IGSF1 Deficiency Leads to Reduced TSH Production Independent of Alterations in Thyroid Hormone Action in Male Mice

Abstract

Loss of function mutations in IGSF1/Igsf1 cause central hypothyroidism. Igsf1 knockout mice have reduced pituitary thyrotropin-releasing hormone receptor, Trhr, expression, perhaps contributing to the phenotype. Because thyroid hormones negatively regulate Trhr, we hypothesized that IGSF1 might affect thyroid hormone availability in pituitary thyrotropes. Consistent with this idea, IGSF1 coimmunoprecipitated with the thyroid hormone transporter monocarboxylate transporter 8 (MCT8) in transfected cells. This association was impaired with IGSF1 bearing patient-derived mutations. Wild-type IGSF1 did not, however, alter MCT8-mediated thyroid hormone import into heterologous cells. IGSF1 and MCT8 are both expressed in the apical membrane of the choroid plexus. However, MCT8 protein levels and localization in the choroid plexus were unaltered in Igsf1 knockout mice, ruling out a necessary chaperone function for IGSF1. MCT8 expression was low in the pituitary and was similarly unaffected in Igsf1 knockouts. We next assessed whether IGSF1 affects thyroid hormone transport or action, by MCT8 or otherwise, in vivo. To this end, we treated hypothyroid wild-type and Igsf1 knockout mice with exogenous thyroid hormones. T4 and T3 inhibited TSH release and regulated pituitary and forebrain gene expression similarly in both genotypes. Interestingly, pituitary TSH beta subunit (Tshb) expression was consistently reduced in Igsf1 knockouts relative to wild-type regardless of experimental condition, whereas Trhr was more variably affected. Although IGSF1 and MCT8 can interact in heterologous cells, the physiological relevance of their association is not clear. Nevertheless, the results suggest that IGSF1 loss can impair TSH production independently of alterations in TRHR levels or thyroid hormone action.

Loss-of-function mutations in the X-linked immunoglobulin superfamily, member 1 (IGSF1), gene are the most common cause of congenital central hypothyroidism (OMIM 300888) (1, 2), often associated with testicular enlargement and dysregulated prolactin and GH (1, 3-6). IGSF1, a type 1 transmembrane glycoprotein of unknown function, is expressed in thyrotrope, somatotrope, and lactotrope cells of the anterior pituitary gland as well as in choroid plexus, developing liver, and, in some species, the testes (2, 7, 8). As in IGSF1-deficient humans, Igsf1 knockout mice are centrally hypothyroid (2, 9). In 2 Igsf1-knockout mouse models, pituitary expression of the thyrotropin-releasing hormone (TRH) receptor (Trhr) is down-regulated (2, 9) and the mice exhibit impaired TRH-stimulated TSH or thyrotropin secretion from the pituitary (9). How loss of IGSF1 leads to impaired Trhr mRNA levels is unclear.

TRH, released from neurons in the hypothalamic paraventricular nucleus, is the principal neuropeptide through which the brain regulates TSH. TSH, in turn, stimulates thyroid hormone (TH) production by the thyroid gland (10, 11). The major THs, T4 and triiodothyronine T3, are critical for growth, development, and metabolism (11). Additionally, THs negatively regulate their own production by inhibiting TRH and TSH synthesis and secretion (10, 11).

THs are zwitterions and require facilitated transport for cellular uptake and efflux (10, 12, 13). THs primarily enter the brain via endothelial cells in the blood-brain barrier. They access circumventricular organs, such as the paraventricular nucleus, via the blood-cerebrospinal fluid barrier at the choroid plexus (14). In humans, the major TH transporter in the brain is monocarboxylate transporter 8 (MCT8; product of the SLC16A2 gene) (15, 16). MCT8 is highly expressed at the apical surface of choroid plexus epithelial cells, where it has been argued to mediate transport TH into cerebrospinal fluid (17). Hemizygous loss-of-function mutations in X-linked SLC16A2 cause Allan-Herndon-Dudley syndrome (18-20). Affected males exhibit moderate-to-severe intellectual disability, hypotonia, and dysregulation of the hypothalamic-pituitary-thyroid (HPT) axis, with elevated serum T3, low T4, and normal to elevated TSH (21, 22). Slc16a2-knockout mice exhibit similar changes in the HPT axis, but normal neural development (23-26). A second transporter, OATP1C1, partially compensates for the absence of MCT8 at the murine blood-brain barrier (27).

Members of the MCT family, such as MCT1, MCT3, and MCT4, depend on a chaperone, CD147, for proper plasma membrane trafficking and apical-basal sorting (28-31). Likewise, MCT2, and at times MCT1, interact with gp70, a CD147-like protein (32-34). Though MCT8 reportedly interacts with the TSH receptor, a bona fide MCT8 chaperone has not yet been described (35). Like MCT8, IGSF1 is expressed in the apical membrane of the choroid plexus (7). Additionally, IGSF1 and CD147 are structurally related members of the Ig superfamily (36). Based on these observations, we hypothesized that MCT8 and IGSF1 may interact and that loss of IGSF1 may alter MCT8 cellular trafficking and/or function, contributing to central hypothyroidism in IGSF1 deficiency. The data establish a physical, but not functional interaction between the proteins. The results also reveal impaired TSH production as a consistent feature of Igsf1 deficiency in mice, independent of thyroid hormone feedback.

Methods

Reagents

TRIzol Reagent (15596018), BCA protein assay kit (23227), DMEM/F12 (11320033), heat inactivated fetal calf serum (26010066), Streptavidin, Texas Red (S872), and Prolong Gold Antifade reagent with DAPI (1266174) were from ThermoFisher Scientific (Waltham, MA, USA). DMEM (319-005-CL) and fetal bovine serum (098159) were from Wisent (St-Bruno, CAN). DMEM/F12 from Life Technologies (Bleiswijk, NL) [¹²⁵I]-T3 (NEX110X100UC) and [¹²⁵I]-T4 (NEX111X100UC) were from PerkinElmer (Boston, MA, USA). The low iodine (LoI) diet supplemented with 0.15% propylthiouracil (PTU; TD.95125) was from Envigo (Indianapolis, IN, USA). Milliplex MAP mouse pituitary magnetic bead panel (MPTMAG-49K [TSH only]) was from MilliporeSigma (Oakville, CAN).

DNA Constructs

Previously described plasmids are listed in Table 1. Murine MCT8 (Slc16a2) and MCT10 (Slc16a10) were amplified from liver cDNA. MCT10 was first cloned into pcDNA3.0 using EcoRI and XhoI. Then, an internal HindIII site was removed by site-directed mutagenesis, without modifying the amino acid sequence. MCT8 and modified MCT10 were cloned into pcDNA3.0 containing a C-terminal triple HA tag with ClaI and HindIII. Murine Crym was amplified from brain cDNA and cloned into pcDNA3.0 with EcoRI and XhoI. All primer sequences are in Table 2. All plasmids were confirmed by sequencing (Genome Quebec, QC, CAN).

Table 1.

Reagents

Reagents	Source or reference	Identifier(s)	Notes
Mouse lines
Igsf1 Δex1 (Igsf1tm1Zuk)	(43)	MGI: 2671049
Igsf1 Δ312 (Igsf1em1Djb)	(9)	MGI: 5779502
Slc16a2 -/y (Slc16a2tm1Sref)	(23)	MGI: 3687949
DNA constructs
pcDNA3.0	Invitrogen (Burlington, CAN)
Murine IGSF1WT	(9)
Human myc-IGSF1WT-HA	(2)
Human myc-IGSF1A713_K721-HA	(2)
Human myc-IGSF1S863F-HA	(2)
Human myc-IGSF1V985A-HA	(3)
Human MCT8	(44)
Human CRYM	(38)
Cell lines
HEK293T	Dr. Terry Hébert, McGill University	CRL-3216
HeLa	Dr. Jason Tanny, McGill University	CCL-2
COS-1	Sigma-Aldrich (Amsterdam, NL)	CVCL_0223
Antibodies
Anti-β-actin (monoclonal mouse)	MilliporeSigma	A5441 RRID:AB_476744	IB (1:10,000)
Anti-GFP (goat polyclonal)	Rockland (Limerick, USA)	600-101-215 RRID:AB_218182	IB (1:1,000)
Anti-HA (mouse monoclonal)	MilliporeSigma	H9658 RRID:AB_260092	IB (1:40,000)
Anti-IGSF1 (rabbit polyclonal)	(45)	RRID:AB_2631165	IB (1:1,000) IF (1:500)
Anti-MCT8 (rabbit polyclonal)	(46)	RRID:AB_2893076	IB (1:1,000) IF (1:500)
Goat anti-rabbit, Alexa Fluor 488	ThermoFisher Scientific	A-11008 RRID:AB_143165	IF (1:500)
Horse anti-goat, Biotinylated	Vector Laboratories (Burlingame, USA)	BA-9500 RRID:AB_2336123	IF (1:150)

Table 2.

Primers

Cloning and mutagenesis primers
Crym amplification Fw	CGGAATTCAAGGCAGGCGGCGAGATGAAGC
Crym amplification Rv	CCGCTCGAGCGGTTCCTTCAACTCACTTGC
Slc16a2 amplification outer Fw	ATATAAGCTTGCCGCGATGGCGCTGCCA
Slc16a2 amplification outer Rv	GCCGATCGATAATGGGCTCTTCAGGTGT
Slc16a10 amplification outer Fw	GCAAGCTTTCTCGGTGCCCGA
Slc16a10 amplification outer Rv	GCAAATCCCGTGAGGTGAGGT
Slc16a10 amplification inner Fw	CGGAATTCTCGGGACATGGTGCCGTCCCA
Slc16a10 amplification inner Rv	CCGCTCGAGCGGCATTAAATAATCGAGGCGGAG
Slc16a10 mutagenesis Fw	CTTCTCCAGGAGAAAGCTCAGTCCTCCAAAAAAAGTC
Slc16a10 mutagenesis Rv	GACTTTTTTTGGAGGACTGAGCTTTCTCCTGGAGAAG
Mutated Slc16a10 amplification Fw	CGGAAGCTTTCGGGACATGGTGCCGTCCCA
Mutated Slc16a10 amplification Rv	CCATCGATAATAATCGAGGCGGAGTC
Quantitative PCR primers
Aldh1a1 Fw	CAGGCTGGGCTGACAAGATT
Aldh1a1 Rv	CCAAATGAACATGAGCATTGGAAA
Cga Fw	TCCCTCAAAAAGTCCAQGAGC
Cga Rv	GAAGAGAATGAAGAATATGCAG
Dio2 Fw	TCTGAGCCGCTCCAAGTC
Dio2 Rv	GGAGCATCTTCACCCAGTTT
Hr Fw	TCCCTGGTATCGAGCACAGA
Hr Rv	CTCCAAGGTTCCTGCTCCAG
Igsf1 Fw	TGAGTTGGGTCAAGAGGATT
Igsf1 Rv	TGAGGAGTTACCAGGATAGAGGA
Rpl19 Fw	CGGGAATCCAAGAAGATTGA
Rpl19 Rv	TTCAGCTTGTGGATGTGCTC
Trhr Fw	CTCCCCAACATAACCGACAG
Trhr Rv	GCAGAGAAACTGGGCTTTGA
Tshb Fw	GAACGGTGGAAATACCAGGA
Tshb Rv	AGAAAGACTGCGGCTTGGTGCA

Cell Culture

Cells were cultured at 37°C in a humidified incubator with 5% CO₂. Human embryonic kidney (HEK) 293T cells and HeLa cells were cultured in DMEM supplemented with 5% or 10% (v/v) fetal bovine serum, respectively. African green monkey kidney fibroblast (COS-1) cells were cultured in DMEM/F12 supplemented with 9% fetal calf serum and 2% penicillin/streptomycin.

Mice and Tissue Preparation

Mouse strains are listed in Table 1. Animals were housed on a 12:12 lights on/lights off cycle with ad libitum access to food and water. Adult animals and pregnant dams were anesthetized with isoflurane and euthanized by CO₂ asphyxiation. Embryonic day 18.5 (E18.5) embryos were decapitated. All animal work was conducted in accordance with federal and institutional guidelines and with the approval of the Facility Animal Care Committee at McGill University (protocol no. 5204).

Brains, choroid plexus, and pituitary glands were dissected from 8-, 12-, or 14-week-old Igsf1^+/y and Igsf1^Δex1/y or Igsf1^Δ312/y males, snap-frozen in liquid nitrogen, and stored at -80°C.

E18.5 embryo heads were fixed overnight in 4% (w/v) paraformaldehyde in PBS at 4°C and embedded in paraffin.

Immunoprecipitation, Deglycosylation, and Immunoblotting

HEK293T cells were seeded at 600 000 cells/well in a 6-well plate. The following day, cells were transfected with a total of 2 µg per well of the indicated plasmids using polyethylenimine at a concentration of 1:3 total DNA:polyethylenimine. Immunoprecipitations were performed as previously described (5).

Choroid plexus and pituitaries were homogenized in RIPA buffer (150 mM NaCl, 50 mM sodium fluoride, 10 mM NaPO₄, 2 mM EDTA, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS), and cleared by centrifugation. Supernatants were collected and protein concentration determined by BCA assay. Where indicated, 50 µg of protein lysate was deglycosylated as previously described (5). In all other cases, 15 µg total protein was denatured at 70°C for 15 minutes. Samples were immunoblotted as previously described (37) using primary antibodies at the concentrations indicated in Table 1.

Immunofluorescence

E18.5 embryo heads were sectioned and stained as previously described (2). Frozen 8-week-old brains were embedded in optimal cutting temperature (OCT) and sectioned at 10 µm. Sections were fixed with 4% paraformaldehyde for 15 minutes, washed with PBS, and blocked with 5% (v/v) goat serum diluted in PBS for 1 hour at room temperature. Next, sections were incubated with primary antibody overnight at 4°C, washed with PBS, incubated with goat anti-rabbit Alexa Fluor 488 1 hour at room temperature, then washed with PBS. Coverslips were mounted using Prolong Gold Antifade with DAPI. All antibodies were diluted in blocking buffer at the concentrations indicated in Table 1. Confocal images were captured with a Leica SP8 microscope and HC PL APO CS3 63X/1.4 numerical aperture oil objective. For Alexa 488, the HPD laser was set at 488 nm, and emission detection at 503 to 575 nm, with a numerical aperture of 1.40, a 1.0 zoom, and a pinhole of 600 μm. DAPI excitation was performed using a multiphoton laser, wavelength at 780 nm and emission was detected at 407 to 569 nm. Images were taken sequentially and overlayed using LAS AF Lite.

TH Uptake Assays

HeLa cells were seeded at 100 000 cells/well in a 12-well plate, and 24 hours later transfected with a combination of pcDNA3.0, µ-crystallin (CRYM), human myc-wild-type murine IGSF1 (IGSF1_WT)-HA, and/or GFP-MCT8 for a total of 3 μg DNA/well. For assays with IGSF1 mutants, COS1 cells were seeded in a 24-well plate and transfected with a combination of pcDNA3.0, CRYM, MCT8, and IGSF1_WT or mutants, for a total of 200 ng/well. In both cases, 24 hours after transfection, cells were equilibrated with PBS with 0.1% (w/v) BSA for 30 minutes at 37°C, then treated with 1 nM of either [¹²⁵I]-T3 or [¹²⁵I]-T4 for 15 or 30 minutes, washed, and then lysed in 0.1 M NaOH. Radioactivity in the cell lysate was measured with a γ-counter and normalized to the total [¹²⁵I] input.

Diet-induced Hypothyroidism and TH Replacement

Eight-week-old males were fed a LoI/PTU diet ad libitum. Blood was collected by submandibular venipuncture before and after 3 weeks on the diet. Next, mice were injected IP daily with either 0.002% (w/v) BSA in PBS (vehicle), T3 (5 ng/g body weight [BW]) or T4 (20, 100, or 200 ng/g BW) following 1 of 2 paradigms. In the first, animals were injected daily with T4 for 3 weeks, increasing the dose every 7 days. Before increasing the dose, blood was collected via submandibular venipuncture. In the second paradigm, animals were injected for 4 consecutive days with a single dose of T3 or T4. Eight hours after the last injection, mice were euthanized, blood collected via cardiac puncture, and tissues collected as described previously. For all collections, blood was coagulated at room temperature for 10 minutes, centrifuged for 10 minutes at 800g, and serum collected. Serum TSH was measured with a Milliplex assay (detection range: 12.2 to 50 000 pg/mL) following manufacturer’s instructions. The intra-assay coefficient of variation was 2.2% and the inter-assay coefficient of variation was 14.0%.

RNA Extraction and Reverse Transcription-quantitative PCR

RNA was extracted from tissues using TRIzol Reagent following manufacturer’s protocol. Two-hundred nanograms of total RNA was reverse transcribed as described (37). The resulting cDNA was used for quantitative PCR as described in (5) using primers in Table 2. mRNA levels were determined using the 2^-ΔΔCT method, and gene expression normalized to ribosomal protein L19 (Rpl19).

Statistical Analysis

Tissue gene expression comparing two genotypes was analyzed by 2-tailed unpaired t tests with Welch’s correction. All other data were analyzed by 2-way ANOVA, followed by a multiple comparisons test. TH uptake data were followed by Dunnett multiple comparisons test, serum TSH data comparing genotypes before and after the LoI/PTU diet, were followed by Šídák multiple comparisons test, and Tukey multiple comparisons test was used in all other cases. Statistical analyses were performed using Prism 9, GraphPad software. P < 0.05 was considered statistically significant.

Results

IGSF1 and MCT8 Interact In Vitro

Because both IGSF1 and MCT8 are expressed at the apical membrane of the choroid plexus (7) and both are involved in the regulation of the HPT axis, we hypothesized that the 2 proteins might interact. When coexpressed in heterologous HEK293T cells, IGSF1_WT coimmunoprecipitated with murine MCT8 (Fig. 1A, top panel, lane 5), but not MCT10 (Fig. 1A, top panel, lane 6), which shares 49% sequence identity with MCT8 (38). We attempted to examine the interaction between IGSF1 and OATP1C1, but we had difficulty expressing the latter protein in heterologous cells (data not shown). Next, we assessed human MCT8 interactions with wild-type and patient-derived mutant forms of IGSF1. As previously reported, IGSF1_{A713_K721} and IGSF1_S863F exhibited markedly impaired plasma membrane trafficking (only the immature glycosylated form [lower band] is expressed), whereas IGSF1_G661R trafficked relative normally (2, 3). MCT8’s interaction with IGSF1_{A713_K721} and IGSF1_S863F was reduced relative to IGSF1_G661R and IGSF1_WT (Fig. 1B, top panel, compare lanes 3-6).

Figure 1.

IGSF1 interacts with MCT8 but not MCT10. (A) HEK293T cells were co-transfected with untagged murine IGSF1_WT and either murine MCT8-HA or MCT10-HA or (B) human GFP-MCT8 and human IGSF1_WT-HA, IGSF1_{A713_K721}-HA, IGSF1_G661R-HA, or IGSF1_S863F-HA. Protein lysates were immunoprecipitated against the HA epitope. Total and precipitated lysates were immunoblotted (IB) with antibodies against HA and IGSF1 or GFP.

IGSF1 Does Not Alter MCT8-Mediated Influx of T3 and T4 In Vitro

To determine the potential functional relevance of the IGSF1-MCT8 interaction, we performed in vitro T3 and T4 uptake assays. CRYM, an intracellular TH-binding protein, was included to prevent TH efflux from cells. MCT8 overexpression augmented uptake of both [¹²⁵I]-T3 (Fig. 2A) and [¹²⁵I]-T4 (Fig. 2B) in HeLa cells. IGSF1, when transfected alone or with MCT8, did not affect TH uptake. In COS-1 cells, [¹²⁵I]-T3 uptake via MCT8 was similarly unaffected by coexpression of IGSF1_WT or IGSF1 mutants (IGSF1_{A713_K721}, IGSF1_S863F, or IGSF1_V985A; Fig. 2C).

Figure 2.

IGSF1 does not affect MCT8 transport of radiolabeled iodothyronines in vitro. HeLa cells were transfected with expression vectors for the indicated combinations of CRYM, MCT8, and IGSF1, and treated with either 1 nM [¹²⁵I]-T3 (A) or [¹²⁵I]-T4 (B) for 0, 15, and 30 minutes. Data represent mean % of the input (± SEM) from 2-3 independent experiments. (C) COS-1 cells were transfected with expression vectors for pcDNA3.0, IGSF1_WT, IGSF1_{A713_K721}, IGSF1_S863F, or IGSF1_V985A and the indicated combination of MCT8 and CRYM and treated with 1 nM [¹²⁵I]-T3 for either 10 (pcDNA3 and MCT8) or 30 (CRYM and MCT8 + CRYM) minutes. For all assays, intracellular counts per minute were measured following treatment and normalized to the input amount. Data were analyzed by 2-way ANOVA followed by Dunnett multiple comparisons test. **P < 0.01; ***P < 0.001; ****P < 0.0001 when comparing all conditions to CRYM (A-B) or pcDNA3.0 (C) control.

IGSF1 Is Not Required for MCT8 Expression and Membrane Localization In Vivo

Because we hypothesized that IGSF1 might act as an MCT8 chaperone, we first confirmed IGSF1 expression at the apical membrane of the choroid plexus in wild-type (Igsf1^+/y) mice by immunofluorescence (Fig. 3A). As revealed by immunoblot, only the mature glycoform (EndoH-insensitive) was observed in choroid plexus of wild-type mice (Fig. 3B, lanes 1-3). For comparison, mature and immature (EndoH-sensitive) IGSF1 glycoforms were detected in protein lysates from IGSF1_WT transfected cells (Fig. 3B, lanes 4-6). Next, we examined MCT8 expression in choroid plexus and pituitary protein lysates from Igsf1^+/y (wild-type; +) and Igsf1^Δex1/y (knockout; Δ) mice by immunoblot (Fig. 3C). MCT8 was expressed in the choroid plexus (Fig. 3C, lanes 1-7) and, to a lesser extent, in the pituitary (Fig. 3C, lanes 8-14, see long exposure), with no clear difference between genotypes. As shown by immunofluorescence, MCT8 was expressed at the apical membrane of the choroid plexus in both Igsf1^+/y (Fig. 3D, top row) and Igsf1^Δex1/y (Fig. 3D, middle row) mice. MCT8 immunoreactivity was not detected in Slc16a2^-/y (MCT8 knockout) mice (Fig. 3D, bottom row), confirming the specificity of the signal.

Figure 3.

MCT8 protein expression and localization in the choroid plexus is not altered in Igsf1 knockout mice. (A) Immunofluorescence for IGSF1 (green) in choroid plexus of Igsf1^+/y (wild-type) and Igsf1^Δex1/y (knockout) E18.5 embryos. (B) Protein lysates from pooled choroid plexus of three 12-week-old Igsf1^+/y males or from HEK293T cells transfected with murine IGSF1_WT were deglycosylated with EndoH or PNGaseF and immunoblotted against IGSF1 (top panel). Arrows indicate mature and immature glycoforms of IGSF1 as well as the deglycosylated protein. (C) Protein lysates from the choroid plexus and pituitary of 12-week-old Igsf1^+/y (+) and Igsf1^Δex1/y (Δ) males were immunoblotted using antibodies against IGSF1 (top panel) or MCT8 (middle panels). MCT8 blots were subjected to both short and long exposures. β-actin was used as the loading control (B-C; bottom panels). (D) Immunofluorescence for MCT8 (left column; green in right column) on coronal brain sections from 8-week-old male Igsf1^+/y (top row), Igsf1^Δex1/y (middle row), or 10-week-old Slc16a1^-/y (bottom row) mice. DAPI (middle column; blue in right column) was used to stain DNA. Scale bar: 50 μm.

Loss of IGSF1 Does Not Alter T4 Negative Feedback to the Brain or Pituitary

Because the previous analyses failed to establish a role for IGSF1 in mediating MCT8 expression or function, we next asked whether IGSF1 deficiency might cause alterations in TH action in the pituitary or forebrain as mediated via other transporters or other mechanisms. We first inhibited endogenous TH production by placing adult male wild-type and Igsf1^Δ312/y mice on a LoI/PTU diet (Fig. 4A). As expected, TSH levels increased markedly in both genotypes, but to a significantly lesser extent in Igsf1^Δ312/y mice (Fig. 4B), as previously described (9). We next treated these mice with exogenous T4, increasing the dose weekly (Fig. 4A). After a week on the lowest dose (20 ng/g BW), TSH levels decreased significantly in both genotypes. The relative decline was modestly, but significantly greater in Igsf1^Δ312/y compared with wild-type mice (Fig. 4C). The genotype difference was not observed at the higher T4 concentrations (100 and 200 ng/g BW) (Fig. 4C). Note that the data in Fig. 4C were normalized to facilitate a direct comparison between genotypes.

Figure 4.

Igsf1 knockout mice exhibit normal sensitivity to exogenous T4. (A) Eight-week-old Igsf1^+/y and Igsf1^Δ312/y mice were subjected to a LoI/PTU diet. After 3 weeks (21 days) on the diet, mice remained on the diet and were injected daily with T4 for 3 weeks. The dose of T4 was increased weekly, starting with 20 ng/g, followed by 100 and 200 ng/g body weight. Blood was collected before starting the diet (day 1) and at the beginning of each week once injections commenced (days 22, 29, and 36). Eight hours following the final injection on day 43, blood was collected and pituitaries and brains were dissected. (B) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice before the diet (day 1) and after 21 days on the diet (day 22). (C) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice over the course of the experiment. TSH levels were normalized relative to the highest value for each individual mouse. (D) Pituitary Tshb, Cga, Trhr, Dio2, and Igsf1 mRNA expression on day 43. (E) Forebrain Hr, Aldh1a1, and Dio2 mRNA expression on day 43. Each dot on the graphs represents an individual mouse. Data in panels B and C were analyzed by 2-way ANOVA followed by either Šídák (B) or Tukey (C) multiple comparisons test. Data in panels D and E were analyzed by 2-tailed unpaired t test with Welch correction. ***P < 0.001; ****P < 0.0001 when comparing between genotypes.

TSH is a heterodimer of the TSHβ (product of the Tshb gene) and glycoprotein hormone α (CGA) subunits. At the end of the experiment (after exposure to the highest T4 concentration), pituitary Tshb expression was significantly lower in Igsf1^Δ312/y relative to wild-type mice. In contrast, there were no significant genotype differences in expression of Cga, Trhr, or Dio2 (encoding the type II iodothyronine deiodinase) (Fig. 4D), though Trhr trended lower in knockouts. There were also no genotype differences in forebrain expression of TH sensitive genes Hr, Aldh1a1, or Dio2 (Fig. 4E).

Wild-type and Igsf1 Knockout Mice Respond Similarly to Low-dose T4

In the previous analysis, the only apparent genotype difference in the serum TSH response to exogenous T4 was observed at the lowest concentration. In addition, given the nature of the experimental design, we were only able to assess pituitary and forebrain gene expression at the highest T4 concentration. Therefore, in a follow-up experiment, we focused specifically on low concentration effects by placing animals on the LoI/PTU diet for 3 weeks (Fig. 5A-B), followed by 4 daily injections of either vehicle or 20 ng/g BW T4 (Fig. 5A). As expected, serum TSH increased on the diet, though to a greater extent in wild-type mice (Fig. 5B). T4 inhibited TSH release; however, in contrast to the earlier experiment (Fig. 4C), the extent of inhibition was comparable between genotypes (Fig. 5C; N.B the serum TSH data were again normalized to facilitate direct comparison between genotypes). In vehicle-treated mice, pituitary expression of Tshb, Cga, Dio2, and Igsf1, but not Trhr, was lower in Igsf1^Δ312/y relative to wild-type mice (Fig. 5D-H). Low-concentration T4 decreased pituitary expression of Tshb, Cga, Trhr, and Dio2, but not Igsf1, in both genotypes (Fig. 5D-H). Tshb, Cga, and Dio2 were lower in Igsf1^Δ312/y mice following T4 treatment, but this was only statistically significant for Cga. Forebrain expression of Hr, Aldh1a1, or Dio2 did not differ between genotypes under these conditions (Fig. 5I-K). T4 significantly, but mildly, stimulated Aldh1a1 (Fig. 5J) and inhibited Dio2 but did not affect Hr expression in the forebrain of both genotypes, which did not differ from one another (Fig. 5I-K).

Figure 5.

Wild-type and Igsf1 knockout mice show similar responses to exogenous T4. (A) Eight-week-old Igsf1^+/y and Igsf1^Δ312/y mice were subjected to a LoI/PTU diet. After 3 weeks (21 days), mice remained on the diet and were injected daily with either vehicle or 20 ng/g body weight T4 for 4 days. Blood was collected before starting the diet (day 1) and before starting injections (day 22). Eight hours following the final injection on day 25, blood was collected, and pituitaries and brains were dissected. (B) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice before the diet (day 1) and after 22 days on the diet. (C) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice on days 1, 22, and 25. Data at the 3-week time point are expanded to show individual data points. TSH levels were normalized relative to the value on day 22 for each individual mouse. Pituitary (D) Tshb, (E) Cga, (F) Trhr, (G) Dio2, and (H) Igsf1 mRNA levels on day 25. Forebrain (I) Hr, (J) Aldh1a1, and (K) Dio2 mRNA levels on day 25. Each dot on the graphs represents an individual mouse. Data were analyzed by 2-way ANOVA followed by either Šídák (B) or Tukey (C-K) multiple comparisons test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Wild-type and Igsf1 Knockout Mice Respond Similarly to Low-dose T3

Finally, we repeated the analysis in Fig. 5, but examined the effects of low concentration T3 (5 ng/g BW; Fig. 6A-B) (25). As with T4, there was no genotype difference in the relative suppression of serum TSH by T3 (Fig. 6C). The effects of T3 on pituitary and forebrain gene expression mirrored those of T4, with some exceptions. T3 significantly inhibited pituitary Trhr and Dio2 expression in both genotypes (Fig. 6F-G). In contrast, Tshb expression was lower in vehicle-treated Igsf1^Δ312/y relative to wild-type mice but was not statistically significantly reduced by T3 in either genotype (Fig. 6D). Cga expression was lower in vehicle-treated Igsf1^Δ312/y than wild-type mice but was only significantly reduced by T3 in the latter (Fig. 6E). Igsf1 expression was decreased in Igsf1^Δ312/y relative to wild-type mice regardless of treatment (Fig. 6H). There was no genotype difference in forebrain Hr, Aldh1a1, or Dio2 in either treatment group (Fig. 6I-K). T3 treatment stimulated Hr and Aldh1a1, but not Dio2 expression compared with vehicle.

Figure 6.

Wild-type and Igsf1 knockout mice show similar responses to exogenous T3. (A) Eight-week-old Igsf1^+/y and Igsf1^Δ312/y mice were subjected to a LoI/PTU diet. After 3 weeks (21 days), mice remained on the diet and were injected daily with either vehicle or 5 ng/g body weight T3 for 4 days. Blood was collected before starting the diet (day 1) and before starting injections (day 22). Eight hours following the final injection on day 25, blood was collected, and pituitaries and brains were dissected. (B) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice before the diet (day 1) and after 21 days on the diet. (C) Serum TSH in Igsf1^+/y (gray) and Igsf1^Δ312/y (yellow) mice on days 1, 22, and 25. Data at the 3-week time point are expanded to show individual data points. TSH levels were normalized relative to the value on day 22 for each individual mouse. Pituitary (D) Tshb, (E) Cga, (F) Trhr, (G) Dio2, and (H) Igsf1 mRNA levels on day 25. Forebrain (I) Hr, (J) Aldh1a1, and (K) Dio2 mRNA levels on day 25. Each dot on the graphs represents an individual mouse. Data were analyzed by 2-way ANOVA followed by either Šídák (B) or Tukey (C-K) multiple comparisons test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Discussion

Here, we show that IGSF1 physically interacts with MCT8, but not MCT10, when the proteins are overexpressed in heterologous cells. We did not examine whether the proteins also interact endogenously in homologous cells. Regardless, there was no apparent effect of IGSF1 on MCT8-mediated TH (T3 or T4) uptake in transfected cells. Therefore, although the proteins may physically interact, it is not evident that this translates into functional changes for either protein. Indeed, T3 and T4 effects in the forebrain and pituitary appeared to be equivalent between wild-type and Igsf1 knockout mice, further casting doubt on a functional interdependence between MCT8 and IGSF1, at least in these tissues. Moreover, MCT8’s distribution and abundance in the choroid plexus was unaltered in Igsf1 knockout mice, suggesting that IGSF1 is not an essential MCT8 chaperone in this tissue.

We attempted to map the IGSF1 interaction domain in MCT8 using MCT8/10 chimeras but were unsuccessful (data not shown) because the chimeric proteins were expressed at low levels relative to wild-type MCT8 and MCT10, complicating interpretation of the results. However, we did establish that the IGSF1-MCT8 interaction likely occurs after the proteins traffic out of the endoplasmic reticulum. That is, missense mutations in IGSF1 often lead to the protein’s misfolding and retention in the endoplasmic reticulum (2, 3, 39, 40), and mutant forms of the protein were reduced in their interactions with MCT8. These results are also consistent with the previous interpretation that IGSF1 does not function as an MCT8 chaperone, given that the proteins likely interact at the plasma membrane.

Given that the IGSF1-MCT8 interaction is specific, at least relative to MCT10, it seems unlikely to be coincidental. It may be that IGSF1 interacts with another transporter with some similarity (either in primary sequence or conformation) to MCT8. Indeed, MCT8 is not appreciably expressed in pituitary thyrotrope cells ((41) and our data), suggesting that if IGSF1 regulates TH transport in these cells, it is via a different transporter or a different mechanism. Rather than systematically interrogating other TH transporters expressed in thyrotropes, we examined exogenous T4 and T3 effects in the pituitary and forebrain more generally. However, we did not observe many or marked differences between wild-type and Igsf1 knockout animals. These data suggest that IGSF1 does not regulate pituitary or forebrain gene expression by altering TH transport, metabolism, or receptor binding.

The results of the present study, nevertheless, shine new light on HPT axis dysfunction in IGSF1 deficiency, with a particular focus on TSH synthesis. In our initial characterization, we found that Igsf1 knockout mice on normal chow had lower pituitary Tshb expression and TSH protein content compared with their wild-type littermates (9). Serum TSH was either unchanged between genotypes or lower in knockouts, depending on the experimental cohort (9). Similarly, in all the in vivo experiments reported here, when mice were on a LoI/PTU (hypothyroid) diet and treated with either vehicle, T4, or T3, we consistently observed lower pituitary Tshb expression in knockout relative to wild-type mice.

How loss of Igsf1 results in decreased TSH production has not been resolved by these or prior studies. Indeed, we previously proposed that the impairment may derive from decreased Trhr expression in the absence of Igsf1, resulting in attenuation of TRH stimulated Tshb mRNA levels (9). Although this model may hold true, in the present work, we did not consistently observe a genotype difference in Trhr expression. Moreover, Tshb expression is not altered in Trhr knockout mice (42). Therefore, it may be that IGSF1 regulates Tshb expression independently of altering Trhr expression and/or TRH signaling. Future studies will interrogate how IGSF1 regulates Tshb transcription. The data presented here suggest that this regulation is likely to be TH independent.

Glossary

Abbreviations

BW

body weight

CRYM

µ-crystallin

E

embryonic

HEK

human embryonic kidney

HPT

hypothalamic-pituitary-thyroid

IGSF1WT

human myc-wild-type murine

IGSF1

LoI, low iodine

MCT

monocarboxylate transporter

PTU

propylthiouracil

TH

thyroid hormone

TRH

thyrotropin-releasing hormone

Funding

This work was supported by Canadian Institutes of Health Research Grant MOP-133557 and Natural Sciences and Engineering Research Council of Canada Discovery Grant 2015-05178 (D.J.B.), as well as by a Natural Sciences and Engineering Research Council of Canada Doctoral Award (E.B.).

Author Contributions

E.B., T.L.S., S.G., and D.J.B. designed the experiments. E.B., T.L.S., Y.W., X.Z., B.B., and S.G. performed experiments. E.B. and D.J.B. wrote the manuscript. All authors reviewed and approved the final manuscript.

Disclosures

The authors have nothing to disclose.

Data Availability

Original data generated and analyzed during this study are included in this published article.