Selective proteasome degradation of C‐terminally‐truncated human WFS1 in pancreatic beta cells

Wolfram syndrome is a monogenic disease mainly caused by mutations in the WFS1 gene. Mutations in the WFS1 gene give rise to diabetes. Here, we characterized mutant WFS1 proteins by studying the stability of full‐length wild‐type (WT) WFS1, a missense mutant P724L, and two C‐terminally truncated mutants, W837X and Y652X. We compared their stability by overexpressing them in MIN6 and HEK293T cells. The C‐terminally truncated mutants W837X and Y652X are degraded more rapidly than the missense P724L mutant or wild‐type WFS1 in MIN6 cells. In contrast, Y652X is more stable than WT or other mutant WFS1 proteins in HEK293T. In conclusion, we found that C‐terminally truncated WFS1 mutants are selectively degraded in a cell type‐specific manner.

Wolfram syndrome is a monogenic disease mainly caused by mutations in the WFS1 gene. Mutations in the WFS1 gene give rise to diabetes. Here, we characterized mutant WFS1 proteins by studying the stability of fulllength wild-type (WT) WFS1, a missense mutant P724L, and two Cterminally truncated mutants, W837X and Y652X. We compared their stability by overexpressing them in MIN6 and HEK293T cells. The Cterminally truncated mutants W837X and Y652X are degraded more rapidly than the missense P724L mutant or wild-type WFS1 in MIN6 cells. In contrast, Y652X is more stable than WT or other mutant WFS1 proteins in HEK293T. In conclusion, we found that C-terminally truncated WFS1 mutants are selectively degraded in a cell type-specific manner.
Wolfram Syndrome is an autosomal recessive disorder characterized by early-onset insulin-dependent diabetes mellitus optic atrophy and progressive neurodegeneration [1]. The WFS1 gene is identified as the main locus mutated in the majority of the patients [2,3]. Wolfram syndrome is characterized as a spectrum disorder. More than 200 different variations in the WFS1 gene have been described in Wolfram syndrome patients [4][5][6][7]. WFS1 encodes a transmembrane protein, also called wolframin, which consists of 890 amino acids and is predicted to have 9 or 10 membrane-spanning domains. The WFS1 protein is ubiquitously expressed but at higher levels in the brain, heart, lung, and pancreas [2,3].
Mutations in the WFS1 gene are reported to affect the WFS1 protein stability [17][18][19]. Wfs1-knockout mice were developed [20][21][22], which exhibit enhanced ER stress, impaired insulin processing and regulation of insulin secretion and develop progressive glucose intolerance and increased pancreatic beta cell death.
Previous studies of genotype-phenotype correlations for WFS1 mutations in Japanese Wolfram syndrome patients revealed that WFS1 mutation could be subdivided into groups in correlation with clinical severity. Variants with nonsense mutations, frameshift mutations, and/or multiple amino acid insertion/deletions in both alleles show more severe symptoms than variants with missense mutations and/or single amino acid insertions in both alleles [7].
In the present study, we introduced one missense and two nonsense mutations in the WFS1 cDNA to gain insights into the relationship between the severity of the mutations. We overexpressed the wild-type (WT) and mutant WFS1 cDNAs and compared their stabilities in mouse insulinoma 6 (MIN6) cells, a mouse pancreatic b-cell line, and HEK293T cells, a human embryonic kidney cell line.

Cell lines
MIN6 cells were obtained from J. Miyazaki in Osaka University and HEK293T cells were from RIKEN BioResource Center (BRC) [23,24].

WFS1 mutant adenovirus vector construction
A cDNA fragment of human WFS1 WT open reading frame (nt. 1-2673; NP_005996.1 aa: 1-890) was cloned from human islet cDNA by PCR. Human WFS1 mutants were generated using inverse PCR and In-Fusion HD (Takara Bio Inc., Shiga, Japan). Human WFS1 was used as the template for PCR. All primers for inverse PCR are listed in Table S1. The WFS1 mutant adenoviral vectors are registered at GenBank and listed in Table S2. Adenoviruses were produced as described [25]. Adenovirus titers were determined by Median Tissue Culture Infectious Dose (TCID50) [26].
HaloTag pulse-chase assay MIN6, HEK293T cells were cultured in Dulbecco's modified Eagle's medium supplied with 25 mM glucose (Thermo Fisher Scientific Inc., Waltham, MA, USA), 10% fetal bovine serum, 100 lM non-essential amino acids, 2 mM Lglutamine, 50 unitsÁmL À1 penicillin, 50 lgÁmL À1 streptomycin, and 100 lM 2-mercaptoethanol, at 37°C, 5% CO 2 . Cells were dissociated with 0.05% Trypsin-EDTA (Thermo Fisher Scientific) and seeded on a 96-well plate coated with Easy iMatrix-511 (Matrixome Inc., Osaka, Japan) at a density of 50 000 cells per well (MIN6 cells) or 10 000 cells (HEK293T cells). MIN6 cells were infected with adenovirus carrying HaloTag-hWFS1 or hWFS1 mutants to examine the steady state of the protein expression in Fig. 1 and in the pulselabeling experiment in Fig. 2. The multiplicity of infection was 3.3. In the pulse-chase experiment, HaloTag-hWFS1 or mutants were inserted into the pCAGGS vector [27] and transfected using Lipofectamine 3000 (Thermo Fisher Scientific) in Figs 3 (MIN6) and 4 (HEK293T). HaloTag-hWFS1 overexpressing vectors are registered at GenBank and listed in Table S2. Labeling using R110 or TMR was done following the manufacturer's protocol. After infection or transfection, cells were cultured for 32 h, then treated with HaloTag R110Direct TM ligand (R110; Promega, Madison, WI, USA) for 16 h to bind to the already produced 'old' HaloTag-hWFS1. After washing with medium to remove the R110, cells were cultured in the presence of a proteasome inhibitor, MG132 (10 lM, in the pulse-chase experiment), or control DMSO (in the pulse labeling). Then, cells were treated with HaloTag TMR ligand (TMR; Promega) for 15 min (AEMG132) to label newly synthesized HaloTag-hWFS1. After pulse labeling with TMR, the steady-state expression levels of the WFS1 proteins were examined after fixing the cells in 4% paraformaldehyde in PBS and counter-stained with DAPI (F. Hoffmann-La Roche Ltd., Basel, Switzerland) (Fig. 2). R110 or TMR fluorescence that labeled the HaloTag-hWFS1 in the cells was analyzed with ImageXpress Micro Confocal High-Content Imaging System and MetaXpress (Molecular Devices, San Jose, CA, USA). To chase the TMR-labeled HaloTag WT or mutant hWFS1 proteins, cells were counter-stained with Hoechst 33342 (Bio-Rad Laboratories Inc., Richmond, CA, USA) and cultured for 8, 16, or 24 h in the absence of MG132 (Figs 3 and 4). The areas of cells expressing the WFS1 proteins were determined as region of interest (ROI) based on the 0 h image labeled with R110. The averaged TMR fluorescence intensity values were obtained for each ROI at 0, 8, 16, and 24 h after TMR treatment. All calculations were performed using the IMAGE J FIJI program [28] for images taken by IMAGEXPRESS. washing with PBST (PBS containing 0.1% Tween20), the membrane was probed with a horseradish peroxidaseconjugated antibody. The membrane was scanned using FUSION-SOLO.4S.WL (M&S Instruments Inc., Tokyo, Japan) after reacting with Luminata Forte Western HRP Substrate (Merck Millipore).

Statistics
Data are presented as mean AE SD (standard deviation) with individual data. Data were analyzed by one-way ANOVA and Dunnett multiple comparisons test (Fig. 1D), Student's t-test (Fig. 2B), or Mann-Whitney's U-test (Figs 3C and 4B). Significant differences are shown as **P < 0.01 or P < 0.05 in the tables. All data were obtained from more than three independent experiments except those specifically indicated.

Results
The hWFS1 C-terminally truncated mutants are unstable in MIN6 cells We investigated one missense and two C-terminally truncated human WFS1 (hWFS1) mutations. The missense mutation corresponds to the luminal C-terminal (P724L), and the C-terminal truncation mutations Y652X and W837X lead to the truncation of a large portion of the C-terminus end (Y652X) or truncation of the last 54 amino acid residues (W837X) (Fig. 1A).
The patients with the P724L mutation were previously described as late onsets of diabetes mellitus, compared to those with the W837X and Y652X mutations, who developed early onset diabetes [7].
To assess the steady-state expression of the hWFS1 mutants, we generated adenoviral vectors carrying human WT or mutant hWFS1 tagged with HaloTag [30] at their N-terminal ends. We infected the MIN6 cell with the above adenoviral vectors and detected by western blot analysis using an anti-HaloTag antibody or anti-WFS1 antibody. The exogenous HaloTag WFS1 WT or the missense hWFS1 mutant P724L gave strong signals at approximately 134 kDa. In contrast, the C-terminally truncated hWFS1 mutants Y652X and W837X expressed in MIN6 cells gave weak signals at 127.7 kDa (W837X) or 106.8 kDa (Y652X) (Fig. 1B). In MIN6 cells, the exogenous HaloTag WFS1 WT and P724L were expressed at almost similar levels with those of the endogenous WFS1, detected by an anti-WFS1 antibody (Fig. 1C).
We also used a synthetic Halo ligand R110 that binds to the HaloTag-hWFS1 protein. The R110 fluorescence visualizes the expression of the WT or mutant WFS1 (Fig. 1D,E). Representative images are shown in Fig. 1D. Average fluorescence intensities in the ROIs are shown in scattered plots (Fig. 1E). The results revealed that the P724L missense mutation was expressed at approximately 0.7-fold of that of the WT. The C-terminally truncated mutants W837X and Y652X were expressed at 0.1-fold of the WT (Fig. 1D,E). The results suggest an instability of the C-terminally truncated hWFS1 mutations expressed in MIN6 cells.

C-terminally truncated hWFS1 proteins are rapidly degraded by the proteasome in MIN6 cells
It is reported that WFS1 mutants are degraded by the ubiquitin-proteasome system [17,18]. We investigate whether the instability of the WFS1 mutant proteins expressed in MIN6 cells is through proteasome degradation. After labeling the already produced 'old' HaloTag-hWFS1 with R110, the HaloTag-hWFS1 overexpressing MIN6 cells were incubated for 7 h with or without proteasome inhibitor MG132, then labeled the newly synthesized HaloTag-hWFS1 protein with TMR ligand (Fig. 2A, pulse labeling with TMR). The results showed that TMR-labeled WT and P724L mutant hWFS1 protein accumulated even without MG132. MG132 treatment increased the accumulation of WT or P724L hWFS1 to about 1.1-fold versus the untreated group (Fig. 2B, right). In contrast, the C-terminally truncated hWFS1 mutants were detected at low levels in the absence of MG132. In the presence of MG132, the C-terminally truncated mutants W837X or Y652X accumulated up to approximately 4.0-fold (W837X) or 3.5-fold (Y652X) in MIN6, compared to without MG132 (Fig. 2B).
Adding MG132 rescued the degradation of the C-terminally truncated mutants, indicating the proteasome-mediated mechanism is involved. To analyze this in detail, we performed another pulse-chase labeling experiment (Fig. 3A). The R110-labeled HaloTag-hWFS1 expressing MIN6 cells (overlapped with the TMR-labeled cells) (0 h) were used to determine the ROIs. TMR-labeled HaloTag-hWFS1 proteins were chased for 8, 16, and 24 h. As a result, WT and P724L proteins were reduced to about 0.65-fold after incubating without MG132 for 24 h. In contrast, C-terminally truncated mutants W837X and Y652X were rapidly degraded to about 0.40-to 0.50-fold at 16 h (W837X) or 24 h (Y624X) in MIN6 cells (Fig. 3B,C).
The results indicate a proteasome-mediated rapid degradation of C-terminally truncated mutants W837X and Y652X in MIN6 cells.
C-terminally truncated mutant Y652X is more stable in HEK293T cells compared to WT or other WFS1 mutants We then examined the stability of WT and mutants in HEK293T cells. We performed similar pulse-chase experiments in HEK293T cells. HEK293T cells expressed WFS1 endogenously at a much lower level than MIN6 cells (Fig. S1). The newly synthesized HaloTag-WT or mutant hWFS1 proteins expressed in the HEK293T cells were labeled with TMR (pulse, 0 h) and were chased for 8, 16, and 24 h in the absence of MG132.
Without MG132, the WT and the missense P724L mutant were degraded to approximately 0.60-fold (Fig. 4A,B), and the C-terminally truncated mutant W837X was about 0.65-fold at 24 h. In contrast, the C-terminally truncated mutant Y652X was about 0.85fold at 24 h. The results indicate that the C-terminally truncated Y652X mutant is the most stable form in HEK293T cells compared to the WT or the P724L mutant (Fig. 4A,B).

Discussions
We compared the stability of WT and mutant WFS1 protein in MIN6 and HEK293T cells. The C-terminus truncated mutants W652X and W837X are degraded by the proteasome pathway in MIN6 cells. In contrast, the C-terminally truncated mutant W652X is more stable than WT or other WFS1 mutants in HEK293T cells. MG132 treatment did not restore the Cterminally truncated mutants to accumulate to the WT level, suggesting the involvement of other nonproteasomal mechanisms. The result suggests that protein stability depends on the tissue or cell environment, which agrees that protein stability varies among tissues [31].
Our results agree with a previous report that the C-terminally truncated mutants exhibited higher stabilities than the WT WFS1 in HEK293T cells [18]. However, the exact mutants used between ours and the previous studies differ. Notably, WT and mutant WFS1 seem to accumulate at similar levels at a steady state in HEK293T cells [18], which agrees with our present results. In MIN6 cells, the steady-state accumulations of C-terminus truncated mutants W837X and T652X were approximately 0.1-fold of the WT WFS1 protein, which was much lower compared to HEK293T cells. The results indicate that WFS1 WT and mutants are more stable in HEK293T than in MIN6 cells. A selective degrading mechanism of Ctruncated WFS1 mutants seems to exist in beta cells. In MIN6 cells, the endogenous WFS1 was not degraded by overexpressing W837X or Y652X mutants, thereby confirming that the degradation system was not activated. The discrepancies between the stability of the WT and mutant WFS1 in MIN6 and HEK293T cells might be due to the different expression levels in regulating proteins involved in the degradation processes in a cell type-specific manner, which remains to be elucidated.
It is reported that the HECT-type ligase ubiquitin ligase (E3) Smad ubiquitination regulatory factor 1 (Smurf1) binds to the C-terminus of WFS1 and degrades WFS1 thus contributing to the instability in HEK293T cells [18]. However, the C-terminally truncated WFS1 mutants are unstable in MIN6 cells, suggesting a selective degrading mechanism of Cterminally truncated WFS1 mutants in beta cells. Expression of other members of the E3 ubiquitin ligase or other targeted protein degrading mechanisms, such as ERAD or SUMOylation, are reported in the pancreatic beta cells and might take part in regulating protein stability. However, we did not find any motif existing in the WFS1 protein [10,[32][33][34][35][36][37].
Understanding the molecular mechanism regulating WFS1 stability might provide a potential therapeutic target for Wolfram syndrome. supported in part by the Takeda Science Foundation, Japan Insulin Dependent Diabetes Mellitus (IDDM) Network.

Conflict of interest
The authors declare no conflict of interest.

Author contributions
HT performed the experiments, collected, analyzed, and discussed the data. DS discussed the data and provided conceptual input and technical advice. KT, YT, and NS discussed the data and gave technical advice. SK provided conceptual input, discussion, writing, and revision of the manuscript; approved the final version of the manuscript; and obtained funding.

Data accessibility
The data supporting this study's findings are openly available in GenBank at https://www.ncbi.nlm.nih.gov/ genbank/, and the accession numbers are listed in Table S2.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. HEK293T cells express WFS1 endogenously at a much lower level compared to MIN6 cells. Table S1. Primers used for generating WFS1 mutants.