Developmental plasticity: Bridging research in evolution and human health

Abstract

Early life experiences can have profound and persistent effects on traits expressed throughout the life course, with consequences for later life behavior, disease risk, and mortality rates. The shaping of later life traits by early life environments, known as 'developmental plasticity', has been well-documented in humans and non-human animals, and has consequently captured the attention of both evolutionary biologists and researchers studying human health. Importantly, the parallel significance of developmental plasticity across multiple fields presents a timely opportunity to build a comprehensive understanding of this phenomenon. We aim to facilitate this goal by highlighting key outstanding questions shared by both evolutionary and health researchers, and by identifying theory and empirical work from both research traditions that is designed to address these questions. Specifically, we focus on: (i) evolutionary explanations for developmental plasticity, (ii) the genetics of developmental plasticity and (iii) the molecular mechanisms that mediate developmental plasticity. In each section, we emphasize the conceptual gains in human health and evolutionary biology that would follow from filling current knowledge gaps using interdisciplinary approaches. We encourage researchers interested in developmental plasticity to evaluate their own work in light of research from diverse fields, with the ultimate goal of establishing a cross-disciplinary understanding of developmental plasticity.

Figure 1.

Polyphenisms (the appearance of discrete phenotypes in response to environmental variation) can arise in two ways: (A) when environmental variation is discontinuous so that only two regions of the reaction norm are ever expressed, or (B) when the organism exhibits a switch point or threshold value at which an alternate morph is produced. Modified from [112]; colored backgrounds indicate the nature of environmental variation, while dots indicate the environments in which organisms are sampled. Most research on the genes and molecular mechanisms underlying developmental plasticity has focused on organisms that naturally exhibit polyphenisms of the type depicted in (A) or (B), or has focused on the extremes of a phenotypic distribution that is naturally continuous, as depicted in (C), such that an ‘artificial polyphenism’ is created for laboratory study. Few studies have examined a range of developmentally induced, continuous phenotypic variation of the sort typically exhibited by humans and other vertebrates, though this has been attempted in some cases (e.g. [87, 115, 169])

Figure 2.

Approaches for mapping genetic variants that contribute to inter-individual differences in developmental plasticity. (A) If individuals of genotype 1 react differently to an early life environmental stressor than individuals of genotype 2, a gene by early environment (G × early E) interaction is implicated. Psychosocial stress has been a major focus of G × early E studies in molecular psychiatry, but the paradigm is generalizable to other early life environments. (B) Response eQTL studies can uncover G × early E interactions using in vitro or in vivo manipulations (such as treatment with dexamethasone, a synthetic glucocorticoid [170]). Under this paradigm, individuals with different genetic backgrounds, or cells cultured from these individuals, are experimentally exposed to two environments and the relationship between genotype and gene expression is assessed in both conditions. An interaction effect (as depicted here) would indicate that environmental sensitivity is genotype-dependent. (C) ASE measures the interaction between allelic imbalance (a difference in the expression levels of the two copies of a given gene within an individual) and the environment. If a particular early life environment alters the degree of allelic imbalance within heterozygotes for a marker variant (blue) but not homozygotes for the marker (red), this indicates that a G × early E interaction exists at a nearby regulatory genetic variant that is also heterozygous