MinION™ Nanopore Sequencing of Skin Microbiome 16S and 16S-23S rRNA Gene Amplicons

Abstract

Human skin microbiome dysbiosis can have clinical consequences. Characterizing taxonomic composition of bacterial communities associated with skin disorders is important for dermatological advancement in both diagnosis and novel treatments. This study aims to analyze and improve the accuracy of taxonomic classification of skin bacteria with MinION™ nanopore sequencing using a defined skin mock community and a skin microbiome sample. We compared the Oxford Nanopore Technologies recommended procedures and concluded that their protocols highly bias the relative abundance of certain skin microbiome genera, most notably a large overrepresentation of Staphylococcus and underrepresentation of Cutibacterium and Corynebacterium. We demonstrated that changes in the amplification protocols improved the accuracy of the taxonomic classification for these three main skin bacterial genera. This study shows that MinION™ nanopore could be an efficient technology for full-length 16S rRNA sequencing; however, the analytical advantage is strongly influenced by the methodologies. The suggested alternatives in the sample processing improved characterization of a complex skin microbiome community using MinION™ nanopore sequencing.

Keywords: 16S rRNA gene sequencing; MinION™; bacterial identification; nanopore sequencing; skin microbiome; skin mock community.

Figure 1

Testing of different amplification methodologies for MinION™ sequencing of human mock skin microbial communities. (A) Comparison of taxonomic profiles of classified reads of the mock community. The Pearson coefficient (r) between sequencing methods was computed to highlight a significant correlation between samples and/or methodologies. ns, not significant; ***P ≤ 0.001; ****P ≤ 0.0001. (B) Similarity matrix and hierarchical clustering of the methodologies based on their relative abundance profiles. (C) Heat map showing the percentage of classified reads to the correct species between the sequencing methods in the mock community.

Figure 2

Testing of different amplification methodologies for MinION™ sequencing of human skin sample microbial communities. (A) Comparison of taxonomic profiles of classified reads of the skin sample communities. The Pearson coefficient (r) between sequencing methods was computed to highlight a significant correlation between samples and/or methodologies. ns, not significant; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001. (B) Similarity matrix and hierarchical clustering of the methodologies based on their relative abundance profiles.

Figure 3

Basic linear regression analysis used to correlate the GC content (%) of mock community skin genera in sequenced samples (x-axis) compared to the number of reads in the MinION™ sequenced samples (y-axis).