TDP-43 dysregulation impairs cholesterol metabolism linked with myelination defects

doi:10.21203/rs.3.rs-6256163/v1

Download PDF

Article

TDP-43 dysregulation impairs cholesterol metabolism linked with myelination defects

https://doi.org/10.21203/rs.3.rs-6256163/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 04 Sep, 2025

Read the published version in Acta Neuropathologica →

You are reading this latest preprint version

TDP-43 is a nuclear protein encoded by the TARDBP gene, which forms pathological aggregates in various neurodegenerative diseases, collectively known as TDP-43 proteinopathies, including amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). These diseases are characterized by multiple pathological mechanisms, with disruptions in lipid regulatory pathways emerging as a critical factor. However, the role of TDP-43 in the regulation of the brain lipid homeostasis and the potential connection of TDP-43 dysfunction to myelin alterations in TDP-43 proteionopathies remain poorly understood, despite the fact that lipids, particularly cholesterol, comprise nearly 70% of myelin. To investigate the causal relationship between TDP-43 dysfunction and disruptions in brain cholesterol homeostasis, we conducted multi-omics analyses (lipidomics, transcriptomics, and functional splicing) on the frontal cortex from the Tardbp^M323K/M323K knock-in mouse model. Lipidomic analysis revealed alterations in lipid pathways related to membrane composition and lipid droplet accumulation, particularly affecting cholesterol-related species. We found higher lipid droplets accumulation in primary fibroblast derived from these mice, as well as in the brain of the mutant mice. Similarly, the immunohistochemical detection of a lipid droplet marker was higher in the post-mortem frontal cortex, gray and white matter, of FTLD-TDP patients compared to non-neurological controls. Transcriptomic analyses showed that TDP-43 pathology led to transcriptional dysregulation of genes essential for myelin production and maintenance. We identified impaired cholesterol metabolism, mainly through the downregulation of endogenous cholesterol synthesis, alongside upregulated cholesterol transport pathways, which we further replicated in FTLD-TDP patients transcriptomic datasets. Collectively, our findings suggest that TDP-43 dysfunction disrupts brain cholesterol homeostasis, potentially compromising myelin integrity.

TDP-43 proteinopathies

cholesterol metabolism

lipidomics

transcriptomics

myelin

lipid droplets.

TDP-43 proteinopathies are a group of neurodegenerative disorders characterised by the mislocalisation and aggregation of transactive response DNA-binding protein 43 (TDP-43) in the cytoplasm of neurons and glial cells. TDP-43, a highly conserved RNA/DNA binding protein, plays a critical role in RNA metabolism, including transcriptional regulation, RNA splicing, transport and stress granule formation [3, 43]. Its pathological accumulation leads to loss of nuclear function and toxic gain of function in the cytoplasm, contributing to neurodegeneration. TDP-43 proteinopathies are most commonly associated with amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD), where TDP-43-positive inclusions are found in affected brain regions such as the motor cortex, spinal cord, and frontal and temporal lobes. Classically, the degeneration of grey matter, atrophy or loss, has been primarily associated with these disorders. In recent years, more studies are showing abnormalities, even reduction, of white matter in TDP-43 proteinopathies, as in the case of ALS [9, 32, 37] and FTD [31, 33]. These abnormalities include axonal degeneration, myelin loss and gliosis, further contributing to the cognitive and motor impairments observed.

Nearly 40% of the brain is white matter, which are highly packed fibres with myelin as the central component. The myelin sheets are lipid rich membranes that isolate the axons of neurons essential for the correct axonal electrical conduction in the brain. The main lipids that make up the myelin are cholesterol, phospholipids, and glycosphingolipids (galactocerebroside) in ratios of ≈ 4:4:2 [25]. It has been estimated that up to 70% of the brain cholesterol is associated with myelin [45]. Cholesterol tight balance regulation is therefore essential for the proper functioning of myelin and brain function. For this reason, normal cholesterol biosynthesis is necessary for the maintenance of oligodendrocytes and is associated with myelin gene expression in vitro and myelin biosynthesis [22, 30, 39].

In TDP-43 proteinopathies, significant alterations in cholesterol metabolism and lipid profiles have been observed in both human patients and animal models. Lipidomics analyses of the frontal cortex of frontotemporal lobar degeneration with TDP-43 pathology (FTLD-TDP) patients had shown a marked disruption of the lipid homeostasis, in particular a reduction in key lipid species such as ether lipids and acylcarnitines, which are essential for maintaining myelin integrity and neuronal function [2, 14]. Reduction in phospholipids were accompanied by an increase in cholesterol ester (CE) and triglycerides in human frontal cortex as well as in neurons derived from C9orf72 iPSC patients [17].

In recent years, more studies are supporting the role of TDP-43 in cholesterol homeostasis. TDP-43 binds the mRNA of the sterol regulatory element-binding protein 2 (SREBP2), a key transcription factor that regulates cholesterol biosynthesis, in oligodendrocytes. Both alterations of TDP-43 levels (depletion and overexpression) lead to reduced SREBP2 and cholesterol transport LDLR expression and activity, resulting in reduced cholesterol production in the spinal cord of mice and patients [12, 20]. This is important because cholesterol is a vital component of neuronal membranes and myelin, and its disruption has been linked to axonal degeneration and cognitive impairment [7, 34]. In addition, oligodendrocytes showed increased susceptibility to oxidative stress due to altered iron and lipid metabolism in the presence of TDP-43 pathology, exacerbating white matter degeneration [5, 21]. Disrupted cholesterol homeostasis in tissues has been detected in the form of cholesterol-ester (CE) in abnormal accumulation in lipid droplets inside the cells. Indeed, lipid droplets accumulations have been found in the spinal cord of SOD1^G93A mice [11], as well as in astrocytes, neurons of TDP-43 transgenic mice and IPS derived motor neurons from C9orf72 ALS/FTD patients, further supporting the role of cholesterol dysregulation in TDP-43 proteinopathies [1, 27, 36, 44].

We have previously characterised a mouse model of TDP-43 proteinopathy, the Tardbp^M323K/M323K mice, which harbours a physiological point mutation in the C-terminal domain of the TDP-43 protein [16, 18]. This mutation recapitulates aspects of human neurodegenerative diseases associated with TDP-43 dysregulation. In this particular model, both cognitive and motor impairments manifest early in life and progressively worsen with age, providing a dynamic system to study the course and molecular mechanism of TDP-43-related alterations. From this model we learned that TDP-43 is clearly crucial for the correct brain development. The Tardbp^M323K/M323K mice have cognitive impairments, including learning and memory deficits with brain structural changes, particularly in regions associated with white matter integrity, including a reduction in the white matter volume, observed by structural MRI [18].

Here, we investigated the role of TDP-43 in the brain cholesterol homeostasis, which impacts on myelin integrity and in the correct neuronal functional activity. We used the Tardbp^M323K/M323K mice and FTLD-TDP patient samples to provide valuable insights into the relation of TDP-43 dysfunction and brain cholesterol dysregulation TDP-43 proteinopathies.

Mice

The Tardbp^M323K/M323K mice were generated in a constant F1 DBA-C57BL hybrid background, as previously described [18]. For all the experiments we use the wild-types (Tardbp^+/+) and homozygous littermates (Tardbp^M323K/M323K), and both sexes.

Mice were kept in auto-ventilated cages and fed “ad libitum” (Rat and Mouse Breeding 3 (RM3), Special Diet Services) with free access to water. Mice were kept under constant conditions with a regular 12 h light and dark cycle, temperature of 21 ± 2 ◦C and humidity 55 ± 10%. Mice were housed in same sex cages of up to five mice. Cages contained Aspen Chips bedding (Datesand), shredded paper and a rodent tunnel for enrichment. Separated male and female cohorts were used in most of the analyses unless specified. All mice were maintained and studied according to international, national and regional guidelines. The study was approved by the local ethical committees of animal care and use of the Hospital Clínico San Carlos (Proex 258.4/20) in accordance with the European and Spanish regulations (2010/63/EU and RD 1201/2005).

At weaning, all mice were genotyped by PCR from an ear biopsy, using a multiplexed PCR assay (two primer/probe pairs, Supplementary Table 1).

Human brain samples

Human tissue blocks were provided by Biobanks BT-CIEN (Madrid) and IdISSC. Control patients had no history of neurological disease or with a cause of death unrelated to any neurological condition, including three women with an age range of 58–72 years and four men from 43 to 65 years of age. FTLD-TDP patients included two women with an age range of 80–84 years and five men from 54 to 75 years of age.

Table 1

Patient demographics data.
	Sex	Age	Dx NP	Sample type
1	Men	65	Control	N1 middle frontal gyrus
2	Men	64	Control	N1 middle frontal gyrus
3	Women	72	Control	N1 middle frontal gyrus
4	Men	58	Control	N1 middle frontal gyrus
5	Women	58	Control	Fixed prefrontal cortex
6	Women	61	Control	Fixed prefrontal cortex
7	Men	43	Control	Fixed prefrontal cortex
8	Men	73	FTLD-TDP	Fixed prefrontal cortex
9	Women	84	FTLD-TDP	Fixed prefrontal cortex
10	Men	75	FTLD-TDP	Fixed prefrontal cortex
11	Men	68	FTLD-TDP	Fixed prefrontal cortex
12	Men	64	FTLD-TDP	Fixed prefrontal cortex
13	Women	80	FTLD-TDP	Fixed prefrontal cortex
14	Men	54	FTLD-TDP	Fixed prefrontal cortex

Immunohistochemical Processing of Patients Brain Tissue

Paraffin-embedded brain tissues were sectioned transversely at a thickness of 4 µm. The sections were subsequently deparaffinized and subjected to antigen retrieval using a citrate buffer. Following this, sections underwent three washes with PBS 1x for 5 minutes each. Endogenous peroxidases were blocked for 2 hours using a solution containing 10% methanol, 2% H₂O₂, and PBS 1x, adjusted with distilled water. After three additional washes with PTB buffer (PBS 1x, 0.02% Triton X-100, and Bovine Serum Albumin [BSA]), sections were washed three times with PBS 1x for 5 minutes each. Autofluorescence was quenched using TrueBlack® Lipofuscin Autofluorescence Quencher, followed by three more washes with PBS 1×. Non-specific epitopes were blocked for 90 minutes using 5% Normal Goat Serum and 0.01% BSA in PBS 1x.

Tissue sections were incubated for 5 days at 4°C with primary antibodies against PLIN2 (1:200, PA5-29099, Thermo Fisher) and IBA1 (1:400, AB300157 Abcam) diluted in PBS 1x with 0.01% BSA. After three 5-minute washes with PBS, sections were incubated overnight with Alexa Fluor 594- and 488-conjugated respectively secondary antibodies (1:400 dilution, Thermo Fisher).

Mouse Tissue Immunostainings

Mice were terminally anaesthetised with 0.4 mg/kg of fentanest and 40 mg/kg of thiopental and cardiac perfused with the fixative 4% PFA without methanol in PBS 1x, pH 7.4. The dissected organs were post-fixed in the same fixative overnight at 4°C and kept on PBS 1x with sodium azide 0.1% until use. Whole brains were cryopreserved by immersion in 30% sucrose. Brains were dried and embedded in OCT using isopentane and dry ice, and frozen at − 20°C until they were processed with a cryostat (Leica). Coronal cryo-sections at 30 µm were collected free-floating through the brain frontal cortex. Free-floating sections were then blocked with PTB (PBS 1x, Triton 0.02% and BSA) and then incubated with the primary antibodies PLIN2 (1:300, PA5-29099, Thermo Fisher) and PLP1 (1:400, AB254363, Abcam) in PBS with Triton X-100 0.02% and the secondary antibodies (Alexa-Fluor, ThermoFisher), and mounted together with Hoechst stain to detect nuclei. Z-Sections images were taken using a confocal microscopy Olympus Fluoview FV1000 v.3.1.1.9. Quantifying at least three to five photos that contains at least 300 cells in the case of PLIN2 staining with Imagej v.1.53q.

Primary ear fibroblasts generation

Mice ears were removed and disinfected with 70% ethanol for 10 minutes, followed by three washes in PBS under a laminar flow hood. The ears were minced and placed in a sterile cryovial with 1-1.5 ml of a filtered enzyme solution containing collagenase D (Worthington) (2.5 mg/ml) and pre-activated pronase (Roche) (Pronasa 20 mg/ml, Tris pH 8 10mM, EDTA 1mM), and activated at 37º C for 30 min in complete RPMI medium (RPMI, 10% FBS, 2.5% HEPES and 1% antibiotic-antimycotic). The suspension was incubated for 95–100 minutes shaking (300 rpm) at 37º C. After enzymatic digestion, the cells were disaggregated and passed through a 70 µm sieve into 20 ml of complete RPMI medium. The cell suspension was centrifuged at 500 g for 10 minutes and the pellet was resuspended in 10 ml of medium, filtered again and supplemented with a further 20 ml of medium before a second centrifugation. The final pellet was resuspended in 10–15 ml RPMI medium and settled in a cell culture flask. The next day, the medium was replaced by DMEM (Thermo Fisher) (DMEM high glucose, 10% FBS, 1% antibiotic-antimycotic, GIBCO) and the cells were cultured for 5 days, until they reached 80–90% confluency.

Flow Cytometry for LD analysis

Primary fibroblast from wild-type and homozygous mice were cultured in 6-well plates for at least 24 hours and then we performed a treatment with oleic acid 250 µM for 24 hours. Following treatment, cells were stained with BODIPY^493/503 (1/1000 dilution, Thermo Fisher Scientific) and incubated for 1 hour and 30 minutes. After incubation, cells were washed with PBS and detached using trypsin for 3–5 minutes at 37°C. The reaction was then halted by adding culture medium, and cells were centrifuged at 1000 rpm for 5 minutes. The supernatant was discarded, and the pellet was resuspended in PBS. This washing step was repeated once more by centrifugation at 1250 g for 5 minutes, followed by resuspension in 300 µL of PBS. Flow cytometry analysis was performed using a FC500 flow cytometer (Beckman Coulter), and results were processed using FlowJo Software (BD Life Sciences) v.10.8.1.

Fibroblasts Immunostainings

Following the 24 hours treatment, the fibroblasts were washed with PBS and fixed with 4% (w/v) paraformaldehyde without methanol in PBS for 15 min at room temperature. For the LD staining, fibroblasts were firstly incubated with the primary antibody of SREBP2 (1/200, PA1-338, Thermo Fisher Scientific) overnight at 4°C. After 3 washes with PBS, the cells were incubated for 1 h at room temperature with the secondary antibody conjugated with fluorophores (1/1500 dilution, Alexa fluor 594 and 647, Thermo Fisher Scientific). Nuclei were stained with HOESCHST 33342 (1/3000 dilution, 11504886/H21492, Fisher scientific) and neutral lipids from LDs were stained with BODIPY^493/503 (1/1000 dilution) for extra 30 min. A final 3 washes with PBS were done to remove the excess, and the coverslips were mounted on a slide with FluorSave Reagent (Merck). Images were taken using a confocal microscopy Olympus Fluoview FV1000 with a version system 3.1.1.9., quantifying at least five photos that contains at least 50 cells with Imagej v.1.53q.

Western blot

Fresh dissected tissues were snap frozen and stored at − 80 ◦C until required. Frontal cortex from 1 hemisphere were homogenised using mechanical disaggregation with beads (Precells) in RIPA buffer (Thermo Fisher) at 4 ◦C with protease and phosphatase inhibitors cocktail (Roche). After 20 min of centrifugation at 13000g at 4 ◦C, the supernatant was collected, and the total amount of protein was quantified using the DC Assay method (BioRaD). 30 µg of total protein was loaded for each of the samples in a 10% bis-acrylamide precast gel (GenScript, US) under reducing conditions, run and transferred into a PVDF low fluorescence membrane (GE healthcare). The membranes were blocked for 1 hour and incubated with the primary antibodies overnight: SREBP2 (1/200, PA1-338, Thermo Fisher Scientific), MOG (1/1000, AB243034, Abcam) and PLP1 (1:2000, AB254363, Abcam). Secondary antibodies labelled for infrared detection and scanned in the infrared scanner Clx (LiCor). The total protein detection kit was used following manufacturer’s instructions (LiCor) for the loading correction. The intensity of the bands was quantified using the image software Image Studio v.5.2 (LiCor).

Gene expression analysis

RNA was extracted from frozen frontal cortex from 1 hemisphere using the RNeasy Lipid Tissue Mini Kit (Qiagen) and DNAsa treatment (Qiagen). cDNA synthesis was performed using the High Capacity cDNA Reverse Transcriptase Kit (Thermo Fisher Scientific) from 1 µg of total RNA. cDNA for qPCR reactions was used at a final concentration of 25 ng per well. Power Sybr Green mastermix (ThermoFisher Scientific) was added, followed by the appropriate pair of primers in a final well volume of 10 µl. A list of all primers used can be found in Supplementary Table 1.

Lipidomic

Snap frozen frontal cortex from one hemisphere from Tardbp^+/+ and Tardbp^M323K/M323K male mice (n = 6 per group) at 9 months of age were used for the lipidome analysis. Briefly, brain tissues we treated using the modified Bligh and Dyer method for lipid extraction. The homogenates were vortexed for 15 seconds, then incubated in 4°C on a mixer (300 rpm) for 1 h. After agitation, 300 µl of ice-chilled chloroform and 250 µl of ice‐chilled MQ water were added sequentially, followed by vortex for 15 s, and centrifugation at 9,000 rpm for 2 min to separate the phases. Bottom organic phases were collected, and the aqueous phases were re‐extracted with 500 µl of chilled chloroform. Collected organic phases were dried in a vacuum concentrator and stored lyophilized at − 80°C.

Samples were then solubilized in 100 µl chloroform/methanol (1:1; v/v) and analyzed in reverse phase using an Acquity UPLC HSS T3 1.8 µm column with the following conditions: mobile phase A (water:acetronitrile, 40:60, with 10 µM ammonium acetate and 0.1% acetic acid), B (water:acetonitrile:isopropanol:acetic acid, 5:10:85:0.1, with 10 µM ammonium acetate and 0.1% acetic acid); flow rate of 300 µl/min; injection volume of 5 µl; column temperature at 55°C; 20% B for 1.5 min; linear change to 100% B over next 16.5min; and maintained at 100% B for 3 min (77). Then, the gradient was reverted back to initial state 20% B for 1 min then held for next 1 min at 20% B. QC samples were injected prior to, and after every 5 samples, to monitor the stability of the instrument. Samples were run under untargeted positive and negative electrospray ionization (ESI) modes in a data-independent manner (MSE mode). The following ESI conditions were used: for positive, capillary at 2 kV, sampling cone at 35 V, source temperature at 100°C, desolvation gas at 500 l/h, and nebulizer at 6.5 bar; and for negative, capillary at 2.2 kV, sampling cone at 40 V, source temperature at 80°C, desolvation gas at 500 l/h, and nebulizer at 6.5 bar. For lock mass correction, leucine enkephain was used at 1 ng/ml in acetonitrile/water (1:1, v/v) with 0.1% formic acid and at a flow rate of 10 µl/min. The low collision energy was set to 4 eV and high collision energy was set between 25 and 40 eV for both positive and negative modes. Pooled samples were run at first and every after 6 samples as QC. Raw data were converted into ABF format using Reifycs Analysis Base File Converter, then used in MS‐Dial (v. 4.9) for peak picking and retention time alignment using default settings. Lipid species were manually verified and named using Lipid Maps abbreviations. The intensities were initially normalized with total ionic current. The corrected readings of identified species were exported into R (v. 4.4.2) to do the batch correction using pooled samples as QC, then calculate the concentrations using the known concentration of spiked internal standards (Avanti Polar Lipids, Alabaster, AL). Lipid levels for each sample were calculated by summing up the total number of moles of all lipid species measured by all three LC-MS methodologies and then normalizing the total to mol %.

For the lipidomic statistical analysis (single factor) we used MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/ accessed on 3 February 2025). Normalisation by sum and Pareto scaling were used. Traditional univariate methods (volcano plot), multivariate statistics (partial least squares discriminant analysis (PLS-DA)) and clustering (heat map) were provided. LION was used to identify altered lipid functions and organelle associations (https://www.lipidontology.com/ accessed 5 February 2025).

RNA-sequencing

RNA was extracted from the frontal brain from one hemisphere of Tardbp^+/+ and Tardbp^M323K/M323K male mice at 9 months of age (n = 4 per group). Quality and quantity were assessed using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies). Sequencing libraries were generated using an NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB) following the manufacturer’s recommendations, and index codes were added to attribute sequences to each sample. The clustering of the index-coded samples was performed on a cBot Cluster Generation System using PE Cluster Kit cBot-HS (Illumina) according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina platform, and 150 bp paired-end reads were generated. Library preparation and RNA-seq was carried out by Novogene. Quality control of FastaQ files was performed using FastQC (https://www.bioinformatics.babraham.ac.uk/projects). Low-quality reads (Phred quality score < 30) and reads too short (length < 15 pb) were removed using Fastp [6]. The alignment to the genome (10 mm mouse reference genome) was achieved using HISAT2 [24].

DEG Expression analysis of the transcriptome

The expression quantification of genes was carried out using FeatureCounts [26]. Only uniquely mapped reads were used for the analysis of differential gene expression quantification with DESeq2 [28]. Raw p-values were adjusted by the Benjamini–Hochberg false discovery rate (FDR) method and the adjusted p-values or p-value less than 0.05 were considered statistically significant. Functional enrichment was done by the Over-Representation Analysis (ORA)[23] method, which performs a statistical evaluation of the fraction of genes in a particular pathway found among the set of genes. A protein–protein interaction network was constructed using STRING[42] from the different comparation we did and the DEGs found. The following parameters were selected: network type: full STRING network; required score: medium confidence (0.4); FDR stringency: medium (5 percent).

Splicing analysis

After alignment of RNA-seq reads by HISAT2, alternative splicing events were identified and quantified by rMATS-turbo 4.3.0 using gene transcript annotations from gencode mus musculus version mm10 mouse reference genome. rMATS was downloaded from the open-source platform SourceForge (https://rnaseq-mats.sourceforge.io/, accessed on 10 December 2023)[40] and rmats2sashimiplot was used to convert the rMATS output into Sashimi-plot. Including novelSS events. The cut-off criteria for statistical significance were adjusted p-value < 0.05 and |IncLevelDifference| > 5%. Venn diagram was used in splicing and transcriptomic analysis to determine which genes o genes with splicing events was common between two or more list of genes [51].

Analysis of RNA sequencing database from FTLD patients and controls

The data used in this study of the frontal cortex transcriptomics from FTLD-TDP patients were obtained from the supplementary file available in this article published in Brain [35]. The data were downloaded and using RStudio v4.3.3., we performed the analysis of differential gene expression quantification with DESeq2 of the frontal cortex samples comparing FTLD type A, B or C with the control patients. The different FTLD patients were categorized as follows: Type A is often associated with GRN mutations, presenting with bvFTD and sometimes non-fluent variant PPA. Type B is linked to patients with C9orf72 expansions, characterized by bvFTD with ALS. Type C, associated with TDP-43 pathology, common in patients without MAPT mutations and is strongly associated with semantic variant PPA. Finally, we selected the fold change value of the desired genes and were plotted in GraphPad.

Table 2

Patients demographics data for the transcriptomic analysis
	Age	N total	Sample type
Control	28–100	24	Frontal cortex
FTLD-TDP-A	67–94	24	Frontal cortex
FTLD-TDP-B	49–97	20	Frontal cortex
FTLD-TDP-C	61–83	22	Frontal cortex

Statistical analyses

Statistical analysis was conducted using GraphPad Prism. Two groups were compared with a single time point using Student’s t-test. Body weights between genotypes, across multiple time-points, were compared using 2-way ANOVA repeated measures/mixed models and Bonferroni correction of multiple testing. Two groups were compared across multiple time points using 2-way ANOVA with Sídak’s multiple comparisons post hoc test. Three or more groups were compared at a single time point using one-way ANOVA with Dunnett’s post hoc test. Statistical analysis of qRT-PCR data was performed on ΔCT values. 2- way ANCOVA (SPSS) was used to correct the energy expenditure data for the lean mass interaction. Statistical significance was defined as p ≤ 0.05 for analysis of phenotyping and molecular biology data. Please see Fig. legends for sample size - n numbers; n numbers refer to biological samples (i.e. number of animals used in animal experiments). Statistical detail for each experiment can be found in the Fig. legends. Where indicated, * = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001. All graphs were generated using GraphPad Prism 8. BioRender.com was used to create original diagrams/Figs.

Tardbp ^M323K/M323K mice display alterations in myelin proteins composition in the frontal cortex

We have previously shown that the M323K mutation in TDP-43 affected the myelin formation and its progressive stability in the homozygous mice[18] macroscopically. Here, we further evaluated the impact of the M323K mutation on myelination at the molecular level conducting an analysis of the expression levels of key myelin-related proteins and genes in the frontal cortex of wild-type and homozygous mutant mice at different ages. The myelin associated protein PLP1 (Proteiolipid protein 1) immunofluorescence staining revealed a significant reduction in the frontal cortex of the Tardbp^M323K/M323K mice compared to their Tardbp^+/+ littermates (Fig. 1a). Western blot analysis confirmed the reduction in PLP1 protein levels in the Tardbp^M323K/M323K frontal cortex, which significantly declined in the Tardbp^M323K/M323K mice at 12 months compared to the wild-type controls (Fig. 1b), as well as a subtle decrease in the expression levels of MOG, another myelin-associated protein (Supplementary Fig. 1a). The progressive reduction of these myelin associated proteins supports the potential effect of TDP-43 mutation in impairing myelin maintenance. Next, we looked at the expression level of key oligodendrocytes and myelin-related genes in the frontal cortex of mutant mice. The expression levels of the genes Plp1, Mbp, Qki, and Myrf were significantly downregulated in Tardbp^M323K/M323K mice compared to their Tardbp^+/+ littermates, whereas Olig2 expression remained unchanged (Fig. 1c and Supplementary Fig. 1b). This could indicate that TDP-43-M323K mutation could be interfering downstream to the oligodendrocyte precursor cells (OPCs) production, affecting the transcriptional regulation of genes essential for myelin maintenance.

Disrupted lipids landscape in the frontal cortex of the Tardbp^M323K/M323K mice

We hypothesized that the myelin alterations observed might, at least partly, be driven by disruptions in lipid metabolism, and that TDP-43 dysfunction could be impairing the regulation of brain lipid homeostasis. To this end, we performed a targeted lipidomic analysis of the frontal cortex of Tardbp^M323K/M323K and wild-type littermate mice (n = 6) at 9 months of age. Partial Least Squares Discriminant Analysis (PLS-DA) (Fig. 2a) showed that 44% of the variability between samples was due to the mutation, and the metabolites that contributed most to this component (VIP > 1) (Supplementary Fig. 2a) were cholesterol, cholesterol ester (CE) species and some dyhidrosphingomyelin species. We found a total of 14 metabolites differentially expressed in the frontal cortex of mutant mice compared to wild-type controls, with log2 fold-change (log2 (FC)) values between − 1 and + 1 and -log10 (p value) > 1. A total of 10 metabolites (LPI 20:2, LPI 20:3, PG 36:4, CE 22:2, CE 16:1, CE 18:1, PEp 32:0, AC C18:0, Cer d18:1/18:0 and GM3 d18:1/18:0) were up-regulated and 4 down-regulated (MhCer d18:1/20:1, NAPS 16:0–38:5, DG 38:2/18:1 and PC 36:0) (Fig. 2b). Among the top significant hits, we identified lipid species of three main families, i.e. the neutral lipids (CE, DG and AC), sphingolipids (Cer, Mhcer and GM3) and phospholipids (Pep, PC, PG, LPI and NAPS), highlighted in the volcano plot (Fig. 2b).

Looking closer at the particular perturbation of the major lipid classes by fold change in the mutant brains compared with the wild-type, the redistribution of lipids suggested changes possibly related to alterations in membrane composition (sphingolipids) and lipid droplet formation (TGs and CEs) (Fig. 2c). These results were more clearly illustrated when we examined the 25 most altered lipids, based on fold change, in a heatmap (Supplementary Fig. 2b), showing that triglycerides and cholesterol esters were the most increased in the mutant mice compared to their wild-type littermates. To investigate this further, we performed a Lipid Ontology (LION) enrichment analysis, which concordantly revealed an increase in lipid droplets, likely driven by elevated TG and CE, and a reduction in membrane components, associated with a reduction in free cholesterol (FC) and changes in sphingolipids (Fig. 2d).

Cholesterol esterification is a process that increases the hydrophobicity of this lipid, thereby facilitating its storage or transport. This mechanism prevents cellular toxicity caused by high levels of free cholesterol in the membrane. From the lipidomic analysis, we determined the rate of cholesterol esterification, the fractional cholesterol esterification rate (FCE, as the ratio between CE and FC). An increase in this ratio was observed in the mutant mice compared to the wild type (Fig. 2e). There was also a significant decrease in CE compared to membrane phospholipids (the ratio between CE and glycerophospholipids (PL)) (Fig. 2e), which may be indicative that there was a reduction in cholesterol levels in cellular membranes in the Tardbp^M323K/M323K mice.

We looked deeper into the regulation of cholesterol biosynthesis in the frontal cortex of these mice in particular through SREBP2, which is retained in the endoplasmic reticulum (ER) in an inactive form bound to SCAP and INSIG proteins. When cholesterol levels drop, SCAP escorts SREBP2 to the Golgi, where it undergoes proteolytic cleavage, releasing its active fragment. This active form translocate to the nucleus, where it binds to sterol regulatory elements (SREs) in target gene promoters, inducing the expression of genes involved in cholesterol biosynthesis ( HMG-CoA reductase (HMGCR), HMG-CoA synthase (HMGCS), and mevalonate kinase (MVK) and DHCR24). SREBP2 activation was measured by the ratio of the nuclear fragment to precursor SREBP2 protein. The ratio of active SREBP2 seems to decrease at 12 months of age in the brain of the Tardbp^M323K/M323K mice compared to their Tardbp^+/+ littermates, which could explain the decreased amounts of free cholesterol observed in the mutants brains.

Altogether, the TDP-43 M323K mutation caused a clear dysregulation of lipids, in particular an imbalance of cholesterol-related species in the frontal cortex of mice, that was an increase in the storage of CE and TG in the form of lipid droplet, while reducing the synthesis of FC and the dysregulation of other lipids involving membrane components (sphingolipids and phospholipids).

Altered TDP-43 Induced accumulation of Lipid Droplets (LD) in cells and in the Frontal Cortex of Mice and FTLD-TDP patients.

The lipidomic analysis of the frontal cortex from the mutant TDP-43 mice revealed an increase in the neutral lipids CE and TGs, which are normally accumulated in the cells in the form of lipid droplets (LD). Consequently, we aimed to investigate whether alterations in the TDP-43 protein, under various conditions and across different tissues in both mice and humans, supported these findings.

Since the analyses of lipid droplets is very challenging in neuronal tissue, as well as being a dynamic process, we first investigated the impact of the TDP-43 M323K mutation directly in the process of lipid droplets intracellular accumulation in a cell-autonomous system using primary cells derived from these mice. Thus, we generated primary fibroblasts from the ears of 3-month-old Tardbp^+/+ and Tardbp^M323K/M323K mice and challenged them with oleic acid to evaluate if the mutation could directly intervene in the cell lipid storage capacity. We stained the lipid droplets with the marker BODIPY and quantified the positive staining by confocal microscopy (Fig. 3a). There were no differences between the two groups at baseline; however, after a 24-hour challenge with oleic acid, the number of lipid droplets per cell, stained with BODIPY, was higher in cells from the Tardbp^M323K/M323K mice compared to those from the Tardbp^+/+ mice. We corroborated these results quantifying the BODIPY fluorescence in the fibroblasts by flow cytometry and confirming that the ratio of Mean Fluorescence Intensity (MFI) obtained between oil-treated and BSA-untreated cells was higher in Tardbp^M323K/M323K cells (Fig. 3b).

Next, we sought to confirm whether this alteration in LD levels could be observed in the brain of the mice. Direct staining of lipid droplets using BODIPY in the brain, which is extremely rich in lipids, is not a reliable and sensitive technique. Thus, we performed immunohistochemical staining of one of the proteins that coats LD in the nervous system, the protein PLIN2 (perilipin 2), in the frontal cortex of 3-month-old mice. PLIN2 assists with the storage of neutral lipids in the LDs, and is highly expressed in neurons and greatly affected by age [10]. Even at early ages, the mutant TDP-43 protein induced a higher amount of PLIN2 protein in the frontal cortex (Fig. 3c), which denoted increased LD accumulation in those mutant brains, as expected from the lipidomic analysis (Fig. 2d).

Furthermore, we evaluated whether the increased LD accumulation was a specific effect of the particular M323K mutation in TDP-43 or whether it is a more general effect of TDP-43 alterations, as in the case of the FTLD-TDP cases without a TARDBP mutation. Hence, we evaluated the staining of PLIN2 in post-mortem frontal cortex white and grey matter tissue from FTLD-TDP patients and non-neurological diseased age-sex matching controls (n = 7 per group) (Table 1). The PLIN2 staining in the post-mortem human tissue was not only intracellular, so we performed a co-staining with the cellular microglia marker IBA1 with PLIN2, both in white and grey matter. The quantification of the microglial-intracellular PLIN2 was higher in both the grey and white matter of the frontal cortex of FTLD-TDP patients compared to the controls (Fig. 3d).

In summary, these results demonstrated that alterations in TDP-43 resulted in increased LD formation in the cortex of mutant mice, which was consistent with the lipidomic data obtained. This effect was also observed in the postmortem frontal cortex of FTLD-TDP patients without a TARDBP mutation. Importantly, our findings showed that the TDP-43 influence on LD formation was TDP-43-dependent, as confirmed through a lipid storage challenge in a cell-autonomous system, rather than simply being a consequence of the neurodegenerative disease.

Disrupted cholesterol metabolism in patients and mice with TDP-43 proteinopathy

To identify which genes and pathways were involved in the myelin and cholesterol homeostasis alterations identified, we performed a transcriptomic analysis by RNA-sequencing of the frontal cortex from Tardbp^+/+ and Tardbp^M323K/M323K 9-months-old male mice (n = 4 per group, using the other half frontal cortex to the one used for the lipidomics). A total of 226 differentially expressed genes (DEGs) using the threshold of p-adjusted < 0.05 were found in the cortex of the mutant mice (Supplementary Table 2). The gene ontology (GO) analysis by ORA (p-adjusted < 0.05 and fold change < 0) showed that the myelination process was significantly downregulated, further confirming our findings (Fig. 4a).

In order to elucidate how the alteration on TDP-43 led to dysregulation of lipid metabolism, and in particular, to the dysregulation of the endogenous cholesterol homeostasis, we looked at the consequence of the dysfunctionality of TDP-43 in RNA metabolism and splicing regulation of lipid regulatory genes. To specifically looked for the lipid regulatory genes dysregulated in the mutant brains, we used a list of all genes that were annotated in the genome within the category of lipid metabolism (1586 lipid genes), as previously done [15] (Supplementary Table 3), to perform a targeted analysis within the total of 1939 DEGs (using threshold of p value < 0,05) (Fig. 4b). A total of 124 lipid genes (56 of them were targets of TDP-43) were deregulated (Supplementary Table 4), and after a STRING pathway analysis, those genes showed to be involved in membrane lipid metabolic process, cholesterol metabolic process and myelination (Fig. 4c), further supporting our previous results. The gene ontology (GO) analyses of those selected 124 lipid genes dysregulated in the mutant frontal cortex indicated an upregulation of cholesterol transport and a downregulation of cholesterol synthesis (Fig. 4d). We validated the gene expression level by quantitative PCR with a different set of biological samples using 9-month-old males, and also, 3-month-old females, to discern the potential age and sex effect. In all of them, we confirmed the downregulation of regulatory genes of the cholesterol biosynthesis (Srebf2, Dhcr24 and Sqle) (Supplementary Fig. 3a), as well as the upregulation of some of the genes of the cholesterol transport (Supplementary Fig. 3b).

Since one of the main functions of TDP-43 is the regulation of RNA splicing, we investigated whether specific cholesterol regulatory genes, and also myelin regulatory genes, could also have modifications in their splicing as consequence of the dysfunction of TDP-43. We performed a splicing analysis of the RNA sequencing data from the mutant and wildtype mice frontal cortex. A total of 2288 genes showed significant alterations in the different types of splicing events (SE, IR, A3SS, A5SS, MXE). From those, a total of 18 genes from the cholesterol metabolism regulatory pathways (12 of those were targets of TDP-43) had significant alterations in their splicing in the Tardbp^M323K/M323K cortex (Fig. 4e). We represented the significant altered splicing exon skipping events by sashimi plots of one of the genes important for the cholesterol biosynthesis signalling pathway, Dhcr7, as well as of another gene related to cholesterol transport, Lpcat3 (Fig. 4f). Interestingly, 22 genes related to the myelination process (14 of them were targets of TDP-43) had alterations in their splicing events induced in the mutant brain.

Finally, we looked at whether the downregulation of the cholesterol synthesis and upregulation of the cholesterol transport pathways were affected in the frontal cortex of patients with different TDP-43 proteinopathies. Thus, we selected all the genes involved in the cholesterol synthesis and transport pathways to see how they were affected in our mouse model of TDP-43 and compared those to other patients of TDP-43 proteinopathies transcriptomic databases. We re-analysed a publicly available transcriptomic database study of the frontal cortex from different types of patients with frontotemporal lobar degeneration (FTLD) with TDP-43 pathology (type A, B, C and D) based on post-mortem histological TDP-43 location and type of aggregates in different brain areas [35]. Both the mice and the different types of FTLD-TDP patients showed the same trend of downregulated genes involved in the cholesterol synthesis and the upregulation of genes involved in cholesterol transport (Fig. 4h).

In conclusion, our findings suggest that TDP-43 alterations had an effect in the myelin stability and maintenance by directly downregulating the myelination process, but mostly through the downregulation of the biosynthesis of the endogenous cholesterol and upregulation of the cholesterols transport, altering the delicate balance of endogenous functional cholesterol which is the essential component of the myelin. Moreover, the dysregulation of the cholesterol metabolism in TDP-43 proteinopathies can be manifested in the form of accumulated lipid droplets in the brain tissue.

The primary objective of this study was to elucidate whether the myelin alterations observed in patients with neurodegenerative diseases associated with TDP-43 proteinopathy, as well as in mouse models carrying TDP-43 mutations, such as our TDP-43-M323K model, were, at least in part, a result of TDP-43 dysfunction that affected lipid metabolism. Specifically, we explored the causal relationship between TDP-43 alterations and the dysregulation of the endogenous brain cholesterol balance, since cholesterol is a main component of myelin and cell membranes and must be tightly regulated to ensure proper myelin formation and, consequently, the correct functioning of neurons.

In the frontal cortex of the TDP-43-M323K mouse model of TDP-43 proteinopathy, our lipidomic analysis revealed that the cholesterol metabolism was clearly altered in the mutant brains, pointing towards an accumulation of cholesterol esters and triglycerides in the form of lipid droplets, as well as a reduction in the levels of free cholesterol. Recent lipidomics analyses of the brain from FTLD-TDP patients reported a remarked loss of myelin related lipids (sphigolipids) and accumulation of cholesterol esters in both frontal and parietal white matter, as a potential indicative of myelin loss [29, 38]. We confirmed the lipidomic findings of accumulated cholesterol esters in the form of LDs, in our mice and as well as in the brains of FTLD-TDP patients without a TARDBP mutation. The fact that these findings could be replicated in patients with TDP-43 pathology, but without a TARDBP mutation, supports the notion that the effects observed in these mice were not specific to this particular mutation, but rather reflected a more general TDP-43 dysfunction.

The endogenous accumulation of cholesterol esters has been suggested to be the consequence of the degeneration process, even the degeneration of the cellular membranes and the myelin. Here, we showed the impact of TDP-43 alteration on lipid droplet accumulation in a cell-autonomous manner in primary fibroblasts carrying the TDP-43-M323K mutation, further suggesting that the accumulation of LDs was also a direct consequence of TDP-43 dysfunction, rather than merely a result of a degenerative myelin or system process. The LD intracellular accumulation in TDP-43 related in vitro models of disease (for instance in iNeurons derived from iPSCs from C9orf72 patients) [4, 27, 41] have been previously reported, further supporting the concept that the dysregulation of endogenous cholesterol is an early pathological event [19, 49], and not exclusively the consequence of a neurodegenerative process.

Through the transcriptomic and splicing analyses, we found that the main molecular mechanism impacting the cholesterol disruption observed was through the dysregulation of the endogenous cholesterol metabolism pathways. The regulation of the brain cholesterol, in particular the endogenous cholesterol biosynthesis, have been previously shown to be downregulated in several neurodegenerative disorders [46], even in the physiological process of aging [47, 50]. For instance, we have previously reported similar transcriptomic alterations of the cholesterol biosynthesis in the spinal cord of SOD1^G93A mice at early presymptomatic disease stages [15]. Other groups have also described very early cholesterol alterations in other models of TDP-43 mutations [13, 48]. Thus, a reduction in the endogenous cholesterol biosynthesis seems to be one of the most important and shared molecular mechanisms from early stages of the neurodegenerative processes. This is crucial, as many neurodegenerative disorders share common pathological mechanisms.

In this study, we demonstrated that the FTLD-TDP postmortem frontal cortices exhibited a more pronounced downregulation of cholesterol biosynthesis genes compared to both the control aged donors and the mutant mice. This could be attributed to the fact that the mice, at that particular age, did not exhibit significant neurodegeneration. The FTLD-TDP brains also show increased lipid droplet accumulation compared to aged controls, as demonstrated by the staining results. On the other hand, alterations in cholesterol transport genes were not consistently upregulated across all FTLD-TDP patient subtypes, as observed in the FTLD-TDP type B group, which included a higher number of ALS patients, potentially with less frontal cortex damage than the other FTLD-TDP types. This was also observed in the mice, which further suggested that more advanced neurodegeneration in the frontal cortex is associated with a broader dysregulation of cholesterol metabolism, with cholesterol biosynthesis being one of the earliest changes observed in all cases. Additionally, these mutant mice exhibited myelin alterations from an early age [18], supporting the idea that cholesterol dysregulation, which impacts myelination, may be an early event triggered by TDP-43 alterations.

Finally, we looked closer at the specific expression and splicing of the TDP-43 target genes to evidence the role of the TDP-43 mutation dysfunction, or at least contributing, to the cholesterol and myelination alterations observed in this study. Other studies have suggested a direct role of TDP-43 dysfunction in lipid regulation alteration and obesity, inferring a general role of TDP-43 in the body lipid metabolism [8, 48].

Closing Remarks and Future Perspectives

This study highlights the critical role of cholesterol dysregulation, particularly in endogenous cholesterol biosynthesis, as a common event in neurodegenerative diseases associated with TDP-43 pathology. The findings suggest that altered cholesterol metabolism, which impacts myelination, could serve as an early indicator of TDP-43 dysfunction, opening new avenues for early diagnostic strategies. Further studies are needed to disentangle the relationship between TDP-43 dysfunction and lipid metabolism. Understanding the broader implications of cholesterol dysregulation in neurodegeneration may offer novel therapeutic targets to address both the underlying pathology and the progression of these diseases, ultimately advancing clinical approaches to treatment.

Acknowledgements

We thank the animal care staff in Hospital Clínico San Carlos (Madrid). We are grateful to Alejandro Rodriguez González for his invaluable help with the analysis of the splicing.

Authors Contribution

Conceptualization: IGT, TJC and SC, Formal Analysis: IGT, EA, JMGC, ZA, LCFB, Funding Acquisition: TJC and SC, Investigation: IGT, JMGC, ZA, IJC, JJ, AGA, KJG, MV, LCFB, Methodology: IJC, KJG, AGA, MV, UGP, JMGC, Project Administration: SC. Resources: TJC, UGP, EA, SC Supervision: SC Visualization: IGT, JMGC, LCFB, EA, Writing-original draft: IGT, JMGC and SC . Writing-review and editing: ALL.

Funding

This research was funded by Ministerio de Ciencia e Innovación (PDI2020-1153-70RB-100/

AEI/10.13039/501100011033 and PID2023-147947OB-I00/AEI) to SC. IGT was supported by a fellowship from Comunidad de Madrid (PIPF-2022/SAL-GL-24282). LCFB was supported by a felloswship from the Universidad Complutense de Madrid (CT58/21-CT59/21).

Declaration of interest

The authors declare that there are no conflicts of interest. The funding institution had no role in data acquisition, analysis, or decision to publish the results.

Data Availability Statement

RNA-sequencing data are in the process of being uploaded at the Gene Expression Omnibus (GEO) database repository.

Afridi R, Rahman MH, Suk K (2022) Implications of glial metabolic dysregulation in the pathophysiology of neurodegenerative diseases. Neurobiol Dis 174:105874. 10.1016/j.nbd.2022.105874
Andrés-Benito P, Gelpi E, Jové M, Mota‐Martorell N, Obis È, Portero‐Otin M, Povedano M, Pujol A, Pamplona R, Ferrer I (2021) Lipid alterations in human frontal cortex in ALS‐FTLD‐TDP43 proteinopathy spectrum are partly related to peroxisome impairment. Neuropathol Appl Neurobiol 47:544–563. 10.1111/nan.12681
Buratti E (2008) Multiple roles of TDP-43 in gene expression, splicing regulation, and human disease. Front Biosci 13:867. 10.2741/2727
Ceron-Codorniu M, Torres P, Fernàndez-Bernal A, Rico-Rios S, Serrano JC, Miralles MP, Beltran M, Garcera A, Soler RM, Pamplona R, Portero-Otín M (2024) TDP-43 dysfunction leads to bioenergetic failure and lipid metabolic rewiring in human cells. Redox Biol 75:103301. 10.1016/j.redox.2024.103301
Chen L, Shen Q, Liu Y, Zhang Y, Sun L, Ma X, Song N, Xie J (2025) Homeostasis and metabolism of iron and other metal ions in neurodegenerative diseases. Signal Transduct Target Ther 10:31. 10.1038/s41392-024-02071-0
Chen S, Zhou Y, Chen Y, Gu J (2018) fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34:i884–i890. 10.1093/bioinformatics/bty560
Cheon SY (2023) Impaired Cholesterol Metabolism, Neurons, and Neuropsychiatric Disorders. Exp Neurobiol 32:57–67. 10.5607/en23010
Chiang P-M, Ling J, Jeong YH, Price DL, Aja SM, Wong PC (2010) Deletion of TDP-43 down-regulates Tbc1d1, a gene linked to obesity, and alters body fat metabolism. Proceedings of the National Academy of Sciences 107:16320–16324. 10.1073/pnas.1002176107
Christidi F, Karavasilis E, Zalonis I, Ferentinos P, Giavri Z, Wilde EA, Xirou S, Rentzos M, Zouvelou V, Velonakis G, Toulas P, Efstathopoulos E, Poulou L, Argyropoulos G, Athanasakos A, Zambelis T, Levin HS, Karandreas N, Kelekis N, Evdokimidis I (2017) Memory-related white matter tract integrity in amyotrophic lateral sclerosis: an advanced neuroimaging and neuropsychological study. Neurobiol Aging 49:69–78. 10.1016/J.NEUROBIOLAGING.2016.09.014
Conte M, Medici V, Malagoli D, Chiariello A, Cirrincione A, Davin A, Chikhladze M, Vasuri F, Legname G, Ferrer I, Vanni S, Marcon G, Poloni TE, Guaita A, Franceschi C, Salvioli S (2022) Expression pattern of perilipins in human brain during aging and in Alzheimer’s disease. Neuropathol Appl Neurobiol 48:e12756. 10.1111/nan.12756
Dodge JC, Jensen EH, Yu J, Sardi SP, Bialas AR, Taksir TV, Bangari DS, Shihabuddin LS (2020) Neutral Lipid Cacostasis Contributes to Disease Pathogenesis in Amyotrophic Lateral Sclerosis. J Neurosci 40:9137–9147. 10.1523/JNEUROSCI.1388-20.2020
Egawa N, Izumi Y, Suzuki H, Tsuge I, Fujita K, Shimano H, Izumikawa K, Takahashi N, Tsukita K, Enami T, Nakamura M, Watanabe A, Naitoh M, Suzuki S, Seki T, Kobayashi K, Toda T, Kaji R, Takahashi R, Inoue H (2022) TDP-43 regulates cholesterol biosynthesis by inhibiting sterol regulatory element-binding protein 2. Sci Rep 12:7988. 10.1038/s41598-022-12133-4
Egawa N, Izumi Y, Suzuki H, Tsuge I, Fujita K, Shimano H, Izumikawa K, Takahashi N, Tsukita K, Enami T, Nakamura M, Watanabe A, Naitoh M, Suzuki S, Seki T, Kobayashi K, Toda T, Kaji R, Takahashi R, Inoue H (2022) TDP-43 regulates cholesterol biosynthesis by inhibiting sterol regulatory element-binding protein 2. Sci Rep 12:7988. 10.1038/s41598-022-12133-4
Feringa FM, Koppes-den Hertog SJ, Wang L, Derks RJE, Kruijff I, Erlebach L, Heijneman J, Miramontes R, Pömpner N, Blomberg N, Olivier-Jimenez D, Johansen LE, Cammack AJ, Giblin A, Toomey CE, Rose IVL, Yuan H, Ward M, Isaacs AM, Kampmann M, Kronenberg-Versteeg D, Lashley T, Thompson LM, Ori A, Mohammed Y, Giera M, van der Kant R (2024) The Neurolipid Atlas: a lipidomics resource for neurodegenerative diseases uncovers cholesterol as a regulator of astrocyte reactivity impaired by ApoE4. bioRxiv 10:2024.07.01.601474
Fernández-Beltrán LC, Godoy-Corchuelo JM, Losa-Fontangordo M, Williams D, Matias-Guiu J, Corrochano S (2021) A Transcriptomic Meta-Analysis Shows Lipid Metabolism Dysregulation as an Early Pathological Mechanism in the Spinal Cord of SOD1 Mice. Int J Mol Sci 22:9553. 10.3390/ijms22179553
Fratta P, Sivakumar P, Humphrey J, Lo K, Ricketts T, Oliveira H, Brito-Armas JM, Kalmar B, Ule A, Yu Y, Birsa N, Bodo C, Collins T, Conicella AE, Mejia Maza A, Marrero‐Gagliardi A, Stewart M, Mianne J, Corrochano S, Emmett W, Codner G, Groves M, Fukumura R, Gondo Y, Lythgoe M, Pauws E, Peskett E, Stanier P, Teboul L, Hallegger M, Calvo A, Chiò A, Isaacs AM, Fawzi NL, Wang E, Housman DE, Baralle F, Greensmith L, Buratti E, Plagnol V, Fisher EM, Acevedo‐Arozena A (2018) Mice with endogenous < scp > TDP ‐43 mutations exhibit gain of splicing function and characteristics of amyotrophic lateral sclerosis. EMBO J 37. 10.15252/embj.201798684
Giblin A, Cammack AJ, Blomberg N, Anoar S, Mikheenko A, Carcolé M, Atilano ML, Hull A, Shen D, Wei X, Coneys R, Zhou L, Mohammed Y, Olivier-Jimenez D, Wang LY, Kinghorn KJ, Niccoli T, Coyne AN, van der Kant R, Lashley T, Giera M, Partridge L, Isaacs AM (2025) Neuronal polyunsaturated fatty acids are protective in ALS/FTD. Nat Neurosci 1–11. 10.1038/s41593-025-01889-3
Godoy-Corchuelo JM, Ali Z, Brito Armas JM, Martins-Bach AB, García-Toledo I, Fernández-Beltrán LC, López-Carbonero JI, Bascuñana P, Spring S, Jimenez-Coca I, Muñoz de Bustillo Alfaro RA, Sánchez-Barrena MJ, Nair RR, Nieman BJ, Lerch JP, Miller KL, Ozdinler HP, Fisher EMC, Cunningham TJ, Acevedo-Arozena A, Corrochano S (2024) TDP-43-M323K causes abnormal brain development and progressive cognitive and motor deficits associated with mislocalised and increased levels of TDP-43. Neurobiol Dis 193:106437. 10.1016/j.nbd.2024.106437
He S, Xu Z, Han X (2025) Lipidome disruption in Alzheimer’s disease brain: detection, pathological mechanisms, and therapeutic implications. Mol Neurodegener 20:11. 10.1186/s13024-025-00803-6
Ho WY, Chang J-C, Lim K, Cazenave-Gassiot A, Nguyen AT, Foo JC, Muralidharan S, Viera-Ortiz A, Ong SJM, Hor JH, Agrawal I, Hoon S, Arogundade OA, Rodriguez MJ, Lim SM, Kim SH, Ravits J, Ng S-Y, Wenk MR, Lee EB, Tucker-Kellogg G, Ling S-C (2021) TDP-43 mediates SREBF2-regulated gene expression required for oligodendrocyte myelination. J Cell Biol 220. 10.1083/jcb.201910213
Horiuchi M, Watanabe S, Komine O, Takahashi E, Kaneko K, Itohara S, Shimada M, Ogi T, Yamanaka K (2024) ALS-linked mutant TDP-43 in oligodendrocytes induces oligodendrocyte damage and exacerbates motor dysfunction in mice. Acta Neuropathol Commun 12:184. 10.1186/s40478-024-01893-x
Khandker L, Jeffries MA, Chang Y-J, Mather ML, Evangelou AV, Bourne JN, Tafreshi AK, Ornelas IM, Bozdagi-Gunal O, Macklin WB, Wood TL (2022) Cholesterol biosynthesis defines oligodendrocyte precursor heterogeneity between brain and spinal cord. Cell Rep 38:110423. 10.1016/j.celrep.2022.110423
Khatri P, Sirota M, Butte AJ (2012) Ten Years of Pathway Analysis: Current Approaches and Outstanding Challenges. PLoS Comput Biol 8:e1002375. 10.1371/journal.pcbi.1002375
Kim D, Paggi JM, Park C, Bennett C, Salzberg SL (2019) Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol 37:907–915. 10.1038/s41587-019-0201-4
Kister A, Kister I (2023) Overview of myelin, major myelin lipids, and myelin-associated proteins. Front Chem 10:1041961. 10.3389/fchem.2022.1041961
Liao Y, Smyth GK, Shi W (2014) featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30:923–930. 10.1093/bioinformatics/btt656
Liu Y, Wang T, Ji YJ, Johnson K, Liu H, Johnson K, Bailey S, Suk Y, Lu Y-N, Liu M, Wang J (2018) A C9orf72–CARM1 axis regulates lipid metabolism under glucose starvation-induced nutrient stress. Genes Dev 32:1380–1397. 10.1101/gad.315564.118
Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:550. 10.1186/s13059-014-0550-8
Marian OC, Teo JD, Lee JY, Song H, Kwok JB, Landin-Romero R, Halliday G, Don AS (2023) Disrupted myelin lipid metabolism differentiates frontotemporal dementia caused by GRN and C9orf72 gene mutations. Acta Neuropathol Commun 11:52. 10.1186/s40478-023-01544-7
Mathews ES, Appel B (2016) Cholesterol Biosynthesis Supports Myelin Gene Expression and Axon Ensheathment through Modulation of P13K/Akt/mTor Signaling. J Neurosci 36:7628–7639. 10.1523/JNEUROSCI.0726-16.2016
McKenna MC, Corcia P, Couratier P, Siah WF, Pradat PF, Bede P (2021) Frontotemporal Pathology in Motor Neuron Disease Phenotypes: Insights From Neuroimaging. Front Neurol 12:723450. 10.3389/FNEUR.2021.723450/XML/NLM
Mollink J, Hiemstra M, Miller KL, Huszar IN, Jenkinson M, Raaphorst J, Wiesmann M, Ansorge O, Pallebage-Gamarallage M, van Cappellen AM (2019) White matter changes in the perforant path area in patients with amyotrophic lateral sclerosis. Neuropathol Appl Neurobiol 45:570. 10.1111/NAN.12555
Neumann M, Kwong LK, Truax AC, Vanmassenhove B, Kretzschmar HA, Van Deerlin VM, Clark CM, Grossman M, Miller BL, Trojanowski JQ, Lee VMY (2007) TDP-43-Positive White Matter Pathology in Frontotemporal Lobar Degeneration With Ubiquitin-Positive Inclusions. J Neuropathol Exp Neurol 66:177–183. 10.1097/01.jnen.0000248554.45456.58
Orth M, Bellosta S (2012) Cholesterol: Its Regulation and Role in Central Nervous System Disorders. Cholesterol 2012:1–19. 10.1155/2012/292598
Pottier C, Mateiu L, Baker MC, Dejesus-Hernandez M, Teixeira Vicente C, Finch NA, Tian S, Van Blitterswijk M, Murray ME, Ren Y, Petrucelli L, Oskarsson B, Biernacka JM, Graff-Radford NR, Boeve BF, Petersen RC, Josephs KA, Asmann YW, Dickson DW, Rademakers R (2022) Shared brain transcriptomic signature in TDP-43 type A FTLD patients with or without GRN mutations. Brain 145:2472–2485. 10.1093/BRAIN/AWAB437
Prater KE, Latimer CS, Jayadev S (2022) Glial TDP-43 and TDP‐43 induced glial pathology, focus on neurodegenerative proteinopathy syndromes. Glia 70:239–255. 10.1002/glia.24096
Raffaele S, Boccazzi M, Fumagalli M (2021) Oligodendrocyte Dysfunction in Amyotrophic Lateral Sclerosis: Mechanisms and Therapeutic Perspectives. Cells 10:1–32. 10.3390/CELLS10030565
Sadler GL, Lewis KN, Narayana VK, De Souza DP, Mason J, McLean C, Gonsalvez DG, Turner BJ, Barton SK (2022) Lipid Metabolism Is Dysregulated in the Motor Cortex White Matter in Amyotrophic Lateral Sclerosis. Metabolites 12:554. 10.3390/metabo12060554
Saher G, Quintes S, Nave K-A (2011) Cholesterol: A Novel Regulatory Role in Myelin Formation. Neuroscientist 17:79–93. 10.1177/1073858410373835
Shen S, Park JW, Lu Z, Lin L, Henry MD, Wu YN, Zhou Q, Xing Y (2014) rMATS: Robust and flexible detection of differential alternative splicing from replicate RNA-Seq data. Proceedings of the National Academy of Sciences 111:E5593–E5601. 10.1073/pnas.1419161111
Sung W, Noh MY, Nahm M, Kim YS, Ki CS, Kim YE, Kim HJ, Kim SH (2024) Progranulin haploinsufficiency mediates cytoplasmic TDP-43 aggregation with lysosomal abnormalities in human microglia. J Neuroinflammation 2024 21(1 21):1–20. 10.1186/S12974-024-03039-1
Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, Simonovic M, Doncheva NT, Morris JH, Bork P, Jensen LJ, von Mering C (2019) STRING v11: protein–protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47:D607–D613. 10.1093/nar/gky1131
Tollervey JR, Curk T, Rogelj B, Briese M, Cereda M, Kayikci M, König J, Hortobágyi T, Nishimura AL, Župunski V, Patani R, Chandran S, Rot G, Zupan B, Shaw CE, Ule J (2011) Characterizing the RNA targets and position-dependent splicing regulation by TDP-43. Nat Neurosci 14:452–458. 10.1038/nn.2778
Velebit J, Horvat A, Smolič T, Prpar Mihevc S, Rogelj B, Zorec R, Vardjan N (2020) Astrocytes with TDP-43 inclusions exhibit reduced noradrenergic cAMP and Ca2 + signaling and dysregulated cell metabolism. Sci Rep 10:6003. 10.1038/s41598-020-62864-5
Vitali C, Wellington CL, Calabresi L (2014) HDL and cholesterol handling in the brain. Cardiovasc Res 103:405–413. 10.1093/cvr/cvu148
Wang N-Q, Sun P-X, Shen Q-Q, Deng M-Y (2025) Cholesterol Metabolism in CNS Diseases: The Potential of SREBP2 and LXR as Therapeutic Targets. Mol Neurobiol 1–25. 10.1007/s12035-024-04672-w
Xu L, Wang J, Zhong J, Lin W, Shen G, He N, Mao X, Fu C, Huang Z, Zhao F, Ye X, Zhu Y, Zheng M, Li H, Wang L, Zhong K, Zhu L, Chen Y (2025) Multiomics identify the gene expression signature of the spinal cord during aging process. Commun Biol 8:193. 10.1038/s42003-025-07475-4
Yang C, Lee G, Bin H, Hu L F (2025) TMEM106B deficiency leads to alterations in lipid metabolism and obesity in the TDP-43Q331K knock-in mouse model. Commun Biol 8:315. 10.1038/s42003-025-07752-2
Yang D, Wang X, Zhang L, Fang Y, Zheng Q, Liu X, Yu W, Chen S, Ying J, Hua F (2022) Lipid metabolism and storage in neuroglia: role in brain development and neurodegenerative diseases. Cell & Bioscience 2022 12:1 12:1–16. 10.1186/S13578-022-00828-0
Zhang J, Liu Q (2015) Cholesterol metabolism and homeostasis in the brain. Protein Cell 6:254. 10.1007/S13238-014-0131-3
Venny 2.1.0 https://bioinfogp.cnb.csic.es/tools/venny/. Accessed 13 Mar 2025

No competing interests reported.

Download PDF

Journal Publication

published 04 Sep, 2025

Read the published version in Acta Neuropathologica →

Editorial decision: Revision requested
07 Apr, 2025
Reviews received at journal
07 Apr, 2025
Reviews received at journal
29 Mar, 2025
Reviewers agreed at journal
24 Mar, 2025
Reviewers agreed at journal
20 Mar, 2025
Reviewers invited by journal
20 Mar, 2025
Editor assigned by journal
20 Mar, 2025
Submission checks completed at journal
20 Mar, 2025
First submitted to journal
18 Mar, 2025

You are reading this latest preprint version

TDP-43 dysregulation impairs cholesterol metabolism linked with myelination defects

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

Mice

Human brain samples

Immunohistochemical Processing of Patients Brain Tissue

Mouse Tissue Immunostainings

Primary ear fibroblasts generation

Flow Cytometry for LD analysis

Fibroblasts Immunostainings

Western blot

Gene expression analysis

Lipidomic

RNA-sequencing

DEG Expression analysis of the transcriptome

Splicing analysis

Analysis of RNA sequencing database from FTLD patients and controls

Statistical analyses

Results

Disrupted cholesterol metabolism in patients and mice with TDP-43 proteinopathy

Discussion

Closing Remarks and Future Perspectives

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1