MBE Advance Access originally published online on January 13, 2007
Molecular Biology and Evolution 2007 24(4):900-908; doi:10.1093/molbev/msm006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2007 The Authors.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Research Articles |
Compensatory Evolution in Response to a Novel RNA Polymerase: Orthologous Replacement of a Central Network Gene
,
* Section of Integrative Biology, University of Texas
Institute for Cellular and Molecular Biology, University of Texas
Section of Molecular Genetics and Microbiology, University of Texas
E-mail: bull{at}bull.biosci.utexas.edu.
 |
Abstract |
|---|
 TOP Abstract IntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
A bacteriophage genome was forced to evolve a new system of regulation by replacing its RNA polymerase (RNAP) gene, a central component of the phage developmental pathway, with that of a relative. The experiment used the obligate lytic phage T7 and the RNAP gene of phage T3. T7 RNAP uses 17 phage promoters, which are responsible for all middle and late gene expression, DNA replication, and progeny maturation, but the enzyme has known physical contacts with only 2 other phage proteins. T3 RNAP was supplied in trans by the bacterial host to a T7 genome lacking its own RNAP gene and the phage population was continually propagated on naive bacteria throughout the adaptation. Evolution of the T3 RNAP gene was thereby prevented, and selection was for the evolution of regulatory signals throughout the phage genome. T3 RNAP transcribes from T7 promoters only at low levels, but a single mutation in the promoter confers high expression, providing a ready mechanism for reevolution of gene expression in this system. When selected for rapid growth, fitness of the engineered phage evolved from a low of 5 doublings/h to 33 doublings/h, close to the expected maximum of 37 doublings/h. However, the experiment was terminated before it could be determined accurately that fitness had reached an obvious plateau, and it is not known whether further adaptation could have resulted in complete recovery of fitness. More than 30 mutations were observed in the evolved genome, but changes were found in only 9 of the 16 promoters, and several coding changes occurred in genes with no known contacts with the RNAP. Surprisingly, the T7 genome adapted to T3 RNAP also maintained high fitness when using T7 RNAP, suggesting that the extreme incompatibility of T7 elements with T3 RNAP is not an invariant property of divergence in these expression systems.
Key Words: experimental • evolution • bacteriophage • genomics • adaptation
|
Introduction |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
The tools of molecular biology are beginning to provide insight into how the different elements of a genome function and evolve to produce various phenotypes. A still novel arena for understanding genomes is offered by a combination of biochemical engineering and experimental evolution. Any genome that is engineered to alter function is likely to be at least partly maladapted at the outset. The engineering may be a deliberate attempt to be disruptive (e.g., create an attenuated strain of a pathogenic organism), or it may be an attempt to create a more fit organism but is designed with imperfect knowledge. In either case, short-term adaptation of the modified genome may select compensatory changes that improve fitness by correcting the defects. The nature of those compensatory changes as well as the overall fitness improvement provide insights into genomic interactions, evolutionary robustness of the genome, and possibly the genome's evolutionary history.
One obvious question from an engineering perspective is how much a genome's fitness can recover by compensatory evolution. If fitness is seriously impaired by a genomic modification, can the genome evolve on its own to fix those problems or will further engineering be necessary? We adopt that perspective here to address the evolution of genome regulation in the obligate lytic bacteriophage T7. This phage and its relatives encode an RNA polymerase (RNAP), which is responsible for expression of all middle and late genes and which also interacts with at least 2 other phage proteins (Zhang and Studier 1995
). Counting each of the 17 phage promoters separately, the T7 RNAP has more known interactions with phage elements than any other phage-encoded protein and is thus the most highly connected protein within the genome network. The experimental design used here rests on the observation that the phage RNAP of T3, a close relative of T7, transcribes from T7 promoters at very low levels (Golomb and Chamberlin 1977; Bailey et al. 1983; Klement et al. 1990). By replacing T7 RNAP with T3 RNAP in an otherwise purely T7 genome, phage growth should be severely debilitated until high levels of expression reevolve across the genome. The most obvious evolutionary pathway for such a phage is evolution of the T3 RNAP to recognize T7 promoters, which can be accomplished with a single amino acid change (Raskin et al. 1992). To avoid that outcome and to force a genome-wide reevolution of regulation, we supply the RNAP in trans to a T7 phage deleted for its RNAP gene, T7
1 (fig. 1). Placement of the RNAP gene within the bacterial host, which is killed by infection, prevents evolution of the RNAP gene the host supplies. Therefore in this study, the only available mechanism to increase the level of transcription by T3 RNAP is through compensatory evolution of phage elements involved in transcriptional regulation.
|
The replacement of a highly connected gene with an ortholog that provides minimal complementation might seem, a priori, to be a profound challenge to evolution. With 18% amino acid sequence difference between T7 and T3 RNAP, and with the protein itself having many interactions with phage elements, reevolution of high fitness might require multiple mutations simultaneously to avoid low-fitness intermediates. Interactions of the RNAP with phage elements may be unusual, however, and may be especially amenable to compensatory evolution to high fitness. Promoters interact primarily with an RNAP. Therefore a mutation in a promoter, which in T7 is usually not in coding DNA, should just affect the rate of transcription initiation. In addition, the T7–T3 RNAP system is favorably disposed to compensatory evolution: a single base change at one position in the T7 consensus promoter sequence allows significant activity by T3 RNAP in vitro (Klement et al. 1990
). Thus, when T7 is forced to utilize T3 RNAP, the phage may be able to recover much of its fitness via stepwise single base promoter changes. However, at this stage any predictions about compensatory evolution and fitness improvement in this system are potentially naive. The progressive benefit of an accumulation of promoter changes will be affected by possible indirect interactions between different promoters via overlapping and polycistronic transcripts. Mutations that increase expression levels from nearest-neighbor promoters may interfere with each other by causing overexpression of genes transcribed from both. Perhaps more problematic is that the phage RNAP directly interacts with other phage proteins, and those interactions may not be fully improved even by a series of single mutational steps. There is thus only a moderately informed basis for predicting evolution; it is some of those predictions that are experimentally tested in this study.
The outcome of experimental adaptations may also help answer a basic question about the evolution of coupled systems. In particular, how does an RNAP gene and its set of cognate promoters coevolve so that proper regulation is maintained while the parental and daughter lineages diverge so they are no longer compatible?
|
Methods and Materials |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
Bacterial and phage strains and the plasmids used are shown in table 1. To create T3
1, a wild-type T3 was grown on IJ1949(pRS-1) and the lysate was then plated on IJ1949 to select phages that had recombined trxA into gene 1. Phages were subsequently screened for dependence of growth on host-supplied T3 RNAP. In order to remove trxA, phage DNA (gene 1::trxA) was digested with EcoRI; the 2 large fragments containing only phage DNA were isolated, ligated, and transfected into IJ1126(pCM56). T31 grows on LM-R (T3 gene 1+ and trxA+), but not on IJ1949 (T3 gene 1+ and trxA-) or BL21 (T3 gene 1- and trxA+).
|
Propagation
Selection was for rapid phage growth. Frozen aliquots of cells were thawed and inoculated into 10 ml Luria-Bertani broth (10 g NaCl, 10 g Bacto Tryptone, and 5 g Bacto yeast extract per liter) at 37 °C with aeration to achieve a density of 108 cells/ml at 60 min. For strains DE3 and LM-R, isopropyl-L-thio-ß-galactoside was added to a final concentration of 1 mM at 50 min. At 60 min, phage were added and grown under the same conditions for 20 min to 60 min before transfer to the next flask. The exact time was generally chosen to allow estimated phage numbers at the point of transfer to be at least 105 but before lysis of the culture; however, lysis was intentionally allowed periodically to encourage recombination of beneficial mutations between phage genomes and thus to avoid clonal interference, which is known to impede adaptation. The combined passage times were 73 h for the T7
1 line on LM-R, 10 h for T31 on LM-R, and 21 h for the recombinant between T710 and T7173 on DE3 (subscripts indicate amount of prior adaptive growth). The adaptation of T71 was terminated at 73 h without immediate regard for whether a plateau had been reached, but after it was clear that fitness had increased substantially and thus that many substitutions likely had occurred. The intent was to focus on mutations of relatively large benefit, with the hope that they would be the most easily interpreted in the context of T7 biology.
Fitness
Assays were carried out under the same conditions used for passage (spanning at least 3 flasks in series) except that lower numbers of phages were transferred (e.g., 103), and transfer to a new flask was done while cell numbers exceeded phage numbers at least 10-fold in order to avoid substantial coinfection. Total and net fitness gain achieved by the adaptation to T3 RNAP was determined by measuring the fitnesses of the initial phage, T710, and phage adapted to T3 RNAP, T7173, on LM-R (providing T3 RNAP). The maximum possible fitness on LM-R (providing T3 RNAP) was estimated by the fitness of T3110, which had been adapted to LM-R for 10 h. This allowed approximation of the maximal fitness on T3 RNAP. To assess whether the adaptation to T3 RNAP resulted in a fitness loss on T7 RNAP (suggestive of a trade-off), we also measured fitness on DE3 (providing T7 RNAP) for phage populations T7173 and for a recombinant containing only mutations in T7173 of general benefit (explained below).
Diagnosis of compensatory mutations
After adaptation, the final culture (T7173) was cross-streaked with T710 to create a zone of mixed infection and thus to allow recombination. The agar containing parental and recombinant phages was collected and used to obtain a phage suspension that was then propagated for 21 h on DE3 (providing T7 RNAP), being supplemented once at an intermediate passage with the initial recombinant mixture. Phage DNA from the final passage of the recombinant population was sequenced over sites of changes in T7173 to determine which changes had been favored during the adaptation to T7 RNAP (table 2). The rationale behind this process is as follows: any mutation in the final culture adapted to T3 RNAP (T7173), which is beneficial regardless of the polymerase present, will also be favored when recombined into the background of a phage lacking any mutations other than the deletion of the RNAP (T710) and grown on a host supplying T7 RNAP. However, those mutations in T7173 that are specific for T3 RNAP and thereby detrimental when T7 RNAP is provided will be selected against. Those changes present in the recombinant population after growth on the bacterial host carrying T7 RNAP are thus likely to be adaptive for the general conditions of adaptive growth and not specific for growth using T3 RNAP.
|
Sequences were determined from phage genomic DNA or polymerase chain reactions (PCR) using chain termination Big Dye 3.1. Reactions were read on an ABI 3100 sequencer. In some cases, sequences were obtained from purified phage stocks; in other cases, sequences were obtained from the mixed population of phage in a lysate at a specific time point. ABI sequence files were analyzed using DNA Star software. The reference T7 genome for nucleotide positions is GenBank V01146.
|
Results |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
Fitness
The immediate effect of replacing T7 RNAP with T3 RNAP was profoundly deleterious for T7 growth. The fitness of the initial phage (T7
10) was 
38 doublings/h on DE3 (T7 RNAP) but only 5 doublings/h on LM-R (T3 RNAP) (fig. 2). After 73 h of selection for rapid growth on LM-R, fitness had increased to 33 doublings/h. The fitness trajectory of T71 on LM-R had not obviously reached an asymptote by 73 h, and it is likely that additional adaptive growth would have led to further increases. However, the rate of fitness increase was diminishing: fitness increased by 20.5 doublings/h in the first 38 h of adaptation but only 7.5 doublings/h in the subsequent 35 h.
|
To get a sense of whether a much higher fitness on LM-R (providing T3 RNAP) was feasible for T7, T3
1 was adapted to LM-R. Because T3 promoters and proteins are already fully compatible with the T3 RNAP supplied by LM-R, the fitness of this phage should approximate the upper limit that might be achieved by a fully adapted T7. After 10 h of adaptation, the fitness of T3110 was 37.2, only 4 doublings/h above that of T7173. It should be realized that this is only an approximate upper limit as intact T7 and T3 phages adapted to the same conditions do not necessarily achieve equal fitness on all host strains (Bull et al. 2004).
Considering that the rate of gain in fitness was declining and the fitness of T7173 is near the maximum indicated by T3110, it is likely that further increases in fitness would have been only gradual. Although there may have been some benefit gained by continuing the adaptation to an obvious plateau, there was already such a substantial fitness improvement (a net 28 doublings/h) and so many substitutions (31, see below) that the major and most easily interpreted patterns had likely been obtained.
The incompatibility of wild-type T7 promoters with T3 RNAP (and conversely, of wild-type T3 promoters with T7 RNAP) may reflect a trade-off in which polymerase specificity for one promoter sequence precludes specificity for others. The possibility of such a trade-off may be evaluated by considering whether the T7 promoters that had changed in T7173 were no longer recognized by T7 RNAP. Overlapping transcripts in T7 (Molineux 2005) preclude accurate measurements of transcription rates from specific promoters. Therefore, in this study, we evaluated the fitness of T7173 on DE3 (providing T7 RNAP) rather than measuring transcription directly. Note that this test is one sided; T7173 had been selected for growth using T3 RNAP, it was not simultaneously or periodically selected for maintained compatibility with T7 RNAP. This test also confounds changes in phage coding sequences that evolved in response to the new RNAP. Such changes could inhibit growth of the evolved phage on T7 RNAP and thus falsely contribute to the appearance of a trade-off between the RNAP and promoter sequences. Yet, when grown on the host expressing T7 RNAP (DE3), fitness of T7173 was essentially the same as of T710, approximately 38 doublings/h (dashed line in fig. 2). This result suggests a complete absence of any trade-off.
However, the comparison is incomplete; mutations that arose during adaptation of T71 to T3 RNAP may include changes that are also beneficial on the comparable bacterial strain harboring T7 RNAP and could mask any that are detrimental. This possibility was resolved by briefly adapting a recombinant between T710 and T7173 to DE3 (T7 RNAP). Adaptation of a mixture of recombinants on DE3 allows selection to retain the changes beneficial on T7 RNAP, whereas eliminating those that are detrimental. The fitness of an isolated recombinant on T7 RNAP was higher than both T710 and T7173 (47 vs. 38–39 doublings/h; fig. 2). The higher fitness of the recombinant is consistent with the existence of a trade-off between RNAP specificity and transcription activity. However, the trade-off is far less than suggested by the initial fitness difference of T710 on DE3 and LM-R. The difference would likely have been even less had selection been maintained simultaneously for both T7 and T3 RNAP. As detailed below, all the promoter changes that evolved in T7173 were selected against in the recombinant, suggesting that promoter changes were indeed only beneficial to growth using T3 RNAP.
The recombinant is thus well adapted to T7 RNAP (host DE3), more so than T710 and T7173. Had the experiment started with this phage, its initial fitness on LM-R might well have been higher than the 5 doublings/h of T710, and we would have attributed less of the adaptive gain to changes specific to T3 RNAP. To establish this important baseline, 3 isolates of the DE3-adapted recombinant population were assayed on LM-R, but instead of a fitness greater than 5, the fitness was negative (–1.34 ± 025, for 1 isolate; –1.43, –0.84 in single determinations of the other 2 isolates). A fitness less than 5 is puzzling. If there were no interactions, any changes that evolved on LM-R (T3 RNAP) that were also beneficial on DE3 (T7 RNAP) would have boosted fitness on LM-R in the original genetic background. The failure to obtain this result points to strong interactions, the natures of which are not clear, and highlights the potential difficulty in predicting evolution in this system.
Molecular evolution
The complete sequence of an isolate, i1, from T7173 revealed 7 substitutions and 1 insertion that were beneficial during growth using either T7 or T3 RNAP and were thus not specifically compensatory for the change to a new RNAP. In addition T7173 i1 contained 21 substitutions, 1 insertion, and 1 deletion of a compensatory nature (table 2). As expected from the interactions known for T7 RNAP, most changes in response to growth using T3 RNAP affected promoters, and most of those tended to change bases favored by T3 RNAP (Klement et al. 1990). However, only 9 of the 16 promoters of T71 acquired changes (the starting phage lacked one promoter of the wild-type genome as part of the gene 1 deletion) (table 2 and fig. 1). Several missense mutations in coding sequences were found; 2 phage proteins, gp3.5 (lysozyme) and gp19 (terminase), are known to interact with the phage RNAP (Zhang and Studier 1995), and both genes acquired missense substitutions that were compensatory for T3 RNAP. Most other missense mutations were either of general adaptive benefit or in genes that have no known interaction with RNAP. Importantly, all the promoter changes that arose in T7173 as a result of adaptation to growth on T3 RNAP were lost during adaptation of the T7173 x T710 recombinant on T7 RNAP. This indicates that the promoter mutations in T7173 were strictly compensatory for T3 RNAP.
To assess the level of promoter polymorphism in the final passage, all promoters were sequenced from 2 other isolates from T7173 (fig. 3). A surprising result is that, although 4 base positions in the promoters were polymorphic across the 3 isolates, all 3 isolates from 73 h of adaptive growth were mutant in the same 8 promoters, whereas 7 promoters remained mutation-free. Only one promoter (ø13) was mutant in one isolate but not in the others.
|
The promoter changes were partly surprising in another way. Relative to the first base transcribed (+1), positions –10, –11, and –12 are considered to be most critical for recognition by T7 RNAP verses T3 RNAP, based on in vitro transcription assays (Klement et al. 1990
). Position –2 was also shown to be important, but only in conjunction with a change at –11 rather than by itself. One of the promoter positions that most commonly evolved in the T7173i1 was –11, which experienced 5 changes (fig. 4). Yet, only one of those mutations was to C, the base most commonly found at –11 in T3 promoters. Six changes also occurred at the –2 position, 5 of them to the T3 base (A). Most of those changes were found with a change at –11 in the same promoter, but 2 were without other promoter changes (
9 in T773il and 2.5 in all 3 isolates at 73 h). Prior work has not tested the specific combination of the most common (–2 and –11) base changes observed in our study, but we can at least assert that those positions match expectations from in vitro studies. However, the most common base change at –11 does not match expectations.
|
To obtain information about the time course of promoter changes and levels of polymorphism, the promoter regions with mutations in T7
173 were sequenced from PCR products of the phage populations at several times from the adaptation (fig. 3). Approximate levels of mutation accumulation were inferred from relative heights of different peaks in the sequencing reaction. Some promoter mutations occurred early in the adaptation and apparently fixed shortly thereafter, but some ascended only slowly through the population (e.g., 1.6 and 2.5).
At intermediate times of the adaptation, promoter regions were also sequenced from isolated phages; the regions sequenced included genome segments adjacent to the promoters, from which additional information was obtained (fig. 5). Several mutations were detected at these intermediate times that were not in the T7173 isolate that was completely sequenced, T7173i1, but they tended to occur in the same genes as the changes in T7173 i1. With only limited sampling, it is not possible to say whether some of those changes were indeed fixed and others lost, but these data reinforce the biological significance of the genic changes during this adaptation to T3 RNAP. To the extent that some changes were lost and others fixed, the evolutionary pattern could represent either clonal interference or merely one of antagonistic epistasis—that different mutations within the same gene produced similar molecular effects and thus were not beneficial in combination.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
Bacteriophage T7 deleted for its RNAP gene (T7
1) was adapted to a host that supplied, in trans, the RNAP from phage T3. T3 is a close relative of T7; the 2 phages have an identical order of essential genes and share homologs of most nonessential genes. However, sequence divergence between most genes approaches 30%, and the protein sequences of the 2 RNA polymerases differ by 18% (Pajunen et al. 2002). Furthermore, in vitro studies show that T3 RNAP has low activity on T7 consensus promoters and vice versa (Golomb and Chamberlin 1977; Bailey et al. 1983; Klement et al. 1990). The specificity of the phage RNAPs for their cognate promoters allowed development of commercial vectors where selective expression of cloned genes can be achieved. Consistent with this prior work, the fitness of T710, which contains only T7 promoters, was initially very low when using T3 RNAP for growth. Compensatory evolution corrected much of this deficit through a combination of genic and promoter substitutions. The trajectory of fitness increase suggests that further adaptation might have resulted in a T7 derivative whose fitness was as high as T31 adapted to the same host.
Although the sample size for comparison is not large, the recovery of fitness in T7173 is much greater than seen in our previous studies with T7 (Rokyta et al. 2002; Bull et al. 2004; Heineman et al. 2005; Springman et al. 2005). One argument for the nearly complete recovery is that the initial phage carried essentially all elements of a wild-type genome and merely needed to fine-tune interactions with the novel RNAP. Yet this argument supposes an absence of strong interactions among the many mutations needed to improve fitness, for which there is little evidence. On the one hand, a promoter change presumably affects little other than rate of transcription initiation, and on these grounds, it would not seem to be subject to interactions with other phenotypes. However, a change in one promoter might have strong interactions with changes in other promoters when their transcripts overlap or by creating imbalances in gene expression levels with other parts of the genome. Supporting this view, changes in the order of genes in T7 have large effects on fitness that are not easily overcome by evolution (Springman et al. 2005). Furthermore, enigmatic interactions were evident from the fact that the recombinant (of T710 and T7173) adapted to DE3 (T7 RNAP) had lower fitness on LM-R (T3 RNAP) than did the starting phage.
The high fitness of T7173 is surprising given that almost half the T7 promoters did not change during adaptation to growth using T3 RNAP. However, most T7 transcripts overlap and thus exhibit some level of redundancy. Furthermore in this study, adaptation of T71 to growth using T3 RNAP was confined to a single environment, where changes in gene expression patterns that would prove beneficial in changing conditions were not selected. Although it is expected that the first few promoter mutations would confer the greatest fitness advantage, it was not expected that activation of so few promoters would confer a fitness benefit so close to the apparent maximum possible on T3 RNAP. However, the high fitness of T7173 despite the absence of change in many promoters can perhaps be justified post hoc.
The total of 31 adaptive changes observed during this brief experimental adaptation (in isolate T7173 i1) is also larger than seen in our prior work (Rokyta et al. 2002; Bull et al. 2004; Springman et al. 2005). The average rate was 1 substitution per 2.4 h of adaptive growth (including both compensatory and noncompensatory mutations). Approximately, half the adaptive substitutions were in genes, rather than promoters, and several genic changes proved compensatory for T3 RNAP, even though only 2 gene products (gp3.5 and gp19) are known to interact with the RNAP.
Surprisingly, T7173 retained high fitness on T7 RNAP. This result suggests a solution to the quandary of how coupled systems diverge in evolution. T7 and T3 have clearly evolved from a common ancestor, yet their RNAPs no longer effectively recognize the heterologous promoters. Thus, both the RNAP and the multiple promoters within a specific genome phage lineage must have coordinately changed recognition during evolution without seriously compromising fitness. The properties of T7173 suggest that intermediates exist that do not exhibit a strong trade-off. Such intermediates may include; 1) promiscuous RNAPs that allow the evolution of different promoter sequences, even in the same genome, 2) promoters that can be recognized by different RNAPs, or 3) promoter redundancy, whereby some promoters evolve different sequences from others in the same genome and transcripts from the promoters with different sequences overlap. These models are not incompatible, and it is likely that T7173 exemplifies the latter two.
Some promoters did not change during adaptation to growth on T3 RNAP, and those must still be recognized normally by T7 RNAP. Yet, the evolution of 2 sets of promoters with different compatibilities cannot be the full story here. In particular, the ø9 and ø10 promoters, which both changed during adaptation to growth on T3 RNAP, must be efficiently recognized by both the T3 and T7 enzymes to achieve the observed levels of fitness on the 2 hosts used. Transcripts from these promoters direct synthesis of the 2 structural proteins that are required at the highest stoichiometry for phage growth; it is extremely doubtful that T7173 would exhibit high fitness with T7 RNAP unless both 9 and 10 were utilized. Nevertheless, some fitness cost remained associated with T7173 on T7 RNAP and T7173 does not represent a stable intermediate in the divergence of T7 and T3. When T710 and T7173 were recombined and grown on DE3 (providing T7 RNAP), phages containing only natural T7 promoters rapidly dominated the population. However, this fitness cost was not high, and in a natural phage system where the RNAP could also evolve, any residual cost may be insignificant. It is even feasible that RNAP variants exist that do not discriminate between T3 and T7 promoters. However, our experimental design precluded RNAP evolution; so novel RNAP molecules could not have been observed here.
At a basic level, the specific promoters that changed and the times during the adaptation at which those changes arose match simple expectations from known T7 biology (Molineux 2005). In a genome lacking the RNAP gene, the first essential genes to enter the cell mainly code for enzymatic activities involved in DNA metabolism, whose absolute levels of expression may not be critical for fitness. In T7, there are several promoters upstream of the DNA metabolism genes; changes in any one could serve to allow adequate levels of transcription by T3 RNAP. Consistent with this idea, the 1.5 promoter changed early and rapidly reached high frequency in the population. Although the 1.6 and 2.5 promoters also changed during the adaptation, those changes were at best slow to ascend through the population. Furthermore, 6 promoters in this region of the genome never changed during the adaptation.
In contrast, T7 late genes mainly code for structural components of the phage. Low levels of synthesis of some proteins would limit phage growth, and it was therefore expected that late promoters would change during adaptation. The first late promoter, 6.5, indeed did change, albeit only late during adaptive growth, but the 9 and 10 promoters changed and fixed early. As mentioned above, these promoters lie upstream of genes whose products are needed at the highest levels during phage growth. Although neither 13 nor 17 promoters changed substantially, it should be noted both that wild-type T3 does not utilize 17 and that T7173 acquired a mutation in the terminator T. The same mutation was observed in a previous study where a reduced termination efficiency was hypothesized (Springman et al. 2005). In the absence of effective transcription termination, all genes downstream of the 9 and 10 promoters should be expressed adequately even in the absence of any activity from downstream promoters. Finally, the promoters closest to the genome ends, OL and OR, both changed early in the adaptation. Neither promoter is normally used for much transcription, but they are thought to be important in DNA replication. The phage RNAP is necessary for both DNA replication initiation and packaging progeny DNA from replicating concatemers, so changes in these promoters during adaptation of T71 to growth using T3 RNAP were anticipated.
The evolution of gene regulation in microbes is actively researched; several patterns of gene regulation have become evident, and the engineering of regulatory circuits has become a reality (e.g., Atsumi and Little 2006; McAdams and Arkin 1998, 2000; Savageau 1998, 2001; Atkinson et al. 2003; Wall et al. 2003, 2004; McAdams et al. 2004). An impressive advance in this arena with respect to T7 was the development of a virtual model (Endy et al. 2000). Unfortunately, this model (v2.5) proved to be uninformative about the fitness values of different promoter combinations and hence about which promoters were most likely to benefit the phage during adaptation to a new RNAP. This is unfortunate, as modeling the pathways of transcription and the switch from host to phage RNAP are among the most robust aspects of the model. However, values of the replication promoters, OR and OL, are set to zero in v2.5 of the simulation, indicating their unimportance in the modeled interactions. In addition, the virtual model is based on growth at 30 °C, rather than the 37 °C used in this work. It is not clear that all intracellular processes and overall phage growth rates will respond in a simple way to a temperature shift. More specifically, implausible fitness values were obtained when genomes were specified to carry a single active promoter or in some cases a pair of active promoters. For example, setting the simulated growth rate of wild type at 33.5 doublings/h, the fitness of a phage with no promoters active was 9.0. This is in fair agreement with the observed value of 5 for T710. However, activating 6.5 or 9 alone yielded fitnesses of 34, the highest values obtained for a single active promoter and slightly in excess of the wild-type value. Activating just 9 and 10 together yielded a fitness of 41, well above wild type. Although other combinations of active promoters might yield even higher fitness than 41, the model would predict that wild type would be vastly outgrown by a phage with only 9 and 10 active, and slightly outgrown by a phage with only 6.5 or 9 active. A priori, the hope for the T7 virtual model was to obtain accurate rankings of fitness values of different promoter combinations in order to explain likely pathways of progressive fitness improvement; however, v2.5 is clearly not able to do that reliably. Hopefully, the new version of the virtual model that is being developed (Endy D, personal communication) will be more comprehensive and flexible in its use of specific transcription parameters.
Overall, more than half the changes specific to T3 RNAP adaptation can be explained on qualitative grounds. Yet our current state of knowledge of T7, although very extensive, is not to the point at which the changes can be justified quantitatively. Although the challenge to the T7 genome imposed by our design was in principle simple, merely requiring the reevolution of activity from a set of tuned-down promoters, the regulatory challenge became genome wide. Potential interactions appeared that lie beyond our current understanding of T7 and as a consequence, are inexplicable by the simple models that were derived from that understanding. It is possible that many of the changes that arose during adaptation to growth using a new RNAP were beneficial specifically because of the unbalanced gene expression. Once fixed, those mutations could constrain further adaptation of upstream promoters or interacting proteins. A precedent for this idea comes from 2 studies in T7 showing that a reduced ligase activity in the infected cell selects for reduction in activity of a phage-coded nuclease. This result was interpreted as helping to coordinate the activities of enzymatic processes with opposing effects (Sadowski 1974; Rokyta et al. 2002).
The results here offer hope for improvement of engineered genomes via corrective evolution. Through a series of point mutations, a genome forced to accept a new RNAP overcame a profound fitness decrement and achieved nearly "wild-type" fitness levels. One must imagine that engineered genomes of the future will not be designed optimally, but will face numerous small or moderate incompatibilities among the heterologous elements. Adaptation of those genomes is a potential method to correct those defects. The evidence here is that substantial corrections can be achieved with experimental evolution.
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
We thank H. Wichman for comments, D. Endy for help using v2.5 of the virtual model, and W. T. McAllister for providing LM-R. This work was supported by National Institutes of Health (NIH) GM57756 (to J.J.B.) and by NIH GM32095 (to I.J.M.). J.J.B. was also supported by the Miescher Regents Professorship at the University of Texas.
Funding to pay the Open Access publication charges for this article was provided by the National Institutes of Health grant GM 57756.
|
Footnotes |
|---|
Jianzhi Zhang, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMethods and MaterialsResultsDiscussionAcknowledgementsReferences |
|---|
Atsumi S and Little JW. (2006) Role of the lytic repressor in prophage induction of phage lambda as analyzed by a module-replacement approach. Proc Natl Acad Sci USA 103:4558–4563.
Atkinson MR, Savageau MA, Myers JT, Ninfa AJ. (2003) Development of genetic circuitry exhibiting toggle switch or oscillatory behavior in Escherichia coli. Cell 59:7–607.
Bailey JN, Klement JF, McAllister WT. (1983) Relationship between promoter structure and template specificities exhibited by the bacteriophage T3 and T7 RNA polymerases. Proc Natl Acad Sci USA 80:2814–2818.
Bull JJ, Badgett MR, Springman R, Molineux IJ. (2004) Genome properties and the limits of adaptation in bacteriophages. Evol Int J Org Evol 58:692–701.
Crooks GE, Hon G, Chandonia JM, Brenner SE. (2004) WebLogo: a sequence logo generator. Genome Res 14:1188–1190.
Endy D, You L, Yin J, Molineux IJ. (2000) Computation, prediction, and experimental tests of fitness for bacteriophage T7 mutants with permuted genomes. Proc Natl Acad Sci USA 97:5375–5380.
Golomb M and Chamberlin MJ. (1977) T7- and T3-specific RNA polymerases: characterization and mapping of the in vitro transcripts read from T3 DNA. J Virol 21:743–752.
Heineman RH, Molineux IJ, Bull JJ. (2005) Evolutionary robustness of an optimal phenotype: re-evolution of lysis in a bacteriophage deleted for its lysin gene. J Mol Evol 61:181–191.[CrossRef][ISI][Medline]
Klement JF, Moorefield MB, Jorgensen E, Brown JE, Risman S, McAllister WT. (1990) Discrimination between bacteriophage T3 and T7 promoters by the T3 and T7 RNA polymerases depends primarily upon a three base-pair region located 10 to 12 base-pairs upstream from the start site. J Mol Biol 215:21–29.[CrossRef][ISI][Medline]
MacDonald L. (1993) Termination by T7 RNA polymerase(State University of New York, Brooklyn (NY)).
McAdams HH and Arkin A. (1998) Simulation of prokaryotic genetic circuits. Annu Rev Biophys Biomol Struct 27:199–224.[CrossRef][ISI][Medline]
McAdams HH and Arkin A. (2000) Towards a circuit engineering discipline. Curr Biol 10:R318–R320.[CrossRef][ISI][Medline]
McAdams HH, Srinivasan B, Arkin AP. (2004) The evolution of genetic regulatory systems in bacteria. Nat Rev Genet 5:169–178.[ISI][Medline]
Molineux I. (2005) The T7 group. In Calendar R (Ed.). The Bacteriophages(Oxford University Press, Oxford) pp. 277–301.
Morris CE, Klement JF, McAllister WT. (1986) Cloning and expression of the bacteriophage T3 RNA polymerase gene. Gene 41:193–200.[CrossRef][ISI][Medline]
Pajunen MI, Elizondo MR, Skurnik M, Kieleczawa J, Molineux IJ. (2002) Complete nucleotide sequence and likely recombinatorial origin of bacteriophage T3. J Mol Biol 319:1115–1132.[CrossRef][ISI][Medline]
Raskin CA, Diaz G, Joho K, McAllister WT. (1992) Substitution of a single bacteriophage T3 residue in bacteriophage T7 RNA polymerase at position 748 results in a switch in promoter specificity. J Mol Biol 228:506–515.[CrossRef][ISI][Medline]
Rokyta D, Badgett MR, Molineux IJ, Bull JJ. (2002) Experimental genomic evolution: extensive compensation for loss of DNA ligase activity in a virus. Mol Biol Evol 19:230–238.
Sadowski PD. (1974) Suppression of a mutation in gene 3 of bacteriophage T7 (T7 endonuclease I) by mutations in phage and host polynucleotide ligase. J Virol 13:226–229.
Savageau MA. (1998) Rules for the evolution of gene circuitry. Pac Symp Biocomput 5:4–65.[Medline]
Savageau MA. (2001) Design principles for elementary gene circuits: elements, methods, and examples. Chaos 11:142–159.[CrossRef][ISI][Medline]
Springman R, Badgett MR, Molineux IJ, Bull JJ. (2005) Gene order constrains adaptation in bacteriophage T7. Virology 341:141–152.[CrossRef][ISI][Medline]
Studier FW and Moffatt BA. (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189:113–130.[CrossRef][ISI][Medline]
Wall ME, Hlavacek WS, Savageau MA. (2003) Design principles for regulator gene expression in a repressible gene circuit. J Mol Biol 332:861–876.[CrossRef][ISI][Medline]
Wall ME, Hlavacek WS, Savageau MA. (2004) Design of gene circuits: lessons from bacteria. Nat Rev Genet 5:34–42.[CrossRef][ISI][Medline]
Zhang X and Studier FW. (1995) Isolation of transcriptionally active mutants of T7 RNA polymerase that do not support phage growth. J Mol Biol 250:156–168.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||