Molecular Characterization of the Response to Conventional Chemotherapeutics in Pro-B-ALL Cell Lines in Terms of Tumor Relapse

Abstract

Little is known about optimally applying chemotherapeutic agents in a specific temporal sequence to rapidly reduce the tumor load and to improve therapeutic efficacy. The clinical optimization of drug efficacy while reducing side effects is still restricted due to an incomplete understanding of the mode of action and related tumor relapse mechanisms on the molecular level. The molecular characterization of transcriptomic drug signatures can help to identify the affected pathways, downstream regulated genes and regulatory interactions related to tumor relapse in response to drug application. We tried to outline the dynamic regulatory reprogramming leading to tumor relapse in relapsed MLL-rearranged pro-B-cell acute lymphoblastic leukemia (B-ALL) cells in response to two first-line treatments: dexamethasone (Dexa) and cytarabine (AraC). We performed an integrative molecular analysis of whole transcriptome profiles of each treatment, specifically considering public knowledge of miRNA regulation via a network-based approach to unravel key driver genes and miRNAs that may control the relapse mechanisms accompanying each treatment. Our results gave hints to the crucial regulatory roles of genes leading to Dexa-resistance and related miRNAs linked to chemosensitivity. These genes and miRNAs should be further investigated in preclinical models to obtain more hints about relapse processes.

Keywords: acute lymphoblastic leukemia; cytarabine; cytostatics; dexamethasone; drug response; tumor relapse.

Figure 1

Schematic diagram for the experimental setup to study relapse mechanisms and identify biomarkers for relapse in pro-B-ALL. Two settings were chosen: (1) drug exposure for 72 h, followed by (2) no drug exposure (“drug release”) for an additional 72 h, noted here as 144 h (+/−). In the experimental readout, trypan-blue staining (cell count), WST-1 assay (cell count and metabolism analysis) and sample collection for following RNA-Seq of RS4;11 and SEM were performed.

Figure 2

Cell count and metabolic activity in RS4;11 (a–c) and SEM (d–f) in response to AraC and Dexa for 72 h. After 72 h, drugs were released and cells cultivated for another 72 h (144 h+/−). Cell count (a,b,d,e) and WST-1 proliferation assay (c,f) were performed after 72 h and 144 h. The control cells and drug exposure times were compared with the Students’ t-test at significance levels of * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, [n ≥ 3].

Figure 3

Dissimilarity of the transcriptional response of Dexa and AraC when applied to SEM and RS4;11. (A) PCA of whole transcriptome profiles of AraC and Dexa applications in the cell lines. (B) Venn diagram showing the differentially expressed genes (DEGs) of each drug treatment in each cell line. The experiments were carried out in biological triplicates.

Figure 4

Over-represented biological processes and molecular functions for the cell lines RS4;11 (A) and for SEM (B) shown separately for the exclusive patterns for the drug releases of AraC and Dexa, as well as the common ones. The molecular functions for RS4;11 (C) and SEM (D) in response to the drug release of AraC and Dexa differentiating between common and exclusive over-representation.

Figure 5

GRNs in response to AraC and Dexa in RS4;11 and SEM. Large red nodes represent the hotspot (driver) genes, blue squares indicate hotspot miRNAs and grey nodes denote the co-regulated miRNAs/genes for (A) AraC in RS4;11, (B) AraC in SEM, (C) Dexa in RS4;11 and (D) Dexa in SEM.

Figure 6

The functional homogeneity within each drug–GRN in the two examined cell lines. For each GRN, the cumulative distribution of GO functional semantic scores of gene pairs of the GRN (red) versus randomly selected genes (black) are depicted for (A) AraC in RS4;11, (B) AraC in SEM, (C) Dexa in RS4;11 and (D) Dexa in SEM. The corresponding p-values were calculated using the Kolgomorov–Smirnov test. The GRNs were constructed with the TFmiR web service (41) for each drug treatment based on the DEG set and the corresponding enriched miRNA set, and visualized using Cytoscape V3.3.0 [45].