PMC

Thyroid follicular epithelia are more extended in mice with Taar1 deficiency. Cryosections through thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice were analyzed by semi-automated morphometry through a Cell Profiler-based pipeline (). Epithelial extensions (EExts) were determined per thyroid mid-section (A). Note that EExts tend to be higher in taar1^-/- mice when compared to the WT controls, thus implying taar1^-/- thyrocytes to be more prismatic, possibly indicating a higher activation state in the thyroid gland upon Taar1 deficiency. The scheme in (B) represents a comparison in sizes of thyrocytes relative to follicle lumen diameters in WT and taar1^-/- mice, respectively, not drawn to scale.

Subcellular localization of Tg-processing cathepsins remains unchanged in mouse thyroid gland epithelia upon Taar1 deficiency. Cryosections through thyroid tissue obtained from C57BL/6 WT and taar1^-/- mice were stained with antibodies against cathepsin B (A,B), cathepsin D (C,D), and cathepsin L (E,F), and analyzed by confocal laser scanning microscopy. Single channel fluorescence micrographs in right panels: top cathepsin B, D, or L, as indicated, middle Draq5^TM, bottom phase contrast. Images represent data obtained from young adult mice, only. Note that subcellular localization of cathepsins was not altered upon Taar1 deficiency and was mainly confined to endo-lysosomal compartments (A–F, arrows). Scale bars represent 20 μm.

Follicle lumen areas decrease upon Taar1 deficiency. Cryosections through thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice were analyzed by semi-automated morphometry through a Cell Profiler-based pipeline () to determine the follicle lumen areas per thyroid mid-section. Note that the average luminal area, depicted as means ± SD, was smaller in taar1^-/- mice when compared to the WT controls (A) [F(1,16) = 14.421, p = 0.002]. A color-coded depiction of luminal area distribution per thyroid mid-section (B) shows considerable heterogeneity among the follicle population per thyroid section, with bigger follicles tending to localize on the thyroid lobe periphery, thus surrounding smaller, more centrally located follicles. This pattern of follicle distribution is maintained in both genotypes and similar for thyroid lobes from young and older adult mice, respectively. Scale bars represent 500 μm.

Thyroid follicles of taar1^-/- mice present a difference in Tg storage capacity. Morphological assessment of intra-luminal Tg was performed by immunolabeling thyroid cryosections from young and older adult WT C57BL/6 and taar1^-/- mice with antibodies against Tg. The homogeneous, dimmer signal denotes cross-linked Tg (A–D, asterisks), as opposed to a higher-intensity labeling of Tg, owing to more accessible Tg epitopes for antibody binding, reflect the multilayered, partially solubilized Tg (A–D, arrows). At 24.9%, the prevalence of cross-linked Tg-containing follicle lumina was highest in the young taar1^-/- thyroid tissue, as compared to 10.0% in the young WT, and 10.2% and 7.0% in the older adult WT and taar1^-/-, respectively (A–D). Proteins isolated from young or older adult WT C57BL/6 and taar1^-/- mice were separated under non-reducing conditions by SDS-PAGE followed by silver staining (E). Note that the less prominent signal of silver staining representing Tg multimers observed for older adult taar1^-/- (E), also represented in the non-significant reduction of multimeric Tg band intensities (F). Data are depicted as means ± SD.

Taar1 deficiency alters Tshr localization in the mouse thyroid glands, but does not affect Mct8 localization. Cryosections through thyroid tissue obtained from young C57BL/6 WT and taar1^-/- mice were labeled with antibodies against TSHR and Mct8 and examined by confocal laser scanning microscopy. Mct8 maintains a basolateral localization in both WT and in taar1^-/- thyroid tissue (A,B, arrows), while Tshr is more prominently localized in intracellular vesicles of taar1^-/- thyrocytes in young mice (D,F, arrowheads), as opposed to predominantly basolateral distribution in the WT (C,E, arrows). (E,F) present a close-up on thyroid epithelia of WT and in taar1^-/- mice, respectively, as merged Tshr and Mct8 images. Secondary antibody controls are shown in (G,H) as indicated. Single channel fluorescence micrographs in right panels: top Mct8, middle Tshr, bottom Draq5^TM as nuclear counter-stain. Insets represent corresponding phase contrast micrographs. Images represent data obtained from 5 to 7 months old mice. Scale bars represent 20 μm in (A,B,E,F) and 50 μm in (C,D,G,H).

Variations in protein amounts of cathepsin B, L, and D in mouse thyroid gland upon Taar1 deficiency. Proteins were isolated from thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice, and separated by SDS-PAGE, followed by immunoblotting with antibodies against cathepsins B, L, or D. Protein amounts of the proform (pro), single chain (SC) and heavy chain of two-chain mature forms (HC) of cathepsin B (A–A3) and cathepsin L (B–B3), and the proform of cathepsin D (C,C1) were analyzed by densitometry and normalized to β-tubulin. Molecular mass markers are indicated in the left margins (A–C). Protein amounts of all forms of cathepsin B were not altered in taar1^-/- thyroids in comparison to WT controls. Two chain mature form protein amounts of cathepsin L were significantly decreased in taar1^-/- thyroids in comparison to WT controls [Genotype F(1,8) = 7.646, p = 0.024; age F(1,8) = 0.874, p = 0.377; interaction F(1,8) = 0.199, p = 0.667]. Similarly, the proform of cathepsin D was reduced in older adult taar1^-/- thyroids, as compared to WT controls (C1), although not reaching the threshold of statistical significance. Data are depicted as means ± SD.

Morphometry of thyroid lobes revealing no gross alteration upon Taar1 deficiency. Cryosections through thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice (A) were analyzed by semi-automated morphometry through a Cell Profiler-based pipeline (). Average follicle area per thyroid mid-section denotes the area covered by the thyroid follicles (B). Follicle area is defined as the external edge of the follicular epithelia, i.e., the collagen IV-positive basal lamina encircling the thyroid follicle lumen. Follicle counts per thyroid mid-section are depicted in (C). Note that there were no differences in follicle areas, but changes were observed in the number of follicles per thyroid mid-section in taar1^-/- and WT genotype [F(1,16) = 0.051, p = 0.824] but a genotype independent [F(1,16) = 0.002, p = 0.966] decrease in young vs. older adult mice [F(1,16) = 5.471, p = 0.033]. Phase contrast and corresponding single channel fluorescence micrographs of Draq5^TM-counterstained nuclei of thyroid tissue are displayed in A as indicated. Scale bars represent 100 μm.

Elevated serum TSH concentrations in young taar1^-/- mice, while body weight gain and thyroid hormones concentrations remain comparable to the WT. Body weight gain was quantified between 8 and 44 weeks of age, revealing no difference between WT and taar1^-/- mice (A). Data are depicted as circles representing means – SD plotted as continuous line below for the WT, and as squares representing means + SD plotted as dotted line above for taar1^-/-. Total T₄ (B) and total T₃ (C) serum concentrations of taar1^-/- vs. WT mice were quantified by radioimmunoassay. Data showed total T₄ and T₃ serum concentrations were not altered in taar1^-/- mice in comparison to WT controls (B,C; WT r = 0.420, p = 0.083; Taar1 r = 0.370, p = 0.109; and WT r = -0.086, p = 0.726; Taar1 r = 0.023, p = 0.916, respectively). Ratios of total T₃ over total T₄ serum concentrations were not altered in taar1^-/- vs. WT mice but declined at older age in both genotypes (D,E) [WT r = -0.445, p = 0.064; Taar1 r = -0.169, p = 0.477; and Genotype F(1,34) = 0.751, p = 0.392; age F(1,34) = 8.735, p = 0.006; interaction F(1,34) = 0.060, p = 0.809]. Serum TSH was determined by ELISA, revealing an age-dependent effect of the genotype on TSH concentration in the serum [WT r = -0.037, p = 0.931; Taar1 r = -0.062, p = 0.908; and Genotype F(1,34) = 8.417, p = 0.012; age F(1,34) = 1.529, p = 0.238; interaction F(1,34) = 8.554, p = 0.012]. The TSH of young taar1^-/- mice is increased, as compared to the age-matched WT controls (F,G). Data are depicted as scatter plots including trendlines in B–D and F, and the bar graphs in E and G display means ± SD.

Protein glycosylation is reduced in older adult Taar1-deficient mouse thyroid tissue, while gross Tg degradation states are not affected. Cryosections through thyroid tissue obtained from young and older adult WT C57BL/6 and taar1^-/- mice were stained with the lectin ConA in order to determine the glycosylation status of luminal Tg (A–D, green). Nuclei were counter-stained with Draq5^TM (red signals). Merged, single channel fluorescence and corresponding phase contrast micrographs are depicted as indicated. Scale bars represent 100 μm. The fluorescence intensities of lectin staining of glycosylated tissue components of young and older adult mice of both genotypes, respectively, as indicated were determined through a Cell Profiler-based pipeline and normalized to the numbers of cells (E); data are depicted in bar charts as means ± SD. Protein lysates prepared from young or older adult WT C57BL/6 and taar1^-/- mice, as indicated, were separated by SDS-PAGE under reducing conditions on a horizontal gel, which was silver-stained. The relative positions of bands representing multi-, di-, and monomeric Tg, as well as Tg fragments of lower molecular masses, are indicated in the right margin. Note that changes in glycosylation states were prevalent in thyroid tissue from young vs. older adult taar1^-/- mice and older adult WT vs. taar1^-/- mice (cf. B with D, and C with D, respectively, and E), but did not affect the extent or pattern of Tg degradation, which was comparable between taar1^-/- and WT mice (F).

Number of cells per thyroid mid-section declines upon Taar1 deficiency, consistent with a higher cell death rate in Taar1-deficient follicles. Cryosections through thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice were analyzed by semi-automated morphometry through a Cell Profiler-based pipeline (). The number of cells is given per 1,000 μm² area of thyroid mid-section (average counts ± SD) in (A). Note that young taar1^-/- mice were characterized by fewer numbers of cells per thyroid mid-section than corresponding WT controls, while there were no differences in the numbers of thyrocytes per 1,000 μm² tissue area observed for older adult taar1^-/- and WT mice. Single channel fluorescence micrographs of Draq5^TM-stained nuclei and corresponding phase contrast micrographs of thyroid follicular epithelia from young WT and taar1^-/- mice are depicted in (B) and (B′), respectively. Note that remnants of dead cells were found to be present in follicle lumina of taar1^-/- but mostly absent from WT controls (B′, arrows). Scale bars represent 20 μm. The average percentage of dead cells (±SD) was found to be significantly higher in taar1^-/- thyroid epithelia than in WT controls [F(1,15) = 51.260, p < 0.001] (C). There is a clear trend toward a decreased cell death rate in older mice [F(1,15) = 6.809, p = 0.020], independent of the genotype [interaction term F(1,15) = 3.636, p = 0.076].

The influence of Taar1 deficiency on liberation of TH in thyroid follicles. Tg (yellow) is synthesized and secreted at the apical plasma membrane of thyrocytes for storage in covalently cross-linked form in the thyroid follicle lumen of WT (left). TSH (violet) binding to basolateral Tshr (pink) triggers retrograde trafficking of endo-lysosomal enzymes (red) for secretion at the apical pole into the peri-cellular follicle luminal space. Subsequently, Tg is solubilized extracellularly by the action of cathepsins B and L before being internalized for endo-lysosomal degradation (orange to red). TH release (yellow) from thyroid follicles is through balanced proteolytic processing of Tg by extra- and intra-cellular means of proteolysis, and subsequent Mct8-mediated translocation across the basolateral plasma membrane. This study asks whether Tg solubilization and processing to yield TH is possibly co-regulated by Taar1 (cyan), which is localized to cilia at the apical plasma membrane of thyrocytes where it can, in principle, interact with intra-follicular generated TH derivatives triggering signaling of this GPCR. Smaller thyroid follicle lumina in male taar1^-/- vs. WT mice reveal thyroglobulin storage in more compacted form (right). Enhanced luminal cystatins C and D (not depicted) render thyroglobulin-solubilizing cathepsin B (red) less active in thyroid tissue of taar1^-/- mice. Cathepsin L amounts (red) are diminished in taar1^-/- vs. WT thyroid follicles featuring more dead cell remnants (dark gray irregular shaped symbols) in the follicle lumen. More extended epithelia do not affect gross thyroid hormone (yellow polygons) release from thyroid follicles of taar1^-/- mice. TSH (violet) concentrations in the blood serum are enhanced upon Taar1 deficiency, while TSH receptors (pink) are non-canonically located in intracellular vesicles. The results indicate Taar1 is necessary to maintain canonical HPT-axis regulation of thyroid function in male mice.

Proteolytic activity of cathepsin B is reduced in older adult mouse thyroid glands upon Taar1 deficiency, consistent with an increase in cystatin levels. Proteins were isolated from thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice, and proteolytic activity of cathepsin B was determined. The graph represents cleavage of Z-Arg-Arg-AMC^∗HCl (A). Cathepsin B activity levels of each sample were obtained by calculating relative fluorescence units (RFU) per protein concentration. Note that cathepsin B activity is genotype-dependently decreased in taar1^-/- mice in comparison to WT controls, irrespective of the age [Genotype F(1,8) = 24.944, p = 0.001; age F(1,8) = 1.256, p = 0.295; interaction F(1,8) = 1.727, p = 0.225]. Cryosections through thyroid tissue obtained from young or older adult C57BL/6 WT and taar1^-/- mice, were labeled with antibodies against cystatin C or D, and analyzed by confocal laser scanning microscopy (B,C,E,F). Single channel fluorescence micrographs in right panels: top cystatin C or D as indicated, middle Draq5^TM, and bottom phase contrast. Images represent data obtained from young mice. The graph in (D) represents quantification of the fluorescence intensity of cystatin C-positive signals per cell given in gray values (GV) for taar1^-/- mice vs. WT control [Genotype F(1,8) = 11.262, p = 0.010; age F(1,8) = 3.173, p = 0.113; interaction F(1,8) = 0.353, p = 0.569]. The graph in (G) represents quantification of fluorescence intensity of cystatin D-positive signals in the lumen, depicted as percentage of total fluorescence intensity of cystatin D per cell for taar1^-/- mice vs. WT control for both ages [Genotype F(1,8) = 1.368, p = 0.276; age F(1,8) = 1.412, p = 0.269; interaction F(1,8) = 7.989, p = 0.022]. Cystatin C was mainly localized in the follicle lumen (B,C, asterisks) in both WT and taar1^-/- mice. Note that protein amounts of cystatin C were increased in taar1^-/- mice in comparison to WT controls. Cystatin D was mainly localized to the peri-cellular space of the lumen (E,F) in both WT and taar1^-/- mice (arrows), while arrowheads point toward intracellular cystatin D-positive vesicles. Scale bars represent 20 μm. Data are depicted as means ± SD. Levels of statistical significance are indicated as ^∗ for P < 0.025.