Genomic landscape of TP53 -mutated myeloid malignancies

Abstract

TP53 -mutated myeloid malignancies are most frequently associated with complex cytogenetics. The presence of complex and extensive structural variants complicates detailed genomic analysis by conventional clinical techniques. We performed whole genome sequencing of 42 AML/MDS cases with paired normal tissue to characterize the genomic landscape of TP53 -mutated myeloid malignancies. The vast majority of cases had multi-hit involvement at the TP53 genetic locus (94%), as well as aneuploidy and chromothripsis. Chromosomal patterns of aneuploidy differed significantly from TP53 -mutated cancers arising in other tissues. Recurrent structural variants affected regions that include ETV6 on chr12p, RUNX1 on chr21, and NF1 on chr17q. Most notably for ETV6 , transcript expression was low in cases of TP53 -mutated myeloid malignancies both with and without structural rearrangements involving chromosome 12p. Telomeric content is increased in TP53 -mutated AML/MDS compared other AML subtypes, and telomeric content was detected adjacent to interstitial regions of chromosomes. The genomic landscape of TP53 -mutated myeloid malignancies reveals recurrent structural variants affecting key hematopoietic transcription factors and telomeric repeats that are generally not detected by panel sequencing or conventional cytogenetic analyses.

Key points: WGS comprehensively determines TP53 mutation status, resulting in the reclassification of 12% of cases from mono-allelic to multi-hit Chromothripsis is more frequent than previously appreciated, with a preference for specific chromosomes ETV6 is deleted in 45% of cases, with evidence for epigenetic suppression in non-deleted cases NF1 is mutated in 48% of cases, with multi-hit mutations in 17% of these cases TP53 -mutated AML/MDS is associated with altered telomere content compared with other AMLs.

Figure 1.

WGS results in the reclassification of some cases of TP53-mutated MDS/AML. A. Combinations of mutation types affecting TP53 in AML/MDS cases, and corresponding proportion in cohort. B. Complex structural variant affecting the TP53 genetic locus is detected using WGS based on resolution of rearrangements involving ch7, chr12, chr2, and NF1 on chr17p. Copy number losses are shown in red. Gray arcs indicate regions linked by structural variants. C. Copy neutral loss of heterozygosity affecting chr17p is detected using WGS based on shifts in B-allele frequency (top) while log2 copy number ratio remains 0 (bottom). Vertical line indicates the position of the TP53 locus.

Figure 2.

Mutational landscape of TP53-mutated myeloid malignancies. TP53-mutated AML (left) and MDS (middle), compared with AML driven by the CBFB::MYH11 or RUNX1::RUNX1T1 fusion proteins. Top: SNV/Indel counts for each sample. Left: frequency of selected genomic alterations in TP53 samples. Middle: Colors represent the types of mutations observed in each sample; asterisks indicate multiple SNV/Indel hits. Bottom: Proportion of each single-nucleotide base change per sample.

Figure 3.

Distinct patterns of copy number alterations are observed in different TP53-mutated malignancies. Y-axis indicates the proportion of samples with a copy gain (blue, positive direction) or loss (red, negative direction) overlapping each gene, genome-wide. Only ‘mostly diploid’ samples without evidence of whole-genome doubling and with at least 50% of autosomes copy neutral are included. For the PCAWG data, all tumor types with at least 20 ‘mostly diploid’, TP53-mutated samples are included. Datasets included (top to bottom): TP53-mutated AML/MDS (N=41), TP53-wild type core-binding factor AML (N=18), and TP53-mutated lymphoid (B-NHL, N=20; CLL, N=6), esophageal (N=24), hepatic (N=56), ovarian (N=23), pancreatic (N=96), and prostate (N=43) malignancies.

Figure 4.

Structural variants and chromothripsis involve very large, complex events in TP53-mutated myeloid malignancies. A. Somatic structural variant call counts for each case. Counts colored according to SV classification: breakends (BND), deletion (DEL), duplication (DUP), insertion (INS), and inversion (INV). B. Distribution of somatic structural variant cluster sizes for each case. Counts colored according to the number of individual structural variant events occurring within the corresponding complex SV cluster. C. Frequency of somatic structural variant breakpoints in 100Kb windows genome-wide. Bar height indicates the fraction of samples with at least one SV breakpoint within each window. (top) Fraction of CBF AML with somatic SV breakpoints in each window. (bottom) Fraction of TP53-mutant AML with somatic SV in each window. In core-binding factor AML, enrichment of breakpoints is clearly present at inv(16) and t(8;21) positions. Breakpoints are present throughout the genome in TP53-mutated AML/MDS cases, with enrichment on chr17 and chr21. D. Large regions of chr5, chr12, chr16, and chr17 are involved in complex rearrangements affecting copy number, including the TP53 genomic locus in a case of TP53-mutated AML (UPN 983349). (outer track) Copy gains (blue) and losses (red). (inner track) Regions involved in high-confidence chromothripsis events (dark blue). Gray arcs indicate novel adjacencies created by structural rearrangements. E. Chromothripsis is detectable in 60% of TP53-mutated myeloid malignancies and is not present in any core-binding factor AML cases. F. Genome-wide distribution of high-confidence chromothripsis calls. Proportion of all TP53-mutated AML/MDS (red; N=42) and PCAWG (blue; N=623) samples with a chromothripsis call on each chromosome.

Figure 5.

ETV6 deletion and decreased expression is common in TP53-mutated myeloid malignancies. A. Genomic locations of copy number losses intersecting the minimally deleted region on chr12p (N=19). Locations of all protein-encoding genes intersecting this region are shown. B. mRNA expression for genes within the minimally deleted region of chr12p, displayed for genes with a minimum mean expression of 10 CPM for at least one group. Boxplot indicates median and first and third quartiles; whiskers indicate the range of all data points falling within 1.5*IQR (interquartile range). CBF, core-binding factor AML (N=11); NK, normal karyotype AML (N=52); TP53-mutated with 12p intact (N=6); TP53-mutated with 12p loss (N=8). C. mRNA expression of ETV6 in TP53-mutated samples with and without 12p loss, compared to NK and CBF AML cases (N=8, 6, 52, and 11, respectively). * Adjusted p-value < 0.05, edgeR exactTestDoubleTail with adjustment for multiple comparisons. D. Extension cohort of additional cases from the previously published TCGA AML data set. mRNA expression of ETV6 in TP53-mutated samples with 12p loss and with 12p intact (N=4 and 11, respectively), NK (N=79), CBF (N=18), and 12p loss (TP53 wildtype, N=4) samples. * Adjusted p-value < 0.05, t-test with multiple testing correction.

Figure 6.

Telomere content is increased in TP53-mutated myeloid malignancies A. Telomere content in units of telomeric reads per GC content-matched million reads (TRPM) was quantified in WGS from tumor specimens (orange) and paired germline control tissues (skin or buccal DNA, blue) for core-binding factor AML (CBF) and TP53-mutated myeloid malignancies. Significant telomere shortening in the CBF sample subset, with a mean shortening of 226 TRPM (telomeric reads per GC content-matched million reads; p=8.3×10⁻⁷; 2-sided, paired t-test) but not in the TP53-mutant subset (mean difference of −2.5 TRPM between tumor and normal) B. Ratio of telomere content for tumor compared to normal tissue. TP53-mutated myeloid malignancies have significantly higher telomere content compared to CBF AML (p=1.7×10⁻⁵, 2-sided student’s t-test). C. Relative abundance of TVR repeats in singleton context. Log2 tumor/normal ratio of singleton TVR repeats in TP53-mutant and CBF subtypes. The relative abundance of singleton TTTGGG repeats is significantly higher in TP53-mutated myeloid malignancies compared to CBF AML (p=1.4×10⁻⁷, 2-sided student’s t-test). D. Example of intrachromosomal insertion of t-type telomeric hexamer sequences seen in the tumor but not the paired normal sequence data from a TP53-mutated case (UPN 387082). E. Fractions of TP53-mutated myeloid malignancies (N=42) and CBF AML patients (N=18) with detectable interstitial (non-telomere) telomeric repeat variants (p=0.006, Fisher’s exact test) F. TP53-mutated cases with detectable interstitial insertions of telomeric variant repeats have higher structural variant counts (p=0.0033, Wilcoxon ranked-sums test). G. Fraction of TP53-mutated cases with detectable interstitial insertions of telomeric variant repeats in sample with (N=25) and without (N=17) chromothripsis (p-value not significant, odds ratio=3.03, 95% CI=(0.61, 20.7), Fisher’s exact test).