Characterising methylation markers in imprinting disorders with direct nanopore sequencing

Prader-Willi syndrome (PWS) and Angelman syndrome (AS) are both imprinting disorders, resulting from aberrant genomic imprinting on chromosome 15q11.2. However, they present with very different phenotypes: PWS is characterised by poor feeding and decreased muscle tone in infancy, followed by later obesity, whilst the characteristics of AS include epilepsy and severe intellectual disability.

Both PWS and AS involve methylation of a CpG island within the promoter region of the gene SNRPN; the conditions are diagnosed by determining the methylation state of this region. In healthy individuals, the maternally derived allele is methylated, whilst the paternal allele is unmethylated. In subjects with PWS, both alleles are methylated; in subjects with AS,  both are unmethylated.

Currently used methods for assessing this methylation include the addition of special library preparation steps such as bisulfite treatment, followed by short-read sequencing. However, bisulfite treatment is a time-consuming process which considerably fragments DNA, whilst the use of PCR can result in bias, leading to poor resolution of areas such as GC-rich regions. Furthermore, the use of short sequencing reads can result in multimapping, especially in repetitive regions. If abnormal imprinting in the SNRPN promoter is identified, the mechanism underlying this must then be resolved — requiring further, complex tests.

‘Use of nanopore adaptive sequencing obviates the need to use the cumbersome bisulfite or methylase pre-treatment steps.¹’

In a proof-of-concept study, Yamada et al. investigated the potential of nanopore sequencing to address these difficulties¹. In contrast to traditional sequencing methods, nanopore sequencing does not require PCR, meaning that intact epigenetic modifications — such as methylation — can be directly detected in sequencing, with no need for special library preparation such as bisulfite or enzymatic treatment. The team extracted DNA from clinical research samples for which PWS or AS had been previously confirmed via fluorescent in-situ hybridisation (FISH), microsatellite marker trio analyses, and methylation-specific PCR. Preparing native DNA for sequencing using the Ligation Sequencing Kit, the team then sequenced on the GridION device.

Here, Yamada et al. made use of adaptive sampling: an entirely bioinformatics-based method of targeted sequencing which is only possible using Oxford Nanopore devices. In this method, targets to be enriched or depleted are provided via a BED file. In real-time sequencing, DNA strands identified as on-target are allowed by the software to continue sequencing, whilst off-target molecules are ejected from the nanopore, providing PCR-free, long-read targeted sequencing. The team selected the relevant region of the SNRPN gene for methylation analysis — but also decided to go further, adding regions covering the entire coding sequences of 3,601 known human disease-causing genes.

Adaptive sampling enabled the target regions — together representing 259 Mb of sequence — to be sequenced to ~17.8x depth of coverage. Structural variants (SVs), SNVs, copy number aberrations, and CpG methylation were then called across the targets from a single dataset.

The nanopore data showed that ‘virtually all of the CpGs within the differentially methylated region of the promoter region of the SNRPN locus were methylated’ in the PWS research samples whereas ‘virtually none’ were methylated in the AS samples, demonstrating the potential of the technology to clearly characterise the epigenomic markers of each disease. Furthermore, the authors described how ‘nanopore adaptive sampling technology was able to correctly delineate the underlying pathogenesis’ behind this abnormal imprinting. In two research samples — one for PWS and one for AS – nanopore sequencing revealed heterozygosity across multiple SNVs, despite previous trio microsatellite marker analysis indicating uniparental heterodisomy. In another PWS research sample, a previously missed 9 kb deletion involving a critical imprinting centre was detected by nanopore sequencing.

‘copy number analysis, homozygosity analysis, and structural variant analysis also allow one to precisely delineate the underlying pathogenic mechanisms, including gross deletion, uniparental heterodisomy, uniparental isodisomy, or imprinting defect due to even a very small deletion of the imprinting center’

Next, the team explored the potential of nanopore sequencing for the characterisation of other imprinting disorders, noting that differentiating imprinting from non-imprinting disorders through medical history and physical examination alone is difficult and were successfully able to resolve differential methylation in regions responsible for diseases including neonatal diabetes mellitus type 1, Russell-Silver syndrome, and Temple syndrome. Critically, all of this was possible to characterise within the same targeted nanopore sequencing dataset.

Concluding their research, Yamada et al. highlighted the potential of nanopore sequencing with adaptive sampling as ‘a very efficient one step assay’ for identifying the molecular markers and underlying mechanisms of PWS and AS. Beyond this, by highlighting the capacity for ‘simultaneous screening of more than 3000 known human disease-causing genes in parallel with methylation analysis’, they demonstrated how this analysis could be ‘combined with targeted gene analysis of all known human genetic disorders in a single assay’.