Title: Early trigeminal and sensory impairment and lysosomal dysfunction in accurate models of Wolfram syndrome Authors: Kerli Tulva Aleksander Pirajev Akbar Zeb Asya E. Aksoy Azizah Bello Benjamin Lee Baldvin F. Guðjónsson Sigridur B. Helgadottir Toomas Jagomäe Andrea García-Llorca Thor Eysteinsson Monika Jürgenson Mario Plaas Eero Vasar Allen Kaasik Miriam A. Hickey Author for correspondence: Miriam A. Hickey Figure 1 Wfs1 expression in trigeminus and in trigeminal motor and sensory nuclei. A) Demonstration of Wfs1 mRNA expression in WT and Wfs1-deficient trigeminus. Tissue was from 8 m old male mice. Data were normalised to HPRT and actin using the 2–ΔΔCt method. Primer sequences are shown in Table 2 and qPCR was conducted according to MIQE guidelines (Bustin et al., 2009 Clin Chem 2009 Apr;55(4):611-22 doi: 10.1373/clinchem.2008.112797). B) Raw Ct values of Wfs1 mRNA expression in WT and Wfs1 deficient trigeminus (mean ± sem). C) Wfs1 protein expression in normal mouse trigeminus using an anti-Wfs1 antibody (1:400, Proteintech Cat# 26995–1-AP, RRID:AB_2880717). D) Wfs1 protein expression in the normal mouse trigeminal motor nucleus and normal mouse trigeminal sensory nucleus (1:400, Proteintech Cat# 26995–1-AP, RRID:AB_2880717). An image of an adjacent cresyl-violet stained section shows anatomy. Figure 2 Volumetric data from ex vivo magnetic resonance imaging of Wfs1 deficient and wildtype mice. Volumes determined manually by a blinded observed using ITK-SNAP (V3.8.0) from ex vivo magnetic resonance imaging sequences using a 9.4T Bruker BioSpec 94/20 USR system connected to a 1 H circular polarized transceiver coil and running ParaVision 6.0.1® software. P10 mice: Tr, 7600 ms TE, 50.03 ms imaging matrix, 320 x 320 x 50 (x,y,z) spatial resolution, 0.04688 x 0.04688 x 0.25 mm P22 mice: Tr, 5256 ms TE, 33 ms imaging matrix, 320 x 320 x 50 (x,y,z) spatial resolution, 0.04688 x 0.04688 x 0.3 mm 8 m mice: Tr, 900 ms TE, 47.13 ms imaging matrix, 360 x 512 x 80 (x,y,z) spatial resolution, 0.0444 x 0.03 x 0.2 mm A) Total brain volume of P10 and P22 mice. Segmentation began at the most rostral cortex overlying the olfactory areas and ended with the last slice to contain both medulla and cerebellum. B) Total brain volume of mice aged 8 months. C) Optic nerve volume of P10 and P22 mice. Optic nerve segmentation began where the nerves exited the foramina and ended where they joined to form the chiasm. D) Optic nerve + optic chiasm + optic tract volume in mice aged 8 months. E) Volume of brainstem in P10 and P22 mice. Segmentation began at the most rostral portion of the pons, ventral to the interpeduncular nucleus and cerebellar peduncle and ended at the termination of the overlying cerebellum. F) Volume of brainstem in mice aged 8 months. Segmentation began at the most rostral portion of the pons, ventral to the interpeduncular nucleus and cerebellar peduncle and ended at the termination of the overlying cerebellum. G) Volume of cerebellum in P10 and P22 mice. H) Volume of cerebellum in mice aged 8 months. Figure 3 Volumetric data from ex vivo magnetic resonance imaging of trigeminus in Wfs1 deficient and wildtype mice. Volumes determined manually by a blinded observed using ITK-SNAP (V3.8.0) from ex vivo magnetic resonance imaging sequences using a 9.4T Bruker BioSpec 94/20 USR system connected to a 1 H circular polarized transceiver coil and running ParaVision 6.0.1® software. P10 mice: Tr, 7600 ms TE, 50.03 ms imaging matrix, 320 x 320 x 50 (x,y,z) spatial resolution, 0.04688 x 0.04688 x 0.25 mm P22 mice: Tr, 5256 ms TE, 33 ms imaging matrix, 320 x 320 x 50 (x,y,z) spatial resolution, 0.04688 x 0.04688 x 0.3 mm 8 m mice: Tr, 900 ms TE, 47.13 ms imaging matrix, 360 x 512 x 80 (x,y,z) spatial resolution, 0.0444 x 0.03 x 0.2 mm 1 yr mice: Tr, 900 ms TE, 47.13 ms imaging matrix, 360 x 512 x 80 (x,y,z) spatial resolution, 0.0444 x 0.03 x 0.2 mm A) Total trigeminal volume of P10 mice and P22 mice. Trigeminal nerve segmentation began at the most rostral cortex overlying the olfactory areas and terminated when bilateral contact between the trigeminae and the pons was made. B) 3D segmentation of a brain from a 1 yr-old wildtype mouse. C) Total trigeminal volume from 8 m-old mice. Segmentation was from just after the end of the eyes and until bilateral contact was made with the pons. D) Trigeminal volume, per coronal slice, in 8 m-old mice. Segmentation was from just after the end of the eyes and until bilateral contact was made with the pons. E) Total trigeminal volume in 1 yr-old mice. Data set from 1-yr-old mice from Cagalinec et al., 2016, PLOS Biol 14, e1002511 doi:10.1371/journal.pbio.1002511 Segmentation was from just after the end of the eyes and until bilateral contact was made with the pons. F) Trigeminal volume, per coronal slice, in 1 yr-old mice. Data set from 1-yr-old mice from Cagalinec et al., 2016, PLOS Biol 14, e1002511 doi:10.1371/journal.pbio.1002511 Segmentation was from just after the end of the eyes and until bilateral contact was made with the pons. Figure 4 Volumetric data from ex vivo magnetic resonance imaging of trigeminus from wildtype and Wfs1 deficient male rats. Volumes determined manually by a blinded observed using ITK-SNAP (V3.8.0) from ex vivo magnetic resonance imaging sequences using a 9.4T Bruker BioSpec 94/20 USR system connected to a 1 H circular polarized transceiver coil and running ParaVision 6.0.1® software. 1 yr rats: Tr, 900 ms TE, 47.13 ms imaging matrix, 360 x 512 x 80 (x,y,z) spatial resolution, 0.0444 x 0.03 x 0.2 mm 17 m rats: Tr, 900 ms TE, 47.13 ms imaging matrix, 320 x 320 x 160 (x,y,z) spatial resolution, 0.125 x 0.125 x 0.25 mm A) Trigeminal volume per slice at 17 months, from ex vivo imaging. Trigeminae were segmented from the level where the eyes ended, until bilateral contact with the pons. B) Total trigeminal volume at 17 months, from ex vivo imaging. Trigeminae were segmented from the level where the eyes ended, until bilateral contact with the pons. C) Trigeminal volume per slice at 15 months of age, from in vivo imaging. Data set from Plaas et al., 2017, Sci. Rep. 7. doi:10.1038/s41598-017-09392-x. Trigeminae were segmented from eruption from foramina until bilateral contact with the pons. B) Total trigeminal volume at 15 months of age, from in vivo imaging. Data set from Plaas et al., 2017, Sci. Rep. 7. doi:10.1038/s41598-017-09392-x. Trigeminae were segmented from eruption from foramina until bilateral contact with the pons. Figure 5 Behavioural analysis of trigeminal function in 8m-old wildtype and Wfs1-deficient mice. A) Response to an air puff to the side of the head and directed towards the eye. Wfs1 mice showed abnormal responses (slow or absent). B) Response to increasing size of von Frey hairs applied to the side of the muzzle. Wfs1 mice required greater stimuli to respond. Figure 6 Demonstration and quantification of inflammation in Wfs1 deficient trigeminal sensory nucleus at 8 months of age compared with wildtype mice. Two top rows: Example GFAP staining (left column), Hoechst staining for nuclei (middle column) and merged images (right columns) Top row: Wildtype mouse Second row: Wfs1-deficient mouse Third row left: IBA1 staining in a wildtype mouse Third row right: IBA1 staining in a Wfs1 deficient mouse Graphs show the mean gray value of GFAP staining (left) and the mean particle size of IBA1-positive staining (right) in the trigeminal sensory nucleus. Figure 7 Calcium transients and mitochondrial membrane potential (ΔΨM) in wildtype or Wfs1-deficient primary dorsal root ganglia (DRGs) isolated from mixed-sex neonate mice, imaged at 3–4 weeks DIV. Calcium transients were based upon relative fluorescent intensity of Fluo-4 (5 μM, ThermoFisher, prepared with pluronic 4 %). Imaging was with an LSM 780 confocal microscope, using a Plan-Apochromat 20x/0.8 objective with excitation at 488 nm and emission at 499-568 nm. Image specifications were 0.69 x 0.69 x 1 μm (pixel size) or 708.49 μm2, with a pixel time of 3.15 μs. Images were taken every 2.5 s for approximately 7 m, and the initial two minutes were used for baseline. To induce depolarisation, 40 mM (final concentration) KCl was added at approximately 120 s. For mitochondrial membrane potential measurements, cells were incubated with medium containing tetramethylrhodamine methyl ester (TMRM) at a final concentration of 10 nM and incubated for 30 minutes. Single-plane images through the soma were taken using an LSM 780 confocal microscope, at excitation 561 nm and emission 566-669 nm using a 40x/1.3 oil DIC M27 objective with image specifications of 0.05 μm x 0.05 μm per pixel, 1024 × 1024 frame. A) Fluo-4 fluorescence intensity, relative to baseline. Peak values were smaller in Wfs1 deficient soma than in wildtype soma. B) The speed of rise to peak was slower in Wfs1-deficient DRG soma compared with wildtype DRGs. Peaks were reached at 15 s for WT soma and 17.5 s for Wfs1 deficient soma. C) The speed of decline from peak fluorescence was slower in Wfs1 deficient DRG soma compared with WT DRGs. Fitted curves (second order polynomial) are shown in green (wildtype) and red (Wfs1-deficient). D) Peak Flou-4 values were smaller in Wfs1-deficient soma compared with wildtype DRG soma. E) TMRM integrated fluorescence intensity per DRG soma (single optical slice) was reduced in Wfs1-deficient soma. F) TMRM fluorescence intensity per DRG soma (single optical slice) was reduced in Wfs1-deficient soma. G) Photomicrographs of example wildtype and Wfs1-deficient DRG somas, stained with TMRM. Figure 8 Lysosome content and acidity in primary DRG soma isolated from wildtype or Wfs1-deficient neonatal mice, imaged at 3–4 weeks DIV. For lysosome content, medium was removed and replaced with medium containing Lysotracker red (100 nM) and CellTracker green (6 μM). Cells were incubated for 1 h, and then imaged. A 10-μm stack through the middle of each DRG soma, identified based upon CellTracker, was taken using an LSM780 confocal LCI Plan-Neofluar 63x/1.3 Imm Korr DIC M27 (pixel size 0.7 μm × 0.7 μm × 0.397 μm). Excitation and emission were 561 nm (LysoTracker) and 488 nm (CellTracker), and LP575nm and BP 505-550 nm, respectively. For lysosome acidity measurements, medium was removed and cells were treated with medium containing 100 nM Lysotracker red for 55 min in the dark, then 1 μM Lysosensor was added. At 60mins, cells washed and then imaged. Images were taken at x63 (1.4 oil DIC M27) using “line” scan to excite both Lysosensor and Lysotracker at the same time at 405 nm and 561 nm, respectively (512 x 512 pixels, 33.74 x 33.74 μm; 0.07 x 0.07 μm with a Z step of 0.34 μm). Emissions were collected at 410-556 nm (Lysosensor) and 566-691 nm (Lysotracker). A) Lysosomes were smaller in soma of primary DRGs from Wfs1-deficient mice. B) Soma content of lysosomes was reduced in DRGs from Wfs1-deficient mice. C) The proportion of lysosomes that were appropriately acidic was reduced in DRG soma from Wfs1 deficient mice. Photomicrographs First row left 3D-projection of Lysotracker-labelled lysosomes in a DRG soma isolated from a wildtype mouse First row middle 3D-projection of Celltracker-labelled DRG soma isolated from a wildtype mouse First row right Merge Second row left 3D-projection of Lysotracker-labelled lysosomes in a DRG soma isolated from a Wfs1-deficient mouse Second row middle 3D-projection of Celltracker-labelled DRG soma isolated from a Wfs1-deficient mouse Second row right Merge Third row left Single optical plane through the soma of a wildtype DRG showing Lysosensor- and Lysotracker-labelled lysosomes Third row right Single optical plane through the soma of a Wfs1-deficient DRG showing Lysosensor- and Lysotracker-labelled lysosomes Figure 9 Publicly available proteomics data from untreated healthy and WS patient fibroblast-derived neural stem cells (Zmyslowska et al., 2021, Cell Commun. Sign. 19, 116. doi:10.1186/s12964-021-00791-2) were analysed. Using a published library of lysosome proteins (Schröder et al., 2010, PROTEOMICS 10, 4053–4076. doi:10.1002/pmic.201000196), all lysosomal matrix proteins (N = 63) that were expressed by the cells (N = 45), and all lysosomal integral membrane proteins or protein complexes of the lysosome membrane (N = 45) that were expressed by the cells (N = 21), were extracted and compared using Morpheus (https://software.broadinstitute.org/morpheus). The untreated healthy and Wolfram syndrome human neural stem cell proteomic dataset was also compared using the False Discovery rate approach (Two-stage step-up method of Benjamini, Krieger and Yekutieli; desired FDR (q) = 5 %). There were 8047 proteins in the dataset, and 3308 were expressed differentially (q value 0.05 or less). Functional enrichment analysis was performed on these 3308 proteins using g:Profiler (version e111_eg58_p18_f463989d) with g:SCS multiple testing correction method applying significance threshold of 0.05(Kolberg et al., 2023, Nucleic Acids Res. 51, W207–W212 doi:10.1093/nar/gkad347) and with DAVID Bioinformatics. Gene ontology cell component (GO CC) terms primary lysosome (GO:0005766), lysosome (GO:0005764) and lysosomal lumen (GO:0043202) were examined among these differentially expressed proteins. Finally, the hits within the KEGG term "lysosome" that were identified by DAVID were then inputted onto the lysosome KEGG pathway 04142 (“lysosome”) to provide a visual representation of the dysregulated lysosome proteins. A) The majority of matrix lysosomal proteins, based upon Schröder et al., 2010, were reduced in WS patient fibroblast neural stem cells. B) The majority of integral lysosomal proteins, based upon Schröder et al., 2010, were reduced in WS patient fibroblast-derived neural stem cells. C) A separate, functional enrichment analysis of all 3308 differentially expressed proteins in DAVID showed that the KEGG pathway “lysosome” was enriched (1.3 % of module, 1.7 fold enrichment, Benjamini p < 0.01). The 39 identified protein hits that were identified in this pathway were placed in Morpheus for visual representation. D) The majority of 39 identified KEGG pathway lysosome proteins were downregulated (blue), although a minority were upregulated (red). Permission for use of map04142 Lysosome was granted by Kanehisa Laboratories (Kanehisa, 2019 Protein Sci. 28, 1947–1951, doi:10.1002/pro.3715; Kanehisa 2000 Nucleic Acids Res. 28, 27–30, doi:10.1093/nar/28.1.27; Kanehisa et al., 2023, Nucleic Acids Res. 51, D587–D592 doi:10.1093/nar/gkac963). Table 1 T2 RARE sequences used for MR imaging. Table shows magnetic resonance imaging specifications. Table 2 Primers used for qPCR. Table shows the primer sequences used for qPCR. qPCR adhered to MIQE guidelines (Bustin et al., 2009 Clin Chem 2009 Apr;55(4):611-22 doi: 10.1373/clinchem.2008.112797). Table 3 Brain volumetric analysis of 8-month-old WT and Wfs1 deficient male mice. Table shows mean total volume of different brain regions, determined from ex vivo magnetic resonance imaging. Confidence intervals, effect sizes and p values, where appropriate, are also shown.