- Triton (Spinward Fringe, Book 3);
- The Bad Book Affair.
- Pre-test Neurology.
- Pocket K No. 15: 'Omics' Sciences: Genomics, Proteomics, and Metabolomics.
- Brocas Brain: Reflections on the Romance of Science;
For some of them, genetic manipulations and competition assays have directly confirmed that mutations indeed improve fitness under the conditions of the LTEE [ 22 , 23 ]. These are also the two non-mutator populations that underwent the most replacements of LTEE-evolved mutations by donor alleles red hash marks , although we do not know whether this relation is coincidental or meaningful.
In the next section, we consider the fate of the presumptively beneficial mutations that were present in the LTEE-derived recipients at the start of the STLE.
These hypotheses are not mutually exclusive, and so two or more of them may contribute to the observed patterns of introgression Fig 1 and Fig 2. Hypothesis 1: Some regions of donor DNA were transferred more often than other regions, leading to overrepresentation of the former regions in the recombinant genomes. Hypothesis 2: Some regions of donor DNA contained alleles that were beneficial to the recipient, leading to overrepresentation of those regions in the recombinant genomes.
Hypothesis 3: Some regions of donor DNA contained alleles that were deleterious to the recipient, leading to underrepresentation of those regions in the recombinant genomes. This hypothesis can be subdivided into two variant hypotheses. According to Hypothesis 3A, the donor alleles were maladaptive regardless of the beneficial mutations that arose during the LTEE. According to Hypothesis 3B, the donor alleles were maladaptive specifically because the recipient genomes had acquired beneficial mutations in those regions during the generations of the LTEE that preceded the STLE.
In addition to the hypotheses above, sequence divergence and mismatch repair are known to reduce recombination efficiency [ 24 ]. However, E. These two source strains were independently isolated from nature many years ago [ 25 ]. Still, about half of their shared genes encode proteins that have identical amino-acid sequences [ 26 ]. In addition to these more or less ancient differences, the four K donors were deliberately modified by transposon mutagenesis to make them auxotrophic for different amino acids and by introducing the F plasmid at different locations into their chromosomes to make them Hfr strains; and, as described above, the B-derived recipients accumulated beneficial mutations during the LTEE.
Therefore, it appears that sequence divergence had a weak effect on the introgression of K alleles into the STLE populations, which mostly influenced the fine-scale mosaic structure of recombinant regions. We lack a priori information about what sites in the K donor genomes could provide beneficial alleles to the recombinant populations. However, if such sites exist, then we would expect them to be in those regions where the introgression scores are high Fig 2. On the other hand, Hypothesis 2 seems unlikely, because Souza et al.
Fig 2 shows the inferred location and direction of the Hfr origins of transfer of the four K donor strains as well as the location of their auxotrophy mutations, which bear on Hypotheses 1 and 3A, respectively. However, not all matings would begin immediately, shaking may disrupt conjugation, and the efficiency of DNA transfer declines with distance from the transfer origin. The peak in the introgression scores between about —1 and 0 Mbp fits very well with the locations and directions of the oriT transfer origin sites for two of the Hfr donors: REL and REL have oriT sites near the edges of this peak that point inward from opposite directions.
Most of the second, less-defined peak in introgression scores seems to fit moderately well with the other two Hfr donors, REL and REL, whose oriT sites are at about 1. However, the other shoulder of the second, less defined peak—from about 1. We also find donor-specific markers in some recombinant clones near the various donor-specific auxotrophic mutations Fig 2B , but almost all of these nearby introgressions involved a different donor. Nonetheless, the near absence of introgression along the circular chromosome between approximately 1.
On balance, then, patterns of introgression provide strong, albeit imperfect, support for Hypothesis 1. We also found compelling evidence of strong purifying selection at the sites of the auxotrophy mutations in the K Hfr donors. Recall that these mutations mean that the cells cannot produce essential amino acids and so cannot grow and persist in the minimal medium used for the STLE. The two donors with transfer properties that can account well for the introgression peak between about —1 and 0 Mbp have auxotrophic mutations located at positions that would sharpen the peak by limiting introgression at each edge.
The other two Hfr donors, REL and REL, have auxotrophy mutations in the argE and leuB genes, respectively, that would contribute to the observed decline in introgression scores on the broader shoulder of the less defined peak from about 0 to 1. REL also has a second auxotrophy mutation in leuB , but this gene is very far from its oriT site and thus probably not relevant to the observed patterns of introgression.
On balance, we also find support for Hypothesis 3A, whereby selection against the effectively lethal auxotrophy mutations in the donor strains reinforces and sharpens the patterns of introgression generated by the mechanics of gene transfer according to Hypothesis 1. Hypothesis 3B offers a different selection-based explanation for the patterns of introgression.
It rests on the idea that selection should also act against donor alleles in those genes where beneficial mutations arose in the LTEE and were present in a particular recipient at the start of the STLE. If this hypothesis were correct, then we would expect to see few, if any cases, where these beneficial mutations were removed and replaced by donor DNA.
The evidence in support of this hypothesis is ambiguous, at best, because of the considerable variation among the recombinant clones, in terms of both the proportion of their DNA that comes from the donor strains and the extent to which the LTEE-derived beneficial mutations have been retained or replaced.
Also, replacements of LTEE-derived beneficial mutations might simply reflect the recent introgression of donor alleles into the recipients i.
Analysis of the genome sequence of the flowering plant Arabidopsis thaliana | Nature
This pattern is consistent with Hypothesis 3B. Across the 12 populations, there is a slight tendency for these presumed beneficial mutations to have been replaced by donor alleles more often than the average genomic site, contrary to this hypothesis. On balance, the evidence does not support Hypothesis 3B. We can also exclude a hypothetical scenario in which beneficial LTEE-derived alleles in the recipients were generally replaced by K alleles that were as or more beneficial. We examined the recombinant clone sequences to determine whether these replacements reverted the gene to its pre-LTEE ancestral state i.
Whole genome and whole exome sequencing
As usual, we summarize the results for the odd-numbered final clones, but the results do not differ substantively for the even-numbered clones. We examined alignments of 60 proteins from both non-mutator and hypermutator clones containing LTEE-evolved alleles in the recipients that were replaced by recombination with the donors.
Fig 3 illustrates the four cases in which new alleles emerged. The yghJ gene in population Ara—1 was evidently affected by at least three recombination events: the whole gene is derived from K except for two regions, spanning residues — and — in the alignment that contain REL markers.
Fig 1. Sequence alignments of all replaced alleles are provided in S1 Dataset.
Only variable sites are shown. Red columns show K markers introduced by a recombination event. A Ara—1 yghJ : this locus appears to have experienced at least 3 recombination events. In contrast, we are confident that LTEE-derived alleles in genes where mutations repeatedly reached high frequency in populations with the ancestral mutation rate were under positive selection Table 1.
We therefore closely examined the alignments of the 30 proteins containing LTEE-evolved mutations in the recipients that were replaced by recombination in the nine odd-numbered non-mutator clones. Eleven of the 22 cases where an LTEE-evolved allele went back to its pre-LTEE ancestral state occurred in genes under strong positive selection in the LTEE [ 21 , 27 ], indicating that many beneficial mutations were removed by recombination with the donors.
The new alleles that evolved after recombination had effectively reconstructed the ancestral states in hslU in Ara—1 and pykF in Ara—4 imply that these reversions of beneficial LTEE-derived mutations by recombination were not adaptive. The following analyses focus, for simplicity, on the odd-numbered recombinant clones from the STLE populations that were not hypermutable; however, the even-numbered clones are similar.
We noticed that these recombinant clones often had more new mutations than did typical LTEE clones that had evolved for generations [ 21 ]. When we looked for evidence of parallel evolution among these new mutations, we found strong but spurious signals in two genes, nohB and waaQ ; in particular, these genes had multiple identical mutations in multiple lineages. The likely explanation for these parallel changes is gene conversion, in which recombination occurred between non-orthologous genes in the K donors and B recipients, such as between diverged paralogs or perhaps even non-homologous sequences.
To investigate further the possibility of gene conversion, we scored all genes that had three or more new mutations in the same recombinant genome as putative gene conversion events Table 3. Because so many apparently multi-mutation events occurred, and usually in multiple lineages, we think each case is best explained by a single gene-conversion event, not by multiple mutations in the same gene. We also used data on positive selection on specific genes in the LTEE [ 21 ] to ask whether new nonsynonymous mutations in the STLE tended to occur in the same genes.
Genes affected by nonsynonymous mutations resulting from putative gene conversions had a mean G -score of 0, while the genes harboring all other new nonsynonymous mutations had a mean G -score of This association suggests that the loss of beneficial mutations to recombination in these lineages also led to stronger selection for new beneficial mutations elsewhere in the genome. In contrast, the Ara—3 STLE population had 8 nonsynonymous replacements, but no new nonsynonymous mutations to compensate. The Ara—3 recombinants are the only ones with genomes that derive primarily from the K donor strains Fig 1 , S2 Fig.
Also, this population underwent unexpected changes in its ecology, which led to a substantial decline in its fitness relative to the common competitor used in the STLE [ 17 ] and the emergence of a negative frequency-dependent interaction between different recombinant genotypes, probably caused by a crossfeeding interaction [ 19 ].
Glossary of Genetics Terms
These striking changes in genetic background and ecological context may well have altered the genetic targets of selection in this population. We examined the distributions of recombinant segment lengths to see whether conjugation left a consistent signature in this respect.
The left column of Fig 4 shows the distribution of lengths of DNA segments derived from the K donor in the recombinant clones, and the right column shows the length distribution of the B-derived segments using the odd-numbered clones.