A cry for kelp: Evidence for polyphenolic inhibition of Oxford Nanopore sequencing of brown algae

Abstract

Genomic resources have yielded unprecedented insights into ecological and evolutionary processes, not to mention their importance in economic and conservation management of specific organisms. However, the field of macroalgal genomics is hampered by difficulties in the isolation of suitable DNA. Even when DNA that appears high quality by standard metrics has been isolated, such samples may not perform well during the sequencing process. We here have compared Oxford Nanopore long-read sequencing results for three species of macroalgae to those of nonmacroalgal species and determined that when using macroalgal samples, sequencing activity declined rapidly, resulting in reduced sequencing yield. Chemical analysis of macroalgal DNA that would be considered suitable for sequencing revealed that DNA derived from dried macroalgae was enriched for polyphenol-DNA adducts (DNA with large polyphenols chemically attached to it), which may have led to sequencing inhibition. Of note, we observed the strongest evidence of sequencing inhibition and reduced sequence output when using samples dried using silica gel-suggesting that such storage approaches may not be appropriate for samples destined for Oxford Nanopore sequencing. Our findings have wide-ranging implications for the generation of genomic resources from macroalgae and suggest a need to develop new storage methods that are more amenable to Oxford Nanopore sequencing or to use fresh flash-frozen tissue wherever possible for genome sequencing.

Keywords: DNA extraction; genome; macroalgae; nanopore; polyphenols; sequencing.

FIGURE 1

(a) Proportion of pores (of maximum available, to account for different numbers of initial pores in MinION and PromethION flow cells) available for sequencing over the course of each sequencing experiment; dashed line indicates generalized additive model (GAM) smoothing across nonmacroalgal samples; solid line indicates smoothing for macroalgal samples. (b, c) Frequency of reads terminating with A/G versus C/T for reads which terminated sequencing because of a blockage (b) and reads which terminated due to completing sequencing the read (c). Triangles represent nonmacroalgal samples, and circles represent macroalgal samples. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

(a) Overlaid UV chromatogram at 280 nm of hydrolyzed DNA for each DNA extract. Stars represent the absorbance peaks for each base, confirmed through analysis of extracted ion chromatograms. C through T is the order of the DNA bases relative to the stars. (b) Same as a, except at 260 nm. (c) Overlaid total ion chromatograms of all DNA samples; note the axis break in intensity for silica‐dried algal DNA (i.e., dried Durvillaea). (d–g) Combined mass spectra between 29 and 30 min for each sample (dried Durvillaea, flash‐frozen Durvillaea, salmon, and chicken, respectively). Boxplots below (d–g) are the hypothesized distribution of polyphenol derived DNA adducts for Durvillaea, calculated as described in the Materials and Methods section. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 3

Proposed mechanism for accumulation of DNA‐polyphenol adducts. (a) Example flavanol polyphenol undergoing oxidation; (b) reaction with guanine DNA base forming a flavanol‐derived o‐quinone DNA adduct (new covalent bond in bold).