<\/p>\n DISRUPTIVE INSERTION MUTATIONS HAVE GREATER DELETERIOUS EFFECTS ON MORE-FIT BACKGROUNDS, CALLED INCREASING COST EPISTASIS. BY CONTRAST, MOST ADAPTATION EXPERIMENTS SHOW A PATTERN OF DIMINISHING RETURNS EPISTASIS, WHEREBY MUTATIONS CONFER SMALLER BENEFIT AS FITNESS INCREASES. THE DIFFERENT CURVATURES MAY REFLECT DIFFERENCES IN HOW DELETERIOUS AND BENEFICIAL MUTATIONS ARE SAMPLED.<\/p>\n <\/p>\n <\/p>\n The modern framing of adaptation imagines genotypes as hyperdimensional Cartesian coordinates and fitness as the elevation above them. This gives rise to a fitness landscape that a population must traverse if it is to adapt (see the figure). The landscape metaphor aids visualization of the trade-off. The capacity for adaptive evolution is determined by the curvature of the uphill terrain. This curvature results from how mutations interact, called epistasis. A wealth of microbial evolution experiments (2<\/em><\/a>\u20136<\/em><\/a>) suggest that the uphill curvature is usually concave, giving rise to \u201cdiminishing returns epistasis\u201d as fitness increases. Thus, the fitness effect of a beneficial mutation is not the same across backgrounds; it gets smaller as background fitness increases. The cost of mutations among offspring is also related to curvature; if the vicinity around a genotype is concave in all directions, the costs will be small and the genotype is considered robust. High robustness occurs when mutations have little effect on phenotype (7<\/em><\/a>). Compared to the uphill terrain, the curvature of the downslope terrain has been less well studied in the context of adaptation.<\/p>\n Johnson\u00a0et al.<\/em>\u00a0sought to characterize the landscape around genetically distinct variants by estimating the fitness effect of insertion mutations in yeast (Saccharomyces cerevisiae<\/em>). They found that most mutations were deleterious (that is, they reduced fitness) and had larger deleterious effects on higher-fitness backgrounds. They called this pattern \u201cincreasing cost epistasis.\u201d<\/p>\n Examining mutations individually, they found that most follow the pattern of increasing cost epistasis. A substantial minority, however, either showed the opposite pattern (diminishing costs) or no trend at all. The same backgrounds were previously the subject of quantitative trait loci (QTL) mapping (8<\/em><\/a>), whereby genetic differences at tens of thousands of sites are used to find genome locations associated with measurable (quantitative) traits. This allowed the authors to ask how well a QTL model explains fitness effects and compare it to models that include background fitness or both. They found that the QTL model is best for some mutations, the background fitness model is best for others, and, for most, the best model includes both QTLs and background fitness. That QTLs were not consistently the best explanation is surprising because they were very good at explaining variation in quantitative traits (8<\/em><\/a>). The implication is that although there is an overarching pattern of increasing cost epistasis, there is also interesting variation at the level of individual mutations. This also suggests that models with mechanistic underpinnings (6<\/em><\/a>,\u00a09<\/em><\/a>) may be more successful than simple models of epistasis (10<\/em><\/a>,\u00a011<\/em><\/a>) in explaining adaptive evolution.<\/p>\n The results of Johnson\u00a0et al.<\/em>\u00a0are unexpected because they imply different curvatures of the fitness landscape depending on what type of mutations are tested: Random insertion mutations follow a convex function on fitness, whereas successful beneficial mutations during adaptation follow a concave one (see the figure). Although epistasis scales with fitness in both cases, the relationship is in opposite directions. How can this be? The authors argue that increasing cost epistasis may arise from metabolic flux, whereby a deleterious mutation in a sequential pathway has lesser negative consequences when other enzymes in the pathway have already been adversely affected. If low-fitness backgrounds have less-functional metabolic pathways, disruptions through mutation can do less damage. But this argument, when reversed, is problematic. It implies that beneficial mutations on higher-flux pathways should generate synergies (or convex surfaces)\u2014such mutations are observed (2<\/em><\/a>), but rarely. Part of the explanation may be that some adaptive mutations in experimental evolution succeed by disrupting expendable pathways; once such pathways are disrupted, further mutations will be of diminishing benefit. Indeed, beneficial disruptive mutations are not uncommon (5<\/em><\/a>,\u00a012<\/em><\/a>).<\/p>\n The different curvatures may also reflect how mutations are sampled. Insertion mutations disrupt random pathways; if disrupting a pathway is usually detrimental, it will tend to generate increasing cost epistasis. To quote the singer-songwriter Townes Van Zandt, \u201cyou don’t need no engine to go downhill.\u201d Conversely, adaptation is biased. When disruptive mutations can confer fitness gains, they will be selected\u2014even if they return only diminishing benefit.<\/p>\n Returning to Darwin’s insight and the question of trade-offs in adaptation, the study of Johnson\u00a0et al.<\/em>\u00a0suggests that as organisms adapt through natural selection of beneficial mutations, they will concurrently suffer an escalating burden of producing less-fit offspring. The authors speculate that increasing costs, paired with the reality of diminishing gains, may arrest adaptation before a fitness optimum is reached. An important test of the authors’ claim will be to assess if increasing cost epistasis is also observed during an adaptive walk, where the genotypes of increasing fitness differ by a far smaller number of mutations. Moreover, it is unclear how often mutation rates will be high enough and beneficial effects small enough for the forces to counterbalance. There is also accumulating evidence that some biological features\u2014gene expression patterns and protein stabilization by chaperones, for example\u2014are robust, meaning they have apparently escaped the treacheries of increasing cost epistasis (7<\/em><\/a>). This raises the question, how have they done so?<\/p>\n <\/p>\n <\/p>\n![](\"https:\/\/science.sciencemag.org\/content\/sci\/366\/6464\/418\/F1.medium.gif\")
<\/span><\/a><\/div>\n\n
GRAPHIC: C. BICKEL\/SCIENCE<\/em><\/q><\/p>\n
![<\/i>\u00a0Download high-res image<\/span><\/a><\/li>\n<\/i>\u00a0Open in new tab<\/span><\/a><\/li>\n<\/i>\u00a0Download Powerpoint<\/span><\/a><\/li>\n<\/ul>\n<\/div>\n<\/div>A FITNESS LANDSCAPE WITH TWO TYPES OF EPISTASIS<\/span><\/p>\nDISRUPTIVE INSERTION MUTATIONS HAVE GREATER DELETERIOUS EFFECTS ON MORE-FIT BACKGROUNDS, CALLED INCREASING COST EPISTASIS. BY CONTRAST, MOST ADAPTATION EXPERIMENTS SHOW A PATTERN OF DIMINISHING RETURNS EPISTASIS, WHEREBY MUTATIONS CONFER SMALLER BENEFIT AS FITNESS INCREASES. THE DIFFERENT CURVATURES MAY REFLECT DIFFERENCES IN HOW DELETERIOUS AND BENEFICIAL MUTATIONS ARE SAMPLED.<\/p>\nGRAPHIC: C. BICKEL\/SCIENCE<\/em><\/q><\/p>\n<\/div>\n<\/figcaption><\/figure>\n <\/p>\n <\/p>\nThe modern framing of adaptation imagines genotypes as hyperdimensional Cartesian coordinates and fitness as the elevation above them. This gives rise to a fitness landscape that a population must traverse if it is to adapt (see the figure). The landscape metaphor aids visualization of the trade-off. The capacity for adaptive evolution is determined by the curvature of the uphill terrain. This curvature results from how mutations interact, called epistasis. A wealth of microbial evolution experiments (2<\/em><\/a>\u20136<\/em><\/a>) suggest that the uphill curvature is usually concave, giving rise to \u201cdiminishing returns epistasis\u201d as fitness increases. Thus, the fitness effect of a beneficial mutation is not the same across backgrounds; it gets smaller as background fitness increases. The cost of mutations among offspring is also related to curvature; if the vicinity around a genotype is concave in all directions, the costs will be small and the genotype is considered robust. High robustness occurs when mutations have little effect on phenotype (7<\/em><\/a>). Compared to the uphill terrain, the curvature of the downslope terrain has been less well studied in the context of adaptation.<\/p>\nJohnson\u00a0et al.<\/em>\u00a0sought to characterize the landscape around genetically distinct variants by estimating the fitness effect of insertion mutations in yeast (Saccharomyces cerevisiae<\/em>). They found that most mutations were deleterious (that is, they reduced fitness) and had larger deleterious effects on higher-fitness backgrounds. They called this pattern \u201cincreasing cost epistasis.\u201d<\/p>\nExamining mutations individually, they found that most follow the pattern of increasing cost epistasis. A substantial minority, however, either showed the opposite pattern (diminishing costs) or no trend at all. The same backgrounds were previously the subject of quantitative trait loci (QTL) mapping (8<\/em><\/a>), whereby genetic differences at tens of thousands of sites are used to find genome locations associated with measurable (quantitative) traits. This allowed the authors to ask how well a QTL model explains fitness effects and compare it to models that include background fitness or both. They found that the QTL model is best for some mutations, the background fitness model is best for others, and, for most, the best model includes both QTLs and background fitness. That QTLs were not consistently the best explanation is surprising because they were very good at explaining variation in quantitative traits (8<\/em><\/a>). The implication is that although there is an overarching pattern of increasing cost epistasis, there is also interesting variation at the level of individual mutations. This also suggests that models with mechanistic underpinnings (6<\/em><\/a>,\u00a09<\/em><\/a>) may be more successful than simple models of epistasis (10<\/em><\/a>,\u00a011<\/em><\/a>) in explaining adaptive evolution.<\/p>\nThe results of Johnson\u00a0et al.<\/em>\u00a0are unexpected because they imply different curvatures of the fitness landscape depending on what type of mutations are tested: Random insertion mutations follow a convex function on fitness, whereas successful beneficial mutations during adaptation follow a concave one (see the figure). Although epistasis scales with fitness in both cases, the relationship is in opposite directions. How can this be? The authors argue that increasing cost epistasis may arise from metabolic flux, whereby a deleterious mutation in a sequential pathway has lesser negative consequences when other enzymes in the pathway have already been adversely affected. If low-fitness backgrounds have less-functional metabolic pathways, disruptions through mutation can do less damage. But this argument, when reversed, is problematic. It implies that beneficial mutations on higher-flux pathways should generate synergies (or convex surfaces)\u2014such mutations are observed (2<\/em><\/a>), but rarely. Part of the explanation may be that some adaptive mutations in experimental evolution succeed by disrupting expendable pathways; once such pathways are disrupted, further mutations will be of diminishing benefit. Indeed, beneficial disruptive mutations are not uncommon (5<\/em><\/a>,\u00a012<\/em><\/a>).<\/p>\nThe different curvatures may also reflect how mutations are sampled. Insertion mutations disrupt random pathways; if disrupting a pathway is usually detrimental, it will tend to generate increasing cost epistasis. To quote the singer-songwriter Townes Van Zandt, \u201cyou don’t need no engine to go downhill.\u201d Conversely, adaptation is biased. When disruptive mutations can confer fitness gains, they will be selected\u2014even if they return only diminishing benefit.<\/p>\nReturning to Darwin’s insight and the question of trade-offs in adaptation, the study of Johnson\u00a0et al.<\/em>\u00a0suggests that as organisms adapt through natural selection of beneficial mutations, they will concurrently suffer an escalating burden of producing less-fit offspring. The authors speculate that increasing costs, paired with the reality of diminishing gains, may arrest adaptation before a fitness optimum is reached. An important test of the authors’ claim will be to assess if increasing cost epistasis is also observed during an adaptive walk, where the genotypes of increasing fitness differ by a far smaller number of mutations. Moreover, it is unclear how often mutation rates will be high enough and beneficial effects small enough for the forces to counterbalance. There is also accumulating evidence that some biological features\u2014gene expression patterns and protein stabilization by chaperones, for example\u2014are robust, meaning they have apparently escaped the treacheries of increasing cost epistasis (7<\/em><\/a>). This raises the question, how have they done so?<\/p>\n <\/p>\n <\/p>\n(\uc6d0\ubb38: \uc5ec\uae30<\/a>\ub97c \ud074\ub9ad\ud558\uc138\uc694~)<\/p>\n <\/p>\n <\/p>\n <\/p>\n <\/p>\n","protected":false},"excerpt":{"rendered":" One of Darwin’s great insights was that he took the widespread observation that organisms are exceptionally well-suited to their environment and turned it(more…)<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[43,38,32,33,29],"tags":[],"class_list":["post-4586","post","type-post","status-publish","format-standard","hentry","category-08----09---","category-38","category-essays-on-science","category-do-biology","category-lets-do-science"],"aioseo_notices":[],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack-related-posts":[{"id":1388,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=1388","url_meta":{"origin":4586,"position":0},"title":"(youtube)\ubaa8\ub4e0 \uc0dd\uba85\uccb4\uc758 \uae30\uc6d0\uc740 \ub2e8 \ud558\ub098\uc758 \uc6d0\uc2dc \uc138\ud3ec\uc774\ub2e4?","author":"biochemistry","date":"August 17, 2018","format":false,"excerpt":"\u00a0 https:\/\/www.youtube.com\/watch?v=cgP10pnHqhU","rel":"","context":"In "'08. \uc0dd\uba85\uccb4\uc758 \uae30\uc6d0\uacfc \uc18d\uc131'\uacfc '09. \uc0dd\uba85\uccb4\uc758 \uc5f0\uc18d\uc131\uacfc \uc720\uc804' \uad00\ub828"","block_context":{"text":"'08. \uc0dd\uba85\uccb4\uc758 \uae30\uc6d0\uacfc \uc18d\uc131'\uacfc '09. \uc0dd\uba85\uccb4\uc758 \uc5f0\uc18d\uc131\uacfc \uc720\uc804' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=43"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/cgP10pnHqhU\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":705,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=705","url_meta":{"origin":4586,"position":1},"title":"Gene to Protein (youtube)","author":"biochemistry","date":"June 4, 2018","format":false,"excerpt":"\u00a0 \u00a0 Gene to Protein \u00a0 https:\/\/www.youtube.com\/watch?v=gG7uCskUOrA \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0","rel":"","context":"In "'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828"","block_context":{"text":"'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=41"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/gG7uCskUOrA\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":707,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=707","url_meta":{"origin":4586,"position":2},"title":"lac Operon (youtube)","author":"biochemistry","date":"June 4, 2018","format":false,"excerpt":"\u00a0 \u00a0 lac Operon \u00a0 https:\/\/www.youtube.com\/watch?v=oBwtxdI1zvk \u00a0 \u00a0 or click here~ \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0","rel":"","context":"In "'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828"","block_context":{"text":"'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=41"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/oBwtxdI1zvk\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":709,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=709","url_meta":{"origin":4586,"position":3},"title":"Gene Expression (youtube)","author":"biochemistry","date":"June 4, 2018","format":false,"excerpt":"\u00a0 \u00a0 Gene Expression \u00a0 https:\/\/www.youtube.com\/watch?v=OEWOZS_JTgk or click here~ \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0","rel":"","context":"In "'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828"","block_context":{"text":"'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=41"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/OEWOZS_JTgk\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":715,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=715","url_meta":{"origin":4586,"position":4},"title":"Polymerase Chain Reaction (PCR) (youtube)","author":"biochemistry","date":"June 4, 2018","format":false,"excerpt":"\u00a0 \u00a0 PCR \u00a0 https:\/\/www.youtube.com\/watch?gl=KR&hl=ko&v=_YgXcJ4n-kQ or click here~ \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0","rel":"","context":"In "'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828"","block_context":{"text":"'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=41"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]},{"id":717,"url":"https:\/\/biochemistry.khu.ac.kr\/lab\/?p=717","url_meta":{"origin":4586,"position":5},"title":"DNA Fingerprinting (youtube)","author":"biochemistry","date":"June 4, 2018","format":false,"excerpt":"\u00a0 \u00a0 DNA Fingerprinting \u00a0 https:\/\/www.youtube.com\/watch?gl=KR&hl=ko&v=drC7rR7CIBg&feature=related or click here~ \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0","rel":"","context":"In "'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828"","block_context":{"text":"'05. \ubb3c\uc9c8\uc758 \uc9c4\ud654' \uad00\ub828","link":"https:\/\/biochemistry.khu.ac.kr\/lab\/?cat=41"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/img.youtube.com\/vi\/\/0.jpg?resize=350%2C200","width":350,"height":200},"classes":[]}],"jetpack_sharing_enabled":false,"jetpack_shortlink":"https:\/\/wp.me\/p9Xo1j-1bY","_links":{"self":[{"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/posts\/4586","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4586"}],"version-history":[{"count":1,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/posts\/4586\/revisions"}],"predecessor-version":[{"id":4587,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=\/wp\/v2\/posts\/4586\/revisions\/4587"}],"wp:attachment":[{"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4586"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4586"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/biochemistry.khu.ac.kr\/lab\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4586"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}](\"https:\/\/science.sciencemag.org\/content\/sci\/366\/6464\/418\/F1.large.jpg?download=true\"){kind=link}