Molecular characterization of identical, novel MLL-EPS15 translocation and individual genomic copy number alterations in monozygotic infant twins with acute lymphoblastic leukemia

Acute lymphoblastic leukemia (ALL) occurring in the first year of life is rare, accounting for 2–5% of pediatric ALL cases. Infant ALL is distinguished by unique clinical and biological characteristics with an aggressive course following a short latency period. The mixed lineage leukemia (MLL) gene, located on chromosome 11q23, is involved in 80% of cases. Over 70 different MLL-fusion partner genes have been molecularly characterized,1 with t(4;11), t(9;11) and t(11;19) occurring most frequently in infant ALL.

The outcome of MLL-rearranged infant ALL remains poor with up to 50% 5-year survival. The uncommon nature of ALL in infancy limits the rate of accrual for clinical trials and evidence for the best therapeutic approach has been conflicting, particularly regarding the benefit of hematopoietic stem cell transplantation. The clinical heterogeneity of ALL in infancy, such as poorer outcome under 90 days of age,2 underlies the need for stratification on these trials. The age-related difference in outcome may be a reflection of underlying molecular characteristics with differences in gene expression profiles according to age.3

Risk stratification according to MLL translocation partners has not been undertaken due to small patient numbers and as the same biological process was thought to occur irrespective of the partner gene, for MLL-mediated leukemogenesis. However, recent identification of distinct molecular differences among the MLL subtypes, including variations in epigenetic4 and gene expression profiles,3 has given rise to the notion that the many fusion partners of MLL should be considered as distinct entities.1 The complex interplay between MLL translocation and these additional molecular differences is not fully understood but further characterization of cases will contribute to the current understanding of the clinical heterogeneity of the disease.

The t(1;11) (MLL-EPS15) translocation is rare with only a small number of clinical cases published and three reports of molecular characterization at the transcriptional level.5–7 The epidermal growth factor receptor pathway substrate 15 (EPS15) gene encodes a protein that is involved in receptor-mediated endocytosis of epidermal growth factor. We present an MLL-EPS15 rearrangement with novel breakpoints at the transcriptional and DNA level in monozygotic infant twins with ALL and characterization of the molecular changes in their leukemic cells. The methods are described in the Online Supplementary Appendix. Both twins were diagnosed at seven weeks of age, with a peripheral blood blast population of 99.96×10/L in Twin One and 154.87×10/L in Twin Two. CD19, CD24, CD10, B-precursor ALL was confirmed on immunophenotyping. Banded chromosomal analysis revealed a 46,XX,t(1;11)(p32;q23)[13]/46,XX[7] karyotype in Twin One and a 46,XX,t(1;11) (p32;q23)[8]/ibid+X[4]/46,XX[8] karyotype in Twin Two, with fluorescence in situ hybridization confirming MLL involvement.8

The MLL-fusion transcripts were sequenced and novel, identical breakpoints were identified in both twins (Figure 1A). For the MLL-EPS15 transcript, the breakpoint was located at the end of exon 8 of MLL which was fused to exon 10 of EPS15. For the EPS15-MLL transcript, the breakpoint was located at the end of exon 9 of EPS15 which was fused to exon 10 of MLL. Skipping of MLL exon 9 was, therefore, evident at the transcriptional level. A novel, identical breakpoint was also identified at the DNA level (Figure 1B). The genomic breakpoint occurred within exon 9 of MLL and intron 9 of EPS15. A deletion of 10bp from exon 9 of MLL coincided with a duplicated 10bp segment from intron 9 of EPS15. These findings are consistent with and provide the first known evidence for the non-homologous end joining mechanism (NHEJ) of DNA repair occurring in the leukemic cells of infants with the t(1;11) translocation. NHEJ has previously been shown in infants with the t(4;11) translocation.9

Table 1.List of tumor-associated genomic aberrations identified using whole-genome cytogenetic arrays in monozygotic twins with t(1;11)(p32;q23) infant acute lymphoblastic leukemia.

In order to detect molecular aberrations that may be unique to each twin, the Affymetrix Cytogenetics Whole-Genome 2.7M Array technology was applied. Each twin had two copy-number alterations (CNAs) (Table 1) that were considered tumor-associated genomic aberrations. One CNA was common in both twins, namely amplification at Xp21.1, which is a gene poor region. This region has previously been identified with a 50kbp deletion in one infant with t(4;11) MLL-rearranged ALL.10 The other CNAs were different for each twin, namely a 47kbp amplification at 6p12.1 which contains the GDNF family receptor alpha like gene (GFRAL), for Twin One and a 28kbp deletion at 9q31.3, containing the muscle, skeletal, receptor tyrosine kinase (MUSK) gene, for Twin Two. Each twin had three regions of tumor-associated copy-number neutral loss of heterozygosity (Table 1), two of which were common between the twins. In addition, Twin Two had Trisomy X occurring in a subpopulation of leukemic cells, evident from both the banded chromosomal and cytogenetic array analyses.

Studies using high-resolution SNP and cytogenetic arrays in singletons with infant t(4;11) MLL-rearranged ALL10,11 revealed an exceptionally low frequency of CNAs. Our results corroborate these findings in a different cytogenetic subtype of MLL-rearranged infant ALL. Together with the concordance shown at the molecular level for the t(1;11) translocation, the identification of additional common aberrations provides further support for the concept of in utero development of a pre-leukemic clone in one twin with transfer to the other by means of the shared placental circulation.12 However, as far as the low frequency of additional aberrations points towards the t(1;11) translocation as the major driver for leukemogenesis, further work is required to establish whether these additional aberrations have a co-operating role. CNAs of the Xp21.1 site have now been demonstrated in 3 infants with MLL-rearranged ALL and so this region merits particular attention in future research. Importantly, this study has shown that several of the additional aberrations present in the leukemic cells of the twins are unique and differ in each patient, providing evidence for molecular heterogeneity for cells that have originated from a common cell of origin. The distinct molecular differences between the twins can be explained by clonal evolution of the leukemic cells from a common precursor or by selection of pre-existing clones. Further clarification of the molecular differences among different MLL-subgroups, and indeed among patients of the same subgroup, will help explain the clinical heterogeneity of the disease.

Figure 1.Molecular analysis of monozygotic twins with cytogenetically confirmed t(1;11)(p32;q23) infant acute lymphoblastic leukemia. (A) Sequence analysis showing the breakpoints at the transcriptional level. (B) Sequence analysis showing the breakpoints at the genomic level.

Acknowledgments

We thank Dr. Silvia Bungaro for providing data to allow a comparison with their previously published MLL-AF4-positive infant ALL patients. We also thank Gillian Northcott for her assistance with the illustrations.

Footnotes

• The online version of this article has a Supplementary Appendix.
• The information provided by the authors about contributions from persons listed as authors and in acknowledgments is available with the full text of this paper at www.haematologica.org.
• Financial and other disclosures provided by the authors using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are also available at www.haematologica.org.