RNA structures facilitate recombination-mediated gene swapping in HIV-1

Many viruses, including retroviruses, undergo frequent recombination, a process which can increase their rate of adaptive evolution. In the case of HIV, recombination has been responsible for the generation of numerous intersubtype recombinant variants with epidemiological importance in the AIDS pandemic. Although it is known that fragments of genetic material do not combine randomly during the generation of recombinant viruses, the mechanisms that lead to preferential recombination at specific sites are not fully understood. Here we reanalyze recent independent data defining (i) the structure of a complete HIV-1 RNA genome and (ii) favorable sites for recombination. We show that in the absence of selection acting on recombinant genomes, regions harboring RNA structures in the NL4-3 model strain are strongly predictive of recombination breakpoints in the HIV-1 env genes of primary isolates. In addition, we found that breakpoints within recombinant HIV-1 genomes sampled from human populations, which have been acted upon extensively by natural selection, also colocalize with RNA structures. Critically, junctions between genes are enriched in structured RNA elements and are also preferred sites for generating functional recombinant forms. These data suggest that RNA structure-mediated recombination allows the virus to exchange intact genes rather than arbitrary subgene fragments, which is likely to increase the overall viability and replication success of the recombinant HIV progeny.     

Detecting balancing selection in genomes: limits and prospects

In spite of the long‐term interest in the process of balancing selection, its frequency in genomes and evolutionary significance remain unclear due to challenges related to its detection. Current statistical approaches based on patterns of variation observed in molecular data suffer from low power and a high incidence of false positives. This raises the question whether balancing selection is rare or is simply difficult to detect. We discuss genetic signatures produced by this mode of selection and review the current approaches used for their identification in genomes. Advantages and disadvantages of the available methods are presented, and areas where improvement is possible are identified. Increased specificity and reduced rate of false positives may be achieved by using a demographic model, applying combinations of tests, appropriate sampling scheme and taking into account intralocus variation in selection pressures. We emphasize novel solutions, recently developed model‐based approaches and good practices that should be implemented in future studies looking for signals of balancing selection. We also draw attention of the readers to the results of recent theoretical studies, which suggest that balancing selection may be ubiquitous but transient, leaving few signatures detectable by existing methods. Testing this new theory may require the development of novel high‐throughput methods extending beyond genomic scans.     

Identification of selection signatures in livestock species

The identification of regions that have undergone selection is one of the principal goals of theoretical and applied evolutionary genetics. Such studies can also provide information about the evolutionary processes involved in shaping genomes, as well as physical and functional information about genes/genomic regions. Domestication followed by breed formation and selection schemes has allowed the formation of very diverse livestock breeds adapted to a wide variety of environments and with special characteristics. The advances in genomics in the last five years have enabled the development of several methods to detect selection signatures and have resulted in the publication of a considerable number of studies involving livestock species. The aims of this review are to describe the principal effects of natural/artificial selection on livestock genomes, to present the main methods used to detect selection signatures and to discuss some recent results in this area. This review should be useful also to research scientists working with wild animals/non-domesticated species and plant biologists working with breeding and evolutionary biology.     

A comparative method for finding and folding RNA secondary structures within protein-coding regions

Existing computational methods for RNA secondary-structure prediction tacitly assume RNA to only encode functional RNA structures. However, experimental studies have revealed that some RNA sequences, e.g. compact viral genomes, can simultaneously encode functional RNA structures as well as proteins, and evidence is accumulating that this phenomenon may also be found in Eukaryotes. We here present the first comparative method, called RNA-DECODER, which explicitly takes the known protein-coding context of an RNA-sequence alignment into account in order to predict evolutionarily conserved secondary-structure elements, which may span both coding and non-coding regions. RNA-DECODER employs a stochastic context-free grammar together with a set of carefully devised phylogenetic substitution-models, which can disentangle and evaluate the different kinds of overlapping evolutionary constraints which arise. We show that RNA-DECODER's parameters can be automatically trained to successfully fold known secondary structures within the HCV genome. We scan the genomes of HCV and polio virus for conserved secondary-structure elements, and analyze performance as a function of available evolutionary information. On known secondary structures, RNA-DECODER shows a sensitivity similar to the programs MFOLD, PFOLD and RNAALIFOLD. When scanning the entire genomes of HCV and polio virus for structure elements, RNA-DECODER's results indicate a markedly higher specificity than MFOLD, PFOLD and RNAALIFOLD.     

Patterns of evolution and host gene mimicry in influenza and other RNA viruses

It is well known that the dinucleotide CpG is under-represented in the genomic DNA of many vertebrates. This is commonly thought to be due to the methylation of cytosine residues in this dinucleotide and the corresponding high rate of deamination of 5-methycytosine, which lowers the frequency of this dinucleotide in DNA. Surprisingly, many single-stranded RNA viruses that replicate in these vertebrate hosts also have a very low presence of CpG dinucleotides in their genomes. Viruses are obligate intracellular parasites and the evolution of a virus is inexorably linked to the nature and fate of its host. One therefore expects that virus and host genomes should have common features. In this work, we compare evolutionary patterns in the genomes of ssRNA viruses and their hosts. In particular, we have analyzed dinucleotide patterns and found that the same patterns are pervasively over- or under-represented in many RNA viruses and their hosts suggesting that many RNA viruses evolve by mimicking some of the features of their host's genes (DNA) and likely also their corresponding mRNAs. When a virus crosses a species barrier into a different host, the pressure to replicate, survive and adapt, leaves a footprint in dinucleotide frequencies. For instance, since human genes seem to be under higher pressure to eliminate CpG dinucleotide motifs than avian genes, this pressure might be reflected in the genomes of human viruses (DNA and RNA viruses) when compared to those of the same viruses replicating in avian hosts. To test this idea we have analyzed the evolution of the influenza virus since 1918. We find that the influenza A virus, which originated from an avian reservoir and has been replicating in humans over many generations, evolves in a direction strongly selected to reduce the frequency of CpG dinucleotides in its genome. Consistent with this observation, we find that the influenza B virus, which has spent much more time in the human population, has adapted to its human host and exhibits an extremely low CpG dinucleotide content. We believe that these observations directly show that the evolution of RNA viral genomes can be shaped by pressures observed in the host genome. As a possible explanation, we suggest that the strong selection pressures acting on these RNA viruses are most likely related to the innate immune response and to nucleotide motifs in the host DNA and RNAs.     

Viruses Roll the Dice: The Stochastic Behavior of Viral Genome Molecules Accelerates Viral Adaptation at the Cell and Tissue Levels

Recent studies on evolutionarily distant viral groups have shown that the number of viral genomes that establish cell infection after cell-to-cell transmission is unexpectedly small (1–20 genomes). This aspect of viral infection appears to be important for the adaptation and survival of viruses. To clarify how the number of viral genomes that establish cell infection is determined, we developed a simulation model of cell infection for tomato mosaic virus (ToMV), a positive-strand RNA virus. The model showed that stochastic processes that govern the replication or degradation of individual genomes result in the infection by a small number of genomes, while a large number of infectious genomes are introduced in the cell. It also predicted two interesting characteristics regarding cell infection patterns: stochastic variation among cells in the number of viral genomes that establish infection and stochastic inequality in the accumulation of their progenies in each cell. Both characteristics were validated experimentally by inoculating tobacco cells with a library of nucleotide sequence–tagged ToMV and analyzing the viral genomes that accumulated in each cell using a high-throughput sequencer. An additional simulation model revealed that these two characteristics enhance selection during tissue infection. The cell infection model also predicted a mechanism that enhances selection at the cellular level: a small difference in the replication abilities of coinfected variants results in a large difference in individual accumulation via the multiple-round formation of the replication complex (i.e., the replication machinery). Importantly, this predicted effect was observed in vivo. The cell infection model was robust to changes in the parameter values, suggesting that other viruses could adopt similar adaptation mechanisms. Taken together, these data reveal a comprehensive picture of viral infection processes including replication, cell-to-cell transmission, and evolution, which are based on the stochastic behavior of the viral genome molecules in each cell.     

Bootstrap, Bayesian probability and maximum likelihood mapping: exploring new tools for comparative genome analyses

Background  

Horizontal gene transfer (HGT) played an important role in shaping microbial genomes. In addition to genes under sporadic selection, HGT also affects housekeeping genes and those involved in information processing, even ribosomal RNA encoding genes. Here we describe tools that provide an assessment and graphic illustration of the mosaic nature of microbial genomes.     

Selection for unequal densities of sigma70 promoter-like signals in different regions of large bacterial genomes

The evolutionary processes operating in the DNA regions that participate in the regulation of gene expression are poorly understood. In Escherichia coli, we have established a sequence pattern that distinguishes regulatory from nonregulatory regions. The density of promoter-like sequences, that could be recognizable by RNA polymerase and may function as potential promoters, is high within regulatory regions, in contrast to coding regions and regions located between convergently transcribed genes. Moreover, functional promoter sites identified experimentally are often found in the subregions of highest density of promoter-like signals, even when individual sites with higher binding affinity for RNA polymerase exist elsewhere within the regulatory region. In order to see the generality of this pattern, we have analyzed 43 additional genomes belonging to most established bacterial phyla. Differential densities between regulatory and nonregulatory regions are detectable in most of the analyzed genomes, with the exception of those that have evolved toward extreme genome reduction. Thus, presence of this pattern follows that of genes and other genomic features that require weak selection to be effective in order to persist. On this basis, we suggest that the loss of differential densities in the reduced genomes of host-restricted pathogens and symbionts is an outcome of the process of genome degradation resulting from the decreased efficiency of purifying selection in highly structured small populations. This implies that the differential distribution of promoter-like signals between regulatory and nonregulatory regions detected in large bacterial genomes confers a significant, although small, fitness advantage. This study paves the way for further identification of the specific types of selective constraints that affect the organization of regulatory regions and the overall distribution of promoter-like signals through more detailed comparative analyses among closely related bacterial genomes.     

Trade-offs in resource allocation in the intracellular life-cycle of hepatitis C virus

Positive sense single-stranded RNA viruses undergo three mutually exclusive processes to replicate within a cell. These are translation to produce proteins, replication to produce RNA viral genomes, and packaging to form virions. The allocation of newly synthesised viral genomes to these processes, which can be regarded as life-history traits, may be subject to natural selection for efficient reproduction. Here, we develop a mathematical model of the process of intracellular viral replication to study alternative strategies for the allocation and reallocation of viral genomes to these processes. We explore four cases of the model: (1) Free Movement, in which viral genomes can freely be allocated and reallocated among translation, replication and packaging; (2) Unidirectional Reallocation, in which allocation occurs freely but reallocation can only proceed from translation to replication to packaging; (3) Conveyor Belt, in which viral genomes are first allocated to translation, then passed on to replication and finally to packaging; and (4) Permanent Allocation in which new genomes are allocated to the three processes but not reallocated between them. We apply this model to hepatitis C virus and study changes in the production of virus as the rates of allocation and reallocation are varied. We find that high viral production occurs when allocation and reallocation of the genome are weighted towards the translation and replication processes. The replication process in particular is favoured. The most productive strategy is a form of the Free Movement model in which genomes are allocated entirely to the replication-translation cycle while allowing some genomes to be packaged through reallocation.     

Synonymous Mutations Reduce Genome Compactness in Icosahedral ssRNA Viruses

Recent studies have shown that single-stranded (ss) viral RNAs fold into more compact structures than random RNA sequences with similar chemical composition and identical length. Based on this comparison, it has been suggested that wild-type viral RNA may have evolved to be atypically compact so as to aid its encapsidation and assist the viral assembly process. To further explore the compactness selection hypothesis, we systematically compare the predicted sizes of >100 wild-type viral sequences with those of their mutants, which are evolved in silico and subject to a number of known evolutionary constraints. In particular, we enforce mutation synonynimity, preserve the codon-bias, and leave untranslated regions intact. It is found that progressive accumulation of these restricted mutations still suffices to completely erase the characteristic compactness imprint of the viral RNA genomes, making them in this respect physically indistinguishable from randomly shuffled RNAs. This shows that maintaining the physical compactness of the genome is indeed a primary factor among ssRNA viruses’ evolutionary constraints, contributing also to the evidence that synonymous mutations in viral ssRNA genomes are not strictly neutral.     

On the nature of spontaneous RNA synthesis by Q beta replicase.

Numerous RNA species of different length and nucleotide sequence grow spontaneously in vitro in Q beta replicase reactions where no RNA templates are added deliberately. Here, we show that this spontaneous RNA synthesis by Q beta replicase is template directed. The immediate source of template RNA can be the laboratory air, but there are ways to eliminate, or at least substantially reduce, the harmful effects of spontaneous synthesis. Solitary RNA molecules were detected in a thin layer of agarose gel containing Q beta replicase, where they grew to form colonies that became visible upon staining with ethidium bromide. This result provides a powerful tool for RNA cloning and selection in vitro. We also show that replicating RNAs similar to those growing spontaneously are incorporated into Q beta phage particles and can propagate in vivo for a number of phage generations. These RNAs are the smallest known molecular parasites, and in many aspects they resemble both the defective interfering genomes of animal and plant viruses and plant virus satellite RNAs.     

Prediction of structured non-coding RNAs in the genomes of the nematodes Caenorhabditis elegans and Caenorhabditis briggsae

We present a survey for non-coding RNAs and other structured RNA motifs in the genomes of Caenorhabditis elegans and Caenorhabditis briggsae using the RNAz program. This approach explicitly evaluates comparative sequence information to detect stabilizing selection acting on RNA secondary structure. We detect 3,672 structured RNA motifs, of which only 678 are known non-translated RNAs (ncRNAs) or clear homologs of known C. elegans ncRNAs. Most of these signals are located in introns or at a distance from known protein-coding genes. With an estimated false positive rate of about 50% and a sensitivity on the order of 50%, we estimate that the nematode genomes contain between 3,000 and 4,000 RNAs with evolutionary conserved secondary structures. Only a small fraction of these belongs to the known RNA classes, including tRNAs, snoRNAs, snRNAs, or microRNAs. A relatively small class of ncRNA candidates is associated with previously observed RNA-specific upstream elements.     

Fossil rhabdoviral sequences integrated into arthropod genomes: ontogeny, evolution, and potential functionality

Retroelements represent a considerable fraction of many eukaryotic genomes and are considered major drives for adaptive genetic innovations. Recent discoveries showed that despite not normally using DNA intermediates like retroviruses do, Mononegaviruses (i.e., viruses with nonsegmented, negative-sense RNA genomes) can integrate gene fragments into the genomes of their hosts. This was shown for Bornaviridae and Filoviridae, the sequences of which have been found integrated into the germ line cells of many vertebrate hosts. Here, we show that Rhabdoviridae sequences, the major Mononegavirales family, have integrated only into the genomes of arthropod species. We identified 185 integrated rhabdoviral elements (IREs) coding for nucleoproteins, glycoproteins, or RNA-dependent RNA polymerases; they were mostly found in the genomes of the mosquito Aedes aegypti and the blacklegged tick Ixodes scapularis. Phylogenetic analyses showed that most IREs in A. aegypti derived from multiple independent integration events. Since RNA viruses are submitted to much higher substitution rates as compared with their hosts, IREs thus represent fossil traces of the diversity of extinct Rhabdoviruses. Furthermore, analyses of orthologous IREs in A. aegypti field mosquitoes sampled worldwide identified an integrated polymerase IRE fragment that appeared under purifying selection within several million years, which supports a functional role in the host's biology. These results show that A. aegypti was subjected to repeated Rhabdovirus infectious episodes during its evolution history, which led to the accumulation of many integrated sequences. They also suggest that like retroviruses, integrated rhabdoviral sequences may participate actively in the evolution of their hosts.     

Replication of lengthened Moloney murine leukemia virus genomes is impaired at multiple stages

It has been assumed that RNA packaging constraints limit the size of retroviral genomes. This notion of a retroviral "headful" was tested by examining the ability of Moloney murine leukemia virus genomes lengthened by 4, 8, or 11 kb to participate in a single replication cycle. Overall, replication of these lengthened genomes was 5- to 10-fold less efficient than that of native-length genomes. When RNA expression and virion formation, RNA packaging, and early stages of replication were compared, long genomes were found to complete each step less efficiently than did normal-length genomes. To test whether short RNAs might facilitate the packaging of lengthy RNAs by heterodimerization, some experiments involved coexpression of a short packageable RNA. However, enhancement of neither long vector RNA packaging nor long vector DNA synthesis was observed in the presence of the short RNA. Most of the proviruses templated by 12 and 16 kb vectors appeared to be full length. Most products of a 19. 2-kb vector contained deletions, but some integrated proviruses were around twice the native genome length. These results demonstrate that lengthy retroviral genomes can be packaged and that genome length is not strictly limited at any individual replication step. These observations also suggest that the lengthy read-through RNAs postulated to be intermediates in retroviral transduction can be packaged directly without further processing.