シングルセルナノポアシークエンスにより 生物学の新たな側面を探索

Single-cell transcriptome sequencing is a powerful tool for profiling transcriptomic heterogeneity in health and disease. However, in the past it has relied on short-read sequencing technologies, which provide little information at the transcript level due to the inherent limitations in linking distal splicing events within the same transcript. With the advent of long-read capable sequencing technologies, such as that provided by Oxford Nanopore Technologies, we are ushering in a new era of single-cell sequencing, where transcripts can be viewed at the isoform level.

Long nanopore sequencing reads facilitate the pairing of alternative splicing with genetic mutations. This was demonstrated by Dan Landau's team, based at the New York Genome Centre, who are exploring the relationship between somatic mutations in the splicing factor gene SF3B1 — which are some of the most prominent drivers in myeloid leukaemias — and their effects on the transcriptome. As a result of aberrant splicing, SF3B1 mutations manifest in disrupted haematopoietic differentiation, but exactly how this occurred was unknown. The difficulties in studying splice-altering mutations in the myeloid lineage are not least because of the array of progenitor cell types that are present during differentiation, but also because wild-type cells and SF3B1 mutant cells lack any distinguishing cell-surface markers. Also, single-cell analysis with 3’- or 5’-biased short-read sequencing lacks resolution in assessing changes at the isoform level.

To that end, the team developed their single-cell workflow, Genotyping of Transcriptomes-Splice (GoT-Splice), for the simultaneous profiling of gene expression, cell-surface protein markers, somatic mutation genotyping, and, due to the long nanopore sequencing reads generated on PromethION™, they could look at the transcript isoforms within the same single cell¹ . Importantly, they observed a four-fold increase in the number of splice junctions per cell detected using full-length nanopore sequencing, which further ‘afforded greater coverage uniformity across the entire transcript, compared to 3’-biased coverage in short-read sequencing’.

They found that SF3B1 mutations were enriched in cells committed towards erythroid progenitors, which is consistent with the literature on SF3B1-driven dyserythropoiesis phenotype. The integration of GoT with nanopore sequencing ‘showed that SF3B1 mutations exert cell-type specific mis-splicing’. That is, in the erythroid lineage, there was obvious cell-type-specific 3’ cryptic splice site usage in SF3B1 mutant cells, affecting genes related to the cell cycle¹.

On the other side of the globe, Dr Rachel Thijssen from the University of Queensland has also been using nanopore sequencing to advance her research into blood cancers, specifically chronic lymphocytic leukaemia (CLL), and the reasons underlying ventoclax drug resistance during relapse². The pharmacologic action of ventoclax is the inhibition of the pro-survival protein BCL2, culminating in apoptosis of the leukaemic cells. Whilst this delivers high remission rates, there is an eventual loss of efficacy. Thijssen et al. investigated CLL samples treated with venetoclax monotherapy using a single-cell sequencing approach, to provide an accurate picture of tumour heterogeneity. Like Landau’s team, Dr Thijssen used long nanopore reads to assess and link mutations to changes in transcriptional read-outs at the single-cell level. They found an array of previously ‘unappreciated’ disrupted splicing events that could be attributed to the ventoclax resistance phenotype, including a non-functional transcript of the pro-apoptotic NOXA gene. Her findings support the notion that ventoclax therapy should be administered in a time-limited period to prevent the development of full-blown resistance².

Both teams have demonstrated the power of linking somatic mutations to transcriptional changes at the single-cell level, and how long nanopore sequencing reads were pivotal in obtaining full-length transcript isoforms.