The terms nature and nurture are used as popular terms for the roles of, respectively, heredity and environment in human development. One extreme view is that our behavior is determined solely by genetic factors, which is referred to as the nature theory of human behavior. The other extreme view is the nurture theory of human behavior, which poses that our behavior is entirely shaped by the environment through experiences. Recently, both theories have been reconciled by assuming that nature and nurture contribute to human behavior, although it is unclear what aspect of behavior is determined by genetic factors and what part by environmental factors. As elegantly described by Charney, the complication to dissect nature versus nurture may be because our genome can change under influence of environmental factors. The genomic plasticity, that is, variability in copy number variations, retrotransposons, and epigenomic variability, may significantly impact genome-wide association studies, which rely on genomic stability. Indeed, these phenomena may account for the so-called missing heritability, which is seen in many complex traits, including behavior and psychiatric conditions. Yet, beyond the important message of Charney, there are two factors that need to be included in the understanding of this missing heritability.

First, not only do environmental stimuli influence the genome, the genome also influences the impact of environmental stimuli. Whereas the “static” diathesis–stress/dual risk view (Monroe & Simons Reference Monroe and Simons1991) adheres to the theory that some individuals are disproportionately likely to be affected adversely by environmental stressors because of a genetic “vulnerability,” the recently introduced “for-better-and-for-worse” concept postulates that “vulnerability” genes actually function as plasticity genes (Belsky et al. Reference Belsky, Jonassaint, Pluess, Stanton, Brummett and Williams2009). These plasticity genes may turn out maladaptive in aversive environments, but adaptive in favorable environments (Belsky et al. Reference Belsky, Jonassaint, Pluess, Stanton, Brummett and Williams2009). A well-known example is the common serotonin transporter-linked polymorphic region (5-HTTLPR). Individuals carrying the short allelic variant of this polymorphism showed maladaptive behavior (e.g., depression) when they were exposed to an aversive environment, but also showed increased adaptive behavior (e.g., improved decision making) in response to a favorable environment (Homberg & Lesch Reference Homberg and Lesch2010). Long-allele carriers, on the other hand, show resilience to aversive environments, but also benefit less from a favorable environment. For instance, female short- (but not long-) allele carriers have increased risk of postpartum depression under poor socioeconomic conditions and less under high socioeconomic conditions (Mitchell et al. Reference Mitchell, Notterman, Brooks-Gunn, Hobcraft, Garfinkel, Jaeger, Kotenko and McLanahan2011). Similarly, in adopted adults with the dopamine D4 receptor (DRD4) 7 repeat (7R) allele, experience with parental problems had the highest scores for trauma, whereas subjects with DRD4-7R who did not experience parental problems showed the lowest ratings. Among participants without this allele, the parental problems during childhood did not make a difference (Bakermans-Kranenburg et al. Reference Bakermans-Kranenburg, van Ijzendoorn, Caspers and Philibert2011). The most plausible explanation for these observations is that plasticity genes shape the sensitivity to environmental stimuli (Belsky et al. Reference Belsky, Jonassaint, Pluess, Stanton, Brummett and Williams2009). Hence, individuals differ by genotype in the extent to which they are affected (either positively or negatively) by environmental exposures. Interestingly, one study showed that the long-allele in combination with high 5-HTT promoter DNA methylation (a type of epigenetic modification) predicted more trauma compared to long-allele subjects with low levels of methylation. On the other hand, the short-allele predicted more trauma, but only when levels of methylation were low (van Ijzendoorn et al. Reference van Ijzendoorn, Caspers, Bakermans-Kranenburg, Beach and Philibert2010). Although it is impossible to dissociate the “chicken and the egg” in cross-sectional association studies, it is tempting to speculate that the impact of early life trauma on DNA methylation varied as function of 5-HTTLPR genotype, and that the level of 5-HTT gene suppression as a consequence of DNA methylation influenced the responses of the children to trauma. The lesson that can be drawn here is that besides that the environment affects the genome (see Charney), the genome also affects the impact of the environment (presumably due to epigenetic processes).

Second, whereas Charney focuses on the importance of the intrauterine and early postnatal environment on changes in the genome and phenotypic plasticity, the impact of later life environment on the genome should not be overlooked. For instance, adult desert locusts have the ability to switch from the solitary phase to the swarming phase depending on population density (Anstey et al. Reference Anstey, Rogers, Ott, Burrows and Simpson2009) through epigenetic modifications (Boerjan et al. Reference Boerjan, Sas, Ernst, Tobback, Lemiere, Vandegehuchte, Janssen, Badisco, Marchal, Verlinden, Schoofs and De Loof2011). This also makes sense from an evolutionary point of view: If early life (epi)genetic programming were unchangeable, the organism would be less able to adapt to environmental changes in later life. Although much of our behavior has been acquired and/or shaped by experiences and epigenetic modifications in early life, it is not the case that these processes end in later life. There is indeed evidence that fear learning, memory, and extinction in adult rodents are associated with histone (Stafford et al. Reference Stafford, Raybuck, Ryabinin and Lattal2012), and DNA methylation (Lubin et al. Reference Lubin, Roth and Sweatt2008) modifications. As stated by Charney, mothers may prepare their offspring for expected future environmental conditions, and if these conditions change, maternal modifications of the offspring's phenotype may turn out to be maladaptive. This may explain common diseases with late-onset phenotypes (Feinberg Reference Feinberg2007). However, if the epigenome continues to be vulnerable to change by environmental influences throughout life, the epigenome associated with a particular disease may also be “treatable.” In support, the histone deacetylase inhibitor trichostatin A rescued impaired fear memory in a mouse model for Alzheimer's disease (Francis et al. Reference Francis, Fa, Ashraf, Zhang, Staniszewski, Latchman and Arancio2009), several pharmacological agents currently used in the treatment of psychiatric disorders act through the epigenome (Wilkinson et al. Reference Wilkinson, Xiao, Kumar, LaPlant, Renthal, Sikder, Kodadek and Nestler2009), and teenagers carrying the 7-repeat version of DRD4 were more positively affected than controls by an intervention targeting substance use (Beach et al. Reference Beach, Brody, Lei and Philibert2010). An intriguing question that arises from this discussion is whether later life environmental factors differentially affect the early life (epi)genetic programming when they match or mismatch the maternal environment. In any case, the examples above clearly illustrate that gene x environment interactions are more complex than described in Charney's article. Combining different theories is necessary to fully understand the “case of the missing heritability.”