Salomon, Matthew P. et al. (2013) International Worm Meeting "The rate and spectrum of spontaneous mutations in experimental populations of the nematode Caenorhabditis remanei."

Ostrow, Dejerianne G., Salomon, Matthew P., Matsuba, Chikako, Baer, Charles F.

[International Worm Meeting, 2013]

To what degree natural selection has shaped the rate of spontaneous mutations among different taxa remains an important unresolved question in evolutionary biology. While mutation rates are known to vary among and within taxa, the relative importance of natural selection versus non-adaptive processes has yet to be determined. Classical theory predicts that the strength of natural selection to reduce the deleterious mutation rate should be stronger in asexual and selfing taxa than in outcrossing sexual taxa, leading to an adaptive decrease in mutation rate in the former. However, other theory predicts (1) that "mutator" alleles can hitchhike to high frequency in asexual/selfing taxa, thereby leading to a (non-adaptive) increase in the mutation rate, and (2) the efficiency of selection to reduce mutation rate may be substantially greater in outcrossing than asexual or selfing taxa due to the larger effective population sizes of the former. Whether general trends in exist in nature is currently unknown. Nematodes in the genus Caenorhabditis provide an ideal system to test questions of how mutation rates vary among closely related species with different reproductive strategies. Within the genus the ancestral reproductive state within is outcrossing (gonochorism), however self-fertilization (hermaphroditism) has evolved independently several times. To examine the role of matting system on the evolution of mutation rates we constructed a set of long-term mutation accumulation (MA) lines of the outcrossing species Caenorhabditis remanei. MA lines were maintained by transferring a single male-female pair of worm per generation. After 122 generations of MA, five randomly selected MA lines along with the ancestral control were re-sequenced using Illumina sequencing technology.

Salomon, Matthew et al. (2009) International Worm Meeting "Persistence time of mutations affecting fitness and body size in Caenorhabditis briggsae and C. elegans."

Upadhyay, Ambuj, Salomon, Matthew, Baer, Charles, Levy, Laura, Keller, Thomas, Phillips, Naomi, Blanton, Dustin, Ostrow, Dejerianne, Bour, Whitney, Sylvestre, Thamar

[International Worm Meeting, 2009]

The level of genetic variation present in a population is a composite function of mutation, population size, and natural selection. Historically, efforts to understand differences (or similarities) between groups in levels of genetic variation have focused on the interplay between population size and natural selection. However, much less attention has been paid to the alternative possibility that differences among groups are due to systematic differences in the underlying rate of mutation. Much of the difficulty in interpreting the role of mutation stems from the fact that most of what is known about genomic mutational properties, for quantitative traits in multicellular eukaryotes, comes from a handful of phylogenetically distant and biologically dissimilar model organisms, making meaningful comparisons difficult. Over the past several years our lab has been investigating the properties of new mutations in a model nematode system within a comparative phylogenetic framework. Mutations have been allowed to accumulate in the (relative) absence of natural selection, thus allowing us to estimate the genetic variance introduced by new mutation (VM) for two species of rhabditid nematodes, Caenorhabditis elegans and C. briggsae. Previous work in this system suggests that the mutation rate in C. briggsae is on the order of twice that of C. elegans for quantitative traits and dinucleotide repeats. Here we report the standing genetic variance (VG) for two quantitative traits, lifetime reproduction and body size, in worldwide collections of C. briggsae and C. elegans natural isolates. Comparisons of VG to VM between the natural isolates and our mutation accumulation lines allow us to infer the magnitude and pattern of constraint on phenotypic evolution in these two species. Taking the results from the two species together, the persistence time (VG/VM) of new mutations affecting fitness is on the order of tens to perhaps hundreds of generations, with an average selection coefficient against homozygotes of a few per-cent. Furthermore, the pattern of persistence time for mutations affecting adult body size is onsistent with that of fitness in both species. These results suggest that idiosyncratic selection, perhaps due to random hitchhiking - "genetic draft" - is paramount in shaping the standing genetic variance of these traits in these species.

Matthew Salomon et al. (2007) International Worm Meeting "Persistence time of mutations affecting fitness and body size in C. elegans and C. briggsae."

Matthew Salomon, Jasmine Brown, Charles Baer, Dejerianne Ostrow, Charlie Shaw

[International Worm Meeting, 2007]

For traits closely tied to fitness, most mutations are deleterious and the ratio t of the standing genetic variance (VG) to the per-generation input of genetic variance from new mutations (VM) provides a measure of the time that a mutation persists in the population and of the strength of selection on mutant alleles. Previous experiments have shown that the deleterious mutation rate in C. elegans is lower than that of Drosophila melanogaster, but how much lower is unresolved. For mutations affecting viability in Drosophila, t ~100 generations. We present data on standing and mutational genetic variance for total fitness and body size in C. elegans. For fitness, t~20 generations, very consistent with the Drosophila results when mating system is taken into account. Preliminary data from C. briggsae suggest that t for fitness is very similar to C. elegans, consistent with an approximately two-fold higher mutation rate in C. briggsae. The difference in mating system between flies and C. elegans suggests that natural selection may have favored a reduction of the mutation rate in the self-compatible taxon, as predicted by theory.

Saxena, Ayush Shekhar et al. (2015) International Worm Meeting "Rapid accumulation of variation in mutation rate under relaxed selection in C. elegans."

Chikako, Matsuba, Salomon, Matthew, Charles, Baer, Saxena, Ayush Shekhar

[International Worm Meeting, 2015]

The rate and spectrum of spontaneous mutation vary at many hierarchical levels, from within individual genomes to among Domains of life. An interesting possibility, for which there is some evidence, is that the mutation rate varies inversely with fitness, such that individuals with low fitness have elevated mutation rates. To systematically investigate that possibility, we allowed mutations to accumulate for approximately 150 generations under relaxed selection in 10 sets of "second-order mutation accumulation" (MA) lines of the nematode C. elegans. Each set of second-order lines was derived from a different "first-order MA line" from a set of MA lines that had accumulated mutations for 250 generations. Of the 10 sets of second-order MA lines, five were derived from a first-order line with high fitness and five were derived from a first-order line with low fitness. Five replicate second-order MA lines from each set of first-order lines and the ten first-order progenitors were sequenced at ~25X genome-wide coverage using standard Illumina technology. The average base-substitution mutation rate does not differ between the High Fitness and Low Fitness treatment, but there is significantly greater variation in mutation rate among first-order lines within the Low Fitness treatment. Those results broadly recapitulate the results for relative fitness itself. In contrast, the High Fitness treatment had a higher rate of small indels. There is an overall deletion bias (3D:2I) which does not differ between the High and Low fitness treatments. Pooled over treatments, there is significant among-line variance in mutation rate, from which we conclude that the genomic mutation rate presents a substantial mutational target. Averaged over all lines, the transition/transversion ratio is 0.72, very close to the previously observed value for and considerably less than the standing Ts/Tv ratio in C. elegans. We observed that single-nucleotide mutations are twice as likely to be found in introns as in exons, and small indels are five-fold more likely to be found in introns as in exons. Those results indicate that a substantial fraction of mutations are removed by purifying selection even at very small effective population size. .

Ostrow, Dejerianne et al. (2009) International Worm Meeting "De-canalizing effect of new mutations evidenced in transcriptome of C. elegans?"

Blanton, Dustin, Ostrow, Dejerianne, Matsuba, Chikako, Salomon, Matthew, Baer, Charles

[International Worm Meeting, 2009]

Why some phenotypic traits are highly plastic or otherwise variable and others are highly invariant (=canalized) in the face of environmental variation remains a puzzle. Data from whole-genome microarray analysis of cDNA collected from large worm cultures suggests that spontaneous mutations reduce the environmental variability of transcript number. In an attempt to better understand the role of transcriptional modulation as a means of maintaining canalized phenotypes, we are investigating the variance between isogenic C. elegans lines as evidenced by the variability in gene expression in much smaller samples. The motivation for using very small samples is to minimize variance propagated by small differences in timing of development. For each of 6 biological replicates, RNA was extracted and pooled from 5 L3 worms, converted to cDNA, amplified, labeled, and hybridized to an Affymetrix GeneChip. Measures of variability between biologically replicated transcriptomes allow us to estimate transcriptional variance when genetic variation is minimized. These data are the foundation for future research comparing ancestral control lines to mutation accumulation lines for which mutational parameters have been well characterized.

Saxena, Ayush et al. (2017) International Worm Meeting "Tempo, mode, and fitness effects of mutation in Caenorhabditis elegans over 400 generations of minimal selection."

Salomon, Matthew, Baer, Charles, Yeh, Shu-Dan, Matsuba, Chikako, Saxena, Ayush

[International Worm Meeting, 2017]

The rate and molecular spectrum of mutations varies within and among genomes, species and higher taxa. The understanding of the mechanistic causes of variation is far from complete; the understanding of evolutionary causes is even more rudimentary. It has been posited that individuals in poor physiological condition will experience higher rates of mutation than those in good condition. Since deleterious mutations lead to poor condition, the possibility exists that individuals carrying a high mutation load will experience an elevated mutation rate. We tested that hypothesis with a set of "second-order mutation accumulation" (O2MA) lines of the nematode C. elegans. MA lines that had accumulated mutations under minimal selection for ~250 generations ("first-order MA lines", O1MA) were sorted into high-fitness and low-fitness groups, replicated into new sets of O2MA lines, and allowed to accumulate mutations for another ~150 generations of minimal selection. Whole-genome sequencing of 48 O2MA lines and their O1MA ancestors revealed significant variation in base-substitution rate and in total mutation rate both among the O1MA lines measured at generation 250 and in the O2MA lines measured at G400. However, O2MA rate does not depend on ancestral fitness, nor does the specific O1MA ancestry explain a significant fraction of the variance among O2MA lines. Thus, the signal of variation in mutation rate decays on the order of a hundred generations. Multiple logistic regression of mutability on a set of predictor variables revealed that local three-base nucleotide context is the most important predictor of mutability, but that GC content of the 1 Kb surrounding a site and - importantly - local recombination rate are also significant predictors. Mutability explains a large fraction of the variance in standing nucleotide diversity. The deletion rate of low-fitness O2MA lines is less than that of high-fitness O2 lines. However, low-fitness O1MA lines carry more indels of putatively large effect than do high-fitness O1MA lines, whereas low-fitness O2MA lines carry fewer indels of putatively large effect than the high-fitness O2MA lines. Consistent with those findings, the mean mutational effect on experimentally-measured fitness is twice as great in the low-fitness O2MA lines as in the high-fitness lines. Taken together, these results strongly imply that, on average, epistasis is synergistic among new deleterious mutations.

Saxena, Ayush Shekhar et al. (2019) International Worm Meeting "Evolution of the rate of copy-number and structural variant mutations under relaxed selection in Caenorhabditis elegans."

Saxena, Ayush Shekhar, Salomon, Matthew, Matsuba, Chikako, Baer, Charles, Rajaei, Moein

[International Worm Meeting, 2019]

Mutation Accumulation (MA) experiments have been a workhorse of evolutionary genetics for over fifty years, and much of what is known about the mutational process at both the phenotypic and molecular level has been learned from MA experiments. The advent of economical whole-genome short-read sequencing greatly increased our ability to characterize the rate and spectrum of base-substitution and short indel mutations. Copy-number variants (CNVs) and structural variants (SVs, e.g., inversions, translocations) have been more difficult to characterize. The paucity of information on the mutational properties of CNVs and SVs is significant because there is compelling evidence that those types of mutations underlie much variation in complex traits. We previously demonstrated by means of a "second-order MA" experiment (i.e., sets of MA lines founded from individual MA lines) with the nematode Caenorhabditis elegans that the rate of base-substitution and (especially) small indel mutations increased over the course of ~250 generations of minimal selection. Preliminary analyses of CNVs based on short-read Illumina sequence data suggested that a similar trend holds for CNVs, but the quantitative estimates of CNV mutation varied by over an order of magnitude depending on the input parameters of the analytical algorithm. To attempt to get a better handle on the CNV and SV mutation rates, we sequenced a small number of MA lines (N=5) with Pacific Biosciences long-read sequencing and used the resulting estimates to attempt to inform the analysis of our more copious short-read data (N>100) MA lines. Preliminary results suggest a conservative long CNV rate of approximately 5% that of small mutations (base subs + short indels). The data are biased toward deletions, but comparisons of the sequence of the ancestor of our MA lines with the C. elegans reference genome indicates that our analytical method has the capacity to detect insertions and inversions as well as deletions. We additionally report comparative MA data from an additional set of 25 C. elegans MA lines derived from a different starting genotype (PB306).

Ostrow, Dejerianne et al. (2011) International Worm Meeting "Proximate and ultimate cumulative effects of temperature on spontaneous mutation in C. briggsae and C. elegans."

Matsuba, Chikako, Regan, Kerry, Sylvestre, Laurence, Upadhyay, Ambuj, Ostrow, Dejerianne, Lewis, Suzanna, Tabman, Brandon, Salomon, Matthew, Baer, Charles F.

[International Worm Meeting, 2011]

A long-standing hypothesis in the field of molecular evolution is that the rate of molecular evolution is positively correlated with metabolic rate. Implicit in this idea (sometimes made explicit) is that the mutation rate is positively correlated with metabolic rate. Unfortunately, several factors are confounded with metabolic rate, including generation time, population size and temperature. Further, to the extent that metabolic rate is correlated with temperature (strongly in ectotherms, less strongly in endotherms), direct mutagenic effects of temperature are often further confounded with environmental "stress". Here we report results from an experiment designed to de-confound the effects of temperature, generation time, evolutionary history, and environmental stress to investigate the relationship between temperature and mutation rate. We allowed spontaneous mutations to accumulate under relaxed selection in C. briggsae (PB800 strain) and C. elegans (N2 strain) at low temperature (18 deg C) for 100 generations and high temperature (26 deg C) for 170 generations. The high temperature environment is markedly stressful for C. elegans by the objective criteria of reduced fecundity and survivorship whereas it is not for C. briggsae. We assayed a demographic measure of fitness relative to the unmutated common ancestor ("control") in both sets of mutation accumulation (MA) lines at both temperatures. The results are both interesting and complicated. In all four cases (both sets of MA lines in both species), the cumulative decline in fitness was greater when assayed at high temperature, although the effect was small in the C. briggsae high temperature MA lines. This result suggests that there is a significant class of mutations with (high) temperature-dependent effects. In C. briggsae, high-temperature MA lines decline in fitness faster than low-temperature MA lines, and that result does not depend (much) on the assay temperature. Thus, it appears that temperature itself is mutagenic in C. briggsae. Conversely, in C. elegans, the cumulative effects of mutations are greater when assayed at high temperature than at low temperature, but there is little difference between high and low-temperature MA lines. This result suggests that in C. elegans, on average, there is no relationship between temperature and mutation rate. Direct characterization of the molecular mutational spectrum is underway and will help resolve the relative contributions of mutation rate and allelic effects to the cumulative mutational decay in fitness.