MBE Advance Access originally published online on June 12, 2008
Molecular Biology and Evolution 2008 25(9):1835-1840; doi:10.1093/molbev/msn131
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Neutrality of Foreign Complex Subunits in an Experimental Model of Lateral Gene Transfer
Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israel
E-mail: urigo{at}post.tau.ac.il.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Lateral gene transfer (LGT) is a powerful force in microbial evolution. However, the barriers that restrict this evolutionary phenomenon are not fully understood. It has long been observed that genes that encode subunits of complexes exhibit relatively compatible phylogenies, implying mostly vertical evolution. This may be explained by the failure of a new gene product to effectively interact with preexisting protein subunits, making its acquisition neutral—a theory termed the "complexity hypothesis." On the other hand, such genes may reduce the fitness of the host by disturbing the stoichiometric balance between complex subunits, resulting in purifying selection against gene retention. To examine these 2 alternative scenarios, we designed an experimental system that mimics the transfer of genes encoding homologs of essential complex subunits into the model bacterium Escherichia coli. In addition, we overexpressed the native E. coli gene in order to examine the contribution of gene dosage effects. We show that accumulation of native or foreign complex subunits in the cell does not result in loss of fitness, except for a minor fitness reduction observed for a single foreign homolog. Indeed, a series of genetic and biochemical assays failed to detect any interaction between the foreign subunits and the native polypeptides of the complex, implying an inability of such transfer events to generate positive selection for gene retention. We conclude that LGT of complex subunits may be mostly neutral and that forces operating against gene retention appear to be moderate.
Key Words: horizontal gene transfer • complexity hypothesis • balance hypothesis • protein complexes • essential genes • Escherichia coli
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Lateral gene transfer (LGT) is a predominant force in microbial evolution and an engine of genetic diversity in microorganisms (Doolittle et al. 2003
). The fact that frequencies of gene transfer vary widely across functional classes (Nakamura et al. 2004), and the observation that some gene families appear to be more "transfer resistant" than others (Jain et al. 1999), sparked great interest in the question of what constitutes a barrier to LGT. It is commonly believed that genes encoding subunits of complexes would be seldom retained after gene transfer, due to lack of positive selection, because they would be unable to form successful interactions with native proteins (Jain et al. 1999). However, because it has been demonstrated that horizontally acquired genes can be fixed in a microbial population, even when they are neutral (Novozhilov et al. 2005), complex membership should probably be regarded as a hurdle, rather than a true barrier, to LGT.
Nevertheless, it is possible that beyond neutrality, selective negative forces do operate against foreign gene retention thus acting as true barriers to LGT. Sorek et al. explored such barriers by examining gaps in clone coverage of genome projects data. These gaps represent genes that cannot be retained by Escherichia coli and constitute lethal transfers (Sorek et al. 2007). Their findings indicate that globally single-copy genes tend to be transfer resistant (Sorek et al. 2007). Along the same lines, an analysis of E. coli phylogenetic data has shown that gene duplicability and transferability are significantly correlated (Wellner et al. 2007). Low duplicability and transferability of a gene are likely to stem from a toxic effect of an increased dosage of its product in the cell, either due to an imbalance in the stoichiometry of protein complex constituents, as suggested by the "balance hypothesis" (Papp et al. 2003), or to other disruptions to the homeostasis of the cell. The former mechanism may not hold true in bacteria because complex membership in E. coli is not correlated with duplicability or transferability (Wellner et al. 2007). Nevertheless, Sorek's work did demonstrate that nearly all universally single-copy genes that could not be transferred from 5 or more genomes encode ribosomal proteins. This observation indicates that the balance hypothesis could still hold true for essential complexes and explain their comparatively low gene transfer rates during the evolution of microorganisms.
Lethality under laboratory conditions represents a very extreme phenotype and one could easily argue that just as "nonessential" genes can actually be essential (Fang et al. 2005), many slightly deleterious acquired genes would never successfully propagate in E. coli in a more natural setting (McInerney and Pisani 2007). We have therefore chosen to focus on more subtle effects of fitness brought about by LGT of subunits of an essential complex. This was performed by artificially introducing a foreign subunit of an essential complex into the model bacterium E. coli and examining the consequences using genetic and biochemical approaches.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
DNA Manipulation
All polymerase chain reaction (PCR) products were amplified using Phusion DNA Polymerase (Finnzyme F-530S/L) with corresponding genomic DNA as templates. Reactions were cycled 31 times (denaturation at 98 °C, 10 s; annealing at varying temperatures [supplementary table S1, Supplementary Material online], 30 s; and elongation at 72 °C, 30 s). Restriction endonucleases used for cloning to pBAD vectors were purchased from Promega (supplementary table S1, Supplementary Material online). All cloned inserts were verified by sequencing.
Knockout of the chromosomal accA gene was carried out using the 1-step inactivation technique (Datsenko and Wanner 2000). Briefly, the kanamycin resistance cassette was amplified using primers accadelF and accadelR from a pKD4 template (annealing at 50 °C for 45 s and elongation at 72 °C for 60 s), with BioTaq Taq polymerase (Bioline Inc., Luckenwalde, Germany). Because pBAD and pKD42 contain the same antibiotic resistance marker, pBAD-accA-E. coli was digested with BspH1, and the resulting 2,930-bp fragment, containing the accA gene with the araBAD promoter, was ligated to a BspH1-digested pACYC-134 plasmid harboring chloramphenicol resistance. Following the genomic knockout, bacteria were grown on arabinose-containing plates to allow complementation from the plasmid-encoded accA gene. Knockout was verified by colony PCR with primers verdelF, verdelR, k1, and Kt (supplementary table S1, Supplementary Material online).
Growth Rate Experiments
Overnight bacterial cultures were diluted 109-fold and were used to inoculate 50 ml of fresh ampicillin-containing Luria-Bertani (LB) broth in a 500-ml flask. After 2 h of growth at 30 °C, each culture (in duplicates) was split into 2 parallel flasks. One culture was supplemented with D-glucose and the other with L-arabinose, both to a final concentration of 0.2%. After an additional 15 h of growth at 30 °C, once the optical density at 600 nm wavelength (OD600) reached a level detectable by a spectrophotometer, OD was measured every 60 min. The experiment was repeated twice.
Western Blotting
Every strain under each growth condition was grown until OD600 reached 0.5, and 1 ml of the culture was centrifuged and resuspended in 50 µl sample buffer (0.5 M Tris–HCl pH 6.8, 10% glycerol, 2% sodium dodecyl sulfate (SDS), 0.01% bromophenol blue). Gel electrophoresis (10% SDS–polyacrylamide gel electrophoresis [SDS-PAGE]) and western blotting were carried out as previously described (Biran et al. 2000). Mouse anti-Flag M2 (Sigma, St. Louis, MO) was used as primary antibody and a horseradish peroxidase–conjugated anti-Mouse IgG from Rabbit (Sigma) as a secondary antibody. Detection was performed with EZ-ECL chemiluminescence detection kit (Biological Industries).
Protein Degradation Assay
A 100 ml of culture of each strain was grown in LB containing ampicillin at 37 °C for 1 h, supplemented with 0.2% arabinose and subsequently grown for another 
1 h at 30 °C until OD600 reached 0.5. Cultures were then supplemented with 30 mg/ml rifampicin and 25 mg/ml spectinomycin to stop protein synthesis. Samples were taken every 30 min, centrifuged, and resuspended in sample buffer. Samples were analyzed by SDS-PAGE followed by western blotting.
Purification of Protein Aggregates
Aggregated proteins were purified from 100 ml cultures as previously described (Tomoyasu et al. 2001), except that Tris–ethylenediaminetetraacetic acid, pH 7.5, containing phenylmethylsulfonyl fluoride (PMSF; 243 µg/l) was used as a buffer for all the purification steps.
Transduction of the accA Knockout
Transductions were performed as previously described (Thomason et al. 2007). Briefly, an overnight culture of E. coli K-12 MG1655 pBAD-AccAE. coli was transduced with 100 µl of several dilutions of a P1 lysate from an E. coli K-12 MG1655 
accA pACYC-AccAE. coli strain. In addition, 2 controls were set up, 1 containing phage lysate and 100 µl LB and the other with recipient cells only and LB. After the most efficient lysate concentration (diluted 20-fold) was established (yielding 25 resistant colonies—verified by PCR) and the negative controls did not yield any colonies, this diluted lysate was used to transduce the strains expressing the foreign subunits in order to examine whether these proteins could complement a deletion in the genomic copy of accA.
Coimmunoprecipitation
Bacterial strains were grown for approximately 1 h to an OD600 of about 0.2, 0.2% L-arabinose was added to the culture, and the cultures were grown for another hour at 30 °C. After harvesting by centrifugation (14,000 x g, 10 min, 4 °C), the pellets were resuspended in 1 ml of 50 mM Tris-base saline (TBS) containing 1 mM PMSF. This procedure of harvesting and resuspending was carried out twice in order to remove any residual LB. Cell lysis was carried out by sonication at 4 Watts for 2 min in an ice bath. The lysates were immediately centrifuged at 20,000 x g for 30 min at 4 °C in order to remove debris composed of aggregates and membranes. Total protein quantity was determined by a Bradford assay, and the lysates were equalized. Each bacterial strain lysate was incubated with 50 µl of refrigerated M2-coated agarose beads, prewashed with 1 ml of 1 mM PMSF-containing TBS for 1 h at 4 °C, in a rotary device which completes a full rotation every 5 s. The tubes were centrifuged at 1,000 x g for 30 s at 4 °C to precipitate the beads, and the supernatant was taken out and kept at 4 °C. Three washes were carried out using 1 ml of 1 mM PMSF-containing TBS. Eventually, the bound proteins were eluted from the lysate by incubation at 80 °C for 10 min with sample buffer.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Expression of Foreign Subunits Can Be Neutral
To test the effect on fitness of laterally acquired subunits of complexes, we chose accA, which encodes the alpha subunit of carboxyl-transferase (CT) as our model for subunit acquisition. CT is a part of a larger complex, acetyl-CoA carboxylase (ACC), which is located in the cytoplasm and catalyzes the first and rate-limiting step of fatty acid synthesis. The ACC complex is conserved from bacteria to animals (Guchhait et al. 1974
). Notably, in E. coli and other bacteria, both gram-positive and gram-negative, this enzyme was shown to have an essential role and could not be deleted (Forsyth et al. 2002; Gerdes et al. 2003; Kobayashi et al. 2003; Knuth et al. 2004). We amplified accA homologs from 4 evolutionarily remote bacteria: E. coli K12 MG1655, Bacillus subtilis strain 168, Listeria monocytogenes strain EGDe, and Agrobacterium tumefaciens strain C58, using PCR (see Materials and Methods). All of those foreign protein homologs had between 50–52% identity and 69–70% similarity to AccAE. coli. All amplicons were cloned into a vector containing a tightly regulated inducible promoter, pBAD, to generate 4 plasmids designated: pBAD-AccAE. coli, pBAD-AccAB. subtilis, pBAD-AccAL. monocytogenes, and pBAD-AccAA. tumefaciens (table 1). Each plasmid was then transformed into the E. coli K12 strain MG1655, and recombinant protein expression was induced by addition of arabinose to the growth medium. The expression of recombinant (transferred) subunits did not result in growth inhibition of any of the plasmid-bearing strains when grown for 7 h, roughly equivalent to 8 generations. In order to detect more subtle changes in growth rate, cultures were diluted 109-fold and then grown for about 30 generations (24 h). During this extended growth period, the strain expressing the B. subtilis AccA homolog was the only clone exhibiting a cumulative reduction in growth rate of 4%, while the other clones showed no growth inhibition (fig. 1). Thus, we conclude that although the expression of a foreign subunit of an essential complex may occasionally be deleterious to the fitness of the host, such expression will be neutral (or very mild) in many other cases.
|
|
One may argue that laboratory conditions, in which no competition takes place, may obscure a reduction in fitness. On the other hand, cloning of the recombinant genes under a strong inducible promoter, such as the BAD promoter, yields strong transcription (verified by quantitative real time PCR, data not shown) resulting in relatively high protein levels in the cell. Conversely, a typical transferred gene, transcribed from its own foreign promoter, will usually be lowly expressed, especially during exponential growth (Doyle et al. 2007
). Thus, in our experimental system, the fitness cost of a foreign protein's expression is by no means minimized.
Both Native and Foreign AccA Subunits Accumulate in the Cytoplasm
Because no substantial effects on growth were observed, we were interested in testing whether the foreign subunits are indeed efficiently expressed. Total cellular protein extracts were obtained from all strains, grown on medium containing either arabinose (induction) or glucose (repression) and compared using SDS-PAGE analysis. Because visualization using Coomassie staining could detect only the accumulation of Acca A. tumefaciens, all homologs were cloned again, this time fusing a Flag epitope to the C-terminal end of their corresponding polypeptides (table 1). Western blotting using antibodies against the Flag epitope demonstrated the presence of substantial quantities of all 4 Flagged proteins, upon induction with arabinose (supplementary fig. S2, Supplementary Material online). Because a substantial protein accumulation was observed for all strains, yet only one of the strains exhibited growth inhibition, we further explored the possibility that these proteins aggregated or rapidly degraded in the host cell. Separation of protein aggregates from soluble cytoplasmic proteins (see Materials and Methods) showed that all the recombinant proteins were detected in the soluble fraction and only some of the Acca A. tumefaciens aggregated (fig. 2A). We then performed protein degradation assays to further test whether lack of growth inhibition was the result of a rapid degradation of the acquired subunits by the host proteolytic machinery. Significantly, the degradation assay (fig. 2B) clearly demonstrated that the acquired subunits are not short lived and persist in the cell long enough to potentially influence its fitness (i.e., their half life is comparable to the generation time of the bacterium).
|
Unlike the Native Homolog, Foreign Subunits Do Not Integrate into the Host's Native Complex
Because former experiments showed that the acquired foreign subunits persisted in the cytoplasm yet had minor or no phenotype, we proceeded to examine whether the lack of fitness cost could be attributed to lack of interaction with the native complex subunits. We first tested whether the cloned subunits can integrate into the native complex using a genetic approach, examining whether they can complement a null accA mutation. This was performed by allele exchange of the native accA with a kanamycin resistance cassette in a strain carrying a plasmid-encoded backup of this essential gene (see Materials and Methods), followed by transduction of the cassette into strains expressing each of the acquired subunits. Strikingly, only the strain carrying pBAD-AccAE. coli (and its Flagged counterpart) could accommodate a genomic accA mutation, resulting in 25 kanamycin-resistant transductant colonies, whereas those carrying the plasmids encoding foreign orthologs could not be transduced. We therefore concluded that the foreign orthologs do not interact properly with the other subunits and do not enable the formation of a functional ACC complex. To test this conclusion in a more direct fashion, we created an accA deletion mutant, complemented by the Flag-fused AccAE. coli expressed from the pBAD plasmid. In this strain, the other subunit (beta) of the CT subcomplex, required for formation of the complete ACC complex, should coprecipitate with AccA-Flag, using anti–Flag-coated agarose beads. Using this coimmunoprecipitation technique, we could test whether similar interactions are formed between the foreign subunits and the residing complex subunits. Indeed, the E. coli subunit could form a stable interaction with the beta subunit (AccD) of the ACC complex, as evidenced by an SDS-PAGE followed by matrix-assisted laser desorption/ionization mass spectrometry of the putative beta subunit band. Conversely, the foreign subunits did not coelute with the native beta subunit (fig. 3), even when a more sensitive silver staining was performed. We therefore conclude that none of the foreign subunits can successfully interact with the host beta subunit and form the CT subcomplex.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Understanding the molecular mechanisms that underlie evolutionary processes can contribute to a more complete understanding of these processes. Therefore, to understand barriers to retention of transferred genes, one should go beyond phylogenomic analysis and investigate experimental models. We have demonstrated that overexpression of the endogenous ACC subunit AccAE. coli did not affect the fitness of an E. coli strain, as reflected by its growth rate in a rich medium, despite the essentiality of the ACC complex. This implies that at least under laboratory conditions, there is no apparent fitness cost due to an imbalance of ACC complex subunits. This is in agreement with our general observation, on E. coli protein data, that in this organism the balance hypothesis, originally formulated for yeast (Papp et al. 2003
), is not supported (Wellner et al. 2007). It therefore appears that transfer resistance due to complex imbalance may be restricted to a handful of protein complexes, especially the ribosome. This may be simply because large hetero-oligomeric complexes having many different subunits are relatively uncommon in bacteria in comparison to eukaryotic cells. Expression of foreign subunits, mimicking an acquisition by LGT showed mixed effects on bacterial growth. Whereas AccAA. tumefaciens and AccAL. monocytogenes did not result in growth inhibition, a third foreign subunit (AccAB. subtilis) caused a minor inhibitory effect. The lack of fitness cost observed in 2 out of 3 experimental transfer events can be attributed to their lack of integration into the native ACC complex as was demonstrated genetically and biochemically (see above). Surprisingly, the AccAB. subtilis, whose expression inhibited growth to a limited extent, also could not form stable interactions with the native subunits (fig. 3). This result can be explained by a weaker binding of AccAB. subtilis to the native CT subcomplex beta subunit, compared with that of the AccAE. coli, resulting in failure to coprecipitate. Alternatively, the fitness cost could be explained by a nonspecific aberrant interaction of AccAB. subtilis with some other cell component that is deleterious to E. coli. We favor the latter explanation because both the taxa relations between the bacterial species and the phylogenetic tree of AccA homologs (supplementary fig. S3, Supplementary Material online) would have predicted that AccAA. tumefaciens, the sister taxon of AccAE. coli, would be the most harmful rather than AccAB. subtilis that has roughly the same branch distance from AccAE. coli as AccAL. monocytogenes.
The observed lack of interaction between foreign subunits and their native complex partners may indicate that the transfer of an individual complex subunit is indeed likely to be neutral, similar to the view of the "complexity hypothesis" (Jain et al. 1999). Unlike other complexes that are encoded on operons or gene clusters and can consequently be cotransferred, escaping the complexity hurdle, the genes encoding the ACC complex are generally not found in a single locus. Specifically, the 4 ACC complex subunits are encoded on an operon in only 31 of the 233 bacterial genomes found in STRING (http://string.embl.de) that contain sequences >40% similar to AccAE. coli. The "phylogenetically discordant sequence" (PDS) estimator measures the extent to which a protein's phylogenetic signal matches most other proteins’ signals in a given genome by examining its similarity to its reciprocal best matches in other genomes (Clarke et al. 2002; Gophna et al. 2005). As may be expected from a non cluster-encoded complex AccA has a PDS value of 1.0, indicating that this gene displays a strong vertical signal and has probably experienced relatively few transfer events throughout microbial evolution.
Genes that are neutral to the recipient organism will often accumulate random mutations, eventually becoming pseudogenes. Nevertheless, it has been demonstrated that even in the absence of purifying selection, some genes can, and will, undergo fixation in a bacterial population (Novozhilov et al. 2005). Subsequently, a small fraction of the formerly neutral genes may even become beneficial following either a change in environmental conditions or additional gene acquisition events. Thus, barriers to LGT among bacteria appear to be low (McInerney and Pisani 2007) and are mere hurdles to the rampant gene exchange evident in some bacterial taxa.
The fact that imbalance of protein complexes does not appear to be a universal barrier to duplication or transfer of genes, leaves the molecular mechanisms behind the conserved single-copy status of specific proteins (Sorek et al. 2007) unknown. Without a general rule, one is left with the arduous task of characterizing each conserved single-copy gene in order to investigate the factors underlying its dosage sensitivity. However, because conserved single-copy genes tend to be essential (Wellner et al. 2007), such investigation is likely to be highly rewarding.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary table S1 and figures S2 and S3 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The authors thank Chen Katz, Yuval Mazor, and Aviram Rasouly for their help in implementing biochemical and genetic techniques and Adi Barzel, Martin Kupiec, and Eliora Z. Ron for critical reading of the manuscript. This work was supported by the Research Networks Program in Bioinformatics of the Ministry of Science and Technology of the State of Israel and the Ministries of Foreign Affairs and National Education and Research of France.
|
Footnotes |
|---|
Dan Graur, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Biran D, Gur E, Gollan L, Ron EZ. Control of methionine biosynthesis in Escherichia coli by proteolysis. Mol Microbiol (2000) 37:1436–1443.[CrossRef][Web of Science][Medline]
Clarke GD, Beiko RG, Ragan MA, Charlebois RL. Inferring genome trees by using a filter to eliminate phylogenetically discordant sequences and a distance matrix based on mean normalized BLASTP scores. J Bacteriol (2002) 184:2072–2080.
Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci USA (2000) 97:6640–6645.
Doolittle WF, Boucher Y, Nesbo CL, Douady CJ, Andersson JO, Roger AJ. How big is the iceberg of which organellar genes in nuclear genomes are but the tip? Philos Trans R Soc Lond B Biol Sci (2003) 358:39–57. discussion. 57–138.
Doyle M, Fookes M, Ivens A, Mangan MW, Wain J, Dorman CJ. An H-NS-like stealth protein aids horizontal DNA transmission in bacteria. Science (2007) 315:251–252.
Fang G, Rocha E, Danchin A. How essential are nonessential genes? Mol Biol Evol (2005) 22:2147–2156.
Forsyth RA, Haselbeck RJ, Ohlsen KL, et al, (22 co-authors). A genome-wide strategy for the identification of essential genes in Staphylococcus aureus. Mol Microbiol (2002) 43:1387–1400.[CrossRef][Web of Science][Medline]
Gerdes SY, Scholle MD, Campbell JW, et al, (21 co-authors). Experimental determination and system level analysis of essential genes in Escherichia coli MG1655. J Bacteriol (2003) 185:5673–5684.
Gophna U, Doolittle WF, Charlebois RL. Weighted genome trees: refinements and applications. J Bacteriol (2005) 187:1305–1316.
Guchhait RB, Polakis SE, Dimroth P, Stoll E, Moss J, Lane MD. Acetyl coenzyme A carboxylase system of Escherichia coli. Purification and properties of the biotin carboxylase, carboxyltransferase, and carboxyl carrier protein components. J Biol Chem (1974) 249:6633–6645.
Jain R, Rivera MC, Lake JA. Horizontal gene transfer among genomes: the complexity hypothesis. Proc Natl Acad Sci USA (1999) 96:3801–3806.
Knuth K, Niesalla H, Hueck CJ, Fuchs TM. Large-scale identification of essential Salmonella genes by trapping lethal insertions. Mol Microbiol (2004) 51:1729–1744.[CrossRef][Web of Science][Medline]
Kobayashi K, Ehrlich SD, Albertini A, et al, (99 co-authors). Essential Bacillus subtilis genes. Proc Natl Acad Sci USA (2003) 100:4678–4683.
McInerney JO, Pisani D. Genetics. Paradigm for life. Science (2007) 318:1390–1391.
Nakamura Y, Itoh T, Matsuda H, Gojobori T. Biased biological functions of horizontally transferred genes in prokaryotic genomes. Nat Genet (2004) 36:760–766.[CrossRef][Web of Science][Medline]
Novozhilov AS, Karev GP, Koonin EV. Mathematical modeling of evolution of horizontally transferred genes. Mol Biol Evol (2005) 22:1721–1732.
Papp B, Pal C, Hurst LD. Dosage sensitivity and the evolution of gene families in yeast. Nature (2003) 424:194–197.[CrossRef][Web of Science][Medline]
Sorek R, Zhu Y, Creevey CJ, Francino MP, Bork P, Rubin EM. Genome-wide experimental determination of barriers to horizontal gene transfer. Science (2007) 318:1449–1452.
Thomason LC, Costantino N, Court DL. E. coli Genome Manipulation by P1 Transduction. In: Current Protocols in Molecular Biology. New York: John Wiley & Sons. Chap. 1. Unit 1.17. (2007).
Tomoyasu T, Mogk A, Langen H, Goloubinoff P, Bukau B. Genetic dissection of the roles of chaperones and proteases in protein folding and degradation in the Escherichia coli cytosol. Mol Microbiol (2001) 40:397–413.[CrossRef][Web of Science][Medline]
Wellner A, Lurie MN, Gophna U. Complexity, connectivity, and duplicability as barriers to lateral gene transfer. Genome Biol (2007) 8:R156.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||