Capturing needles in haystacks: a comparison of B-cell receptor sequencing methods

Abstract

BackgroundDeep-sequencing methods are rapidly developing in the field of B-cell receptor (BCR) and T-cell receptor (TCR) diversity. These promise to revolutionise our understanding of adaptive immune dynamics, identify novel antibodies, and allow monitoring of minimal residual disease. However, different methods for BCR and TCR enrichment and amplification have been proposed. Here we perform the first systematic comparison between different methods of enrichment, amplification and sequencing for generating BCR and TCR repertoires using large sample numbers.ResultsResampling from the same RNA or cDNA pool results in highly correlated and reproducible repertoires, but resampling low frequency clones leads to stochastic variance. Repertoires generated by different sequencing methods (454 Roche and Illumina MiSeq) and amplification methods (multiplex PCR, 5’ Rapid amplification of cDNA ends (5’RACE), and RNA-capture) are highly correlated, and resulting IgHV gene frequencies between the different methods were not significantly different. Read length has an impact on captured repertoire structure, and ultimately full-length BCR sequences are most informative for repertoire analysis as diversity outside of the CDR is very useful for phylogenetic analysis. Additionally, we show RNA-based BCR repertoires are more informative than using DNA.ConclusionsRepertoires generated by different sequencing and amplification methods are consistent, but we show that read lengths, depths and error profiles should be considered in experimental design, and multiple sampling approaches could be employed to minimise stochastic sampling variation. This detailed investigation of immune repertoire sequencing methods is essential for informing basic and clinical research.Electronic supplementary materialThe online version of this article (doi:10.1186/s12865-014-0029-0) contains supplementary material, which is available to authorized users.

Eye on the B-ALL: B-cell receptor repertoires reveal persistence of numerous B-lymphoblastic leukemia subclones from diagnosis to relapse

“…Detection of very low-level B-ALL sequences in some qPCR-negative samples was reproducible in replicates of the same sample, with the exception of one sample (Supplementary Table S4, patient 1611, day 19 sample), from which clonotypic sequences were detected at low frequency (0.0656%) in only one of two repeats in keeping with stochastic ‘loss' of a rare variant. 8, 25, 26, 27 Likewise, the percentage of matching BCR sequences in DNA samples (UKALL XI) correlated with the % blasts in PB (Figure 1b). Notably, we detected clonotypic sequences in 6 of 10 patients at day 28, where blasts were not detected (<1%).…”

“…We have previously shown that there is a strong linear correlation in the frequencies of functional BCRs between RNA and DNA from the same sample. 27 Here, by studying three patients with both DNA and RNA available from the same B-ALL sample, we find that DNA-amplified B-ALL BCR sequences represented a much higher percentage of total BCR sequences than RNA-derived ones (Supplementary Figure S2), most likely due to differences in RNA expression between B-cell subsets and differences in RNA stability the B-ALL clone compared with their healthy B-cell counterparts. 30, 31 Therefore, the RNA BCR repertoire, although very powerful in detecting low-level MRD, appears to underestimate the absolute fraction of residual B-ALL cells in a sample.…”

“…MiSeq libraries were generated and reads filtered as described previously (detail in Supplementary information). 10 The network generation algorithm and network properties were calculated as in Bashford-Rogers et al : 8 each vertex represents a unique sequence, where relative vertex size is proportional to the number of identical reads. Edges join vertices that differ by single nucleotide non-indel differences and clusters are collections of related, connected vertices.…”

“…7, 8 The BCR sequence repertoire of an individual thus represents a snapshot of their B-cell population structure and can identify the presence of clonal proliferations, making it useful in the diagnosis and monitoring of B-cell malignancies. 9 Next-generation sequencing of BCR repertoires 10 can therefore facilitate the longitudinal study of the clonal dynamics of malignant B-cell populations from diagnosis to relapse.…”

Longitudinal analysis of peripheral blood T cell receptor diversity in patients with systemic lupus erythematosus by next-generation sequencing

“…Comprehensive assessment of the T cell repertoire entails knowing both the diversity (the number of unique clonotypes) and the relative abundance of these clonotypes (number of reads per clonotype), which can be obtained simultaneously by NGS using genomic DNA (gDNA) or mRNA templates. Although differences in TCRβ mRNA quantity per cell exist [42, 43], clonal rank derived from mRNA abundance correlates well with those quantified from sorted cell population [42, 44] or its corresponding gDNA [45]. Therefore, in this study we used mRNA extracted from blood to evaluate the T cell repertoire of HC and SLE subjects.…”