Keywords

1 Introduction

Childhood leukemia is the commonest form of childhood cancer and represents clonal proliferation of transformed hematopoietic cells as a result of genomic and proteomic changes [1,2,3]. Treatment of childhood leukemia has been extensively successive during the recent years. Patient complete remission has reached 80–85%, yet a remaining 15–20% still relapses. In the treatment of childhood leukemia, glucocorticoids (GC) still remain as first-line drugs as well as response to GC treatment still remains a significant prognostic factor [4,5,6,7]. Glucocorticoids enter the cell passively and bind to a 97 KiloDalton protein, the glucocorticoid receptor (GR) [8,9,10,11], causing through allosteric effects exposure of nuclear localization signals on the surface of GR. GR next enters the nucleus, binding to the glucocorticoid response element (GRE) followed by activated transcription of target genes caused by the effect of the receptor on the genome. The action of the GR depends solely on the type of tissue or cell that it affects. In the case of neoplasmatic lymphocytes (e.g., ALL), it is known that it sets the cell on its programmed death (apoptosis) pathway. Therefore, Prednisolone is widely used for leukemia treatment [11]. To reduce the complexity due to feedback mechanisms affecting gene expression, a snapshot of early direct effects on gene expression is necessary to be obtained at an early stage under glucocorticoid treatment. Such an analysis is expected to include key genetic regulators and initiators of further downstream pathways.

The significance of Prednisolone resistance in acute lymphoblastic leukemia (ALL) is thoroughly discussed in the literature. Prednisolone is a glucocorticoid and one of the most important agents in the treatment of ALL, especially in children, a very sensitive group of patients [2]. Resistance to Prednisolone is considered to be of crucial importance for disease prognosis [2]. Toward the identification of resistance mechanisms, several tools and approaches can be deployed. Gene expression and gene regulatory mechanisms are considered crucial for the study of diseases and, in particular, ALL. Gene expression analysis of GC-treated ALL cells may allow discovery of drugs which in combination with glucocorticoids may increase the effectiveness of anti-leukemia therapies. The use of high-throughput technologies provided new insights into cancer, in general, and to leukemia specifically as far as, diagnosis, prognosis, treatment response, and new therapeutic targets is concerned.

Previous studies have indicated that the cell line model used manifested higher proliferation rates at very low Prednisolone concentrations [1, 12]. These findings indicated that the two lowest concentrations of Prednisolone used (10 nM and 1 μM) had a mitogenic effect. Interestingly, it has been also found that total cell death after exposure to 1 μM Prednisolone is similar to that observed for higher doses (>100 μM). However, cells treated with 1 μM Prednisolone grew faster than untreated cells and cells treated with higher doses. Cells exposed to >1 μM Prednisolone manifested the same growth potential as untreated cells [1, 12]. Previous cytotoxicity data indicated that the CCRF-CEM cell line exhibited resistance to corticosteroid-induced apoptosis, 72 h after treatment. The highest dosage (700 μM, which was ~8-fold higher from the average in vivo dosage, which was calculated to be 100 μM) manifested a combined apoptotic and necrotic effect. This was in contrast to the case of GC-sensitive cells where up to 80% total cell death has been reported after treatment with only 20 nM dexamethasone. In the same cell model, it appeared that the first 48 h showed no dose-dependent cumulative cytotoxic effect of Prednisolone. Upon 48 h of incubation, a dose-dependent escalation of cytotoxicity started to become evident, which was manifested clearly as a dose-dependent effect after 72 h of Prednisolone stimulation. Specifically, upon 72 h, apoptosis exhibited a dose-dependent effect while necrosis peaked at 1 μM, dropped sharply between 1 μM and 40 μM to increase dose-dependently again. Interestingly, bioinformatics analysis has shown that CRE-BP1/c-Jun (AP-1) heterodimer transcription factor was overexpressed transcription factors of differentially expressed genes under Prednisolone treatment, where c-Jun is induced by GCs at the transcriptional level. C-Jun expression it has been shown that it is important in the GC-induced apoptosis. However, at protein level there is an inhibition of c-Jun causing a delay in its expression. Another exhibited effect of Prednisolone on the leukemic cell system CCRF-CEM was the observation of high proliferation levels. Exhibited proliferation levels were in contrast with the significant total cell death observed at low concentrations (<1 μM) in a previous study [1, 12]. Even at high Prednisolone concentrations (>100 μM, the proliferation rates were similar to those observed for untreated cells. Apoptotic death is at lower levels than that of untreated cells. The small peak in total cell death at 1 μM of Prednisolone concentration was mainly an effect of a peak in necrosis for the same concentration. Despite the increased necrosis, cell growth exceeded that corresponding to the control and higher doses due to the mitogenic effect. At higher doses (>1 μM), Prednisolone operated as a dose-dependent cytotoxic agent but still proliferation remained at the same levels as that of untreated cells. The dose dependency of mitogenic behavior was similar to that reported in several in vitro and in vivo studies for different cell types [13,14,15].

It has been originally reported that the NF-κB(p65) regulation is altered due to GR-mediated I-κB regulation [16]. NF-κB nuclear translocation, when stimulated, is expected to take place within the first 30–60 min after stimulation [17]. At the same time it has been also reported that NF-κB(p65) manifests a constitutive presence in the nucleus of lymphoblast cells. Possible reasons for the constitutive nuclear presence of NF-κB RelA might be a) the inability of the GR to inhibit NF-κB from entering the nucleus and b) the absence of the I-κBα protein in this cell system. It has been reported that the Iκ-Bα protein is degraded through the ubiquitin-proteasome pathway [18, 19]. This led to the hypothesis that proteasome plays a role in the phenotype of this cell line. Proteasome inhibition was previously reported as selectively lethal to leukemia stem cells causing a dose-dependent growth inhibition of CCRF-CEM cells [1, 12, 20]. From a previous work, it has become evident that Prednisolone works in a dual mechanism activating different GR transactivation or transrepression pathways, which hinted about a potential role of NF-κB in the CCRF-CEM cells. Bioinformatics analysis has shown that NF-κB was prevalent as a transcription factor binding motif (TFBM) in genes regulated by Prednisolone [1, 12]. There is an integrated circuit between GR, Iκ-Bα, and NF-κB reported previously as a dual mechanism of action of GCs. This means that the GR induces Iκ-Bα expression, which inhibits NF-κB transactivation, while GR inhibits NF-κB transactivation via physical interaction. Another piece of evidence is the prevalence of GR/NF-κB-related TFBMs in the analysis of the microarray data. Protein expression analysis showed that the glucocorticoid receptor (GR) did not immediately inhibit the NF-κB entry in the nucleus, directly or through the inhibitory protein I-κB [1, 12]. From previous reports we have highlighted the role of specific genes in acute lymphoblastic leukemia. In particular, AML1 [12, 21], IRF4 [12, 22], MEIS1 [12, 23], HOXA9 [12, 23], and GRIM19 [1, 12, 20] were found to play a significant role in ALL pathogenesis. We have further examined the role of these genes for their participation in resistance to GC-induced apoptosis.

2 Materials and Methods

2.1 Cell Culture

The CCRF-CEM cell line was used as the model, obtained from the European Collection of Authenticated Cell Cultures (ECACC). This cell line is a T lymphoblast leukemia cell line, known to be resistant to glucocorticoids (GC). CCRF-CEM is a human T-cell line originally isolated from a child with acute lymphoblastic leukemia [24]. Cell culture conditions have been extensively discussed previously [1, 20,21,22,23]. Twenty-four hours (24 h) before the application of Prednisolone (which is called −24 h time in the present study) cells were harvested by centrifugation at 1000 rpm for 10 min on a KUBOTA centrifuge. Cells were seeded at an initial concentration of 1.0 × 103 to 1.3 × 103 cells/μl in a final medium volume of 10 ml. Cell population counts were determined with the use of a Nihon Kohden CellTac-α hematology analyzer.

2.2 Prednisolone Treatment

Concentrations of Prednisolone treatment were selected on the basis of the average in vivo dosage administrated intravenously to children at ages between 1 month and 12 years old as previously reported [1]. Finally, the concentrations, control (0 μM), 100 nM, 100 μM, and 700 μM were chosen for further cell treatment, where 100 μM is estimated to be the mean in vivo concentration in pediatric patients between 1 month and 12 years old. Cell number was then determined at −24 h, 0 h, 1 h, 4 h, and subsequently every 24 h.

2.3 Experimental Setup

Experiments were performed in 10 cm2 cell culture flasks. Experimental setup included cell culture flasks, where (a) cells were grown and no Prednisolone was added (control or reference experiment), (b) cells were grown under 10 nM Prednisolone, (c) cells were grown under 100 μM Prednisolone, and (d) cells were grown under 700 μM Prednisolone. This experimental setup was used for the investigation of gene expression of specific genes. Another exactly similar setup was used for the implementation of the gene silencing experiments.

Cell cycle distribution and DNA content was determined with standard PI (propidium iodide, Invitrogen Inc.) staining as described previously on a Beckman Coulter flow cytometer Flow-Count XL [1, 20]. All concentrations and time point experiments consist at least of triplicate experiments.

RNA was isolated with TRIzol (Invitrogen Inc.) as described from the manufacturer. Gene expression was evaluated with qRT-PCR. Genes examined included the GRIM19 (NDUFA13) (Refseq NM_015965), AML1 transcript variant 1 (Refseq NM_001754.4), AML1 transcript variant 2 (Refseq NM_001001890.2), IRF4 (Refseq NM_001195286.2), MCL1 (Refseq NM_021960.5), MEIS1 (Refseq NM_002398.3), HOXA9 (Refseq NM_152739), GAPDH (Refseq NM_002046), and b-actin (Refseq X00351). Investigated genes were tested for three samples; control, 10 nM, 100 μM, and 700 μM Prednisolone at 1 h, 4 h, and 72 h treatment, using the one-step Plexor™ qRT-PCR kit (Promega Inc.) [25, 26]. Primer sequences are summarized in Table 1. The qRT-PCR conditions used have been described previously [1, 20,21,22,23].

Table 1 Primer sequences of investigated genes
Full size table

Standard curves were created in order to check for the method consistency gene expression accuracy. Two experiments were performed: one with the GRIM19 (Accession NM_015965) primers, using a dilution series, for the standard curve, from the control RNA (0 μM Prednisolone at 4 h) comprising six points, each in a duplicate (2 × 50 ng, 2 × 5 ng, 2 × 0.5 ng, 2 × 0.05 ng, 2 × 5 pg, 2 × 0.5 pg), and a second experiment comprising both sets of primers for GRIM19 (Accession NM_015965), b-actin (Accession X00351), and GAPDH (Accession NM_002046). Experiments were performed either in multiplexing between GRIM19 and b-actin or GAPDH or without any multiplexing. The results of the standard curves are presented in Fig. 1.

Fig. 1
figure 1

Standard curves of GRIM19, GAPDH, and b-actin genes in multiplexing or not mode. In particular, standard curve experiments included a six-point standard curve for un-multiplexed GAPDH (R2 = 0.96) (a), a six-point standard curve of b-actin (R2 = 0.96) (b), a six-point standard curve of un-multiplexed GRIM19 (R2 = 0.96) (c), three-point GRIM19 multiplexed with GAPDH (R2 = 0.98) (d), and multiplexed GRIM19 with b-actin (R2 = 0.97) (e)

Full size image

2.4 MCL1 Silencing (siRNA Experimentation)

Cells under Prednisolone treatment were further processed for MCL1 silencing. Cells were cultured in 10 cm2 flasks, up to a volume of 5 ml, and were transfected according to the instructions of the Lipofectamine®3000 (ThermoScientific, New York, CA). Thereafter, the treated cells were assigned into the following categories: Negative Control 1 (NC1) (no siRNA, no transfection reagent), Negative Control 2 (NC2) (no siRNA, with transfection reagent), Control 1 (C1) (cells transfected with MCL1 silencing sequence only), cells treated with MCL1-siRNA and 10 nM Prednisolone (Prednisolone Treatment 1 (Treat1)), cells treated with MCL1-siRNA and 100 μM Prednisolone (Prednisolone Treatment 2 (Treat2)), cells treated with MCL1-siRNA treated and 700 μM Prednisolone (Prednisolone Treatment 3 (Treat3)), and cells treated with 700 μM Prednisolone without any MCL1-siRNA (Prednisolone Treatment 4 (Predni)). The transfected cells were cultured in a 5% CO2 incubator at 37 °C. Following incubation for 6–8 h, the original medium was replaced by complete medium for another 24 h incubation for further experimentation. Relative gene expression was considered with respect to the NC1 and NC2 experiments.

2.5 Bioinformatics Analyses

Gene network was formed by interactions identified using the Coremine web toolFootnote 1. GRIM19, AML1, MEIS1, IRF4, HOXA9, MCL1, and Prednisolone were incorporated into the web tool and known interactions were identified. Genes were inserted as a model in the SimBiology Toolbox of the MATLAB® Computational Environment (the MathWorks, Inc.). Gene ontology and pathway annotation were performed with the WebGestalt web tool [27,28,29].

2.6 Statistical Analysis

Continuous data are presented as mean ± standard deviation (SD) unless otherwise stated. Flow cytometry and cell cycle data were analyzed using the algorithms proposed by Watson et al. and Ormerod et al. [30, 31]. Statistical analysis was performed using the T-test for the proliferation, cytotoxic, cell cycle, and gene expression data. Relative gene expression was estimated with the 2−Ct with respect to the expression of GAPDH and actin genes. Data pre-processing was performed with Microsoft® Excel. Statistical and data analyses have been performed using the MATLAB computational environment (the MathWorks, Inc.). All differences were considered statistically significant if they obtained a p-value of p < 0.05 unless otherwise defined. Post hoc comparisons (adjusted with Bonferroni criterion) were also performed when significant differences (p < 0.05) of the gene expression data.

2.7 Ethics Statement

No human or animal samples and/or subjects were used in the present study.

3 Results

3.1 Cell Proliferation and Apoptosis

Cell proliferation showed that cells are resistant to Prednisolone since they significantly continued to grow after 72 h as compared to 0 h and 4 h treatment (Fig. 2a). At the same time, cells manifested relatively low levels of apoptosis, yet cells under 700 μM and 100 μM manifested a late apoptotic effect at 48 h, which increased at 72 h (Fig. 2b). Apoptotic effect was significant at 48 h and 72 h as compared to 0 h of treatment.

Fig. 2
figure 2

Cell proliferation and apoptosis of CCRF-CEM cells. Cell proliferation has been measured under Prednisolone treatment (a). In addition, apoptosis has been estimated and it has been found that Prednisolone manifests a late apoptotic effect at 48 h and 72 h (b) (asterisks denote a significance at the p < 0.05 level between estimated factors at 48 h and 72 h and 0 h)

Full size image

3.2 Functional Annotation of Genes Under Investigation

Our first attempt included the investigation of our genes with respect to their functional annotation that is for their known functions as well as known participation in cellular signaling pathways. Known connections were investigated with the Coremine web tool, and we have found that examined genes, Prednisolone, and leukemia did manifest a connection (Fig. 3a). Additionally, gene ontology (GO) annotation revealed that genes participate in hematopoetic processes, as well as MCL1 and GRIM19 participate in apoptotic and metabolic processes (Fig. 3b). Similarly, pathway annotation analysis revealed that MCL1 and GRIM19 do also participate in apoptotic and metabolic pathways (Fig. 3c). These findings constituted the first hint that MCL1 and GRIM19 were of two genes of interest probably participating in resistance to GC-induced apoptosis.

Fig. 3
figure 3

Functional annotation of the genes under investigation. Results from Coremine (a), WebGestalt gene ontology (b), and WebGestalt pathway analysis (c) are presented

Full size image

3.3 Gene Expression

Cells under Prednisolone treatment were examined for the gene expression profiles of GRIM19 (Fig. 4a), AML1 (Fig. 4b), IRF4 (Fig. 4c), MCL1 (Fig. 4d), HOXA9 (Fig. 4e), and MEIS1 (Fig. 4f) at 1 h of treatment. Significant differences were manifested between concentrations of Prednisolone for all genes, except for MCL1. Almost all genes manifested significantly different expression levels within the same Prednisolone treatment (Fig. 4g).

Fig. 4
figure 4

Expression levels of AML1 (a), IRF4 (b), HOXA9 (c), MEIS1 (d), GRIM19 (e), and MCL1 (f) were estimated for the CCRF-CEM cells under Prednisolone treatment at 1 1 h. Interestingly, significant differences were manifested between all genes at different concentrations except for MCL1 (f). Similarly, significant differences were manifested between gene expression in each concentration (g) (*asterisk denotes a significance at the p < 0.05 level and **denote a significance at the p < 0.01 level)

Full size image

Interestingly, there was no significant difference in the expression levels between GRIM19 and MCL1 genes, indicating similar expression levels for these two genes. This hinted us to a possible common function between the two genes and we investigated their expression levels at 4 h which is presented in Fig. 5. It appeared that GRIM19 and MCL1 manifested very similar expression patterns, except for a marginal difference between MCL1 and GRIM19 at 100 μM and 700 μM (Fig. 4a), with respect to Prednisolone concentration at 4 h. Interestingly, GRIM19 expression manifested a linear behavior (R2 = 0.91) (Fig. 5a), and MCL1 manifested a bell-shaped regression (R2 = 0.95) indicating their biological role in GC regulation. In order to further investigate this finding, correlation analysis has shown that GRIM19 and MCL1 manifested the highest correlation coefficient with respect to each other, but also with respect to apoptosis at 72 h as compared to the other genes (Table 2). In addition, GRIM19 levels manifested significant correlation with respect to cell proliferation at 72 h (r = 0.64).

Fig. 5
figure 5

Gene expression of GRIM19 and MCL1 in cells under Prednisolone treatment at 4 h with respect to the Prednisolone concentration (a) and the genes (b)

Full size image
Table 2 Correlation analysis of GRIM19 and MCL1 with respect to each other’s expression as well as with respect to apoptosis at 4 h and 72 h (correlation coefficient (r) in bold depicts significant correlation between variables)
Full size table

We have further examined the gene expression profiles by attempting to sort gene expression levels using a hierarchical clustering (HCL) approach. In particular, we have found that apoptosis levels at 72 h were strongly correlated with MCL1 expression levels at 4 h and in addition with GRIM19 also at 4 h (Fig. 6), reinforcing our previous observation of regression results (Fig. 5a). Based on our observations we have come up with the hypothesis that MCL1 could probably play a significant role in resistance to GC-induced apoptosis, or MCL1 could be a gene directly regulated by Prednisolone, i.e., the GR. Examining the expression levels of GRIM19 and MCL1 at 72 h, we have found that both genes manifested similar patterns as in previous time points (Fig. 7).

Fig. 6
figure 6

Hierarchical clustering (HCL) of investigated genes. A correlation between apoptosis at 72 h and MCL1 and GRIM19 at 4 h was observed

Full size image
Fig. 7
figure 7

Gene expression of GRIM19 (a) and MCL1 (b) in cells under Prednisolone treatment at 1 h, 4 h, and 72 h

Full size image

In particular, GRIM19 at 72 h did manifest significant correlation with respect to MCL1 (r = 0.77) and apoptosis at 72 h (r = 0.72) (Table 3), yet MCL1 expression levels at 72 h did not manifest a significant correlation to apoptosis at 72 h (r = 0.2) (Table 3). However, when accounting for the GRIM19 and MCL1 expression levels of cells under Prednisolone treatment, excluding control samples, we have found that both GRIM19 and MCL1 manifested significant correlation to apoptosis at 72 h (r = 0.93, r = 0.99, respectively) (Table 3). This interesting finding was probably due to the observed anti-apoptosis manifested by the cells under Prednisolone treatment. Finally, it also appeared that GRIM19 expression levels at 72 h manifested a significant negative correlation with cell proliferation also at 72 h (r = −0.88).

Table 3 Correlation analysis of GRIM19 and MCL1 with respect to each other’s expression at 72 h as well as with respect to apoptosis at 1 h, 4 h, and 72 h (correlation coefficient (r) in bold depicts significant correlation between variables, as well as *depicts gene expression after excluding the control expression values for those genes)
Full size table

Our previous observations have led us to the hypothesis that MCL1 could probably play a significant role in resistance to GC-induced apoptosis and especially under Prednisolone treatment. For that reason we have attempted to silence the MCL1 gene by transfection, and we have found that indeed MCL1 is regulated by Prednisolone in a dose-dependent manner. Cells were transfected at 48 h and MCL1 levels were evaluated 24 h later.

In particular, it appeared that silenced cells for the MCL1 gene manifested a gradual decrease in MCL1 expression with respect to Prednisolone from 10 nM to 700 μM (Fig. 8a). Non-transfected cells manifested significant difference with respect to the expression levels of MCL1 in cells under 10 nM and 100 μM but not under 700 μM (Fig. 8a). Similarly, significant differences were observed between the mean expression levels of all treatments and cells treated with Prednisolone without transfection (Fig. 8b).

Interestingly, cells treated with different concentrations of Prednisolone and silenced for the MCL1 gene manifested a significant reverse correlation (r = −0.83) with respect to the observed apoptosis at 72 h.

Fig. 8
figure 8

Gene expression of MCL1 after silencing (a) and the mean expression of MCL1 with respect to the silenced and not silenced experiments (b) at 72 h (Legend: Ctrl, Control experiment, which included cells treated with the silencing MCL1 sequence and no Prednisolone; Treat1, Cells treated with 10 nM Prednisolone and the MCL1 silencing sequence; Treat2, Cells treated with 100 μM Prednisolone and the MCL1 silencing sequence; Treat3, Cells treated with 700 μM Prednisolone and the MCL1 silencing sequence; siRNA+Predni, The mean expression levels of Treat1, Treat2, and Treat3 experiments; Predni, Cells treated with 100 μM Prednisolone and no MCL1 silencing sequence)

Full size image

4 Discussion

Corticosteroids are used as therapeutics for almost over half a century. They include some of the most studied substances, especially in leukemia treatment. It is generally accepted that corticosteroids inhibit growth and induce apoptosis of immune system cells. Several studies have attempted to identify gene targets responsible for GC action as well as resistance to GC-induced apoptosis. Interesting examples include the identification of mpk-1, which appeared to sensitize cells to glucocorticoids while interference with bim expression resulted in inhibition of apoptosis [32, 33]. Elucidation of the mechanisms of GC action may lead to identification of gene targets responsible for glucocorticoid resistance. This may allow discovery of drugs, such as inhibitors of overexpressed genes, which in combination with glucocorticoids may increase the effectiveness of anti-leukemia therapies. Key tools in this process are high-throughput technologies such as microarray-based gene expression analysis.

In this study, we have observed resistance behavior consistent with these studies as well as more recent reports for CCRF-CEM cells [34, 35]. Although it is possible that these cells are clonally inhomogeneous, that is the case too with the in vivo situation in patients. Moreover, the large number of the CCRF-CEM subclone studies in the literature makes it difficult to do comparisons of different subclone behaviors and provide an absolute frame of reference in terms of a resistant cell model. Thus, we believe that the cell line used for this study is useful in simulating glucocorticoid action and resistance in leukemic cells.

In the present report we have observed two significant phenomena: the resistance of the CCRF-CEM cell line under Prednisolone treatment and the mitogenic effect Prednisolone has on these cells. Two genes were identified as interesting in this cell line and in particular MCL1 and GRIM19. To the best of our knowledge, there are no previous reports concerning the role of GRIM19 in leukemia except our previous report where we highlighted the role of GRIM19 in acute leukemia [1, 12]. On the other hand, numerous reports have investigated the role of MCL1 in acute lymphoblastic leukemia. In particular, all recent and previous studies agree on the fact that MCL1 is an anti-apoptotic gene, which when inhibited it promotes cell death and inhibits cell proliferation [36, 37]. MCL1 is a gene whose physiological function concerns the natural homeostatic response of survival against oxidative stress. At the same time, MCL1 is believed to be degraded by the proteasome, a mechanism that is accelerated by the action of GCs [38]. Our findings agreed with previous studies, which have indicated that silencing of MCL1 reduced the gene’s expression levels as well as sensitized leukemia cells to Prednisolone [39]. However, in our study MCL1 silencing did not result into cell sensitization to Prednisolone, indicating the presence of additional mechanisms of resistance to GC-induced apoptosis. This finding was in agreement with recent reports, which stated that not all MCL1-inhibiting molecules are able to sensitize leukemic cells to GCs [40]. At the same time, it has been shown that MCL1 downregulation is accompanied by BCL2 family of gene downregulation suggesting a common signaling mechanism for both genes and GC-induced apoptosis [41]. In the present study, our observed MCL1 expression was in agreement with the expected MCL1 function since we have found that MCL1 levels decreased with increasing concentration and increasing apoptosis. Interestingly, cells treated with 10 nM Prednisolone manifested the highest levels of expression as compared to the other treatments. At the same time, MCL1 silencing at 10 nM had the least effect on the gene’s levels indicating a competitive phenomenon between Prednisolone and MCL1 expression.

4.1 Conclusions

To conclude, the current study showed that GRIM19 and MCL1 were two genes whose early expression was closely related to GC’s late effect. Also, the fact that the very early expression of both genes indicated that cells probably were not inherently resistant, but they adapted to the challenging environment reacting in such a way that probably triggered a resistance mechanism. Finally, our results are a probable hint that MCL1 and GRIM19 are possible therapeutic targets for GC-resistant leukemia as suggested also by previous reports.