Nuffield Department of Medicine, University of Oxford, Oxford, UK.
NIHR Health Protection Unit in Antimicrobial Resistance and Healthcare-associated Infection, University of Oxford, Oxford, UK.
Microb Genom. 2024 Oct;10(10). doi: 10.1099/mgen.0.001296.
Whole-transcriptome (long-read) RNA sequencing (Oxford Nanopore Technologies, ONT) holds promise for reference-agnostic analysis of differential gene expression in pathogenic bacteria, including for antimicrobial resistance genes (ARGs). However, direct cDNA ONT sequencing requires large concentrations of polyadenylated mRNA, and amplification protocols may introduce technical bias. Here we evaluated the impact of direct cDNA- and cDNA PCR-based ONT sequencing on transcriptomic analysis of clinical . Four bloodstream infection-associated isolates (=2 biological replicates per isolate) were sequenced using the ONT Direct cDNA Sequencing SQK-DCS109 and PCR-cDNA Barcoding SQK-PCB111.24 kits. Biological and technical replicates were distributed over eight flow cells using 16 barcodes to minimize batch/barcoding bias. Reads were mapped to a transcript reference and transcript abundance was quantified after depletion of low-abundance and rRNA genes. We found there were strong correlations between read counts using both kits and when restricting the analysis to include only ARGs. We highlighted that correlations were weaker for genes with a higher GC content. Read lengths were longer for the direct cDNA kit compared to the PCR-cDNA kit whereas total yield was higher for the PCR-cDNA kit. In this small but methodologically rigorous evaluation of biological and technical replicates of isolates sequenced with the direct cDNA and PCR-cDNA ONT sequencing kits, we demonstrated that PCR-based amplification substantially improves yield with largely unbiased assessment of core gene and ARG expression. However, users of PCR-based kits should be aware of a small risk of technical bias which appears greater for genes with an unusually high (>52%)/low (<44%) GC content.
全长（长读）RNA 测序（Oxford Nanopore Technologies，ONT）有望实现对致病细菌差异基因表达的无参考分析，包括对抗微生物药物耐药基因（ARGs）的分析。然而，直接 cDNA ONT 测序需要大量聚腺苷酸化 mRNA，并且扩增方案可能会引入技术偏差。在这里，我们评估了直接 cDNA 和基于 cDNA 的 ONT 测序对临床 的转录组分析的影响。使用 ONT 直接 cDNA 测序 SQK-DCS109 和 cDNA PCR 条形码 SQK-PCB111.24 试剂盒对 4 株血流感染相关分离株（每个分离株=2 个生物学重复）进行测序。使用 16 个条形码将生物和技术重复分配到 8 个流动池中，以最大限度地减少批次/条形码偏差。将读数映射到转录本参考，并在去除低丰度和 rRNA 基因后定量转录本丰度。我们发现，两种试剂盒的读数之间存在很强的相关性，并且当限制分析仅包括 ARGs 时也是如此。我们强调，GC 含量较高的基因的相关性较弱。与 PCR-cDNA 试剂盒相比，直接 cDNA 试剂盒的读取长度更长，而 PCR-cDNA 试剂盒的总产量更高。在对使用直接 cDNA 和 PCR-cDNA ONT 测序试剂盒测序的分离株的生物和技术重复进行的这项小型但方法严谨的评估中，我们证明了基于 PCR 的扩增可大大提高产量，并且核心基因和 ARG 表达的评估基本无偏。然而，基于 PCR 的试剂盒的使用者应该意识到存在技术偏差的小风险，对于 GC 含量异常高（>52%）/低（<44%）的基因，这种风险似乎更大。
