Mutational pressure drives differential genome conservation in two bacterial endosymbionts of sap feeding insects

Abstract

Compared to free-living bacteria, endosymbionts of sap-feeding insects have tiny and rapidly evolving genomes. Increased genetic drift, high mutation rates, and relaxed selection associated with host control of key cellular functions all likely contribute to genome decay. Phylogenetic comparisons have revealed massive variation in endosymbiont evolutionary rate, but such methods make it difficult to partition the effects of mutation vs. selection. For example, the ancestor of auchenorrhynchan insects contained two obligate endosymbionts, Sulcia and a betaproteobacterium (BetaSymb; called Nasuia in leafhoppers) that exhibit divergent rates of sequence evolution and different propensities for loss and replacement in the ensuing ∼300 Ma. Here, we use the auchenorrhynchan leafhopper Macrosteles sp. nr. severini, which retains both of the ancestral endosymbionts, to test the hypothesis that differences in evolutionary rate are driven by differential mutagenesis. We used a high-fidelity technique known as duplex sequencing to measure and compare low-frequency variants in each endosymbiont. Our direct detection of de novo mutations reveals that the rapidly evolving endosymbiont (Nasuia) has a much higher frequency of single-nucleotide variants than the more stable endosymbiont (Sulcia) and a mutation spectrum that is even more AT-biased than implied by the 83.1% AT content of its genome. We show that indels are common in both endosymbionts but differ substantially in length and distribution around repetitive regions. Our results suggest that differences in long-term rates of sequence evolution in Sulcia vs. BetaSymb, and perhaps the contrasting degrees of stability of their relationships with the host, are driven by differences in mutagenesis.SIGNIFICANCE STATEMENTTwo ancient endosymbionts in the same host lineage display stark differences in genome conservation over phylogenetic scales. We show the rapidly evolving endosymbiont has a higher frequency of mutations, as measured with duplex sequencing. Therefore, differential mutagenesis likely drives evolutionary rate variation in these endosymbionts.

Phylogenomics of flavobacterial insect nutritional endosymbionts with implications for Auchenorrhyncha phylogeny

“…However, the dataset included only 20 Sulcia strains representing a limited number of Auchenorrhyncha hosts and the phylogeny was based on data from a single gene region. Recent genomic studies of many organisms harboring Sulcia endosymbionts (Wu et al, 2006;McCutcheon and Moran, 2007;McCutcheon et al, 2009;McCutcheon and Moran, 2010;Woyke et al, 2010;Bennett and Moran, 2013;Bennett et al, 2014;Koga and Moran, 2014;Van Leuven et al, 2014;Campbell et al, 2015;Chang et al, 2015;Bennett et al, 2016;Jia et al, 2017;Mao et al, 2017;Ankrah et al, 2018;Bennett and Mao, 2018;Matsuura et al, 2018;Shih et al, 2019;Waneka et al, 2020) provide an ideal opportunity to conduct a comprehensive phylogenomic study of Sulcia, examine its relationships to related bacterial lineages, and rigorously test the degree to which these endosymbionts have codiversified with their hosts. Our main questions were: (i) how is Sulcia related to free-living flavobacteria and various other lineages of flavobacterial insect endosymbionts?…”

Mitochondrial mutations inCaenorhabditis elegansshow signatures of oxidative damage and an AT-bias

“…Given the dominance of CG→TA transitions and CG→AT transversions in the spectrum (Figure 1a), it was not surprising that the AT-increasing count of 173 was significantly greater than the AT- decreasing count of 52 (binomial test, p << 0.0001). This difference is not driven by differences in AT vs. GC sequencing coverage, as the coverage-adjusted AT-decreasing count (normalized by the ratio of AT coverage per base pair over GC coverage per base pair, as in Waneka et al . 2021) of 61 was still significantly less than the adjusted AT-increasing count of 164 (binomial test, p << 0.0001).…”

“…2020; Broz et al . 2021; Waneka et al . 2021), some of which have shown very low relative frequencies of CG→AT transversions.…”

“…2020; Broz et al . 2021; Waneka et al . 2021), where counts of each putative mtDNA mutation are tabulated from a total-cellular shotgun DNA library.…”

“…NUMT-derived library fragments can map to the mitochondrial genome, and variants that have accumulated in NUMTs can be difficult to distinguish from true mitochondrial mutations (Hazkani-Covo et al 2010). To overcome this challenge, we have previously implemented a k-mer count-based approach (Wu et al 2020;Broz et al 2021;Waneka et al 2021), where counts of each putative mtDNA mutation are tabulated from a total-cellular shotgun DNA library. NUMT-derived variants are expected to have k-mer counts in the shotgun library that are similar to the counts for rest of the nuclear genome.…”