Analysis of Copy Number Abnormalities of 86 Acute Lymphoblastic Leukemia Cell Lines Based on the Genetic Subtypes

Introduction: Genetic abnormalities are important to predict prognosis and sometimes can be therapeutic target in pediatric acute lymphoblastic leukemia (ALL). Although the cell lines with recurrent chromosomal abnormalities or leukemic fusions derived from ALL patients are useful tool for various in vitro experiments, it has not been fully investigated whether there is the difference of genetic alterations between clinical samples and cell lines.Here, we performed MLPA analysis of 86 ALL cell lines to determine copy number abnormalities (CNA) and compare with those of the patient's clinical samples.

Methods: We performed MLPA analysis of 86 cell lines of ALL (14 with BCR-ABL, 11 with MLL rearrangement, 18 with TCF3-PBX1, 4 with TCF3-HLF, 4 with ETV6-RUNX1 and 35 B-other ALL cell lines) to determine CNA of IKZF1, PAX5, CDKN2A, CDKN2B, ETV6, RB1, BTG1 and EBF1. Then, CNAs were compared to those of patients' samples such as UK cohort (Schwab C, et al. Haematologica, 2013) and Japanese cohort (Asai D, et al. Cancer Med, 2013) according to each specific genetic abnormality, such as BCR-ABL, ETV6-RUNX1, TCF3-PBX1 and MLL-related fusions. In addition, we performed multiplex PCR and RNA-seq to determine fusion transcripts related to Ph-like ALL for the six Ph-negative cell lines with IKZF1 deletions. To determine the expression level of IKZF1 isoform in these cell lines, we performed real time PCR analysis of IKZF1 isoform 1 (IK1) and isoform 6 (IK6).

Results: In the BCR-ABL positive cell lines, the frequencies of CDKN2A/2B and BTG1 deletion significantly higher than those in UK cohort (CDKN2A/2B: 100 vs 48%, P <0.01, BTG1; 57 vs 3%, P<0.01). In the TCF3-PBX1 or ETV6-RUNX1 positive group, the frequencies of CDKN2A/2B and BTG1 deletion were increased, but did not obtain statistically significant difference. CNAs of transcriptional factors were not significantly different between cell lines and patient's samples in each genetic subtype. However, IKZF1 deletion tended to increase in BCR-ABL positive (13/14 vs 21/33 in UK cohort, P=0.07) and MLL rearranged cell lines (3/11 vs 2/28 in UK cohort and 0/12 in Japanese cohort, P=0.13 and 0.09). Interestingly, we determined six B other ALL cell lines with IKZF1 deletion. Multiplex PCR analysis determined 5 of six cell lines expressed P2RY8-CRLF2 which was confirmed by Sanger sequence. We also determined high expression of IK1 was observed in half of the BCR-ABL positive cell lines. In contrast, the expression level of IK1 was not increased in the cell lines other than BCR-ABL positive cell lines irrespective of IKZF1 deletion.

Discussions: The current analysis revealed the high frequencies of CDKN2A/2B and BTG1 deletion in the BCR-ABL1 positive cell lines compared to those in patents' samples, suggesting that these genetic alterations contributed to the growth advantage. We also determined that CNAs of transcriptional factors were stable in cell lines. Although these findings suggest that CAN of transcriptional factors such as EBF1, PAX5 and ETV6 were less related to growth and survival of leukemic cells, the frequency of IKZF1 deletion tended to be higher in cell lines with BCR-ABL and MLL rearrangement, suggesting IKZF1 deletion might be related to survival of leukemic cells.

We also found six B-other cell lines with IKZF1 deletion and determined 5 of the six cell lines expressing P2RY8-CRLF2 fusion. Currently, we are planning comprehensive genetic analysis of these six cell lines using target sequence and RNA-seq.