Accurate inference of orthologous genes is a pre-requisite for most comparative genomics studies, and is also important for functional annotation of new genomes. Identification of orthologous gene sets typically involves phylogenetic tree analysis, heuristic algorithms based on sequence conservation, synteny analysis, or some combination of these approaches. The most direct tree-based methods typically rely on the comparison of an individual gene tree with a species tree. Once the two trees are accurately constructed, orthologs are straightforwardly identified by the definition of orthology as those homologs that are related by speciation, rather than gene duplication, at their most recent point of origin. Although ideal for the purpose of orthology identification in principle, phylogenetic trees are computationally expensive to construct for large numbers of genes and genomes, and they often contain errors, especially at large evolutionary distances. Moreover, in many organisms, in particular prokaryotes and viruses, evolution does not appear to have followed a simple ‘tree-like’ mode, which makes conventional tree reconciliation inapplicable. Other, heuristic methods identify probable orthologs as the closest homologous pairs or groups of genes in a set of organisms. These approaches are faster and easier to automate than tree-based methods, with efficient implementations provided by graph-theoretical algorithms enabling comparisons of thousands of genomes. Comparisons of these two approaches show that, despite conceptual differences, they produce similar sets of orthologs, especially at short evolutionary distances. Synteny also can aid in identification of orthologs. Often, tree-based, sequence similarity- and synteny-based approaches can be combined into flexible hybrid methods.

We compare the sets of experimentally validated long intergenic non-coding (linc)RNAs from human and mouse and apply a maximum likelihood approach to estimate the total number of lincRNA genes as well as the size of the conserved part of the lincRNome. Under the assumption that the sets of experimentally validated lincRNAs are random samples of the lincRNomes of the corresponding species, we estimate the total lincRNome size at approximately 40,000 to 50,000 species, at least twice the number of protein-coding genes. We further estimate that the fraction of the human and mouse euchromatic genomes encoding lincRNAs is more than twofold greater than the fraction of protein-coding sequences. Although the sequences of most lincRNAs are much less strongly conserved than protein sequences, the extent of orthology between the lincRNomes is unexpectedly high, with 60 to 70% of the lincRNA genes shared between human and mouse. The orthologous mammalian lincRNAs can be predicted to perform equivalent functions; accordingly, it appears likely that thousands of evolutionarily conserved functional roles of lincRNAs remain to be characterized.

Author Summary

Genome analysis of humans and other mammals reveals a surprisingly small number of protein-coding genes, only slightly over 20,000 (although the diversity of actual proteins is substantially augmented by alternative transcription and alternative splicing). Recent analysis of the mammalian genomes and transcriptomes, in particular, using the RNAseq technology, shows that, in addition to protein-coding genes, mammalian genomes encode many long non-coding RNAs. For some of these transcripts, various regulatory functions have been demonstrated, but on the whole the repertoire of long non-coding RNAs remains poorly characterized. We compared the identified long intergenic non-coding (linc)RNAs from human and mouse, and employed a specially developed statistical technique to estimate the size and evolutionary conservation of the human and mouse lincRNomes. The estimates show that there are at least twice as many human and mouse lincRNAs than there are protein-coding genes. Moreover, about two third of the lincRNA genes appear to be conserved between human and mouse, implying thousands of conserved but still uncharacterized functions.

Evolution of prokaryotes involves extensive loss and gain of genes, which lead to substantial differences in the gene repertoires even among closely related organisms. Through a wide range of phylogenetic depths, gene frequency distributions in prokaryotic pangenomes bear a characteristic, asymmetrical U-shape, with a core of (nearly) universal genes, a “shell” of moderately common genes, and a “cloud” of rare genes. We employ mathematical modeling to investigate evolutionary processes that might underlie this universal pattern. Gene frequency distributions for almost 400 groups of 10 bacterial or archaeal species each over a broad range of evolutionary distances were fit to steady-state, infinite allele models based on the distribution of gene replacement rates and the phylogenetic tree relating the species in each group. The fits of the theoretical frequency distributions to the empirical ones yield model parameters and estimates of the goodness of fit. Using the Akaike Information Criterion, we show that the neutral model of genome evolution, with the same replacement rate for all genes, can be confidently rejected. Of the three tested models with purifying selection, the one in which the distribution of replacement rates is derived from a stochastic population model with additive per-gene fitness yields the best fits to the data. The selection strength estimated from the fits declines with evolutionary divergence while staying well outside the neutral regime. These findings indicate that, unlike some other universal distributions of genomic variables, for example, the distribution of paralogous gene family membership, the gene frequency distribution is substantially affected by selection.

Background

Collections of Clusters of Orthologous Genes (COGs) provide indispensable tools for comparative genomic analysis, evolutionary reconstruction and functional annotation of new genomes. Initially, COGs were made for all complete genomes of cellular life forms that were available at the time. However, with the accumulation of thousands of complete genomes, construction of a comprehensive COG set has become extremely computationally demanding and prone to error propagation, necessitating the switch to taxon-specific COG collections. Previously, we reported the collection of COGs for 41 genomes of Archaea (arCOGs). Here we present a major update of the arCOGs and describe evolutionary reconstructions to reveal general trends in the evolution of Archaea.

Results

The updated version of the arCOG database incorporates 91% of the pangenome of 120 archaea (251,032 protein-coding genes altogether) into 10,335 arCOGs. Using this new set of arCOGs, we performed maximum likelihood reconstruction of the genome content of archaeal ancestral forms and gene gain and loss events in archaeal evolution. This reconstruction shows that the last Common Ancestor of the extant Archaea was an organism of greater complexity than most of the extant archaea, probably with over 2,500 protein-coding genes. The subsequent evolution of almost all archaeal lineages was apparently dominated by gene loss resulting in genome streamlining. Overall, in the evolution of Archaea as well as a representative set of bacteria that was similarly analyzed for comparison, gene losses are estimated to outnumber gene gains at least 4 to 1. Analysis of specific patterns of gene gain in Archaea shows that, although some groups, in particular Halobacteria, acquire substantially more genes than others, on the whole, gene exchange between major groups of Archaea appears to be largely random, with no major ‘highways’ of horizontal gene transfer.

Conclusions

The updated collection of arCOGs is expected to become a key resource for comparative genomics, evolutionary reconstruction and functional annotation of new archaeal genomes. Given that, in spite of the major increase in the number of genomes, the conserved core of archaeal genes appears to be stabilizing, the major evolutionary trends revealed here have a chance to stand the test of time.

Reviewers

This article was reviewed by (for complete reviews see the Reviewers’ Reports section): Dr. PLG, Prof. PF, Dr. PL (nominated by Prof. JPG).

ABSTRACT

Bacteria and archaea face continual onslaughts of rapidly diversifying viruses and plasmids. Many prokaryotes maintain adaptive immune systems known as clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated genes (Cas). CRISPR-Cas systems are genomic sensors that serially acquire viral and plasmid DNA fragments (spacers) that are utilized to target and cleave matching viral and plasmid DNA in subsequent genomic invasions, offering critical immunological memory. Only 50% of sequenced bacteria possess CRISPR-Cas immunity, in contrast to over 90% of sequenced archaea. To probe why half of bacteria lack CRISPR-Cas immunity, we combined comparative genomics and mathematical modeling. Analysis of hundreds of diverse prokaryotic genomes shows that CRISPR-Cas systems are substantially more prevalent in thermophiles than in mesophiles. With sequenced bacteria disproportionately mesophilic and sequenced archaea mostly thermophilic, the presence of CRISPR-Cas appears to depend more on environmental temperature than on bacterial-archaeal taxonomy. Mutation rates are typically severalfold higher in mesophilic prokaryotes than in thermophilic prokaryotes. To quantitatively test whether accelerated viral mutation leads microbes to lose CRISPR-Cas systems, we developed a stochastic model of virus-CRISPR coevolution. The model competes CRISPR-Cas-positive (CRISPR-Cas+) prokaryotes against CRISPR-Cas-negative (CRISPR-Cas−) prokaryotes, continually weighing the antiviral benefits conferred by CRISPR-Cas immunity against its fitness costs. Tracking this cost-benefit analysis across parameter space reveals viral mutation rate thresholds beyond which CRISPR-Cas cannot provide sufficient immunity and is purged from host populations. These results offer a simple, testable viral diversity hypothesis to explain why mesophilic bacteria disproportionately lack CRISPR-Cas immunity. More generally, fundamental limits on the adaptability of biological sensors (Lamarckian evolution) are predicted.

IMPORTANCE

A remarkable recent discovery in microbiology is that bacteria and archaea possess systems conferring immunological memory and adaptive immunity. Clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated genes (CRISPR-Cas) are genomic sensors that allow prokaryotes to acquire DNA fragments from invading viruses and plasmids. Providing immunological memory, these stored fragments destroy matching DNA in future viral and plasmid invasions. CRISPR-Cas systems also provide adaptive immunity, keeping up with mutating viruses and plasmids by continually acquiring new DNA fragments. Surprisingly, less than 50% of mesophilic bacteria, in contrast to almost 90% of thermophilic bacteria and Archaea, maintain CRISPR-Cas immunity. Using mathematical modeling, we probe this dichotomy, showing how increased viral mutation rates can explain the reduced prevalence of CRISPR-Cas systems in mesophiles. Rapidly mutating viruses outrun CRISPR-Cas immune systems, likely decreasing their prevalence in bacterial populations. Thus, viral adaptability may select against, rather than for, immune adaptability in prokaryotes.

A fundamental observation of comparative genomics is that the distribution of evolution rates across the complete sets of orthologous genes in pairs of related genomes remains virtually unchanged throughout the evolution of life, from bacteria to mammals. The most straightforward explanation for the conservation of this distribution appears to be that the relative evolution rates of all genes remain nearly constant, or in other words, that evolutionary rates of different genes are strongly correlated within each evolving genome. This correlation could be explained by a model that we denoted Universal PaceMaker (UPM) of genome evolution. The UPM model posits that the rate of evolution changes synchronously across genome-wide sets of genes in all evolving lineages. Alternatively, however, the correlation between the evolutionary rates of genes could be a simple consequence of molecular clock (MC). We sought to differentiate between the MC and UPM models by fitting thousands of phylogenetic trees for bacterial and archaeal genes to supertrees that reflect the dominant trend of vertical descent in the evolution of archaea and bacteria and that were constrained according to the two models. The goodness of fit for the UPM model was better than the fit for the MC model, with overwhelming statistical significance, although similarly to the MC, the UPM is strongly overdispersed. Thus, the results of this analysis reveal a universal, genome-wide pacemaker of evolution that could have been in operation throughout the history of life.

Author Summary

A central concept of evolution is Molecular Clock according to which each gene evolves at a characteristic, near constant rate. Numerous studies support the Molecular Clock hypothesis in principle but also show that the clock is indeed very approximate. Genome-wide comparative analysis of phylogenetic trees described here reveals a distinct, more general feature of genome evolution that we called Universal Pacemaker. Under this model, when the rate of evolution changes, the change occurs synchronously in many if not all genes in the evolving genome. In other words, the relative rates of gene evolution remain constant across long evolutionary spans: if a gene is slow relative to the rest of the genes in the given lineage, it is always slow, and if it evolves fast, it is always fast. We show here that the Universal Pacemaker model fits the available data much better than the traditional Molecular Clock model. These findings are compatible with the previously observed accelerations and decelerations of evolution in individual lineages but we show that synchronous, genome-wide change of evolutionary rates is a global feature of genome evolution that appears to pervade the entire history of life.

Orthologous relationships between genes are routinely inferred from bidirectional best hits (BBH) in pairwise genome comparisons. However, to our knowledge, it has never been quantitatively demonstrated that orthologs form BBH. To test this “BBH-orthology conjecture,” we take advantage of the operon organization of bacterial and archaeal genomes and assume that, when two genes in compared genomes are flanked by two BBH show statistically significant sequence similarity to one another, these genes are bona fide orthologs. Under this assumption, we tested whether middle genes in “syntenic orthologous gene triplets” form BBH. We found that this was the case in more than 95% of the syntenic gene triplets in all genome comparisons. A detailed examination of the exceptions to this pattern, including maximum likelihood phylogenetic tree analysis, showed that some of these deviations involved artifacts of genome annotation, whereas very small fractions represented random assignment of the best hit to one of closely related in-paralogs, paralogous displacement in situ, or even less frequent genuine violations of the BBH–orthology conjecture caused by acceleration of evolution in one of the orthologs. We conclude that, at least in prokaryotes, genes for which independent evidence of orthology is available typically form BBH and, conversely, BBH can serve as a strong indication of gene orthology.

There are no known RNA viruses that infect Archaea. Filling this gap in our knowledge of viruses will enhance our understanding of the relationships between RNA viruses from the three domains of cellular life and, in particular, could shed light on the origin of the enormous diversity of RNA viruses infecting eukaryotes. We describe here the identification of novel RNA viral genome segments from high-temperature acidic hot springs in Yellowstone National Park in the United States. These hot springs harbor low-complexity cellular communities dominated by several species of hyperthermophilic Archaea. A viral metagenomics approach was taken to assemble segments of these RNA virus genomes from viral populations isolated directly from hot spring samples. Analysis of these RNA metagenomes demonstrated unique gene content that is not generally related to known RNA viruses of Bacteria and Eukarya. However, genes for RNA-dependent RNA polymerase (RdRp), a hallmark of positive-strand RNA viruses, were identified in two contigs. One of these contigs is approximately 5,600 nucleotides in length and encodes a polyprotein that also contains a region homologous to the capsid protein of nodaviruses, tetraviruses, and birnaviruses. Phylogenetic analyses of the RdRps encoded in these contigs indicate that the putative archaeal viruses form a unique group that is distinct from the RdRps of RNA viruses of Eukarya and Bacteria. Collectively, our findings suggest the existence of novel positive-strand RNA viruses that probably replicate in hyperthermophilic archaeal hosts and are highly divergent from RNA viruses that infect eukaryotes and even more distant from known bacterial RNA viruses. These positive-strand RNA viruses might be direct ancestors of RNA viruses of eukaryotes.

Spliceosomal introns are one of the principal distinctive features of eukaryotes. Nevertheless, different large-scale studies disagree about even the most basic features of their evolution. In order to come up with a more reliable reconstruction of intron evolution, we developed a model that is far more comprehensive than previous ones. This model is rich in parameters, and estimating them accurately is infeasible by straightforward likelihood maximization. Thus, we have developed an expectation-maximization algorithm that allows for efficient maximization. Here, we outline the model and describe the expectation-maximization algorithm in detail. Since the method works with intron presence–absence maps, it is expected to be instrumental for the analysis of the evolution of other binary characters as well.

Abstract

The widespread exchange of genes among prokaryotes, known as horizontal gene transfer (HGT), is often considered to “uproot” the Tree of Life (TOL). Indeed, it is by now fully clear that genes in general possess different evolutionary histories. However, the possibility remains that the TOL concept can be reformulated and remain valid as a statistical central trend in the phylogenetic “Forest of Life” (FOL). This article describes a computational pipeline developed to chart the FOL by comparative analysis of thousands of phylogenetic trees. This analysis reveals a distinct, consistent phylogenetic signal that is particularly strong among the Nearly Universal Trees (NUTs), which correspond to genes represented in all or most of the analyzed organisms. Despite the substantial amount of apparent HGT seen even among the NUTs, these gene transfers appear to be distributed randomly and do not obscure the central tree-like trend.

Archaeal and bacterial ribosomes contain more than 50 proteins, including 34 that are universally conserved in the three domains of cellular life (bacteria, archaea, and eukaryotes). Despite the high sequence conservation, annotation of ribosomal (r-) protein genes is often difficult because of their short lengths and biased sequence composition. We developed an automated computational pipeline for identification of r-protein genes and applied it to 995 completely sequenced bacterial and 87 archaeal genomes available in the RefSeq database. The pipeline employs curated seed alignments of r-proteins to run position-specific scoring matrix (PSSM)-based BLAST searches against six-frame genome translations, mitigating possible gene annotation errors. As a result of this analysis, we performed a census of prokaryotic r-protein complements, enumerated missing and paralogous r-proteins, and analyzed the distributions of ribosomal protein genes among chromosomal partitions. Phyletic patterns of bacterial and archaeal r-protein genes were mapped to phylogenetic trees reconstructed from concatenated alignments of r-proteins to reveal the history of likely multiple independent gains and losses. These alignments, available for download, can be used as search profiles to improve genome annotation of r-proteins and for further comparative genomics studies.

The recently discovered CRISPR-Cas adaptive immune system is present in almost all archaea and many bacteria. It consists of cassettes of CRISPR repeats that incorporate spacers homologous to fragments of viral or plasmid genomes that are employed as guide RNAs in the immune response, along with numerous CRISPR-associated (cas) genes that encode proteins possessing diverse, only partially characterized activities required for the action of the system. Here, we investigate the evolution of the cas genes and show that they evolve under purifying selection that is typically much weaker than the median strength of purifying selection affecting genes in the respective genomes. The exceptions are the cas1 and cas2 genes that typically evolve at levels of purifying selection close to the genomic median. Thus, although these genes are implicated in the acquisition of spacers from alien genomes, they do not appear to be directly involved in an arms race between bacterial and archaeal hosts and infectious agents. These genes might possess functions distinct from and additional to their role in the CRISPR-Cas-mediated immune response. Taken together with evidence of the frequent horizontal transfer of cas genes reported previously and with the wide-spread microscale recombination within these genes detected in this work, these findings reveal the highly dynamic evolution of cas genes. This conclusion is in line with the involvement of CRISPR-Cas in antiviral immunity that is likely to entail a coevolutionary arms race with rapidly evolving viruses. However, we failed to detect evidence of strong positive selection in any of the cas genes.

Multiple constraints variously affect different parts of the genomes of diverse life forms. The selective pressures that shape the evolution of viral, archaeal, bacterial and eukaryotic genomes differ markedly, even among relatively closely related animal and bacterial lineages; by contrast, constraints affecting protein evolution seem to be more universal. The constraints that shape the evolution of genomes and phenomes are complemented by the plasticity and robustness of genome architecture, expression and regulation. Taken together, these findings are starting to reveal complex networks of evolutionary processes that must be integrated to attain a new synthesis of evolutionary biology.

Summary

Posttranslational arginylation mediated by arginyltransferase (ATE1) plays an important role in cardiovascular development, cell motility and regulation of cytoskeleton and metabolic enzymes. This protein modification was discovered decades ago, however, the arginylation reaction and the functioning of ATE1 remained poorly understood due to the lack of good biochemical models. Here we report the development of an in vitro arginylation system, in which ATE1 function and molecular requirements can be tested using purified recombinant ATE1 isoforms supplemented with a controlled number of components. Our results show that arginylation reaction is a self-sufficient, ATP-independent process that can affect different sites in a polypeptide, and that arginyltransferases form different molecular complexes in vivo, associate with components of the translation machinery, and have distinct, partially overlapping subsets of substrates, suggesting that these enzymes play different physiological functions.

When Charles Darwin formulated the central principles of evolutionary biology in the Origin of Species in 1859 and the architects of the Modern Synthesis integrated these principles with population genetics almost a century later, the principal if not the sole objects of evolutionary biology were multicellular eukaryotes, primarily animals and plants. Before the advent of efficient gene sequencing, all attempts to extend evolutionary studies to bacteria have been futile. Sequencing of the rRNA genes in thousands of microbes allowed the construction of the three- domain “ribosomal Tree of Life” that was widely thought to have resolved the evolutionary relationships between the cellular life forms. However, subsequent massive sequencing of numerous, complete microbial genomes revealed novel evolutionary phenomena, the most fundamental of these being: (1) pervasive horizontal gene transfer (HGT), in large part mediated by viruses and plasmids, that shapes the genomes of archaea and bacteria and call for a radical revision (if not abandonment) of the Tree of Life concept, (2) Lamarckian-type inheritance that appears to be critical for antivirus defense and other forms of adaptation in prokaryotes, and (3) evolution of evolvability, i.e., dedicated mechanisms for evolution such as vehicles for HGT and stress-induced mutagenesis systems. In the non-cellular part of the microbial world, phylogenomics and metagenomics of viruses and related selfish genetic elements revealed enormous genetic and molecular diversity and extremely high abundance of viruses that come across as the dominant biological entities on earth. Furthermore, the perennial arms race between viruses and their hosts is one of the defining factors of evolution. Thus, microbial phylogenomics adds new dimensions to the fundamental picture of evolution even as the principle of descent with modification discovered by Darwin and the laws of population genetics remain at the core of evolutionary biology.

Experimental studies on enzyme evolution show that only a small fraction of all possible mutation trajectories are accessible to evolution. However, these experiments deal with individual enzymes and explore a tiny part of the fitness landscape. We report an exhaustive analysis of fitness landscapes constructed with an off-lattice model of protein folding where fitness is equated with robustness to misfolding. This model mimics the essential features of the interactions between amino acids, is consistent with the key paradigms of protein folding and reproduces the universal distribution of evolutionary rates among orthologous proteins. We introduce mean path divergence as a quantitative measure of the degree to which the starting and ending points determine the path of evolution in fitness landscapes. Global measures of landscape roughness are good predictors of path divergence in all studied landscapes: the mean path divergence is greater in smooth landscapes than in rough ones. The model-derived and experimental landscapes are significantly smoother than random landscapes and resemble additive landscapes perturbed with moderate amounts of noise; thus, these landscapes are substantially robust to mutation. The model landscapes show a deficit of suboptimal peaks even compared with noisy additive landscapes with similar overall roughness. We suggest that smoothness and the substantial deficit of peaks in the fitness landscapes of protein evolution are fundamental consequences of the physics of protein folding.

Author Summary

Is evolution deterministic, hence predictable, or stochastic, that is unpredictable? What would happen if one could “replay the tape of evolution”: will the outcomes of evolution be completely different or is evolution so constrained that history will be repeated? Arguably, these questions are among the most intriguing and most difficult in evolutionary biology. In other words, the predictability of evolution depends on the fraction of the trajectories on fitness landscapes that are accessible for evolutionary exploration. Because direct experimental investigation of fitness landscapes is technically challenging, the available studies only explore a minuscule portion of the landscape for individual enzymes. We therefore sought to investigate the topography of fitness landscapes within the framework of a previously developed model of protein folding and evolution where fitness is equated with robustness to misfolding. We show that model-derived and experimental landscapes are significantly smoother than random landscapes and resemble moderately perturbed additive landscapes; thus, these landscapes are substantially robust to mutation. The model landscapes show a deficit of suboptimal peaks even compared with noisy additive landscapes with similar overall roughness. Thus, the smoothness and substantial deficit of peaks in fitness landscapes of protein evolution could be fundamental consequences of the physics of protein folding.

The arms race between cellular life forms and viruses is a major driving force of evolution. A substantial fraction of bacterial and archaeal genomes is dedicated to antivirus defense. We analyzed the distribution of defense genes and typical mobilome components (such as viral and transposon genes) in bacterial and archaeal genomes and demonstrated statistically significant clustering of antivirus defense systems and mobile genes and elements in genomic islands. The defense islands are enriched in putative operons and contain numerous overrepresented gene families. A detailed sequence analysis of the proteins encoded by genes in these families shows that many of them are diverged variants of known defense system components, whereas others show features, such as characteristic operonic organization, that are suggestive of novel defense systems. Thus, genomic islands provide abundant material for the experimental study of bacterial and archaeal antivirus defense. Except for the CRISPR-Cas systems, different classes of defense systems, in particular toxin-antitoxin and restriction-modification systems, show nonrandom clustering in defense islands. It remains unclear to what extent these associations reflect functional cooperation between different defense systems and to what extent the islands are genomic “sinks” that accumulate diverse nonessential genes, particularly those acquired via horizontal gene transfer. The characteristics of defense islands resemble those of mobilome islands. Defense and mobilome genes are nonrandomly associated in islands, suggesting nonadaptive evolution of the islands via a preferential attachment-like mechanism underpinned by the addictive properties of defense systems such as toxins-antitoxins and an important role of horizontal mobility in the evolution of these islands.

Background

The CRISPR-Cas adaptive immunity systems that are present in most Archaea and many Bacteria function by incorporating fragments of alien genomes into specific genomic loci, transcribing the inserts and using the transcripts as guide RNAs to destroy the genome of the cognate virus or plasmid. This RNA interference-like immune response is mediated by numerous, diverse and rapidly evolving Cas (CRISPR-associated) proteins, several of which form the Cascade complex involved in the processing of CRISPR transcripts and cleavage of the target DNA. Comparative analysis of the Cas protein sequences and structures led to the classification of the CRISPR-Cas systems into three Types (I, II and III).

Results

A detailed comparison of the available sequences and structures of Cas proteins revealed several unnoticed homologous relationships. The Repeat-Associated Mysterious Proteins (RAMPs) containing a distinct form of the RNA Recognition Motif (RRM) domain, which are major components of the CRISPR-Cas systems, were classified into three large groups, Cas5, Cas6 and Cas7. Each of these groups includes many previously uncharacterized proteins now shown to adopt the RAMP structure. Evidence is presented that large subunits contained in most of the CRISPR-Cas systems could be homologous to Cas10 proteins which contain a polymerase-like Palm domain and are predicted to be enzymatically active in Type III CRISPR-Cas systems but inactivated in Type I systems. These findings, the fact that the CRISPR polymerases, RAMPs and Cas2 all contain core RRM domains, and distinct gene arrangements in the three types of CRISPR-Cas systems together provide for a simple scenario for origin and evolution of the CRISPR-Cas machinery. Under this scenario, the CRISPR-Cas system originated in thermophilic Archaea and subsequently spread horizontally among prokaryotes.

Conclusions

Because of the extreme diversity of CRISPR-Cas systems, in-depth sequence and structure comparison continue to reveal unexpected homologous relationship among Cas proteins. Unification of Cas protein families previously considered unrelated provides for improvement in the classification of CRISPR-Cas systems and a reconstruction of their evolution.

Open peer review

This article was reviewed by Malcolm White (nominated by Purficacion Lopez-Garcia), Frank Eisenhaber and Igor Zhulin. For the full reviews, see the Reviewers' Comments section.

Severity of seasonal influenza A epidemics is related to the antigenic novelty of the predominant viral strains circulating each year. Support for a strong correlation between epidemic severity and antigenic drift comes from infectious challenge experiments on vaccinated animals and human volunteers, field studies of vaccine efficacy, prospective studies of subjects with laboratory-confirmed prior infections, and analysis of the connection between drift and severity from surveillance data. We show that, given data on the antigenic and sequence novelty of the hemagglutinin protein of clinical isolates of H3N2 virus from a season along with the corresponding data from prior seasons, we can accurately predict the influenza severity for that season. This model therefore provides a framework for making projections of the severity of the upcoming season using assumptions based on viral isolates collected in the current season. Our results based on two independent data sets from the US and Hong Kong suggest that seasonal severity is largely determined by the novelty of the hemagglutinin protein although other factors, including mutations in other influenza genes, co-circulating pathogens and weather conditions, might also play a role. These results should be helpful for the control of seasonal influenza and have implications for improvement of influenza surveillance.

Background

It is common belief that all cellular life forms on earth have a common origin. This view is supported by the universality of the genetic code and the universal conservation of multiple genes, particularly those that encode key components of the translation system. A remarkable recent study claims to provide a formal, homology independent test of the Universal Common Ancestry hypothesis by comparing the ability of a common-ancestry model and a multiple-ancestry model to predict sequences of universally conserved proteins.

Results

We devised a computational experiment on a concatenated alignment of universally conserved proteins which shows that the purported demonstration of the universal common ancestry is a trivial consequence of significant sequence similarity between the analyzed proteins. The nature and origin of this similarity are irrelevant for the prediction of "common ancestry" of by the model-comparison approach. Thus, homology (common origin) of the compared proteins remains an inference from sequence similarity rather than an independent property demonstrated by the likelihood analysis.

Conclusion

A formal demonstration of the Universal Common Ancestry hypothesis has not been achieved and is unlikely to be feasible in principle. Nevertheless, the evidence in support of this hypothesis provided by comparative genomics is overwhelming.

Reviewers

this article was reviewed by William Martin, Ivan Iossifov (nominated by Andrey Rzhetsky) and Arcady Mushegian. For the complete reviews, see the Reviewers' Report section.

Phylogenetic trees of individual genes of prokaryotes (archaea and bacteria) generally have different topologies, largely owing to extensive horizontal gene transfer (HGT), suggesting that the Tree of Life (TOL) should be replaced by a “net of life” as the paradigm of prokaryote evolution. However, trees remain the natural representation of the histories of individual genes given the fundamentally bifurcating process of gene replication. Therefore, although no single tree can fully represent the evolution of prokaryote genomes, the complete picture of evolution will necessarily combine trees and nets. A quantitative measure of the signals of tree and net evolution is derived from an analysis of all quartets of species in all trees of the “Forest of Life” (FOL), which consists of approximately 7,000 phylogenetic trees for prokaryote genes including approximately 100 nearly universal trees (NUTs). Although diverse routes of net-like evolution collectively dominate the FOL, the pattern of tree-like evolution that reflects the consistent topologies of the NUTs is the most prominent coherent trend. We show that the contributions of tree-like and net-like evolutionary processes substantially differ across bacterial and archaeal lineages and between functional classes of genes. Evolutionary simulations indicate that the central tree-like signal cannot be realistically explained by a self-reinforcing pattern of biased HGT.

Evolutionary binary characters are features of species or genes, indicating the absence (value zero) or presence (value one) of some property. Examples include eukaryotic gene architecture (the presence or absence of an intron in a particular locus), gene content, and morphological characters. In many studies, the acquisition of such binary characters is assumed to represent a rare evolutionary event, and consequently, their evolution is analyzed using various flavors of parsimony. However, when gain and loss of the character are not rare enough, a probabilistic analysis becomes essential. Here, we present a comprehensive probabilistic model to describe the evolution of binary characters on a bifurcating phylogenetic tree. A fast software tool, EREM, is provided, using maximum likelihood to estimate the parameters of the model and to reconstruct ancestral states (presence and absence in internal nodes) and events (gain and loss events along branches).

Motivation: Identifying orthologous genes in multiple genomes is a fundamental task in comparative genomics. Construction of intergenomic symmetrical best matches (SymBets) and joining them into clusters is a popular method of ortholog definition, embodied in several software programs. Despite their wide use, the computational complexity of these programs has not been thoroughly examined.

Results: In this work, we show that in the standard approach of iteration through all triangles of SymBets, the memory scales with at least the number of these triangles, O(g3) (where g = number of genomes), and construction time scales with the iteration through each pair, i.e. O(g6). We propose the EdgeSearch algorithm that iterates over edges in the SymBet graph rather than triangles of SymBets, and as a result has a worst-case complexity of only O(g3log g). Several optimizations reduce the run-time even further in realistically sparse graphs. In two real-world datasets of genomes from bacteriophages (POGs) and Mollicutes (MOGs), an implementation of the EdgeSearch algorithm runs about an order of magnitude faster than the original algorithm and scales much better with increasing number of genomes, with only minor differences in the final results, and up to 60 times faster than the popular OrthoMCL program with a 90% overlap between the identified groups of orthologs.

Availability and implementation: C++ source code freely available for download at ftp.ncbi.nih.gov/pub/wolf/COGs/COGsoft/

Contact: dmk@stowers.org

Supplementary information: Supplementary materials are available at Bioinformatics online.