• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of molbiolevolLink to Publisher's site
Mol Biol Evol. Apr 2010; 27(4): 833–839.
Published online Dec 2, 2009. doi:  10.1093/molbev/msp290
PMCID: PMC2877537

Slip into Something More Functional: Selection Maintains Ancient Frameshifts in Homopolymeric Sequences


Mutational hotspots offer significant sources of genetic variability upon which selection can act. However, with a few notable exceptions, we know little about the dynamics and fitness consequences of mutations in these regions. Here, we explore evolutionary forces shaping homopolymeric tracts that are especially vulnerable to slippage errors during replication and transcription. Such tracts are typically eliminated by selection from most bacterial sequences, yet persist in genomes of endosymbionts with small effective population sizes (Ne) and biased base compositions. Focusing on Blochmannia, a bacterial endosymbiont of ants, we track the divergence of genes that contain frameshift mutations within long (9–11 bp) polyA or polyT tracts. Earlier experimental work documented that transcriptional slippage restores the reading frame in a fraction of messenger RNA molecules and thereby rescues the function of frameshifted genes. In this study, we demonstrate a surprising persistence of these frameshifts and associated tracts for millions of years. Across the genome of this ant mutualist, rates of indel mutation within homopolymeric tracts far exceed the synonymous mutation rate, indicating that long-term conservation of frameshifts within these tracts is inconsistent with neutrality. In addition, the homopolymeric tracts themselves are more conserved than expected by chance, given extensive neutral substitutions that occur elsewhere in the genes sampled. These data suggest an unexpected role for slippage-prone DNA tracts and highlight a new mechanism for their persistence. That is, when such tracts contain a frameshift, transcriptional slippage plays a critical role in rescuing gene function. In such cases, selection will purge nucleotide changes interrupting the slippery tract so that otherwise volatile sequences become frozen in evolutionary time. Although the advantage of the frameshift itself is less clear, it may offer a mechanism to lower effective gene expression by reducing but not eliminating transcripts that encode full-length proteins.

Keywords: homopolymer, selection, indel, transcriptional slippage, endosymbiont


DNA regions with intrinsically high mutation rates, or mutational hotspots, include particularly mutable dinucleotides, bases with unusual chemistry, and repeat regions prone to expansion and contraction (Benzer 1961; Rogozin and Pavlov 2003). Among these hotspots, homopolymeric tracts are exceptionally unstable, as they are prone to insertions or deletions due to misalignment of the DNA strands (Streisinger et al. 1966). By clarifying the dynamics and fitness consequences of mutations in unstable DNA motifs, we can begin to understand the evolutionary significance of these volatile regions.

Like any mutations, changes within hotspots are occasionally beneficial and fuel adaptation. For example, reversible frameshifts within homopolymers allow bacteria to toggle the expression of contingency genes (Koch 2004) and allow for phase variation of pathogen surface molecules (Bayliss 2009). However, indels in homopolymers more often have deleterious effects, and frequent frameshifts along such tracts are typically eliminated by selection (Moran et al. 2009). Genes that contain homopolymers are also vulnerable to transcriptional slippage, an enzymatic error that generates a heterogeneous pool of messenger RNA (mRNA) molecules that differ in the number of nucleotides in the tract and therefore have varied reading frames (Tamas et al. 2008). In rare cases, the translated products may represent functional, alternative proteins or subunits; however, usually “slipped” polypeptides are truncated and expected to be detrimental (Baranov et al. 2005). The deleterious consequences of frequent indels during replication and transcriptional slippage may explain the dearth of homopolymers in most bacterial genes (Baranov et al. 2005) and their nonrandom location within genes where they persist (van Passel and Ochman 2007).

Among bacteria, obligate endosymbionts of insects typically show several distinct genome features, including exceptional genome reduction, strong AT compositional bias, and rapid rates of sequence evolution (Moran et al. 2008). These apparent signs of genome degradation reflect shifts in selection, mutation, and genetic drift resulting from their host-dependent lifestyle. Their genomes also show unusually high abundance of polyA and polyT homopolymers, even compared with similarly AT-rich bacteria (Tamas et al. 2008). This persistence of slippery tracts may reflect reduced efficacy of purifying selection in bacteria with reduced Ne (Moran 1996). Small indels in these regions disrupt the reading frame and can trigger the process of gene erosion (Moran et al. 2009).

In exceptional cases, however, transcriptional slippage can provide the advantage of rescuing the functionality of genes with frameshifts in the DNA sequence. For example, in the ant mutualist, Blochmannia, five loci (hisH, ybiS, ytfM, gmhB, and ubiF) are known to contain frameshifts along homopolymeric tracts in one of the two published Blochmannia genomes (Gil et al. 2003; Degnan et al. 2005). Using experimental approaches, transcriptional slippage was shown to restore the correct reading frame in a portion of mRNA molecules for three Blochmannia loci sampled (hisH, ybiS, and ytfM) as well as frameshifted genes of the aphid mutualist, Buchnera (Tamas et al. 2008). In this earlier analysis, signatures of purifying selection at such frameshifted loci also supported the restoration of gene function (Tamas et al. 2008).

In the current study, we further explore the relationship between replication and transcription errors by tracking the evolution of homopolymers and associated frameshifts. We explore the molecular evolution of three of the five Blochmannia loci known to contain frameshifts along a polyA or polyT tract. In addition, we quantify rates of indel mutations along such tracts by analyzing polymorphisms among strains of Blochmannia pennsylvanicus, the mutualist of the ant species Camponotus pennsylvanicus. Combined, these data let us test the following hypothesis: Frameshifts can persist in coding regions when these frameshifts are rescued by transcriptional slippage. This hypothesis predicts that a frameshift can be conserved over evolutionary time, if the slippery tract is also conserved. Possible selective pressures maintaining the frameshift itself are uncertain, but might involve a favorable reduction in effective gene expression by lowering but not eliminating transcripts that encode full-length proteins.

Materials and Methods

Molecular Methods

We target three of the five Blochmannia genes known to contain frameshifts along polyA or polyT tracts. We selected two genes that we previously demonstrated undergo transcriptional slippage along a polyA tract (Tamas et al. 2008): hisH (histidine and purine biosynthesis) and ybiS (involved in attachment of peptidoglycan to the outer membrane). In addition, we sampled gmhB (involved in lipopolysaccharide biosynthesis) to explore whether similar patterns occur at frameshifts along a polyT (rather than polyA) tract.

Ant specimens included a subset of those in an earlier phylogenetic study (Degnan et al. 2004) as well as Blochmannia from Polyrhachis species and two additional C. pennsylvanicus samples (from Massachusetts and Wisconsin). Species names are listed in figure 1. Briefly, genomic DNA prepared from individual ants was used as template in polymerase chain reaction (PCR) with gene-specific primers (primer sequences available upon request). PCR products were sequenced directly using an ABI 3730xl automated sequencer (Applied Biosystems). We confirmed that long homopolymeric tract lengths were identical across independent PCR products sequenced with varying sequencing chemistry (Big Dye and dGTP). New sequences from this study are deposited in GenBank (hisH: GU214028-GU214046; gmhB: GU214047-GU214057; ybiS: GU219480-GU219486).

FIG. 1.
Divergence of frameshifted homopolymers in the ant mutualist, Blochmannia, within coding regions of (a) hisH, (b) ybiS, and (c) gmhB. Phylogenies were estimated using Bayesian analysis and match the tree estimated in earlier work (Degnan et al. 2004). ...

DNA Sequence Analysis

Sequence assemblies were curated in Consed (Gordon et al. 1998) and aligned using ClustalX (Chenna et al. 2003). Phylogenetic analyses were performed using MrBayes v3.1.2 (Ronquist and Huelsenbeck 2003) with a general time reversible model, run for 10 million generations, and 90% burnin. Orthologs from the related endosymbiont Sodalis glossinidus (NC007712) were used as outgroups.

We used PAML 4.2 (Yang 2007) to estimate pairwise nonsynonymous substitutions per nonsynonymous site (dN), synonymous substitutions per synonymous site (dS), and the ratio of these values. We also tested whether ω (or dN/dS) differs across a given gene phylogeny by implementing branch models in PAML. We calculated ω as a single ratio across all branches in the phylogeny using the M0 (one-ratio) model and allowed distinct ω values for branches with and without the frameshift using a two-ratio model. We compared the ln L of the models with a likelihood ratio test. For each gene, we performed the analysis across the open reading frame (ORF), as well as the regions upstream and downstream of the homopolymeric tract.

Testing for Conservation of the Frameshift

Estimation of Indel Mutation Rates.

Homopolymeric regions are known to experience high rates of indel mutations (Gomez-Valero et al. 2008; Moran et al. 2009). In order to confirm this pattern and to quantify indel mutation rates, we reanalyzed polymorphism data for B. pennsylvanicus. The B. pennsylvanicus genome represents a population sample consisting of endosymbiont DNA isolated from five C. pennsylvanicus colonies collected at two sites in Falmouth, MA (Degnan et al. 2005). We calculated polymorphisms from well-supported variants (≥40 Phred score) that occurred in two or more independent clones of the genome assembly, which was sequenced to 12-fold coverage (Degnan et al. 2005).

Although these data do not represent the full extent of polymorphism in this endosymbiont species, the sample provides a valuable intraspecific data set to approximate the relative rates of distinct mutation types. Specifically, we compared rates of synonymous mutations within ORFs versus rates of indel mutations within polyA (or polyT) tracts ≥5 bp long. To calculate the indel rate per tract, we divided the observed indel polymorphisms by the total number of tracts present. We performed calculations across the entire genome, as well as distinct regions including intergenic spacers where indels are most likely to reflect mutational processes. When considering ORF sequences, any overlap between adjacent ORFs was trimmed from one gene in order to represent these regions just once in the analysis.

Tracking Synonymous Divergences across Frameshifted Clades.

In light of these mutation rate data, we then assessed the extent to which indels would be expected across the divergent lineages that maintain the frameshift. (For the purposes of these sequence analyses, we consider the “frameshifted clade” of ybiS as the group of seven isolates sharing the ancient frameshift.) We quantified synonymous divergence across frameshifted clades using two approaches. First, we calculated pairwise dS values between the most divergent sequences, using PAML. These pairwise values span the ancestral node of the frameshifted group, but they do not account for the extensive change along independent branches within that group. Therefore, we also tallied synonymous changes across all branches in the frameshifted clade in PAML, by summing dS × S for all relevant branches.

Testing for Conservation of Homopolymeric Tracts

Estimating the Chance of Zero Nucleotide Changes within Tract.

To explore whether tracts are significantly conserved, we tested the null hypothesis that such tracts evolve neutrally. Based on per-site substitution rates at 4-fold degenerate A’s (or T’s) in the genes considered, we calculated the chance that the observed number of consecutive A’s (or T’s) in a given polyA (or polyT) tract did not experience a base substitution by chance alone. In more detail, we performed the following steps for hisH and ybiS, the two genes with a conserved polyA tract: 1) Using a codon-based model in PAML, we performed a marginal reconstruction of nucleotide changes across the tree and identified those changes that occurred at 4-fold degenerate sites; 2) We counted the number of 4-fold degenerate bases in each reconstructed ancestral node. Combined, these data let us calculate the per-site rate of change, or the proportion of each 4-fold degenerate base that underwent a substitution, along each individual branch. For brevity, we use P4A to refer to the proportion of 4-fold degenerate A’s that underwent a substitution across a given branch. Compared with more complex models of sequence divergence, this simple proportion will underestimate actual substitution rates, thus making our test more conservative; and 3) We estimated the chance that the observed number of consecutive A’s in the homopolymeric tract of length L did not undergo a substitution. For the purpose of this calculation, we treated substitutions across the sequence and across branches as independent events. We estimated the probability that one A did not change as (1 − P4A), and the probability that L number of A’s did not change as (1 − P4A)L. This value applies to a single branch. We then multiplied these values across relevant branches to estimate the probability that the observed polyA tract did not change throughout the divergence of the clade. The same analysis as performed for the polyT tract of gmhB by estimating P4T, etc.

Estimating the Chance of No Synonymous Substitution within Tract.

We accounted for the formal possibility that selection to maintain consecutive lysine (AAA/AAG) or phenylalanine (TTT/TTC) residues could, in principle, contribute to the maintenance of a homopolymeric region. See Results and Discussion for reasons why this possibility is unlikely. Even if this were true, the exclusive use of AAA (rather than AAG) and TTT (rather than TTC) codons is striking. We tested the null hypothesis that synonymous sites within the homopolymeric tract evolve neutrally, using an approach similar to that described above. Based on per-site substitution rates at 4-fold degenerate sites, we focused on the two specific substitution types in question: the proportion of 4-fold degenerate A’s that changed to a G or proportion of 4-fold degenerate T’s that changed to a C. We then calculated the chance that the three synonymous sites in the polyA or polyT tract did not experience a synonymous substitution by chance alone.

Results and Discussion

Preservation of Frameshifts: Evidence for Selection

Frameshifts Persist across Deep Divergences.

We discovered that frameshifts persist within endosymbiont genes at several taxonomic levels: within host ant colonies, among populations, and most surprisingly, across ancient species divergences. Specifically, our polymorphism data for closely related B. pennsylvanicus strains showed no variation in the length of frameshifted tracts (i.e., all sequence reads in the genome assembly contained the frameshift). Similarly, we found that hisH sequences from geographically distinct (Massachusetts and Wisconsin) populations of B. pennsylvanicus share the frameshift. In addition, we show here that frameshifts have persisted across ancient species divergences (fig. 1). Based on the estimated divergence times of ant hosts (Degnan et al. 2004), shared frameshifts are several million years old (except for a potentially young, independent frameshift in ybiS of Polyrhachis turneri; fig. 1b).

Genome-Wide, Indels within Homopolymers Exceed Synonymous Mutation Rates.

Our analysis of the B. pennsylvanicus genome, assembled from a population sample, revealed that polyA and polyT tracts experience exceptionally high rates of indel mutations across the genome, consistent with previous work (Gomez-Valero et al. 2008; Moran et al. 2009). Indels are most likely to be neutral in intergenic regions (but see Dunbar et al. 2007), so we use these regions to estimate the indel mutation rate. Within intergenic regions, ~1% of polyA or polyT tracts ≥5 bp showed indel polymorphisms. For longer tracts (≥8 and ≥9 bp), indel mutation rates were far higher (~17% and 21%, respectively) (table 1). Thus, the frequency of indels increases with tract length, as documented previously (Moran et al. 2009). However, this polymorphism data showed no indels in the tracts of the four B. pennsylvanicus frameshifted loci (hisH and ybiS studied here, as well as ubiF and ytfM), nor across the numerous in-frame tracts within other ORFs.

Table 1.
Among Related Strains of Blochmannia pennsylvanicus, Indel Polymorphisms along PolyA or PolyT Tracts Far Exceed the Genome-Wide Synonymous Mutation Rate (of 0.01%, see Results and Discussion).

The rate of indel mutations per tract far exceeds the rate of synonymous mutations per synonymous site. Of the 188,946 synonymous sites in all ORFs, polymorphisms occurred at just 176 sites (or 0.00093 changes per synonymous site). This very low level of synonymous polymorphism (<0.1%) reflects the close relationship among the bacterial strains included in the genome library but is dramatically lower than the rate of indel mutations. For long tracts ≥9 bp, such as those in the frameshifted genes considered here, the indel rate within intergenic regions is 226-times the synonymous mutation rate (i.e., 0.21/0.00093). When considering the entire genome (a sample that conservatively underestimates indel mutation rate), this ratio is still quite high, with indels occurring 27-times more often than synonymous mutations (i.e., 0.025/0.00093).

Dearth of Indels Is Unexpected, Given the High Divergence among Frameshifted Lineages.

In our analysis of divergent Blochmannia species, frameshifts and associated polyA/T tracts persist in lineages that have diverged for millions of years. For example, the Blochmannia lineages sharing the ancient frameshift in ybiS diverged about 16.2–19.9 Ma (Degnan et al. 2004). Synonymous changes along individual branches and between most sister species are not saturated (e.g., see pairwise dS values calculated across the entire gene; fig. 1). However, the synonymous divergence for taxa spanning a frameshifted clade is well above saturation (table 2). Divergences in this range are difficult to estimate precisely, but clearly substantial synonymous change has accumulated. Likewise, a tally of all synonymous changes along branches illustrates that, on average, each synonymous site has experienced several substitutions during the diversification of the frameshifted clades (table 3).

Table 2.
Saturation of Pairwise Synonymous Divergences between Taxa Spanning the Frameshifted Clade.
Table 3.
Tally of Synonymous Substitutions during the Divergence of the Frameshifted Clade.

Because rates of indel mutations within long tracts far exceed rates of synonymous mutations, we would expect numerous (on the order of many hundreds to thousands of) indel substitutions within these clades if substitutions reflected neutral processes. The fact that we observe zero indel substitutions in hisH and gmhB, and just two indels in the ancient frameshifted clade of ybiS, is therefore inconsistent with neutrality. This striking difference in rates of indels versus synonymous changes within versus between species suggests that selection preserves the length of frameshifted tracts.

Preservation of Homopolymeric Tracts: Evidence for Selection

Homopolymeric tracts that contain frameshifts show a striking level of nucleotide conservation compared with other regions of these genes and compared with the homologous region in taxa that lack the frameshift (fig. 1). To explore this pattern quantitatively, we determined whether the homopolymeric tracts themselves show unusual nucleotide conservation. We tested the null hypothesis that such tracts are neutral and evolve as do 4-fold degenerate sites in the gene considered. We approached this by estimating per-site substitution rates at 4-fold degenerate A’s (or T’s) across each branch (supplementary fig. 1, Supplementary Material online) and then determining the chance that the observed number of consecutive A’s (or T’s) in the tract did not experience a substitution as this group diversified. For each gene, the chance of zero nucleotide substitutions in the tract region was vanishingly small (<0.0005 in each case, and as low as 10−11; supplementary table 1, Supplementary Material online). Therefore, we can rule out the possibility that the homopolymeric regions tracts are conserved by chance alone.

We accounted for the remote possibility that selection to maintain consecutive residues of lysine (AAA/AAG) or phenylalanine (TTT/TTC) could, in principle, contribute to the maintenance of polyA and polyT tracts, respectively. Although is it conceivable that selection favors these particular amino acids in the tract region, this is unlikely because amino acid variations occur in related Blochmannia lineages that lack the frameshift (fig. 1). In this sense, selection favoring lysine or phenylalanine would involve a clade-specific selective pressure. Furthermore, selection for particular amino acids cannot explain the absence of synonymous changes in the homopolymeric tracts (AAA → AAG or TTT → TTC). Based on A → G and T → C substitutions at 4-fold degenerate sites in these genes, it is improbable that synonymous sites would not have undergone a substitution within the tracts of hisH (P = 0.041) and ybiS (P = 0.052), though this factor cannot be ruled out for gmhB (0.39) (supplementary table 1, Supplementary Material online).

Signs of Transcriptional Slippage, a Proposed Advantage of Homopolymers.

We detected clear signatures of purifying selection at frameshifted genes, supporting the functionality of encoded proteins. Pairwise dN/dS values calculated for each gene are far less than 1 (fig. 1), consistent with patterns observed at frameshifted Buchnera genes (Tamas et al. 2008). In addition, at hisH and gmhB, a branch model suggests that lineages with the frameshift actually have a slight (but nonsignificant) reduction in ω, just the opposite of the prediction under relaxed constraint (table 4). At ybiS, the branch test suggests lower ω along the sole Blochmannia lineage with an intact gene (P. macropus); however, extremely high dS (>3) along this lineage makes the inference unreliable. These results were not altered when we considered only the regions upstream or downstream of the tract (supplementary table 2, Supplementary Material online). In sum, the disrupted coding regions show signatures of functional constraint that is comparable with their intact relatives.

Table 4.
Maximum Likelihood Estimates of Nonsynonymous/Synonymous Divergences (ω) for Blochmannia Genes.

Further supporting a role of transcriptional slippage, frameshifts and homopolymers show a strong association. Frameshifts persist only in sequences that also retain long homopolymeric tracts (A11 for hisH, A9-10 for ybiS, and T9 for gmhB). By contrast, genes of related Blochmannia species that lack the frameshift often show interruption of the polyA or polyT tract with bases other than A or T, respectively (fig. 1).


Several lines of evidence demonstrate an exceptional conservation of frameshifts and long homopolymeric tracts in which they occur. First, frameshifts and associated tracts have persisted for millions of years across genetically divergent lineages showing extensive substitutions outside of the tract region. Second, patterns of polymorphism at intergenic regions demonstrate that indel mutations are common within polyA and polyT tracts of Blochmannia. In this ant mutualist, the rate of indel mutations in long (≥9 bp) tracts exceeds the synonymous mutation rate by >200-fold. We conclude that, in the absence of purifying (negative) selection, numerous indel substitutions would be expected along lineages maintaining the frameshift. However, we observe very few indels in these ancient groups: zero indels for two genes and one indel for ybiS. This exceptional conservation of tract length argues that selection maintains the observed frameshifts.

Likewise, the A and T bases within the homopolymeric tracts are far more conserved than would be expected under neutrality. Given the high sequence divergences for the genes sampled, the tract would not be conserved in the absence of selection. Based on rates of 4-fold degenerate substitutions elsewhere in the genes considered, we can rule out the possibility that the tracts evolve neutrally. The unlikely possibility of clade-specific selection favoring particular amino acids (lysine or phenylalanine) is difficult to rule out, but this explanation cannot explain the improbable conservation of synonymous sites within the tract.

We propose that purifying selection maintains the frameshift in these bacterial loci, as well as the polyA or polyT tract required for restoration of gene function via transcriptional slippage. Under this model, any base change within the slippery tract would inhibit the restoration of gene function by slippage and, as an effective knockout mutation, experience intense purifying selection. Consistent with a rescuing effect of slippage, the frameshifted genes show low dN/dS ratios and no signatures of relaxed functional constraint. Further supporting an important role of slippage, frameshifts persist only in sequences that also retain long homopolymeric tracts. These results suggest a new mechanism for the maintenance of homopolymeric regions within bacterial genomes: The slippery tracts are essential to rescue functionality.

It is less clear what mechanisms may explain the persistence of the frameshift itself. Compared with genes with an intact reading frame, the frameshift may reduce effective gene expression by lowering the number of transcripts for full-length proteins. Genes that contain frameshifts have varied functions, and there is no clear proposal for why reduced expression levels could be advantageous. However, alternative explanations, that frameshifts compensate for indels elsewhere or that truncated proteins are functional, are not supported (see Tamas et al. 2008). Thus, although speculative, current data suggest that the maintenance of frameshifts involves reduction in effective gene expression. Exploiting frameshifts as a mechanism to alter expression may be especially important in bacteria such as endosymbionts that lack many gene regulation capabilities (Wilcox et al. 2003).

Supplementary Material

Supplementary tables 1 and 2 and figure 1 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).

Supplementary Material

[Supplementary Data]


We thank Simon Robson for contributing Polyrhachis specimens used for this study. We are grateful to Adam B. Lazarus and Jeremy Brozek for assistance with data collection and analysis and thank Siv G. E. Andersson for helpful discussion. We also appreciate the valuable comments of two anonymous reviewers. This work was supported by grants from the National Science Foundation (MCB-0604177) and the National Institutes of Health (R01GM062626) to J.J.W.


  • Baranov PV, Hammer AW, Zhou J, Gesteland RF, Atkins JF. Transcriptional slippage in bacteria: distribution in sequenced genomes and utilization in IS element gene expression. Genome Biol. 2005;6:R25. [PMC free article] [PubMed]
  • Bayliss CD. Determinants of phase variation rate and the fitness implications of differing rates for bacterial pathogens and commensals. FEMS Microbiol Rev. 2009;33:504–520. [PubMed]
  • Benzer S. Genetic fine structure. Harvey Lect. 1961;56:1–21. [PubMed]
  • Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD. Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res. 2003;31:3497–3500. [PMC free article] [PubMed]
  • Degnan PH, Lazarus AB, Brock CD, Wernegreen JJ. Host-symbiont stability and fast evolutionary rates in an ant-bacterium association: cospeciation of Camponotus species and their endosymbionts, Candidatus Blochmannia. Syst Biol. 2004;53:95–110. [PubMed]
  • Degnan PH, Lazarus AB, Wernegreen JJ. Genome sequence of Blochmannia pennsylvanicus indicates parallel evolutionary trends among bacterial mutualists of insects. Genome Res. 2005;15:1023–1033. [PMC free article] [PubMed]
  • Dunbar HE, Wilson AC, Ferguson NR, Moran NA. Aphid thermal tolerance is governed by a point mutation in bacterial symbionts. PLoS Biol. 2007;5:e96. [PMC free article] [PubMed]
  • Gil R, Silva FJ, Zientz E, et al. (13 co-authors) The genome sequence of Blochmannia floridanus: comparative analysis of reduced genomes. Proc Natl Acad Sci U S A. 2003;100:9388–9393. [PMC free article] [PubMed]
  • Gomez-Valero L, Latorre A, Gil R, Gadau J, Feldhaar H, Silva FJ. Patterns and rates of nucleotide substitution, insertion and deletion in the endosymbiont of ants Blochmannia floridanus. Mol Ecol. 2008;17(19):4382–4392. [PubMed]
  • Gordon D, Abajian C, Green P. Consed: a graphical tool for sequence finishing. Genome Res. 1998;8:195–202. [PubMed]
  • Koch AL. Catastrophe and what to do about it if you are a bacterium: the importance of frameshift mutants. Crit Rev Microbiol. 2004;30:1–6. [PubMed]
  • Moran NA. Accelerated evolution and Muller’s rachet in endosymbiotic bacteria. Proc Natl Acad Sci U S A. 1996;93:2873–2878. [PMC free article] [PubMed]
  • Moran NA, McCutcheon JP, Nakabachi A. Genomics and evolution of heritable bacterial symbionts. Annu Rev Genet. 2008;42:165–190. [PubMed]
  • Moran NA, McLaughlin HJ, Sorek R. The dynamics and time scale of ongoing genomic erosion in symbiotic bacteria. Science. 2009;323:379–382. [PubMed]
  • Rogozin IB, Pavlov YI. Theoretical analysis of mutation hotspots and their DNA sequence context specificity. Mutat Res. 2003;544:65–85. [PubMed]
  • Ronquist F, Huelsenbeck JP. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003;19:1572–1574. [PubMed]
  • Streisinger G, Okada Y, Emrich J, Newton J, Tsugita A, Terzaghi E, Inouye M. Frameshift mutations and the genetic code. Cold Spring Harb Symp Quant Biol. 1966;31:77–84. [PubMed]
  • Tamas I, Wernegreen JJ, Nystedt B, Kauppinen SN, Darby AC, Gomez-Valero L, Lundin D, Poole AM, Andersson SG. Endosymbiont gene functions impaired and rescued by polymerase infidelity at poly(A) tracts. Proc Natl Acad Sci U S A. 2008;105:14934–14939. [PMC free article] [PubMed]
  • van Passel MW, Ochman H. Selection on the genic location of disruptive elements. Trends Genet. 2007;23:601–604. [PubMed]
  • Wilcox JL, Dunbar HE, Wolfinger RD, Moran NA. Consequences of reductive evolution for gene expression in an obligate endosymbiont. Mol Microbiol. 2003;48:1491–1500. [PubMed]
  • Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol. 2007;24:1586–1591. [PubMed]

Articles from Molecular Biology and Evolution are provided here courtesy of Oxford University Press
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...