Meet the bloggers:
Mitchel Daniel is a postdoctoral fellow at Florida State University. He is an evolutionary and behavioral ecologist, and is especially interested in sexual selection, kin selection, and kin recognition. Follow Mitchel’s work @MitchelJDaniel.
Walid Mawass is an evolutionary geneticist, currently a PhD candidate at the Université du Québec à Trois-Rivières, studying the evolution of life histories in French-Canadian historical populations using quantitative genetic and genomic approaches. Walid is interested in evolutionary genetics theory in general, with a current focus on contemporary evolution in natural populations and the role of interactions between environment and genetics on evolutionary trajectories. Follow Walid’s work @MawassWalid.
Sabrina Heiser is a PhD Candidate in Dr. Charles D. Amsler’s lab at the University of Alabama at Birmingham. Her research focuses on the factors driving the geographic distribution of chemical defenses in a red seaweed. For her sample and data collection, she gets to go and SCUBA dive in Antarctica. She received her B.Sc. in Marine Biology from Plymouth University (UK) and is originally from Germany. Follow her adventures on her website.
Mitchel
Kimberly Hughes’ President’s Introduction kicked off the symposium with a brief history of the American Genetics Association (AGA) – formerly the American Breeder’s Association, which first met to discuss the then new science of genetics for practical breeding of animals and plants. The AGA’s Journal of Heredity is the oldest journal devoted to research on the synthesis of Mendelian inheritance and Darwin’s theory of evolution. This history noteworthy, in part because it parallels how the topic of this symposium – genes as environments – has developed. The study of genes as environments was initially motivated by agricultural breeding problems, such as artificial selection often not producing the expected results, and crop strains that do well in mixed stands often doing poorly in pure stands. New theory was needed to explain these phenomena, and so models of genetic effects at the individual level were extended to explicitly consider social environment, including indirect genetic effects (IGEs). As a relative newcomer to the IGE literature, this talk piqued my excitement for the rest of the symposium. IGEs can explain why the magnitude or even the direction of the response to selection is different than what you would expect from classical quantitative genetics. The IGE framework has also been argued to provide a unifying framework for many different problems in social evolution at multiple levels of biological organization. It will be exciting to see how this framework is applied to disparate study organisms and diverse research questions in the coming talks!
Allen Moore’s foundational contributions to the IGE literature benefitted from a series of collaborations and exchanges with others studying social evolution. A reviewer of his earlier work on the genetics of social dominance in cockroaches pointed out that “dominance doesn’t exist in individuals; it exists in the space between individuals.” So, how can we talk about its genetics (and hence, its evolution)? Following the thinking of classical quantitative genetics, different components contribute to a phenotype, and these components can be subdivided, giving us increasingly complex models with which to represent and predict social selection. Despite their complexity, these models are valuable because they indicate the things we need to measure to understand how social traits evolve. The key is to understand the genotype-phenotype map, which, for social traits, requires incorporating the genotypic effects of social partners. Allen stressed repeatedly that this is more easily said than done. To paraphrase Allen very loosely, what even is a phenotype? Identifying and measuring a phenotype that is biologically meaningful is a challenge that many of us are familiar with.
Amelie Baud expanded on a point raised by Allen Moore: IGEs likely have considerable human health implications, and medical care could be enhanced by accounting for social environment. But, IGEs in humans are difficult to study and fraught to discuss. Instead, Amelie is using lab rodents as a model, and thinks of IGEs as a toolfor examining social effects on health-related traits. She’s found that the genotype of an individual’s cage mate affects many disease-related traits, including some not expected to be altered by social interactions. She also showed that the genome-wide association study of IGEs (igeGWAS) can identify loci not detected by GWAS of direct genetic effects. But it can become harder to detect IGEs when the two are strongly correlated (such as when individuals tend to interact with their kin).
Piter Bijma discussed the importance of IGEs for heritable variation in the prevalence of infectious diseases. Classical quantitative genetics says that the heritability of susceptibility to a disease approaches 0 as the prevalence of the disease in the population approaches 0. This suggests that selection should be ineffective at eradicating genetic susceptibility to disease. When accounting for IGEs, however, this dynamic changes; heritable variation increasesas prevalence nears 0. This implies that most of the heritable variation in the prevalence of endemic diseases is probably due to IGEs. Heritability of disease prevalence is typically estimated by solely modelling direct genetic effects (DGEs); the true heritability is probably much higher!
Julia Saltz – Complex social environments are difficult to study, so it is common to average across genetic and phenotypic heterogeneity when estimating indirect genetic effects. But, this averaging may not work well if there is variation in the strength or type of interactions among individuals. Individuals might interact more strongly with a neighbor than a random individual, say, and there are clear differences between how an individual interacts with its mother versus its sibling. Julia’s experiments using fruit flies show that heterogeneity indeed matters! The genotype of an individual’s social partner on day 1 of the experiment affected how aggressively they behaved with a new social partner on day 2. What’s more, the genotype of the day 1 social partner also impacted plasticity in behavior towards the new social partner, affecting patterns of retaliation. Partner genotype also matters when it comes to exploratory behavior, though there was no consistent effect, indicating substantial genotype-by-genotype epistasis. This complexity is intriguing – and distressing – because it means that averaging across group members leaves out an important part of the puzzle of IGEs. This concern resonated during the Q&A; how do we effectively study IGEs (and social evolution broadly) when these complicated interactions are so important?! Julia’s advice: future work may reveal certain kinds of social interactions, or interactions during certain developmental periods, that are most impactful. If so, we can focus on those.
Adria LeBoeuf – Trophallaxis is the transfer of food or other fluids among members of a community, often via regurgitation. However, much more than just nutrients can be transferred. Trophallaxis fluids can contain nestmate recognition cues, small RNAs, and hormones that regulate growth. Adria’s work suggests that Trophallaxis functions as a way for colonies to effectively regulate larval development, possibly enhancing the collective behavior of colonies.
Ian Brettell – Medaka are resilient to inbreeding, which makes them a great system for establishing isogenic lines with which to test for IGEs. Harnessing this, Ian observed the swimming behavior of pairs of females and found robust evidence of IGEs.
Eric Wice – Most organisms exist within a social network. To fully understand the quantitative genetic basis of social network position (and thus, how network position evolves) one must account for the genetic makeup of all individuals within that network. Eric placed individuals in groups comprised of one of several different genetic lines. Many different components of network position were affected by the individual’s own genotype and the genotype of its group members, further demonstrating the pervasiveness of IGEs.
Reflecting on the first day of the symposium – There were some clear themes from the day’s talks, and the discussion that followed, about where the field of IGEs stands and what is most urgent for moving the field forward. It was widely agreed upon that IGEs are relevant to a broad set of traits in a wide range of organisms, and that they can (at least in some cases) be very large in magnitude. It is also agreed that, when present, IGEs can have dramatic consequences for social evolution, altering the rate or even the direction of the response to selection. One problem motivating much of work presented today is the problem of missing heritability – can IGEs help to fill this gap?
Perhaps the liveliest discussion centered around how to go about measuring IGEs so that we can better apply this framework. Identifying a coherent phenotype, and measuring it reliably, can be deceptively tricky tasks. And what is the best approach for studying IGEs? Controlled lab environments are useful for precisely understanding the details of things. Field studies are inherently messier, but are important for determining the relevance of what we are studying. So, the two approaches are complementary. Once we’ve identified a sensible phenotype, how do we go about measuring IGEs without getting overwhelmed by all the ways in which IGEs can change depending on focal genotype, timing, group size, etc.? I sensed a combination of dismay, excitement, and optimism when discussing how to grapple with the complexities of IGEs in realistic social settings. It was also interesting to learn that IGEs have been discovered in parallel many times (albeit in somewhat different ways) in different fields, including agricultural breeding, evolutionary genetics, psychology, and medicine. I think this speaks to the broad and substantial impacts that advances in the field of IGEs will likely have for both basic and applied research.
Walid
Today is the first day of the AGA 2020 President’s symposium. The theme this year being “Genes as Environment”. The conference is going to be held for the first time in a virtual manner over 4 days. There will be 20 main talks given by 18 invited speakers from all over the world, with Prof. Allen J. Moore the key distinguished speaker for this year’s symposium. Before the end of each day, there will be several lightning talks given by students and postdocs on their research work. Open discussions and socializing periods during each day are planned as well.
The theme of this year’s symposium “Genes as Environment” implies an important recent biological insight: Genes (or genotypes) of other individuals can constitute the environment affecting the focal individual’s phenotype. In 1997, Moore et al. published their influential study in Evolution focusing on the origin of these effects in social interactions and how they relate to direct genetic effects. Most of the talks given today will focus on indirect genetic effects (IGEs) which is telling of how important studying these effects has become in the field of genetics, whether it is in an evolutionary or ecological framework. It is important to note as well the work of Griffing (1967) which laid the groundwork for understanding how associate or indirect effects can arise in a population with interacting individuals and the need to incorporate these effects in our explicative models.
I will be listening to every talk today – and each day, as I am personally interested in IGEs – and will do my best to write up for each a summary of the contents, results and conclusions.
Allen Moore – Why we need to understand indirect genetic effects
Allen starts out with giving 1 star to the 2020 year, and I would totally agree with that statement.
Allen’s presentation is divided into three parts:
Why was the model developed: The beginning of Allen’s work started with his PhD on the study of social dominance, then going into studying and developing quantitative genetic models for studying maternal effects. Allen with the collaboration of Brodie worked on developing their model based on quantitative genetic framework and the understanding that we need to incorporate selection, inheritance and a proper understanding of the phenotypes.
How have the models been applied: The choice of phenotype was ones that arise from social interactions. Social partners would constitute an environment, but this environment is heritable and can evolve. From this, we can say that the social partner provides an IGE. The important element of the model that was missing at the time was the interaction effect coefficient which determined the importance of the social environment on the focal individual. This coefficient can reflect different social interactions (competition, aggression…). It is important to measure the strength of the interaction, need to know the traits of the interacting individuals that affect expression, measure their genetic variation, and define the phenotype based on the social context. From social interactions, social selection can contribute to the evolution of traits. It is possible to define selection with IGEs. There needs to be a proper measurement of the covariance between the phenotypes of the interactants which can include both a heritable and an environmental source.
Where might we go from here: The applications of IGEs models extends first to the study of sexual conflict, sexual selection and mate choice, and social selections. It can be applied in evolutionary studies that include pedigreed populations allowing for estimating the genetic effects. It can be applied as well to agriculture studying the social interactions between domesticated species. More recently, it is now being applied to health studies, host resistance for example, anxiety and wound healing.
The need for the theory of IGEs rests on the fact that evolutionary models with IGEs can lead to better understanding of changes in phenotypes in light of interactions, insight on the rate of adaptation, better agricultural practices, as well as health and medicinal developments by incorporating IGEs.
Amelie Baud – Indirect genetic effects arising from interactions between cage mates in laboratory rodents
Amelie starts her talk with explaining that her rodent models are used as a model for humans, i.e. phenotypes relevant to human health and disease, focusing on adults and adolescents and interactions between same-sex cage mates.
The goal of her study is to understand if the similar effects are at play in humans and if it is possible to better control the social environment of rodent models. Amelie considers the use of IGE as tools to study social effects. So Amelie considers that there is a conserved element in IGEs and stresses the fact that there is no need for prior knowledge of phenotypic traits and the underlying mechanisms to estimate the IGEs.
The focus is on the impact of IGEs on the phenotypic variance and the DGEs. Then the focus is on the mechanisms underlying the IGEs. The models constructed estimate jointly IGEs and DGEs on the focal phenotype.
The benefit of using the rodent lab populations is that there is little or no confounding bias on the estimates from population structure, cryptic relatedness and non-random partner choice. Additionally, there is both a large sample and a large number of phenotypes included. Found both expected and unexpected IGEs explaining variation in certain phenotypes. And in most of the phenotypes, the IGEs are stronger than the DGEs which might bias heritability estimates which can be overestimated if IGEs are not modeled.
Used correlation between DGEs and IGEs to determine if phenotypic contagion is sufficient to explain IGE, which she found that it is not enough. Hence there could be loci underlying IGEs that are different than those for DGEs for the same phenotype. The igeGWAS model includes the direct effect of the tested locus. The results show no overlap with important DGE loci for the same phenotype with fewer significant IGE loci.
Amelie then focused on a loci Epha4 underlying forced swim test which was only associated with IGE effects and no DGEs for this phenotype. Understanding the mechanisms of this pathway of effect will be tested using knocked-out breeds vs knocked-in.
Piter Bijma – Indirect genetic effects contribute the vast majority of the heritable variation in the prevalence of infectious disease
Piter explains how the phenotype is explained by DGEs and IGEs and we can define total breeding value as well which play a role in the response to selection (heritable variation). The trait studied is the prevalence of an infectious disease hence a binary trait (infected 1 vs non-infected 0).
According to classic theory when prevalence is 0, heritability should be 0 and then selection cannot lead to any change. However from an epidemiological perspective, prevalence of disease is an emergent property. Infecting someone can be considered an indirect effect. Prevalence here is determined by the basic reproduction number showing that there would be an exponential growth of the number of infected in a population. Piter shows that prevalence is a function of the reproduction number, which in of itself depends on other parameters. From this, the genetic traits which are of interest for group immunity: susceptibility, infectivity, and recovery.
Which of these three traits show IGEs. Infectivity is an obvious candidate for creating IGEs, and so is susceptibility and recovery. IGEs everywhere! The presence of IGEs makes a difference between calculating the DGE and total breeding value for prevalence of disease, with a difference of a factor P which is the prevalence. Most of the heritability in prevalence due to susceptibility is due to IGEs. So another example of how important it is to incorporate IGEs in calculating heritability.
Julia Saltz – IGEs in groups
According to Julia, to account for multiple social patterns in the IGE model is to average the phenotypes of all interacting social partners. Though this might not apply when some social partners have greater effects on the focal individual. This would be okay for averaging if we are able to identify these individuals. However, another possibility is the difference in the way a social partner influences the focal individual. In this case, the averaging would not work. So Julia wants to understand though her models of fruit flies how the heterogeneity in the social interactions can affect IGEs.
The first variable is the order of the individual. Meaning when did the interaction occur in time. Early interactions can have long lasting effects. Aggression has order effects, since who won and lost can influence the result of the next interaction for an individual. Studying IGEs over time based on the aggressive behavior of flies. The goal is to see how IGEs can carry over from the interaction in day 1 into an interaction on day 2 against a standard partner. Results show that IGEs can affect mean aggression on the next day, meaning prior experience with different genotypes on day 1 can carry over to the interaction of day 2 against a standard partner. Additionally, the response of the focal individual in day 2 is affected by the IGEs of the first partner.
The second variable is the individual identity of the partner. The trait focused on is the exploratory behavior in the flies. In this case, Julie looked at the variation in IGE within groups. Results show that different group members have different effects on focal exploratory behavior. Though it seems that same-genotype effects are present and it can influence the behavior of the focal individual. Across groups, different partner genotypes have different effects on the exploratory behavior of the focal individual.
So the interaction coefficient meaning the dynamic of the interactions between the focal individual and the interactant depends on genotype-by-genotype interactions or epistasis. So the magnitude of IGEs can vary, for example based on order and identity, and the process of interaction between a focal individual and an interaction depends on other partners as well.
Sabrina
Amelie Baud presented work on the impacts of social interactions on model rodents. Which phenotypes are more affected and what traits have the strongest effects?
Piter Bijma discussed how indirect genetic effects contributed to heritable variation in the prevalence of endemic infectious diseases. Factors such as susceptibility, infectivity, and recovery all impacted prevalence. Before this talk, I always thought of susceptibility as how strong the immune system is. However, there is a lot more to it as it also matters how many social interactions the organism has, which can be driven by its underlying genotype.
Julie Saltz talked about her lab’s research on indirect genetic effects using fruit flies. Having fruit flies in your living space is irritating most of the time and I had not thought of them having different personalities before. The behavior of a fruit fly had a clear effect on a second fruit fly’s behavior. However, whether aggressive behavior made it more or less aggressive the next day also depended on the genotype of the second fruit fly. Similarly, exploratory behaviors of fruit flies are impacted by the genotypes of the focal individual and the other individuals.