Deep insights into the immune response with multiomic nanopore sequencing

Multiomic approaches are critical in gaining as much information as possible from precious clinical research samples. Traditionally, this has required the use of multiple platforms and lengthy, complex workflows. Even with numerous data sources, important information can still be missed.

In a collaboration between Benhur Lee Laboratory (Icahn School of Medicine at Mount Sinai, USA) and the Applications Team at Oxford Nanopore Technologies, multiomic nanopore sequencing was used to study the adaptive immune response. By characterising antibody diversity at both the DNA and RNA level from a single donor, the team were able to study an individual’s immune system in detail — and push the boundaries of how much data can be obtained from one research sample¹.

Unravelling the complexities of adaptive immunity

Presenting the research at the Biopharma Day following the Nanopore Community Meeting 2024 (Boston, USA), Scott Hickey (Director Genomic Applications, Oxford Nanopore Technologies) described the complex processes involved in the human adaptive immune response.

Upon encountering an unfamiliar antigen, B cells become activated, leading to their proliferation and clonal expansion. This results in the differentiation of B cells into two cell types: plasmablasts, which secrete antibodies, and memory B cells, which provide long-term immunological memory through the B cell receptor. Antibodies make up part of the adaptive immune response against invading pathogens and are encoded by the immunoglobulin (IG) loci.

Antibodies are highly diverse — a feature critical to their ability to identify a vast range of antigens — and understanding the relationship between the germline IG haplotypes and the expressed antibody repertoire could therefore provide important insights into the immune response to pathogens, autoimmune disease, and cancer. To study these processes in detail, the team performed genomic and single-cell transcriptomic nanopore sequencing of a blood research sample from an individual who had received the measles, mumps, and rubella (MMR) vaccine six days prior.

Whole-genome assembly, variant calling, methylation analysis, and haplotyping — from one sequencing run

To explore the germline IG haplotypes present, Scott and his colleagues sequenced genomic DNA from the research sample on one PromethION Flow Cell. Generating 45x depth of coverage and a read length N50 of over 20 kb, the data enabled the team to perform de novo human genome assembly.

The group then focused on the highly polymorphic IG heavy (IGH) locus, which encodes the heavy chain of antibodies. Within this locus are three key genes, V, D, and J, which encode the variable domain of the heavy chain that is responsible for antigen recognition. With nanopore reads of unrestricted length, the team were able to assemble the IGH locus in a single, fully phased contig, allowing thorough characterisation of the V, D, and J genes for a ‘super-high resolution, comprehensive look’ at the alleles present.

Mapping these reads to a reference database and calling single nucleotide polymorphisms (SNPs) in the locus allowed production of a ‘personalised’ reference genome from the research sample, featuring only the alleles present in the individual’s genome — including novel alleles.

Finally, the researchers directly detected epigenetic modifications in the native DNA sequencing data. Scott illustrated how this revealed hypomethylation signatures around the gene bodies and transcription start sites of the V genes, with similar patterns observed for the D and J genes.

Gene and isoform expression analysis at the single-cell level

Having explored the germline component of the antibody repertoire of an individual, the group then turned to single-cell transcriptome sequencing to assess expression. The team used the 10x Genomics Chromium platform to prepare cell-barcoded libraries from memory B cells and plasmablasts isolated from the same research sample; they then performed single-cell nanopore sequencing on a PromethION and analysed the data using the EPI2ME workflow wf-single-cell. Single-cell sequencing of the same samples was also performed using a legacy short-read sequencing technology for comparison purposes.

Scott shared how the short-read single-cell sequencing data showed clusters based on differential gene expression (Fig. 1A). However, with nanopore sequencing of full-length transcripts, Scott explained that identification is taken beyond the gene level, allowing differential isoform expression analysis — revealing more distinct clusters, with greater separation between them (Fig. 1B).

Bringing together multiomic approaches from one technology

Next, the team generated high-quality consensus sequences for full-length B cell receptor transcripts. B cell receptors consist of transmembrane-bound antibodies, and through nanopore sequencing, the team identified the antibody isotypes present in the memory B cells and plasmablasts — including subtypes that would not be possible to distinguish with legacy short-read sequencing.

The researchers then harnessed the data from both their single-cell transcriptomic and genomic approaches by mapping the antibody transcript sequences to their ‘personalised’ reference genome. This delivered ‘really high-quality assignment and annotation of the V gene alleles’, which were both haplotyped and isotyped.

By comparing the transcript sequences with that of the V gene alleles in the reference genome, they were able to characterise somatic hypermutation in the antibody transcripts, an important element of the adaptive immune response to an antigen. They identified a 10% frequency of somatic mutations — and which specific cells contained mutations — providing in-depth data for clonal expansion analysis.

Finally, Scott described how the team utilised the high-accuracy antibody consensus sequences to generate functional antibodies. After cloning the sequences into an expression vector, they tested the antibodies against MMR antigens. This revealed two antibodies that were active against the measles virus fusion protein, one that was active against both rubella and mumps viral antigens, and one that was ‘truly neutralising’ against the measles virus. Scott concluded that ‘this is … really incredible, that you can go just from sequencing data to actually [synthesising] an antibody that’s effective against the virus’.

What will you discover from a single research sample?

Find out more about multiomic nanopore sequencing