Progranulin Deficiency Induces Mitochondrial Dysfunction in Frontotemporal Lobar Degeneration with TDP-43 Inclusions

Loss-of-function (LOF) mutations in GRN gene, which encodes progranulin (PGRN), cause frontotemporal lobar degeneration with TDP-43 inclusions (FTLD-TDP). FTLD-TDP is one of the most common forms of early onset dementia, but its pathogenesis is not fully understood. Mitochondrial dysfunction has been associated with several neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD) and amyotrophic lateral sclerosis (ALS). Here, we have investigated whether mitochondrial alterations could also contribute to the pathogenesis of PGRN deficiency-associated FTLD-TDP. Our results showed that PGRN deficiency induced mitochondrial depolarization, increased ROS production and lowered ATP levels in GRN KD SH-SY5Y neuroblastoma cells. Interestingly, lymphoblasts from FTLD-TDP patients carrying a LOF mutation in the GRN gene (c.709-1G > A) also demonstrated mitochondrial depolarization and lower ATP levels. Such mitochondrial damage increased mitochondrial fission to remove dysfunctional mitochondria by mitophagy. Interestingly, PGRN-deficient cells showed elevated mitochondrial mass together with autophagy dysfunction, implying that PGRN deficiency induced the accumulation of damaged mitochondria by blocking its degradation in the lysosomes. Importantly, the treatment with two brain-penetrant CK-1δ inhibitors (IGS-2.7 and IGS-3.27), known for preventing the phosphorylation and cytosolic accumulation of TDP-43, rescued mitochondrial function in PGRN-deficient cells. Taken together, these results suggest that mitochondrial function is impaired in FTLD-TDP associated with LOF GRN mutations and that the TDP-43 pathology linked to PGRN deficiency might be a key mechanism contributing to such mitochondrial dysfunction. Furthermore, our results point to the use of drugs targeting TDP-43 pathology as a promising therapeutic strategy for restoring mitochondrial function in FTLD-TDP and other TDP-43-related diseases.


Introduction
Heterozygous loss-of-function (LOF) mutations in granulin (GRN) gene leading to progranulin (PGRN) happloinsufficiency have been identified as a major cause of familial frontotemporal lobar degeneration with TDP-43 accumulation (FTLD-TDP) [1][2][3][4][5][6]. FTLD-TDP patients exhibit behavioral changes and language difficulties associated with the neuronal death in the frontal and temporal lobar brain cortex. Currently, the mechanisms by which reduced levels of PGRN lead to neurodegeneration are still unknown. Mitochondrial dysfunction has been shown to contribute to neuronal death in neurodegenerative disorders such as Alzheimer's disease (AD), Parkinson's disease (PD) and amyotrophic lateral sclerosis (ALS) [7][8][9][10][11]. Recent evidence has shown that PGRN plays an important role in regulating mitochondrial homeostasis and activity [12,13]. Furthermore, previous results from our lab showed that PGRN insufficiency regulated the intrinsic/mitochondrial apoptosis pathway in GRN knockdown (KD) SH-SY5Y neuroblastoma cells [14,15] and peripheral cells from FTLD-TDP patients carrying a LOF GRN mutation (c.709-1G > A) [16]. Altogether, this evidence suggests that mitochondrial dysfunction could also contribute to neurodegeneration in FTLD-TDP linked to PGRN deficiency. Therefore, understanding the molecular mechanisms by which PGRN deficiency leads to mitochondrial dysfunction in FLD-TDP may lead to the identification of new therapeutic strategies.
In this work, we have investigated the effect of PGRN deficit on mitochondrial function, dynamics and degradation, using a cellular model of FTLD-TDP with PGRN deficiency (SH-SY5Y GRN KD cells) and lymphoblasts from FTLD-TDP patients carrying a LOF mutation in the GRN gene (c.709-1G > A). Our results showed that PGRN deficiency induced mitochondrial damage in both FTLD-TDP models. Interestingly, we found that because of the autophagy impairment associated with PGRN loss, the damaged mitochondria in PGRN-deficient cells failed to be degraded, leading to mitochondrial accumulation. Importantly, the treatment with TDP-43 phosphorylation inhibitors rescued mitochondrial function in GRN KD cells, suggesting a key role for TDP-43 pathology in the mitochondrial dysfunction observed in PGRN-deficient cells. Together, our results indicate that mitochondrial impairment might contribute to neuronal death in FTLD-TDP associated with LOF GRN mutations and that modulating TDP-43 phosphorylation might represent a good therapeutic approach to rescue the mitochondrial dysfunction and the consequent neurodegeneration in FTLD-TDP and other TDP-43 proteinopathies.

Cell Lines Culture and Treatments
The control and stable GRN KD human neuroblastoma SH-SY5Y cells (Clone # 207) were a generous gift from Drs. Alvin P. Joselin and Jane Y. Wu from the Center for Genetic Medicine (Northwestern University, Chicago, IL, USA). These lines were generated using the pSUPERIOR RNAi construct containing a sequence of 19 nucleotides targeting human GRN, as was previously described [14,17]. Cells were cultured in Dulbecco's Modified Eagle Medium (DMEM; Thermo Fisher Scientific, Waltham, MA, USA), supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; Thermo Fisher Scientific, MA, USA) and 1% penicillin/streptomycin (P/S; Thermo Fisher Scientific, MA, USA). When necessary, DMEM was supplemented with other compounds as follows: (i) to block autophagy, SH-SY5Y cells were cultured for 12 h in DMEM containing 100 nM of V-ATPase inhibitor bafilomycin A1 (BafA1; Sigma-Aldrich, MO, USA) and (ii) to block TDP-43 phosphorylation, SH-SY5Y cells were cultured for 48 h in DMEM containing the casein kinase 1 (CK1) inhibitors IGS-2.7 and IGS-3.27 (5 µM). These two small molecules were synthetized in our laboratories according to procedures previosly described [18,19].
The lymphoblastic cell lines used in this study (Table 1) were generated in our laboratory by infecting peripheral blood lymphocytes from FTLD-TDP patients and control subjects with Epstein Barr virus (EBV), as previously described [20]. Lymphoblastic cell lines were grown in suspension using RPMI-1640 medium supplemented with 1% P/S and 10 % (v/v) FBS. PGRN levels were measured in plasma samples from all subjects involved in this study using the PGRN ELISA kit AG-45A-0018YEK-KI01 (AdipoGene, Füllinsdorf, Switzerland), following the manufacturer's protocol. Peripheral blood samples of all the individuals enrolled in this studio were taken after obtaining written informed consent of the patients or their relatives. All study protocols were approved by the Donostia Hospital and the Spanish Council of Higher Research Institutional Review Board (01/01/2006) and are in accordance with National and International Guidelines (Declaration of Helsinki).

Quantitative Real-Time PCR
Total RNA from lymphoblasts was extracted using Trizol (Invitrogen, Alcobendas, Madrid, Spain) and was used to perform a qPCR as previously described [21]. Briefly, RNA was treated with DNase I Amplification Grade (Invitrogen, Alcobendas, Madrid, Spain) and then transcribed into cDNA using the Superscript III Reverse Transcriptase kit (Invitrogen, Alcobendas, Madrid, Spain). Quantitative real-time polymerase chain reaction (PCR) was performed in triplicates using the TaqMan Universal PCR MasterMix No Amperase UNG reagent (Applied Biosystems, Alcobendas, Madrid, Spain) and the Bio-Rad iQ5 system with a thermal profile of an initial 5 min melting step at 95 • C followed by 40 cycles at 95 • C for 10 s and 60 • C for 60 s. GRN relative messenger RNA (mRNA) levels were normalized to β-actin expression using the simplified comparative threshold cycle delta-delta CT method (2-[∆CT PGRN -∆CT actin]). Primers were designed using the Universal ProbeLibrary for Human (Roche Applied Science, Madrid, Spain) and used at a final concentration of 20 µM (GRN primers: 5 -tctgtagtctgagcgctaccc-3 and 5 -agggtccacatggtctgc-3 ; β-actin primers: 5 -ccaaccgcgagaagatga-3 and 5 -ccagaggcgtacagggatag-3 ).

Cell Viability and Apoptosis Measurement
Cell viability was assessed using the MTT assay. Cells were seeded in triplicate in 96-well plates, and 72 h later, 10 µL of 5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide reactive (MTT; Sigma-Aldrich, St. Louis, MO, USA) was added to each well containing 100 µL of media and incubated for 4 h at 37 • C to allow the viable cells to reduce the MTT to formazan. Purple formazan crystals were then dissolved in 100 µL DMSO and the absorbance was measured at 595 nm using a microplate reader (EnSpire, PerkinElmer Waltham, MA, USA).
Apoptosis was assessed by the microscopic examination of nuclei morphology. To do so, cells were stained with DAPI (4,6-diamidino-2 phenylindole) (Thermo Fisher Scientific, MA, USA) and then imaged using a Leica TCS SP5 confocal microscope (Leica Microsystems, Wetzlar, Germany) with an excitation peak at 359 nm and an emission peak at 457 nm. Cells displaying highly condensed nuclei, or pyknotic nuclei, were considered apoptotic cells.

Measurement of Mitochondrial Membrane Potential (∆Ψ m )
Mitochondrial membrane potential (∆Ψ m ) was analyzed using the cell-permeant fluorescent dyes tetramethylrhodamine ethyl ester (TMRE, Thermo Fisher Scientific, MA, USA) and tetramethylrhodamine methyl ester (TMRM, Thermo Fisher Scientific, MA, USA), according to previously established protocols [22,23]. Briefly, cells were seeded either in 96-well plates or in 6-well plates on 25 mm coverslips and 48 h after seeding, cells were incubated with 40 nM TMRE or TMRM in a HEPES-buffered salt solution (HBSS) (156 mM NaCl, 3 mM KCl, 2 mM MgSO 4 , 1.25 mM KH 2 PO 4 , 2 mM CaCl 2 , 10 mM glucose and 10 mM HEPES; pH adjusted to 7.35 with NaOH) for 40 min at 37 • C. Then, TMRE/TMRM fluorescence was assessed using either a POLARstar Galaxy spectrofluorimeter (BMG Labtechnologies, Offenburg, Germany) or a Zeiss 510 confocal microscope equipped with META detection system (Zeiss, Oberkochen, Germany) with 40× oil immersion objective. In both cases, excitation wavelength was 560 nm and emission was detected above 580 nm. Microscope images were analyzed using the Volocity software (Quorum Technologies, Ontario, Canada) and TMRM values for untreated cells were set to 100%.

Reactive Oxygen Species (ROS) Measurement
Intracellular accumulation of ROS was determined using the fluorescent probe CM-H 2 DCFDA (Thermo Fisher Scientific, MA, USA). To do so, control and GRN KD SH-SY5Y cells were seeded in 96-well plates, and 48 h later, cells were loaded with 10 µM CM-Antioxidants 2023, 12, 581 5 of 20 H 2 DCFDA for 30 min. Fluorescence measurements were carried out using a POLARstar Galaxy spectrofluorimeter (BMG Labtechnologies, Offenburg, Germany). The Excitation wavelength used was 495 nm and emission was detected above 510 nm.
Mitochondrial ROS levels were measured using MitoSox (Thermo Fisher Scientific, MA, USA), a fluorogenic dye specifically targeted to mitochondria in living cells, which produces red fluorescence when oxidized by superoxide. SH-SY5Ycells were cultured in 6-well plates on 25 mm coverslips during 48 h and then incubated with 5 µM MitoSox in HBSS for 30 min at room temperature. Z-stack images were obtained using a Zeiss 510 confocal microscope equipped with META detection system (Zeiss, Oberkochen, Germany) and 40× oil immersion objective using an excitation wavelength of 510 nm. Emission was detected at 580 nm. Fluorescence intensity was quantified using the Volocity software (Quorum Technologies, Ontario, Canada).

Measurement of Cellular Oxygen Consumption
Cellular oxygen consumption rate (OCR) was measured using a Seahorse XF24 Extracellular Flux Analyzer (Seahorse, Agilent Technologies, CA, USA). A total of 20,000 control and GRN KD SH-SY5Y cells per well were plated in DMEM supplemented with 10% FBS and 1% P/S. Then, 24 h later, cell growing media was replaced by 25 mM glucose, 1 mM Pyruvate and 2 mM L-glutamine containing XF Base medium. Cells were incubated for 1 h in a CO 2 -free incubator at 37 • C allowing temperature and pH equilibration before loading into the XF24 analyzer. Mitochondrial function was determined through sequential addition of 1 µM oligomycin (Sigma-Aldrich, MO, USA), 0.5 µM carbonylcyanide-ptrifluoromethoxyphenylhydrazone (FCCP; Sigma-Aldrich, MO, USA) and 1 µM antimycin (Sigma-Aldrich, MO, USA)/1 µM rotenone (Sigma-Aldrich, MO, USA). This sequential addition allowed us to determine the basal oxygen consumption, oxygen consumption-linked to ATP synthesis (ATP), maximal respiration and the cellular spare capacity.

ATP Levels Measurement
In SH-SY5Y cells, ATP was measured using a FRET-based ATP-plasmid indicator (AT1.03 sensor) kindly provided by Dr. H. Imamura, following previously established protocols [24]. Briefly, control and GRN KD SH-SY5Y cells were seeded in 6-well plates on 25 mm coverslips and then transfected with the FRET-based ATP-plasmid indicator using Effectene transfection reagent (Qiagen, Hilden, Germany) according to the manufacturer's instructions. One day after transfection, cell culture media were replaced with HBSS medium plus Ca 2+ and Mg 2+ and then cells were subjected to a time-dependent fluorescence imaging using a Zeiss 510 LSM confocal microscope with META detection system. Images were obtained using a 63× oil-immersion objective. Excitation of cyan fluorescent protein was 405 nm and emission was detected between 460 and 510 nm. Yellow fluorescent protein was excited using the 405 nm laser line and emission was detected using a band-pass filter from 515 to 580 nm. Minimal illumination intensity was kept to avoid phototoxicity (at 0.1-0.2% of laser output) and the pinhole was adjusted to give an optical slice of~2 µm. The ratio-metric analysis of the yellow-and cyan-fluorescent proteins was assessed using the ZEN software (Zeiss, Oberkochen, Germany) and allowed the estimation of ATP kinetics within single cells.
In lymphoblasts, ATP basal levels were measured by using the Vialight plus assay kit (Lonza, Verviers, Belgium). This assay allows the measurement of ATP present in all metabolically active cells and is based upon the bioluminescent measurement of ATP. Lymphoblasts were seeded in 24-well plate and after 48 h, cells were lysed using the cell lysis buffer provided by the kit. Protein concentration of the cell lysates was estimated using the BCA assay (Thermo Fisher Scientific, MA, USA) and samples were diluted in cell lysis buffer to obtain 0.1 µg/µL protein concentration. Then, 100 µL of the protein solutions were transferred to each well of a 96-well plate containing 100 µL of ATP monitoring reagent plus (in triplicates). Plates were incubated for 2 min at room temperature and then luminescence was measured using an EnSpire microplate reader (PerkinElmer Waltham, MA, USA).

Mitochondrial Mass
Mitochondrial mass was measured using the cell permeable mitochondria-selective dye MitoTracker Green FM (Thermo Fisher Scientific, MA, USA). MitoTracker is a fluorescent dye that localizes to the mitochondrial matrix regardless of the mitochondrial membrane potential and covalently binds to mitochondrial proteins by reacting with free thiol groups of cysteine residues. GRN KD and control SH-SY5Y cells were seeded in 96-or 24-well plates. For this assay, the same number of GRN KD and control SH-SY5Y cells were seeded in each well. Then, 72 h later, SH-SY5Y cells were incubated for 15 min with 400 nM of MitoTracker Green FM and then fluorescence was determined by using either a confocal microscope Leica TCS SP5 (Leica Microsystems, Wetzlar, Germany) or a spectrofluorimeter (BMG Labtechnologies, Offenburg, Germany). In both cases, excitation wavelength was 490 nm and emission was detected at 516 nm.
Mitochondrial mass was also assessed by analyzing the levels of the mitochondrial structural protein complex V-β subunit (CxVβ) by immunoblotting or immunofluorescence, as previously described [25][26][27]. Briefly, to assess mitochondrial mass by immunoblotting, equal amounts of total protein were loaded for all the samples and CxVβ levels were normalized using the cytoskeleton protein β-actin. To assess mitochondrial mass by immunofluorescence, cells were seeded in 24-well plates on 13 mm coverslips and stained using an anti-CxVβ antibody. Then, Z-stack images were obtained using a Zeiss 510 LSM confocal microscope with META detection system with a 40× oil immersion objective, which were analyzed using the Volocity software (Quorum Technologies, Ontario, Canada), which allowed us to visualize, analyze and quantify 3D fluorescence images.

Immunofluorescence and Colocalization Analysis
SH-SY5Y cells were seeded at an initial density of 1 × 10 5 cells/well in 24-well plates containing 12 mm coverslips, and 48 h after seeding, half of the coverslips were treated with BafA1 (100 nM) for 12 h and then fixed with 4% PFA. For the immunostaining, cells were blocked with 2% BSA in PBS, permeabilized with 0.1% saponin, incubated with primary antibodies (rabbit anti-p62/SQSTM1, 1:200 and mouse anti-CxVβ, 1:200) at room temperature. Then, 1 h later, cells were incubated with the corresponding secondary antibodies (Alexa-fluor 488 goat anti-rabbit, 1:200, and Alexa-fluor 568 goat anti-mouse, 1:200, both from Thermo Fisher Scientific, MA, USA) for 45 min at room temperature. Preparations were mounted using proLong Gold antifade mountant with DAPI (4 ,6diamidino-2-phenylindole) (Thermo Fisher Scientific, MA, USA) and visualized using a Zeiss 510 LSM confocal microscope with META detection system and a 40× oil immersion objective. Fluorescence intensity and colocalization were analyzed using Volocity software (Quorum Technologies, Ontario, Canada). Pearson's correlation coefficient was used to estimate the colocalization between green and red channels.

Statistical Analysis
Student's t test, one-way and two-way analysis of variance (ANOVA) statistical analyses were performed using GraphPad Prism 6. Bonferroni's analysis was used to analyze the statistical significance between multiple groups. Plots show means ± Standard Error of the Mean (SEM) of all experiments performed. Differences were considered statistically significant when p < 0.05.

GRN KD SH-SY5Y Cells Recapitulate Pathological Characteristics of FTLD-TDP
Previous reports have demonstrated that PGRN depletion induces cytosolic TDP-43 accumulation in several cell models [18,28,29], suggesting that PGRN-deficient cells could be used to study FTLD-TDP. Here, we have investigated if GRN KD SH-SY5Y cells also recapitulate key aspects of FTLD-TDP pathophysiology such as TDP-43 phosphorylation/cleavage, neuronal death and oxidative stress [30][31][32]. We found that GRN depletion in SH-SY5Y cells led to increased TDP-43 protein phosphorylation at S409/410 and cleavage into 25 kDa C-terminal fragments ( Figures 1A-D and S1). Furthermore, GRN KD cells exhibited decreased cell viability ( Figure 1E) and increased apoptosis ( Figure 1F), as was indicated by the presence of morphological features of apoptotic cell death such as increased chromatin condensation and the formation of pyknotic nuclei ( Figure 1F). To determine if GRN deficiency induced changes in oxidative stress status in SH-SY5Y cells, we assessed the cytosolic and mitochondrial levels of reactive oxygen species (ROS) by measuring fluoresce of CM-H 2 DCFDA and MitoSox probes, respectively. Both cytosolic ( Figure 1G) and mitochondrial ( Figure 1H) ROS production were increased in PGRN-deficient cells, compared with control SH-SY5Y cells. Together, these results indicate that GRN KD SH-SY5Y cells are an adequate model to study FTLD-TDP, as they mimic some of the main hallmarks of FLTD-TDP. significant when p < 0.05.

GRN KD SH-SY5Y Cells Recapitulate Pathological Characteristics of FTLD-TDP
Previous reports have demonstrated that PGRN depletion induces cytosolic TDP-43 accumulation in several cell models [18,28,29], suggesting that PGRN-deficient cells could be used to study FTLD-TDP. Here, we have investigated if GRN KD SH-SY5Y cells also recapitulate key aspects of FTLD-TDP pathophysiology such as TDP-43 phosphorylation/cleavage, neuronal death and oxidative stress [30][31][32]. We found that GRN depletion in SH-SY5Y cells led to increased TDP-43 protein phosphorylation at S409/410 and cleavage into 25 kDa C-terminal fragments ( Figure 1A-D and S1). Furthermore, GRN KD cells exhibited decreased cell viability ( Figure 1E) and increased apoptosis ( Figure 1F), as was indicated by the presence of morphological features of apoptotic cell death such as increased chromatin condensation and the formation of pyknotic nuclei ( Figure 1F). To determine if GRN deficiency induced changes in oxidative stress status in SH-SY5Y cells, we assessed the cytosolic and mitochondrial levels of reactive oxygen species (ROS) by measuring fluoresce of CM-H2DCFDA and MitoSox probes, respectively. Both cytosolic ( Figure  1G) and mitochondrial ( Figure 1H) ROS production were increased in PGRN-deficient cells, compared with control SH-SY5Y cells. Together, these results indicate that GRN KD SH-SY5Y cells are an adequate model to study FTLD-TDP, as they mimic some of the main hallmarks of FLTD-TDP.

PGRN Insufficiency Impairs Mitochondrial Bioenergetics in SH-SY5Y Cells and FTLD-TDP Patient's Lymphoblasts
Because mitochondria play a key role in ROS production and apoptotic cell death, we studied the mitochondrial function in our in vitro GRN KD cellular model. To do so, we analyzed the mitochondrial membrane potential (∆Ψ m ), which reflects the mitochondrial health and function, in control and GRN KD SH-SY5Y cells using the TMRE probe. TMRE is a cell permeant, positively charged fluorescent dye that accumulates in active mitochondria. GRN KD cells exhibited a significant reduction of the TMRE signal (Figure 2A), indicating that PGRN deficiency induced mitochondrial depolarization. Consistent with the reduced ∆Ψ m , we found that GRN KD cells exhibited reduced mitochondrial ATP levels ( Figure 2B).  To further investigate how PGRN deficiency affects mitochondrial bioenergetics, we estimated the oxygen consumption rate (OCR) using the Seahorse XF analyzer ( Figure 2C). GRN KD cells showed reduced basal OCR ( Figure 2D). Consistent with the above results, after F o F 1 -ATP synthase inhibition with oligomycin ( Figure 2C) PGRN-deficient cells demonstrated lower ATP production linked to respiration ( Figure 2E). In addition, GRN KD cells displayed lower maximal respiration ( Figure 2F) and spare capacity ( Figure 2G), both obtained after addition of the mitochondrial uncoupler FCCP ( Figure 2C). Together, these findings indicate that PGRN deficiency could be associated with reduced activity or lack of substrates for the mitochondrial respiratory complexes I or II.
We then investigated if the mitochondrial bioenergetics deficits observed in the GRN KD model could be extensible to FTLD-TDP patients. To do so, we used lymphoblastoid cell lines generated from FTLD-TDP patients carrying the c.709-1G > A heterozygous mutation in the GRN gene. This mutation is predicted to cause exon eight skipping, frameshift and premature translation termination, resulting in nonsense-mediated mRNA decay [33]. As expected, FTLD-TDP patients carrying this mutation exhibited decreased PGRN levels in plasma, compared with control subjects ( Figure S2A). Furthermore, lymphoblastoid cell lines generated from the c.709-1G > A GRN mutation carriers exhibited reduced GRN mRNA and PGRN protein levels ( Figure S2B-D). Importantly, similarly to GRN KD SH-SY5Y cells, lymphoblasts from FTLD-TDP patients carrying the c.709-1G > A GRN mutation exhibited depolarized mitochondria ( Figure 2H) and reduced ATP levels when compared with lymphoblasts from healthy subjects ( Figure 2I). These observations suggest that mitochondrial impairment might be a pathological feature of FTLD-TDP.

Progranulin Deficiency Increases Mitochondrial Mass
To explore whether the impairment of mitochondrial bioenergetics in GRN KD cells could be explained by a reduced amount of mitochondria, we measured mitochondrial mass in control and PGRN-deficient cells. Interestingly, GRN KD SH-SY5Y cells exhibited higher mitochondrial mass as assessed using the MitoTracker Green FM fluorescence signal ( Figure 3A,B). The increased mitochondrial mass in PGRN-deficient cells was then validated by immunoblot by measuring the levels of the mitochondrial structural protein complex V-β subunit (CxVβ) (Figures 3C and S3A) and normalized by the levels of the cytoskeleton protein β-actin. Remarkably, the accumulation of mitochondria in GRN KD cells was not the result of increased mitochondrial biogenesis, as indicated by the presence of equal levels of the mitochondrial biogenesis marker PGC1α in both control and GRN KD cells (Figures 3D and S3B).

Progranulin Deficiency Increases Mitochondrial Mass
To explore whether the impairment of mitochondrial bioenergetics in GRN KD cells could be explained by a reduced amount of mitochondria, we measured mitochondrial mass in control and PGRN-deficient cells. Interestingly, GRN KD SH-SY5Y cells exhibited higher mitochondrial mass as assessed using the MitoTracker Green FM fluorescence signal ( Figure 3A-B). The increased mitochondrial mass in PGRN-deficient cells was then validated by immunoblot by measuring the levels of the mitochondrial structural protein complex V-β subunit (CxVβ) ( Figure 3C and S 3A) and normalized by the levels of the cytoskeleton protein β-actin. Remarkably, the accumulation of mitochondria in GRN KD cells was not the result of increased mitochondrial biogenesis, as indicated by the presence of equal levels of the mitochondrial biogenesis marker PGC1α in both control and GRN KD cells (Figures 3D and S 3B). These results suggested that the PGRN deficiency-induced increase in mitochondrial mass is not due to enhanced mitochondrial biogenesis but may be the result of impaired degradation of damaged mitochondria. These results suggested that the PGRN deficiency-induced increase in mitochondrial mass is not due to enhanced mitochondrial biogenesis but may be the result of impaired degradation of damaged mitochondria.

Impaired Autophagy in GRN KD Cells Blocks the Removal of Damaged Mitochondria
Mitochondria are dynamic organelles that constantly fuse and divide. The processes of mitochondrial fusion and fission, known as mitochondrial dynamics, are key mechanisms for the mitochondrial quality control as they regulate the removal of damaged mitochondria by mitophagy. We studied the mitochondrial dynamics in the control and GRN KD SH-SY5Y cells by analyzing the levels of fusion and fission proteins such as mitofusin 1 and 2 (Mfn1-2), Opa1, FIS1 and Drp1 (Figures 4A,B, S4 and S5). PGRN-deficient cells showed decreased levels of the mitochondrial fusion protein Opa1 (Figures 4A and S4C) and increased levels of the mitochondrial fission proteins FIS1 and Drp1 in GRN KD cells, compared with control cells (Figures 4B and S5), demonstrating an imbalance in the mitochondrial fusion/fission dynamics towards increased mitochondrial fission. To study whether the increased mitochondrial fission targeted the mitochondria for their disposal by mitophagy, we assessed the colocalization of mitochondria with the mitophagy marker p62 [34]. PGRN deficiency increased the colocalization of the mitochondrial marker CxVβ with p62 ( Figures 4C and S6), showing that damaged mitochondria were targeted for mitophagy in PGRN-deficient cells.
of mitochondrial fusion and fission, known as mitochondrial dynamics, are key mechanisms for the mitochondrial quality control as they regulate the removal of damaged mitochondria by mitophagy. We studied the mitochondrial dynamics in the control and GRN KD SH-SY5Y cells by analyzing the levels of fusion and fission proteins such as mitofusin 1 and 2 (Mfn1-2), Opa1, FIS1 and Drp1 ( Figure 4A-B, S4 and S5) . PGRN-deficient cells showed decreased levels of the mitochondrial fusion protein Opa1 ( Figure 4A and S 4C) and increased levels of the mitochondrial fission proteins FIS1 and Drp1 in GRN KD cells, compared with control cells ( Figure 4B and S 5), demonstrating an imbalance in the mitochondrial fusion/fission dynamics towards increased mitochondrial fission. To study whether the increased mitochondrial fission targeted the mitochondria for their disposal by mitophagy, we assessed the colocalization of mitochondria with the mitophagy marker p62 [34]. PGRN deficiency increased the colocalization of the mitochondrial marker CxVβ with p62 ( Figure 4C and S 6), showing that damaged mitochondria were targeted for mitophagy in PGRN-deficient cells.  The colocalization between the two fluorophores was assessed by measuring the Pearson's correlation coefficient. Plot represents the average ± SEM of the Pearson's correlation coefficient measured in eleven images from three independent experiments. Statistical analysis was performed using Student's t-test. * p < 0.05. (D) Fluorescence images showing the levels of CxVβ (in red) before and after the treatment with 100 nM bafilomycin A1 (BafA1) in control and GRN KD SH-SY5Y cells. CxVβ staining was used as a marker of mitochondrial accumulation. Scale bar = 9 µm. CxVβ fluorescence intensity was measured using Volocity software. Plot represents the average ± SEM of CxVβ fluorescence intensity measured in eight different images from three independent experiments. Statistical analysis was performed using two-way ANOVA's tests followed by Bonferroni s correction. ** p < 0.01 significant difference between untreated control and GRN KD cells. † † † p < 0.001 significant difference between untreated control cells and BafA1 treated control cells.
It has been previously demonstrated that PGRN plays an important role in regulating autophagy [35] and that PGRN depletion leads to autophagy blockage [36]. To address whether autophagy was also impaired in GRN KD SH-SY5Y cells, we measured autophagic flux by monitoring changes in the levels and localization of the autophagy adaptor p62 and the autophagosome marker LC3II (microtubule-associated protein 1 light chain 3B), before and after bafilomycin A1 (BafA1) treatment ( Figure S7) [37]. Under basal conditions, GRN KD cells showed increased p62 levels together with decreased LC3II levels, compared with control cells (Figure S7A,B). Notably, when we added Baf1A, a V-ATPase inhibitor that inhibits autophagosome-lysosome fusion and blocks autophagosome degradation, the rate of LC3 II formation was lower in GRN KD cells than in control cells ( Figure S7A). Together, these results suggest that GRN KD SH-SY5Y cells have reduced autophagy flux, probably associated with a failure in autophagosome formation. Thus, we asked if the mitochondrial accumulation associated with PGRN deficiency could be a consequence of the autophagy failure observed in GRN-deficient cells. To do so, we measured mitochondrial mass before and after blocking autophagy with BafA1. Consistent with the above results ( Figure 3A-C), GRN KD cells showed increased CxVβ staining compared with control cells ( Figure 4D). BafA1 treatment induced mitochondrial accumulation in control cells but did not modify CxVβ levels in GRN KD cells ( Figure 4D). These results confirmed that mitochondrial accumulation in PGRN-deficient cells was a consequence of a general failure of autophagy.

Inhibition of TDP-43 Phosphorylation Restores Mitochondrial Bioenergetics in GRN KD Cells
GRN KD SH-SY5Y cells accumulated S409/S410 phosphorylated C-terminal fragments of TDP-43 protein. It has been reported that casein kinase-1 δ (CK-1 δ) is the kinase that phosphorylates TDP-43 at these residues [38]. We previously developed two brainpenetrant CK-1δ inhibitors inhibitors (IGS2.7 and IGS3.27) and demonstrated that both compounds decreased TDP-43 phosphorylation and accumulation as well as prevented neuronal death in FTLD-TDP patient-derived lymphoblasts [18]. Because TDP-43 pathology in FTLD could be related to mitochondrial impairment [10,[39][40][41], here we investigated if the inhibition of TDP-43 phosphorylation could have an effect in the mitochondrial bioenergetics of PGRN-deficient cells. Treatment with both CK-1δ inhibitors, IGS2.7 and IGS3.27, restored the ∆Ψ m in SH-SY5Y GRN KD cells, with no effect on control cells ( Figure 5). These results suggested that the accumulation of phosphorylated forms of TDP-43 might be responsible for the mitochondrial impairment observed in GRN KD cells. Mitochondrial membrane potential was measured by assessing TMRM fluorescence using a confocal microscope with 40× water-immersion objective (Ex/Em = 549/575 nm). Scale bar = 9µ m. Top panel shows representative images of TMRM staining in control and GRN KD cells before and after the treatment with both pTDP-43 inhibitors. Plot represents the average ± SEM of at least eleven images for each condition. The experiment was performed three times. Statistical analysis was performed using twoway ANOVA´s tests followed by Bonferroni´s correction. *** p < 0.001 significant difference between untreated control and GRN KD cells. † † † p < 0.001 significant difference between untreated and treated GRN KD cells.

Discussion
Subjects carrying heterozygous GRN LOF mutations develop early onset FTLD-TDP, a neurodegenerative disease considered the second most common cause of dementia after AD [42]. However, the pathological mechanisms resulting in the clinical and cellular Mitochondrial membrane potential was measured by assessing TMRM fluorescence using a confocal microscope with 40× water-immersion objective (Ex/Em = 549/575 nm). Scale bar = 9 µm. Top panel shows representative images of TMRM staining in control and GRN KD cells before and after the treatment with both pTDP-43 inhibitors. Plot represents the average ± SEM of at least eleven images for each condition. The experiment was performed three times. Statistical analysis was performed using two-way ANOVA s tests followed by Bonferroni s correction. *** p < 0.001 significant difference between untreated control and GRN KD cells. † † † p < 0.001 significant difference between untreated and treated GRN KD cells.

Discussion
Subjects carrying heterozygous GRN LOF mutations develop early onset FTLD-TDP, a neurodegenerative disease considered the second most common cause of dementia after AD [42]. However, the pathological mechanisms resulting in the clinical and cellular features of FTLD-TDP associated with GRN mutations are still not well understood. There is growing evidence that mitochondrial abnormalities are involved in the pathogenesis of common neurodegenerative diseases such as AD, PD and ALS [40,[43][44][45], but little is known about the role of mitochondrial dysfunction in the pathogenesis of FTLD. This work was undertaken to investigate the link between mitochondrial dysfunction and FTLD-TDP associated with PGRN deficiency using a neuronal model of FTLD-TDP based on GRN gene silencing and lymphoblasts from FTLD-TDP patients carrying a GRN LOF mutation. Our results indicated that PGRN deficiency impaired mitochondrial bioenergetics in both the FTLD-TDP neuronal model and the FTLD-TDP patient's derived lymphoblasts. Interestingly, previous reports from our lab demonstrated that PGRN deficiency induced alterations in mitochondrial/intrinsic apoptotic cell death [14][15][16]. Since apoptosis has been largely related to neuronal cell death in FTLD [46] it is likely that the mitochondrial impairment caused by PGRN deficiency may be one of the factors contributing to neuronal death in FTLD-TDP.
Further analysis of the bioenergetics status of our cell model showed that PGRN deficiency was associated with lower oxygen consumption. We also observed that GRN KD cells reached poor maximal respiration rates upon addition of the FCCP uncoupler, compared with control cells. These results along with the mitochondrial depolarization and the increased ROS production in GRN KD cells suggested that mitochondrial respiration might be inhibited in a complex I-dependent manner [47,48]. Previous reports demonstrated that in ALS and FTLD models, TDP-43 bound to the mitochondrial mRNA and impaired the expression of the complex I subunits ND3 and ND6 causing complex I disassembly. Similar to our results, in these reports the dysfunctional complex I resulted in increased ROS production, mitochondrial depolarization and reduced ATP production [40,41]. Together, this evidence demonstrates that in FTLD-TDP associated with PGRN deficiency, the inhibition of mitochondrial respiration may be due to a complex I deficiency caused by the accumulation of aberrant forms of TDP-43 protein. Interestingly, lymphoblasts from FTLD-TDP patients carrying a GRN LOF mutation also exhibited mitochondrial depolarization and reduced ATP levels, suggesting that the mitochondrial impairment could be a main feature of FTLD-TDP associated with GRN mutations.
Because mitochondria are crucial organelles for maintaining the physiological activity of cells, damaged mitochondria are rapidly degraded. Interestingly, PGRN-deficient cells showed accumulation of depolarized mitochondria. Among other causes, the accumulation of dysfunctional mitochondria could be associated with a defect in mitochondrial degradation. Mitochondrial degradation is regulated by mitochondrial dynamics and mitophagy [49]. Mitochondria are dynamic organelles that constantly undergo fission and fusion events. Whereas fusion helps maintain mitochondrial function, the fission process enables damaged mitochondria to be removed from the mitochondrial network for degradation by mitophagy. Our results showed that PGRN deficiency favored mitochondrial fission and the initiation of the mitophagy process, allocating the damaged mitochondria of GRN KD cells to degradation. These results agree with previous reports showing that depolarized mitochondria were degraded by mitophagy in vivo and in vitro [50][51][52]. However, although in PGRN-deficient cells the depolarized mitochondria initiated the mitophagy process, their degradation in the lysosomes was not completed. It has been demonstrated that PGRN regulates the autophagy-lysosomal pathway and that PGRN deficiency induces autophagy impairment [35,36]. In agreement with these reports, we found that PGRN deficiency induced autophagy failure in SH-SY5Y cells, implying that the accumulation of damaged mitochondria in GRN KD cells was due to a defect in the autophagy-lysosomal pathway and not to a failure of mitochondrial dynamics or mitophagy initiation.
Our findings of altered mitochondrial bioenergetics, dynamics and mitophagy in PGRN-deficient cells agree with previous reports showing that PGRN acts as a regulator of mitochondrial homeostasis [13] and activity [12]. However, these reports did not demonstrate whether the effect of PGRN in regulating mitochondrial function was direct or indirect. The fact that the inhibition of TDP-43 phosphorylation restored the mitochondrial membrane potential in GRN KD cells suggests that the accumulation of phospho-TDP-43 protein might be the responsible for the mitochondrial impairment observed in PGRN-deficient cells. Several studies using murine and cell models of ALS and FTLD overexpressing wild type or mutant TDP-43 have demonstrated a link between TDP-43 and mitochondria [10,[39][40][41]53], describing that TDP-43 localizes to mitochondria, causing mitochondrial damage and a reduction in mitochondrial ATP synthesis [40,41]. Our results are consistent with these previously published data and support the hypothesis that TDP-43 pathology could play an important role inducing mitochondrial dysfunction in FTLD-TDP patients carrying LOF GRN mutations. On the other hand, PGRN deficiency might also affect the mitochondrial homeostasis through other pathways unrelated to TDP-43. For example, we previously reported an overactivation of Wnt signaling in GRN KD SH-SY5Y cells and lymphoblasts from FTLD-TDP patients carrying a LOF GRN mutation [15,54,55]. Both canonical and non-canonical Wnt signaling pathways have been implicated in mitochondrial dynamics and biogenesis [56][57][58]. Thus, the impairment of Wnt signaling in GRN KD cells might also contribute to the mitochondrial dysfunction. Interestingly, more recent reports have demonstrated a mitochondrial-initiated regulation of Wnt signaling [59,60], implying a bidirectional crosstalk between mitochondria and the Wnt pathway and suggesting that mitochondrial impairment might also be responsible for the alterations in Wnt pathway observed in PGRN-deficient cells.
As is summarized in Figure 6, this study describes that the partial loss of PGRN provokes the imbalance of mitochondrial bioenergetics in a neuronal-like cell model and patient-derived lymphoblasts, which might contribute to the neuronal death in FTLD-TDP. Furthermore, it demonstrates that the autophagy failure associated with PGRN deficiency blocks the degradation of impaired mitochondria in the lysosomes leading to the aberrant accumulation of damaged mitochondria in FTLD-TDP cellular models. Our results also point out that the TDP-43 pathology contributes to the mitochondrial damage observed in GRN KD cells. Interestingly, the treatment with brain penetrant phospho-TDP-43 inhibitors restores mitochondrial function in PGRN-deficient cells, suggesting that the regulation of TDP-43 pathology might prevent neuronal death in FTLD-TDP and other TDP-43-related pathologies by reverting or preventing mitochondrial dysfunction.  [10,[39][40][41]53], describing that TDP-43 localizes to mitochondria, causi mitochondrial damage and a reduction in mitochondrial ATP synthesis [40,41]. Our sults are consistent with these previously published data and support the hypothesis t TDP-43 pathology could play an important role inducing mitochondrial dysfunction FTLD-TDP patients carrying LOF GRN mutations. On the other hand, PGRN deficien might also affect the mitochondrial homeostasis through other pathways unrelated TDP-43. For example, we previously reported an overactivation of Wnt signaling in GR KD SH-SY5Y cells and lymphoblasts from FTLD-TDP patients carrying a LOF GRN m tation [15,54,55]. Both canonical and non-canonical Wnt signaling pathways have be implicated in mitochondrial dynamics and biogenesis [56][57][58]. Thus, the impairment Wnt signaling in GRN KD cells might also contribute to the mitochondrial dysfuncti Interestingly, more recent reports have demonstrated a mitochondrial-initiated regulat of Wnt signaling [59,60], implying a bidirectional crosstalk between mitochondria and t Wnt pathway and suggesting that mitochondrial impairment might also be responsi for the alterations in Wnt pathway observed in PGRN-deficient cells.
As is summarized in Figure 6, this study describes that the partial loss of PGRN p vokes the imbalance of mitochondrial bioenergetics in a neuronal-like cell model and p tient-derived lymphoblasts, which might contribute to the neuronal death in FTLD-TD Furthermore, it demonstrates that the autophagy failure associated with PGRN deficien blocks the degradation of impaired mitochondria in the lysosomes leading to the aberra accumulation of damaged mitochondria in FTLD-TDP cellular models. Our results a point out that the TDP-43 pathology contributes to the mitochondrial damage observ in GRN KD cells. Interestingly, the treatment with brain penetrant phospho-TDP-43 hibitors restores mitochondrial function in PGRN-deficient cells, suggesting that the r ulation of TDP-43 pathology might prevent neuronal death in FTLD-TDP and other TD 43-related pathologies by reverting or preventing mitochondrial dysfunction.

Conclusions
This study demonstrates that PGRN deficiency causes mitochondrial dysfunction in an FTLD-TDP cell model and in lymphoblasts derived from FTLD-TDP patients, suggesting that mitochondria may be damaged in FTLD-TDP associated with LOF GRN mutations. Furthermore, we found that the autophagy failure associated with PGRN deficiency affects mitochondrial degradation, leading to the accumulation of damaged mitochondria. Importantly, our results show that the treatment with phospho-TDP-43 inhibitors restores mitochondrial function in PGRN-deficient cells, suggesting that the mitochondrial depolarization could be a consequence of TDP-43 pathology. Furthermore, this study points to the use of drugs targeting TDP-43 as promising therapies to restore mitochondrial function in FTLD-TDP and other TDP-43-related diseases.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/antiox12030581/s1. All data generated during this study are included in this published article and its supplementary information files. The raw datasets and western blot images are available from the corresponding authors upon reasonable request.  Institutional Review Board Statement: Peripheral blood samples of all the individuals enrolled in this study were taken after obtaining written informed consent of the patients or their relatives. All study protocols were approved by the Donostia Hospital and the Spanish Council of Higher Research Institutional Review Board and are in accordance with National and International Guidelines (Declaration of Helsinki).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data is contained within the article and supplementary materials.

Conflicts of Interest:
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.