Primary Cilia Mediate TSH-Regulated Thyroglobulin Endocytic Pathways

Primary cilia are sensory organelles with a variety of receptors and channels on their membranes. Recently, primary cilia were proposed to be crucial sites for exocytosis and endocytosis of vesicles associated with endocytic control of various ciliary signaling pathways. Thyroglobulin (Tg) synthesis and Tg exocytosis/endocytosis are critical for the functions of thyroid follicular cells, where primary cilia are relatively well preserved. LRP2/megalin has been detected on the apical surface of absorptive epithelial cells, including thyrocytes. LRP2/megalin on thyrocytes serves as a Tg receptor and can mediate Tg endocytosis. In this study, we investigated the role of primary cilia in LRP2/megalin expression in thyroid gland stimulated with endogenous TSH using MMI-treated and Tg-Cre;Ift88flox/flox mice. LRP2/megalin expression in thyroid follicles was higher in MMI-treated mice than in untreated control mice. MMI-treated mice exhibited a significant increase in ciliogenesis in thyroid follicular cells relative to untreated controls. Furthermore, MMI-induced ciliogenesis accompanied increases in LRP2/megalin expression in thyroid follicular cells, in which LRP2/megalin was localized to the primary cilium. By contrast, in Tg-Cre;Ift88flox/flox mice, thyroid with defective primary cilia expressed markedly lower levels of LRP2/megalin. Serum Tg levels were elevated in MMI-treated mice and reduced in Tg-Cre;Ift88flox/flox mice. Taken together, these results indicate that defective ciliogenesis in murine thyroid follicular cells is associated with impaired LRP2/megalin expression and reduced serum Tg levels. Our results strongly suggest that primary cilia harbors LRP2/megalin, and are involved in TSH-mediated endocytosis of Tg in murine thyroid follicles.


INTRODUCTION
Thyroglobulin (Tg), the most abundant thyroid-specific protein synthesized by follicular cells (thyrocytes), serves as the molecular template for the synthesis of thyroid hormones T4 and T3 at the thyrocyte-colloid interface. A major regulatory step in thyroid hormone release in mammalian thyroid follicular cells is Tg endocytosis. This process requires micropinocytosis, which includes nonspecific fluid-phase pinocytosis and receptor-mediated endocytosis. Tg internalized by receptors is handled by post-endocytic pathways that sort Tg molecules to undergo lysosomal degradation, transcytosis, or recycling. Effective Tg endocytosis is primarily regulated by TSH and plays an important role in thyroid hormone release. In thyroid follicular cells, clathrin-coated pits, caveolaedependent endocytosis, and low-density lipoprotein receptor protein 2 (LRP2, also known as megalin) are involved in receptor-mediated endocytic pathways of Tg (1,2). LRP2/ megalin has been detected on the apical surface of thyrocytes; it serves as a Tg receptor and can mediate Tg endocytosis (1,2). Megalin knockout mice exhibit hypothyroidism, which is associated with reduced levels of serum Tg and free T4 (fT4) levels, and significantly elevated levels of serum TSH (3).
The primary cilia concentrate proteins, hormones, and ions so that they can exert their effects on the primary ciliary membrane. The ciliary pocket, a cytoplasmic invagination of the periciliary membrane, is a crucial site for exocytosis/endocytosis of vesicles for delivery and retrieval of ciliary membrane components; in addition, receptor-mediated endocytosis takes place at the ciliary pocket (4). Previously, we reported that the primary cilia of the mammalian thyroid follicles protrude from the apical surface of follicular cells toward the luminal colloid and present at the cell-colloid interface (5). In murine thyroid follicular cells, the primary cilium plays important roles in maintaining the globular follicular structure of the thyroid (6). Consequently, defective primary cilia in the murine thyroid results in irregular dilation of follicles and colloid Tg depletion (6). Interestingly, the morphological and functional alterations of the thyroid in Tg-Cre;Ift88 flox/flox mice with defective primary cilia resembled those in megalin knockout mice (3).
The primary cilium is the key machinery involved in the transduction of the sonic Hedgehog (Shh) signaling pathway. The components of the Shh pathway, smoothened (Smo), patched 1 (Ptch1), GLI family zinc finger 1 (Gli1), GLI family zinc finger 2 (Gli2), and GLI family zinc finger 3 (Gli3), exhibit dynamic movements along the primary cilium (7,8). In the central nervous system, the endocytic receptor LRP2/megalin mediates Shh signaling. LRP2/megalin forms a co-receptor complex with Ptch1 that promotes Shh binding and internalization of the Shh/Ptch1 complex and Shh pathway activation in the ciliary pocket of neuroepithelial cells (9). Therefore, we propose that induction of LRP2/megalinmediated Tg endocytosis is followed by activation of the Shh signaling pathway in primary cilia of thyroid follicular cells.
In this study, we investigated whether the primary cilium of thyroid follicular cells plays a role in Tg endocytosis. We observed the primary cilia or ciliogenesis and LRP2/megalin expression in the murine thyroid gland with endogenous TSH stimulation and high rates of Tg endocytosis. In addition, we observed LRP2/megalin expression in thyroid follicular cells of Tg-Cre;Ift88 flox/flox mice, which have no functional primary cilia due to loss of the Ift88 gene.

Mice
Mouse experiments were approved by the Institutional Animal Care and Use Committee of the Catholic Univ. of Korea Daejeon St. Mary's Hospital (approval ID, CMCDJ-AP-2019-002). Male C57BL/6J mice were purchased from DooYeol Biotech (Seoul, Korea). Mice were housed in temperature-controlled (22 ± 2°C) and light-controlled conditions (12 hours light/12 hours dark cycle, lights on at 7 am), and had free access to food and water. Twelve-week-old C57BL/6J mice were divided into two groups: Group 1 was an untreated control group; Group 2 received 0.05% methimazole (MMI, Sigma-Aldrich, 301507) in distilled drinking water for 4 weeks. MMI is used to establish hypothyroidism in experimental animals. Mouse weights were measured before the start of the experiment and after 4 weeks of MMI exposure.

Generation of Thyroid-Specific Ift88-Knockout Mice
Ift88 flox/flox mice were obtained from Dr. Kim J (Korea Advanced Institute of Science and Technology, Republic of Korea), and Thyroglobulin-Cre/+ (Tg-Cre) transgenic mice were obtained from Dr. Jukka Kero (University of Turku, Finland). The mice were maintained on the C57BL/6 genetic background. Ift88 flox/ flox mice were crossed with Tg-Cre transgenic mice to generate thyroid follicle-specific Ift88-knockout (Tg-Cre;Ift88 flox/flox ) mice. Only 35-week-old male mice were used in this study (6 Tg-Cre; Ift88 flox/flox and 6 Tg-Cre;Ift88 +/+ mice). The experiments using Tg-Cre;Ift88 floxed mice received prior approval by the Institutional Animal Care and Use Committee of the Catholic Medical Center (approval ID, CRCC-BE-CMC-17013391).

Blood Collection and Thyroid Function Test
Retro-orbital blood collected from mice was allowed to clot by leaving the sample undisturbed for 30 minutes at room temperature. The clotted blood was centrifuged at 3000g for 10 minutes. Sera were separated and stored at -80°C prior to the hormonal assay. Serum fT4 and serum Tg levels were measured using radioimmunoassay (RIA) by Dr. Kun-Ho Kim (Chungnam National University Hospital, Republic of Korea). Serum TSH was measured using a specific mouse TSH RIA provided by Dr. Cheng SY (Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA).

Thyroid Extraction
We extracted mouse thyroid using a stereo microscope (Leica EZ4). The right lobe of the dissected thyroid gland was fixed in 10% neutral buffered formalin for 24 hours at room temperature, and then the tissue was embedded in a paraffin block. Tissue slices were subjected to hematoxylin and eosin (H&E) staining and immunohistochemistry. The left lobe of the thyroid gland was stored at -80°C prior to RNA isolation.

RNA Isolation and RT-qPCR
Total RNA was extracted using TRIzol (Invitrogen). Complementary DNA (cDNA) was synthesized from the total RNA using M-MLV Reverse Transcriptase and oligo-dT primers (Invitrogen). Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was performed using QuantiTect SYBR Green PCR Master Mix (QIAGEN). Each reaction was performed in triplicate. PCR primers are presented in Supplementary Table S2. Amplification conditions were as follows: 10 minutes at 95°C for enzyme activation, followed by 40 cycles of 95°C denaturation for 10 seconds, 60°C annealing for 30 seconds, and 72°C extension for 30 seconds. The cycle threshold (Ct) values for Gapdh RNA and RNA of target genes were measured and calculated using the Life Technologies 7500 software (Life Technologies, Foster City, CA USA). The bar graph data of target genes were normalized to Gapdh mRNA levels.

Detection of Primary Cilia With Immunofluorescence Staining
Paraffin-embedded 7 mm-thick tissue sections were incubated at 56°C for 5 hours. The sections were then deparaffinized in xylene and rehydrated through a graded series of ethanol baths. Antigens were retrieved in antigen retrieval buffer (0.01 M citric acidsodium citrate, pH 6.0) by heating the sections in an autoclave at 121°C for 25 minutes. After washing, the sections were air-dried for 30 minutes and rewashed with 1× PBS. The sections were fixed with 4% paraformaldehyde in PBS for 15 minutes and then permeabilized with 0.5% Triton X-100 in PBS for 10 minutes at room temperature. Tissue sections were blocked with 5% bovine serum albumin in PBS for 30 minutes at room temperature. Thereafter, the sections were incubated with primary antibodies for 24 hours at 4°C. On the following day, the tissue-section slides were washed three times with 1× PBS and incubated at 4°C for 12 hours with secondary antibodies. Primary antibodies against acetylated a-tubulin (Cell signaling), ARL13B (ProteinTech Group), polyglutamylation modification (GT335, AdipoGen), g-tubulin (Sigma-Aldrich), and thyroglobulin (Dako) were used. Goat anti-mouse and goat anti-rabbit secondary antibodies conjugated to Alexa Fluor 488 or 568 (Invitrogen/Life Technologies) were used for indirect immunofluorescence. The stained slides were observed under a FluoView FV1000 microscope equipped with a charge-coupled device camera (Olympus Corp.). The frequency of primary cilia was determined as follows: 100 follicles of similar size were selected; follicular cells with acetylated a-tubulin-positive and g-tubulinpositive cilia within each thyroid follicle were counted; and the average number of primary cilia per one follicle was calculated.

Transmission Electron Microscopy
The mouse thyroid tissues were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer at 4°C for 4 hours. After washing in 0.1 M cacodylate, samples were post-fixed in 1% OsO 4 in cacodylate buffer (pH 7.2) containing 0.1% CaCl 2 for 1 hour at 4°C. Samples were analyzed by electron microscopy (Tecnai G2 Spirit Twin; FEI Company; Korea Basic Science Institute).

Statistical Data Analysis
Serum levels of fT4, TSH, and Tg were assessed by comparing each group to control mice by one-way ANOVA Dunnett's post hoc test. Student's t-test was used to compare mRNA levels between experimental groups versus controls. Data are presented as means ± standard deviation (SD). P-value < 0.05 was considered statistically significant.

Histopathological and Functional Analysis of Thyroid in Methimazole (MMI)-Treated Mice
To investigate the characteristics of LRP2/megalin expression and Tg endocytosis in the murine thyroid with endogenous TSH stimulation, we observed thyroid follicles and colloid Tg in control and MMI-treated mice.
Oral administration of MMI elicited a significant decrease in serum levels of fT4 (control = 1.47 ± 0.31 ng/dL; MMI = 0.63 ± 0.20 ng/dL; P = 0.032) and an increase in serum TSH (control = 34.91 ± 20.53 ng/mL; MMI = 232.98 ± 187.91 ng/mL; P = 0.00004), resulting in primary hypothyroidism ( Figure 1A and Supplementary Table S1). Serum Tg levels were 0.21 ± 0.13 ng/ mL in the control group and 0.34 ± 0.029 ng/mL in the MMItreated group, respectively. The serum Tg levels were higher in MMI-treated mice than in untreated controls ( Figure 1A). The thyroid gland and body weight were larger in the MMI-treated group than in the control group ( Figures 1B-a, C-a and Supplementary Table S1).
The thyroid of MMI-treated mice exhibited irregular-shaped follicles, enlarged follicular cells exhibiting cytoplasmic vacuoles with centrally located nuclei, and depletion of luminal colloid ( Figures 1B-b, c, d, C-b, c, d). Luminal colloid of control thyroid follicles was homogeneously stained by Tg ( Figure 1D-a, b), whereas in the MMI-treated group, luminal colloid showed reduced Tg staining density ( Figure 1E-a, b). At times, little to no luminal colloid Tg was observed in the thyroid follicles of MMI-treated mice, yet the cytoplasm of follicular cells was densely stained with Tg ( Figure 1E-b). These findings indicate that thyroid follicular cells of MMI-treated mice actively take up more colloid Tg due to elevated endogenous TSH stimulation in response to reduced concentrations of serum thyroid hormone, resulting in a smaller follicular lumen, reduced luminal colloid, and elevated cuboidal/columnar follicular cell formation.
Electron microscopic examination of thyroid glands in MMItreated mice confirmed the increase in cytoplasmic vesicles. More endocytic vesicles and electron-dense lysosomes were found under the apical plasma membrane of thyroid follicular cells in MMI-treated mice than in the control group ( Figures 1F, G).
Further, thyroid follicular cells of MMI-treated mice exhibited hypertrophy relative to the control group ( Figures 1F, G). In the thyroid follicular cells of MMI-treated mice, the rough endoplasmic reticulum (rER) exhibited distended cisternae, which were visible as cytoplasmic vacuoles under light microscopy ( Figure 1C-d). Immunohistochemistry revealed cytoplasmic accumulation of Tg in mice treated with MMI ( Figure 1E-b), but it was not prominent in control mice ( Figure 1D-b).
Collectively, these morphological changes in MMI-treated mice resulted from increased TSH-stimulated functional activity of thyroid follicular cells.

Elevated Expression of Lrp2/Megalin in MMI-Treated Mouse Thyroid
Based on the morphological and functional characteristics, we investigated whether LRP2/megalin exhibited differential expression under the two experimental conditions. To assess the effects of MMI on receptor-mediated Tg endocytosis of thyroid follicular cells, we first measured the mRNA levels of Lrp2/megalin, Clta (clathrin light chain A), Cltb (clathrin light chain B), Cltc (clathrin heavy chain), and Cav1 (caveolin-1). mRNA levels of Lrp2/megalin, Clta, and Cltb were higher in the thyroid of MMItreated mice than the control group (Figure 2A). mRNA levels of Cav1 and Cltc were elevated, but not significantly (Figure 2A).
Next, we observed the expression pattern of LRP2 in thyroid follicles in mice treated with MMI. LRP2 immunohistochemistry is restricted to the apical plasma membrane of thyroid follicular cells. However, LRP2 expression increased in both the apical plasma membrane and the cytoplasm in follicular cells of the MMI-treated group ( Figure 2B). Interestingly, Tg immunofluorescence increased in the cytoplasm of follicular cells of MMI-treated mice ( Figure 2C). These results are consistent with previous observations that Lrp2/megalin mediates Tg uptake under intense TSH stimulation, resulting in transcytosis of Tg from the colloid to the bloodstream (11).

Increased Ciliogenesis in Thyroid Follicles of MMI-Treated Mice
We next investigated whether there was an association between LRP2/megalin expression and ciliogenesis of primary cilia in vivo. To this end, we performed immunofluorescence analysis to assess changes in ciliogenesis associated with Tg endocytosis. Anti-acetylated a-tubulin and anti-g-tubulin were used as proteins of the ciliary axoneme and the basal body, respectively ( Figure 3A). The average frequencies of primary cilia in thyroid of control and MMI-treated mice were 37.80 ± 16.75% and 52.65 ± 9.57%, respectively ( Figure 3B). Thus, the thyroid follicles in MMI-treated mice exhibited a significant increase in ciliogenesis relative to control thyroids (P = 0.025). At the same time, we examined changes in the mRNA expression of genes associated with ciliogenesis. The primary cilium is a dynamic organelle that repeatedly undergoes assembly and disassembly.  Cenpj (centromere protein J) and Sept7 (septin 7), which are involved in positive regulation of primary cilia assembly, were expressed at significantly higher levels in the thyroid of MMItreated mice than in the control group ( Figure 3C). In addition, a gene encoding a negative regulator of primary cilia assembly, Dnm2 (dynamin 2), was expressed at lower levels in the thyroid of MMI-treated mice than in the control group. However, Map4 (microtubule-associated protein 4) was upregulated in the thyroid of MMI-treated mice ( Figure 3D).
To clarify the relationship between ciliogenesis and LRP2/ megalin expression in the thyroid follicles of MMI-treated mice, we evaluated LRP2 localization to primary cilia. Immunofluorescence microscopy revealed that LRP2 was localized with the basal body of primary cilia marked by g-tubulin ( Figure 3E). To further confirm the association between primary cilia and Tg endocytosis, we conducted transmission electron microscopy (TEM) in MMI-treated and untreated murine thyroid follicles. Interestingly, abundant endocytic vesicles and lysosomes were primarily located near the basal body in thyroid follicles of MMI-treated mice ( Figure 3F).
Taken together, endogenous TSH stimulation resulted in increased ciliogenesis and expression of LRP2/megalin in thyroid follicular cells.
To clarify the relationship between defective primary cilia and Lrp2/megalin expression, we compared the expression of genes related to receptor-mediated Tg endocytosis in 35-week-old Tg-Cre;Ift88 flox/flox thyroid relative to Tg-Cre;Ift88 +/+ control thyroids. First, murine thyroids with ciliary loss mediated by Ift88 deficiency exhibited significant downregulation of mRNA levels of Clta, Cltb, and Cltc relative to wild-type controls, although the difference in Cltc expression was not statistically significant ( Figure 4D). mRNA levels of caveolin-1 (Cav1) and caveolin-2 (Cav2) were lower in the thyroid follicles of Tg-Cre; Ift88 flox/flox than in those of Tg-Cre;Ift88 +/+ mice ( Figure 4D). In particular, mRNA levels of Lrp2/megalin were markedly lower in 35-week-old Tg-Cre;Ift88 flox/flox than Tg-Cre;Ift88 +/+ thyroids ( Figure 4E). Immunohistochemistry revealed that the LRP2 was present on the apical plasma membrane of control thyroid follicles, whereas LRP2 was less expressed in the thyroid follicles of Tg-Cre;Ift88 flox/flox mice ( Figure 4F).
Therefore, loss of primary cilia in murine thyroid follicles was correlated with reduced expression of Lrp2/megalin, which was also associated with irregularly dilated follicles with colloidal Tg depletion and lower levels of serum Tg.
Interaction Between LRP2/Megalin, Tg Endocytosis, and SHH Signaling in the Primary Cilium of Thyroid Follicles LRP2/megalin-mediated endocytosis controls Shh trafficking and signaling (13)(14)(15). The primary cilia, which play critical roles in signal transduction, are central organelles in the Shh signaling pathway (7,8). The primary cilium can act as a positive or negative regulator of Shh signaling (16,17). To elucidate the molecular mechanism of LRP2/megalin-mediated Tg endocytosis in primary cilia, we analyzed the expression levels of genes and proteins related to the SHH signaling pathway in the thyroid of MMItreated and untreated control mice. Although Shh mRNA levels exhibited no significant change, the mRNA levels of Smo, Ptch1, Gli1, Gli2, and Gli3 in the thyroid were higher in MMI-treated mice than in the control group ( Figure 5A). Next, we monitored the expression of SHH signaling pathways in the thyroid of MMItreated mice by immunohistochemistry and immunoblot assays for SHH, PTCH1, SMO, and GLI1. GLI1 immunohistochemistry was consistently higher in the cytoplasm and nucleus of thyroid follicular cells of MMI-treated mice relative to control, whereas SHH was barely detectable in the thyroid of MMI-treated mice and controls ( Figure 5B). Immunoblotting revealed an increase of PTCH1 and GLI1 expressions in the thyroid of MMI-treated mice ( Figure 5C), and it showed an increase in the expression of PTCH1 without a significant change in SHH ( Figure 5C). These findings indicate that Shh signaling was upregulated in the thyroid of TSH-stimulated mice. Consistent with this, the PTCH1 is detected in primary cilium of controls, whereas it is not observed in primary cilium of MMI-treated mice ( Figure 5D).

DISCUSSION
Here, we demonstrated that Lrp2/megalin is localized in the primary cilium of thyroid follicular cells. Endogenous TSH stimulation in MMI-treated mice increased Lrp2/megalin expression and Tg endocytosis. In addition, thyroid-specific cilium-deficient Tg-Cre;Ift88 flox/flox mice exhibited a significant loss of Lrp2/megalin and a reduction in serum Tg despite the high TSH level. Hence, these results may link the functional role of thyroid primary cilium with Lrp2/megalin-mediated Tg endocytosis in vivo.
LRP2/megalin plays a role in Tg uptake in thyroid follicular cells. TSH-stimulated Tg uptake is an important determinant of thyroid hormone and Tg release into the bloodstream (18)(19)(20). We found that Lrp2/megalin in Tg-Cre;Ift88 flox/flox mice exhibited a significant reduction in serum Tg levels despite having high levels of TSH. These findings indicate that ciliumdeficient thyroid follicular cells impaired Tg uptake with TSH stimulation. In this study, we showed that Tg-Cre;Ift88 flox/flox mice developed as irregularly dilated thyroid follicles with depleted colloid Tg. These follicular changes, accompanied by a reduced serum Tg and elevated serum TSH, suggest that absence of LRP2 leads to reduced thyroid hormone and Tg release.
We observed a significant increase in Tg endocytosis and serum TSH levels in the thyroid follicular cells of MMI-treated mice. In addition, both the frequency of primary cilia and LRP2/megalin expression were elevated in the thyroid of MMI-treated mice. High TSH stimulation in follicular cells of MMI-treated mice may be responsible for the elevated frequency of primary cilia. TSH is required not only for the differentiated function of thyrocytes, but also for the stimulation of cell cycle progression and proliferation in various species (26). It remains to be determined how TSH stimulation is directly linked to increased ciliogenesis and LRP2/ megalin expression of thyroid follicular cells.
In addition, our findings indicate that Shh/Gli1 signaling pathway contributes to Lrp2/megalin-mediated Tg endocytosis in primary cilia of thyroid follicular cells. The co-localization of (C) Immunoblot analysis revealed that PTCH1 and GLI1 levels were higher in MMI-treated mice than in controls. We observed no difference in the expressions of SHH and SMO. Immunoblot data are normalized against GAPDH levels in the same sample. (D) PTCH1 was localized in the primary cilium (arrow head) of thyroid follicular cells in the control group. In MMI-treated mice, loss of PTCH1 from the primary cilium (arrow head) was confirmed by immunofluorescence staining.
(E) Proposed mechanisms for the relationship between receptor-mediated Tg endocytosis and Shh signaling at the primary cilium.
Lrp2 and Tg promoted activation of Ptch1 and translocation of Gli1 to the nucleus. We speculate that Lrp2/megalin-mediated Tg endocytosis activated by TSH stimulation may promote the internalization of Ptch1 near Lrp2/megalin, alleviating Ptch1dependent inhibition of Smo activity and activating the Shh signaling pathway ( Figure 5E).
In conclusion, we have demonstrated that ciliogenesis and LRP2/megalin expression are significantly elevated in the thyroid follicular epithelium of endogenous TSH-stimulated mice. Furthermore, these TSH-stimulated mice showed that LRP2 is localized to the primary cilium. Tg-Cre;Ift88 flox/flox mice, which exhibited thyroid-specific ciliary loss, expressed dramatically lower levels of LRP2/megalin expression. Together, our results strongly suggest that LRP2/megalin-mediated endocytosis of Tg in murine thyroid follicles is regulated by ciliogenesis. This is the first in vivo study showing that the primary cilia of thyroid follicular cells are the site of LRP2/megalin-mediated Tg endocytosis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by CMCDJ-AP-2019-002.

AUTHOR CONTRIBUTIONS
JL, JYC, and MS conceived the work, designed and performed the experiments, and wrote the manuscript. HJS provided human thyroid tissue samples. Serum-free thyroxine and Tg levels were measured by K-HK. All authors contributed to the article and approved the submitted version.