Mutations contribute to genetic variation in all living systems. Thus, precise estimates of mutation rates and spectra across a diversity of organisms are required for a full comprehension of evolution. Here, a mutation-accumulation (MA) assay was carried out on the endosymbiotic bacterium Teredinibacter turnerae.After 3,025 generations, base-pair substitutions (BPSs) and insertion–deletion (indel) events were characterized by whole-genome sequencing analysis of 47 independent 9 10 MA lines, yielding a BPS rate of 1.14 10 per site per generation and indel rate of 1.55 10 events per site per generation, which are among the highest within free-living and facultative intracellular bacteria. As in other endosym- bionts, a signiﬁcant bias of BPSs toward A/T and an excess of deletion mutations over insertion mutations are observed for these MA lines. However, even with a deletion bias, the genome remains relatively large (5.2 Mb) for an endosymbiotic bacterium. The estimate of the effective population size (N )in T. turnerae is quite high and comparable to free-living bacteria (4.5 10 ), suggesting that the heavy bottlenecking associated with many endosymbiotic relationships is not prevalent during the life of this endosymbiont. The efﬁciency of selection scales with increasing N and such strong selection may have been operating against the deletion bias, preventing genome erosion. The observed mutation rate in this endosymbiont is of the same order of magnitude of those with similar N , consistent with the idea that population size is a primary determinant of mutation-rate evolution within endosymbionts, and that not all endosymbionts have low N . Key words: mutation-accumulation (MA) assay, endosymbiosis, Teredinibacter turnerae, drift-barrier hypothesis. Introduction mutation-accumulation (MA) assay (Denver 2009; Lee et al. Spontaneous mutations contribute largely to the input of ge- 2012; Sung et al. 2012). Initially proposed by Mukai (1964), netic variation into living systems, and compose a major force an MA experiment uses continuous individual passages of in driving the evolutionary process. Accurate estimates of the several lineages derived from a single ancestral colony. The spontaneous rate and spectrum of mutations across a large reduction in effective population size, N ,reduces the efﬁ- number of species are required to create a better comprehen- ciency of selection on spontaneous mutations, allowing for sion of evolutionary patterns (Lynch et al. 2016). The least the accumulation of all but the most deleterious mutations. biased approach for mutation-rate estimation is the Genome-wide sequencing of multiple MA lines has generated The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Genome Biol. Evol. 10(3):723–730. doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 723 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Senra et al. GBE unbiased estimates of the rate and spectrum of spontaneous Elshahawi et al. 2013). Yet, despite its intracellular life-style, mutations across eukaryotes and bacteria (Lynch et al. 2016). this bacterium can be cultivated independent of its host Mutation rates vary 1,000-fold across species (10 – (Waterbury et al. 1983), allowing an evaluation of the mutation 10 mutations per site per generation) and the evolution pattern of this unique intracellular bacterium by MA assays. of mutation rate has been linked to the population size of the organisms (Lynch 2010a; Sung et al. 2012, 2016). Speciﬁcally, Materials and Methods the effective population size determines the power of random MA Assay genetic drift, which deﬁnes the lower limit to which selection can promote reduction of deleterious mutation rates (Lynch The T. turnerae strain used in this work (CS30) was isolated 2008, 2010b; Sung et al. 2012; Lynch et al. 2016). from Neoteredo reynei (Teredinidae) sampled in a Mangrove Although the number of MA genome-wide sequencing in Rio de Janeiro, Brazil (Trindade-Silva et al. 2009). Starting experiments has increased dramatically in the last few years, from a single cell, 100 independent MA lines were derived. these experiments have been restricted to free-living model Every 2 days, a single individual colony from each MA line was organisms. Endosymbiotic bacteria that are phylogenetically isolated and transferred to a fresh solid BMS (Trindade-Silva diverse (Moya et al. 2008), ecologically ubiquitous, and central et al. 2009) and incubated at 30 C over the course of the to host evolution (Margulis and Bermudes 1985; Schink 1997; experiment. The imposed bottlenecking process ensures that Minic and Herve 2004; Zientz et al. 2004; Croft et al. 2005; mutations accumulate in an effectively neutral fashion, as Stewart et al. 2005; Kneip et al. 2007) have not been subject demonstrated by Kibota and Lynch (1996). Generation to such assays, mostly because of methodological reasons. As time was estimated monthly using an entire colony from consequence of their long-term and intimate intracellular 12 randomly selected MA lines, transferred to 1 PBS associations, endosymbionts are thought to display elevated saline buffer, vortexed, serially diluted, and replated for mutation rates in comparison with their free-living close rela- CFU counting. The generation time for each MA line was tives (Moran 1996; Itoh et al. 2002; Woolﬁt and Bromham calculated using the harmonic mean of the cell divisions 2003) and eroded genomes (often <1 Mb coding for 300 per transfer over the course of the experiment. After an genes), with the loss of essential cell functions forcing most of average of 3,025 generations, the MA assay was con- them to rely on host cells for survival (McCutcheon and cluded and the genomic DNA of the MA lines were Moran 2011). This pattern of evolution is thought to start extracted using the wizard DNA extraction kit (Promega) just after the establishment of the host-restricted association to Illumina library standards. (Moran 1996). Such a lifestyle can force the bacterium through constant population bottlenecks (between-genera- Sequencing and Alignment tion host transmissions) and drive a reduction of the N of 101-bp paired-end Illumina (Illumina Hi-Seq platform) sequenc- the symbiont (now limited to the host abundance). Under this ing was applied to randomly selected 47 T. turnerae MA lines. view, the increased power of genetic drift (which is inversely The coverage depth of each MA line was 100 and the av- related to N ) reduces the efﬁciency of purifying selection at erage library fragment size (distance between paired end reads) removing slightly deleterious mutations, leading to a deleteri- was 175 bp. The paired-end reads from each MA line were ous pattern of genome evolution (Lynch et al. 2016) through individually mapped against the reference genome T. turnerae processes associated with Muller’s ratchet (Moran 1996). T7901 (Yang et al. 2009;available at https://www.ncbi.nlm.nih. Here, we have performed an MA experiment on the en- gov) using two different mappers: BWA v0.7.4 (Li and Durbin dosymbiotic bacterium Teredinibacter turnerae. This cellulo- 2009) and NOVOALIGN v2.08.02 (available at www.novocraft. lytic and nitrogen-ﬁxing c-proteobacterium colonizes com). The generated pileup ﬁles were converted to SAM for- specialized structures called glands of Deshayes in both demi- mat using SAMTOOLS v0.1.18 (Li 2011). We used in-house perl branchs of mollusks of the family Teredinidae (Bivalve: scripts to parse the alignments and to produce forward and Pholadoidea). Teredinibacter turnerae supplies cellulolytic reverse mapping information at each site, resulting in a conﬁg- enzymes and nitrogen compounds (Carpenter and Culliney uration of 8 numbers for each line (A,a,C,c,G, g,T,t)corre- 1975; Trytek and Allen 1980; Gallager et al. 1981; Waterbury sponding to the number of reads mapped at each genomic et al. 1983; Distel et al. 1991, 2002; Lechene et al. 2007)that position in the reference sequence. A separate ﬁle was also allow their hosts to feed on wooden material during their generated to display sites that had insertion–deletion (indels) juvenile and adulthood stages. The complete genome se- mutation calls from the two alignment algorithms. quencing of this endosymbiont has revealed that 7% of its genome is devoted to proteins involved in the biosynthesis Mutation Calling of secondary metabolites (Yang et al. 2009), including anti- biotics, suggesting that this bacterium is highly involved in Base-pair substitutions (BPSs) identiﬁcation was carried out as the synthesis of bioactive metabolites required for the previously described (Sung et al. 2015) using a consensus host survival (Trindade-Silva et al. 2009; Han et al. 2013; approach to identify putative mutations by comparing each 724 Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Mutation Rate and Spectrum in Teredinibacter turnerae GBE individual MA line (focal line) against the consensus of all the silent-sites from complete and draft genomes of T. turnerae other MA lines. Indels were identiﬁed as in Lynch (2010a)and strains publicly available in the GenBank (http://www.ncbi. brieﬂy, followed the criteria: 1) At each position, a consensus nlm.nih.gov/genbank/) as described (Sung et al. 2012). indel requires 30% of the reads spanning a position in a line to indicate the same indel (size and motif); 2) Each consensus Multinucleotide Mutations Probability Estimation indel requires a minimum of two forward and two reverse reads spanning the indel; 3) At a single site, any identical indel If we assume mutations are random with respect to their event that is shared across >50% of the lines is considered position, the chance of occurrence of one or more adjacent either a progenitor indel that existed prior to the initiation of mutations in any giving genome can be estimated. The the MA line, or a genome assembly error, and not included in expected probability of occurrence of multinucleotide muta- the ﬁnal indel list; 4) BWA and NOVOALIGN alignment algo- tions (MNMs) within window sizes of 20, 50, or 100 nt in the rithms must both independently identify the site as a putative genome of T. turnerae MA lines, is the number of observed indel. The use of two separate alignment algorithms ensures BPSs (779) divided by the genome size (5.19 10 ), and by that algorithm-speciﬁc alignment errors are not responsible the number of MA lines (47), and multiplied the correspond- for false-positive mutation calls. Short-read mapping algo- ing window sizes (20, 50, or 100 nt). The chance of occur- rithms have difﬁculties mapping indel events >10 bp, so we rence of one MNM in at least one window in a single MA line fed the BWA and NOVOALIGN alignment output into PINDEL is then (Schrider et al. 2011): (Ye et al. 2009), a short-read realignment algorithm used to 1 cdfð1; kÞ ; (4) identify indels through paired-end information. Our criteria for PINDEL for indels included the following: 1) Each indel where cdf is the cumulative distribution function for a Poisson requires a minimum of six forward and six reverse reads indi- process, k is the per-window rate, and n is the number of cating the indel, with a minimum of 20 reads overall support- windows in the genome. The probability to observe even one ing the indel call; and 2) any indel shared across >50% of the MNM in the genome of these MA lines was <0.01 for all lines is considered a progenitor indel that existed prior to the tested window lengths (only data for 50 nt shown, supple- initiation of the MA line, and excluded from the ﬁnal indel list. mentary table S4, Supplementary Material online). Mutation Rate Calculation Results The base-substitution (BPS) mutation rate per site per gener- ation (l ) for each MA line is estimated as equation (1): bs MA Assay The T. turnerae strain CS30 used in this work was originally l ¼ ; (1) bs nT isolated from a N. reynei (Teredinidae, Bivalve) host sampled in a mangrove in Rio de Janeiro, Brazil (Trindade-Silva et al. where m is the number of observed base substitutions, n is 2009). 47 MA lines of T. turnerae underwent an average of the number of nucleotide sites analyzed, and T is the total 3,025 generations of parallel single-individual passages from number of generations in the MA assay. The standard error the ancestral CS30 strain on solid basal medium with sucrose for each MA line was calculated using equation (2): (BMS) at 30 C(Trindade-Silva et al. 2009). By the end of this rﬃﬃﬃﬃﬃﬃﬃ l process, a number of MA lines were displaying smaller colony bs SE ¼ : (2) sizes, consistent with reduced ﬁtness from the accumulation nT of deleterious load (Kibota and Lynch 1996). The T. turnerae MA lines were subjected to 101-bp paired-end high-through- The total standard error of the BPS mutation rate is estimated put whole-genomic sequencing (Illumina Hi-Seq platform); by equation (3): the resulting reads were mapped against the T. turnerae SE ¼ pﬃﬃﬃﬃ ; (3) pooled T7901 (NC_012997) reference genome; and mutations were identiﬁed as previously described (Sung et al. 2015). Spontaneous-mutation data are summarized in table 1 where s represents the standard deviation of the mutation (with further details in supplementary tables S1–S3, rates across all lines, and N is the number of lines analyzed. Supplementary Material online). As in other MA experiments The same calculation was used to calculate indel mutation with other organisms (Denver et al. 2012; Lee et al. 2012; rate, with l replaced with l . bs indel Sung et al. 2012), the number of BPSs in coding regions (v ¼ 5.748, df¼ 1, P¼ 0.016) and the expected ratio of non- Effective Population Size 2 synonymous to synonymous mutations (v ¼ 3.220, df¼ 1, Under the assumption of neutrality, T. turnerae’s N was in- P¼ 0.070) are not signiﬁcantly different from random expect- directly derived from the average nucleotide heterozygosity at ations. Because a large proportion of nonsynonymous Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 725 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Senra et al. GBE Table 1 Summarized Mutation-Accumulation (MA) Assay Data 210 23 MA Gen. BPSs Indels Total Mutation BPSs Rate (310 )/ BPSs Rate (310 )/ Ts/Tv Lines (n) Events Site/Generation Genome Replication Total 47 3,025 779 106 885 — — — Ave. — — 16.57 2.26 18.83 11.4 5.59 3.48 Max — — 28 6 34 19.8 9.26 17 SEM — — 0.81 0.24 1.05 0.55 1.61 0.47 NOTE.—Ave., average; Max, maximum; SEM, standard error of the mean; Gen., number of generations; indels, insertion/deletion events; Ts/Tv, transition/transversion ratio. mutations are expected to be removed by natural selection if (during the course of this MA assay). These data are consistent effective, a random distribution of nonsynonymous and syn- with the idea that spontaneous mutations may not be inde- onymous mutations indicates that selection played a minimal pendent with respect to position (Amos 2010; Schrider et al. role on the mutation process during this MA experiment. 2011; Sung et al. 2015). Mutation Rate and Distribution of Mutations Mutation Spectrum A total of 885 spontaneous mutations accumulated within Of the total 779 BPSs, 71.6% (558/779) are transitions (sup- the genomes of the 47 tested T. turnerae MA lines, of which plementary table S1, Supplementary Material online), yielding 779 (88.1%) are BPSs and 106 (11.9%) are indels (ﬁg. 1A, a Transition/Transversion (Ts/Tv) BPS ratio of 3.48 (SEM 0.47; table 1, supplementary tables S1–S3, Supplementary Material table 1). The observed transition bias is consistent with that online). The genome-wide BPSs rate of 1.14 10 (standard observed in other MA experiments (Ochman 2003; Lynch errorofthe mean [SEM]¼ 0.55) per site per generation, or 2010b). BPSs toward A/T (ﬁg. 2A) account for 68.4% (533/ 5.59 10 (SEM¼ 1.61) mutations per genome per gener- 779) of all BPSs, which is also commonly observed in other ation, and the indel rate 1.55 10 (SEM¼ 0.16) events bacterial systems (Lynch et al. 2008; Denver 2009; Keightley per site per generation are among the highest within free et al. 2009; Hershberg and Petrov 2010; Hildebrand et al. living and facultative intracellular bacteria (ﬁg. 1B). Although 2010; Sung et al. 2012), and thought to be linked to the the BPSs are distributed randomly with respect to function, high observed rate of spontaneous deamination of cytosine the distribution of mutations with respect to chromosomal and the conversion of guanine to 8-oxo-guanine (Duncan and position is not homogeneous. Similar to Escherichia coli Miller 1980). Considering deletion mutation events, 64% (47/ (Foster et al. 2013), a peak of mutations is observed near 73) are short (1–3 bp) and 36% (26/73) long (4 bp or more; the replication terminus located at chromosome position ﬁg. 2B), including a 12,581-bp long deletion in MA line 56. 2.5 Mb (Yang et al. 2009). As shown in supplementary ﬁg- Short insertions account for 75% (25/33) of all insertions ure S1, Supplementary Material online, a 50-kb bin containing (ﬁg. 2B). Overall, there is a signiﬁcant excess of deletions the replication terminus of T. turnerae accumulated signiﬁ- (68.8%, 73/106) over insertions in this MA experiment 2 3 cantly more BPSs than the other bins elsewhere in the ge- (v ¼ 15.09, df¼ 1, P¼ 0.10 10 ), with deletion account- nome of this bacterium (one sample-t test¼ 36.50, df¼ 103, ing for a loss of 9.89 10 nucleotides per generation per P¼ 3.15 10 ; supplementary ﬁg. S1, Supplementary MA line (175.93-fold higher than the gain by insertions). Material online). Moreover, we noticed an excess of adjacent These data are consistent with a general deletion bias ob- mutations (within single MA lines) corresponding to 21 BPSs served in bacteria (Mira et al. 2001; Kuo and Ochman and 16 indels (4.2% of the total de novo mutations) arising 2009; Koskiniemi et al. 2012). within <50 nucleotides from each other (supplementary table S4, Supplementary Material online). Under the assumption Effective Population Size (N ) Estimates that mutations are randomly distributed across the genome, the probability of the occurrence of MNMs (Schrider et al. The effective population size (N ) is inversely proportional to 2011), deﬁned here as two or more mutations within win- the power of random genetic drift, which has been suggested dows of 50 nucleotides is expected to be extremely low to have an inﬂuence on the evolution of mutation rate (Lynch (P¼ 1.96 10 ). Yet, MNMs are observed in over a third 2010a; Sung et al. 2012; Lynch et al. 2016). Accordingtothe (34.0%) of the MA lines. Most clustered mutations are found drift-barrier hypothesis (DBH), the genome-wide mutation as pairs, with a single instance of three closely neighboring rate affecting ﬁtness (estimated by the total protein coding mutations in MA line 29 (supplementary table S4, nucleotides) is expected to be inversely related to N .In hap- Supplementary Material online). Our data do not allow us loid organisms and under the assumption of neutrality, N can to determine whether these MNMs arose as a single event, be estimated using the formula p ¼ 2N l ,where p is equal S e bs S but, if not, they had at least occurred in a short period of time to the variation at silent sites among natural isolates, and l is bs 726 Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Mutation Rate and Spectrum in Teredinibacter turnerae GBE FIG.1.—(A) Distribution of mutations, BPSs, and indels mapped in the 47 Teredinibacter turnerae CS30 MA lines. From the outer ring to inner ring scaled to total genome size: gene density (grey), signiﬁcantly elevated (1 kb blocks >2 standard deviations from the mean) G/C (blue) or A/T content (red), position of base substitutions in each line (black—intergenic substitution; grey—synonymous substitution; blue—nonsynonymous substitution; red—insertions; green—deletions), and base-substitution density (25 kb blocks, red > orange > yellow). Circos plot (Krzywinski et al. 2009) was used to create this ﬁgure. Please access the online version for color information on this ﬁgure. (B) Mutation rates (BPSs and indels per site per generation) across different bacterial species. The data for this analysis were extracted from Sung et al. (2016). FIG.2.—Mutation spectrum of Teredinibacter turnerae MA lines. (A) Conditional BPSs rate normalized to the number of AT or GC base pairs in the genome. (B) Indels. Short indels are deﬁned here as indels ranging from 1 to 3 bp, whereas long indels are 4 bp. the BPSs rate per site per generation. Teredinibacter turnerae’s 4.5 10 .As shown in ﬁgure 3,this N is slightly lower than p was measured by comparing silent-site diversity in draft T. comparable measures in most free-living bacteria (e.g., E. coli turnerae genomes (deposited in GenBank) according to Sung N is estimated to be on the order of 10 [Sung et al. 2016]). et al. (2012). By substituting the estimated l (1.14 10 Moreover, the mutation rate in this endosymbiont is quite bs per site per generation) in the formula, we estimate N to be similar to rates in other species with similar N (ﬁg. 3), e e Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 727 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Senra et al. GBE “Candidatus Tremblaya princeps” that has a complete ge- nome length of 139 kb coding only for 110 genes (Lopez- Madrigal et al. 2011; McCutcheon and Moran 2011). In MA lines of the endosymbiont T. turnerae, we found a rate of BPSs (1.14 10 per-site per generation) higher than those observed in most free-living and facultative intracellular pathogenic bacteria and of the same order of magnitude of other nonfree-living bacteria such as Buchnera aphidicola (4 10 per site per generation; Moran et al. 2009). The observed indel rate in the present study (1.55 10 events per site per generation) is also one of the highest among bacteria (with established indel rates based on whole- genome sequencing data generated after mutation accumu- lation assay; ﬁg. 1B). Furthermore, we observed a strong de- FIG.3.—Average total genome-wide BPSs rate in protein coding DNA 2 3 letion bias (v ¼ 15.09, df¼ 1, P¼ 0.10 10 ), with 2.2 0.59 per generation as a function of N . Trend line F(x)¼ 64.35 more spontaneous deletion events than insertion events, (r ¼ 0.693). The data for this analysis were extracted from Lynch et al. resulting in a loss of 9.83 10 nucleotides per generation (2016). Labels: Eubacteria: At, Agrobacterium tumefaciens;Bs, Bacillus per MA line. This observation suggests that genome erosion subtilis;Bc, Burkholderia cenocepacia;Ec, Escherichia coli;Hp, might be strong and have a role in T. turnerae evolution. Helicobacter pylori;Mf, Mesoplasma ﬂorum; Pa, Pseudomonas aerugi- Although T. turnerae exhibits mutational properties that nosa; Sen, Salmonella enterica; Se, Staphylococcus epidermidis;Tth, Thermus thermophila;Tt, Teredinibacter turnerae;and Vc, Vibrio cholerae. are similar to other endosymbionts (higher mutation rate Unicellular Eukaryotes: Cr, Chlamydomonas reinhardtii;Nc, Neurospora and strong deletion bias), T. turnerae retains a relatively large crassa;Pf, Plasmodium falciparum;Pt, Paramecium tetraurelia;Sc, genome (5.2 Mb), comparable in size to many free-living Saccharomyces cerevisiae;Sp, Schizosaccharomyces pombe;Tb, bacteria such as E. coli (Blattner 1997), Pseudomonas aerugi- Trypanosoma brucei. nosa (Stover et al. 2000), and Bacillus subtilis (Kunstetal. 1997). Because genomic erosion is a time dependent process, it may be suggested that this symbiotic relationship has arisen reinforcing the idea that N (Sung et al. 2012)has a large recently. Yet, fossil records reveal that this is not a recent effect in determining the mutation rate evolution in endo- association, with the family Teredinidae dating back to lower symbiontsasinother species (McCutcheon and Moran 2011). Cretaceous (100–145 million years [Myr]; Lopes et al. 2000), similar to the 150 Myr proposed for the symbiotic association Discussion between aphids and B. aphidicola that has a genome size of Endosymbiotic bacteria have a unique lifestyle of coexistence 0.6 Mb (Bennett and Moran 2014). However, population data with their hosts, and the forces driving their evolution differ show that the measured effective population size of T. turn- from those of free-living bacteria and have a profound impact erae (4.5 10 ) lies between the N measured in the free- 7 8 on their patterns of genome evolution (Wernegreen 2015). living bacteria B. subtilis (10 )and E. coli (10 )(Sung et al. From evolutionary theory, one may predict that endosym- 2012), suggesting that T. turnerae may not be subject to bionts are subject to relaxed purifying selection on metabolic heavy bottlenecking events such as those theoretically pre- pathways that are redundant in host’s genomes (Moran dicted to occur along the life history of host-restricted endo- 2003), resulting in a reductive genome evolution, and in- symbionts (Wernegreen 2015). Thus, although T. turnerae creased ﬁxation of slightly deleterious mutations by genetic exhibits a statistically signiﬁcant deletion bias, it has not expe- drift due to N reduction caused by population bottlenecks rienced the large-scale genome erosion usually found in endo- intrinsic to the host-restricted lifestyle (for more exhaustive symbionts, presumably because of more effective selection in review on this topic, see Wernegreen 2015). removing deletions from population. Our ﬁndings are consis- Conﬁrming theoretical predictions on the role of the en- tent with the expectation that effective population size has a dosymbiotic lifestyle in genome evolution, prior mutation large inﬂuence on the mutation rates (both indel and BPSs; studies using phylogenetic comparisons and recent genomic Sung et al. 2012, 2016) and genome size (Moran 2003; Kuo data from a number of insect endosymbiotic species have and Ochman 2009; McCutcheon and Moran 2011; brought to light that these organisms have rapid sequence Wernegreen 2015). evolution in comparison to closely related free-living bacteria Teredinibacter turneae is an intracellular bacterium that can (Moran 1996; Itoh et al. 2002). In addition, they often display be cultivated under in vitro conditions. Moreover, based on severely eroded genomes (Kuo and Ochman 2009), where all morphological data and the high sequence identity of the 16S but essential genes for the maintenance of the host associa- rDNA from T. turnerae isolated from many Teredinidae spe- tion are lost. An extreme example of this process is cies from unrelated geographical regions, this bacterium 728 Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Mutation Rate and Spectrum in Teredinibacter turnerae GBE Denver DR, et al. 2012. Variation in base-substitution mutation in exper- might be subject to horizontal transfer across hosts imental and natural lineages of Caenorhabditis nematodes. Genome (Waterbury et al. 1983; Distel et al. 1991). In fact, many en- Biol Evol. 4(4):513–522. dosymbiotic bacteria can be horizontally transmitted between Distel DL, DeLong EF, Waterbury JB. 1991. Phylogenetic characterization hosts (Sandstro ¨ m et al. 2001; Haselkorn et al. 2009; Russell and in situ localization of the bacterial symbiont of shipworms et al. 2009; Duron et al. 2010), and some intracellular oppor- (Teredinidae: bivalvia) by using 16S rRNA sequence analysis and oligo- deoxynucleotide probe hybridization. Appl Environ Microbiol. tunistic pathogenic bacteria such as P. aeruginosa (LaBauve 57(8):2376–2382. and Wargo 2012) are capable of surviving in the environment Distel DL, Morrill W, MacLaren-Toussaint N, Franks D, Waterbury J. 2002. as free-living organisms. Through similar mechanisms, these Teredinibacter Turnerae gen. nov., sp. nov., a dinitrogen-ﬁxing, cellu- endosymbionts might be able to maintain large N during lolytic, endosymbiotic gamma-proteobacterium isolated from the gills their life cycles. Thus, we propose that the endosymbiont T. of wood-boring molluscs (Bivalvia: teredinidae). Int J Syst Evol Microbiol. 52(Pt 6):2261–2269. turnerae is maintained in large enough populations when Duncan BK, Miller JH. 1980. Mutagenic deamination of cytosine residues transferred from one generation to another, with selection in DNA. Nature 287(5782):560–561. acting efﬁciently enough to counteract a deletion bias and Duron O, Wilkes TE, Hurst GD. 2010. Interspeciﬁc transmission of a male- to maintain mutation rates at levels that lie between those killing bacterium on an ecological timescale. Ecol Lett. of free-living bacteria and endosymbiotic bacteria that un- 13(9):1139–1148. Elshahawi SI, et al. 2013. Boronated tartrolon antibiotic produced by sym- dergo frequent bottlenecking. biotic cellulose-degrading bacteria in shipworm gills. Proc Natl Acad Sci U S A. 110(4):E295–E304. Foster PL, Hanson AJ, Lee H, Popodi EM, Tang H. 2013. On the mutational Supplementary Material topology of the bacterial genome. G3 (Bethesda) 3(3):399–407. Supplementary data areavailableat Genome Biology and Gallager SM, Turner RD, Berg CJ. 1981. Physiological aspects of wood Evolution online. consumption, growth, and reproduction in the shipworm Lyrodus pedicellatus Quatrefages (Bivalvia: teredinidae). J Exp Mar Bio Ecol. 52(1):63–77. Han AW, et al. 2013. Turnerbactin, a novel triscatecholate siderophore Acknowledgments from the shipworm endosymbiont Teredinibacter turnerae T7901. PLoS ONE. 8(10):e76151. This work was supported by Instituto Nacional de Ci^ encia e Haselkorn TS, Markow TA, Moran NA. 2009. Multiple introductions of the Tecnologia (INCT) para o Controle do Cancer; Conselho Spiroplasma bacterial endosymbiont into Drosophila. Mol Ecol. Nacional de Desenvolvimento Cientıﬁco e Tecnologico 18(6):1294–1305. (CNPq, Brazil) (484005/2013-8); the Marie Curie Actions, Hershberg R, Petrov DA. 2010. Evidence that mutation is universally biased towards AT in Bacteria. PLoS Genet. 6(9):e1001115. European Commission FP7-PEOPLE-2009-IRSES project Hildebrand F, Meyer A, Eyre-Walker A. 2010. Evidence of selection upon CINAR PATHOBACTER (project 247658); and by the genomic GC-content in Bacteria. PLoS Genet. 6(9):e1001107. Multidisciplinary University Research Initiative Award Itoh T, Martin W, Nei M. 2002. Acceleration of genomic evolution caused W911NF-09-1-0444 from the US Army Research Ofﬁce to by enhanced mutation rate in endocellular symbionts. Proc Natl Acad M.L., H. Tang, and S. Finkel. This work was also supported Sci U S A. 99(20):12944–12948. by NIH (R01-GM036827 and R35-GM122566) and by Keightley PD, et al. 2009. Analysis of the genome sequences of three Drosophila melanogaster spontaneous mutation accumulation Lines. Coordenac¸ao de Aperfeic¸oamento de Pessoal de Nıvel Genome Res. 19(7):1195–1201. Superior (CAPES, Brazil) (Post-Doctoral fellowship, “PNDP” Kibota TT, Lynch M. 1996. Estimate of the genomic mutation rate dele- to M.S. project 006995/2011-51). terious to overall ﬁtness in E. coli. Nature 381(6584):694–696. Kneip C, Lockhart P, Voß C, Maier U-G. 2007. Nitrogen ﬁxation in Eukaryotes – new models for symbiosis. BMC Evol Biol. 7(1):55. Literature Cited Koskiniemi S, Sun S, Berg OG, Andersson DI. 2012. Selection-driven gene Amos W. 2010. Even small SNP clusters are non-randomly distributed: is loss in Bacteria. PLoS Genet. 8:1–7. this evidence of mutational non-independence? Proc Roy Soc Lond B Krzywinski M, et al. 2009. Circos: an information aesthetic for comparative Biol. 277(1686):1443–1449. genomics. Genome Res. 19(9):1639–1645. Bennett G, Moran NA. 2014. Heritable symbiosis: the advantages and Kunst F, et al. 1997. The complete genome sequence of the Gram-positive perils of an evolutionary rabbit hole. Proc Natl Acad Sci U S A. bacterium Bacillus subtilis. Nature 390(6657):249–256. 112(33):10169–10176. Kuo C-H, Ochman H. 2009. Deletional bias across the three domains of Blattner FR. 1997. The complete genome sequence of Escherichia coli K- life. Genome Biol Evol. 1:145–152. 12. Science 277(5331):1453–1462. LaBauve AE, Wargo MJ. 2012. Growth and Laboratory Maintenance of Carpenter EJ, Culliney JL. 1975. Nitrogen ﬁxation in marine shipworms. Pseudomonas aeruginosa. Curr Protoc Microbiol. Unit-6E.1. Science 187(4176):551–552. Lechene CP, Luyten Y, McMahon G, Distel DL. 2007. Quantitative imaging Croft MT, Lawrence AD, Raux-Deery E, Warren MJ, Smith AG. 2005. of nitrogen ﬁxation by individual Bacteria within animal cells. Science Algae acquire vitamin B12 through a symbiotic relationship with bac- 317(5844):1563–1566. teria. Nature 438(7064):90–93. Lee H, Popodi E, Tang H, Foster PL. 2012. Rate and molecular spectrum of Denver DR. 2009. A genome-wide view of Caenorhabditis elegans base- spontaneous mutations in the bacterium Escherichia coli as deter- substitution mutation processes. Proc Natl Acad Sci U S A. mined by whole-genome sequencing. Proc Natl Acad Sci U S A. 106(38):16310–16314. 109(41):E2774–E2783. Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 729 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Senra et al. GBE Li H, Durbin R. 2009. Fast and accurate short read alignment with Sandstro ¨ m JP, Russell JA, White JP, Moran NA. 2001. Independent origins burrows-wheeler transform. Bioinformatics 25(14):1754–1760. and horizontal transfer of bacterial symbionts of aphids. Mol Ecol. Li H. 2011. A statistical framework for SNP calling, mutation discovery, 10(1):217–228. association mapping and population genetical parameter estimation Schink B. 1997. Energetics of syntrophic cooperation in methanogenic from sequencing data. Bioinformatics 27(21):2987–2993. degradation. Microbiol Mol Biol Rev. 61(2):262–280. Lopes SGBC, Domaneschi O, De Moraes DT, Morita M, Meserani GDLC. Schrider DR, Hourmozdi JN, Hahn MW. 2011. Pervasive multinucleotide 2000. Functional anatomy of the digestive system of Neoteredo reynei mutational events in Eukaryotes. Curr Biol. 21(12):1051–1054. (Bartsch, 1920) and Psiloteredo healdi (Bartsch, 1931) (Bivalvia: tere- Stewart FJ, Newton ILG, Cavanaugh CM. 2005. Chemosynthetic endo- dinidae). Geol Soc SP. 177(1):257–271. symbioses: adaptations to oxic–anoxic interfaces. Trends Microbiol. Lopez-Madrigal S, Latorre A, Porcar M, Moya A, Gil R. 2011. Complete 13(9):439–448. genome sequence of ‘Candidatus Tremblaya princeps’ strain PCVAL, Stover CK, et al. 2000. Complete genome sequence of Pseudomonas an intriguing translational machine below the living-cell status. J aeruginosa PAO1, an opportunistic pathogen. Nature Bacteriol. 193:5587–5588. 406(6799):959–964. Lynch M, et al. 2008. A genome-wide view of the spectrum of spontane- Sung W, Ackerman MS, Miller SF, Doak TG, Lynch M. 2012. Drift-barrier ous mutations in yeast. Proc Natl Acad Sci U S A. 105(27):9272–9277. hypothesis and mutation-rate evolution. Proc Natl Acad Sci U S A. Lynch M. 2010a. Evolution of the mutation rate. Trends Genet. 109(45):18488–18492. 26(8):345–352. Sung W, et al. 2015. Asymmetric context-dependent mutation patterns Lynch M. 2010b. Rate, molecular spectrum, and consequences of human revealed through mutation–accumulation experiments. Mol Biol Evol. mutation. Proc Natl Acad Sci U S A. 107(3):961–968. 32(7):1672–1683. Lynch M, et al. 2016. Genetic drift, selection and the evolution of the Sung W, et al. 2016. Evolution of the insertion–deletion mutation rate mutation rate. Nat Rev Genet. 17(11):704–714. across the tree of life. G3 (Bethesda) 6(8):2583–2591. Margulis L, Bermudes D. 1985. Symbiosis as a mechanism of evolution: Trindade-Silva AE, et al. 2009. Physiological traits of the symbiotic bacte- status of cell symbiosis theory. Symbiosis 1:101–124. rium Teredinibacter turnerae isolated from the mangrove shipworm McCutcheon JP, Moran NA. 2011. Extreme genome reduction in symbi- Neoteredo reynei. Genet Mol Biol. 32(3):572–581. otic Bacteria. Nat Rev Microbiol. 10(1):13–26. Trytek RE, Allen WV. 1980. Synthesis of essential amino acids by bacterial Minic Z, Herve G. 2004. Biochemical and enzymological aspects of the symbionts in the gills of the shipworm Bankia setacea (Tryon). Comp symbiosis between the deep-sea tubeworm Riftia pachyptila and its Biochem Physiol A Physiol. 67(3):419–427. bacterial endosymbiont. Eur J Biochem. 271(15):3093–3102. Waterbury JB, Calloway CB, Turner RD. 1983. A cellulolytic nitrogen-ﬁxing Mira A, Ochman H, Moran NA. 2001. Deletional bias and the evolution of bacterium cultured from the gland of Deshayes in shipworms (Bivalvia: bacterial genomes. Trends Genet. 17(10):589–596. teredinidae). Science 221(4618):1401–1403. Moran NA. 1996. Accelerated evolution and Muller’s ratchet in endosym- Wernegreen JJ. 2015. Endosymbiont evolution: predictions from theory biotic Bacteria. Proc Natl Acad Sci U S A. 93(7):2873–2878. and surprises from genomes. Ann NY Acad Sci. 1360(1):16–35. Moran NA. 2003. Tracing the evolution of gene loss in obligate bacterial Woolﬁt M, Bromham L. 2003. Increased rates of sequence evolution in symbionts. Curr Opin Microbiol. 6 (5):512–518. endosymbiotic bacteria and fungi with small effective population sizes. Moran NA, McLaughlin HJ, Sorek R. 2009. The dynamics and time scale of Mol Biol Evol. 20(9):1545–1555. ongoing genomic erosion in symbiotic Bacteria. Science Yang JC, et al. 2009. The complete genome of Teredinibacter turnerae 323(5912):379–382. T7901: an intracellular endosymbiont of marine wood-boring bivalves Moya A, Pereto J, Gil R, Latorre A. 2008. Learning how to live together: (shipworms). PLoS ONE. 4(7):e6085. genomic insights into Prokaryote–animal symbioses. Nat Rev Genet. Ye K, Schulz MH, Long Q, Apweiler R, Ning Z. 2009. Pindel: a pattern 9(3):218–229. growth approach to detect break points of large deletions and me- Mukai T. 1964. The genetic structure of natural populations of Drosophila dium sized insertions from paired-end short reads. Bioinformatics melanogaster. I. spontaneous mutation rate of polygenes controlling 25(21):2865–2871. viability. Genetics 50(500):1–19. Zientz E, Dandekar T, Gross R. 2004. Metabolic interdependence of obli- Ochman H. 2003. Neutral mutations and neutral substitutions in bacterial gate intracellular Bacteria and their insect hosts. Microbiol Mol Biol genomes. Mol Biol Evol. 20(12):2091–2096. Rev. 68(4):745–770. Russell JA, et al. 2009. Specialization and geographic isolation among Wolbachia symbionts from ants and lycaenid butterﬂies. Evolution 63(3):624–640. Associate editor: Howard Ochman 730 Genome Biol. Evol. 10(3):723–730 doi:10.1093/gbe/evy027 Advance Access publication February 3, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/3/723/4838064 by Ed 'DeepDyve' Gillespie user on 16 March 2018
Genome Biology and Evolution – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera