The ability of biological systems to adapt to genetic and environmental perturbations is a fundamental but poorly understood process at the molecular level. By quantifying metabolic fluxes and global mRNA abundance, we investigated the genetic and metabolic mechanisms that underlie adaptive evolution of four metabolic gene deletion mutants of Escherichia coli (Δpgi, Δppc, Δpta, and Δtpi) in parallel evolution experiments of each mutant. The initial response to the gene deletions was flux rerouting through local bypass reactions or normally latent pathways. The principal effect of evolution was improved capacity of already active pathways, whereas new flux distributions were not observed. Combinatorial changes in capacity and pathway activation, however, led to different intracellular flux states that enabled evolution in three of the four parallel cases tested. The molecular bases of the evolved phenotypes were then elucidated by global mRNA transcript analyses. Activation of latent pathways and flux changes in the tricarboxylic acid cycle were found to correlate well with molecular changes at the transcriptional level. Flux alterations in other central metabolic pathways, in contrast, were apparently not connected to changes in the transcriptional network. These results give new insight into the dynamics of the evolutionary process by demonstrating the flexibility of the metabolic network of E. coli to compensate for genetic perturbations and the utility of combining multiple high throughput data sets to differentiate between causal and noncausal mechanistic changes. The ability of biological systems to adapt to genetic and environmental perturbations is a fundamental but poorly understood process at the molecular level. By quantifying metabolic fluxes and global mRNA abundance, we investigated the genetic and metabolic mechanisms that underlie adaptive evolution of four metabolic gene deletion mutants of Escherichia coli (Δpgi, Δppc, Δpta, and Δtpi) in parallel evolution experiments of each mutant. The initial response to the gene deletions was flux rerouting through local bypass reactions or normally latent pathways. The principal effect of evolution was improved capacity of already active pathways, whereas new flux distributions were not observed. Combinatorial changes in capacity and pathway activation, however, led to different intracellular flux states that enabled evolution in three of the four parallel cases tested. The molecular bases of the evolved phenotypes were then elucidated by global mRNA transcript analyses. Activation of latent pathways and flux changes in the tricarboxylic acid cycle were found to correlate well with molecular changes at the transcriptional level. Flux alterations in other central metabolic pathways, in contrast, were apparently not connected to changes in the transcriptional network. These results give new insight into the dynamics of the evolutionary process by demonstrating the flexibility of the metabolic network of E. coli to compensate for genetic perturbations and the utility of combining multiple high throughput data sets to differentiate between causal and noncausal mechanistic changes. All biological systems are capable of short term response to environmental changes and, on longer time scales, to evolutionary adaptation. Due to the high growth rates and large numbers of individuals, adaptive evolution of microbes under laboratory conditions rapidly leads to improved growth phenotypes. This principle is exploited for evolutionary engineering (1Sauer U. Adv. Biochem. Eng. Biotechnol. 2001; 73: 129-169PubMed Google Scholar, 2Zelder O. Hauer B. Curr. Opin. Microbiol. 2000; 3: 248-251Crossref PubMed Scopus (32) Google Scholar) and experimental testing of general evolutionary principles (3Fong S.S. Palsson B.O. Nat. Genet. 2004; 36: 1056-1058Crossref PubMed Scopus (247) Google Scholar, 4Cooper T.F. Rozen D.E. Lenski R.E. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1072-1077Crossref PubMed Scopus (338) Google Scholar, 5Adams J. Res. Microbiol. 2004; 155: 311-318Crossref PubMed Scopus (30) Google Scholar). Combining experimental evolution with targeted genetic perturbations then allows one to pose specific questions on robustness and adaptability of biological networks. In response to such externally introduced deletions, microorganisms can invoke a number of different strategies to adjust their functionality. For metabolic networks, these strategies can involve a local bypass of the deleted reaction, complete redirection of flux, reassignment of enzymes to catalyze the deleted reaction, activation of silent genes, or activation of otherwise down-regulated pathways. Whereas all of these mechanisms could potentially cope with genetic perturbations, the exact mechanisms utilized during evolution are poorly understood. Thus, a central question in adaptation and evolution is to determine whether evolving cells refine their existing pathway usage or whether they invoke major metabolic changes such as the activation of latent pathways. The molecular basis of such evolutionary processes is now experimentally traceable with the ability to rapidly improve microbial phenotypes using laboratory evolution (from weeks to a few months) (6Ibarra R.U. Edwards J.S. Palsson B.O. Nature. 2002; 420: 186-189Crossref PubMed Scopus (654) Google Scholar, 7Fong S.S. Marciniak J.Y. Palsson B.O. J. Bacteriol. 2003; 185: 6400-6408Crossref PubMed Scopus (94) Google Scholar, 8Elena S.F. Lenski R.E. Nat. Rev. Genet. 2003; 4: 457-469Crossref PubMed Scopus (956) Google Scholar) coupled with the principal accessibility of the underlying causes through various “omics” methods or genome resequencing (9Honisch C. Raghunathan A. Cantor C.R. Palsson B.O. van den Boom D. Genome Res. 2004; 14: 2495-2502Crossref PubMed Scopus (37) Google Scholar). As a particularly popular tool, simultaneous transcript level monitoring of all genes within the genome by DNA microarray technology was used to identify altered gene expression in evolved Escherichia coli and Saccharomyces cerevisiae strains (4Cooper T.F. Rozen D.E. Lenski R.E. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 1072-1077Crossref PubMed Scopus (338) Google Scholar, 10Ferea T.L. Botstein D. Brown P.O. Rosenzweig R.F. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9721-9726Crossref PubMed Scopus (385) Google Scholar, 11Sonderegger M. Jeppsson M. Hahn-Hägerdal B. Sauer U. Appl. Environ. Microbiol. 2004; 70: 2307-2317Crossref PubMed Scopus (106) Google Scholar). Altered expression levels, however, do not distinguish between cause and effect and thus cannot directly reveal mechanistic links between altered expression and phenotype. In particular, when considering evolution of metabolic functions, more direct information on intracellular flux rerouting would be necessary to reveal the molecular mechanisms that cause a given improved phenotype. Such in vivo reaction rates are accessible through methods of 13C-based metabolic flux analysis (12Sauer U. Curr. Opin. Biotechnol. 2004; 15: 58-63Crossref PubMed Scopus (201) Google Scholar), which have been used successfully to identify functional flux states in various microbes (13Perrenoud A. Sauer U. J. Bacteriol. 2005; 187: 3171-3179Crossref PubMed Scopus (226) Google Scholar, 14Fischer E. Sauer U. Nat. Genet. 2005; 37: 636-640Crossref PubMed Scopus (262) Google Scholar, 15Blank L.M. Kuepfer L. Sauer U. Genome Biol. 2005; 6: R49Crossref PubMed Google Scholar, 16Fuhrer T. Fischer E. Sauer U. J. Bacteriol. 2005; 187: 1581-1590Crossref PubMed Scopus (284) Google Scholar, 17Gunnarsson N. Mortensen U.H. Sosio M. Nielsen J. Mol. Microbiol. 2004; 52: 895-902Crossref PubMed Scopus (26) Google Scholar). Potentially, flux data can fill the gap between the intrinsically noisy and indirect transcriptome, proteome, or metabolome data and the actual phenotype (18Hellerstein M.K. Annu. Rev. Nutr. 2003; 23: 379-402Crossref PubMed Scopus (87) Google Scholar, 19Bailey J.E. Nat. Biotechnol. 1999; 17: 616-618Crossref PubMed Scopus (128) Google Scholar). Thus, the combination of transcript profiles and quantitative intracellular flux data can provide greater insight into biological processes at the molecular level by implicating gene expression changes to altered phenotypes through association with flux growth rates of strains on through adaptive evolution (6Ibarra R.U. Edwards J.S. Palsson B.O. Nature. 2002; 420: 186-189Crossref PubMed Scopus (654) Google Scholar, 7Fong S.S. Marciniak J.Y. Palsson B.O. J. Bacteriol. 2003; 185: 6400-6408Crossref PubMed Scopus (94) Google Scholar, M. Sauer U. Appl. Environ. Microbiol. 2003; PubMed Scopus Google Scholar, M. van Res. 2005; PubMed Scopus Google Scholar), of high growth rates was for metabolic gene deletion mutants of E. coli (3Fong S.S. Palsson B.O. Nat. Genet. 2004; 36: 1056-1058Crossref PubMed Scopus (247) Google Scholar). of evolved mutants that the of different mechanisms during parallel evolution under The molecular however, multiple flux could the improved growth phenotypes. we evolved four E. coli mutants in metabolic and for under growth conditions on these be by at different we used metabolic flux and global gene expression analysis to identify the metabolic mechanisms for the improved phenotypes. and mutants of were (3Fong S.S. Palsson B.O. Nat. Genet. 2004; 36: 1056-1058Crossref PubMed Scopus (247) Google Scholar), and the and mutants were by gene deletions by the Proc. Natl. Acad. Sci. U. S. A. 2000; PubMed Scopus Google Scholar) with the E. coli The and were used to the with the and was by with of deletion mutants of E. coli was in of with of in using for at of of of of and of The were and then of of of of and of a of of of of and of the of initial of each were in to for adaptive evolution experiments were the deletion mutant. In evolution cells were and to growth with at by into The at each was to for changes in growth The was at This process of growth and was for for the mutants for the mutants and for the and the and evolution experiments were (3Fong S.S. Palsson B.O. Nat. Genet. 2004; 36: 1056-1058Crossref PubMed Scopus (247) Google Scholar). This process of evolution in evolved mutants with as and The number of was on a basis by the of each and during growth into were evolved a growth was for more This process of a of growth that were evolved under were used to of at and were used to that were for of for or of were in and on a at and For was as the or as a of and growth was by the and were using or in were by high analysis at a of using a at and a of acid at a of The were during the growth as U. J. M. T. J.E. J. Bacteriol. 1999; PubMed Google growth on specific and specific using a of of and Flux used metabolic flux by for analysis were as E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar). of were during the growth were in at for in The were under a of at and then at in of and of with for E. N. Sauer U. Biochem. 2004; PubMed Scopus Google Scholar). were on a with The distributions of were then for E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar). The distributions were to the in vivo metabolic with and data which of fluxes through reactions and pathways to the of intracellular E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar). In experiments with in of but in the and The flux of analysis E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar) would to the This however, would at the of The normally which to in contrast, would such at the of A. PubMed Scopus Google Scholar). the of through the bypass through we the of which is to the of that of are The which to a and one was the of bypass to the of and are used as The of was not however, the for be to the one of of be the the of can be and can be used to determine This however, the of through the the pathway into we The of through the pathway was the flux and the of and the of a with one as the of of the through the and the through of the pathway for the of in was by the of a combination of and The of was and the process was for of the of and of through the pathway were The for these flux were all in the have E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar), normally for these were using the with the that the of the of a is to These new were used to determine the by the process in a The and the for the were these The were within of the and the was used as the for the in the of in the the used to the flux through the pathway and through pathway is not a was used to determine the of the pathway to In such a all through the pathway a at whereas the pathway the of the pathway to can be as of the of the pathway to and is the of the of the pathway to The on was using E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar). Flux fluxes were with a that all major pathways of central E. N. Sauer U. Biochem. 2004; PubMed Scopus Google Scholar). For the the was with the bypass on the The network was to the one in For all the deleted reactions were in the network to for their in vivo for the by for the mutants was the reaction to the analysis a complete of tricarboxylic acid cycle The reaction for the different of fluxes and the three experimentally rates of and of with of the flux were used as as was E. N. Sauer U. Biochem. 2004; PubMed Scopus Google through the through the and and through the pathway The four were used as whereas the were used as on the the was in the and the was as A. and U. for The through the pathway and through the bypass were used as for the using the the of through the and the of through the into were the for M. of the Scholar) and the growth and M. M. A. J. M. T. J.E. Sauer U. J. Bacteriol. 2002; PubMed Scopus Google Scholar). The of the of the and flux was using the The were by through the experimental E. N. Sauer U. Biochem. 2004; PubMed Scopus Google Scholar). The was at with initial flux distributions to of the global and the to the of the was used to all metabolic flux and fluxes N. Fischer E. Sauer U. 2005; 6: PubMed Scopus Google Scholar). mRNA E. coli genome were used for all transcriptional analyses. experimental was in using and the of the on were used for the were to growth of was to of and was using the were using a and was by on and by the and were as by were using B. U. 2003; 4: PubMed Scopus Google Scholar) for and of were then for expression using between evolved mutants and genes a were as changes in gene This in of of genes for each evolution mutant. These of genes were then into J. S. J. S. M. Res. 2005; PubMed Scopus Google Scholar) for The of the of genes was using the D. L. M. 2004; PubMed Scopus Google Scholar), to the number of E. coli genes on the is the number of genes that are a of the is the number of genes, and is the number of genes that are and a of the In for in were of were in of and three through a and at In of the bypass was by for and coupled for and For the of T. A. Biol. Scholar) was and of of The for was to the with the the of of For the of and Biochem. J. PubMed Scopus Google Scholar) was and of of For all three the reaction was by the of or and of and of was in the and The for and were and The in was with the of identify the metabolic by adaptive evolution to cope with a gene we four gene that the flux at of and parallel of each were evolved for under conditions of growth in with The of evolution were as for and these were as and to the and evolution experiments (3Fong S.S. Palsson B.O. Nat. Genet. 2004; 36: 1056-1058Crossref PubMed Scopus (247) Google Scholar), the and mutants rapidly evolved improved phenotypes the specific growth and but all evolved mutants the rates at the of Whereas all mutants evolved to improved parallel evolved different the evolved mutants and different and and the and mutants in their rates Thus, the evolutionary are and with to the for high growth but the underlying states are not the growth of evolved and during growth on The parallel evolved mutants are by the and of for for experiments in a new Flux are the actual metabolic mechanisms by adaptive and do all mutants evolved in parallel on the mechanisms to cope with a given these intracellular fluxes were experiments E. N. Sauer U. Biochem. 2004; PubMed Scopus Google Scholar). As E. Sauer U. J. Biochem. 2003; PubMed Scopus Google Scholar, U. S. A. Fischer E. J. Biol. 2004; PubMed Scopus Google Scholar, C. T. J. Bacteriol. 2003; 185: PubMed Scopus Google Scholar), in the is the to the and pathways to bypass the of the otherwise in the was with C. T. J. Bacteriol. 2003; 185: PubMed Scopus Google Scholar, E. Sauer U. J. Biol. 2003; PubMed Scopus Google Scholar). Whereas the altered flux was evolution more the flux level to of the In to the initial through the pathway in evolved mutants In the of however, the evolved network Whereas the flux was to the with active and the was more to the with and tricarboxylic acid cycle flux of the Thus, the mutants evolved to improved phenotypes through different intracellular flux these flux the deleted the for metabolic in to the flux of through the in the of the cells U. S. A. Fischer E. J. Biol. 2004; PubMed Scopus Google Scholar). in to the the to to the of the in the the the to growth on as the reaction tricarboxylic acid cycle that are for U. B. Microbiol. Rev. 2005; PubMed Scopus Google Scholar). mutants however, and the growth phenotype of the was through adaptive evolution with in the evolved mutants In all the normally the of as was L. Microbiol. 2004; PubMed Google Scholar). The flux through the however, was in of the for the and In the and the flux was through the cycle with the and as reactions E. Sauer U. J. Biol. 2003; PubMed Scopus Google Scholar). to cycle in all three mutants by to Thus, the latent the and was deletion of of the mutants led to in the metabolic network by using or as was necessary to the through the active the in mutants was by of As been for different of were D.E. S. J.S. J. Bacteriol. 1999; PubMed Google Scholar). the phenotype of the was otherwise to the growth and network were in the evolved mutants and the flux states were different in the evolved mutants with fluxes and in the mutant. The tricarboxylic acid cycle and pathway which been for a D.E. J.E. T. Mol. Microbiol. 2003; PubMed Scopus Google Scholar), were not in the flux states of the of the in the a of the flux into and of mutants to through the normally bypass A. PubMed Scopus Google Scholar, S. Microbiol. PubMed Scopus Google Scholar) the data provide indirect for bypass activation was through in with the the in in the of the bypass more the flux level to of the This is a direct of improved bypass fluxes through the with in that the is more for fluxes through the bypass the reactions in the Thus, the normally latent bypass the introduced Whereas the evolution led to improved growth through of metabolic of the introduced to different to the of which in in of mutants and and and and and experiments in a new mRNA the flux data metabolic mechanisms that cope with and the changes by they cannot reveal the actual genetic In to the molecular mechanisms for the changes in network during we profiles of the evolved and mutants that major flux All results and were to expression changes by for a of the with the evolved genes were All changes in gene expression were with the flux changes through the reaction by of evolved expression and flux changes the In the and fluxes genes and fluxes genes, when with the between flux and expression changes in and of these genes for mutants and and the in expression of expression of the genes and was and with the flux in with the in phenotype and flux a large number of expression were found between the mutants in the of central In particular, expression of the tricarboxylic acid cycle genes and and, of all enzymes in was with the flux expression of the gene with the in and in expression was in the with Thus, is a genetic basis for the tricarboxylic acid cycle and in the and the of changes in the mutant. E. coli a and a which are used to the and J. Biochem. J. PubMed Scopus Google Scholar, S. J. Biochem. PubMed Scopus Google Scholar, B. J. Bacteriol. 1999; PubMed Google Scholar). The by to whereas the by to and in of the for of the U. S. A. Fischer E. J. Biol. 2004; PubMed Scopus Google Scholar). changes for the with the flux for evolved the expression of the a in was in to the whereas the expression of the but was for that the to In evolved altered expression of genes was with flux changes and to mutants was altered expression in the of central with expression of and the tricarboxylic acid cycle genes and and expression of and the gene Thus, activation of the and tricarboxylic acid cycle fluxes were directly through altered expression of genes in these pathways. In the and fluxes genes and fluxes genes, when with the between expression and flux however, was for genes of which was expression of and four genes in and the major flux in the mutants with the activation of the normally latent was mutants expression of This is by the in of the The and tricarboxylic acid fluxes were not by changes in gene fluxes to have a genetic we were to whether was to a mechanistic the genes or other expression changes were in genes or in a we for genes with expression changes in the in the and mutants with active of genes was and expression of genes was in the and mutants but not in the other three mutants these genes were in metabolic pathways that the tricarboxylic acid expression of with and and expression in of the pathways for and In was in the mutants with expression of the and genes that are in the tricarboxylic acid cycle This of expression changes one or more in the transcriptional network that expression of the and other Whereas the active was for the whether the changes to the phenotypes. in transcriptional were investigated by using expression changes in each to the that a in a for each evolved were for using the at a of for changes that could to we found that evolved mutants of the with and that all three mutants the and down-regulated the with Whereas of the and is that a in the network was to the acid with other biological In the of the of to could a in to to other metabolic In the of mutants the activation of the cycle E. Sauer U. J. Biol. 2003; PubMed Scopus Google Scholar) was which to and thus could the of Thus, in adaptive mechanisms in the transcriptional network that are connected to the changes in the metabolic changes in the and using the of expression changes within each The of the expression level is given as the changes for each gene within the are given as in the in a new strategies can be in response to externally introduced genetic perturbations that genetic robustness to the network. The initial response to the four investigated deletions was a local rerouting of fluxes the which activation of latent pathways in were for was not a evolution a new activation of a silent gene or reassignment of the effect of evolution the time was in the capacity of already active pathways. As a the fluxes in mutants that were by the and In the metabolic response in the which the activation of pathways that are or such as the or the was during adaptive are the of in and the evolved mutants and the of local flux rerouting through the pathway in the evolved the by E. coli in response to externally introduced genetic is not as by the parallel evolved mutants that in reactions and tricarboxylic acid cycle Whereas are a number of for to compensate for a gene one is that introduced genetic a of fluxes through the network as and evolution would then a process of initial D. D. Proc. Natl. Acad. Sci. U. S. A. 2002; PubMed Scopus Google Scholar). results but we were to large in the flux states of parallel evolved This to the robustness of the metabolic network and ability to different to not but to improve growth These results provide experimental to results that metabolic have different of Eng. 2003; PubMed Scopus Google Scholar). is that biological and in the of these flux but is these are into evolutionary and is not The of molecular fluxes is by multiple mechanisms at that gene and determine whether transcriptional were the cause of the altered flux gene expression were with flux particularly high was found for the complete activation or of latent pathways, such as the and the In expression of at one gene with the flux changes through these pathways. flux changes in the tricarboxylic acid cycle with altered gene the fluxes in of the tricarboxylic acid cycle for and evolved The of in the mutants be to the changes in tricarboxylic acid cycle fluxes for the other three and fluxes were and was found for the three pathways of initial not in the mutants with pathway and expression with the pathway flux in evolved and tricarboxylic acid cycle flux changes to be at in at the transcriptional level. Flux through the and pathways and in contrast, were not at the level. a between gene expression and metabolic fluxes for the been in S. whereas tricarboxylic acid cycle and pathway fluxes van J. Biol. 2004; PubMed Scopus Google Scholar). of flux and gene expression for E. coli under conditions on and a of for the however, in to the pathway a expression Biotechnol. 2002; PubMed Scopus Google Scholar). The of in data be to the changes in pathway gene expression changes are not in the such as or changes in not be in mRNA transcript Thus, the analysis of gene expression data with flux data is one of mRNA transcript and changes that be causal to phenotypes. In the of the gene expression data in the of flux to on a number of expression changes of the of expression changes. E. coli is to rapidly to growth a phenotype initial genetic perturbations by of existing pathways parallel to the This can be to in the central of E. a network of reactions but for with and for in of the network to capacity of the parallel pathways to be to the pathways used in the however, metabolic are to refine usage of the new pathways. These could be to intracellular new pathway usage and was that strains evolved in parallel utilized different of to large in flux states and we have found that the metabolic network of E. coli is in response to a genetic with metabolic in initial flux rerouting to compensate for the and a to flux analysis of multiple data was to mechanistic of these with
No takes yet. Share an insight, caveat, or question.
Fong et al. (2005) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: