Background: Current standard karyotyping methods are labor-intensive and have severe limitations in terms of resolution and duration to results. For hematologic malignancies, but not restricted to, this represents a relevant obstacle in translating molecular findings into time critical treatment decisions. Aims: To develop a sequencing based approach for rapid and scalable cytogenetics that can be applied to a wide spectrum of hematologic malignancies to detect clinically relevant molecular markers within 48hrs Methods: For copy number alteration detection, whole genome sequencing was performed on a GridION sequencer (Oxford Nanopore Technologies, ONT, using the SQK-LSK109 kit) in a cohort of 53 hematologic neoplasms, including acute myeloid leukemia (AML, n=18), chronic lymphocytic leukemia (CLL, n=8), multiple myeloma (MM, n=17) and pediatric acute lymphocytic leukemia (ALL, n=10) samples. For the identification of balanced alterations, transcriptome sequencing data were generated using ONT (SQK-DCS109 kit), and we applied an in-house developed analysis pipeline based on minimap2 alignment followed by Blast with an ensuing filtering algorithm based on orientation, inter-read distance and length filtering. For targeted sequencing of structural aberrations, we developed a CRISPR-Cas9 based tiling approach using sgRNA pools (Integrated DNA Technologies, IDT). This approach allows to cover AML relevant targets such as t(8;21), the KMT2A (MLL) breakpoint area, and the FLT3-ITD region as well as lymphoma relevant regions such as the immunoglobulin heavy chain locus (library preparation was performed using SQK-CS9109 kit). Results: For copy number profiling a median whole-genome coverage of 2.6 was reached (range 0.27-7 fold). ONT sequencing and conventional karyotyping approaches showed a high concordance with Pearson correlation coefficients >0.95 for copy number alteration comparisons between conventional cytogenetics and ONT sequencing results for all investigated disease entities. Regarding the detection of structural aberrations transcriptome sequencing and fusion gene analysis using the above described analysis and filtering pipeline we could e.g. reliably detect the t(9;22) translocation from the K562 cell line and the t(8;21) translocation from Kasumi-1. Additionally, using this method we were able to correctly identify the primary AML sample harboring a t(8;21) translocation that was among the n=12 AML patient samples. In addition to an RNA-based work-flow, we established a CRISPR-Cas9 based DNA enrichment approach for the detection of recurrent structural aberrations from specified genomic DNA loci. The enrichment of genomic regions of interest using sgRNA libraries led e.g. to a median coverage of 44 reads (range 5-136-fold) of the t(8;21) region-of-interest in primary AML samples, which allow the detection of the exact coordinates of the breakpoints. Summary/Conclusion: Long read based sequencing based copy number alteration and structural variation detection based on combination of different long read sequencing approaches like low-coverage whole genome sequencing, transcriptome sequencing and CRISPR-Cas9 based sequencing of genomic loci of interest represents a highly promising tool for high resolution and high speed cytogenetics that has the potential to overcome many limitations of conventional cytogenetics.
Read full abstract