MBE Advance Access originally published online on September 22, 2008
Molecular Biology and Evolution 2008 25(12):2557-2565; doi:10.1093/molbev/msn210
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
The Proteomic Constraint and Its Role in Molecular Evolution
Molecular Biology Department, University of Wyoming
E-mail: stevenmassey{at}gmail.com.
 |
Abstract |
|---|
TOP
Abstract
Definition of the Proteomic...Recently, the concept of a "Proteomic Constraint" was introduced to explain the frequency of genetic code deviations in mitochondrial genomes. The Proteomic Constraint was proposed to be proportional to the size of the mitochondrially encoded proteome, hence small proteomes are expected to experience smaller total numbers of errors resulting from genetic code deviations, leading to less likelihood of causing lethality. The concept is now extended to encompass several other aspects of the genetic information system. When the Proteomic Constraint is small, it is proposed that there is little selective pressure to evolve or maintain error correction mechanisms, as a result of the smaller total number of errors that accumulate. Conversely, a large Proteomic Constraint is proposed to result in a correspondingly large selective pressure to evolve or maintain error correction mechanisms. Differences in the size of the Proteomic Constraint can help to explain differences in replicational, transcriptional, and translational fidelities between genomes. A key piece of evidence is the existence of negative power law relationships between proteome size and error rates; these are demonstrated to be diagnostic of the action of the Proteomic Constraint. The Proteomic Constraint is argued to be a major factor determining mutation rates in a diverse range of DNA genomes, implying that mutation rates are clock like. A small Proteomic Constraint partly explains why RNA viruses possess high mutation rates. A reduced Proteomic Constraint in intracellular pathogenic bacteria predicts a drift upwards in mutation rates. Differences in the Proteomic Constraint also appear to be linked to differences in recombination rates between eukaryotes. In addition, a reduced Proteomic Constraint may explain features of resident genomes, such as loss of DNA repair pathways, increased substitution rates, and AT biases, in addition to the occurrence of genetic code deviations. Thus, it is argued that the Proteomic Constraint is a universal factor that influences a wide range of properties of the genetic information system.
Key Words: Proteomic Constraint • error rate • mutation rate • codon reassignment • resident genome
|
Definition of the Proteomic Constraint |
|---|
I propose that the selective pressure to minimize informational error rates corresponds to the size of the proteome; this selective pressure is termed the "Proteomic Constraint." The upper case symbol
is used to represent the Proteomic Constraint (for the "P" of "Proteome"). The Proteomic Constraint is proportional to the size of the proteome; the larger the proteome, the larger the Proteomic Constraint:
 | (1) |
A variety of additional factors may affect the magnitude of the Proteomic Constraint. A refined expression for the Proteomic Constraint is
 |
 | (2) |
Nbinding is expected to be proportional to Nproteins, so may be taken out of the expression. Nlethal may be influenced by factors such as the average functional redundancy per open reading frame (ORF), the average number of pleiotropic interactions per ORF, and intrinsic and extrinsic mutational robustness. The latter refer to mutational robustness resulting from sequence properties (intrinsic robustness) or other factors such as the expression level of heat shock proteins or ploidy (extrinsic robustness; after Elena et al. 2006
). Nlethal is a simple way of describing the deleterious effects of mutations on an organism; clearly many mutations do not cause lethality, but do cause a significant loss of fitness.
As increases, the selective pressure to minimize error rates is increased as the size of the target for errors is increased (assuming that the majority of errors are deleterious). This will lead to a corresponding decrease in error rates and results in a negative power function between error rates and , assuming that the proportion of errors that are deleterious is constant in different proteomes. This may be illustrated as follows. Suppose that a proteome can tolerate one mutation per generation. A proteome of 10,000 codons would tolerate an error rate of 1 x 10–4 per codon. A proteome of 100,000 codons would tolerate an error rate of 1 x 10–5 per codon; likewise a proteome of 1,000,000 codons would tolerate an error rate of 1 x 10–6 per codon. This leads to a negative power relationship between proteome size and error rate:
 | (3) |
The exponent would be –1; c is referred to as the "Constraint Factor." The lower the value of the Constraint Factor, the higher the proportion of errors that are deleterious, that is, the greater the constraint on error rates. The error rate is expressed in terms of number of errors per informational unit (base pair, codon, or amino acid). A negative power function is therefore diagnostic of the influence of the Proteomic Constraint on fidelity. There follows an examination of the experimental evidence for the influence of the Proteomic Constraint on several features of the genetic information system.
|
The Proteomic Constraint and Variation in Mutation Rates |
|---|
The concept of a Proteomic Constraint provides an explanation for the variation in mutation rate per genome replication, among genomes encoding proteomes of different sizes. Most fitness affecting mutations are deleterious (Morgan 1903
, etc.); therefore, it is reasonable to expect that natural selection acts to minimize the occurrence of mutations. The more ORFs in a genome, the more ORFs that will be subject to mutation as the mutational target is larger. This means that the larger the proteome, the larger the likelihood of a deleterious mutation occurring. Therefore, the larger the proteome, the higher the selective pressure to minimize mutations. This should lead to a reduction in mutation rates in genomes encoding large proteomes.
RNA genes and promoter- and transcription factor–binding sites should also exert a selective pressure on mutation rates. The effect of RNA genes on mutation rates (
) may be expressed as follows:
 | (4) |
The mutation rate may be expressed as follows:
 | (5) |
As protein-coding genes comprise the large majority of the information content of the genome, the Proteomic Constraint is therefore likely to be the major determinant of mutation rate:
 | (6) |
From equation (3), we would expect:
 | (7) |
There follows an examination of the empirical evidence supporting the influence of the Proteomic Constraint on mutation rates. A distinction is made between substitution rate, which measures the rate of fixation of a mutation in a population, and the underlying mutation rate. Substitution rates reflect the underlying mutation rate and potentially additional population effects.
Mutation rates are higher in prokaryotes than in eukaryotes (Drake et al. 1998); this is consistent with the greater size of eukaryotic proteomes compared with prokaryotic proteomes. An interesting analysis was conducted by Ciccarelli et al. (2006). The authors demonstrate a negative correlation between the number of genes in a genome and substitution rates (as denoted by tree branch lengths) in eukaryotes and eubacteria. This finding is consistent with the influence of the Proteomic Constraint on underlying mutation rates.
A reduced Proteomic Constraint explains the elevation in substitution rates that have occurred in resident genomes, which have reduced proteomes, as the selective pressure to maintain low mutation rates is reduced, and mutation rates will be expected to drift upwards. "Resident" genome refers to nonviral genomes that have a nonnuclear location, that is, organelle, nucleomorph, obligate endosymbiotic bacterial, obligate intracellular pathogenic bacterial, and obligate intracellular protist genomes (after Andersson and Kurland 1998). Obligate endosymbiotic bacteria have undergone reductions in the size of their proteomes and experienced an elevation in substitution rates. For example, Buchnera aphidicola, an obligate bacterial endosymbiont of aphids, encodes only 638 proteins (van Ham et al. 2003) and has experienced an elevation in substitution rates (Itoh et al. 2002). Candidatus blochmannia, an obligate bacterial endosymbiont of ants, encodes 610 proteins (Degnan et al. 2005) and has also experienced an elevation in substitution rates (Degnan et al. 2004). A range of other obligate bacterial endosymbionts of invertebrates have also undergone reductions in the sizes of their proteomes and concomitant elevations in their substitution rates (Woolfit and Bromham 2003; Canback et al. 2004).
Nucleomorphs are residual nuclei arising from secondary endosymbioses of eukaryotic algae which have also undergone reductions in the size of their proteomes, encoding 331 (Gilson et al. 2006) and 464 (Douglas et al. 2001) proteins. These genomes also undergo a concomitant elevation in substitution rates (Patron et al. 2006). The obligate intracellular eukaryotic parasite Encephalitozoon cuniculi has also undergone an extreme reduction in proteome size (1,994 proteins, Katinka et al. 2001) and experiences high substitution rates (Keeling and Fast 2002). In addition to resident genomes, members of the Prochlorococcus genus of free-living cyanobacteria have undergone a reduction in proteome size (1716–2273 ORFs) and a concomitant elevation in substitution rate (Dufresne et al. 2005).
The absence of DNA repair pathways from genomes that encode small proteomes may explain why they experience high substitution rates. DNA repair genes are absent from a variety of genomes that encode small proteomes, for example Helicobacter pylori, Borrelia burdorferi, Chlamydia trachomatis, and Rickettsia prowazeki (Stepkowski and Legocki 2001), the mycoplasmas (Carvalho et al. 2005), B. aphidicola (Moran and Mira 2001), members of the Prochlorococcus genus (Dufresne et al. 2005), E. cuniculi (Gill and Fast 2007), and nucleomorphs (Douglas et al. 2001; Gilson et al. 2006). I propose that the reduction in the Proteomic Constraint reduces the selective pressure to maintain these pathways, resulting in loss of the pathways, and a concomitant increase in the underlying mutation rates in these genomes, leading to elevated substitution rates.
Mitochondria generally have high substitution rates, reflecting high underlying mutation rates (Denver et al. 2000, Caenorhabditis elegans; Lynch et al. 2008, yeast; Haag-Liautard et al. 2008, Drosophila melanogaster; these data are not plotted in fig. 1 below as the respective number of mitochondrial divisions undergone are not known). These observations are consistent with the highly reduced size of the mitochondrially encoded proteome, which is expected to exert a low Proteomic Constraint on mitochondrial mutation rates. This perspective at least partially answers the causes of mitochondrial "hypermutability," pondered by Nabholz et al. (2008). However, plant mitochondria tend to exhibit low substitution rates (Palmer and Herbon 1988), which seem to indicate a low underlying mutation rate. Determining whether the underlying mutation rate of plant mitochondria is indeed low, and the mechanistic causes if this is the case, will shed light on the thesis presented here.
|
Evidence from viruses also supports the influence of the Proteomic Constraint on mutation rates. DNA viruses that encode their own DNA polymerases have small proteomes and high mutation rates compared with nonviral genomes (Drake et al. 1998
). The ability of point mutations to induce an antimutator phenotype in Herpes simplex DNA polymerase (Hall et al. 1984) is consistent with the Proteomic Constraint being insufficiently large to select and maintain an antimutator phenotype. Many DNA viruses utilize host DNA polymerases, which are under a high Proteomic Constraint, due to the large size of the host proteome. This explains why these DNA viruses experience lower mutation rates than DNA viruses that encode their own DNA polymerases. RNA viruses possess the smallest of all proteomes and the highest mutation rates of any genome (for reviews, see Holland et al. 1982; Katz and Skalka 1990). RNA viruses code for only a few proteins, including a polymerase which can copy RNA into a complementary nucleic acid (usually RNA or, in the case of retroviruses, DNA) and structural proteins. Most RNA viruses encode an RNA-dependent RNA polymerase (RDRP), involved in replication of the virus (Ortin and Parra 2006, a review). RDRP lacks exonucleolytic editing, which causes high mutation rates (Domingo et al. 1997; Freistadt et al. 2007). RNA viruses that use reverse transcriptase (retroviruses) also have high mutation rates due to the lack of 3'- to 5'-proofreading activity of retroviral reverse transcriptase (Steinhauser et al. 1992). I suggest that the high mutation rates of RNA viruses occur partly as a consequence of an extremely small Proteomic Constraint.
The theory of a Proteomic Constraint implies that if the proportion of essential codons is the same in different genomes, then these genomes will possess a mutation rate proportional to the size of the proteome (eq. 6), resulting in a negative power relationship (eq. 7). This in turn implies that the number of mutations per proteome per replication will be equal for different genomes. Strong support comes from a consideration of the mutation rates from a range of genomes. Previously, a negative relationship was demonstrated between genome size and the number of mutations per generation in microbes with DNA genomes (Drake 1991; Drake et al. 1998). Consistent with these analyses, when µ is plotted against genome size a negative power function relates the two parameters (fig. 1a). A notable outlier to the correlation is bacteria Oenococcus oeni, which has a high mutation rate, probably due to the loss the mutS and mutL DNA repair genes (Marcobal et al. 2007). These losses are probably linked to a recent reduction in the proteome size of O. oeni, linked with a reduction in metabolic capacity (how a reduction of the Proteomic Constraint may lead to lead of DNA repair pathways is discussed above and below). When multicellular eukaryote mutation rates are added to the analysis, they deviate from the negative power relationship (fig. 1a). A simple explanation is that the genomes of multicellular eukaryotes are composed of a large proportion of noncoding DNA. Consistent with this, when µ is plotted against proteome size, the proteomes of both microbes and multicellular eukaryotes show a strong correlation to a negative power function (fig. 1b). The analysis is consistent with the interpretation that the coding portion of the genome exerts a selective constraint on mutation rates and implies that noncoding DNA does not exert a selective pressure to reduce mutations. This in turn is consistent with the observation that a large amount of this so-called "junk" DNA is nonfunctional in a sequence-specific sense. The analysis constitutes strong evidence for the influence of the Proteomic Constraint on mutation rates.
The correlation displayed in figure 1b implies that mutation rates have a negative power relationship to the total number of codons in a genome:
 | (8) |
This is analogous to equation (7): µ = c–1.
Inserting values derived from figure 1b, gives:
 | (9) |
 | (10) |
The approximate –1 value of the exponent for this function is strong evidence for the influence of the Proteomic Constraint on mutation rates. In this equation, the Constraint Factor is equivalent to 0.0007. Equation (9) allows µ to be predicted from proteome size and implies that underlying mutation rates are clock like, in that they follow a predictable relationship to proteome size, in DNA genomes that encode their own replication machinery. Deviations from the best-fit line may partly be due to inaccuracies in predicting proteome size.
An important feature of the analysis is that the mutation rates of multicellular organisms are corrected for the number of cell divisions that germ line cells undergo, so that the mutation rate is per genome replication (after Drake et al. 1998). When the correction is not conducted, the mutation rates of multicellular eukaryotes deviate markedly from the best-fit line (fig. 1c). An implication of this observation is that in the context of mutation rates a "generation" in multicellular eukaryotes is not the event of producing an offspring but better describes the number of cell divisions that a germ cell undergoes before syngamy occurs. This may seem paradoxical as mutations occurring in a germ cell are generally not expected to exert a phenotypic effect, so the number of cell divisions that it undergoes cannot directly influence the evolution of mutation rates. However, even though mutations in somatic cells or tissues do not enter the germ line, constituting the Weismann barrier, they do exert a selective pressure to minimize replicational errors via phenotypic effects in the soma. The germ line mutation rate per cell division, despite some differences, should theoretically reflect that of somatic cells as they utilize the same replicational machinery, but selection for reduced mutation rates occurs in the soma. Differences in mutation rates per organismal generation in multicellular eukaryotes largely result from differences in the number of genome replications undergone by germ cells (as well as differences in proteome size). This has implications for the definition of generation in multicellular eukaryotes: "Generation" may be defined as the number of genome replications undergone by germ cells in the context of underlying mutation rates, whereas "generation" may be defined as the organismal generation in the context of substitution rates.
When the mutation rates of RNA genomes (11 RNA viruses and 1 retrotransposon) are plotted against their respective proteome sizes, they are high, consistent with a low Proteomic Constraint, but deviate from the negative power relationship observed for DNA genomes, being elevated in comparison (fig. 1d). RNA virus mutation rates do not appear to display a negative power relationship with RNA virus proteome size; however, the data are unclear and might become clearer when more mutation rate data becomes available. The elevated nature of RNA virus mutation rates suggests that there is a factor in addition to the size of the Proteomic Constraint, leading to an additional elevation of mutation rates in RNA genomes. At a mechanistic level, this appears to be due to the lack of proofreading activities of viral reverse transcriptase and RDRP; however, the reasons for the lack of proofreading remain to be determined.
The proposal that the Proteomic Constraint influences viral mutation rates stands in direct contrast to the hypothesis that viral mutation rates limit the size of RNA virus genomes (Eigen et al. 1988). The Proteomic Constraint implies that if a virus has no need to evolve new genes, thereby increasing the size of its proteome, then it has no need to evolve a more accurate polymerase. Clearly, the observation that DNA and RNA viruses with similar proteome sizes display very different mutation rates is evidence against the hypothesis of Eigen et al. In addition, RNA viruses of similar proteome sizes vary substantially in their mutation rates, which is further evidence against the hypothesis. Other factors are known to influence genome size in viruses, such as the capsid packaging constraint. The hypothesis that mutation rates limit genome size has also been invoked to account for prokaryotic genome size (Eigen and Schuster 1978). The observation that O. oeni possess a substantially elevated mutation rate compared with other prokaryotes of similar proteome size is evidence against this hypothesis.
The analysis implies that the proportion of total mutations that are deleterious is similar in a diverse range of genomes. The analysis is also consistent with most mutations arising from replicational errors. The Proteomic Constraint is expected to exert a selective pressure to minimize the occurrence of nonreplicational mutations also. To conclude, the Proteomic Constraint is a major factor determining mutation rates in a diverse range of genomes. The theory of a Proteomic Constraint implies that once mutation rates have been reduced to a certain level in nature, there will be insufficient selective pressure to reduce them further, answering the question of why mutation rates do not fall to zero (Sturtevant 1937).
|
The Proteomic Constraint and Genomic AT Bias |
|---|
The great majority of resident genomes are AT biased (Gilson et al. 1997
; Moran and Wernegreen 2000). The loss of DNA repair genes has been proposed as a cause of AT bias in resident genomes (Wernegreen and Funk 2004). If a reduction in the Proteomic Constraint in these genomes leads to a reduction in the selective pressure to maintain these DNA repair pathways, as discussed above, a reduction in the Proteomic Constraint may indirectly explain the strong mutational biases experienced by resident genomes. This explanation also applies to the genomes of the free-living Prochlorococcus cyanobacteria, which have also undergone marked reductions in proteome size and concomitant elevations in AT biases. When GC/AT content is plotted against proteome size in eubacteria, a cluster of bacteria with small genomes and elevated AT contents is revealed (fig. 2). The majority of these are resident genomes, indicated on the plot in red. Twenty-one different genuses are found in the cluster of resident genomes; this is a case of convergent evolution at the molecular level and indicative of the potential influence of the Proteomic Constraint on diverse lineages of bacteria.
|
|
The Proteomic Constraint and Variations in Translational Fidelity |
|---|
A simple scenario illustrates the potential effects of the Proteomic Constraint on translational fidelity. Suppose a mutation in the translation system results in the misincorporation of amino acid "a" for amino acid "b" once in every mRNA transcript. If 1 in 10 of the mRNAs produce an inactive protein product on average and a genome encodes 20 essential ORFs that are equally expressed, for 200 mRNAs there are 20 inactivated proteins. Alternatively, if a genome encodes 200 essential ORFs that are equally expressed, the sum is 200 inactivated proteins. In the latter case, there will be a larger fitness loss, making the assumption that the effects of translational errors are cumulative. The lower the transcriptional redundancy, the more likely a misincorporation event will be to exert a phenotypic effect. Therefore, the more proteins that are expressed and the lesser the transcriptional redundancy, the greater the negative effects of translational infidelity on fitness. Hence, it may be understood how the Proteomic Constraint can exert a selective pressure on translational fidelity, and a reduction in the Proteomic Constraint can lead to reduced selection against translational errors.
Empirical evidence supports the influence of the Proteomic Constraint on translational fidelity. Gauging the differences in translational fidelity between prokaryotes and eukaryotes is difficult as it depends on the anticodon properties of the tRNAs responsible for decoding different codons, codon context, and relative tRNA abundance. This is reflected in differing error rates observed for differing codons (e.g., Ortego et al. 2007). However, a reduced Proteomic Constraint can explain features of the translation systems of resident genomes that have hitherto been difficult to explain. For example, mitochondrially encoded tRNAs and rRNAs show gross changes in secondary structure, undergoing large numbers of deletions, and an increase in the number of base mispairings; the stabilities of mitochondrially encoded rRNAs and tRNA are consequently reduced (Lynch 1996; Lynch and Blanchard 1998). A reduced Proteomic Constraint in mitochondria reduces the selective pressure to maintain the integrity of mitochondrially encoded tRNAs and rRNAs and may explain these observations. A reduced Proteomic Constraint also provides an explanation for the accumulation of slightly deleterious mutations in mitochondrial and chloroplast tRNA and rRNA genes (Lynch 1996; Lynch and Blanchard 1998), in B. aphidicola 16S rRNA genes (Lambert and Moran 1998), and in the gene encoding B. aphidicola elongation factor (Brynnel et al. 1998). This is in addition to the proposed influence of Muller's ratchet on resident genome evolution (Lynch 1996; Moran 1996). The Proteomic Constraint is not expected to affect operational genes encoded by resident genomes, whereas Muller's ratchet should affect both informational and operational genes.
Aminoacyl-tRNA synthetases charge tRNAs with amino acids; the accuracy of tRNA and amino acid recognition by these enzymes is vital for translational fidelity. The proofreading activity of mitochondrial leucyl-tRNA synthetase (leuRS) (which is encoded in the nucleus but targeted to the mitochondrion) is dispensable for viability in Saccharomyces cerevisiae; however, the proofreading activity of cytoplasmic leuRS is vital for the viability of Escherichia coli (Karkhanis et al. 2006). Loss of proofreading activity in mouse cytoplasmic alanyl-tRNA synthetase (alaRS) leads to tRNA mischarging, protein misfolding, and neurodegeneration (Lee et al. 2006), and the loss of proofreading in E. coli cytoplasmic valyl-tRNA synthetase (valRS) leads to loss of viability (Nangle et al. 2002). However, S. cerevisiae mitochondrial pheRS has lost its proofreading capacity, resulting in reduced aminoacylation fidelity compared with its cytoplasmic counterpart (Roy et al. 2005), consistent with the reduced Proteomic Constraint of the mitochondrial proteome. These data provide evidence for the existence of a reduced Proteomic Constraint on translational fidelity in mitochondria and a higher Proteomic Constraint on cytoplasmic translational fidelity.
An additional feature that may reflect a reduced Proteomic Constraint in resident genomes is the reduced tRNA complements present in these genomes (e.g., Muto et al. 1990; Knight et al. 2001). The tRNA complements of resident genomes do not accurately reflect their codon usage biases, which mainly result from strong AT biases. This is expected to result in suboptimal translational efficiency. Codon usage biases in free-living organisms appear largely to have evolved to minimize the detrimental effects of mistranslation (Archetti 2004; Najafabadi et al. 2007; Stoletski and Eyre-Walker 2007), that is, increase translational fidelity. A reduced Proteomic Constraint may reduce the selective pressure to optimize tRNA complements and codon usage biases.
|
The Proteomic Constraint and Variations in Transcriptional Fidelity |
|---|
The same arguments invoked for the influence of the Proteomic Constraint on replicational and translational fidelities also apply to transcriptional fidelity. This assumes that a significant proportion of errors in mRNAs result in expressed proteins that have deleterious effects and that such effects are additive. Although transcriptional fidelity has not been examined in great detail, there is some supporting evidence. The Proteomic Constraint predicts that the fidelities of DNA virus RNA polymerases are less than that of their cellular counterparts. Supporting this, the fidelity of T7 RNA polymerase (Huang et al. 2000
) is lower than that of eukaryotic RNA polymerases II and III (Alic et al. 2007).
A corollary to this discussion is a consideration of the relative fidelities of eukaryotic RNA polymerases II and III (RNA pol II and RNA pol III). RNA pol III, which transcribes tRNAs, has higher fidelity than RNA pol II, which transcribes mRNAs, indicating that there has been a stronger selective pressure to increase the fidelity of RNA pol III compared with RNA pol II (Alic et al. 2007). This may be understood by considering that the effects of transcriptional errors on the proteome are likely to be greater for RNA pol III than RNA pol II because the effects of transcriptional errors resulting from RNA pol II are amplified via tRNAs with altered translational fidelities. A clear example is where a tRNA anticodon is mutated to code for a different amino acid according to the genetic code, which is usually lethal. Therefore, the total proteomic information that is affected is larger resulting from a transcriptional error in an individual tRNA rather than in an individual mRNA coding for an operational protein. Likewise, the fidelity of eukaryotic RNA pol I is also expected to be higher than that of RNA pol II as it is responsible for transcribing rRNA genes, which are involved in protein translation. For these reasons, fidelity of prokaryotic RNA polymerases (which transcribe all three RNAs; rRNAs, tRNAs, and mRNAs) are difficult to compare directly to those of eukaryotic RNA polymerases, which have a polymerase allocated to each of these RNAs individually. This means that indirect in vivo measurement of mRNA transcriptional error rates using reporter systems is more useful for comparing error rates.
A pertinent question to this work is why mutation rates are typically several orders of magnitude lower than those of translational and transcriptional error rates (Burger et al. 2006). As discussed, the Proteomic Constraint is likely to be the main factor in determining mutation rates and is likely to be a factor in determining translational/transcriptional error rates. Additional factors are likely to be responsible for the elevated character of translational/transcriptional error rates compared with mutation rates. One important difference between the two types of error rates is that redundancy (in terms of elevated expression of mRNAs and the number of protein molecules produced per transcript) can be directly selected for in order to ameliorate the deleterious effects of translational/transcriptional errors. A related argument is made by Burger et al. (2006). In contrast, pure functional gene redundancy is rare at the DNA level (Wagner 2005) and explicit selection for gene redundancy at the DNA level in order to produce mutational robustness is hard to envisage due to the need to invoke group selection. Expression redundancy is unlikely to be able to compensate for the occurrence of mutations in DNA encoded ORFs as all resulting transcripts and proteins are affected. In the presence of mRNA and protein redundancy, considering that transcriptional/translational errors are random, feedback mechanisms play a role in regulating the optimal amount of protein in a cell.
|
The Proteomic Constraint and Variation in Recombination Rates |
|---|
The Proteomic Constraint also appears to influence variation in recombination rates between eukaryotes. Previously, a negative relationship between genome size and recombination rate was noted (Lynch 2006
). When recombination rate is plotted against proteome size, a negative power function relates the two parameters (fig. 3a). In the analysis, recombination rate is plotted per organismal generation, as opposed to per genome replication (fig. 1). This is because meiosis occurs only once per organismal generation.
|
The negative power relationship is indicative of the action of the Proteomic Constraint. The relationship displayed in figure 3 implies that the Proteomic Constraint is a major factor affecting the frequency of recombination events in eukaryotes, implying that the Proteomic Constraint exerts a selective pressure to minimize recombination rate. There is a requirement for at least one crossover per chromosome arm during meiosis; hence, it was investigated whether the number of chromosomes influences recombination rates (fig. 3b). No meaningful correlation is observed, reinforcing the conclusion that the major factor influencing recombination rate is the Proteomic Constraint. The Proteomic Constraint may affect recombination rates either directly or indirectly. Consistent with a direct effect is the observation that there is an increasing elevated deviation from the best-fit line with increasing genome size (Lynch 2006
), similar to the elevated deviation of Mus and Homo observed when mutation rates are plotted against genome size (fig. 1a). These observations might contribute to considerations of the evolution of sexual recombination.
Inserting values from figure 3a leads to the following:
 | (11) |
The relationship is not as strong as for mutation rates (R = –0.94 for mutation rates, fig. 1b; R = –0.80 for recombination rates, fig. 3a). This is consistent with factors additional to the Proteomic Constraint influencing recombination rates.
|
The Proteomic Constraint and Variations in the Mitochondrial Genetic Code |
|---|
The concept of a Proteomic Constraint was initially formulated to explain the occurrence of genetic code deviations (codon reassignments) in mitochondrial and mycoplasma genomes (Massey and Garey 2007
), which have experienced reductions in proteome size. Evidence was presented that a reduced Proteomic Constraint allows genetic code changes to occur in mitochondrial genomes with less likelihood of causing lethality. A negative power relationship was revealed between the number of codon reassignments undergone by mitochondrial genomes (Massey and Garey 2007) and is also observed with mitochondrial proteome sizes (data not shown). The Proteomic Constraint was initially formulated in qualitative terms and the reason for the negative power relationship was not elucidated. Consideration of the effects of the Proteomic Constraint on other aspects of the information system led to this work. Exegesis of the negative power function presented here provides some further evidence for the role of the Proteomic Constraint in influencing the number of codon reassignments undergone by mitochondrial genomes. In this context, the number of codon reassignments undergone by a mitochondrial genome is analogous to an error rate, in that these have occurred over a period of time, that is since the single endosymbiotic event that gave rise to mitochondria in all extant eukaryotes. This perspective regards mitochondrial codon reassignments as having a deleterious component. |
Conclusion: , a Universal Factor in Molecular Evolution
|
|---|
The purpose of this paper is to lay a groundwork for studying the occurrence of errors in the genetic information system. The Proteomic Constraint paradigm applies to all proteomes and provides a rationale for several aspects of molecular evolution that have hitherto been difficult to explain. One of these is differences in mutation rates in eukaryotes, prokaryotes, viruses, and resident genomes. Consideration of the Proteomic Constraint facilitates incorporation of a protein functional dimension and prompts a redefinition of the term generation in the context of multicellular organisms. Other aspects include differences in transcriptional and translational fidelities and recombination rates between genomes and the occurrence of codon reassignments in mitochondrial and mycoplasma genomes. The negative power relationship revealed between
and mutation rates with an exponent of –1, and the negative power relationships with recombination rates and the number of codon reassignments in mitochondrial genomes are key evidence for the existence of a Proteomic Constraint. The theory and analyses presented here imply that the Proteomic Constraint is an important factor in molecular evolution. |
Supplementary Material |
|---|
Supplementary tables are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
I would like to thank Dr J.W. Drake (National Institutes of Health) for valuable comments on mutation rates, Dr D. Jarvis (University of Wyoming) for valuable comments on virus molecular biology, and two anonymous referees for their insightful comments.
|
Footnotes |
|---|
Present address: Biology Department, University of Puerto Rico, San Juan, Puerto Rico.
Michele Vendruscolo, Associate Editor
|
References |
|---|
Alic N, Ayoub N, Landrieux E, Favry E, Baudouin-Cornu P, Riva M, Carles C. Selectivity and proofreading both contribute significantly to the fidelity of RNA polymerase III transcription. Proc Natl Acad Sci USA (2007) 104:10400–10405.
Andersson SG, Kurland CG. Reductive evolution of resident genomes. Trends Microbiol (1998) 6:263–268.[CrossRef][Web of Science][Medline]
Archetti M. Selection on codon usage for error minimization at the protein level. J Mol Evol (2004) 59:400–415.[CrossRef][Web of Science][Medline]
Brynnel EU, Kurland CG, Moran NA, Andersson SG. Evolutionary rates for tuf genes in endosymbionts of aphids. Mol Biol Evol (1998) 15:574–582.[Abstract]
Burger R, Willensdorfer M, Nowak MA. Why are phenotypic mutation rates much higher than genotypic mutation rates? Genetics (2006) 172:197–206.
Canback B, Tamas I, Andersson SG. A phylogenomic study of endosymbiotic bacteria. Mol Biol Evol (2004) 21:1110–1122.
Carvalho FM, Fonseca MM, Batistuzzo de Medeiros S, Scortecci KC, Blaha CA, Agnez-Lima LF. DNA repair in reduced genome: the Mycoplasma model. Gene (2005) 360:111–119.[CrossRef][Web of Science][Medline]
Ciccarelli FD, Doerks T, von Mering C, Creevey CJ, Snel B, Bork P. Toward automatic reconstruction of a highly resolved tree of life. Science (2006) 311:1283–1287.
Degnan PH, Lazarus AB, Brock CD, Wernegreen JJ. Host-symbiont stability and fast evolutionary rates in an ant-bacterium association: cospeciation of Camponotus species and their endosymbionts, Candidatus blochmannia. Syst Biol (2004) 53:95–100.
Degnan PH, Lazarus AB, Wernegreen JJ. Genome sequence of Blochmannia pennsylvanicus indicates parallel evolutionary trends among bacterial mutualists of insects. Genome Res (2005) 15:1023–1033.
Denver DR, Morris K, Lynch M, Vassilieva LL, Thomas WK. High direct estimate of the mutation rate in the mitochondrial genome of Caenorhabditis elegans. Science (2000) 289:2342–2344.
Douglas S, et al. The highly reduced genome of an enslaved nucleus. Nature (2001) 410:1091–1096.[CrossRef][Web of Science][Medline]
Drake JW. A constant rate of spontaneous mutation in DNA-based microbes. Proc Natl Acad Sci USA (1991) 88:7160–4164.
Drake JW, Charlesworth B, Charlesworth D, Crow JF. Rates of spontaneous mutation. Genetics (1998) 148:1667–1686.
Dufresne A, Garczarek L, Partensky F. Accelerated evolution associated with genome reduction in a free-living prokaryote. Genome Biol (2005) 6:R14.[CrossRef][Medline]
Eigen M, McCaskill J, Schuster P. Molecular quasispecies. J Phys Chem (1988) 92:6881–6891.[CrossRef][Web of Science]
Eigen M, Schuster P. The hypercycle: a principle of natural self-organization. Part A: emergence of the hypercycle. Naturwissenschaften (1978) 64:541–565.
Elena SF, Carrasco P, Daros J, Sanjuan R. Mechanisms of genetic robustness in RNA viruses. EMBO Rep (2006) 7:168–173.[CrossRef][Web of Science][Medline]
Freistadt MS, Vaccaro JA, Eberle KE. Biochemical characterization of the fidelity of poliovirus RNA-dependent RNA polymerase. Virol J (2007) 4:44.[CrossRef][Medline]
Gill EE, Fast NM. Stripped-down DNA repair in a highly reduced parasite. BMC Mol Biol (2007) 8:24.[CrossRef][Medline]
Gilson PR, Meier U, McFadden GI. Size isn‘t everything: lessons in genetic miniaturisation from nucleomorphs. Curr Opin Genet Dev (1997) 7:800–806.[CrossRef][Web of Science][Medline]
Gilson PR, Su V, Slamovits CH, Reith ME, Keeling PJ, McFadden GI. Complete nucleotide sequence of the chlorarachniophyte nucleomorph: nature's smallest nucleus. Proc Natl Acad Sci USA (2006) 103:9566–9571.
Haag-Liautard C, Coffey N, Houle D, Lynch M, Charlesworth B, Keightley PD. Direct estimation of the mitochondrial DNA mutation rate in Drosophila melanogaster. PLoS Biol (2008) 6:e204.
Hall JD, Coen DM, Fisher BL, Weisslitz M, Randall S, Almy RE, Gelep PT, Schaffer PA. Generation of genetic diversity in herpes simplex virus: an antimutator phenotype maps to the DNA polymerase locus. Virology (1984) 132:26–37.[CrossRef][Web of Science][Medline]
Holland J, Spindler K, Horodyski F, Grabau E, Nichol S, VandePol S. Rapid evolution of RNA genomes. Science (1982) 215:1577–1585.
Huang J, Brieba LG, Sousa R. Misincorporation by wild-type and mutant T7 RNA polymerases: identification of interactions that reduce misincorporation rates by stabilizing the catalytically incompetent open conformation. Biochemistry (2000) 39:11571–11580.
Itoh T, Martin W, Nei M. Acceleration of genomic evolution caused by enhanced mutation rate in endocellular symbionts. Proc Natl Acad Sci USA (2002) 99:12944–12948.
Karkhanis VA, Bonlecki MT, Poruri K, Martinis SA. A viable amino acid editing activity in the leucyl-tRNA synthetase CP1-splicing domain is not required in the yeast mitochondria. J Biol Chem (2006) 281:33217–33225.
Katinka MD, et al. Genome sequence and gene compaction of the eukaryotic parasite Encephalitozoon cuniculi. Nature (2001) 414:450–453.[CrossRef][Web of Science][Medline]
Katz RA, Skalka AM. Generation of diversity in retroviruses. Annu Rev Gen (1990) 24:409–445.[CrossRef][Web of Science][Medline]
Keeling PJ, Fast NM. Microsporidia: biology and evolution of highly reduced intracellular parasites. Annu Rev Microbiol (2002) 56:93–116.[CrossRef][Web of Science][Medline]
Knight RD, Landweber LF, Yarus M. How mitochondria redefine the code. J Mol Evol (2001) 53:299–313.[CrossRef][Web of Science][Medline]
Lambert JD, Moran NA. Deleterious mutations destabilize ribosomal RNA in endosymbiotic bacteria. Proc Natl Acad Sci USA (1998) 95:4458–4462.
Lee JW, et al. Editing-defective tRNA synthetase causes protein misfolding and neurodegeneration. Nature (2006) 443:50–55.[CrossRef][Web of Science][Medline]
Lynch M. Mutation accumulation in transfer RNAs: molecular evidence for Muller's ratchet in mitochondrial genomes. Mol Biol Evol (1996) 13:209–220.[Abstract]
Lynch M. The origins of eukaryotic gene structure. Mol Biol Evol (2006) 23:450–468.
Lynch M, Blanchard JL. Deleterious mutation accumulation in organelle genomes. Genetica (1998) 102–103:29–39.[CrossRef]
Lynch M, Sung W, Morris K, et al, (11 co-authors). A genome-wide view of the spectrum of spontaneous mutations in yeast. Proc Natl Acad Sci USA (2008) 105:9272–9277.
Marcobal AM, Sela D, Wolf YI, Makarova KS, Mills DA. Role of hypermutability in the evolution of the genus Oenococcus. J Bacteriology (2007) 190:564–570.
Massey SE, Garey JR. A comparative genomics analysis of codon reassignments reveals a link with mitochondrial proteome size and a mechanism of genetic code change via suppressor tRNAs. J Mol Evol (2007) 64:399–410.[CrossRef][Web of Science][Medline]
Moran NA. Accelerated evolution and Muller's ratchet in endosymbiotic bacteria. Proc Natl Acad Sci USA (1996) 93:2873–2878.
Moran NA, Mira A. The process of genome shrinkage in the obligate symbiont Buchnera aphidicola. Genome Biol (2001) 2:054.
Moran NA, Wernegreen JJ. Lifestyle evolution in symbiotic bacteria: insights from genomics. Trends Ecol Evol (2000) 15:321–326.[CrossRef][Medline]
Morgan TH. Evolution and adaptation (1903) New York: MacMillan.
Muto A, Andachi Y, Yuzawa H, Yamao F, Osawa S. The organization and evolution of transfer RNA genes in Mycoplasma capricolum. Nucleic Acids Res (1990) 18:5037–5043.
Nabholz B, Glemin S, Galtier N. Strong variations of mitochondrial mutation rate across mammals—the longevity hypothesis. Mol Biol Evol (2008) 25:120–130.
Najafabadi HS, Lehmann J, Omidi M. Error minimization explains the codon usage of highly expressed genes in Escherichia coli. Gene (2007) 387:150–155.[CrossRef][Web of Science][Medline]
Nangle LA, de Crecy Lagard V, Doring V, Schimmel P. Genetic code ambiguity. Cell viability related to the severity of editing defects in mutant tRNA synthetases. J Biol Chem (2002) 277:45729–45733.
Ortego BC, Whittenton JJ, Li H, Tu S-C, Willson RC. In vivo translational inaccuracy in Escherichia coli: missense reporting using extremely low activity mutants of Vibrio harveyi luciferase. Biochemistry (2007) 46:13864–13873.
Ortin J, Parra F. Structure and function of RNA replication. Annu Rev Microbiol (2006) 60:305–326.[CrossRef][Web of Science][Medline]
Palmer JD, Herbon LA. Plant mitochondrial DNA evolves rapidly in structure, but slowly in sequence. Proc Natl Acad Sci USA (1988) 28:87–97.
Patron NJ, Rogers MB, Keeling PJ. Comparative rates of evolution in endosymbiotic nuclear genomes. BMC Evol Biol (2006) 6:46.[CrossRef][Medline]
Roy H, Ling J, Alfonzo J, Ibba M. Loss of editing activity during the evolution of mitochondrial phenylalanyl-tRNA synthetase. J Biol Chem (2005) 280:38186–38192.
Steinhauser DA, Domingo E, Holland JJ. Lack of evidence for proofreading mechanisms associated with an RNA virus polymerase. Gene (1992) 122:281–288.[CrossRef][Web of Science][Medline]
Stepkowski T, Zegocki AB. Reduction of bacterail genome size and expansion resulting from obligate intracellular lifestyle and adaptation to soil habitiat. Acta Biochim Pol. (2001) 48:367–381.[Web of Science][Medline]
Stoletski N, Eyre-Walker A. Synonymous codon usage in Eschetichia coli selection for translational accuracy. Mol Biol Evol (2007) 24:374–381.
Sturtevant AH. Essays on evolution. I. On the effects of selection on the mutation rate. Q Rev Biol (1937) 12:464–476.[CrossRef]
van Ham RC, et al. Reductive genome evolution in Buchnera aphidicola. Proc Natl Acad Sci USA (2003) 100:581–586.
Wagner A. Distributed robustness versus redundancy as causes of mutational robustness. Bioessays (2005) 27:176–188.[CrossRef][Web of Science][Medline]
Wernegren JJ, Funk DJ. Mutation exposed: a neutral explanation for extreme base composition of an endosymbiont genome. J Mol Evol (2004) 59:849–858.[CrossRef][Web of Science][Medline]
Woolfit M, Bromham L. Increased rates of sequence evolution in endosymbiotic bacteria and fungi with small effective population sizes. Mol Biol Evol (2003) 20:1545–1555.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||