The serotonin transporter gene‐linked polymorphism and negative emotionality: placing single gene effects in the context of genetic background and environment

Mutations and polymorphisms in brain monoamine systems are attractive candidate genes for neuropsychiatric disorders given the weight of evidence implicating these neurotransmitters in various mental illnesses (Cravchik & Goldman 2000). An important regulatory component of these systems is the presynaptic transporter, such as the serotonin transporter (5-HTT), which terminates signaling by recovering neurotransmitter from the synapse (Blakely et al. 1991). Humans carry a common 44-base pair insertion/deletion polymorphism in the promoter region of the 5-HTT gene (serotonin transporter linked polymorphic region, HTTLPR ) that is found in two forms, long (l) and short (s). Individuals with either one or two copies of the s allele (40–70% of a given population) appear to have fewer serotonin transporters than individuals with an l/l genotype (Heinz et al. 2000; Lesch et al. 1996). As 5-HT has been implicated in the mediation of normal and abnormal emotional states, recent research has been directed at testing whether the HTTLPR contributes to individual differences in negative emotional states and risk for affective disorders. In an influential early study, Lesch et al. (1996) found that individuals possessing the s, low-expressing, HTTLPR variant showed high trait anxiety; a finding subsequently confirmed by other large sample studies. However, a number of negative findings have questioned the strength and reliability of the s variant association with negative emotionality (Anguelova et al. 2003). Negative findings could imply that the HTTLPR does not have a major influence on negative emotion. However, a degree of inconsistency has proven common in human behavioral genetic studies and likely reflects the genetically complex (polygenic) nature of a specific behavioral trait; with any one single gene variant contributing a small (e.g. < 10%) amount of the overall interindividual variance (McGuffin et al. 2001). Thus, the effect of the HTTLPR on emotional behavior may be critically dependent on the presence of other interacting genetic variants and environmental factors, each conferring either susceptibility or protection. In this Special Interest Section, four contributions discuss various approaches to identifying genetic and environmental factors interacting with the HTTLPR to influence emotion. The large samples required to obtain sufficient statistical power to discover possible gene–gene interactions affecting behavior is a limiting factor in human genetic studies. The mouse provides a valuable alternative system, given the ability to utilize isogenic strains and manipulate the mouse genome with techniques such as gene targeting. Murphy et al. (2003) review recent research on a 5-HTT (htt) null mutant mouse that has been crossed with mice lacking other major monoamine and neurotrophic genes. This multimutant approach has already provided novel insights and testable hypotheses regarding how the HTTLPR could interact with variants in other candidate genes to impact emotional behavior. An alternative approach to identifying htt–gene interactions, as discussed by Holmes et al. (2003), is to backcross the mouse htt null mutation onto different congenic strains and test whether the behavioral effects of the mutation vary across strains. For example, the htt null mutant mouse is known to exhibit abnormalities in anxiety-like behavior (Holmes et al. in press). The presence of an anxiety-like phenotype in htt null mutant mice on one background, but not another, could provide a starting point for the discovery of genetic modifiers of 5-HTT function, with relevance to understanding variability in HTTLPR s variant penetrance in humans. The HTTLPR s variant, in combination with other genes, may confer increased risk for mood and anxiety disorders, but is unlikely to be determinative in the absence of appropriate epigenetic and environmental factors (Fig. 1). However, there is currently little understanding of how the HTTLPR might interact with stress and trauma to affect risk for neuropsychiatric disease. In an important recent study, Caspi et al. (2003) found that the HTTLPR s variant increased susceptibility for major depression, but only in those individuals that had suffered multiple stressful life events. As reviewed by Barr et al. (2003), similar findings have been acquired in non-human primates. The Rhesus macaque homologue of the HTTLPR s variant is associated with abnormalities in 5-HT function and emotional behavior but, consistent with Caspi et al. (2003), only in monkeys exposed to a stressful early life. In the context of these encouraging findings, an important avenue for future research will be to understand whether there are critical periods of early life development during which interactions between genetic variation in 5-HTT function and environmental insult are most pronounced (Rutter 2002). A given emotional trait, as defined by personality questionnaires, or an emotional disorder, as defined by diagnostic criteria, undoubtedly represents genetically heterogeneous phenomena mediated by multiple neural systems. Such genetic and neural complexity will necessarily compromise the ability to detect the effect of any one gene variant, such as the HTTLPR, at the level of behavior. Being more discrete phenomena, endophenotypes potentially provide a more practical measure of disease diathesis (Gottesman & Gould 2003). Hariri and Weinberger (2003) propose that a more direct approach to assessing the effect of the HTTLPR on emotional responsivity is via direct measurement of neural responses to fear-provoking stimuli. In support of this idea, these authors recently reported that the HTTLPR s variant is associated with greater functional activation of the amygdala following exposure to fearful faces (Hariri et al. 2002). More recent findings indicate that this relationship is further modulated by gender, with greater effects of the s allele evident in females A.R. Hariri, E.M. Drabant, K.E. Munoz, B.S. Kolachana, V.S. Mattay, M.F. Egan & D.R. Weinberger, unpublished). The research strategies presented in this Special Interest Section provide promise for better understanding the influence of the HTTLPR on negative emotionality within the context of an individual's genetic background and environment. More generally, these approaches could provide a useful template for future studies that seek to elucidate the genetic basis of individual differences in behavior and risk for neuropsychiatric disease.