Abstract
The phosphodiesterase 8B (PDE8B) gene encodes the high affinity cAMP-hydrolyzing PDE8B (1) and is strongly expressed in the thyroid and to a lower extent in the pituitary and the hypothalamus (2). It is primarily active in the thyroid in catalyzing the hydrolysis and inactivation of cAMP after thyrotropin (TSH) signaling (1). Besides its role in the thyroid gland, PDE8B is also highly expressed in the adrenal gland, where it is a major regulator of steroidogenesis, and PDE8B mutations have been implicated in adrenal tumorigenesis (3). Common genetic variation in this gene, in particular the A-allele of the rs4704397 polymorphism, has been related to higher TSH levels (1 –5). A proxy of this single nucleotide polymorphism (SNP), rs2046045, also located in intron 1 and in high linkage disequilibrium (r 2=0.98) with rs4704397, was associated with TSH levels in the study of Arnaud-Lopez et al. (1) and in the deCODE cohort as well (6). Nevertheless, the effect of common genetic variance in PDE8B on circulating thyroid hormone levels remains controversial (1,2,4 –6). The initial study of Arnaud-Lopez et al. (1) did not investigate associations with free thyroid hormones, whereas other studies reported a lack of association (4 –6). Taylor et al. (2), in contrast, described a small negative association between the presence of rs4704397 and free thyroxine (FT4) levels.
In this study, we investigated the influence of rs4704397 in the PDE8B gene on TSH and thyroid hormones (FT4 and free triiodothyronine [FT3]) levels in a population of 2524 healthy middle-aged (between 35 and 55 years) men and women (Asklepios study). Study design and baseline characteristics have been described before (7).
Subjects on thyroid medication or with positive thyroid peroxidase autoantibodies (TPOAbs) were excluded, leaving 2416 subjects. Thyroid hormone function tests (TSH, FT3, and FT4) and TPOAbs were determined using immuno-electrochemiluminescence (Roche reagents) on Cobas 411 (Roche Diagnostics GmbH, Mannheim, Germany). Genotyping was performed by KBiosciences using KASPar technology (
In the Asklepios population, the presence of the rs4704397 A-allele in PDE8B was strongly associated with higher TSH levels. Beside this effect on TSH, our results also suggest lower levels of both FT4 and FT3 with the presence of this SNP, although this association was smaller and much less significant (Table 1). Analyses were sex-adjusted. Besides genetic factors, environmental factors are known to determinate thyroid hormone concentrations. Age is an important determinant of TSH, especially at older ages. Nevertheless, in this study, the age range was narrow and additional adjustment for age, as well as for height, weight, and current smoking, did not alter the significance of observed associations (data not shown). The observation of modest inverse associations with circulating free thyroid hormones might add to the understanding of the mechanism whereby this polymorphism influences thyroid hormone homeostasis. Our results, with reciprocal associations for TSH and both FT4 and FT3, support the hypothesis, postulated by Arnaud-Lopez et al. (1) and Taylor et al. (2), that PDE8B plays a role in TSH signaling in the thyroid gland rather than operating at the central control of TSH secretion. This PDE8B polymorphism might reduce cAMP activity in the thyroid, leading to a decreased thyroid stimulatory response to TSH and hence lower subsequent thyroid hormone production (5). A central mechanism altering the TSH set-point in the brain would be expected to cause higher TSH levels together with unaltered thyroid hormone levels (2,4). This hypothesis is further supported by the expression pattern of PDE8B, with predominant expression in the thyroid. However, the associations with free thyroid hormones were small and had a much lower level of significance than the associations with TSH. In the larger cohort of Gudmundsson et al. (6), no associations between rs2046045, a proxy of rs4704397, and free thyroid hormones were observed. The strength of our study is that blood sampling and thyroid assays were highly standardized and (pre)analytical variability was lower, compared with large multicenter studies, thereby allowing detection of smaller associations with free thyroid hormones.
Values for the total group and per rs 4704397 genotype are mean±SD; for TSH, median levels with interquartile range (because of nonnormal distribution) are reported. p-Values result from linear regression analysis under an additive model, adjusting for sex.
In conclusion, we have replicated the previously observed strongly positive association between the A-allele of rs 4704397 and TSH levels. Furthermore, we provide new evidence for negative associations with both FT4 and FT3, although this could also be an indirect effect of the demonstrated strong positive association with TSH, given the log-linear relationship between TSH and free thyroid hormones. Additional validation of the associations with free thyroid hormones is needed.