Thyrotropin Receptor p.N432D Retained Variant Is Degraded Through an Alternative Lysosomal/Autophagosomal Pathway and Can Be Functionally Rescued by Chemical Chaperones

Abstract

Background:

Loss-of-function mutations of thyrotropin receptor (TSHR) are one of the main causes of congenital hypothyroidism. As for many disease-associated G-protein coupled receptors (GPCRs), these mutations often affect the correct trafficking and maturation of the receptor, thus impairing the expression on the cell surface. Several retained GPCR mutants are able to effectively bind their ligands and to transduce signals when they are forced to the cell surface by degradation inhibition or by treatment with chaperones. Despite the large number of well-characterized retained TSHR mutants, no attempts have been made for rescue. Further, little is known about TSHR degradation pathways. We hypothesize that, similar to other GPCRs, TSHR retained mutants may be at least partially functional if their maturation and membrane expression is facilitated by chaperones or degradation inhibitors.

Methods:

We performed in silico predictions of the functionality of known TSHR variants and compared the results with available in vitro data. Western blot, confocal microscopy, enzyme-linked immunosorbent assays, and dual luciferase assays were used to investigate the effects of degradation pathways inhibition and of chemical chaperone treatments on TSHR variants' maturation and functionality.

Results:

We found a high discordance rate between in silico predictions and in vitro data for retained TSHR variants, a fact indicative of a conserved potential to initiate signal transduction if these mutants were expressed on the cell surface. We show experimentally that some maturation defective TSHR mutants are able to effectively transduce Gs/cAMP signaling if their maturation and expression are enhanced by using chemical chaperones. Further, through the characterization of the intracellular retained p.N432D variant, we provide new insights on the TSHR degradation mechanism, as our results suggest that aggregation-prone mutant can be directed toward the autophagosomal pathway instead of the canonical proteasome system.

Conclusions:

Our study reveals alternative pathways for TSHR degradation. Retained TSHR variants can be functional when expressed on the cell surface membrane, thus opening the possibility of further studies on the pharmacological modulation of TSHR expression and functionality in patients in whom TSHR signaling is disrupted.

Introduction

Loss-of-function (LOF) mutations of thyrotropin receptor (TSHR) are one of the principal causes of congenital hypothyroidism (1,2). TSHR is a G-protein coupled receptor (GPCR) that is characterized by a seven-transmembrane alpha-helix B-subunit and an extracellular A-subunit, linked by disulfide bonds (3,4).

TSHR post-translational modifications are required for correct trafficking, maturation, and activity (3,5,6). The N-linked glycosylation, happening in the endoplasmic reticulum (ER), is fundamental (Fig. 1A). Acquisition of mannose-type carbohydrates permits the interactions with molecular chaperones (Fig. 1B), required for correct receptor folding, homodimerization, passage through the ER quality control system, and translocation to cis-Golgi (Fig. 1C) (3,5,7). In trans-Golgi, TSHR finally acquires the complex-type carbohydrates (Fig. 1D) that characterize the mature form expressed on the cell surface and undergoes tyrosine sulfation, which is fundamental for activation (6,8). On the plasma membrane, TSHR is cleaved by an unidentified enzyme with loss of a short sequence of variable size, peptide C (Fig. 1E). The receptor is finally composed by an extracellular A-subunit and a transmembrane B-subunit linked by disulfide bonds (3,5,9,10).

FIG. 1.

TSHR maturation and the hypothesized degradation pathways. THSR comprises a seven-transmembrane alpha helix B-subunit (green) and an extracellular A-subunit (purple). It is synthesized in the ER as a full-length unglycosylated protein [(A), immature form]. The acquisition of mannose-type carbohydrates [(B), high-mannose form] permits the interactions with molecular chaperones such as calnexin and calreticulin, thus facilitating protein folding and is required for receptor homodimerization, allowing passage through the ER quality control system and translocation to cis-Golgi (F). The TSHR maturation continues (C) as it passes to the trans-Golgi, where it acquires the complex-type carbohydrates that characterize the mature form of the protein [(D), complex carbohydrates form] that is expressed on the cell surface. The majority of mature TSHR present on the cell membrane is cleaved with loss of a small amino acid sequence of variable size, named peptide C. The receptor is thus finally composed of a transmembrane B-subunit linked by disulfide bonds to the extracellular A-subunit that can have slightly different dimensions depending on the cleavage sites and carries all the carbohydrate side chains [(E), gray discontinued line indicates that only the A-subunit is shown in the Western blot image]. The TSHR can be degraded through different systems. The misfolded receptors that are not able to proceed through the ER may be directed toward the proteasome (G) or the endolysosome (H) systems, depending on the nature of the alterations. Receptors that are not able to proceed through maturation in the Golgi can be retrotranslocated to the ER (F) for another round of control or directed toward degradation (I). The turnover of normal membrane TSHR also occurs through the endolysosme system after receptor endocytosis (I). ER, endoplasmic reticulum; TSHR, thyrotropin receptor. Color images are available online.

For many disease-associated GPCRs, including TSHR, LOF is most often due to poor cell surface expression, rather than from intrinsic deficiencies in signal transduction. The abnormal mutant conformation leads to interactions with alternative molecular chaperones (3,11,12), ER blockage, and degradation by proteasome or autophagosome (Fig. 1G, H) (13,14). Different retained GPCR mutants are able to effectively bind their ligands and transduce intracellular signals when forced to the cell surface (15 –17). The use of chemical chaperones is a well-explored area to overcome ER retention of various membrane receptors (18 –20). Currently, little is known about TSHR degradation pathways and no attempts have been made to rescue TSHR mutants.

The aim of our work was to better elucidate TSHR degradation pathways and the possibility of TSHR mutants being rescued with chemical chaperones. We concentrated our attention on two different mutants that we previously described: the TSHR p.N432D, which is retained in the ER as a high-mannose form, and the p.P449L, which is usually expressed on the plasma membrane but has impaired signaling (21). We then validate our findings in two other retained TSHR mutants (22,23). Our results show for the first time that maturation-defective TSHR mutants are able to transduce Gs/cAMP signaling when rescued by the chemical chaperone Trimethylamine-N-oxide (TMAO). Moreover, we provide new insights on the TSHR degradation mechanism, as our results suggest that aggregation-prone mutants are directed toward the autophagosomal pathway instead of the canonical proteasome system.

Materials and Methods

Patients

Patients bearing the p.N432D and p.P449L variants were previously described (21). The study had been approved by the local ethics committee (code: 05C002_2010). Informed written consent for genetic analyses had been obtained.

Chemicals

Cell culture reagents, ProLong Gold Antifade Reagent with 4′,6-diamidino-2-phenylindole, LysoTracker Red DND-99, ER-Tracker Green, Alexa-Fluor conjugated and HRP-conjugated antibodies, and Restore Western blot Stripping reagent were purchased from Thermo-Fisher (Waltham, MA). Mouse anti-Actin Ab-5 was purchased from BD (Franklin Lakes, NJ). Anti-TSHR antibodies BA8 (Cat#SC_BA8, RRID:AB_2716681), 3G4 (Cat#SC_3G4, RRID:AB_2716682), and 28.1 (Cat#SC_28.1, RRID:AB_2716683) were a gift of Dr. S. Costagliola (Free University of Brussels, Belgium) (24 –28). Anti-E-Cadherin antibody was purchased from Abcam (Cambridge, United Kingdom), and anti-VDAC antibody was purchased from Santa Cruz (Dallas, TX). bTSH, Anti-green fluorescent protein antibody, TMAO, dimethyl sulfoxide, and MTT were purchased from Sigma-Aldrich (St. Louis, MO).

In silico prediction

The TSHR variants membrane expression and functionality was assessed through the TSHR mutation database (29). Fifty-five variants were subjected to in silico predictions and assigned as damaging or not damaging as specified in Supplementary Methods.

Cell culture, transfection, treatments, and viability assay

COS-7 cells were grown in Dulbecco's modified Eagle medium (Thermo-Fisher) supplemented with 10% fetal bovine serum and penicillin–streptomycin (Sigma-Aldrich). TSHR expression vectors were previously described (21). pSVL plasmids containing wild type (WT), p.E34K, and p.R46P TSHR variants were a gift of Dr. Tonacchera (University of Pisa, Italy) (22,23).

Transfection, degradation modulation, and rescue were performed as described in Supplementary Methods. Cell viability was tested by using the MTT assay (30).

Western blotting

Cells were lysed in sodium dodecyl sulfate (SDS) buffer (62.5 mM Tris-HCl pH 6.8, 2% SDS) supplemented with protease, phosphatase, and proteasome inhibitors. Membrane preparations were obtained with Plasma Membrane Protein Extraction Kit (Abcam) following the manufacturer's instructions. Total Cellular Membranes and Plasma Membranes fractions were then processed as described in Supplementary Methods. Band intensity was quantified with ImageJ software (31).

Immunofluorescence and confocal microscopy

Samples were processed as previously described (32), and a detailed description is in Supplementary Methods. Images were acquired with EclipseTi-E inverted microscope with IMA10 × Argon-ion laser System by Melles Griot; images were acquired with CFI Plan Apo VC 60 × Oil (Nikon).

Flow cytometry

Samples were processed as previously described (32), and a detailed description is in Supplementary Methods. Measurements were performed with an FACSCalibur flow cytometer (BD Biosciences). Data were analyzed with Flowing Software 2.

Functional assays

cAMP pathway activity was assessed with Cignal CRE Reporter (luc) Kit (Qiagen, Hilden, Germany), while Gq11/IP3 pathway activity was measured with IP-One ELISA assay kit (Cisbio, Waltham, MA), following the manufacturer's instructions, as described in Supplementary Methods.

Statistical analyses

All experiments were independently repeated at least three times, as indicated in the figure legends. After normal distribution and variance similarity evaluation, two-sided unpaired t-test (eventual Welch's correction for groups with different variances), one-way analysis of variance with Bonferroni post hoc test, Kruskal-Wallis H test with Dunn's post hoc test, and Chi-square test were used as indicated in the figure legends.

For concentration–effect curves of Gs/cAMP signaling, a log(agonist) versus normalized response–variable slope equation was used for curve interpolation and parameters definition.

For confocal experiments, the degree of colocalization was quantified through Pearson's correlation coefficient, as measured with Nikon NIS Elements software. Correlation was defined as strong for a Pearson's correlation coefficient higher than 0.8, moderate when higher than 0.5 but <0.8, and weak when higher than 0.2 but <0.5. In all figures, data are shown as mean ± standard error of the mean. Data were analyzed by using GraphPad Prism 5 software, and significance was expressed as p values (*p < 0.05, **p < 0.01, ***p < 0.001).

Results

In silico prediction and in vitro data of receptor functionality are significantly discordant in retained mutants

We obtained complete information about in vitro functionality and subcellular localization of 55 LOF TSHR variants (29) and categorized them as intracellular-retained or membrane-expressed (Supplementary Table S1). These mutations were subjected to in silico predictions and assigned as functional or nonfunctional.

The comparison of in vitro and in silico data revealed a significantly higher discordance rate among the retained group (12/24 and 7/31 mutants with positive prediction but in vitro LOF for intracellular-retained and membrane-expressed, respectively, p = 0.0471) (Table 1). Such a discrepancy may indicate that some ER-retained TSHR mutants can potentially transduce signals if expressed on the cell surface. We, thus, explored the degradation mechanisms and chaperone rescue on two previously reported (21) TSHR LOF variants: the intracellular-retained p.N432D and the membrane-expressed p.P449L.

Table 1.

In Vitro and In Silico Functionality Concordance of Thyrotropin Receptor Variant with Different Subcellular Localization

		Localization
		Retained	Membrane	Total
Prediction	Concordant	12	24	36
Prediction	Discordant	12	7	19
	Total	24	31	55

Distribution of TSHR variants from in vitro data and in silico predictions concordance of function in relation to subcellular localization. For 55 different variants, data about in vitro subcellular localization and functionality were obtained through literature review. In silico prediction were obtained with six different online tools (polyphen-2, PROVEAN, SIFT, PhD-SNP, PANTHER, and SNPs&GO) and TSHR variants were then assigned as damaging (more than 50% of predictions were concordant as functionally damaging or impaired functionality) or not damaging (50% or more of predictions were concordant as neutral or benign), as described in Supplementary Methods and in Supplementary Table S1.

Each variant was then assigned to one of the four groups: membrane localization with in vitro and in silico concordance on functionality, membrane localization with in vitro and in silico discordance on functionality, intracellular retainment with in vitro and in silico concordance on functionality, and intracellular retainment with in vitro and in silico discordance on functionality.

TSHR, thyrotropin receptor.

N432D variant is arrested in the ER and forms different aggregates

We performed confocal microscopy with two different antibodies: the BA8 directed against a conformational epitope on the mature A-subunit and the 3G4 raised against a linear epitope in the C-peptide that recognizes principally immature forms (24,25).

WT TSHR and p.P449L variant had a normal membrane expression in all transfected cells (Fig. 2A), whereas p.N432D had a variable pattern detected by BA8 antibody, with three main morphologies: small intracellular aggregates (SA), perinuclear signal (PS), and cytoplasmic macroaggregate (MA) (Fig. 2B). SA and PS were the most frequent ones, while in around 10% p.N432D pattern had the characteristics of more than one morphology (mixed morphology) (Fig. 2B). In contrast, p.N432D variant staining with 3G4 antibody revealed a constant pattern of diffuse PS that was not detectable with BA8 antibody (Fig. 2B). This difference may indicate the presence of a significant amount of immature or incorrectly folded receptors recognized only by the 3G4 antibody but not by BA8 (24). Interestingly, SA were similar to the puncta that characterize misfolded GPCR mutants degraded by autophagocytosis (33,34), while MA were suggestive of perinuclear aggregates related to the proteasome degradation pathway (34,35).

FIG. 2.

p.N432D mutant is retained in the ER and lysosomes in different aggregates. (A) Representative images of WT TSHR, p.P449L mutant, and empty vector transfected cells stained with 3G4 and BA8 antibodies. (B) Representative images of the different morphological presentations of p.N432D mutant after anti-TSHR 3G4 or anti-TSHR BA8 antibodies staining. For BA8 antibody: PS, perinuclear signal; SA, small intracellular aggregates; MA, macroaggregates; MM, mixed morphology, concomitant presence of PS, SA, and/or MA detected with BA8 antibody. 3G4 antibody detected a constant mix of perinuclear signal and small aggregates. (C) Quantification of the relative frequencies of the different p.N432D patterns detected with BA8 antibody as in (B). (D) Representative images of colocalization experiment showing ER (red) and N432D mutant (green) stained with either BA8 or 3G4 antibodies. (E) Analysis of Pearson's coefficient for colocalization. The graph represents the averages of the Pearson's coefficients for colocalization detected for N432D mutant stained with either BA8 or 3G4 antibodies and ER. (F) Representative images of colocalization experiments showing late endosome/lysosomes (LYSO, red) and N432D mutant (green) stained with either BA8 or 3G4 antibodies. (G) Analysis of Pearson's coefficient for colocalization. The graph represents the averages of the Pearson's coefficients for colocalization detected for N432D mutant stained with either BA8 or 3G4 antibodies and late endosome/lysosomes. Statistical analysis: (C) n = 14 (2018 cells analyzed), (E) n = 12 (248 cells analyzed), (G) n = 6 (110 cells analyzed). Statistical significance was determined with one-way ANOVA in (C) and t-test with Welch's correction in (E) and (G). *p < 0.05 and ***p < 0.001 as indicated. ANOVA, analysis of variance; WT, wild type. Color images are available online.

Co-staining with p.N432D variant and ER or late endosome/lysosome markers showed that the majority of the protein recognized by 3G4 antibody was, indeed, localized in the ER (Fig. 2D, E). Different features of ER stress, such as vacuoles and enlarged morphology (36), were also detected in transfected cells. On the other hand, the aggregates recognized by the BA8 antibody showed a mild co-localization with endosomes/lysosomes (Fig. 2F, G).

TSHR mutants are degraded through different pathways

For many GPCRs, the ubiquitin-proteasome system is the main degradation system (37,38), while mutants prone to form aggregates are directed toward autophagic degradation (33,39). We evaluated whether that was the case by performing Western blot analysis in different conditions, with the 28.1 antibody that recognizes the full-length receptor at different stages of maturation and the cleaved A-subunit (28,40).

MG132 proteasome inhibitor induced a significant accumulation of mature WT TSHR and p.P449L variant, confirming the fundamental role of this pathway. However, only a strong accumulation of the immature form was detected for p.N432D (Fig. 3A, B). NH₄Cl, an autophagocytosis inhibitor, did not cause significant alterations in the total WT TSHR, although a change in the amount of mature form could be appreciated, as previously reported (5). On the other hand, endolysosomal inhibition induced a more effective accumulation of p.P449L and of immature p.N432D than the proteasomal inhibitor (Fig. 3A, B).

FIG. 3.

WT and mutant TSHRs are degraded through different pathways. (A) Representative Western blot images of TSHR expression and maturation after treatment with 20 mM NH₄Cl, 10 μM MG132 and a combination of the two inhibitors. TSHR was stained with 28.1 antibody, GFP was used as transfection efficiency control, and actin was used as loading control. (B) Densitometric quantification of Western blot experiments showing complex carbohydrates, high mannose, and A subunit forms of TSHR after treatment with NH₄Cl, MG132 and a combination of the two inhibitors. (C) Representative images and relative quantification of confocal microscopy experiments showing anti-TSHR BA8 antibody staining of p.N432D mutant after treatment with NH₄Cl, MG132 and a combination of the two inhibitors. For each treatment, the signal pattern with a bigger fold change increase in respect to the control is shown. White, PS; light gray, SA; intermediate gray, MA; dark gray, MM. (D) Representative Western blot images of TSHR expression and maturation after treatment with 10 mM LiCl. TSHR was stained with 28.1 antibody, GFP was used as transfection efficiency control, and actin was used as loading control. (E) Densitometric quantification of Western blot experiments showing complex carbohydrates, high mannose, and A subunit forms of TSHR after treatment with LiCl. Statistical analysis: (B) n = 4, (C) n = 14, (E) n = 3. Statistical significance was determined with one-way ANOVA (non-parametric Kruskal-Wallis H test) followed by Dunn's post hoc test. *p < 0.05, **p < 0.01, and ***p < 0.001 as indicated. GFP, green fluorescent protein. Color images are available online.

Confocal microscopy experiments revealed a significant increase in SA after autophagocytosis inhibition, while a significant increase in MA was seen after proteasome inhibition; concomitant inhibition had an intermediate effect (Fig. 3C), thus confirming the Western blot data.

Autophagocytosis activation with LiCl induced an almost complete degradation of the p.N432D variant, with milder effects on p.P449L and no effects on WT (Fig. 3D, E). Moreover, only p.N432D expression induces JNK 1/2 phosphorylation, an event linked to autophagocytosis activation (41) and significantly reduced cell viability (Supplementary Fig. S1A, B), thus confirming the role of autophagocytosis in misfolded TSHR degradation.

The chemical chaperone TMAO restores p.N432D mutant membrane expression

The p.N432D mutant does not maturate even if protein degradation is inhibited, but as immature TSHR can signal when expressed on the plasma membrane (42,43), we investigated whether this was the case with the use of different chemical chaperones.

Western blot experiments showed that, unlike in other GPCRs (44), treatment with glycerol was not effective in the p.N432D variant rescuing (Supplementary Fig. S3). Nevertheless, TMAO treatment (45) increased the maturation of all TSHR variants, but with major effects on p.N432D whose A-subunit intensity reaches levels similar to the WT control, indicating possible membrane expression; the high-mannose form had a larger increase in the WT and p.P449L TSHR (Fig. 4A, B). Hybridization with 3G4 antibody revealed that TMAO treatment caused a significant increase in a high-molecular-weight band (around 200 kDa) (Supplementary Fig. S2A, B) that has been identified as dimers of high-mannose forms (43), whose formation was fundamental for the passage through ER quality control.

FIG. 4.

TMAO restores p.N432D mutant trafficking and membrane expression. (A) Representative images of Western blotting experiments showing the expression and maturation of the TSHR variants without and after treatment with TMAO. TSHR was stained with 28.1 antibody, GFP was used as transfection efficiency control, and actin was used as loading control. (B) Densitometric quantification of Western blot experiments showing complex carbohydrates, high mannose, and A subunit forms of TSHR without and after TMAO treatment. (C) Representative flow cytometric histograms showing BA8 antibody signal in unpermeabilized cells with or without TMAO treatment. The R-1 markers indicate cells expressing TSHR and are used to quantitate receptor expression based on the mean fluorescence intensity. Untreated samples, light blue area; TMAO-treated samples, red area; overlapping area, green. (D) Mean fluorescence intensity quantification of the TSHR variants' expression on cell membrane without or after TMAO treatment, with 3G4 antibody labeling. Values are expressed as percentage of untreated WT. (E) Representative flow cytometric histograms showing 3G4 antibody signal in unpermeabilized cells with or without TMAO treatment. The R-1 markers indicate cells expressing TSHR and are used to quantitate receptor expression based on the mean fluorescence intensity. Untreated samples, light blue area; TMAO-treated samples, red area; overlapping area, green. (F) Mean fluorescence intensity quantification of the TSHR variants' expression on cell membrane without or after TMAO treatment, with 3G4 antibody labeling. Values are expressed as percentage of untreated WT. (G) Representative images of confocal microscopy experiments performed with BA8 staining, showing TSHR variants in normal conditions and after TMAO treatment. TSHR, green staining, nuclei, blue 4′,6-diamidino-2-phenylindole staining. Statistical analysis: (B) n = 3, (D) n = 6, (F) n = 6. Statistical significance was determined with one-way ANOVA followed by Dunns (D, F) or Bonferroni (B) post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 in respect to untreated TSHR WT. TMAO, trimethylamine-N-oxide. Color images are available online.

Fluorescence-activated cell sorting experiments in non-permeabilized cells confirmed cell-surface expression of TMAO-treated p.N432D. Interestingly, membrane expression resulted in 75% of the WT control with the BA8 antibody (Fig. 4C, D), and only around 50% of the WT control with the 3G4 antibody (Fig. 4E, F) with a BA8:3G4 ratio of 1.62 ± 0.25 (p < 0.05 vs. treated WT), a finding that indicated increased cleavage of the p.N432D mutant (25).

Immunofluorescence experiments with BA8 staining confirm p.N432D membrane expression after TMAO treatment. The increased intracellular staining for TMAO-treated TSHRs is in agreement with the increased immature forms detected in Western blot experiments (Fig. 4G).

Cellular membrane fractionations revealed that the A-subunit was the predominant form on the cell surface. Moreover, TMAO promoted WT TSHR translocation on the cell surface, as we detected decreased levels of all TSHR maturation forms in the total membrane extracts and increased levels in the plasma membrane extracts. This effect was not seen in the p.P449L variant, which was also insensitive to TMAO effects (Supplementary Fig. S2C, D).

Membrane expression uncovers the functional potential of p.N432D and other retained variants

The evaluation of rescued p.N432D variant signaling transduction abilities through Gs/cAMP and Gq11/IP3 pathways revealed that, indeed, the mutant was partially functional when expressed on the plasma membrane. In fact, although the Gq11/IP3 pathway remained greatly compromised (Fig. 5A), the maximal Gs/cAMP response was almost completely rescued (Fig. 5B). Concentration–effect curves (Fig. 5C and Supplementary Fig. S4A; Table 2) showed that TMAO treatment had virtually no effect on the Gs/cAMP signaling of either the WT or p.P449L mutant receptor, while the TMAO-treated p.N432D curve was right-shifted, indicating higher EC50 values.

FIG. 5.

TMAO treatment unveiled partial functionality of N432D and other known retained mutants. (A) Gq11/IP3 pathway activity after maximal dose stimulation of TSHR WT, p.N432D, and p.P449L variants on normal conditions and after TMAO treatment, measured as IP1 accumulation and expressed as percentage of stimulated WT activity. (B) Gs/cAMP pathway activity after maximal dose stimulation of TSHR WT, p.N432D, and p.P449L variants on normal conditions and after TMAO treatment, measured as cAMP reporter luciferase activity and expressed as percentage of stimulated WT activity. (C) Dose–response curves for Gs/cAMP signaling of TSHR WT, p.N432D, and p.P449L variants after TMAO treatment. Each variant's curve is expressed as a percentage of its own E _max. (D) Representative images of Western blotting experiments showing the expression and maturation of the TSHR variants without and after treatment with TMAO. TSHR was stained with 28.1 antibody, and actin was used as loading control. (E) Densitometric quantification of Western blot experiments showing the expression and maturation of the TSHR variants without and after treatment with TMAO. (F) Gq11/IP3 pathway activity after maximal dose stimulation of TSHR WT, p.E34K, and p.R46P variants on normal conditions and after TMAO treatment, measured as IP1 accumulation and expressed as percentage of stimulated WT activity. (G) Gs/cAMP pathway activity after maximal dose stimulation of TSHR WT, p.E34K, and p.R46P variants on normal conditions and after TMAO treatment, measured as cAMP reporter luciferase activity and expressed as percentage of stimulated WT activity. (H) Dose–response curves for Gs/cAMP signaling of TSHR WT, p.E34K, and p.R46P variants after TMAO treatment. Each variant's curve is expressed as a percentage of its own Emax. Statistical analysis: (A–C) n = 8, (E–H) n = 3. Statistical significance was determined with one-way ANOVA followed by Bonferroni's post hoc test (A–C) or Kruskal-Wallis H test followed by Dunn's post hoc test (E–G). *p < 0.05, **p < 0.01, ***p < 0.001 in respect to TMAO-untreated TSHR WT or as indicated.

Table 2.

Membrane Expression and Functional Parameters of the Loss-of-Function Thyrotropin Variants

	Membrane expression (3G4)	Membrane expression (BA8)	IP1 E _max (% WT)	cAMP E _max (% WT)	cAMP EC₅₀ (U/L)
WT	100.00 ± 1.09	100.00 ± 0.61	100.00 ± 10.41	100.00 ± 1.80	0.10 ± 0.04
WT + TMAO	102.90 ± 1.31	97.17 ± 2.75	115.10 ± 13.61	101.00 ± 6.33	0.04 ± 0.05
p.N432D	1.75 ± 2.06^***	0.31 ± 0.55^***	9.12 ± 2.08^***	1.53 ± 0.01^***	—
p.N342D + TMAO	51.03 ± 9.19^*** ^,§§§	77.3 ± 4.021^*** ^,§§§	10.36 ± 1.74^***	76.61 ± 3.50^*** ^,§§§	0.75 ± 0.09^***
p.P449L	128.8 ± 5.37^***	118.3 ± 4.64^**	30.96 ± 6.96^***	67.95 ± 1.26^***	0.62 ± 0.07^***
p.P449L + TMAO	133.80 ± 4.99^***	109.1 ± 5.51^*	22.25 ± 6.21^***	67.24 ± 2.98^***	1.09 ± 0.04^*** ^, ^§§
EMPTY	0.36 ± 4.40^***	0.41 ± 0.40^***	6.16 ± 1.98^***	2.85 ± 2.23^***	—
EMPTY + TMAO	1.06 ± 2.06^***	0.95 ± 0.87^***	7.17 ± 7.40^***	2.88 ± 2.25^***	—

The table summarizes the main characteristics of the LOF THSR variants. Membrane expression was examined with flow cytometry as in Figure 4C and E. Maximal stimulation and EC50 were obtained from the experiment as in Figure 5. Values are expressed as mean ± SD.

Statistical analysis: Statistical significance was determined with one-way ANOVA. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001 versus WT TSHR, ^§§ p < 0.01, ^§§§ p < 0.001 versus the respective untreated TSHR variant.

LOF, loss of function; SD, standard deviation; WT, wild type; TMAO, trimethylamine-N-oxide.

As a last step, we investigated two additional retained variants that had discordant in silico and in vitro functionality: the p.E34K, which has a reported membrane expression of 30% of WT, and the p.R46P, which is reported to be almost totally retained and with a very low ability to signal through the cAMP pathway (22,23).

TMAO treatment induced an increase in the cleaved A-subunit levels in the p.E34K variant and greatly enhanced the maturation of the retained p.R46P one (Fig. 5E, D), with effects similar to the ones observed in p.N432D. Accordingly, functional assays revealed a significant increase in both Gs/cAMP and Gq11/IP3 pathways for p.E34K and a significant rescue of the signaling abilities of p.R46P (Fig. 5F, G), with concentration–effect curves and EC50 similar to those of the WT (Fig. 5H).

Discussion

In the present work, we reveal two important issues regarding the possible intracellular destiny of the folding-defective TSHR mutants. First, they may be degraded not only through the proteasomal pathway but also through an alternative autophagosomal-like pathway that kicks in as an emergency exit after retention in the ER. Second, they can at least partially function if forced to the cell surface by using chemical chaperones. Our data provide a possible explanation for the observed lack of concordance between the in silico prediction of receptor functionality and in vitro findings, as misfolded mutants that retain signaling abilities may have a premature maturation arrest, intracellular retention, and subsequent degradation.

The involvement of the lysosomal system in the degradation of misfolded TSHR mutants is a new and interesting finding. In particular, p.N432D has such structural changes that prevent it from passing the ER quality control. In the ER, the mutant is likely to form aggregates (SA) that cannot be retro-translocated to the cytoplasm where the proteasome system operates, but are instead degraded by alternative autophagocytosis (Figs. 1–3 and Supplementary Fig. S1), a mechanism similar to the one previously described for gonadotropin-releasing hormone receptor mutant p.E90K (33).

The TMAO treatment likely inhibits the formation of ER aggregates while promoting receptor homodimerization, sheltering p.N432D from ER quality control and allowing advancement to the Golgi compartment and finally to the plasma membrane (3,46), as also indicated by the appearance of the A-subunit bands in plasma membrane preparations. Nevertheless, its maturation does not seem to follow regular steps even after TMAO treatment as we detected very low levels of complex carbohydrates form (Fig. 4A, B).

There are two possible explanations for this. The first and most likely hypothesis is that only a small percentage of plasma membrane p.N432D mutant reach full maturation, while most of it is still blocked at the high-mannose stage. Membrane expression of immature TSHR has already been described (26,42,43,47), and TSHR with reduced glycosylation sites has TSH binding affinity and EC50 for cAMP that are indistinguishable from the mature one (3). In this case, the p.N432D maturation limiting factor may be the ability to form dimers in the ER compartment, as the staining with 3G4 antibody showed a significant increase in the levels of immature TSHR dimers after TMAO treatment (Supplementary Fig. S2A, B). The increased cleavage indicated by the variation in BA8:3G4 ratio (Fig. 4C, E; Table 2) can then be explained by the already known higher sensitivity to protease action of immature TSHR (25,42).

The second possible explanation is that the TMAO-treated p.N432D mutant reaches full maturation, but all the mature receptors undergo proteolytic cleavage and thus only the A-subunit is detected. This may be explained by an increased sensitivity of the TSHR mutant to proteases or because a lower amount of mutant TSHR on the membrane is more effectively processed by proteases.

Irrespective of these considerations, functional assays show that p.N432D mutant is able to bind TSH and signal when expressed on the plasma membrane (Fig. 5A–C; Table 2). The lack of Gq11/IP3 pathway activity may be explained by the intrinsic differences between Gs and Gq interaction with TSHR.

First of all, the Gq11/IP3 pathway is more dependent on the total amount of cleaved receptor (10) and on TSHR homodimerization abilities (48), and TMAO-treated p.N432D has an absolute amount of cleaved receptor present on the plasma membrane that is definitely lower than WT (Fig. 4 and Supplementary Fig. S2).

In addition, interactions between TSHR and Gq are more demanding than the ones with Gs (4,49), and an in silico model predicted that p.N432D mutation results in more severe modifications that can affect the interaction with G-proteins (21). The TMAO treatment can either mask these conformational alterations or force the mutant through a more correct conformation, enough to achieve a partial rescue of Gs interactions and cAMP signaling but not enough to restore the more demanding interactions with Gq.

These speculations are also supported by the findings in two other discordant mutants, where TMAO treatment more efficiently rescued Gs/cAMP than the Gq11/IP3 signaling (Fig. 5D–H and Supplementary Fig. S4B).

In conclusion, TSHR can be degraded through proteasome or autophagosome pathways depending on specific structural defects. The chaperone TMAO allows TSHR mutants to pass ER quality control, increasing cell surface expression. As in other GPCR-related diseases, TSHR LOF mutations mainly cause ER retention, as detected by the discrepancy between in silico predictions and in vitro data. As demonstrated here, retained mutants that are brought to the cell surface are able to effectively transduce intracellular signals. These findings open the possibility of further studies on pharmacological modulation of TSHR expression and functionality in patients with disrupted TSHR signaling.

Footnotes

Authors' Contributions

E.S.G., A.L., and V.G. designed and performed the experiments. V.V. and M.B. contributed to the experiment planning. L.P. supervised the experimental work and provided research funds. All authors contributed to the writing and revision of the article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

The work was partially supported by Ricerca Corrente funds of Istituto Auxologico Italiano (project: EPIPOT; funding code: 05C002_2010). E.S.G. was partially supported by Fondazione Anna villa e Felice Rusconi (CF: 80008670129) PhD scholarship.

Supplementary Material

Supplementary Method

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Table S1

References

Persani

, Rurale

, de Filippis

, et al. 2018. Genetics and management of congenital hypothyroidism. Best Pract Res Clin Endocrinol Metab, 32:387–396.

Persani

, Calebiro

, Cordella

, et al. 2010. Genetics and phenomics of hypothyroidism due to TSH resistance. Mol Cell Endocrinol, 322:72–82.

Kursawe

, Paschke

. 2007. Modulation of TSHR signaling by posttranslational modifications. Trends Endocrinol Metab, 18:199–207.

Kleinau

, Neumann

, Grüters

, et al. 2013. Novel insights on thyroid-stimulating hormone receptor signal transduction. Endocr Rev, 34:691–724.

Siffroi-Fernandez

, Giraud

, Lanet

, et al. 2002. Association of the thyrotropin receptor with calnexin, calreticulin and BiP. Effects on the maturation of the receptor. Eur J Biochem, 269:4930–4937.

Costagliola

, Panneels

, Bonomi

, et al. 2002. Tyrosine sulfation is required for agonist recognition by glycoprotein hormone receptors. EMBO J, 21:504–513.

Bulenger

, Marullo

, Bouvier

. 2005. Emerging role of homo- and heterodimerization in G-protein-coupled receptor biosynthesis and maturation. Trends Pharmacol Sci, 26:131–137.

Bonomi

, Busnelli

, Persani

, et al. 2006. Structural differences in the hinge region of the glycoprotein hormone receptors: evidence from the sulfated tyrosine residues. Mol Endocrinol, 20:3351–3363.

Rapoport

, McLachlan

. 2016. TSH receptor cleavage into subunits and shedding of the A-subunit; a molecular and clinical perspective. Endocr Rev, 37:114–134.

10.

M-TH

, Radu

, Ghinea

. 2009. The cleavage of thyroid-stimulating hormone receptor is dependent on cell-cell contacts and regulates the hormonal stimulation of phospholipase c. J Cell Mol Med, 13:2253–2260.

11.

Mizrachi

, Segaloff

. 2004. Intracellularly located misfolded glycoprotein hormone receptors associate with different chaperone proteins than their cognate wild-type receptors. Mol Endocrinol, 18:1768–1777.

12.

Young

, Wertman

, Dupré

. 2015. Regulation of GPCR anterograde trafficking by molecular chaperones and motifs. Prog Mol Biol Transl Sci, 132:289–305.

13.

Milligan

. 2010. The role of dimerisation in the cellular trafficking of G-protein-coupled receptors. Curr Opin Pharmacol, 10:23–29.

14.

Conn

, Ulloa-Aguirre

, Ito

, et al. 2007. G Protein-coupled receptor trafficking in health and disease: lessons learned to prepare for therapeutic mutant rescue in vivo. Pharmacol Rev, 59:225–250.

15.

Newton

, Whay

, McArdle

, et al. 2011. Rescue of expression and signaling of human luteinizing hormone G protein-coupled receptor mutants with an allosterically binding small-molecule agonist. Proc Natl Acad Sci U S A, 108:7172–7176.

16.

Rivero-Müller

, Chou

Y-Y

, Ji

, et al. 2010. Rescue of defective G protein-coupled receptor function in vivo by intermolecular cooperation. Proc Natl Acad Sci U S A, 107:2319–2324.

17.

Janovick

, Pogozheva

, Mosberg

, et al. 2012. Rescue of misrouted GnRHR mutants reveals its constitutive activity. Mol Endocrinol, 26:1179–1188.

18.

Sato

, Ward

, Krouse

, et al. 1996. Glycerol reverses the misfolding phenotype of the most common cystic fibrosis mutation. J Biol Chem, 271:635–638.

19.

Baskakov

, Kumar

, Srinivasan

, et al. 1999. Trimethylamine N-oxide-induced cooperative folding of an intrinsically unfolded transcription-activating fragment of human glucocorticoid receptor. J Biol Chem, 274:10693–10696.

20.

Granell

, Mohammad

, Ramanagoudr-Bhojappa

, et al. 2010. Obesity-linked variants of melanocortin-4 receptor are misfolded in the endoplasmic reticulum and can be rescued to the cell surface by a chemical chaperone. Mol Endocrinol, 24:1805–1821.

21.

Lábadi Á, Grassi

, Gellén

, et al. 2015. Loss-of-function variants in a hungarian cohort reveal structural insights on TSH receptor maturation and signaling. J Clin Endocrinol Metab, 100:E1039–E1045.

22.

Agretti

, De Marco

, Capodanno

, et al. 2007. A fast method to detect cell surface expression of thyrotropin receptor (TSHr): the microchip flow cytometry analysis. Thyroid, 17:861–868.

23.

De Marco

, Agretti

, Camilot

, et al. 2009. Functional studies of new TSH receptor (TSHr) mutations identified in patients affected by hypothyroidism or isolated hyperthyrotrophinaemia. Clin Endocrinol (Oxf), 70:335–338.

24.

Costagliola

, Rodien

, Many

, Ludgate

, et al. 1998. Genetic immunization against the human thyrotropin receptor causes thyroiditis and allows production of monoclonal antibodies recognizing the native receptor. J Immunol, 160:1458–1465.

25.

Costagliola

, Khoo

, Vassart

. 1998. Production of bioactive amino-terminal domain of the thyrotropin receptor via insertion in the plasma membrane by a glycosylphosphatidylinositol anchor. FEBS Lett, 436:427–433.

26.

Alberti

, Proverbio

, Costagliola

, et al. 2001. A novel germline mutation in the TSH receptor gene causes non-autoimmune autosomal dominant hyperthyroidism. Eur J Endocrinol, 145:249–254.

27.

Urizar

, Montanelli

, Loy

, et al. 2005. Glycoprotein hormone receptors: link between receptor homodimerization and negative cooperativity. EMBO J, 24:1954–1964.

28.

Minich

, Lenzner

, Morgenthaler

. 2004. Antibodies to TSH-receptor in thyroid autoimmune disease interact with monoclonal antibodies whose epitopes are broadly distributed on the receptor. Clin Exp Immunol, 136:129–136.

29.

Lüblinghoff

, Nebel

, Huth

, et al. 2012. The leipzig thyrotropin receptor mutation database: update 2012. Eur Thyroid J, 1:209–210.

30.

Grassi

, Vezzoli

, Negri

, et al. 2015. SP600125 has a remarkable anticancer potential against undifferentiated thyroid cancer through selective action on ROCK and p53 pathways. Oncotarget, 6:36383–36399.

31.

Schindelin

, Arganda-Carreras

, Frise

, et al. 2012. Fiji: an open-source platform for biological-image analysis. Nat Methods, 9:676–682.

32.

Lábadi Á, Grassi

, Gellén

, et al. 2015. Loss-of-function variants in a hungarian cohort reveal structural insights on TSH receptor maturation and signaling. J Clin Endocrinol Metab, 100:E1039–E1045.

33.

Houck

, Ren

, Madden

, et al. 2014. Quality control autophagy degrades soluble ERAD-resistant conformers of the misfolded membrane protein GnRHR. Mol Cell, 54:166–179.

34.

, Echeverri

, Moyer

. 2003. Endoplasmic reticulum retention, degradation, and aggregation of olfactory G-protein coupled receptors. Traffic, 4:416–433.

35.

Saliba

, Munro

PMG

, Luthert

, et al. 2002. The cellular fate of mutant rhodopsin: quality control, degradation and aggresome formation. J Cell Sci, 115:2907–2918.

36.

D'Agostino

, Crespi

, Polishchuk

, et al. 2014. ER reorganization is remarkably induced in COS-7 cells accumulating transmembrane protein receptors not competent for export from the endoplasmic reticulum. J Membr Biol, 247:1149–1159.

37.

Cook

, Zhu

C-C

, Hinkle

. 2003. Thyrotropin-releasing hormone receptor processing: role of ubiquitination and proteasomal degradation. Mol Endocrinol, 17:1777–1791.

38.

Petaja-Repo

, Hogue

, Laperriere

, et al. 2001. Newly synthesized human opioid receptors retained in the endoplasmic reticulum are retrotranslocated to the cytosol, deglycosylated, ubiquitinated, and degraded by the proteasome. J Biol Chem, 276:4416–4423.

39.

Lahuna

, Quellari

, Achard

, et al. 2005. Thyrotropin receptor trafficking relies on the hScrib-betaPIX-GIT1-ARF6 pathway. EMBO J, 24:1364–1374.

40.

, Van Sande

, Lefort

, et al. 2001. Effects of mutations involving the highly conserved S281HCC motif in the extracellular domain of the thyrotropin (TSH) receptor on TSH binding and constitutive activity. Endocrinology, 142:2760–2767.

41.

Ding

, Yin

. 2008. Sorting, recognition and activation of the misfolded protein degradation pathways through macroautophagy and the proteasome. Autophagy, 4:141–150.

42.

Siffroi-Fernandez

, Costagliola

, Paumel

, Giraud

, et al. 2001. Role of complex asparagine-linked oligosaccharides in the expression of a functional thyrotropin receptor. Biochem J, 354(Pt 2):331–336.

43.

Sura-Trueba

, Aumas

, Carre

, et al. 2009. An inactivating mutation within the first extracellular loop of the thyrotropin receptor impedes normal posttranslational maturation of the extracellular domain. Endocrinology, 150:1043–1050.

44.

Chen

, Ma

, Liu

, et al. 2014. Functional rescue of Kallmann syndrome-associated prokineticin receptor 2 (PKR2) mutants deficient in trafficking. J Biol Chem, 289:15518–15526.

45.

Robben

, Sze

, Knoers

NVAM

, et al. 2006. Rescue of vasopressin V2 receptor mutants by chemical chaperones: specificity and mechanism. Mol Biol Cell, 17:379–386.

46.

Nagayama

, Nishihara

, Namba

, et al. 2000. Identification of the sites of asparagine-linked glycosylation on the human thyrotropin receptor and studies on their role in receptor function and expression. J Pharmacol Exp Ther, 295:404–409.

47.

Nagayama

, Namba

, Yokoyama

, et al. 1998. Role of asparagine-linked oligosaccharides in protein folding, membrane targeting, and thyrotropin and autoantibody binding of the human thyrotropin receptor. J Biol Chem, 273:33423–33428.

48.

Allen

, Neumann

, Gershengorn

. 2011. Occupancy of both sites on the thyrotropin (TSH) receptor dimer is necessary for phosphoinositide signaling. FASEB J, 25:3687–3694.

49.

Kleinau

, Jaeschke

, Worth

, et al. 2010. Principles and determinants of G-protein coupling by the rhodopsin-like thyrotropin receptor. PLoS One, 5:e9745.