‘Dig in to the dark matter of the genome’: faster, more comprehensive disease research using nanopore sequencing

Steven Jones is the Co-director and Head of Bioinformatics at the Michael Smith Genome Sciences Centre, Canada. In research shared in a recent preprint, Steven and his colleagues utilised a PromethION 24 to perform whole-genome sequencing of 189 tumour and 41 matched normal research samples from the Personalized Oncogenomics (POG) program, producing data spanning genomic variants and methylation, and enabling haplotype phasing (1). The authors highlighted ‘the potential of long-read sequencing for resolving complex cancer-related structural variants, viral integrations, and extrachromosomal circular DNA’.

In a previous GenomeWeb webinar, sponsored by Oxford Nanopore Technologies (November 2023), Steven shared how he and his team are using high-output nanopore sequencing on PromethION to simultaneously study copy number variants (CNVs), complex structural variants (SVs), epigenetic changes, and chromosomal phasing in a variety of hereditary genetic diseases.

In his presentation, Stephen began by describing the analysis of a Duchenne muscular dystrophy (DMD) clinical research sample. Steven explained how a rearrangement had previously been identified within the DMD1 gene using an alternative technology; however, the significance remained unknown. Using long nanopore reads, Steven and his team were able to ‘more fully and more accurately’ characterise a duplication and its breakpoints within the gene, and ascertain that a functional, wild-type copy of the gene also existed downstream of this event in the DMD research sample. Steven was particularly impressed by the speed at which they were able to determine this.

Explaining that he and his team are also interested in using long nanopore sequencing to characterise a number of genes associated with hereditary cancer, which have ‘non-obvious or…confusing changes in genes’, Steven shared examples where the long reads enabled the team to ‘more accurately determine…the structure, composition and breakpoints’ within regions of interest, helping to determine potential pathogenic events. He shared one of their recent publications, in which he and his team suggest that SVs could be an underestimated cause of hereditary cancer (2). Steven suggested that the ability to accurately identify and characterise complex genetic aberrations in the human genome has the future potential to help guide downstream decisions for suspected hereditary cancer families.

Moving on to the field of neonatal care, Steven explained that the time taken to gain a comprehensive understanding of regions of interest within the genome is very important. Currently, a whole battery of tests is required: exome sequencing, single-gene tests, cytogenetic testing, chromosome microarray (CMA), and breakage analysis. Steven shared examples of genetic variants which are challenging to fully characterise using a single technology, such as variants in the almost identical SMN1 and SMN2 genes, which can affect the type and severity of spinal muscular atrophy. Describing how the variants are difficult to detect using short-read sequencing and require ‘bespoke’ analysis due to their high similarity, Steven revealed that ‘standard’ nanopore sequencing of a research sample enabled them to see ‘a clear predicted region of homozygous loss on chromosome 5’.

Rapid and accurate detection of potentially pathogenic changes shows the potential to help inform further assay developments to assess this locus in other cases of congenital myopathy. In another research sample, CMA had identified a rearrangement in chromosome 13, but no further characterisation was possible. Using long nanopore reads, the team identified a complex de novo SV, involving losses, triplications, and duplications; Steven noted that ‘what was nice about…nanopore is that you’re able to define the actual breakpoints’.

‘One thing that’s very exciting about having long reads is the concept of phasing’

The ability to determine which parent an allele was inherited from is very important, as this impacts the risk of disease development. Steven shared an example of the SDHD gene, where inheritance of a pathogenic allele from the father means the offspring have a very high risk of developing cancer, whereas there is no heightened risk compared to the average person if a pathogenic allele is inherited from the mother. Steven expressed how exciting it is that long nanopore reads enable variants to be segregated according to parent-of-origin, even without parental data. In the future, this could mean that trio sequencing would no longer be required, ensuring that sequencing is accessible to samples where parental data is not available, leading to fairer, more inclusive analysis, saving time, and reducing costs.

Nanopore technology enables native DNA to be sequenced directly, so that DNA modifications remain intact and can be detected alongside nucleotide sequence with no further library prep. Steven demonstrated how they had been utilising direct methylation detection, sharing their research on imprinted genes where differential methylation is expected; he presented data showing how ‘methylation clearly segregates…with the haplotype’.

‘methylation detection…through Oxford Nanopore Technology was equivalent, if not better than…bisulfite sequencing’.

Steven and his team have now published an improved map of human differential methylation, which confirmed previously detected imprinted regions and identified novel differentially methylated regions. Expressing how ‘exciting’ it is that nanopore sequencing enables the phasing of methylated regions, Steven shared an example of a Lynch syndrome research sample in which mutations in the protein coding region had not been detected in previous analyses.

Using nanopore sequencing, Steven and his team detected differential methylation between haplotypes in the promoter region of the MLH1 gene, indicating that one of the promoters was ‘broken’, suggesting a loss-of-function allele. Again, describing this as ‘quite exciting’ information, Steven explained that certain methylation signatures associate with certain diseases; they are currently investigating the potential of nanopore sequencing to profile these signatures.

‘All kinds of potential now that we have long reads and…methylation’

Steven highlighted that information available from a nanopore sequencing run *‘[allows] us to dig in to the dark matter: all the regulatory elements and all the sequences upstream of a gene or downstream of a gene that might be regulating it in cis’. For the first time, genes that were poorly understood can now be comprehensively characterised. Steven expressed his view that applying long nanopore reads and methylation detection to transcriptome analysis provides the potential to map allele-specific expression and determine the role this plays in disease.

‘The beauty of the approach is really when you can haplotype’

Steven shared examples focusing on hereditary cancer, demonstrating how phasing using long nanopore reads enabled a more comprehensive view of the genetic mechanisms involved. He described how understanding hereditary cancer is a ‘challenge’ using short reads. The two-hit hypothesis states that a tumour suppressor gene must be inactivated on both alleles; using short-read haplotype blocks, determining whether two mutations are acting in cis or trans is difficult. Steven admitted that ‘historically, we would have just guessed’, and this can have major impacts on downstream decisions. Sharing an example from a hereditary cancer research sample, Steven showed how the PTEN gene had two mutations that, previously, they ‘would have presumed they were on different chromosomes’; using long nanopore reads, they were able to precisely determine that the mutations were on the same chromosome.

Steven also shared examples of paired tumour-normal research samples where methylation of the promoter region of BRCA1 was much higher in the tumour research samples. Phasing using long nanopore reads revealed that one haplotype was methylated and the other unmethylated, indicating a ‘pathogenic event’. Steven concluded his talk by explaining that the phasing of variants is ‘very important information as we interpret genomes. If we can get [the information] easily by phasing and having the methylation phased, then I think that’s a very exciting route forward for us’.

Read the preprint: Long-read sequencing of an advanced cancer cohort resolves rearrangements, unravels haplotypes, and reveals methylation landscapes

View the presentation: Nanopore sequencing: insights from neonatal intensive care to cancer