Using repeated lysis steps fractionates between heterotrophic and cyanobacterial DNA extracted from xenic cyanobacterial cultures

Abstract

Extracting DNA from cyanobacteria can be a challenge because of their diverse morphologies, challenging cellular structure, and the heterotrophic microbiome often present within cyanobacterial cultures. As such, even with high DNA yields, the percentage of reads coming from the cyanobacterial host can be low, leading to an incomplete cyanobacterial genome assembly. In this research, we optimized a DNA isolation protocol using three iterative cell lysis steps to enrich the portion of DNA isolated coming from the cyanobacterial host rather than the heterotrophic microbiome. In order to utilize in-house nanopore sequencing, we faced a challenge using our lysis protocol: the iterative lysis approach led to more DNA shearing than is ideal for this sequencing technology. To solve this, we used two bead-based size selection steps to remove shorter molecules of DNA before nanopore sequencing. Analysis of the sequenced reads showed that, in the first lysis, the cyanobacterial sequences were only 35% of all reads. In the repeated lysis steps, however, the proportion of reads coming from the cyanobacterium increased to 75% or higher. Using our iterative lysis protocol, we were able to sequence the genomes of two fresh water cyanobacteria isolated from northern Mississippi, namely Leptolyngbya sp. BL-A-14 and Limnothrix sp. BL-A-16. The genomes of these isolates were assembled as closed chromosomes of 7.2 and 4.5 Mb for BL-A-14 and BL-A-16, respectively. As it is not always possible to prepare axenic cultures of cyanobacteria, we hope our approach will be useful for sequencing other xenic cultures of cyanobacteria.

Keywords: cyanobacteria; microbiomes; whole genome sequencing; xenic cultures.

Fig. 1.

The iterative lysis workflow. The initial lysis step is incomplete, after collecting the pellet there is still DNA available to isolate in the solid cell matter. By repeating the lysis steps more vigorously, more DNA can be collected. In addition, because the cyanobacteria are generally more difficult to lyse than heterotrophic bacteria, the proportion of cyanobacterial DNA increases in each lysis fraction.

Fig. 2.

The Geneious mapping of reads from sequencing runs to the final BL-A-16 genome. The proportion of reads from fraction A to B to C mapped to the final cyanobacterial genome shows an increase with each successive fractionation. Which we argue represents the winnowing of heterotrophic DNA from the sample during the iterative lysis. Total refers to total number of reads.

Fig. 3.

The antiSMASH annotation of the BL-A-16 lim pathway. a) The final complete assembly allowed annotation of the full pathway, vs the clipped version from the draft Illumina-derived assembly. In addition to detecting four limE homologs, the complete assembly showed a transposase as was seen before in the CACIAM 69d lim cluster, Zhang Y et al. (2022) but also genes which may be involved in a type II secretion system (e.g. a type II secretion system F family protein gene that we dub limH). We hypothesize based on the adjacency of limH-L to the core biosynthetic genes that they may be responsible for excreting the final cyanobactins. AntiSMASH annotations of genes and homologs in the lim pathway (parentheticals) are given in the color-coded key. b) Comparing the predicted protein sequences for BL-A-16 limE1, limE2, and limE4 to the prior limE1 and limE3 homologs, discussed by Zhang and coworkers (Zhang Y et al. 2022), showed identical leader peptides, recognition sequences for A-family proteases (RSII) domains, and recognition sequences for G-family macrocyclases (RSIII), but different core peptide (CP) sequences. While limE3 from BL-A-16 had differences in the RSII, CP, and RSIII domains from the other homologs.