titration with the BCL2 inhibitor in FL/Doxo + MEK1(CA) cells reduced the IC 50 for the amount of BCL2 inhibitor approximately 500-fold. While FL/Doxo + TP53(DN) cells expressed high levels of MDR1 as did FL/Doxo + MEK1(CA) cells, but they differed in their sensitivities to BCL2 and MEK1 inhibitors.
In conclusion, our studies document the effects of signaling pathways and TP53 on the drug-resistance of hematopoietic cells. By using a single parental cell line and three drug resistant derivatives of it which either have or lack TP53 (DN) or MEK1(CA), we can identify some of the key components in resistance to doxorubicin a common, clinically used chemotherapeutic drug. Namely increased expression of MDR1 and altered BCL2 family member expression which contribute to altered sensitivity to doxorubicin as well as targeted therapeutics. Our results may have clinical significance as some of the drugs and signal transduction inhibitors and their next generation counter-parts are being used to treat various cancers.

Drug efflux assays
Drug Efflux assays were performed by FACS analysis as described previously [83].

Side population analysis
Side population analysis was performed as described [88][89][90]. Briefly, FL5.12, FL/Doxo, FL/Doxo + TP53(DN) and FL/Doxo + MEK1(CA) Cells were resuspended at 1 × 10 6 /mL in RPMI 1640 with 10% FBS and 10% WEHI-3B supernatant in the presence and absence of 25 nM verapamil. Hoechst 33342 dye (Sigma-Aldrich, Saint Louis, Missouri) was added to a final concentration of 5 μg/mL. The cells were incubated at 37 0 C for 90 minutes. Cells were then centrifuged and the pellet resupended in cold PBS containing 2% FBS. Cells were counterstained with propidum iodide (10 μg/ml). Hoechst dye was excited at 355 nM. The side-population positive analysis was then performed as described [88][89][90] on a BD FACSVantage. The side-population population, essentially disappeared in FL5.12 and FL/Doxo cells when they were cultured in the presence of verapamil. Experiments were set up in triplicate in 6-well plates.
Quantitative RT-PCR and normalization to FL5.12 cells TaqMan Gene Expression assays were described previously [80]. Levels of mRNA transcripts were normalized to the levels of the particular gene transcripts detected in FL5.12 cells which was set at 1.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest with publication of this manuscript.

FUNDING
JAM and SLA were supported in part by grants from East Carolina University Grants (#111104 and # 111110-668715-0000). LC and SR were supported in part by grants from: Intesa San Paolo Foundation. AMM was supported in part by grants from: MIUR FIRB 2011 (RBAP11ZJFA_001).