AML1 mutation and its coexistence with different transcription factor gene families in de novo acute myeloid leukemia: redundancy or synergism

Abstract

AML1 mutations were identified in 6.3% of AML patients with chromosomal translocations involving CBF, PML-RARα, HOX, or ETS transcription factor (TF) gene families. Rare chromosomal abnormalities, t(16;21) and t(7;11), were also found. This study represents the first series to demonstrate the coexistence of known and novel AML1 mutations with different TF gene mutations. Although the occurrence of two TF gene mutations may appear unnecessary, the possible synergistic mechanism between different TF gene families cannot be excluded and needs to be further explored.

Different families of transcription factor genes have been functionally characterized in AML including core binding factor (CBF), retinoic acid receptor α (RARα), and members of the homeobox (HOX) and ETS gene family.1–4 Although loss of function of transcription factor leads to impaired differentiation, a single transcription factor gene mutation by itself is not sufficient to cause acute leukemia in animal models.1 The human AML1 gene encodes the major α subunit of the heterodimeric CBF complex that plays an important role in normal hematopoiesis.1,2 Translocation and non-translocation mechanism of AML1 deregulation have been reported.2, 5–6 This study evaluates whether AML1 point mutation could be an additional genetic event associated with leukemic transformation in 80 de novo AML cases with chromosomal translocations involving CBF, PML-RARα, HOX or ETS gene families. Thirty-eight were males and 42 were females with a median age of 41 years (range 15–83 years). Chromosomal abnormalities were described according to the International System for Cytogenetic Nomenclature (ISCN).7 Exons 3–5 covering the most commonly mutated region of the AML1 gene were investigated by polymerase chain reaction-single strand conformation polymorphism (PCR-SSCP) techniques.5 PCR amplifications of AML1-ETO and PML-RARα variants were also performed.8,9 FLT3 internal tandem duplication (ITD) and tyrosine kinase domain (TKD) mutations were examined in cases with mutated AML1 using our previously established protocol.10 AML1 mutations were found in 6.3%. Three novel mutations were identified including c.412_413insTTTTG, c.292delC, and c.359C→A. All AML1 mutations occurred in exon 4. 3.2% (1/31) of patients with t(8;21)/AML1-ETO and 5.4% (2/37) of patients with t(15;17)/PML-RARα cases had AML1 mutations. AML1 mutations were also found in 2 patients with rare karyotypic abnormalities, i.e. t(7;11)(p15;p15) involving HOX gene family and t(16;21)(p11;q22) involving ETS gene family as shown in Table 1. Four out of five patients carrying AML1 mutations and transcription factor fusion genes were males and three of them were over sixty years old. The patient with t(7;11)(p15;p15) and a novel AML1 mutation was a 78-year-old woman (no. 875) who also had trisomy 8 and FLT3-TKD mutation as shown in Figures 1A. Another novel AML1 mutation was found in a patient carrying AML1-ETO fusion gene (no. 597) as shown in Figure 1B. The presence of bcr1 gene variant is shown in Figure 1C in two patients with AML1 mutations. Figure 1D shows the sequencing analysis of AML1-ETO fusion gene that coexisted with AML1 mutation (c.359C→A), indicating that both mutations occurred in the same allele.

Figure 1.A. FLT3-TKD mutation in a patient (case no. 875) with t(7;11) and trisomy 8. B. AML1-ETO fusion gene in a patient (case no. 597) with co-existing AML1 mutation. C. PCR products of PML-RARα fusion gene (bcr1) in 2 patients with coexisting AML1 mutations. D) sequencing analysis of AML1-ETO fusion gene with co-existing AML1 mutation (c.359C→A) in the same allele. NC represents a negative control, PC represents a positive control, and WT represents a wild type.

Table 1.Frequency of AML1 mutation in de novo AML patients with chromosome translocations involving transcription factor genes.

A two-hit model of AML emphasizes the clear collaboration between inactivating mutations of transcription factors and mutations affecting receptor tyrosine kinases. In this study, we challenge the above model by looking for collaboration between two transcription factors from different gene families. Interestingly, despite its relatively low incidence (6.3%), AML1 mutation could be identified across all different families of transcription factor genes. This present study is unique in many aspects. It is the first to report (i) three novel AML1 mutations in non-M0 de novo AML patients, (ii) the coexistence of AML1 mutation with PML-RARα, and (iii) the coexistence of AML1 mutation with rare karyotypes involving HOX or ETS gene family including t(7;11)(p15;p15) and t(16;21)(p11;q22). Although this study represents a small and selective cohort, it is interesting to find such co-occurrence of AML1 mutation with different members of transcription factor gene families. It could be speculated that abnormalities in more than one transcription factor genes result in the activation of cellular pathways that could together initiate and propagate the leukemic transformation without help from a different class of gene mutations such as tyrosine kinases or their downstream effectors. The fact that rare karyotypes such as t(7;11)(p15;p15) and t(16;21)(11;q22) were identified in this series of Southeast Asian AML patients is also of interest. This phenomenon has been reported to be more frequent in Asian populations and results in the replacement of the transcriptional regulatory region of HOXA9 by a region of NUP98.3 In this study, AML1 mutation also occurred in 1 out of 3 patients with t(16;21)(p11;q22). The t(16;21)(p11.2;q21) leads to the production of the FUS-ERG fusion gene.4 Given that t(16;21)(p11;q22) is such a rare karyotype and only <50 cases were reported in literature, it is interesting that AML1 mutation occurred frequently in this subgroup. In conclusion, rare karyotypes and three novel AML1 mutations were identified in this study. Although an emerging paradigm of AML emphasizes the concept of collaboration between transcription factors and tyrosine kinases, the collaborative mechanism between various different transcription factor gene families may exist and requires further studies.

Acknowledgments

we thank the staff of the Division of Hematology, Department of Medicine at the Faculty of Medicine Siriraj Hospital for the excellent care given to the patients in this study

Footnotes

• Funding: CUA received the faculty award from the Siriraj Chalermprakiat Fund and the principal investigator of the leukemia project was supported by Mahidol University and the National Research Council of Thailand.