Missense variants in CMS22 patients reveal that PREPL has both enzymatic and nonenzymatic functions

Abstract

Congenital myasthenic syndrome-22 (CMS22, OMIM 616224) is a rare genetic disorder caused by deleterious genetic variation in the prolyl endopeptidase-like (PREPL) gene. Previous reports have described patients with deletions and nonsense variants in PREPL, but nothing is known about the effect of missense variants in the pathology of CMS22. In this study, we have functionally characterized missense variants in PREPL from 3 patients with CMS22, all with hallmark phenotypes. Biochemical evaluation revealed that these missense variants do not impair hydrolase activity, thereby challenging the conventional diagnostic criteria and disease mechanism. Structural analysis showed that the variants affect regions most likely involved in intraprotein or protein-protein interactions. Indeed, binding to a selected group of known interactors was differentially reduced for the 3 variants. The importance of nonhydrolytic functions of PREPL was investigated in catalytically inactive PREPL p.Ser559Ala cell lines, which showed that hydrolytic activity of PREPL is needed for normal mitochondrial function but not for regulating AP1-mediated transport in the transgolgi network. In conclusion, these studies showed that CMS22 can be caused not only by deletion and truncation of PREPL but also by missense variants that do not necessarily result in a loss of hydrolytic activity of PREPL.

Keywords: Endocrinology; Genetic variation; Genetics.

Figure 1. CMS22 patient PREPL variants remain catalytically active.

(A) The 3 CMS22 mutants and the catalytically inactive PREPL p.Ser559Ala mapped on the PREPL structure. (B) Catalytic activity using FP-Biotin activity-based probe binding. Blot: Streptavidin = FP-biotin signal, Flag = PREPL signal. Graph: quantitative reactivity of WT and variant PREPL to the FP-biotin probe normalized to total PREPL abundance (n = 4) (C) Isothermal titration calorimetry measurements of the binding of inhibitor 8 to WT, p.Ser559Ala, p.Arg243Cys, and p.Arg647Gln PREPL. Shown are single representative traces and each stated K_D value is the mean from n = 3 technical replicates. (D) Representative graph of DIFMUO cleavage by WT and variant PREPL and calculated relative activity (n = 3). RFU, relative fluorescence units. Statistical analysis was performed using 1-way ANOVA. Significance levels are shown as *P ≤ 0.05 **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001.

Figure 2. Protein stability of CMS22 patient PREPL variants.

(A) Recombinant p.Arg243Cys, p.Arg647Gln mutants and the catalytically inactive p.Ser559Ala PREPL variant are soluble and monomeric by size exclusion chromatography. Recombinant p.Ile412Arg PREPL mutant aggregates in vitro, eluting in the void volume (8 mL). (B) CD spectra of CMS22 mutants, which share the same secondary structure as the WT protein. (C) Normalized thermal denaturation curves monitored by measuring the CD signal at 217 nm across a temperature range from 4°C to 95°C. The melting temperature (T_m) of PREPL WT and the p.Ser559Ala and p.Arg647Gln mutants is calculated to be 63.5°C, while the melting temperature of p.Arg243Cys is 61.5°C. The unfolding process of all PREPL variants displays a 2-step pattern, with the first unfolding step occurring at approximately 50°C. (D) Structure of PREPL p.Arg243Cys mutant solved by cryo-EM in the open conformation, sharing the same fold as the WT protein. (E) Protein halflife measured by pulse chase. Representative blot of PREPL abundance after 0, 5, 10, and 15 hours of chase and quantified halflife time for WT and variant PREPL. (n = 3) Statistical analysis was performed using 1-way ANOVA.

Figure 3. Overview of protein-protein interactions from 11 target proteins with PREPL variants.

Protein-protein interaction scores with 11 previously identified PREPL interactors were determined using MAPPIT and relative interaction signals plotted for (A) WT PREPL, and PREPL variants p.Arg243Cys, p.Ile412Arg, and p.Arg647Gln (B) WT PREPL and inactivated PREPL p.Ser559Ala an WT PREPL inactivated by addition of 50 μM total concentration of inhibitor 8 (C) WT PREPL and the catalytic domain or β-propeller domain separately.

Figure 4. PREPL^S559A mutant cells have a normal TGN morphology and display mitochondrial respiratory dysfunction.

(A) PREPL expression levels in 20 μg of HEK293T cell lysate from WT and PREPL variant cell lines. (B) PREPL activity evaluated by FP-TAMRA in lysates from WT and PREPL variant cell lines. (C) Evaluation of TGN size by confocal microscopy between WT and PREPL variant cell lines (n = 37–75). (D) Evaluation of mitochondria function by Seahorse cell mito stress test between WT and PREPL variant cell lines (n = 3). Data were analyzed by Mann-Whitney U test (multiple comparison corrected by Dunnett’s test). Significance levels are shown as **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001.

Figure 5. Location of CMS22 PREPL variants on the PREPL structure.

(A) Location of the CMS22 patient mutations within the PREPL crystal structure (PDB: 7OBM). (B) Arg412 is located on the β-propeller in the vicinity of the hydrophobic residues Val498, Leu504, Ser441, Ile437, and Phe461. (C) Gln647 is on the catalytic domain close to the catalytic triad residues Ser559, Asp645, and His690. (D) Cys243 and nearby PREPL residues are found on an exposed surface of the β-propeller domain.

Figure 6. A proposed functional cycle for PREPL incorporating catalytic and noncatalytic activity.

Based on our in vitro results, we propose a functional cycle for PREPL where PREPL first binds a target protein (A), potentially facilitated by a PTM on the target protein. Next, it can follow 2 paths: (a) PREPL cleaves a PTM from the target protein (B), breaking the complex (C) and thereby completing its physiological function. Alternatively, (b) the protein interaction can continue independent of catalytic activity and could expand to bigger complexes (red protein) (D) (noncatalytic function). These complexes can dissociate without the need for catalytic activity or by the removal of a potential PTM by PREPL (E).