Engineering CRISPR/Cas9 to mitigate abundant host contamination for 16S rRNA gene-based amplicon sequencing

Abstract

Abstract Background High-throughput sequencing of bacterial 16S rRNA gene (16S-seq) is a useful and common method for studying bacterial community structures. However, contamination of the 16S rRNA genes from the mitochondrion and plastid hinders the sensitive bacterial 16S-seq in plant microbiota profiling, especially for some plant species such as rice. To date, efficiently mitigating such host contamination without a bias is challenging in 16S rRNA gene-based amplicon sequencing. Results We developed Cas-16S-seq method to reduce abundant host contamination for plant microbiota profiling. This method utilizes the Cas9 nuclease and specific guide RNA (gRNA) to cut 16S rRNA targets during library construction, thereby removing host contamination in 16S-seq. We used rice as an example to validate the feasibility and effectiveness of Cas-16S-seq. We established a bioinformatics pipeline to design gRNAs that specifically target rice 16S rRNA genes without bacterial 16S rRNA off-targets. We compared the effectiveness of Cas-16S-seq with that of the commonly used 16S-seq method for artificially mixed 16S rRNA gene communities, paddy soil, rice root, and phyllosphere samples. The results showed that Cas-16S-seq substantially reduces the fraction of rice 16S rRNA gene sequences from 63.2 to 2.9% in root samples and from 99.4 to 11.6% in phyllosphere samples on average. Consequently, Cas-16S-seq detected more bacterial species than the 16S-seq in plant samples. Importantly, when analyzing soil samples, Cas-16S-seq and 16S-seq showed almost identical bacterial communities, suggesting that Cas-16S-seq with host-specific gRNAs that we designed has no off-target in rice microbiota profiling. Conclusion Our Cas-16S-seq can efficiently remove abundant host contamination without a bias for 16S rRNA gene-based amplicon sequencing, thereby enabling deeper bacterial community profiling with a low cost and high flexibility. Thus, we anticipate that this method would be a useful tool for plant microbiomics.