Socio-economic status covaries with central nervous system serotonergic responsivity as a function of allelic variation in the serotonin transporter gene-linked polymorphic region

Abstract

It was reported recently that exposure to an adverse rearing environment lowers central nervous system (CNS) serotonergic activity in a nonhuman primate (rhesus monkeys), but only among animals having the shorter variant of a functional, biallelic repeat polymorphism in the regulatory region of the serotonin transporter (5-HTT) gene. Because repeat variants of the same core sequence affect transcriptional efficiency of the 5-HTT gene in humans, we examined whether biallelic variation in the 5-HTT gene-linked polymorphic region (5-HTTLPR) acts analogously to modulate a previously described association between socio-economic status (SES) and CNS serotonergic function. The 5-HTTLPR was genotyped in 139 adult men and women (n=75 and 64) who were administered a standard neuroendocrine challenge to assess central serotonergic responsivity (plasma prolactin (PRL) response to the serotonin releasing agent, fenfluramine). Socio-economic status was estimated from reported income and years of education. Hierarchical linear regression showed serotonergic responsivity to be predicted by the interaction of 5-HTTLPR genotype and SES (p=0.018). Individuals of lower income and less education had lower peak PRL concentrations following administration of fenfluramine than did subjects ranking higher on these dimensions, but only among persons possessing at least one 5-HTTLPR short allele. Within genotype, SES covaried moderately with the PRL response to fenfluramine among subjects who were homozygous for the short allele (r₁₈=0.50, p<0.03). A similar association was present at lesser magnitude in heterozygotes (r₇₀=0.24, p<0.05) and absent among subjects homozygous for the long allele (r₄₅=−0.04, n.s.). Findings were comparable for men and women and persisted on re-analysis restricted to persons without current Axis I psychopathology. We conclude that allelic variation at 5-HTTLPR moderates the influence of social position on CNS serotonergic responsivity.

Introduction

A 44-bp insertion/deletion polymorphism in the 5′ regulatory region of the serotonin transporter (5-HTT) gene may be associated with behavioral phenotypes as diverse as anxiety-related personality traits, various psychopathologies (e.g., affective disorders, alcoholism, bulimia nervosa, autism), late onset Alzheimer’s disease, and cardiovascular reactions to psychological stress (Lesch et al., 1996, Collier et al., 1996, Cook et al., 1997, Li et al., 1997, Oliveira et al., 1998, Rosenthal et al., 1998, Hallikainen et al., 1999, Di Bella et al., 2000, Williams et al., 2001). However, attempts to replicate behavioral associations in the 5-HTT gene-linked polymorphic region (5-HTTLPR) have occasioned both success and failure (e.g., Ball et al., 1997, Ebstein et al., 1997, Klauck et al., 1997, Rees et al., 1997, Esterling et al., 1998, Hoehe et al., 1998, Gelernter et al., 1998, Jorm et al., 1998, Mazzanti et al., 1998, Ohara et al., 1998, Ricketts et al., 1998, Rosenthal et al., 1998, Deary et al., 1999, Flory et al., 1999, Frisch et al., 1999, Gustavsson et al., 1999, Murakami et al., 1999, Zhong et al., 1999, Benjamin et al., 2000, Hu et al., 2000, Jorm et al., 2000, Kunugi et al., 2000, Mann et al., 2000, Osher et al., 2000, Melke et al., 2001, Mossner et al., 2001, Tordjman et al., 2001, Tsai et al., 2001, Yirimya et al., 2001, Steffens et al., 2002). Reasons for these inconsistencies remain unclear, but may entail a variety of methodological differences among studies, such as variability in participant age, sex, and ethnicity, concomitant psychiatric morbidity or other sources of sample heterogeneity, limited statistical power, or differences in study design.

Located ~1 kb upstream of the coding sequence of the 5-HTT gene on chromosome 17q11.2, the 5-HTTLPR contains functional variation known to modulate activity of the 5-HTT gene promoter. Relative to the longer variant of this biallelic repeat, the deletion, or short, allele reduces transcriptional efficiency by about twofold when fused to a luciferase reporter gene and transfected into human placental choriocarcinoma (JAR) cells (Lesch et al., 1996, Heils et al., 1996). In cultured lymphoblasts, the short allele is associated with lower 5-HTT mRNA production and decreased serotonin uptake (Lesch et al., 1996). Other studies addressing functional properties of the 5-HTTLPR are informative as well, although not always consistent. For instance, the short allele may be associated with lower midbrain 5-HTT mRNA expression and transporter availability (Little et al., 1998, Heinz et al., 2000), reduced uptake of serotonin into platelets (Greenberg et al., 1999, Nobile et al., 1999), and lower whole blood serotonin content (Hanna et al., 1998), but these findings are not reported in all studies (Willeit et al., 2001, Stoltenberg et al., 2002, Kaiser et al., 2002). In accord with reduced serotonin turnover in brain, cerebrospinal fluid (CSF) concentrations of the serotonin metabolite, 5-hydroxyindoleacetic acid (5-HIAA), may be lower in persons homozygous for the short allele, but again, this association was seen in only one of two investigations (Jonsson et al., 1998, Williams et al., 2001). Each of two other recent studies suggest that neuroendocrine responses to pharmacologic challenges that enhance serotonergic neurotransmission also vary by 5-HTTLPR genotype. Women homozygous for the short allele, for instance, exhibited smaller elevations in plasma prolactin concentration (i.e. diminished central serotonergic responsivity) on acute administration of clomipramine, an inhibitor of neuronal serotonin reuptake, relative to women carrying two copies of the long allele (Whale et al., 2000). In the second study, males having at least one short variant of the promoter polymorphism also showed a blunted prolactin response to the 5-HT releasing agent, fenfluramine, compared to men likewise homozygous for the long allele (Riest et al., 2001).

Length variation in the 5-HTTLPR of rhesus monkeys (Macaca mulatta) has been described as well (Lesch et al., 1997, Bennett et al., 2002). The rhesus polymorphism (rh5-HTTLPR) is similarly biallelic, and as in humans, luciferase reporter gene constructs containing the short allele demonstrate reduced in vitro transcriptional efficiency. Interestingly, the short allele is also associated with lower CSF 5-HIAA concentrations in rhesus macaques, but only among peer-reared (as opposed to mother-reared) monkeys (Bennett et al., 2002). Peer-rearing constitutes a model of early maternal deprivation that presages developmental abnormalities persisting into adulthood (Higley et al., 1992). These findings suggest that environmental adversity moderates effects of allele-dependent variation in 5-HTT activity on brain serotonergic function. Analogous interactions might conceivably account, in part, for apparent inconsistencies among prior studies involving human subjects, in which study populations often vary markedly. First, however, it may be asked whether attributes of the environment or subject characteristics moderate effects of 5-HTTLPR variation on central nervous system (CNS) serotonergic activity in humans. In a nonpatient sample of community volunteers, we recently observed that rises in plasma prolactin concentration elicited by fenfluramine challenge are attenuated among subjects of lower socio-economic status (SES) (as indexed by income and educational attainment), compared to persons of more advantaged social position (Matthews et al., 2000). In this paper, we report that this association varies in magnitude as a function of allelic variation in the 5-HTTLPR, with lower socio-economic status predicting diminished central serotonergic responsivity only among individuals carrying the short allele of this regulatory polymorphism.