Background

The underlying molecular defects in myelodysplastic syndromes (MDS), which are a heterogeneous group of malignant clonal hematologic disorders, are not well understood. Recently, next generation sequencing (NGS) based whole genome and exome sequencing highlighted the oligoclonal nature of persistent MDS clones that are present already at early disease stages. The reconstruction of mutational hierarchies in MDS clones and distinction of primary founder from subsequently acquired lesions has yet to be thoroughly interrogated and is likely to aid dissecting the molecular pathogenesis of MDS.

Methods

An amplicon-based NGS assay using the Roche 454 GS Junior system was established within the IRON-II framework study in order to screen for 17 commonly mutated genes in MDS. Genomic DNA from purified mononuclear bone marrow (BM) cells of 23 MDS IPSS low/int1 risk subjects was screened for somatic mutations. Called variants were compared to dbSNP and COSMIC database entries to rule out germline polymorphisms. In addition, copy number variation analysis was performed by Affymetrix SNP 6.0 array profiling. Custom pyrosequencing assays and interphase-FISH were applied for sensitive quantification of lesion burdens in FACS-sorted myeloid, erythroid, lymphoid and stem/progenitor cells. These were isolated from patients’ primary BM as well as their long-term engrafted human xenotransplants using our recently established MDS xenograft model.

Results

In this work, we identified 12 oligoclonal BM samples with ≥2 molecular lesions. Of note, varying frequencies of individual mutations between different sorted cell subsets from primary or human xenografted BM support the notion that distinct MDS (sub-)clones from these subjects contributed to hematopoiesis simultaneously and lead to differential engraftment between xenografts. Comparison of variable subset-specific mutation burdens allowed deciphering the individual hierarchical architecture of the mutational landscape from 9 individuals. ASXL1, SF3B1 and SRSF2 were detected as a primary lesion for 2 patients each. In contrast, large-scale genomic alterations such as del(5q), del(RUNX1) or trisomy 8 occurred as late-end lesion or even defined distinct clones which coexist with others harboring different mutations as detected for 2 subjects. Surprisingly, CD19+ and CD3+ lymphocytes from primary and/or xenografted BM displayed significant mutational burden of at least 1 mutation in 50% of the MDS cohort (5/10). Moreover, mutations were detected simultaneously in lymphocytes (hCD19+) as well as myeloid (hCD33+) and erythroid (hCD235a+) cells from three xenografted samples indicating a potent multilineage engraftment capability of MDS hematopoietic stem cells.

Interestingly, one individual presented with high RUNX1 mutational frequency in the primary early progenitor fraction (CD34+CD38+), which was absent in the stem-cell enriched fraction (CD34+CD38-), whereas TET2, ZRSR2 and ASXL1 mutations were detected in both fractions and their xenografts. Intriguingly, only xenotransplantation of primary CD34+38- BM cells lead to long-term engraftment of RUNX1 wild type human BM cells in mice, while CD34+CD38+ BM cells gave rise to short term engraftment of RUNX1 mutated human BM cells indicating that mutated RUNX1might originate in a more committed progenitor fraction with limited self-renewal potential.

Conclusion

Molecular characterization of oligoclonal mutation patterns in primary and xenograft BM allowed the establishment of individual mutational hierarchies and indicates a relatively random order in the mutational evolution of MDS clones, although spliceosome mutations appear as rather early events. Furthermore, our analysis revealed engraftment of independent MDS clones in different mice xenografted with the same subject material, which opens the door to the in vivo study of isolated clones with respect to their pathomechanisms and response to treatment. Our data also suggests that the occurrence of large-scale genomic aberrations is frequently preceded by small-scale gene mutations, emphasizing their potential role in disease diagnosis and risk stratification. Finally, detection of MDS specific mutations in the lymphocytic compartment might be involved in facilitating impaired immune functionality and needs to be investigated prospectively.

Disclosures:

Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Staller:MLL Munich Leukemia Laboratory: Employment. Kohlmann:MLL Munich Leukemia Laboratory: Employment; Roche Diagnostics: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution