Mitotic gene conversion can be as important as meiotic conversion in driving genetic variability in plants and other species without early germline segregation.

In contrast to common meiotic gene conversion, mitotic gene conversion, because it is so rare, is often ignored as a process influencing allelic diversity. We show that if there is a large enough number of premeiotic cell divisions, as seen in many organisms without early germline sequestration, such as plants, this is an unsafe position. From examination of 1.1 million rice plants, we determined that the rate of mitotic gene conversion events, per mitosis, is 2 orders of magnitude lower than the meiotic rate. However, owing to the large number of mitoses between zygote and gamete and because of long mitotic tract lengths, meiotic and mitotic gene conversion can be of approximately equivalent importance in terms of numbers of markers converted from zygote to gamete. This holds even if we assume a low number of premeiotic cell divisions (approximately 40) as witnessed in Arabidopsis. A low mitotic rate associated with long tracts is also seen in yeast, suggesting generality of results. For species with many mitoses between each meiotic event, mitotic gene conversion should not be overlooked.

High level of somatic mutations detected in a diploid banana wild relative Musa basjoo.

Plants are thought to lack an early segregating germline and often retain both asexual and sexual reproduction, both of which may allow somatic mutations to enter the gametes or clonal progeny, and thereby impact plant evolution. It is yet unclear how often these somatic mutations occur during plant development and what proportion is transmitted to their sexual or cloned offspring. Asexual "seedless" propagation has contributed greatly to the breeding in many fruit crops, such as citrus, grapes and bananas. Whether plants in these lineages experience substantial somatic mutation accumulation is unknown. To estimate the somatic mutation accumulation and inheritance among a clonal population of plant, here we assess somatic mutation accumulation in Musa basjoo, a diploid banana wild relative, using 30 whole-genome resequenced samples collected from five structures, including leaves, sheaths, panicle, roots and underground rhizome connecting three clonal individuals. We observed 18.5 high proportion de novo somatic mutations on average between each two adjacent clonal suckers, equivalent to ~ 2.48 × 10-8 per site per asexual generation, higher than the per site per sexual generation rates (< 1 × 10-8) reported in Arabidopsis and peach. Interestingly, most of these inter-ramet somatic mutations were shared simultaneously in different tissues of the same individual with a high level of variant allele fractions, suggesting that these somatic mutations arise early in ramet development and that each individual may develop only from a few apical stem cells. These results thus suggest substantial mutation accumulation in a wild relative of banana. Our work reveals the significance of somatic mutation in Musa basjoo genetics variations and contribute to the trait improvement breeding of bananas and other asexual clonal crops.

Ancient rapid functional differentiation and fixation of the duplicated members in rice Dof genes after whole genome duplication.

Whole genome duplication (WGD) in plants is typically followed by genomic downsizing, where large portions of the new genome are lost. Whether this downsizing is accompanied by increased or decreased evolutionary rates of the remaining genes is poorly known, not least because homeolog pairings are often obscured by chromosomal rearrangement. Here, we use the newly published genome from a sedge, namely Kobresia littledalei, and CRISPR/Cas-9 editing to investigate how the Rho WGD event 70 million years ago (MYA) affected transcription factor evolutionary rates, fates, and function in rice (Oryza sativa) and sorghum (Sorghum bicolor). We focus on the 30-member DNA-binding with one zinc finger (Dof) transcription factor family in both crops due to their agronomic importance. Using the known speciation dates of rice from Kobresia (97 MYA) and sorghum (50 MYA), we find that rates of amino acid substitution in the critical Dof domain region were over twofold higher during the 20-million-year period following the WGD than before or afterward. Through comparison of synteny blocks, we report that at least 11% of Dof genes were purged from 70 to 50 MYA, while only 6% have been lost in the most recent 50-million-year interval. CRISPR/Cas9 editing revealed widespread fitness-related defects in flowering and lack of redundancy of paired members, as well as significant differences in expression between gene pairs. Together these findings demonstrate the strength of Dof genes as a model for deep evolutionary study and offer one of the most detailed portraits yet of the Rho WGD impact on a gene lineage.

Somatic Mutation Analysis in Salix suchowensis Reveals Early-Segregated Cell Lineages.

Long-lived plants face the challenge of ever-increasing mutational burden across their long lifespan. Early sequestration of meristematic stem cells is supposed to efficiently slow down this process, but direct measurement of somatic mutations that accompanies segregated cell lineages in plants is still rare. Here, we tracked somatic mutations in 33 leaves and 22 adventitious roots from 22 stem-cuttings across eight major branches of a shrub willow (Salix suchowensis). We found that most mutations propagated separately in leaves and roots, providing clear evidence for early segregation of underlying cell lineages. By combining lineage tracking with allele frequency analysis, our results revealed a set of mutations shared by distinct branches, but were exclusively present in leaves and not in roots. These mutations were likely propagated by rapidly dividing somatic cell lineages which survive several iterations of branching, distinct from the slowly dividing axillary stem cell lineages. Leaf is thus contributed by both slowly and rapidly dividing cell lineages, leading to varied fixation chances of propagated mutations. By contrast, each root likely arises from a single founder cell within the adventitious stem cell lineages. Our findings give straightforward evidence that early segregation of meristems slows down mutation accumulation in axillary meristems, implying a plant "germline" paralog to the germline of animals through convergent evolution.

The architecture of intra-organism mutation rate variation in plants.

Given the disposability of somatic tissue, selection can favor a higher mutation rate in the early segregating soma than in germline, as seen in some animals. Although in plants intra-organismic mutation rate heterogeneity is poorly resolved, the same selectionist logic can predict a lower rate in shoot than in root and in longer-lived terminal tissues (e.g., leaves) than in ontogenetically similar short-lived ones (e.g., petals), and that mutation rate heterogeneity should be deterministic with no significant differences between biological replicates. To address these expectations, we sequenced 754 genomes from various tissues of eight plant species. Consistent with a selectionist model, the rate of mutation accumulation per unit time in shoot apical meristem is lower than that in root apical tissues in perennials, in which a high proportion of mutations in shoots are themselves transmissible, but not in annuals, in which somatic mutations tend not to be transmissible. Similarly, the number of mutations accumulated in leaves is commonly lower than that within a petal of the same plant, and there is no more heterogeneity in accumulation rates between replicate branches than expected by chance. High mutation accumulation in runners of strawberry is, we argue, the exception that proves the rule, as mutation transmission patterns indicate that runner has a restricted germline. However, we also find that in vitro callus tissue has a higher mutation rate (per unit time) than the wild-grown comparator, suggesting nonadaptive mutational "fragility". As mutational fragility does not obviously explain why the shoot-root difference varies with plant longevity, we conclude that some mutation rate variation between tissues is consistent with selectionist theory but that a mechanistic null of mutational fragility should be considered.

Large-scale identification and functional analysis of NLR genes in blast resistance in the Tetep rice genome sequence.

Tetep is a rice cultivar known for broad-spectrum resistance to blast, a devastating fungal disease. The molecular basis for its broad-spectrum resistance is still poorly understood. Is it because Tetep has many more NLR genes than other cultivars? Or does Tetep possess multiple major NLR genes that can individually confer broad-spectrum resistance to blast? Moreover, are there many interacting NLR pairs in the Tetep genome? We sequenced its genome, obtained a high-quality assembly, and annotated 455 nucleotide-binding site leucine-rich repeat (NLR) genes. We cloned and tested 219 NLR genes as transgenes in 2 susceptible cultivars using 5 to 12 diversified pathogen strains; in many cases, fewer than 12 strains were successfully cultured for testing. Ninety cloned NLRs showed resistance to 1 or more pathogen strains and each strain was recognized by multiple NLRs. However, few NLRs showed resistance to >6 strains, so multiple NLRs are apparently required for Tetep's broad-spectrum resistance to blast. This was further supported by the pedigree analyses, which suggested a correlation between resistance and the number of Tetep-derived NLRs. In developing a method to identify NLR pairs each of which functions as a unit, we found that >20% of the NLRs in the Tetep and 3 other rice genomes are paired. Finally, we designed an extensive set of molecular markers for rapidly introducing clustered and paired NLRs in the Tetep genome for breeding new resistant cultivars. This study increased our understanding of the genetic basis of broad-spectrum blast resistance in rice.

Parent-progeny sequencing indicates higher mutation rates in heterozygotes.

Mutation rates vary within genomes, but the causes of this remain unclear. As many prior inferences rely on methods that assume an absence of selection, potentially leading to artefactual results, we call mutation events directly using a parent-offspring sequencing strategy focusing on Arabidopsis and using rice and honey bee for replication. Here we show that mutation rates are higher in heterozygotes and in proximity to crossover events. A correlation between recombination rate and intraspecific diversity is in part owing to a higher mutation rate in domains of high recombination/diversity. Implicating diversity per se as a cause, we find an ∼3.5-fold higher mutation rate in heterozygotes than in homozygotes, with mutations occurring in closer proximity to heterozygous sites than expected by chance. In a genome that is a patchwork of heterozygous and homozygous domains, mutations occur disproportionately more often in the heterozygous domains. If segregating mutations predispose to a higher local mutation rate, clusters of genes dominantly under purifying selection (more commonly homozygous) and under balancing selection (more commonly heterozygous), might have low and high mutation rates, respectively. Our results are consistent with this, there being a ten times higher mutation rate in pathogen resistance genes, expected to be under positive or balancing selection. Consequently, we do not necessarily need to evoke extremely weak selection on the mutation rate to explain why mutational hot and cold spots might correspond to regions under positive/balancing and purifying selection, respectively.

Identifying a large number of high-yield genes in rice by pedigree analysis, whole-genome sequencing, and CRISPR-Cas9 gene knockout.

Repeated artificial selection of a complex trait facilitates the identification of genes underlying the trait, especially if multiple selected descendant lines are available. Here we developed a pedigree-based approach to identify genes underlying the Green Revolution (GR) phenotype. From a pedigree analysis, we selected 30 cultivars including the "miracle rice" IR8, a GR landmark, its ancestors and descendants, and also other related cultivars for identifying high-yield genes. Through sequencing of these genomes, we identified 28 ancestral chromosomal blocks that were maintained in all the high-yield cultivars under study. In these blocks, we identified six genes of known function, including the GR gene sd1, and 123 loci with genes of unknown function. We randomly selected 57 genes from the 123 loci for knockout or knockdown studies and found that a high proportion of these genes are essential or have phenotypic effects related to rice production. Notably, knockout lines have significant changes in plant height (P < 0.003), a key GR trait, compared with wild-type lines. Some gene knockouts or knockdowns were especially interesting. For example, knockout of Os10g0555100, a putative glucosyltransferase gene, showed both reduced growth and altered panicle architecture. In addition, we found that in some retained chromosome blocks several GR-related genes were clustered, although they have unrelated sequences, suggesting clustering of genes with similar functions. In conclusion, we have identified many high-yield genes in rice. Our method provides a powerful means to identify genes associated with a specific trait.

High-resolution insight into recombination events at the SD1 locus in rice.

Recombination during meiosis plays an important role in genome evolution by reshuffling existing genetic variations into fresh combinations with the possibility of recovery of lost ancestral genotypes. While crossover (CO) events have been well studied, gene conversion events (GCs), which represent non-reciprocal information transfer between chromosomes, are poorly documented and difficult to detect due to their relatively small converted tract size. Here, we document these GC events and their phenotypic effects at an important locus in rice containing the SD1 gene, where multiple defective alleles contributed to the semi-dwarf phenotype of rice in the 'Green Revolution' of the 1960s. Here, physical separation of two defects allows recombination to generate the wild-type SD1 gene, for which plant height can then be used as a reporter. By screening 18 000 F2 progeny from a cross between two semi-dwarf cultivars that carry these different defective alleles, we detected 24 GC events, indicating a conversion rate of ~3.3 × 10-4 per marker per generation in a single meiotic cycle in rice. Furthermore, our data show that indels and single-nucleotide polymorphisms (SNPs) do not differ significantly in GC rates, at least at the SD1 locus. Our results provide strong evidence that GC by itself can regain an ancestral phenotype that was lost through mutation. This GC detection approach is likely to be broadly applicable to natural or artificial alleles of other phenotype-related functional genes, which are abundant in other plant genomes.

Direct Determination of the Mutation Rate in the Bumblebee Reveals Evidence for Weak Recombination-Associated Mutation and an Approximate Rate Constancy in Insects.

Accurate knowledge of the mutation rate provides a base line for inferring expected rates of evolution, for testing evolutionary hypotheses and for estimation of key parameters. Advances in sequencing technology now permit direct estimates of the mutation rate from sequencing of close relatives. Within insects there have been three prior such estimates, two in nonsocial insects (Drosophila: 2.8 × 10-9 per bp per haploid genome per generation; Heliconius: 2.9 × 10-9) and one in a social species, the honeybee (3.4 × 10-9). Might the honeybee's rate be ∼20% higher because it has an exceptionally high recombination rate and recombination may be directly or indirectly mutagenic? To address this possibility, we provide a direct estimate of the mutation rate in the bumblebee (Bombus terrestris), this being a close relative of the honeybee but with a much lower recombination rate. We confirm that the crossover rate of the bumblebee is indeed much lower than honeybees (8.7 cM/Mb vs. 37 cM/Mb). Importantly, we find no significant difference in the mutation rates: we estimate for bumblebees a rate of 3.6 × 10-9 per haploid genome per generation (95% confidence intervals 2.38 × 10-9 and 5.37 × 10-9) which is just 5% higher than the estimate that of honeybees. Both genomes have approximately one new mutation per haploid genome per generation. While we find evidence for a direct coupling between recombination and mutation (also seen in honeybees), the effect is so weak as to leave almost no footprint on any between-species differences. The similarity in mutation rates suggests an approximate constancy of the mutation rate in insects.

Selective sweep with significant positive selection serves as the driving force for the differentiation of japonica and indica rice cultivars.

BACKGROUND: Asian cultivated rice (Oryza sativa L.), including japonica and indica, is unarguable the most important crop in Asia as well as worldwide. However, a decisive conclusion of its origination and domestication processes are still lacking. Nowadays, the ever-increasing high-throughput sequencing data of numerous rice samples have provided us new opportunities to get close to the answer of these questions. RESULTS: By compiling 296 whole-genome sequenced rice cultivars and 39 diverse wild rice, two types of domesticated regions (DR-I and DR-II) with strong selective sweep signals between different groups were detected. DR-I regions included 28 blocks which significantly differentiated between japonica and indica subspecies, while DR-II regions were consisted of another 28 blocks which significantly differentiated between wild and cultivated rice, each covered 890 kb and 640 kb, respectively. In-depth analysis suggested that both DR-Is and DR-IIs could have originated from Indo-China Peninsula to southern China, and DR-IIs might be introgressed from indica to japonica. Functional bias with significant positive selection has also been detected in the genes of DR-I, suggesting important role of the selective sweep in differentiation of japonica and indica. CONCLUSIONS: This research promoted a new possible model of the origin of the cultivated rice that DR-Is in japonica and indica maybe independently originated from the divergent wild rice in the Indo-China Peninsula to southern China, and then followed by frequent introgression. Genes with significant positive selection and biased functions were also detected which could play important roles in rice domestication and differentiation processes.

Dissimilar Manifestation of Heterosis in Superhybrid Rice at Early-Tillering Stage under Nutrient-Deficient and Nutrient-Sufficient Condition.

Heterosis has long been exploited for crop breeding; however, the genetic mechanisms, particularly the initial establishment of heterosis during the early vegetative growth phase, remain elusive. The biggest challenge for that is to exclude noise genes from the identified heterosis-related candidates. Herein, we use nutrient-deficient hybrid with no measurable growth heterosis as control. After filtering these irrelevant genes, only 336 differentially expressed genes (DEGs), which is significantly lower than in previous reports, were identified as heterosis-related genes in a superhybrid rice of Liang-You-Pei-Jiu (LYP9) at early-tillering stage. Among the DEGs that could be mapped to quantitative trait loci (QTL), approximately 72.8% could be covered by yield or growth vigor-related QTL, thereby suggesting that our DEGs were reliable and may have potential value to rice breeding. Among the 336 DEGs identified, a majority showed intermediate expression relative to that of its parental lines (i.e. additive effects), particularly, expression was frequently more similar to the paternal line rather than the maternal line (44.1% versus 32.7%); the remaining 27.1% were exclusively up- or down-regulated between the hybrid and either parent. Interestingly, up-regulated genes encoded various enzymes, whereas down-regulated genes were enriched in responses to stress, indicating that hybrids may benefit from both activating metabolism-related pathways and alleviating fitness cost through allelic interactions. Furthermore, a significantly larger proportion of divergent genes and higher nonsynonymous substitutions rates were detected in these DEGs, suggesting a potential contribution to the establishment of heterosis in superhybrid rice LYP9.

A genome-wide survey reveals abundant rice blast R genes in resistant cultivars.

Plant resistance genes (R genes) harbor tremendous allelic diversity, constituting a robust immune system effective against microbial pathogens. Nevertheless, few functional R genes have been identified for even the best-studied pathosystems. Does this limited repertoire reflect specificity, with most R genes having been defeated by former pests, or do plants harbor a rich diversity of functional R genes, the composite behavior of which is yet to be characterized? Here, we survey 332 NBS-LRR genes cloned from five resistant Oryza sativa (rice) cultivars for their ability to confer recognition of 12 rice blast isolates when transformed into susceptible cultivars. Our survey reveals that 48.5% of the 132 NBS-LRR loci tested contain functional rice blast R genes, with most R genes deriving from multi-copy clades containing especially diversified loci. Each R gene recognized, on average, 2.42 of the 12 isolates screened. The abundant R genes identified in resistant genomes provide extraordinary redundancy in the ability of host genotypes to recognize particular isolates. If the same is true for other pathogens, many extant NBS-LRR genes retain functionality. Our success at identifying rice blast R genes also validates a highly efficient cloning and screening strategy.

Mutation rate analysis via parent-progeny sequencing of the perennial peach. II. No evidence for recombination-associated mutation.

Mutation rates and recombination rates vary between species and between regions within a genome. What are the determinants of these forms of variation? Prior evidence has suggested that the recombination might be mutagenic with an excess of new mutations in the vicinity of recombination break points. As it is conjectured that domesticated taxa have higher recombination rates than wild ones, we expect domesticated taxa to have raised mutation rates. Here, we use parent-offspring sequencing in domesticated and wild peach to ask (i) whether recombination is mutagenic, and (ii) whether domesticated peach has a higher recombination rate than wild peach. We find no evidence that domesticated peach has an increased recombination rate, nor an increased mutation rate near recombination events. If recombination is mutagenic in this taxa, the effect is too weak to be detected by our analysis. While an absence of recombination-associated mutation might explain an absence of a recombination-heterozygozity correlation in peach, we caution against such an interpretation.

Mutation rate analysis via parent-progeny sequencing of the perennial peach. I. A low rate in woody perennials and a higher mutagenicity in hybrids.

Mutation rates vary between species, between strains within species and between regions within a genome. What are the determinants of these forms of variation? Here, via parent-offspring sequencing of the peach we ask whether (i) woody perennials tend to have lower per unit time mutation rates compared to annuals, and (ii) hybrid strains have high mutation rates. Between a leaf from a low heterozygosity individual, derived from an intraspecific cross, to a leaf of its selfed progeny, the mutation rate is 7.77 × 10-9 point mutations per bp per generation, similar to Arabidopsis thaliana (7.0-7.4 × 10-9 point mutations per bp per generation). This suggests a low per unit time mutation rate as the generation time is much longer in peach. This is supported by our estimate of 9.48 × 10-9 point mutations per bp per generation from a 200-year-old low heterozygosity peach to its progeny. From a more highly heterozygous individual derived from an interspecific cross to its selfed progeny, the mutation rate is 1.38 × 10-8 mutations per site per generation, consistent with raised rates in hybrids. Our data thus suggest that (i) peach has an approximately order of magnitude lower mutation rate per unit time than Arabidopsis, consistent with reports of low evolutionary rates in woody perennials, and (ii) hybridization may, indeed, be associated with increased mutation rates as considered over a century ago.

Rapidly evolving R genes in diverse grass species confer resistance to rice blast disease.

We show that the genomes of maize, sorghum, and brachypodium contain genes that, when transformed into rice, confer resistance to rice blast disease. The genes are resistance genes (R genes) that encode proteins with nucleotide-binding site (NBS) and leucine-rich repeat (LRR) domains (NBS-LRR proteins). By using criteria associated with rapid molecular evolution, we identified three rapidly evolving R-gene families in these species as well as in rice, and transformed a randomly chosen subset of these genes into rice strains known to be sensitive to rice blast disease caused by the fungus Magnaporthe oryzae. The transformed strains were then tested for sensitivity or resistance to 12 diverse strains of M. oryzae. A total of 15 functional blast R genes were identified among 60 NBS-LRR genes cloned from maize, sorghum, and brachypodium; and 13 blast R genes were obtained from 20 NBS-LRR paralogs in rice. These results show that abundant blast R genes occur not only within species but also among species, and that the R genes in the same rapidly evolving gene family can exhibit an effector response that confers resistance to rapidly evolving fungal pathogens. Neither conventional evolutionary conservation nor conventional evolutionary convergence supplies a satisfactory explanation of our findings. We suggest a unique mechanism termed "constrained divergence," in which R genes and pathogen effectors can follow only limited evolutionary pathways to increase fitness. Our results open avenues for R-gene identification that will help to elucidate R-gene vs. effector mechanisms and may yield new sources of durable pathogen resistance.

Widely distributed hot and cold spots in meiotic recombination as shown by the sequencing of rice F2 plants.

Numerous studies have argued that environmental variations may contribute to evolution through the generation of novel heritable variations via meiotic recombination, which plays a crucial role in crop domestication and improvement. Rice is one of the most important staple crops, but no direct estimate of recombination events has yet been made at a fine scale. Here, we address this limitation by sequencing 41 rice individuals with high sequencing coverage and c. 900 000 accurate markers. An average of 33.9 crossover (c. 4.53 cM Mb(-1) ) and 2.47 non-crossover events were detected per F2 plant, which is similar to the values in Arabidopsis. Although not all samples in the stress treatment group showed an increased number of crossover events, environmental stress increased the recombination rate in c. 28.5% of samples. Interestingly, the crossovers showed a highly uneven distribution among and along chromosomes, with c. 13.9% of the entire genome devoid of crossovers, including 11 of the 12 centromere regions, and c. 0.72% of the genome containing large numbers of crossovers (> 50 cM Mb(-1) ). The gene ontology (GO) categories showed that genes clustered within the recombination hot spot regions primarily tended to be involved in responses to environmental stimuli, suggesting that recombination plays an important role for adaptive evolution in rapidly changing environments.

Great majority of recombination events in Arabidopsis are gene conversion events.

The evolutionary importance of meiosis may not solely be associated with allelic shuffling caused by crossing-over but also have to do with its more immediate effects such as gene conversion. Although estimates of the crossing-over rate are often well resolved, the gene conversion rate is much less clear. In Arabidopsis, for example, next-generation sequencing approaches suggest that the two rates are about the same, which contrasts with indirect measures, these suggesting an excess of gene conversion. Here, we provide analysis of this problem by sequencing 40 F(2) Arabidopsis plants and their parents. Small gene conversion tracts, with biased gene conversion content, represent over 90% (probably nearer 99%) of all recombination events. The rate of alteration of protein sequence caused by gene conversion is over 600 times that caused by mutation. Finally, our analysis reveals recombination hot spots and unexpectedly high recombination rates near centromeres. This may be responsible for the previously unexplained pattern of high genetic diversity near Arabidopsis centromeres.

Allele-mining of rice blast resistance genes at AC134922 locus.

The AC134922 locus is one of the most rapidly evolving nucleotide binding site-leucine-rich repeat (NBS-LRR) gene family in rice genome. Six rice blast resistance (R) genes have been cloned from this locus and other two resistance candidate genes, Pi34 and Pi47, are also mapped to this complex locus. Therefore, it seems that more functional R genes could be identified from this locus. In this study, we cloned 22 genes from 12 cultivars based on allele-mining strategy at this locus and identified 6 rice blast R genes with 4 of them recognizing more than one isolates. Our result suggests that gene stacking might be the evolutionary strategy for complex gene locus to interact with rapidly evolving pathogens, which might provide a potential way for the cloning of durable resistance genes. Moreover, the mosaic structure and ambiguous ortholog/paralog relationships of these homologous genes, caused by frequent recombination and gene conversion, indicate that multiple alleles of this complex locus may serve as a reservoir for the evolutionary novelty of these R genes.

Single-nucleotide mutation rate increases close to insertions/deletions in eukaryotes.

Mutation hotspots are commonly observed in genomic sequences and certain human disease loci, but general mechanisms for their formation remain elusive. Here we investigate the distribution of single-nucleotide changes around insertions/deletions (indels) in six independent genome comparisons, including primates, rodents, fruitfly, rice and yeast. In each of these genomic comparisons, nucleotide divergence (D) is substantially elevated surrounding indels and decreases monotonically to near-background levels over several hundred bases. D is significantly correlated with both size and abundance of nearby indels. In comparisons of closely related species, derived nucleotide substitutions surrounding indels occur in significantly greater numbers in the lineage containing the indel than in the one containing the ancestral (non-indel) allele; the same holds within species for single-nucleotide mutations surrounding polymorphic indels. We propose that heterozygosity for an indel is mutagenic to surrounding sequences, and use yeast genome-wide polymorphism data to estimate the increase in mutation rate. The consistency of these patterns within and between species suggests that indel-associated substitution is a general mutational mechanism.