Illustrations
Figure 1
Types of products generated by recombination. Homologous recombinants have the same genomic organization as the parental molecules from which they are derived. Non-homologous recombinants, on the other hand, contain a deletion (recombinants shorter than the parental molecules) or a duplication (recombinants longer than the parental molecules) of one of the regions of the parental molecules. Some authors speak rather of “precise” (homologous) and “imprecise” (non-homologous) recombinants.
Figure 1
Figure 2
Recombination mechanisms involving the activity of the viral polymerase. A. The so-called “copy choice” or “template switching” mechanism consists of a change in templates by the polymerase during the extension of an RNA strand. It is undoubtedly favored by the formation of a heteroduplex between two viral genomes, for example at the level of inverted repeated sequences forming the stem-loop type structures. B. The primer alignment and extension mechanism (PAAE) is based on the presence of viral RNA fragments whose truncated 3’ end hybridizes to a complementary sequence carried by a complete genome of opposite polarity. The 3’ end of a fragment then serves as a primer and is extended by the viral polymerase as a recombinant genome.
Figure 2
Figure 3
Experimental models based on the Qβ phage genome. A. RQ RNAs are short RNAs, with specific sequences present at their 5’ and 3’ ends, that are replicated by the Qβ phage replicase. The two recombination partners are synthetic transcripts each of which has only one end of an RQ RNA. They themselves cannot be replicated. Each partner also carries an exogenous extension. B. In the initial model, the two partners are preincubated together in the absence of any other biological molecules. The mixture is then deposited on an agarose gel which contains the viral replicase capable of replicating the recombinant molecules having the two ends of the RQ RNA. The ions and NTPs necessary for the replicase activity are delivered by diffusion from a nylon membrane applied against the gel. After 60 min, the RNA molecules replicated in the gel are transferred to a new membrane and revealed by radiolabeled probes. A second series of experiments adds two additional steps. At the end of the pre-incubation phase, the mixture of the two partners is heated in order to destabilize any intermolecular bonds that may have formed, and the 3’-OH ends are oxidized in order to block their extension by the polymerase. These steps aim to prevent any replicative recombination once the RNA molecules are placed in the presence of the replicase in the gel. C. Transesterification mechanism that was proposed from the results of the first series of experiments. Most of the recombinant molecules observed contain the integral 5’ partner (including its exogenous extension) and the 3’ partner deprived of its 5’ region. D. Non-enzymatic mechanism of cleavage/ligation catalized by Mg2+ ions proposed after the second series of experiments.
Figure 3
Figure 4
In vitro model used to highlight the establishments of covalent bonds between RNAs after their cleavage in the presence of Mg2+ ions. A. The model relies on two synthetic oligoribonucleotides (O1 and O2), each of which contains two short sequences (in blue and red) that are complementary to sequences harboured by the other. B. One molecule of O2 is able to form a complex with two copies of O1. In presence of Mg2+ ions, O1-O1 dimeric products accumulate. They result from the ligation of two copies of O1 after the cleavage of the polyA stretch of one of them.
Figure 4
Figure 5
Schematic representation of synthetic transcripts from eukaryotic viruses that have demonstrated NRR. A. The first model based on poliovirus consists of two defective transcripts: the 5’ partner consists of the 5’ UTR region of the genome while the 3’ partner contains the entire genome except for the 5’ part of the 5’UTR region. In the second poliovirus-based model, the 5’ partner spans the whole genome except its 3’ part, which includes sequences that encode domains that are crucial for the virus polymerase 3D; the 3’ partner contains a part of the polymerase-encoding sequence and the 3’UTR. B. In the BVDV-based model, none of the two partners contains the whole sequence that encodes the virus polymerase NS5B.
Figure 5
Figure 6
Effects of the nature of the termini on the frequency of recombination and the location of recombination sites in a model based on poliovirus and one based on BVDV. The presence of a hydroxyl (OH) or phosphate (PO4 3- ) group at the end of the partners influences the number of recombinants generated, but also the relative abundance of each type of recombinant.
Figure 6
Figure 7
Schematic representation of putative mechanisms of NRR observed during co-transfection experiments with pairs of defective poliovirus or BVDV genomes. A. Mechanism of spontaneous transesterification without enzymatic catalysis. B. Replicative mechanisms (by copy choice or PAAE) catalyzed by an unidentified cellular polymerase. C. A multiphasic mechanism involving a phase of reverse transcription of the recombination partners into double-stranded DNA, a phase of DNA/DNA recombination and a phase of transcription of the recombinant DNA into RNA. The cellular enzymes which would be responsible for the different phases have not been identified. D. Breakage/ligation mechanism in which fragmentation of the recombination partners would produce fragments reassembled by an unidentified ligase; the breakage step could generate different types of ends: 3’-OH, 3’-phosphate (3’-P), 2’, 3’ cyclic phosphate (2’, 3’-cP), 5’-OH and 5’-phosphate (5’-P).
Figure 7
Auteurs
1 Institut Pasteur, Populations virales & pathogenèse, CNRS UMR 3569, Paris, France
2 Université de Paris, Populations virales et pathogenèse, Paris, France
Une affaire non résolue : la recombinaison non réplicative chez les virus à ARN de polarité positive. La recombinaison génétique est un moteur important de l’évolution de certaines espèces de virus à ARN de polarité positive. Les évènements de recombinaison se produisent lorsqu’au moins deux virus infectent simultanément la même cellule ; ils donnent naissance à des génomes originaux constitués de séquences génétiques provenant des génomes parentaux. Le mécanisme principal de la recombinaison implique la polymérase virale qui génère une copie chimérique lorsqu’elle change de matrice lors de la synthèse d’un génome viral. Différents modèles expérimentaux suggèrent l’existence d’évènements de recombinaison indépendants de l’activité de la polymérase virale. L’origine de ces évènements et leur fréquence restent à ce jour indéterminées. De même, on ne sait pas si la recombinaison non réplicative engendre des produits qui diffèrent des recombinants générés par la polymérase virale. Si c’est le cas, la recombinaison non réplicative pourrait jouer un rôle spécifique dans l’évolution des virus à ARN de polarité positive. Enfin, les rares données disponibles suggèrent que la recombinaison non réplicative ne nécessiterait pas de séquence d’origine spécifiquement virale. Il est ainsi possible que la recombinaison non réplicative observée dans des modèles d’étude basés sur des ARN viraux nous révèle en fait un mécanisme plus général non spécifique des génomes viraux.