We provide the first detailed proteomic analysis of thyroid impact on the heart in male and female aged mice. Previous studies evaluating thyroid effects in the hearts of elderly mice specifically excluded females or did not provide analyses according to sex [25–28]. This is an important gap in knowledge as thyroid hormone interacts with sex hormones at multiple levels ranging from transcriptional to post-translational. We show TH state elicits differential sex-based proteomic changes, especially in the female hypothyroid heart.
Thyroid hormone is known to modulate cardiac phenotype. While hyperthyroidism generally induces myocardial hypertrophy, hypothyroidism can cause a dilated cardiomyopathy [5]. These conditions are highly prevalent in aging populations [10, 29]. The impact of therapy can be highly variable, particularly on the heart [29]. Thus, the principal objective of this study was to assess the proteomic changes due to TH status in the heart for both sexes. Echocardiography allows us to make longitudinal analyses of thyroid effect on cardiac function and morphology. Both sexes appeared exquisitely sensitive to the hyperthyroid state, which induced significant elevations in heart rate. The chronotropic response occurred concomitant with increasing left posterior wall thickness, but with no change in functional parameters such as fractional shortening and ejection fraction. Although, we could not directly measure stroke volume, the increase in heart rate along with stable cardiac chamber size and ejection fraction suggests that TH induced a high cardiac output state. In contrast, the hypothyroid state caused a significant decrease in heart rate in males; however, there was limited chronotropic effect in females. With similar logic and no change in echocardiographic derived dimensions, males presumably had a low cardiac output, while females exhibited no overall change due hypothyroidism. Thus, the data suggest that the hemodynamics of the female aged hearts are more resistant to the effects of low TH levels.
Altered TT4 serum levels, including echocardiographic and morphometric data confirmed that we successfully altered thyroid state in the heart with our perturbations, establishing the basis for our proteomic analyses. We then further defined the proteome for each sex in our aging hearts according to thyroid state. Surprisingly, prior to our analysis, detailed proteomic analyses have not been performed in hearts subjected to thyroid treatment.
A prior investigation [30] reported that sex modulates TH action in both 12 and 20 months old C5/BL/7 mice. Those investigators studied effects of thyroid state on multiple systemic parameters such as body temperature and weight, as well as muscle strength and locomotor coordination. However, they did not perform end-organ specific evaluation other than heart rate. Additionally, molecular analyses were not included. Those authors also alluded to their finding of sex-based differences in TT4 serum levels in response to treatment as a potential reason for phenotypic differences. However, we found no such differences in our aged mice. The TT4 responses to treatments were virtually identical in the two sexes. This discrepancy could be related to differences in protocol; mainly as we administered daily thyronine/thyroxine in chow, and they provided T4 by intraperitoneal injection every 48 hours. Regardless, sex related effects, noted in our study, would not be due to differences in serum TT4 levels.
2D-DIGE in conjunction with mass spectrometry identified 55 cardiac proteins which responded to alterations in thyroid status. Searches of pathway analyses databases, STRING and DAVID, indicated that they are involved in processes relating to mitochondrial function, structural and inflammatory responses, each of which is integral to the function and remodeling of the heart. Additionally, a search of the GeneHancer database identified 29/55 (53%) of the proteins as potential TH receptor targets. This agrees with Ayers et al. [31] report that ~ 50% of TH induced genes display THRβ binding sites.
Prior studies in our laboratory have shown that both aging and chronic hypothyroidism separately impairs myocardial fatty acid oxidation in male rats [32, 33]. Accordingly, we chose the ACAA2 protein to validate our 2D-DIGE MS/MS data given this enzyme regulates the critical final step of mitochondrial fatty acid oxidation. We found that hyperthyroidism decreased ACAA2 in aged hearts only in males. Similarly, a prior study showed decreased ACAA2 protein expression in male rat hyperthyroid liver, supporting our findings [34].
The female hypothyroid hearts appear to undergo a more dynamic response compared to their male counterparts, with 26/55 affected proteins compared 15/55 observed for the male hearts. Many of the proteins altered are members of the mitochondrial respiratory chain, such as QCR1, a subunit of the ubiquinol-cytochrome c reductase complex, and the NDUB or NDUS proteins which participate in the mitochondrial membrane respiratory chain NADH dehydrogenase (Complex 1). This together with our physiological data suggest the female aged heart is more accommodating to the hypothyroid state. Therefore, the finding of more metabolic proteins undergoing altered expression in female hypothyroid heart may yield insight into potential mechanistic difference between sexes.
Inflammatory response proteins were also altered by thyroid status, many of which are glycoproteins. For example, A1BG, a novel thyroid hormone target, is a plasma glycoprotein of unknown function. Glycoproteins play key roles in inflammatory diseases, are known to be altered in cardiomyopathies and thyroid disorders, and have the potential to be exploited for diagnosis and/or treatment targets [35, 36]. Glycoproteins identified in this study, such as Serpina3K (SPA3K) and A1BG, have been shown to be differentially expressed in myocardial pathology [37, 38]. In this study A1BG protein was absent in the MCON hearts and highly expressed in the FCON. The FPTU A1BG expression plunged to the non-detectable levels of MCON and MPTU. Thyroid dysfunction is primarily diagnosed by the levels of thyroid stimulating hormone (TSH), a regulator of TH production. However, bioactivity of TSH is related to glycosylation modifications and age [39]; current clinical assays do not distinguish between TSH glycosylated forms. The expression of varied modified TSH protein in the aged population may account for discrepancies in diagnosis and prevalence observed in studies of aging patients, emphasizing the need for identifying other diagnostic targets. Xiao et al. reported on zinc-α2- glycoprotein increase and its correlation with TH levels in hyperthyroid patients; however, that study did not analyse male and females separately [40]. We have identified several glycoproteins altered in the aged mouse by TH state and differentially expressed between the sexes. Hence, our findings show the importance of examining both males and females for the identification and use of potential biomarkers towards better discernment of subclinical thyroid states or indications for treatment.
TH regulates cardiovascular remodelling, affecting size, shape and function in the heart [41, 42]. Our study identified several cardiac remodelling proteins further validating our data. Thyroid disorders affect the myocardium similarly to that observed by the aging process, such as changes to cardiac contractility, cardiac output and systemic vascular resistance [43]. Additionally, it is well recognized that structural and functional differences exist between male and female, such as anatomical size [44] and electrical activity [45]. We identified myosin heavy chain α (MHC6), for instance, a major protein in cardiac muscle and contractile function that is known to be regulated by TH via the TH hormone nuclear receptor sites in its promoter region [46]. Other structural proteins identified included regulatory myosin light chain 2 and 3 which interact directly with MHC6. Structure-related proteins identified, such as alpha B-crystallin interact with desmin/ actin cytoskeletal complexes, and are predicted to play a protective role during stress conditions [47]. The structural proteins identified in addition to being altered by TH state also showed preferential sex expression. Further investigations of these proteins and their cellular interactions will aid in better understanding any sexual dimorphisms.
An interesting disparity was observed in our study in that the echocardiographic measurements showed modest phenotypic differences between males and females compared to the key differences observed for the proteomic changes. One explanation is that the proteins we observed altered by TH status are predominantly involved in myocardial substrate utilization as opposed to structural or hypertrophic changes. Supporting this rationale, previous work in our lab showed TH treated aged male mouse heart experienced significant alterations in substrate utilization with fatty acids, but exhibited limited or unchanged cardiac functions compared to baseline or aged untreated [30, 32, 33]. Thus, we would need to do specific studies on cardiac substrate utilization for fatty acids to define these sex-differences.
In conclusion, we demonstrate that thyroid hormone differentially regulates the cardiac proteomic profile in an age and sex influenced manner. Although thyroid dysfunction is more prevalent in women previous animal studies have excluded females. In our study, the aged female heart had a greater number of proteomic changes in response to thyroid hormone state. The identification of numerous sex-based differentially expressed proteins altered by thyroid hormone status shows the importance of examining both sexes. The integration of sex-specific data will yield a better understanding of thyroid hormone regulation in the aged heart.
Limitations of the study. First, the proteins involved in thyroid hormone regulation identified by the 2D-DIGE and LC-MS/MS method used in this study are by no means meant to be an exhaustive list. Second, as a screening study, it is not within the scope of the present research to determine if the changes in protein expression levels are due to direct or indirect modulation of the proteins by thyroid hormone. Third, small sample size limits the statistical power of our data [48]. Fourth, it is important to note a major difference between aging female mice and postmenopausal aged women; aging female mice do not have the equivalent of menopause. Aging female mice become acyclic by 11–16 months and develop estrogen deficiency but not as profoundly as aged women [49].