Transcriptome analysis offers a comprehensive illustration of the genetic background of pediatric acute myeloid leukemia

Abstract

Recent advances in the genetic understanding of acute myeloid leukemia (AML) have improved clinical outcomes in pediatric patients. However, ∼40% of patients with pediatric AML relapse, resulting in a relatively low overall survival rate of ∼70%. The objective of this study was to reveal the comprehensive genetic background of pediatric AML. We performed transcriptome analysis (RNA sequencing [RNA-seq]) in 139 of the 369 patients with de novo pediatric AML who were enrolled in the Japanese Pediatric Leukemia/Lymphoma Study Group AML-05 trial and investigated correlations between genetic aberrations and clinical information. Using RNA-seq, we identified 54 in-frame gene fusions and 1 RUNX1 out-of-frame fusion in 53 of 139 patients. Moreover, we found at least 258 gene fusions in 369 patients (70%) through reverse transcription polymerase chain reaction and RNA-seq. Five gene rearrangements were newly identified, namely, NPM1-CCDC28A, TRIP12-NPM1, MLLT10-DNAJC1, TBL1XR1-RARB, and RUNX1-FNBP1. In addition, we found rare gene rearrangements, namely, MYB-GATA1, NPM1-MLF1, ETV6-NCOA2, ETV6-MECOM, ETV6-CTNNB1, RUNX1-PRDM16, RUNX1-CBFA2T2, and RUNX1-CBFA2T3. Among the remaining 111 patients, KMT2A-PTD, biallelic CEBPA, and NPM1 gene mutations were found in 11, 23, and 17 patients, respectively. These mutations were completely mutually exclusive with any gene fusions. RNA-seq unmasked the complexity of gene rearrangements and mutations in pediatric AML. We identified potentially disease-causing alterations in nearly all patients with AML, including novel gene fusions. Our results indicated that a subset of patients with pediatric AML represent a distinct entity that may be discriminated from their adult counterparts. Based on these results, risk stratification should be reconsidered.

Graphical abstract

Figure 1.

Clinical and genetic profiles of the 369 patients enrolled in the JPLSG AML-05 trial. (A) Pie chart represented the frequency of each gene alterations. (B) Landscape of 369 de novo pediatric AML patients. Each column indicates 1 patient. *^,#Both gene fusions were observed in the same patient. RUNX1-R, RUNX1 rearrangement; NPM1-R, NPM1-rearrangement.

Figure 2.

Novel/rare fusion genes identified in patients with de novo pediatric AML. (A-D) The structure of the detected fusion proteins. (A) Five RUNX1 rearrangements. (B) Three ETV6 rearrangements. (C) Three NPM1 rearrangements. (D) Other novel gene fusions. R1, R2, and R3 are 51 ± 52 amino acid tandem repeats that comprise the DNA-binding domain Ext–leukemia-associated protein (LAP). 14-3-3 BD, 14-3-3 binding domain; AD, acidic domains; AT-hook, adenine-thymine hook; CCD, coiled-coil domain; CDC15, CDC15 homology region; CID, CBP interaction domain; CtBP, C-terminal binding protein; ETS, ETS domain; FCHD, FER-CIP1 homology domain; HLH, helix-loop-helix; ID, transcription inhibition domain; LiSH, Lis homology domain; MBD, metal-binding domain; MBR, moderately basic region; MTG16, myeloid translocation gene 16; NES, nuclear export signal; NHR2, nervy homology region 2; NLS, nuclear localization signal; PRBR, putative ρ-binding region; RD, runt domain; RHD, runt homolog domain; SH3D, Src homology 3 domain; TAD, transcription activation domain; TAFH, TAF homology; ZF, zinc finger; ZFD, zinc-finger domain; Znf-MYND, MYND zinc-finger domain.

Figure 3.

Unsupervised hierarchical clustering analysis of the 139 patients with pediatric AML. (A) Two-dimensional hierarchical clustering analysis of the 139 patients was performed using 3106 probe sets that were differentially expressed in 139 patients. (B) We showed the detail of gene expression pattern focused on the FLT3, RUNX1, HOXAs, HOXBs, and HIST1s in 139 patients. Each column represents a patient and each row represents a probe set. FAB classification, biallelic CEBPA, NPM1, WT1, RUNX1, KMT2A-PTD, KMT2A rearrangements, adverse gene fusions, PRDM16 and MECOM gene expression status, and outcome of each patient are indicated. Relative expression levels normalized to the average for each probe set are indicated by color, where red and green represent high and low expression, respectively. (C) Messenger RNA expression levels of FLT3, RUNX1, PRDM16, HOXA7, HOXB3, HOXB6, HIST1H2BD, and HIST1H1C in 139 AML patients. KMT2A-R, KMT2A rearrangement.

Figure 3.

Unsupervised hierarchical clustering analysis of the 139 patients with pediatric AML. (A) Two-dimensional hierarchical clustering analysis of the 139 patients was performed using 3106 probe sets that were differentially expressed in 139 patients. (B) We showed the detail of gene expression pattern focused on the FLT3, RUNX1, HOXAs, HOXBs, and HIST1s in 139 patients. Each column represents a patient and each row represents a probe set. FAB classification, biallelic CEBPA, NPM1, WT1, RUNX1, KMT2A-PTD, KMT2A rearrangements, adverse gene fusions, PRDM16 and MECOM gene expression status, and outcome of each patient are indicated. Relative expression levels normalized to the average for each probe set are indicated by color, where red and green represent high and low expression, respectively. (C) Messenger RNA expression levels of FLT3, RUNX1, PRDM16, HOXA7, HOXB3, HOXB6, HIST1H2BD, and HIST1H1C in 139 AML patients. KMT2A-R, KMT2A rearrangement.

Figure 4.

Correlations between established risk factors for pediatric AML. The diagram indicates co-occurrence or mutual exclusivity between pairs of clinical features, genetic alterations, gene expressions, and outcomes. The circles indicate that crossed factors are statistically co-occurrent (yellow) or exclusive (green).