Clonal Progression Including Evolution Of An iAMP21 Chromosome In Xenograft Models Of BCP-ALL

Insight into the role of recurrent genomic changes arising in leukemia has depended largely on the functional analysis of cell lines and transgenic animals. However not all acquired abnormalities are represented in available cell lines and a few are too complex to be faithfully engineered in animal models. One such abnormality is intrachromosomal amplification of chromosome 21 (iAMP21), a heterogeneous cytogenetic rearrangement, with a distinct clinical profile, occurring in 2% of childhood B-cell precursor acute lymphoblastic leukemia (BCP-ALL). The most highly amplified segment of iAMP21 varies in size and copy number but has always included a 5.1 Mb common region of amplification (CRA). Regions flanking the most highly amplified segment have profiles that may be step-like, or more complex combinations of lower level amplification, normal copy number and deletion. Abnormalities at other chromosomal locations such as deletions of IKZF1, CDKN2A/B or RB1 and rearrangements activating CRLF2 co-occur with iAMP21 but have never been shown to precede its formation.

To investigate clonal evolution and to provide a resource for future functional studies, we established xenografts from three cases of iAMP21 BCP-ALL in NOD.Cg-Prkdc^scidIl2rg^tm1Wjl/SzJ (NSG) mice. Signs of disease were seen between 6 and 8 months after intrafemoral injection and autopsy revealed marked splenomegaly in all cases. Cells isolated from the bone marrow and spleen, were positive for human B cell markers CD19 and CD10. Disease onset was more rapid (3-6 months) when primografts were transplanted and cells from these secondary mice were used to establish leukemia in a third generation. Whole body imaging of mice transplanted with xenograft cells, transduced with a lentivirus expressing luciferase, revealed the rapid spread of disease from the injection site to the contralateral femur, sternum, pelvis, ribs, spine and skull, with the spleen becoming the major site of engraftment at late stage disease.

Comparison of SNP 6.0 array profiles of presentation and xenografts samples revealed several examples of post-transplant genomic progression which, with the exception of a biallelic deletion of the CDKN2A/B locus, had not previously been reported in iAMP21 patients. Additional copies of chromosomes 9 and 12 in patient samples were lost in xenografts resulting in copy number neutral loss of heterozygosity (CNN-LOH) for chromosome 12 but not 9. Analysis of further primary iAMP21 BCP-ALL cases identified two with CNN-LOH for part of the long arm of chromosome 12, suggesting involvement of an imprinted locus in this region.

iAMP21 chromosomes were retained in all xenografts, supporting evidence that it is a primary abnormality. However, as these rearrangements have been considered to be stable once formed, we were surprised to observe acquisition of a non–contiguous deletion of chromosome 21 in one secondary mouse and all of its tertiary recipients. The deletion was proximal to the CRA and resulted in extension of pre-existing haploinsufficient regions by 6.6 Mb. Evolution of this iAMP21 chromosome was confirmed by fluorescence in-situ hybridisation (FISH) using probes targeting the deletion, with loss of signal observed in all cells from mice carrying the deletion but in no cells from the presentation sample or unaffected mice. Although we cannot rule out that the deletion arose in response to adaptive pressure specific to the xeno-environment, opportunities to observe similar progression in patients have been limited because FISH analysis has generally targeted only the CRA and few paired presentation/relapse samples have been analysed by genomic array. While it remains likely that overrepresentation of oncogenes within the CRA acts as a primary driver of iAMP21 BCP-ALL, the additional deletion may highlight the position of tumour suppressor genes involved in leukemia progression. As these patients are treated as a single clinical entity, an important implication of these data is that the position and extent of deletions in iAMP21 rearrangements may influence response to therapy.

In conclusion, xenografts of iAMP21 BCP-ALL closely resemble human primary disease and will make valuable pre-clinical models for basic and translational research. As models for the study of clonal evolution they support previous evidence that iAMP21 is a primary abnormality and have revealed novel forms of clonal progression.