Refine
Year of publication
- 2017 (2)
Document Type
- Doctoral Thesis (2)
Language
- English (2) (remove)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2)
Keywords
- Schilddrüse (2) (remove)
The thyroid as the largest endocrine gland mainly produces and secretes the thyroid hormones (TH): 3,3’,5-triiodo-L-thyronine (T3) and its pro-hormone L-thyroxine (T4). Besides the impact on growth, normal development, bone marrow structure, the cardiovascular system, body weight and thermogenesis, TH play a vivid role in many metabolic regulatory mechanisms in almost all tissues. Thyroid diseases are relatively prevalent and cause, due to the resulting TH imbalances, a broad spectrum of effects. Many of them manifest in pathologically increased or decreased TH levels defined as hyperthyroidism or hypothyroidism, respectively. Routinely, determination of the thyroid state is based on the assessment of the classical markers TSH and free T4. However, this practice has several drawbacks. Moreover, elucidation of the pleiotropic effects of TH on multiple molecular pathways is mostly based on cell culture, tissue and rodent models. Analysis of animal biofluids like serum and urine using metabolomics approaches demonstrated the extensive impact of TH on other body compartments. In contrast, proteome profiling has not been exploited for the comprehensive characterization of the general metabolic effects of TH. Plasma as a large and diverse compartment of the human proteome provides a great opportunity to identify novel protein markers of thyroid function as well as to characterize metabolic effects of TH in humans.
Therefore, a study of experimental thyrotoxicosis was performed with 16 male volunteers treated with 0.25 mg/d levothyroxine (L-T4) for 8 weeks to induce a hyperthyroid state. Plasma samples were collected before the L-T4 application started, two times during the treatment and additionally two times after withdrawal. Proteome analysis revealed remarkable alterations including increased levels of two known proteins known to correlate with TH levels (sex hormone-binding globulin and cystatin C). The correlation with free T4 levels revealed 76 out of 437 detected proteins with a Pearson correlation coefficient of r ≥ |0.9|. One prominent signature included 10 coagulation cascade proteins exhibiting significantly increased plasma levels during thyrotoxicosis, thereby revealing a trend towards a hypercoagulative state in hyperthyroidism. To overcome the statistical drawbacks of the Pearson correlation analysis, additionally a mixed-effect linear regression model using serum free T4 concentrations as exposure and protein abundances as outcome while controlling for age, BMI, and batch was implemented. Application of this model resulted in the detection of 63 proteins with significant associations to free T4 levels. Besides the already mentioned augmented coagulation, a significant drop in the amounts of three apolipoproteins (ApoD, ApoB-100 and ApoC3) was observed. Furthermore, an increased abundance of proteins assigned to the complement system was detected.
Experimental studies in humans were complemented by corresponding analyses in murine models. In the current work, plasma samples of two murine studies including male C57BL/6 wildtype mice were analyzed to elucidate the impact of thyroid dysfunction on the plasma proteome. The first study was similarly designed as the human model of experimentally induced thyrotoxicosis and assigned the animals to three groups: a control group, a T4 treatment group, and a T4 recovery group, whereupon the latter first received T4 followed by a subsequent TH normalization period. A high proportion of plasma proteins exhibited significantly different protein levels during T4 application (n = 120), where 90 of these also showed a corresponding reverse trend after T4 withdrawal (T4 recovery vs. T4), thereby displaying transient alterations. The molecular pattern of hyperthyroidism in the murine model indicated, as in the human study, a pronounced decrease in apolipoproteins. However, in clear contrast to the human data, the levels of proteins related to the coagulation cascade and complement system were also transiently decreased in mice, while being increased in humans.
The second murine analysis focused on the impact of hyper- and hypothyroidism caused by T3 or T4 treatment and MMI/KClO4 application, respectively. In general, compared to the first murine study less clear alterations of protein levels were detected. Proteins related to the complement system revealed fewer changes in the T3 group and only marginal changes after T4 induction. Unexpectedly, the MMI/KClO4-induced hypothyroidism caused a reduction of the levels of several proteins assigned to the complement system, although different components and factors were affected.
Generally, rodent studies partially provided a divergent picture of TH action as compared to human studies. However, in spite of inconsistent results in studies regarding the effects of TH that are possibly due to species-specific differences, an important role of TH on several metabolic and other pathways, e.g. in the process of blood coagulation and apolipoprotein regulation, is evident. The results from both murine and human studies presented here provide novel insights into changes in the plasma proteome in the context of thyroid diseases which might contribute to a better understanding of TH action on metabolism and other pathways.
The thyroid gland is of crucial importance in human metabolism. Its main secretion products, L-thyroxine (T4) and 3,3’,5-triiodo-L-thyronine (T3), are essential for proper development of multiple tissues and organs as well as for their functioning in the adult organism. The secretion of thyroid hormones (TH) is stimulated by thyrotropin (TSH) released from the pituitary gland. This tight connection between both hormones is of crucial importance for the clinical diagnosis of thyroid dysfunction. During the last two decades the concept of TH action developed to increased complexity. However, most of the recent advances in the field of TH research are based either on cell culture, tissue or animal models or stem from studies investigating specific hypotheses in humans. Thus, experimental approaches for the comprehensive, hypothesis-free characterization of metabolic effects of classical and non-classical TH in human are urgently needed. This holds true in particular for the TH derivative 3,5-diiodothyronine (3,5-T2). It was described to alleviate the typical detrimental metabolic consequences of a high-fat diet and even reversed hepatic steatosis. To replicate these experimental findings from rodents in humans, comprehensive data from the population-based Study of Health in Pomerania (SHIP) was analyzed in the present work. Based on a euthyroid, diabetes-free SHIP-subsample (N=761), non-linear associations between the serum concentrations of 3,5-T2 and glucose as well as TSH were detected. In contrast, no significant 3,5-T2 associations with several anthropometric markers or blood lipid parameters were observed, partially questioning the transferability of the beneficial metabolic 3,5-T2 effects reported for pharmacological intervention studies on rodents to humans. Recent advances in technological development now allow for the use of high-throughput spectrometric platforms to characterize the small molecule content (metabolome) of blood and urine samples. The detected metabolome constituents can be associated with any relevant parameters of interest, thereby extending the scope of classical association studies. Therefore, in the second part of the present thesis, the metabolic fingerprints of FT4, TSH as well as the ratio log(TSH)/FT4 as markers of thyroid function were profiled. Strong differences between the metabolic fingerprints of FT4 and TSH were observed, partially alleviated by the log(TSH)/FT4 ratio. These findings not only emphasize the high diagnostic value of the combined evaluation of TSH and FT4 in the assessment of thyroid function but additionally argue for a holistic approach in the diagnosis of thyroid function. More moderate endogenous effects of 3,5-T2 were evaluated by comparing its urinary metabolic fingerprint with that of the classical TH. A number of associations became apparent, indicating a function of endogenous 3,5-T2 in intermediary metabolism. Besides partially confirming associations with respect to the presented findings in animal studies, the strongest 3,5-T2-association was observed with trigonelline, a metabolite described earlier to exhibit similar beneficial effects as 3,5-T2 on glucose metabolism when used as a pharmacological agent in animal studies. An association towards hippurate indicated a partial overlap with the metabolic profile of TSH and hence consolidated results from the first two projects in the sense of a thyromimetic role of 3,5-T2 in the feedback regulation of TH. The diagnosis of thyroid disorders based on the classical markers TSH and FT4 suffers from restricted sensitivity in the subclinical range as both parameters have broad reference ranges in the general population. Therefore, in an approach to detect novel peripheral biomarkers of thyroid function, sixteen healthy young men were challenged with 250 µg of levothyroxine (L-T4) over a period of eight weeks in the fourth project presented here as part of this thesis. Monitoring of the volunteers over a period of sixteen weeks allowed delineation of the metabolic shifts first towards thyrotoxicosis and later in the context of the restoration of euthyroidism. The use of mass spectrometry for the comprehensive characterization of the metabolite as well as the protein content of samples taken at the different time points revealed profound molecular alterations, despite the lack of any clinical symptoms in the volunteers. Molecular signatures of thyrotoxicosis indicated increased energy expenditure, pronounced defense against systemic oxidative stress, a general drop in apolipoproteins, as well as increased abundances of proteins related to the coagulation cascade and the complement system. Good and robust classification of the thyroid state independent of TSH and FT4 was achieved using random forest analysis with a subset of fifteen metabolites and proteins, indicating new options in the individualized diagnosis of thyroid disorders.