Pathogenic MCT8^V235L creates a steric clash that is alleviated by a compensating mutation of MCT8^F285A

Introduction

Monocarboxylate transporter (MCT) 8 is a specific thyroid hormone (TH) transporter that facilitates the transport of the active TH 3,3′,5-triiodothyronine (T3), its prohormone 3,3′,5,5′-tetraiodothyronine (T4, thyroxine) as well as the inactive 3,3′,5′-triiodothyronine (reverse T3, rT3) across plasma membranes (1, 2). Inside the cell, TH are metabolized by three enzymes called deiodinase type 1–3. These enzymes have different substrate specificities and either activate T4 to T3 or inactivate T4 and T3 to rT3 or T2 (3). T3 binds to nuclear T3 receptors, modulating their function as transcription factors (4).

The solute carrier family 16A2 (SLC16A2) gene, encoding MCT8, is located on the X chromosome. Thus, pathogenic variants lead to MCT8 deficiency in male patients, formerly addressed as Allan-Herndon-Dudley syndrome (AHDS) (5). We recently described a patient with MCT8 deficiency carrying a substitution of valine 235 to leucine (MCT8^V235L) (6). The patient shows the main characteristics of MCT8 deficiency including developmental delay accompanied by muscle hypotonia and dystonia. Thyroid function tests revealed a typical endocrine phenotype with high T3 and low T4 as well as borderline increased thyroid-stimulating hormone. Biochemical analyses of the MCT8^V235L mutant overexpressed in Madin-Darby canine kidney (MDCK) 1 cells demonstrated a completely abolished T3 transport activity of the mutant (6).

Other pathogenic variants known at amino acid position 235 are the substitution of valine by methionine (MCT8^V235M) as well as an insertion of an additional valine (MCT8^insV236) (5, 7). Patients with these mutations similarly display severe developmental delay and muscle hypotonia. Biochemical analyses in overexpressing cell systems revealed that the TH transport activity of MCT8^V235M is completely abolished while the activity of MCT8^insV236 depends on the cell type (8, 9, 10).

Over the years, we and others developed several homology models to understand the structure–function relations of MCT8. We initially used the glycerol-3-phosphate transporter (GlpT) to create the first MCT8 homology model in the inward-open conformation (2). Updated versions of the inward-open homology model as well as newly created homology models in the outward-open and partially occluded outward-open conformations were generated based on the xylose transporter XylE, the glucose transporter GLUT3 and the fucose transporter FucP (11, 12). Our homology models helped to rationalize the functions of His415, Arg301, His192 and Arg445 involved in substrate binding as well as amino acids important for substrate specificity (e.g. Ala224) and conformational changes (e.g. Lys418 and Tyr419) (11, 13, 14). The homology model of Groeneweg et al. (12) identified amino acids involved in the formation of the substrate-binding pocket (e.g. Phe287).

Here, we studied our outward-open homology model (11) to understand the effect of the valine to leucine replacement at position 235 and identified steric clashes and van der Waals (VDW) interactions between Leu235 and Phe285 as well as Leu235 and Gln288 possibly responsible for the transport inactivity of the pathogenic V235L mutation. While we were able to exclude interactions with Gln288 as a cause for impaired T3 transport activity by testing a series of MCT8^V235 mutants in the present study, we speculated that steric clashes and VDWs interactions between Leu235 and Phe285 might cause undesirable conformational changes that impair the function of another Phe at position 287. Thus, we tested a series of Phe287 mutants to understand its role in T3 transport. Groeneweg et al. have previously shown that the mutation of the aromatic Phe287 to the smaller aliphatic alanine (MCT8^F287A) reduces T3 and T4 transport activity in overexpressing cells (12). Here, we demonstrate that transport activity of MCT8^V235L is rescued by the compensating mutation of Phe285 to alanine that most likely relieves steric clashes.

Materials and methods

Results

Structure analyses of the MCT8 homology model in the outward-open conformation (11) predict that Val235 is present in the second transmembrane domain (TMD) without immediate access to either the transport channel or molecular surface. Rather, it packs closely to TMD4, which is known to contribute amino acids interacting with the substrate (14). In our model, the substitution of valine to leucine at position 235 leads to steric clashes with Phe285 and Gln288 in TMD4 (Fig. 1). Using RING (18) and Cytoscape (https://cytoscape.org), we discovered potentially detrimental van der Waals (VDW) interactions between Leu235 and Phe285 as well as Gln288, which did not occur with Val235 (Table 2). In contrast to Leu235, pathogenic Met235 (MCT8^V235M) forms VDW interactions with Phe285 but not with Gln288 (Table 2).

View Full Size

Structure of the outward-open homology model of MCT8 (11). (A) Overview of the MCT8 monomer with highlights on Val235, Phe285, Phe287 and Gln288. The potential transport cavity is highlighted in light blue (top view). (B and C) Magnified view of position 235 and its molecular environment for Val235 (B) and pathogenic Leu235 (C). Analyses of the homology model suggest steric clashes between Leu235 and Phe285 as well as Gln288 that are highlighted with light red circles.

Citation: European Thyroid Journal 14, 3; 10.1530/ETJ-25-0009

• Download Figure
• Download figure as PowerPoint slide

We have previously shown that pathogenic MCT8^V235L is unable to transport T3 (6), although it is expressed at the plasma membrane of MDCK1 cells (Fig. 2A). To understand, which interactions contribute to the loss of MCT8^V235L transport function, we mutated MCT8^V235 to alanine, threonine and isoleucine. We selected these amino acids to test for different biochemical side chain properties. The substitution of valine to alanine creates a hole by shortening the side chain, while the mutation of valine to threonine replaces a methyl group with a polar hydroxy group. Both mutations avoid VDW interactions with Phe285 and Gln288, respectively (Table 2). The exchange of valine with isoleucine elongates the branched side chain, moves the chiral C_β to C_γ, and introduces two potential VDW interactions with Gln288, while VDW interactions with Phe285 are absent (Table 2).

View Full Size

Expression and function of MCT8^V235 mutants. (A) Plasma membrane expression of MCT8 wild-type (top) and pathogenic V235L (bottom) in overexpressing MDCK1 cells analyzed by immunocytochemistry. MCT8 expression is shown in red (RRID: AB_307019 for HA antibody and RRID: AB_2338060 for Alexa Fluor 594 goat anti-rabbit secondary antibody). DAPI staining of the cell nuclei is shown in blue. Scale bar: 15 μM. (B) 10 μg of whole cell lysates from MDCK1 cells stably expressing control and mutant MCT8 were used for western blot analysis. Empty vector transfected MDCK1 cells served as negative controls (pcDNA3). MCT8 was detected by anti-HA antibody (RRID:AB_307019), β-ACTIN (RRID:AB_262011) served as loading control. (C) Quantification of western blot analyses (n = 4) as shown in (B). (D) T3 transport activity of MCT8^V235 mutants compared to control MCT8 expressed in MDCK1 cells. 2 nM ¹²⁵I-T3 and 10 nM T3 were applied to the cells for 20 min. Empty vector transfected MDCK1 cells served as negative controls and were subtracted as background. Results for each cell clone were normalized to MCT8 expression levels determined in (C). The normalized results for clones A and B of each MCT8 wild-type and mutant were combined in one column. The MCT8 wild-type was set to 100% to combine multiple experiments in one graph (n = 4). Statistical analysis was done using one-way ANOVA with Dunnett’s multiple comparisons. NS, not significant; SD, standard deviation.

Citation: European Thyroid Journal 14, 3; 10.1530/ETJ-25-0009

• Download Figure
• Download figure as PowerPoint slide

Western blot analyses confirmed the expression of all MCT8^V235 mutants in stably expressing MDCK1 cells (Fig. 2B). We quantified western blotting results to normalize the T3 transport activity of each mutant to the respective expression levels (Fig. 2C). Uptake experiments with radioactively labeled T3 showed that none of our mutations at position 235 affected the transport activity of MCT8^V235 mutants (Fig. 2D). Thus, neither reduction (Val235Ala) nor the change in polarity (Val235Thr) of the amino acid at position 235 abolished transport function. Furthermore, increasing the size of the side chain (Val235Ile) and the associated introduction of two predicted VDW interactions between Ile235 and Gln288 had no influence on transport activity. Thus, we speculated that VDW interactions between Leu235 and Gln288 are not the reason for the loss of transport function of MCT8^V235L. These findings are in accordance with absent VDW interactions between pathogenic Met235 and Gln288 (Table 2). This leaves the movement of the chiral center from C_β to C_γ and the associated movement of the methyl group as the culprit in V235L.

Here, we hypothesized that the interaction of Leu235 with Phe285 leads to steric clashes that result in an undesired shift of Phe287 into the potential transport cavity (Fig. 1C).

To test which side chain properties at amino acid position 287 are critical for T3 transport, we mutated Phe287 to tryptophan, tyrosine and valine. The tryptophan substitution increases the size of the aromatic side chain, while the tyrosine adds a polar hydroxy group. Substitution with valine removes the aromatic properties and reduces the size at this position.

Biochemical analyses revealed that all MCT8^F287 mutants are expressed in MDCK1 cells (Fig. 3A). Again, we quantified the western blot signal to normalize the results obtained from T3 transport measurements (Fig. 3B). We observed increased T3 transport activity for MCT8^F287W, unchanged activity for MCT8^F287Y and decreased T3 transport activity for MCT8^F287V (Fig. 3C). Surface biotinylation experiments showed that the mutation of phenylalanine to valine at position 287 affects plasma membrane expression of the mutant, possibly responsible for the reduction in T3 uptake (Fig. 3C and D). These experiments demonstrate that the polarity (Phe287Tyr) of the side chain at position 287 is not critical for transport function, while the enlargement (Phe287Trp) supports transport activity. The experiments further confirmed the results by Groenweg et al. who showed that a mutation of Phe287 to alanine decreased T3 and T4 transport function (12). We observed a similar decrease in T3 transport function when we reduced the side chain size by mutating phenylalanine to valine. The reduction in T3 transport is likely due to reduced plasma membrane expression of F287V. Thus, we concluded that a certain size and/or the aromatic properties of the side chain at position 287 are pivotal for proper plasma membrane expression and T3 transport activity in MDCK1 cells.

View Full Size

Expression and function of MCT8^F287 mutants. (A) 10 μg of protein lysates of MDCK1 cells stably expressing mutant MCT8 were used for western blot analysis. pcDNA3 transfected MDCK1 cells served as negative controls. MCT8 detection was performed with anti-HA antibody (RRID:AB_307019). Anti-β-ACTIN antibody (RRID:AB_262011) was used as loading control. (B) Quantification of western blot analyses (n = 4) from cell lysates and surface biotinylating experiments show a reduced expression of F287V at the plasma membrane compared to cell lysate. Statistics: two-way ANOVA with Sidak’s multiple comparisons. (C) T3 transport of MCT8^F287 mutants compared to control MCT8 after 20 min. T3 uptake of pcDNA3 transfected cells was subtracted as background. Results for each cell clone were normalized to MCT8 expression levels in cell lysate determined in (B). The normalized results for clones A and B of each MCT8 wild-type and mutant were combined in one column. The mean of MCT8 wild-type was set to 100% to combine multiple experiments in one graph (n = 4). Statistics: one-way ANOVA with Dunnett’s multiple comparisons. (D) Plasma membrane expression of MCT8 and F287 mutants in MDCK1 cells investigated by surface biotinylation and western blotting. **P < 0.01. ***P < 0.001. NS, not significant; SD, standard deviation.

Citation: European Thyroid Journal 14, 3; 10.1530/ETJ-25-0009

• Download Figure
• Download figure as PowerPoint slide

RING and Cytoscape analyses predicted that the mutagenesis of Phe285 to alanine might abolish VDW interactions with Leu235. Thus, we decided to establish an MCT8^V235L,F285A double mutant to prevent unwanted VDW forces and to mitigate steric clashes possibly responsible for a conformational change of Phe287. Western blot analysis of MCT8^V235L,F285A double mutants as well as MCT8 and MCT8^F285A controls demonstrated the expression of all tested cell clones in MDCK1 cells (Fig. 4A). We quantified the western blot signal to normalize the results obtained for T3 uptake measurements (Fig. 4B). T3 uptake experiments confirmed that MCT8^F285A and MCT8^V235L,F285A were able to transport T3 similarly to MCT8-expressing control cells (Fig. 4C). The MCT8^V235L,F285A double mutant presented a similar Michaelis constant (K_M) for T3 as wild-type MCT8, with K_M (MCT8) = 4.7 ± 0.8 μM and K_M (MCT8^V235L,F285A) = 4.3 ± 0.9 μM (Fig. 4D).

View Full Size

Expression and function of MCT8^V235L,F285A double mutants. (A) Western blot analyses of MCT8, MCT8^F285A and MCT8^V235L,F285A were performed with 20 μg whole cell lysates. pcDNA3 transfected cells were used as negative control. Antibodies anti-HA (RRID:AB_307019) and β-ACTIN (RRID:AB_262011) were used for detection of MCT8 and the loading control, respectively. (B) Quantification of western blot analyses (n = 4) as shown in (A). (C) Relative T3 uptake of MCT8^V235L,F285A double mutant in comparison to wild-type MCT8 and MCT8^F285A after 20 min. Results for each cell clone were normalized to MCT8 expression levels determined in (B). Background T3 uptake of empty-vector transfected MDCK1 cells was subtracted. The normalized results of each MCT8 wild-type and mutant cell clone were combined in one column. The MCT8 wild-type was set to 100% to combine multiple experiments in one graph (n = 2–3). Statistics: one-way ANOVA with Dunnett’s multiple comparisons. (D) K_M value determination of MCT8^{V235L, F285A} (n = 2). Values were corrected for background uptake and normalized to expression levels determined in (B). The normalized results of MCT8 wild-type cell clones as well as MCT8^V235L,F285A cell clones were combined in the two curve fits. NS, not significant; SD, standard deviation.

Citation: European Thyroid Journal 14, 3; 10.1530/ETJ-25-0009

• Download Figure
• Download figure as PowerPoint slide

Discussion

MCT8 deficiency is caused by mutations in the SLC16A2 gene coding for the specific TH transporter MCT8. The pathogenic Val235Leu variant recently reported in an MCT8-deficient patient replaces a hydrophobic branched-chain valine with a larger, hydrophobic branched-chain leucine (6). The key difference between both amino acids is the chiral, methylated carbon that is shifted by one C atom. While this mutation appears to be a mild exchange, the MCT8^V235L mutant completely lost the ability of T3 transport in overexpressing MDCK1 cells (6). Analyses of our homology model helped us to unravel the mechanism leading to the inability of MCT8^V235L to transport T3.

Our homology model (11) implies that the amino acids Phe285 and Gln288 (located in TMD4) form a pocket for Val235 (located in TMD2). We mutated Val235 to alanine, threonine and isoleucine to assess the side chain characteristics necessary for the interactions within the pocket. Neither the reduction (MCT8^V235A) nor the change in polarity (MCT8^V235T) of the side chain impeded T3 transport. Although isoleucine and leucine have similar side chain characteristics, the mutation of Val235 to leucine completely abolished T3 transport (6), while the transport activity of MCT8^V235I was not affected. Investigations of our outward-open homology model suggested that Leu235 forms VDW interactions with Phe285 that are absent with Ile235. Similar VDW interactions are formed with Met235, another inactive pathogenic MCT8 mutation (8, 9). We speculate that both MCT8 mutants may be inactive due to the formation of unwanted VDW interactions between the mutated amino acid at position 235 and Phe285.

Analyses of our homology model predicted steric clashes between Leu235 and Phe285, which possibly push helix 4 aside and thus shift Phe287 into the potential substrate binding cavity. The importance of Phe287 for TH transport was previously described in overexpressing COS1 and JEG3 cells (12). The reduction in size and the removal of the aromatic properties of the side chain by mutating Phe287 to alanine reduced the T3 and T4 transport activity (12). Our experiments further investigated the side chain characteristics necessary at this position by mutating Phe287 to tyrosine, tryptophan and valine. Tyrosine introduced a polar hydroxy group, while tryptophan led to an enlargement of the side chain. Valine removed the aromatic properties and reduced the size of the side chain. All three tested mutants were capable of T3 transport, although the valine substitution reduced the T3 transport activity due to disturbed plasma membrane expression. The enlargement of the side chain increased the T3 transport activity. The results suggest that not polarity but a certain size and/or aromatic properties influence the side chain characteristic at position 287. Different from the diminished plasma membrane expression of MCT8^F287V in MDCK1 cells, the surface expression of MCT8^F287A in COS1 cells was not affected (12). We and others have already shown that the plasma membrane expression of MCT8 mutants depends on cellular context (8, 10). The corresponding amino acid to Phe287 in MCT10 is Tyr184. We have previously shown that Tyr184 is crucial for the uptake of aromatic amino acids and that the substitution to phenylalanine, among others, enables MCT10 to transport T4 (14).

We hypothesized that a shortening of the amino acid side chain at position 285 might reduce the steric clash with Leu235 and the unwanted conformational shift of Phe287. Thus, we generated an MCT8^V235L mutant with an additional mutation of Phe285 to alanine. The MCT8^V235L,F285A double mutant rescued T3 transport in overexpressing MDCK1 cells, which confirmed the presence of the predicted steric clash between Leu235 and Phe285. The K_M values for wild-type MCT8 and MCT8^V235L,F285A were in a similar range.

In conclusion, we show that in a 12-TMD transporter, a simple shift of a chiral carbon far from the substrate binding site can affect the transport activity if the packing of helices is altered, even slightly.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This work was supported by Sherman Family Foundation and Deutsche Forschungsgemeinschaft (DFG): BR6514/1 to D B and SCHW914/7 to U S.

Author contribution statement

D B designed the experiments. N S analyzed the homology model. N S and D B performed the experiments. N S, D B and U S analyzed the data. N S and D B wrote the manuscript. U S and F S commented on the manuscript.