Skip to main content
NCBI home page
Microbial Biotechnology logoLink to Microbial Biotechnology
. 2021 Mar 17;15(3):817–831. doi: 10.1111/1751-7915.13793

Cecal microbial transplantation attenuates hyperthyroid‐induced thermogenesis in Mongolian gerbils

Saeid Khakisahneh 1,2, † , Xue‐Ying Zhang 1,2, † ,✉, Zahra Nouri 1,2, De‐Hua Wang 1,2,✉
PMCID: PMC8913869  PMID: 33729663

Summary

Endothermic mammals have a high energy cost to maintain a stable and high body temperature (Tb , around 37°C). Thyroid hormones are a major regulator for energy metabolism and Tb . The gut microbiota is involved in modulating host energy metabolism. However, whether the interaction between the gut microbiota and thyroid hormones is involved in metabolic and thermal regulations is unclear. We hypothesized that thyroid hormones via an interaction with gut microbiota orchestrate host thermogenesis and Tb . l‐thyroxine‐induced hyperthyroid Mongolian gerbils (Meriones unguiculatus) increased resting metabolic rate (RMR) and Tb , whereas Methimazole‐induced hypothyroid animals decreased RMR. Both hypothyroid and hyperthyroid animals differed significantly in faecal bacterial community. Hyperthyroidism increased the relative abundance of pathogenic bacteria, such as Helicobacter and Rikenella, and decreased abundance of beneficial bacteria Butyricimonas and Parabacteroides, accompanied by reduced total bile acids and short‐chain fatty acids. Furthermore, the hyperthyroid gerbils transplanted with the microbiota from control donors increased type 2 deiodinase (DIO2) expression in the liver and showed a greater rate of decline of both serum T3 and T4 levels and, consequently, a more rapid recovery of normal RMR and Tb . These findings indicate that thyroid hormones regulate thermogenesis depending on gut microbiota and colonization with normal microbiota by caecal microbial transplantation attenuates hyperthyroid‐induced thermogenesis. This work reveals the functional consequences of the gut microbiota‐thyroid axis in controlling host metabolic physiology and Tb in endotherms.

In this paper, we demonstrate that hyperthyroidism showed increases in resting metabolic rate (RMR) and body temperature (T b ), and were associated with a disturbed gut microbial community, with increased relative abundances of pathogenic bacteria, such as Helicobacter and Rikenella, and decreased abundances of beneficial bacteria Butyricimonas and Parabacteroides. Colonization with normal microbiota by caecal microbial transplantation attenuated hyperthyroid‐induced thermogenesis through regulation of type 2 deiodinase (DIO2) expression. These findings emphasize the functional consequences of the gut microbiota‐thyroid axis in controlling host metabolic physiology and T b .

graphic file with name MBT2-15-817-g006.jpg

Introduction

Endothermic mammals have a high cost to maintain a relatively stable and high body temperature (Tb , around 37°C). Thyroid hormones play a key role in regulating energy metabolism and thermoregulation in endothermic mammals (Silva, 2003; Mullur et al., 2014). The active 3,5,3’‐tri‐iodothyronine (T3) is usually derived from inactive thyroxine (T4) by type 2 deiodinase (DIO2) which is localized widely in the hypothalamus and peripheral tissues (Mullur et al., 2014). T3, acting via its nuclear receptors, modifies gene expression related to metabolism and development. In humans, serum thyroid hormone levels are clinically important indicators of thyroid diseases, such as hyperthyroidism and hypothyroidism. The factors influencing the conversion from T4 to T3 are diverse, but one of the important mechanisms may rely on the gastrointestinal tract and commensal microbiota (de Herder et al., 1989).

The gastrointestinal tract of a man of standard body weight is colonized by 3.9 × 1013 bacteria, along with bacteriophages, archaea, viruses and fungi (Sender et al., 2016). This huge community of gut bacteria plays crucial functions not only in digestion but also in hormonal, immunological and metabolic homeostasis (Tremaroli and Backhed, 2012; Fetissov, 2017). Research in microbial endocrinology indicates that gut microbiota can produce the same neuroactive compounds as the host and possess cognate receptors, which enable bi‐directional communications between the host and the gut microbiota (Lyte, 2014). For example, certain bacteria like Bacteroides and Lactobacillus can produce neurochemicals such as γ‐aminobutyric acid that are major inhibitory neurotransmitters of both the vertebrates and microbes, and regulate physiological functions (Barrett et al., 2012; Strandwitz et al., 2019). The deiodinases are widely distributed in the gastrointestinal tract and this may contribute substantially to the whole body T3 pool, due to the large gut surface, and the gut microbiota can hydrolyse iodothyronine conjugates (Nguyen et al., 1993). Thus, the gut microbiota may affect deiodinase activity and alter the availability of T3 and T4, and thus may modulate hormone recycling and thyroid homeostasis (Fröhlich and Wahl, 2019; Shin et al., 2020). In addition, thyroid function may affect gut motility and bacterial growth (Lauritano et al., 2007). Bacterial overgrowth in the small intestine was found in a high percentage of humans with a history of hypothyroidism, presumed to be due to slower transit time (Lauritano et al., 2007), whereas hyperthyroidism was associated with significantly accelerated intestinal transit compared to controls (Wegener et al., 1992). Although phenotypic data support a possible role of microbiota in the peripheral metabolism of thyroid hormones and thyroid function (Köhling et al., 2017; Virili and Centanni, 2017; Feng et al., 2019; Knezevic et al., 2020), the interaction between gut microbiota and thyroid homeostasis in regulating metabolic rate and Tb is still unclear.

Wild Mongolian gerbils inhabit semi‐arid steppes, desert grasslands, and agricultural fields of northern China, Mongolia, and Russia (Wilson and Reeder, ). They show seasonal changes in body mass, resting metabolic rate (RMR), and nonshivering thermogenesis (NST) (Zhang and Wang, 2007), and exhibit plastic responses in body mass, body composition, morphology of digestive tract, energy intake, and serum leptin level to cope with variations in temperature, photoperiod and food availability (Li and Wang, 2005; Zhao and Wang, 2006, 2009). Thyroid hormones are involved in regulating metabolic rate at different temperatures (Khakisahneh et al., 2019). Our recent study showed that alterations in gut microbiota mediated huddling‐ or cold‐induced metabolic and thermal homeostasis (Zhang et al., 2018). However, it is not known whether the interaction between gut microbiota and thyroid hormones is involved in metabolic and thermal regulations. We first investigated the changes in RMR, Tb , and gut microbiota composition in drug‐induced hypothyroid or hyperthyroid gerbils and then examined whether the gut microbiota could attenuate hyperthyroid‐ or hypothyroid‐induced metabolic phenotypes. We hypothesized that thyroid hormones via an interaction with gut microbiota orchestrate host metabolism and thermogenesis. We demonstrate that hyperthyroid‐induced increases in RMR and Tb were associated with a disturbed gut microbial community, and colonization with normal microbiota could attenuate hyperthyroid‐induced thermogenesis.

Results

Hyperthyroid and hypothyroid states alter metabolic phenotypes and bacterial metabolites

To test the role of thyroid hormones in regulating metabolic phenotypes and bacterial metabolites, the gerbils drank tap water with 0.0036% l‐thyroxin (Hyper), 0.04% Methylimazol (to inhibit thyroid hormone synthesis, Hypo) or just water (control) for 4 weeks (more details about the experimental design in the method section). There was no change in body mass among groups or with time (Fig. S1A). Food intake in the Hyper group was 67% higher than in the control (P < 0.001; Fig. 1A). At the end of experiment, both RMR and maximum NST (NSTmax) were higher in the hyperthyroid (by 167% in RMR and by 72% in NSTmax), and were lower in the hypothyroid gerbils (by 38% in RMR and by 36% in NSTmax) than controls (P < 0.001; Fig. 1B). The regulatory NST (NSTreg) showed a significant difference only between the Hyper and Hypo groups (P < 0.001; Fig. 1B). Hyper gerbils had a higher Tb (by 0.4–0.7°C) from the 15th day after T4 treatment till the end than the other two groups (P < 0.05; Fig. 1C), and showed a slight but not significant increase in activity in the first week of treatment (Fig. 1D). DIO2 expression in the liver was lower in the Hyper group than in the control (Fig. 1E and F). The Hyper gerbils had lower levels of serum total bile acids compared to the control and Hypo groups (P = 0.007; Fig. 1G). There was no difference in serum tumour necrosis factor alpha (TNF‐α) among groups (Fig. S1B). The concentrations of faecal total short‐chain fatty acids (SCFAs) (P = 0.007), particularly acetic acid, butyric acid and valeric acid, were significantly lower in the hyperthyroid gerbils (Fig. 1H and I, S1C), whereas the concentration of isobutyric acid was higher in the hypothyroid than in the control group (P = 0.022; Fig. 1I).

Fig. 1.

Fig. 1

Open in a new tab

Metabolic phenotypes and metabolites in the hyperthyroid (Hyper), hypothyroid (Hypo) and control gerbils.

A. Food intake.

B. Resting metabolic rate (RMR), maximum nonshivering thermogenesis (NSTmax) and regulatory NST (NSTreg).

C, D. Average core body temperature (Tb ) and activity during the experiment (n = 5 per group).

E, F. The expression of type 2 deiodinase (DIO2) in the liver, measured by Western blot (WB) and ELISA.

G. The levels of total bile acids in serum.

H, I. The concentrations of faecal short‐chain fatty acids (SCFAs). Data are presented as means ± SEM (n = 7‐8 per group, except Tb and activity). *P < 0.05, Hyper vs Control. Different letters above columns indicate significant differences among groups (P < 0.05).

Hyperthyroid and hypothyroid states alter the diversity and composition of faecal microbiota

To determine whether thyroid status affected gut microbiota, we analysed 16S rRNA gene sequencing in faecal samples in Control, Hyper and Hypo groups. The sequencing resulted in a total of 3 210 280 valid reads which were identified into 739 907 unique OTUs at a threshold of 97% sequence identity. The rarefaction curve of Goods coverage for the samples reached saturation (Fig. S1D), indicating that most bacteria were identified in the study. The Hypo gerbils showed higher phylogenetic diversity of the microbial community than the control and Hyper groups, but the Hyper group did not differ from the control (α‐diversity, Table S1). The β‐diversity indicated by Principal Coordinate Analysis (PCoA) based on unweighted (ANOSIM, R = 0.319, P = 0.001) and weighted UniFrac distances (ANOSIM, R = 0.212, P = 0.011) showed that the structure of the microbial communities differed from each other (Fig. 2A).

Fig. 2.

Fig. 2

Open in a new tab

Thyroid hormone treatments shape bacterial composition.

A. Principal coordinate analyses (PCoA) plots based on unweighted and weighted UniFrac distances in faecal microbiota of different groups (unweighted, ANOSIM, R = 0.319, P = 0.001; weighted, ANOSIM, R = 0.212, P = 0.011). The ellipses cover 68% of the data for each group.

B. Relative abundance of different bacteria at the phylum and genus levels (‘+’indicates the mean of data).

C Differential bacterial taxonomy selected by LEfSe analysis with LDA score > 2 in the faecal microbiota community (Differences between groups are represented by the colour of the most abundant class).

D. Venn diagram based on OTUs distribution between groups.

E. Heatmap of correlation between specific OTUs and physiological parameters. Data are presented as means ± SEM (n = 7‐8 per group). Different letters above columns indicate significant differences among groups (P < 0.05). Hyper, hyperthyroid; Hypo, hypothyroid.

The dominating bacterial phyla and families are shown in Fig. S1E, F. At the phylum level, the relative abundance of Bacteroidetes tended to increase in Hyper gerbils and decreased in the Hypo group compared to the control, and Firmicutes showed inverse patterns (Fig. 2B). The proportion of Proteobacteria was significantly higher in the Hyper and Hypo groups than in the control group (One‐way ANOVA, F 2,19 = 8.734, P = 0.002; post hoc tests, P < 0.05; Fig. 2B). At the genus level (Fig. 2B, Table S2), the Hyper gerbils showed increased proportions of Rikenella (F 2,19 = 5.335, P = 0.014), Helicobacter (F 2,19 = 7.968, P = 0.003), and Coprococcus (F 2,19 = 4.162, P = 0.032), but decreased proportions of Butyricimonas (F 2,19 = 4.994, P = 0.018) and Parabacteroides (F 2,19 = 3.046, P = 0.071) when compared with the control group. To assess differences in microbial communities affected by thyroid hormone treatment, we applied the linear discriminant analysis (LDA) effect size (LEfSe) method and identified four distinctive bacteria in the control gerbils, 10 in the Hyper, and 23 in the Hypo gerbils [LDA score (log 10) > 2] (Fig. 2C; Fig. S1G).

The Venn diagram identified the specific genus in the OTUs of control as Devosia, specific genera of the Hyper as Staphylococcus and Veillonella, and specific genera of the Hypo as Rubrobacter, Syntrophomonas, Luteimonas and Mycoplasma, and there were 52 common genera belonging to all groups (Fig. 2D). Pearson correlation analyses revealed that both food intake and RMR were correlated with OTUs of Coprococcus, Odoribacter, Oscillospira, Prevotella and Ruminococcus, and serum T3 and/or T3/T4 levels were correlated with OTUs of Coprococcus, Helicobacter and Rikenella (Fig. 2E; Table S3).

Caecal microbiota transplantation attenuates hyperthyroid thermogenesis

To test the role of gut microbiota in regulating thyroid hormones and metabolism, we performed caecal microbiota transplantation (CMT) in the control and Hyper gerbils (Fig. 3A). The Hyper gerbils were colonized via an intragastric gavage with the microbiota from the control donors (Hyper‐Con) or with sterile saline (Hyper, sham CMT); while the control gerbils were colonized with microbiota from the Hyper animals (Con‐Hyper) or with sterile saline (Con, sham CMT). Food intake both in the Hyper and Hyper‐Con gerbils increased dramatically compared to that in the control group by the end of the T4 treatment (F 3,15 = 20.875, P < 0.001; Fig. 3B). After CMT, food intake in the Hyper‐Con group decreased significantly and showed no difference from the control group in the third week after CMT (P = 0.117, Fig. 3B), whereas food intake in the Hyper gerbils remained higher than both the control and Hyper‐Con gerbils till the end of experiment (the 6th week) (P < 0.05, Fig. 3B). All Hyper animals showed increased RMR by the end of the T4 treatment when compared with the control (F 3,14 = 27.768, P < 0.001; Fig. 3C). The RMR in Hyper‐Con gerbils decreased markedly and returned to the control level by the 6th week after CMT (P = 0.212). In contrast, the Hyper gerbils showed a steady higher RMR even in the 6th week after termination of the T4 treatment (P = 0.011). At the end of the T4 treatment (0th week), all the Hyper gerbils showed significantly higher Tb compared with their control counterparts (F 2,11 = 8.713, P = 0.005; Fig. 3D). The Hyper‐Con gerbils started to decrease Tb immediately following the CMT and did not differ from the control group by 5 weeks after the CMT (P = 0.649), while Tb in the Hyper group remained higher than in the control gerbils till the end (P = 0.02). Both T3 and T4 were at higher levels after T4 treatment (week 0, Fig. 3E and F), and then started to decline after termination of drug treatment. The Hyper gerbils still showed higher T3 and T4 levels 6 weeks after drug treatment, whereas the Hyper‐Con gerbils returned to the control level by the first week after CMT (Fig. 3E and F). The DIO2 expression in the liver was higher in the Hyper‐Con group than in the Hyper group (P < 0.075; Fig. 3G and H).

Fig. 3.

Fig. 3

Open in a new tab

Gut microbiota buffer hyperthyroidism‐induced hyperphagia and hypermetabolism.

A. Schematic overview of experimental design.

B. Food intake in different groups during the experiment.

C. Resting metabolic rate (RMR) of every group during the experimental period.

D. Average core body temperature (Tb ) of gerbils from different groups.

E, F. Serum tri‐iodothyronine (T3) and thyroxine (T4) levels.

G, H. Type 2 deiodinase (DIO2) by Western blot (WB) and ELISA. Values are means ± SEM (n = 5 per group). *P < 0.05, Hyper versus Con. # P < 0.05, Hyper‐Con versus Con. + P < 0.05, Hyper‐Con versus Hyper. CMT, caecal microbiota transplantation; Hyper‐Con, Hyper gerbils were colonized with control microbiota; Hyper, Hyper gerbils with sterile saline; Con‐Hyper, control gerbils were colonized with Hyper microbiota; Con, control gerbils with sterile saline.

One week after CMT, the α‐diversity of microbial community showed a difference between Hyper and Con‐Hyper groups, indicated by observed OTUs and PD whole tree (Table S4, Fig. 4A). The β‐diversity showed separation only between Hyper‐Con and control measured by unweighted UniFrac distances (ANOSIM, R = 0.181, P = 0.005), but did not separate among groups indicated by weighted UniFrac distances (ANOSIM, R = 0.083, P = 0.817; Fig. 4B). Microbiota transplantation led to an alteration in the microbial composition of recipients in the Hyper‐Con gerbils with a 21% increase in the Firmicutes compared to the Hyper group (Fig. 4C and D), and the Hyper‐Con gerbils increased the proportions of Allobaculum (F 3,16 = 6.221, P = 0.005) and Lactobacillus (F 3,16 = 8.186, P = 0.002; Fig. 4D) one week after CMT, when compared with the Hyper group. We also observed that the Con‐Hyper gerbils showed a higher abundance of Roseburia compared to the control group (F 3,16 = 4.046, P = 0.026; Fig. 4D). The genus Anaeroplasma maintained a relatively higher proportion even 6 weeks after CMT (Fig. 4E). The LEfSe method identified distinctive microbial biomarkers in each group (Fig. S2A and B).

Fig. 4.

Fig. 4

Open in a new tab

Caecal microbiota transplantation (CMT) alters faecal microbiota composition in hyperthyroid gerbils.

A. Phylogenetic diversity (PD)—whole tree analysis for the samples.

B. Principal coordinate analyses (PCoA) plots based on unweighted and weighted UniFrac distances in faecal microbiota of different groups (unweighted, ANOSIM, R = 0.181, P = 0.005; weighted, ANOSIM, R = 0.083, P = 0.817).

C. Relative abundance at the phylum levels in faecal microbiota community one week after CMT in each individual animal of the four groups.

D. The relative abundance of different bacteria at the phylum and genus levels in different experimental groups at week 1 after CMT.

E. The relative abundance of some specific bacteria in different experimental groups before (week 0) and after CMT (weeks 1 and 6). Data are presented as means ± SEM (n = 5 per group). Different letters above columns indicate significant differences among groups (P < 0.05). Hyper‐Con, Hyper gerbils were colonized with control microbiota; Hyper, Hyper gerbils with sterile saline; Con‐hyper, control gerbils were colonized with hyper microbiota; Con, control gerbils with sterile saline.

Microbiota transplantation has no effects in the hypothyroid gerbils

To explore whether gut microbiota could modulate the hypothyroid‐induced reduction in metabolic rate, we induced hypothyroidism in Mongolian gerbils with administrated methimazole in the drinking water (Hypo), and transplanted caecal microbiota from the control donors to hypo gerbils or from hypo donors to control gerbils (for more details see follow in method section) (Fig. 5A). Neither body mass nor food intake was influenced by CMT among groups (P > 0.05) or with time (P > 0.05; Fig. 5B and C). RMR also showed no difference among groups (F 3,17 = 1.328, P = 0.298; Fig. 5D). Methimazole treatment led to a significant decrease in serum T3 levels in the gerbils before CMT (F 3,17 = 5.547, P = 0.007; Fig. 5E), but one week after CMT till the last week of experiment no significant difference was observed in serum T3 levels among groups (week 1, F 3,17 = 1.448, P = 0.257; week 6, F 3,17 = 3.055, P = 0.051; Fig. 5E).

Fig. 5.

Fig. 5

Open in a new tab

Caecal microbiota transplantation (CMT) did not affect metabolic traits in hypothyroid or control gerbils.

A. Schematic overview of experimental design.

B. Body mass.

C. Food intake.

D. Resting metabolic rate (RMR).

E. Serum tri‐iodothyronine (T3) levels. Values are means ± SEM (n = 5 per group). *P < 0.05, Hypo versus Con. Hypo‐Con, Hypo gerbils were colonized with control microbiota; Hypo, Hypo gerbils with sterile saline; Con‐hypo, control gerbils were colonized with hypo microbiota; Con, control gerbils with sterile saline.

Microbiota transplantation did not affect the α‐diversity of the microbial community in the recipients (Fig. S3A). The β‐diversity based on unweighted UniFrac distances showed clear separation between the control and Hypo groups (ANOSIM, R = 0.275, P = 0.001), but no separation between CMT and sham CMT (Fig. S3B). The β‐diversity based on weighted UniFrac distances showed no separation among groups (ANOSIM, R = 0.070, P = 0.995; Fig S3B). The specific bacteria in each group were identified by LEfSe method (Fig. S3C and D).

Discussion

Increasing evidence indicates that the gut microbial community is crucial for metabolic homeostasis (Tremaroli and Backhed, 2012). To understand the relationship between gut microbiota, thyroid hormones and metabolism, we determined the changes in gut microbiota diversity in the hyper and hypothyroid gerbils, and further evaluated the role of the gut microbiota in regulating thyroid hormones and metabolic phenotypes in Mongolian gerbils. We observed, as expected, that the hyperthyroid gerbils had higher RMR and Tb compared with the control and hypothyroid groups. Both hyper‐ and hypothyroid animals showed significant alterations in the bacterial composition. Interestingly, CMT of control bacteria to the hyperthyroid gerbils had a profound effect in reducing RMR, Tb and thyroid hormone levels. These results indicated that the interaction between thyroid hormones and gut microbiota contributes to regulating host thermogenesis.

Thyroid hormones shaped metabolic phenotype and microbiota diversity

It is conventionally thought that thyroid hormones regulate not only basal metabolic rate, but also the thermogenesis from brown adipose tissue (BAT) (Mullur et al., 2014). As expected, our data showed that thyroid hormone levels have direct effects in controlling metabolism and Tb . The hyperthyroid gerbils had higher RMR (by 167%), NSTmax (by 72%), and Tb (by 0.4–0.7°C) compared to the control. However, the regulatory NST (directly from BAT thermogenesis) both in the hyper and hypothyroid gerbils did not differ significantly from the control, although there was a difference between the hyper and hypothyroid groups. Our data are supported by evidence that thyroid thermogenesis is independent of uncoupling protein 1 (UCP1, a thermogenic protein found in the mitochondrial inner membrane of BAT) (Cannon and Nedergaard, 2004; Dittner et al., 2019; Johann et al., 2019; Mittag, 2020). The higher energy expenditure in hyperthyroid gerbils was compensated by increased food intake (by 67%) and these gerbils could maintain body mass stable. Diet drives quick changes in gut microbial community in humans and animals (Gentile and Weir, 2018; Zmora et al., 2019). Hyperthyroid‐induced changes in gut microbiota might be induced by increased food consumption and transit time. We did not restrict food intake of animals to exclude this possibility. Our previous study, via a paired‐feeding protocol, supported that the alteration in diversity and composition of gut microbiota in cold‐acclimated animals can be induced by ambient temperature, not due to increased food intake (Bo et al., 2019). Gut microbes produce a variety of neuroendocrine hormones and possess cognate host receptors (Lenard, 1992; Roshchina, 2010; Lyte, 2014). Therefore, changes in host thyroid hormones could induce corresponding responses both in the host and gut microbiota. No significant difference in serum TNF‐α (one inflammatory factor) was observed in the different groups, implying that the higher Tb in the hyperthyroid gerbils was probably not due to an inflammatory response. We observed an initial increase followed by a decrease to the control levels in the activity of hyperthyroid gerbils. The change in activity curves could be interpreted to suggest that the hypothyroid gerbils increase activity to stay warm and the hyperthyroid animals decrease activity to counteract body temperature increases. These data indicate that hyperthyroidism induces increases in thermogenesis, which leads to the elevated Tb .

The gastrointestinal tract is the largest endocrine organ and also a metabolic organ, with a wide range of enzymes, hormones and neurotransmitters being expressed in the intestinal epithelium, and thus it can interact with host cells and contribute to the regulation of metabolism (Rastelli et al., 2019). Besides genetics, the diets as well as host hormones or neurotransmitters, can affect gut microbial diversity of the hosts (Koren et al., 2012). Studies in humans showed that hyperthyroid patients had significant enrichments in pathogenic bacteria such as Enterobacteriaceae and Clostridium in the gut in comparison with a healthy group (Zhou et al., 2014), while hypothyroidism was accompanied by a modest intestinal bacterial overgrowth (Lauritano et al., 2007; Zhao et al., 2018). Both rat (Shin et al., 2020) and gerbil models (in the present study) of thyroid dysfunction revealed that profile of gut microbiota varied with different thyroid functional status. We also observed that the hyperthyroid gerbils showed an increase in pathogenic bacteria, such as Rikenella, Helicobacter, Staphylococcus and Veillonella, which have been reported as causes of serious infections and inflammation (Alkadhi et al., 2014). Moreover, the hyperthyroid gerbils had a lower abundance in beneficial SCFA‐producing bacteria such as Butyricimonas and Parabacteroides. Furthermore, the correlation analyses support the concept that these specific bacteria such as Allobaculum, Coprococcus, Odoribacter, Oscillospira, Parabacteroid, Prevotella, Rikenella and Ruminococcus were involved in the regulation of host RMR and food intake. These data indicate that thyroid hormones shift the gut microbiota, which may affect host thermogenesis.

The alteration of the gut microbial community leads to changes in SCFAs, bacterial fermentation products. Bile acid metabolism is also related to gut microbiota, which via bile salt hydrolases enzymes deconjugate the glycine or taurine from the sterol core of the primary bile acids. And subsequently, it undergoes a variety of microbiota‐encoded transformations like 7alpha/beta‐dehydroxylation that generate secondary bile acids, which can have a wide effects on the host metabolism (Ridlon et al., 2006; Yao et al., 2018; Foley et al., 2019). Bile acids have a critical role in regulating lipid metabolism and adaptive thermogenesis (Watanabe et al., 2006; Worthmann et al., 2017). Both SCFAs and bile acids exert systemic impacts on host appetite, lipid metabolism and inflammation (Smith et al., 2013; Foley et al., 2019). We observed decreases both in SCFAs and total bile acids in the hyperthyroid animals, which may lead to increased feeding and diet‐induced thermogenesis. In support of our results, a study by Ellis showed that treatment of hepatocytes with T3 decreased total bile acid formation to 44% of controls (Ellis, 2006), and another study showed that the decrease in thyroid hormones was closely related to the increase in plasma bile acid levels (Ockenga et al., 2012). The decreases seen in total SCFAs and bile acids would also promote the expression of virulent genes in the bacteria, and affect the diversity and structure of the gut microbiota (Worthmann et al., 2017; Lin et al., 2018). Consequently, the changes in gut microbiota and bacterial metabolites would influence host metabolism and thermal homeostasis. The thyroid‐microbiota linkage will be further verified in non‐orally‐manipulated models, since the administration of thyroxine and/or methimazole orally may cause intestinal microbiota to interact artificially with these components even before absorption.

Normal microbiota attenuates hyperthyroid‐induced thermogenesis of the host

Gut microbiota‐host interaction is crucial in maintaining hormonal and metabolic homeostasis. The hyperthyroid gerbils still had elevated serum T3 and T4 levels in the 6th week after termination of the T4 treatment. We measured the enzyme DIO2, which can mediate T4‐to‐T3 conversion, and found that DIO2 expression in the liver was reduced with T4 treatment and still tended to decrease even after T4 treatment stopped. The study by (Watanabe et al. (2006) also showed a hyperthyroid‐associated reduction in DIO2. The other enzyme, type 1 deiodinase (DIO1), may also contribute to changes in thyroid hormone levels, since it was upregulated in the hyperthyroid state (Worthmann et al., 2017). This indicates that T4 treatment exerted a long‐lasting effect on thyroid metabolism and the conversion of extra T4 was delayed in these animals.

We observed a greater rate of decline of both T4 and T3 in the hyperthyroid gerbils receiving microbiota from the control donors, which was accompanied by the return of both RMR and Tb to control levels. Hyper gerbils with the CMT showed an increase of DIO2 expression in the liver compared to sham CMT. Hyper‐Con gerbils had a reduced Firmicutes/Bacteroidetes ratio and increased the abundance of SCFA‐producing bacteria such as Allobaculum and Lactobacillus compared to sham CMT. The greater rate of decline of both T4 and T3 in the Hyper‐Con group demonstrates an effect of the CMT on the concentrations of both T4 and T3 and this may be what affects thermogenesis. In contrast, the CMT from hyperthyroid to control did not have an effect. One possibility is that the microbes from hyperthyroid gerbils failed to colonize control gerbils. The gut microbiome via their metabolites, SCFAs and bile acids may regulate the expression of both DIO1 and DIO2 and thus affect serum T3 and T4 levels (Ito et al., 2011; Virili and Centanni, 2017). Hyperthyroidism was associated with reduced SCFAs and bile acids, and thus reduced DIO2. It is also possible that the changes in gut microbiome can affect enterohepatic metabolism of thyroid hormones (Virili and Centanni, 2017), further contributing to the faster rate of decline in T4 and T3 in the Hyper‐Con group. Therefore, gut microbiota may regulate the metabolism of thyroid hormones by modulating the deiodinase activity and the enterohepatic circulation, and the changes in thyroid hormone levels may contribute to the changes in RMR and Tb . Besides thyroid hormones, other endogenous regulators such as catecholamines and glucocorticoids can also mediate the changes in metabolic rate and Tb (Cannon and Nedergaard, 2004; Luijten et al., 2019), and these regulators may interact to control thermogenesis in endothermic species (Ribeiro et al., 2001; Mullur et al., 2014).

Although the hyperthyroid animals receiving control microbiota could reduce metabolic rate and Tb , we observed that CMT of hyperthyroid microbiota to control recipients did not cause any changes in metabolic traits. These results support the idea that the changes in metabolism and Tb may be the consequence of interactions among a number of bacterial species and depend on both the diversity of the microbial community and their host. We did not observe any effects of CMT in hypothyroid to control recipients or control to hypothyroid recipients on metabolic traits, bacterial metabolites or the bacteria in different genera. Whether normal gut microbiota can also regulate hypothyroidism should be further investigated.

Conclusion

The current study illustrates the interaction between gut microbiota and thyroid hormones in regulating host thermogenesis. We observed that hyperthyroid‐induced increases in metabolic rate and Tb were associated with altered gut microbial composition and metabolites, such as SCFAs and bile acids. We then confirmed for the first time that normal microbiota could regain normal conditions in a hyperthyroid animal by regulation of DIO2 expression in the liver and buffer hyperthyroid‐induced thermogenesis. We demonstrate that thyroid hormones regulate thermogenesis depending on gut microbiota and CMT attenuates hyperthyroid‐induced thermogenesis. These findings emphasize the functional and evolutional consequences of the thyroid‐microbiota interface in controlling host metabolic physiology and Tb in endotherms.

Experimental procedures

Animals and housing

The breeding colonies of Mongolian gerbils were outbred and rejuvenated with wild populations from the desert grasslands of Inner Mongolia or with gerbils from other laboratories every 3 years. Female Mongolian gerbils between 6 months and one year of age were housed individually in plastic cages (30 × 15 × 20 cm) that contained sawdust bedding in the laboratory at the Institute of Zoology, the Chinese Academy of Sciences (CAS). All animals were maintained under a 16L: 8D photoperiod with lights on at 0400 h at a room temperature of 23 ± 1°C. The gerbils were fed a standard rat pellet chow (Beijing KeAo Bioscience Co., Beijing, China) and were provided with water ad libitum. All procedures on animals in this study were approved under the Animal Care and Use Committee of the Institute of Zoology, CAS.

Experimental designs

Experiment 1 examined the role of thyroid hormones in regulating metabolic rate, Tb , and gut microbiota. A total of 24 adult Mongolian gerbils were randomly divided into three groups (n = 8 per group) for 4 weeks. In the hyperthyroid (Hyper) group, the gerbils drank tap water with 0.0036% l‐thyroxine (T4, weight per volume, w/v) (Ahmed et al., 2012; Jena et al., 2013). In the control group, the animals drank only tap water. In the hypothyroid (Hypo) group, the animals drank tap water with 0.04% methimazole (w/v) to inhibit thyroid hormone synthesis (Ahmed et al., 2012). Fresh faeces were collected, frozen in liquid nitrogen and then stored at −80°C for later DNA extraction. The detailed methods for measuring RMR, NST, Tb and activity are available in the supplementary materials.

Experiment 2 examined the role of gut microbiota colonization in buffering hyperthyroid‐induced changes in metabolic phenotypes. This experiment included three processes, T4 treatment for 3 weeks, and then caecal microbiota transplantation (CMT) for 3 days and then recovery for 6 weeks (Fig. 3A). In the T4 treatment phase, 26 Mongolian gerbils were induced to become hyperthyroid with 0.0036% L‐thyroxine (T4, weight per volume, w/v) in the drinking water or drank only tap water as the Control group (n = 13 per group). Three animals in each group were donors for CMT. After 3 weeks, T4 treatment was stopped. The hyperthyroid gerbils were colonized via intragastric gavage with the microbiota from the control donors (Hyper‐Con) or sterile saline (Hyper, sham CMT) for 3 days (once every day, five gerbils per group), and then kept for 6 weeks. The control gerbils were colonized with hyperthyroid microbiota (Con‐Hyper) or sterile saline (Con, sham CMT) at the same time. Blood was collected from the infraorbital vein of animals in the last week before CMT and the first week after CMT, and also at the end of treatment (the 6th week after CMT). Fresh faeces were collected before CMT and also in the 1st and 6th weeks after CMT, frozen in liquid nitrogen and then stored at −80°C for later DNA extraction.

In experiment 3 we investigated the role of gut microbiota in buffering hypothyroid‐induced changes in metabolic phenotypes. In the treatment phase, one set of 26 adult Mongolian gerbils were induced to become hypothyroid with 0.04% methimazole (w/v) in the drinking water for 3 weeks or drank only tap water as the control (n = 13 per group) (Fig. 5A). For the subsequent 6 weeks in the CMT phase (five gerbils per group), the hypothyroid gerbils were colonized with control microbiota (Hypo‐Con) or sterile saline (Hypo) for 3 days (once every day). The control gerbils were colonized with hypothyroid microbiota (Con‐Hypo) or sterile saline (Con). Blood was collected at the same time points (before CMT, the 1st and 6th weeks after CMT) with the same method as experiment 2.

Body temperature, activity and metabolic trials

Core Tb and activity were recorded during the treatment period through transponders (G2 EMitter, to ± 0.1°C, STARR life sciences). Animals were anaesthetized with an intraperitoneal injection of pentobarbital sodium (1%) with a dose of 50 mg kg−1 and the EMitter was implanted in the abdomen of the gerbils. After surgery, the animals were allowed 10 days to recover. All receivers for collecting data were connected to a computer with the Vital View software.

An open‐flow respirometry system (TSE LabMaster, Bad Homburg, Germany) was used to determine metabolic rates, as described previously (Zhang et al., 2018). The temperature of the metabolic chamber was maintained at 30 ± 0.5°C (within the thermal neutral zone). For each individual, the average of three consecutive and minimum readings of oxygen consumption was taken as the RMR after at least 1 h acclimation in the metabolic chamber. NST was measured one day after the RMR measurement. Norepinephrine (NE), with a dosage calculated according to the equation, NE (mg kg‐1) = 6.6W−0.458 (W is body mass in gram), was subcutaneously injected to every individual gerbil to induce the maximum NST (NSTmax) (Zhang and Wang, 2007). Oxygen consumption was recorded for 60 min at 5 min intervals and the temperature was maintained at 25 ± 1°C. The three highest consecutive readings of oxygen consumption 15–20 min post‐injection were taken to calculate the maximum NST during the 1‐h measurement. The regulatory NST, (NSTreg, which is from brown adipose tissue, BAT), was calculated as NSTmax minus the RMR determined the previous day. The detailed methods are available in the supplementary materials.

DIO2 measurement by Western blot

After the experiment, the gerbils were sacrificed by CO2 asphyxiation between 0900–1100 h. Blood was collected from the superior vena cava to get the serum for measurement of thyroid hormones, and the liver was collected and frozen in liquid nitrogen. About 0.1 g of liver tissue was homogenized in RIPA buffer (10 mM Tris (pH 7.0), 158 mM NaCl, 1% TritonX‐100, 5 mM EDTA, 1 mM DTT, 1 mM PMSF (P7626‐5G; Sigma‐Aldrich Corp., St. Louis, MO, USA), and PIC (P8340, 1:1000; Sigma‐Aldrich Corp., St. Louis, MO, USA), and centrifuged at 14 000 rpm at 4°C for 10 min to collect the supernatant. Protein concentration was measured by the Folin phenol method.

Total liver protein was separated in a discontinuous sodium dodecyl sulphate – polyacrylamide gel electrophoresis (SDS‐PAGE), and was blotted to a PVDF membrane (IPVH00010; Merck Millipore Ltd., Darmstadt, Germany) (270 mA, 2 h). Membranes were blocked in 5% non‐fat dry milk and incubated with rabbit anti‐DIO2 antibody (1:3000; PAC903Ra01; Cloud‐Clone Corp., Wuhan, China) and rabbit anti‐β‐Tubulin (30302ES60, Shanghai Yisheng Biotechnology Co., Shanghai, China) overnight at 4°C. Peroxidase‐conjugated goat anti‐rabbit IgG (1:5000; 111‐035‐003; Jackson ImmunoResearch, West Grove, PA, USA) was used as the secondary antibody. Blots were quantified with Quantity One Software (BioRad Laboratories, Hercules, CA, USA, Ver.4.4.0) and analysed using Image LabTM Software (Bio‐Rad Laboratories), and normalized to β‐Tubulin (a protein marker) as relative units (RU).

DIO2 measurement by ELISA

The DIO2 level in the liver was also measured using an ELISA kit (SEC903Ra, Cloud‐Clone Corp., Wuhan, China) according to the manufacturer’s instructions. Absorbance was measured at 450 nm against a blank using an ELISA reader (RayBiotech, Canada). The minimum detectable value of the kit was 0.059 ng ml‐1. The intra‐ and inter‐assay coefficients of variation (CV) for DIO2 were 10% and 12%, respectively. Assay results were expressed as ng ml‐1.

Serum total thyroid hormones assays

Serum total T3 and T4 levels were quantified with a radioimmunoassay using RIA kits (China Institute of Atomic Energy, Beijing, China), according to the instructions as previously described (Khakisahneh et al., 2019). The lower limit of the assay when using a 50 μl sample was 0.25 ng ml‐1 for T3 and 3.96 ng ml‐1 for T4. Intra‐ and inter‐assay coefficients of variation were 2.4% and 8.8% for T3, and 4.3% and 7.6% for T4, respectively.

Serum total bile acid concentration

Serum concentrations of total bile acids were measured using Total Bile Acids Assay Kit (Colorimetric) (ab239702; Abcam, Cambridge, UK). According to the manufacturer’s instructions, the standards and the serum (25 μl) were added to a 96‐well plate and adjusted to 50 μl with ddH2O. After that, the probe mix (100 μl) was added to the wells of each sample and standard, and the plate was incubated at 37°C for 10 min. Finally, 50 μl of reaction mix was added to the wells and absorbance was measured at 405 nm in a kinetic mode at 37°C for 60 min and protected from light.

Serum TNF‐α levels

Serum TNF‐α levels were detected using an ELISA kit (SEA133Ra, Cloud‐Clone Corp., Wuhan, China) according to the manufacturer’s instructions. Absorbance was measured at 450 nm against a blank using an ELISA reader (RayBiotech, Canada). Assay results were expressed as pg ml‐1.

Short‐chain fatty acids (SCFAs)

The faecal pellets (0.2 g) were extracted with ddH2O (200 ml) and centrifuged at the speed of 13 000 rpm at 4°C for 20 min. H3PO4 (25%, 6 μl) was added to the supernatant (54 μl) with the ratio 1:9 of each sample and the supernatant was then filtered through a centrifugal filter (0.22 μm) for measurement. The identification and quantification of six SCFAs, acetate, propionate, butyrate, isobutyrate, valerate, and isovalerate were carried out by gas chromatography (GC, Agilent7890A; Agilent Technologies, Palo Alto, CA, USA) with a GC autosampler and a FID system (Zhang et al., 2018). The GC was equipped with a DB‐WAX column [Polyethylene Glycol 20000, 30 m × 0.25 mm (ID), film thickness 0.25 µm, Agilent Technologies]. Helium (> 99.999%) was used as carrier gas at a constant flow rate of 1 ml min‐1. The system was operated at a maximum temperature of 250°C. A 0.5 µl sample was injected using a splitless injection and the injection temperature was 230°C. The temperature of the oven was programmed from 60°C for 1 min then increased at 5°C min‐1 to 200°C, and then at 10°C min‐1 to 230°C, and the total running time lasted for 32 min for each sample. The FID detector was set at 300oC, H2 flow rate was 40 ml min‐1, air flow rate was 400 ml min‐1, and make‐up N2 flow rate was 25 ml min‐1. The SCFAs were identified by comparing their retention times with those of authentic reference compounds and quantified by the abundance relative to that of the standards.

CMT

In this study, the strategy for CMT was developed based on wild rodents and without use of antibiotics. Caecal contents were collected and mixed from three donors of the Control, Hyper and Hypo groups, respectively. The caecal contents (200 mg) were diluted in 2 ml sterile saline (0.9% sodium chloride), and then a 200 μl suspension was delivered by intragastric gavage to the recipient gerbils for 3 days (once every day) with a modified method (Zhang et al., 2018). For sham CMT, the recipient gerbils received an intragastric gavage of 200 μL sterile saline on the same 3 days.

16S rRNA gene amplicon sequencing analysis

Total DNA was extracted from 180 to 220 mg faecal pellets using QIAamp® fast DNA stool kit (Germany) according to the manufacturer’s protocols. The V3–V4 hypervariable regions of the 16S rRNA gene were amplified using two universal primers (341F – 805R) (Table S5) (Zhang et al., 2018). The sequencing was performed on an Illumina HiSeq 2500. The 16S sequence paired‐end data set was joined and quality was filtered using the FLASH method (Magoc and Salzberg, 2011). We performed all sequences analysis using QIIME (version 1.9.1) software suite, according to the Qiime tutorial (http://qiime.org/) with some modified methods (Caporaso et al., 2010). Chimeric sequences were removed using usearch61 with de novo models (Edgar, 2010). Sequences were clustered against the 2013 Greengenes (13_8 release) ribosomal database's 97% reference dataset. Sequences which did not match any entries in this reference were subsequently clustered into de novo operation taxonomic units (OTUs) at 97% similarity with UCLUST. Taxonomy was assigned to all OTUs using the RDP classifier within QIIME and the Greengenes reference data set. We calculated the rarefaction and rank abundance curves from OTU tables using α diversity and rank abundance scripts within the QIIME pipeline. The hierarchical clustering based on population profiles of the most common and abundant taxa was performed through UPGMA clustering (Unweighted Pair Group Method with Arithmetic Mean, also known as average linkage) on the distance matrix of OTU abundance and a Newick formatted tree was obtained using the QIIME package.

Statistical analysis

The data of RMR, NSTmax, NSTreg and food intake were analysed by one‐way or two‐way ANCOVA or by repeated measure ANCOVA with body mass as a covariate. Tb , activity, DIO2, serum hormones and others were analysed by one‐way or two‐way ANOVA. Significant group differences were further evaluated using Tukey’s post hoc tests. The software SPSS 17.0 was used for statistical analyses. Results are presented as mean ± SEM, and P < 0.05 was considered to be statistically significant.

Gut microbial diversity was analysed using α diversity and β diversity. The α diversity was evaluated by calculating the Chao1, number of observed OTUs, Shannon, Simpson’s, and PD whole tree. Comparing the differences of α diversity between groups was analysed by parametric t‐test using the t‐distribution at rarefaction depth 24 000. The β diversity was shown by PCoA based on unweighted and weighted UniFrac distances using evenly sampled OTU abundances at rarefaction depth 24 000, and statistical significances between groups were analysed by ANOSIM by 999 permutations. The LEfSe method was used to assess differences in microbial communities using an LDA score threshold of 2 (Segata et al., 2011). The significant group differences of relative abundances of specific bacteria were analysed by one‐way ANOVA followed by Tukey’s post hoc tests where required. Venn diagrams were made by jvenn (Bardou et al., 2014). The correlation between specific OTUs and metabolic traits was based on Bootstrapping (−permutations 1000) and calculated via Pearson correlation. The level of statistical significance was set at P < 0.05 (False Discovery Rate (FDR)‐corrected).

Conflict of interest

The authors disclose no conflicts.

Authors’ contributions

X.Y.Z. and D.H.W. conceived the study and design the experiments. S.K. performed most of the experiments. Z.N. performed the measurements by Western blot and ELISA kits. X.Y.Z. and S.K. analysed the data and wrote the manuscript. All authors have read and critically revised the manuscript.

Acknowledgements

We would like to thank Professors Barbara Cannon and Jan Nedergaard, Stockholm, for helpful comments during the preparation of the manuscript, and the anonymous reviewers for their constructive comments on the manuscript. We thank Dr. Qing‐Sheng Chi for his suggestions in metabolism measurements and all the members of Animal Physiological Ecology Group for discussion. We also thank Jianfeng Wang from Beijing Nebula Medical Laboratory Co., Ltd. for helps in 16S rRNA gene sequencing.

Microb. Biotechnol. (2022) 15(00), 817–831

Funding Information

This study was supported by the National Natural Science Foundation of China (grant numbers 31770440, 31772461 and 32090020); the Fourth Pilot‐Research Program for Human Spaceflight (020103); and SKH was supported by UCAS Scholarship program for international students.

Contributor Information

Xue‐Ying Zhang, Email: zhangxy@ioz.ac.cn.

De‐Hua Wang, Email: wangdh@ioz.ac.cn.

Data Availability Statement

Raw sequence data are deposited in the NCBI Sequence Read Archive under accession PRJNA557008, PRJNA557014 and PRJNA557010.

References

  1. Ahmed, O.M. , Ahmed, R.G. , El‐Gareib, A.W. , El‐Bakry, A.M. , and El‐Tawab, S.A. (2012) Effects of experimentally induced maternal hypothyroidism and hyperthyroidism on the development of rat offspring: II—The developmental pattern of neurons in relation to oxidative stress and antioxidant defense system. Int J Dev Neurosci 30: 517–537. [DOI] [PubMed] [Google Scholar]
  2. Alkadhi, S. , Kunde, D. , Cheluvappa, R. , Randall‐Demllo, S. , and Eri, R. (2014) The murine appendiceal microbiome is altered in spontaneous colitis and its pathological progression. Gut Pathogens 6: 25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bardou, P. , Mariette, J. , Escudié, F. , Djemiel, C. , and Klopp, C. (2014) jvenn: an interactive Venn diagram viewer. BMC Bioinformatics 15: 293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barrett, E. , Ross, R.P. , O'Toole, P.W. , Fitzgerald, G.F. , and Stanton, C. (2012) γ‐Aminobutyric acid production by culturable bacteria from the human intestine. J Appl Microbiol 113: 411–417. [DOI] [PubMed] [Google Scholar]
  5. Bo, T.B. , Zhang, X.Y. , Wen, J. , Deng, K. , Qin, X.W. , and Wang, D.H. (2019) The microbiota‐gut‐brain interaction in regulating host metabolic adaptation to cold in male Brandt's voles (Lasiopodomys brandtii). ISME J 13: 3037–3053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cannon, B. , and Nedergaard, J. (2004) Brown adipose tissue: function and physiological significance. Physiol Rev 84: 277–359. [DOI] [PubMed] [Google Scholar]
  7. Caporaso, J.G. , Kuczynski, J. , Stombaugh, J. , Bittinger, K. , Bushman, F.D. , Costello, E.K. , et al. (2010) QIIME allows analysis of high‐throughput community sequencing data. Nat Methods 7: 335–336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dittner, C. , Lindsund, E. , Cannon, B. , and Nedergaard, J. (2019) At thermoneutrality, acute thyroxine‐induced thermogenesis and pyrexia are independent of UCP1. Mol Metab 25: 20–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Edgar, R.C. (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. [DOI] [PubMed] [Google Scholar]
  10. Ellis, E.C. (2006) Suppression of bile acid synthesis by thyroid hormone in primary human hepatocytes. World J Gastroenterol 12: 4640. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Feng, J. , Zhao, F. , Sun, J. , Lin, B.Q. , Zhao, L. , Liu, Y. , et al. (2019) Alterations in the gut microbiota and metabolite profiles of thyroid carcinoma patients. Int J Cancer 144: 2728–2745. [DOI] [PubMed] [Google Scholar]
  12. Fetissov, S.O. (2017) Role of the gut microbiota in host appetite control: bacterial growth to animal feeding behaviour. Nat Rev Endocrinol 13: 11–25. [DOI] [PubMed] [Google Scholar]
  13. Foley, M.H. , O’Flaherty, S. , Barrangou, R. , and Theriot, C.M. (2019) Bile salt hydrolases: gatekeepers of bile acid metabolism and host‐microbiome crosstalk in the gastrointestinal tract. PLoS Pathog 15: e1007581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fröhlich, E. , and Wahl, R. (2019) Microbiota and thyroid interaction in health and disease. Trends Endocrinol Metab 30: 479–490. [DOI] [PubMed] [Google Scholar]
  15. Gentile, C.L. , and Weir, T.L. (2018) The gut microbiota at the intersection of diet and human health. Science 362: 776–780. [DOI] [PubMed] [Google Scholar]
  16. de Herder, W.W. , Hazenberg, M.P. , Pennock‐Schröder, A.M. , Oosterlaken, A.C. , Rutgers, M. , and Visser, T.J. (1989) On the enterohepatic cycle of triiodothyronine in rats; importance of the intestinal microflora. Life Sci 45: 849–856. [DOI] [PubMed] [Google Scholar]
  17. Ito, M. , Toyoda, N. , Nomura, E. , Takamura, Y. , Amino, N. , et al. (2011) Type 1 and type 2 iodothyronine deiodinases in the thyroid gland of patients with 3,5,3'‐triiodothyronine‐predominant Graves' disease. Eur J Endocrinol 164: 95–100. [DOI] [PubMed] [Google Scholar]
  18. Jena, S. , Dandapat, J. , and Chainy, G.B. (2013) Curcumin differentially regulates the expression of superoxide dismutase in cerebral cortex and cerebellum of l‐thyroxine (T4)‐induced hyperthyroid rat brain. Neurol Sci 34: 505–510. [DOI] [PubMed] [Google Scholar]
  19. Johann, K. , Cremer, A.L. , Fischer, A.W. , Heine, M. , Pensado, E.R. , Resch, J. , et al. (2019) Thyroid‐hormone‐induced browning of white adipose tissue does not contribute to thermogenesis and glucose consumption. Cell Rep 27: 3385–3400. [DOI] [PubMed] [Google Scholar]
  20. Khakisahneh, S. , Zhang, X.Y. , Nouri, Z. , Hao, S.Y. , Chi, Q.S. , and Wang, D.H. (2019) Thyroid hormones mediate metabolic rate and oxidative, anti‐oxidative balance at different temperatures in Mongolian gerbils (Meriones unguiculatus). Comp Biochem Physiol C: Toxicol Pharmacol 216: 101–109. [DOI] [PubMed] [Google Scholar]
  21. Knezevic, J. , Starchl, C. , Tmava Berisha, A. , and Amrein, K. (2020) Thyroid‐Gut‐Axis: how does the microbiota influence thyroid function? Nutrients 12: 1769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Köhling, H.L. , Plummer, S.F. , Marchesi, J.R. , Davidge, K.S. , and Ludgate, M. (2017) The microbiota and autoimmunity: their role in thyroid autoimmune diseases. Clin Immunol 183: 63–74. [DOI] [PubMed] [Google Scholar]
  23. Koren, O. , Goodrich, J.K. , Cullender, T.C. , Spor, A. , Laitinen, K. , Bäckhed, H.K. , et al. (2012) Host remodeling of the gut microbiome and metabolic changes during pregnancy. Cell 150: 470–480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lauritano, E.C. , Bilotta, A.L. , Gabrielli, M. , Scarpellini, E. , Lupascu, A. , Laginestra, A. , et al. (2007) Association between hypothyroidism and small intestinal bacterial overgrowth. J Clin Endocrinol Metab 92: 4180–4184. [DOI] [PubMed] [Google Scholar]
  25. Lenard, J. (1992) Mammalian hormones in microbial cells. Trends Biochem Sci 17: 147–150. [DOI] [PubMed] [Google Scholar]
  26. Li, X.S. , and Wang, D.H. (2005) Seasonal adjustments in body mass and thermogenesis in Mongolian gerbils (Meriones unguiculatus): the roles of short photoperiod and cold. J Comp Physiol B 175: 593–600. [DOI] [PubMed] [Google Scholar]
  27. Lin, Y.H. , Chen, Y. , Smith, T.C. , and Staels, B. (2018) Short‐chain fatty acids alter metabolic and virulence attributes of Borrelia burgdorferi . Infect Immun 86: e00217–e218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Luijten, I.H.N. , Cannon, B. , and Nedergaard, J. (2019) Glucocorticoids and brown adipose tissue: do glucocorticoids really inhibit thermogenesis? Mol Aspects Med 68: 42–59. [DOI] [PubMed] [Google Scholar]
  29. Lyte, M. (2014) Microbial endocrinology: host‐microbiota neuroendocrine interactions influencing brain and behavior. Gut Microbes 5: 381–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Magoc, T. , and Salzberg, S.L. (2011) FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 27: 2957–2963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mittag, J. (2020) More than fever – Novel concepts in the regulation of body temperature by thyroid hormones. Exp Clin Endocrinol Diab 128: 428–431. [DOI] [PubMed] [Google Scholar]
  32. Mullur, R. , Liu, Y.Y. , and Brent, G.A. (2014) Thyroid hormone regulation of metabolism. Physiol Rev 94: 355–382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nguyen, T.T. , DiStefano, J.J. 3rd , Huang, L.M. , Yamada, H. , and Cahnmann, H.J. (1993) 5'‐ and 5‐deiodinase activities in adult rat cecum and large bowel contents inhibited by intestinal microflora. Am J Physiol 265(3 Pt 1): E521–E524. [DOI] [PubMed] [Google Scholar]
  34. Ockenga, J. , Valentini, L. , Schuetz, T. , Wohlgemuth, F. , Glaeser, S. , Omar, A. , et al. (2012) Plasma bile acids are associated with energy expenditure and thyroid function in humans. J Clin Endocrinol Metab 97: 535–542. [DOI] [PubMed] [Google Scholar]
  35. Rastelli, M. , Cani, P.D. , and Knauf, C. (2019) The gut microbiome influences host endocrine functions. Endocr Rev 40: 1271–1284. [DOI] [PubMed] [Google Scholar]
  36. Ribeiro, M.O. , Carvalho, S.D. , Schultz, J.J. , Chiellini, G. , Scanlan, T.S. , Bianco, A.C. , et al. (2001) Thyroid hormone‐sympathetic interaction and adaptive thermogenesis are thyroid hormone receptor isoform‐specific. J Clin Investig 108: 97–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ridlon, J.M. , Kang, D.J. , and Hylemon, P.B. (2006) Bile salt biotransformations by human intestinal bacteria. J Lipid Res 47: 241–259. [DOI] [PubMed] [Google Scholar]
  38. Roshchina, V.V. (2010) Evolutionary considerations of neurotransmitters in microbial, plant and animal cells. In Microbial Endocrinology: Interkingdom Signaling in Infectious Disease and Health. Lyte, M. , and Freestone, P.P. (eds). New York: Springer, pp. 17–52. [Google Scholar]
  39. Segata, N. , Izard, J. , Waldron, L. , Gevers, D. , Miropolsky, L. , Garrett, W.S. , et al. (2011) Metagenomic biomarker discovery and explanation. Genome Biol 12: R60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sender, R. , Fuchs, S. , and Milo, R. (2016) Are we really vastly outnumbered? Revisiting the ratio of bacterial to host cells in humans. Cell 164: 337–340. [DOI] [PubMed] [Google Scholar]
  41. Shin, N.R. , Bose, S. , Wang, J.H. , Nam, Y.D. , Song, E.J. , Lim, D.W. , et al. (2020) Chemically or surgically induced thyroid dysfunction altered gut microbiota in rat models. FASEB J 34: 8686–8701. [DOI] [PubMed] [Google Scholar]
  42. Silva, J.E. (2003) The thermogenic effect of thyroid hormone and its clinical implications. Ann Intern Med 139: 205–213. [PubMed] [Google Scholar]
  43. Smith, P.M. , Howitt, M.R. , Panikov, N. , Michaud, M. , Gallini, C.A. , Bohlooly‐Y, M. , et al. (2013) The microbial metabolites, short‐chain fatty acids, regulate colonic Treg cell homeostasis. Science 341: 569–573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Strandwitz, P. , Kim, K.H. , Terekhova, D. , Liu, J.K. , Sharma, A. , Levering, J. , et al. (2019) GABA‐modulating bacteria of the human gut microbiota. Nat Microbiol 4: 396–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tremaroli, V. , and Backhed, F. (2012) Functional interactions between the gut microbiota and host metabolism. Nature 489: 242–249. [DOI] [PubMed] [Google Scholar]
  46. Virili, C. , and Centanni, M. (2017) “With a little help from my friends”‐the role of microbiota in thyroid hormone metabolism and enterohepatic recycling. Mol Cell Endocrinol 458: 39–43. [DOI] [PubMed] [Google Scholar]
  47. Watanabe, M. , Houten, S.M. , Mataki, C. , Christoffolete, M.A. , Kim, B.W. , Sato, H. , et al. (2006) Bile acids induce energy expenditure by promoting intracellular thyroid hormone activation. Nature 439: 484–489. [DOI] [PubMed] [Google Scholar]
  48. Wegener, M. , Wedmann, B. , Langhoff, T. , Schaffstein, J. , and Adamek, R. (1992) Effect of hyperthyroidism on the transit of a caloric solid‐liquid meal through the stomach, the small intestine, and the colon in man. J Clin Endocrinol Metab 75: 745–749. [DOI] [PubMed] [Google Scholar]
  49. Worthmann, A. , John, C. , Rühlemann, M.C. , Baguhl, M. , Heinsen, F.A. , Schaltenberg, N. , et al. (2017) Cold‐induced conversion of cholesterol to bile acids in mice shapes the gut microbiome and promotes adaptive thermogenesis. Nat Med 23: 839–849. [DOI] [PubMed] [Google Scholar]
  50. Yao, L. , Seaton, S.C. , Ndousse‐Fetter, S. , Adhikari, A.A. , DiBenedetto, N. , Mina, A.I. , et al. (2018) A selective gut bacterial bile salt hydrolase alters host metabolism. Elife 7: e37182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zhang, X.Y. , Sukhchuluun, G. , Bo, T.B. , Chi, Q.S. , Yang, J.J. , Chen, B. , et al. (2018) Huddling remodels gut microbiota to reduce energy requirements in a small mammal species during cold exposure. Microbiome 6: 103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhang, Z.Q. , and Wang, D.H. (2007) Seasonal changes in thermogenesis and body mass in wild Mongolian gerbils (Meriones unguiculatus). Comp Biochem Physiol A: Mol Integr Physiol 148: 346–353. [DOI] [PubMed] [Google Scholar]
  53. Zhao, F. , Feng, J. , Li, J. , Zhao, L. , Liu, Y. , Chen, H. , et al. (2018) Alterations of the gut microbiota in Hashimoto's thyroiditis patients. Thyroid 28: 175–186. [DOI] [PubMed] [Google Scholar]
  54. Zhao, Z.J. , and Wang, D.H. (2006) Effects of photoperiod on energy budgets and thermogenesis in Mongolian gerbils (Meriones unguiculatus). J Therm Biol 31: 323–331. [Google Scholar]
  55. Zhao, Z.J. , and Wang, D.H. (2009) Plasticity in the physiological energetics of Mongolian gerbils is associated with diet quality. Physiol Biochem Zool 82: 504–515. [DOI] [PubMed] [Google Scholar]
  56. Zhou, L. , Li, X. , Ahmed, A. , Wu, D. , Liu, L. , Qiu, J. , et al. (2014) Gut microbe analysis between hyperthyroid and healthy individuals. Curr Microbiol 69: 675–680. [DOI] [PubMed] [Google Scholar]
  57. Zmora, N. , Suez, J. , and Elinav, E. (2019) You are what you eat: diet, health and the gut microbiota. Nat Rev Gastroenterol Hepatol 16: 35–56. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Raw sequence data are deposited in the NCBI Sequence Read Archive under accession PRJNA557008, PRJNA557014 and PRJNA557010.

Articles from Microbial Biotechnology are provided here courtesy of Wiley

RESOURCES