EUVIMEDObjevte EUVIMED
KommentarOpen access · CC BY 4.0

Bridging quasispecies theory and social evolution models for sociovirology insights: a commentary on Leeks et al. 2023

Santiago F Elena iD

Journal of Evolutionary Biology · 2023

DOI: 10.1111/jeb.14229 ↗

At a glance

Kommentar in Journal of Evolutionary Biology, 2023. Written by Santiago F Elena. Institution: Instituto de Biología Integrativa de Sistemas (I2SysBio), CSIC-Universitat de València, Paterna (Valencia), Spain. Major MEDLINE topics: Social Evolution, Onions. Cited 3 times so far (Crossref); 31 references. The full text is freely available (open access, CC BY 4.0). Funded by: Conselleria de Innovación, Universidades, Ciencia y Sociedad Digital, Generalitat Valenciana, Ministerio de Ciencia e Innovación.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Due to their high mutation and recombination rates and large population sizes, RNA viruses generate highly genetically complex populations during the infection of single hosts. Within these genetically heterogeneous populations, individual genomes engage in diverse interactions, ranging from competition for cellular resources, sharing of viral gene products and functional trans‐complementation (public goods game), interference of defective genomes with wild‐type (WT) virus, phenotypic mixing and hiding of mutant genomes inside WT virions (cheating game), and even the collective transmission of particles. As reviewed by Leeks et al. (2023), all these diverse and pervasive interactions can be seen as “social traits”, and advocate for viruses as exceptional, yet rarely considered, tools for the students of social evolution. Some of these interactions have been mechanistically studied by virologists for many years without paying particular attention to their evolutionary implications. However, this disconnection between molecular virology and evolution is fortunately changing. Several pioneering attempts to bring evolutionary biology into this field had experimentally tested whether kin selection plays a role during nonrandomly assorted transmission inocula in shaping virulence (Miralles et al., 1997) and explored the conflicts arising within cells when being coinfected with multiple viruses (Turner & Chao, 1999, 2003). In their remarkable study, Turner and Chao (1999) showed that bacteriophage ϕ6 populations evolved at high multiplicity of infection (MOI; i.e., many viruses coinfecting the same cell) gained an added advantage during coinfection with the WT genotype, indicating that these viruses evolved a defection strategy for intracellular competition. By contrast, ϕ6 populations evolved at low MOI never showed such behaviour. However, when the high MOI‐evolved viruses were grown by themselves, they showed lowered fitness, as expected for a population composed only of defectors. These observations matched the predictions of the Prisoner's Dilemma game (Turner & Chao, 1999). The evolution of ϕ6 at high MOI resulting in the generation of cheaters that engaged in a Prisoner's Dilemma can be attributed to the absence of clonal structure at high MOI. The absence of clonal structure fosters the evolution of selfishness. By contrast, at low MOI, competition did not occur or involved closely related kin. Later on, these observations were extended to competition between two different viral species coinfecting the same plant cells and using each other's proteins that represented public goods; in particular viral suppressors of RNA silencing that block innate immune responses (Chen et al., 2012; Martín & Elena, 2009). A particular mention in the context of this commentary is the so‐called defective interfering particles (DIP). Since their description by Von Magnus and Gard (1947) in influenza‐infected eggs, DIPs, generated by erroneous replication of the WT virus, have attracted attention from both experimentalists and theoreticians (reviewed in Vignuzzi & López, 2019). DIPs behave as cheater hyperparasites stealing viral proteins and negatively impacting viral replication and transmission, even driving the viral population to extinction, which opens the possibility of using them as antiviral therapeutics (Notton et al., 2014). Indeed, game theoretical approaches have been used to explain the evolution of DIPs from competing viral strains, switching from a Prisoner's Dilemma strategy to a Snowdrift one in which lineages of ultradefectors evolved at the cost of lineages of ultracooperators (Chao & Elena, 2017). In recent years, researchers have made noteworthy observations about cooperative traits in viruses: the collective escape from interferon‐based innate immunity (Domingo‐Calap et al., 2019) and the advantages that collective transmission units offer to viruses (Andreu‐Moreno & Sanjuán, 2018; Sagredo‐Otero & Sanjuán, 2019). These findings have sparked an interest in studying viral populations from a social perspective, leading to the coining of the term “Sociovirology” (Díaz‐Muñoz et al., 2017; Leeks et al., 2021, in press; Sanjuán, 2017, 2021). In their review, Leeks et al. (2023) effectively highlight the extensive and varied nature of social interactions among co‐infecting viruses. These interactions occur within the same cells, hosts, or even host populations. This spectrum of interactions ranges from cooperation to conflict and plays a role across multiple phases of the viral lifecycle. Consequently, these interactions have a profound impact on the behaviour of viral populations, shaping their evolution, influencing their paths, and even influencing the clinical course of infections. Despite their significance, as underscored by Leeks et al., (2023) the field of sociovirology remains relatively limited within the broader domains of evolutionary biology and virology. To address this gap, Leeks et al. (2023) advocate for addressing three key aspects: (1) provide a more comprehensive characterization of the various mechanisms through which viruses engage in social interactions and an understanding of how these interactions contribute to their evolution and adaptability. (2) Identification of pertinent questions within the realm of sociovirology that challenge existing theories of social evolution. And (3) provision of concrete instances, data resources, and thought‐provoking inquiries that facilitate the exploration of the natural history of viruses engaged in social behaviours. Leeks et al. (2023) identify the development of new theories for sociovirology both as an opportunity and as a challenge, especially due to the disconnection between the theoretical frameworks commonly used by virologists and evolutionary biologists. Eigen et al. (1989) quasispecies theory has permeated into the virological literature as a convenient framework to study the dynamical behaviour of the mutant swarm of viral genotypes generated within infected individuals (Andino & Domingo, 2015; Bull et al., 2005; Domingo et al., 2021). Quasispecies theory provides a solid mathematical framework to addresses the genetic diversity present within populations of organisms that mutate rapidly, such as RNA viruses. Quasispecies show some interesting emergent properties, e.g., lethal mutagenesis and error catastrophe, mutational robustness, and high evolvability (Andino & Domingo, 2015; Solé et al., 2021). Interestingly, quasispecies theory and classical population genetics are essentially equivalent, although they differ in their assumptions about mutation rates and mutational effects (Wilke, 2005). The classic approach to the mutation‐selection balance assumes mutation rates are small and new mutations are of minor magnitude (Crow & Kimura, 1970), whereas the quasispecies theory relaxes these assumptions (Wilke, 2005). By doing so, the quasispecies theory posits that the unit of selection is not the individual genome, but rather the constantly produced swarms of mutants that emerge due to high mutation rates (Wilke, 2005). In this context, quasispecies theory could be likened to kin selection theories. Another strong implication is that while the average fitness for an asexual population at the mutation‐selection balance only depends on the genomic mutation rate according to classic models (Kimura & Maruyama, 1966), in the quasispecies context average fitness also depends on the actual distribution of mutational effects (i.e., the fraction of lethal, deleterious, neutral, and beneficial mutations), that is, on the topography of the fitness landscape (Wilke & Adami, 2003). Hamilton's rule, a key concept in evolutionary biology, delves into the circumstances that allow altruistic behaviour (assisting others at a personal cost) to develop within a population (Hamilton, 1964). This rule employs relatedness (r), benefit to the recipient (b), and cost (c) to the actor as variables to predict whether acts of altruism will be favoured by natural selection. In mathematical terms, Hamilton's rule simply states than br – c > 0. Although a direct correlation between quasispecies mutation‐selection replicator models and Hamilton's rule has yet to be established, given the nature of the quasispecies concept itself, a reasonable hypothesis is that the rule should naturally emerge from the quasispecies theory. My goal here is to explore this potential relationship and showing whether quasispecies theory may provide a convenient framework to study sociovirology. Now A contains the growth rates of different viral genotypes, ai. Q denotes the mutational coupling (adjacency) matrix of the viral genotypic network that represents the connectivity among genotypes by mutation (Eigen et al., 1989). Elements in the Q matrix are given by qij=μHij1−μl−Hij⁠, where μ is the per site mutation rate, Hij the Hamming distance between strains i and j and l the genome length. In other words, Hij provides a direct measure of the degree of genetic relatedness between pairs of strains and therefore, Q provides a convenient algebraic representation of these relationships among all n genotypes. This simple equation illustrates that the steady‐state occurrence of cooperators is evidently influenced by both the coefficients within the payoff matrix and the mutation rate that affects the cooperative trait. This expression is very similar to the previous one but now considers both factors promoting cooperation (i.e., the benefit for cooperating B and the penalty when two defectors are facing each other P) weighted by genetic similarity vs both factors promoting defection (i.e., the cost of cooperating S and the temptation to cheat T). To illustrate the dependency of the equilibrium frequency of the cooperator, xC*⁠, on the expected mutational load per genome u and the specific interaction strategy between strains (structure of the A payoff matrix), we run a simple numerical simulation in which u varied between zero and one and for four different structures of the payoff matrix. The results of this simulation are shown in Figure 1. First, as expected for a quasispecies model, a critical u value, uc, exists at which a phase transition (Solé et al., 2021) occurs and the WT strain, in this example the cooperator, is displaced by the defector mutant genotype. Most interestingly for the purpose of this commentary, the actual value of the critical mutation rate at which phase transition occurs, depends on the specific structure of the payoff matrix: interaction strategies in which cooperation is penalized, such as the Prisoner's Dilemma or the Snowdrift, move the critical mutation rate towards lower values (e.g., from uc ~ 0.5 for the standard quasispecies to uc ~ 0.46 for Prisoner's Dilemma), with the case of Snowdrift, in which ultradefectors are favoured (Chao & Elena, 2017), showing the phase transition at the lowest critical mutation rate (uc ~ 0.32). On the other hand, a Mutual Cooperation strategy results in perdurance of the cooperator strain at higher mutation rates (uc ~ 0.75). Worden and Levin (2007) and Chao & Elena (2017) showed that whenever the quantitative values of the different components of the payoff matrix are affected by mutations, the populations can switch among strategies. In the case studied by Worden and Levin (2007), populations switched from Prisoners Dilemma to stable cooperation while. As already mentioned, in the case of viral DIPs studied by Chao & Elena (2017), the population diversified into ultradefectors and ultracooperators. Dependence of critical mutation rate on the interaction strategies used by competing quasispecies. The relationship between the parameters in the payoff matrix are: B = S = T = P for the standard quasispecies model; T > B > P > S for the Prisoner's Dilemma game; S > P > T > B for the Snowdrift game; and B > T > S > P for the Mutual Cooperation game. In this example, the smallest value was always set to 1 and doubled for each of the following parameters in the ordered lists: e.g., for Prisoner's dilemma T = 8, B = 4, P = 2, and S = 1, while for Mutual Cooperation B = 8, T = 4, S = 2, and P = 1. Additional parameters fixed for all simulations: ε = 0.1, xC(0) = 10 and xD(0) = 1. In this commentary, my aim is not to delve into an exhaustive and advanced mathematical analysis of the intricate connection between quasispecies models and classical models in the context of the evolution of social traits. Instead, I take a deliberately simplified approach to illustrate a rather intuitive point: the fundamental principle of social evolution models, often referred to as Hamilton's rule, can be directly extrapolated from the quasispecies replicator equations. This derivation hinges on two assumptions: (i) the matrix of replication coefficients can be likened to a payoff matrix and (ii) genetic relatedness among competing strategies solely results from mutation in a well‐mixed population and not from transmission bottlenecks, spatial aggregation or other physical features of the environment, as commonly assumed by other models (Leeks et al., 2021). It is worth noting that my intention is not to comprehensively explore this concept for A matrices of all dimensions (n × n), as this would necessitate a more nuanced examination carried out by qualified theoreticians. I have just focused into a toy model with a 2 × 2 A matrix and a symmetric Q matrix. The A matrix can accommodate highly diverse situations by simply choosing the combination of coefficients of interests that better describe the many social viral interactions described by Leeks et al. (2023). In summary, my point is that the framework of quasispecies replicator equations can serve as a convenient initial stepping stone for modelling social interactions among viral entities. Consequently, this foundational approach may contribute to the development of a theoretical framework underpinning the field of sociovirology, where viral dynamics are studied within the broader context of social evolution. Santiago F. Elena: Conceptualization (lead); formal analysis (lead); funding acquisition (lead); investigation (lead); methodology (lead); software (lead); visualization (lead); writing – original draft (lead); writing – review and editing (lead). I thank Lin Chao, Xavier Jarque, J. Tomás Lázaro, Rafael Sanjuán, and Josep Sardanyés for extensive and enlightening discussions about the origin and evolutionary consequences of cooperation and defection in the virosphere. I also thank Asher Leeks and Samuel L. Díaz‐Muñoz for inviting me to participate in the mind‐blowing meeting Social Viruses funded by ESEB. This work was supported by grant PID2022‐136912NB‐I00 funded by MCIN/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR and by Generalitat Valenciana grant CIPROM/2022/59. The author declares no conflict of interests. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/jeb.14229. The R script used for the numerical simulations is available at https://git.csic.es/sfelenalab/quasispecies‐playing‐games.

Subjects

MeSH terms (MEDLINE)

Social EvolutionOnionsQuasispeciesCooperative BehaviorSelection, GeneticBiological EvolutionGame Theory

Author keywords

Hamilton's ruleRNA virusesevolutionary biologygame theoryreplicator equationssocial interactions

Authors and institutions

  • Instituto de Biología Integrativa de Sistemas (I2SysBio), CSIC-Universitat de València, Paterna (Valencia), Spain.
  • The Santa Fe Institute, Santa Fe, New Mexico, USA.

ORCID profiles: Santiago F Elena

Funding

  • Conselleria de Innovación, Universidades, Ciencia y Sociedad Digital, Generalitat Valenciana
  • Ministerio de Ciencia e Innovación

Cite

APA

Elena, S. F. (2023). Bridging quasispecies theory and social evolution models for sociovirology insights: a commentary on Leeks et al. 2023. Journal of Evolutionary Biology. https://doi.org/10.1111/jeb.14229

Vancouver

Elena SF. Bridging quasispecies theory and social evolution models for sociovirology insights: a commentary on Leeks et al. 2023. Journal of Evolutionary Biology. 2023. doi:10.1111/jeb.14229

BibTeX
@article{Elena2023,
  author = {Elena, Santiago F},
  title = {Bridging quasispecies theory and social evolution models for sociovirology insights: a commentary on Leeks et al. 2023},
  journal = {Journal of Evolutionary Biology},
  year = {2023},
  doi = {10.1111/jeb.14229}
}
Start for free →