TRPV4 Deficiency Shifts Mitochondrial Dynamics Toward a Fragmented Morphology in Primary Microglia

Highlights

What are the main findings?

• Trpv4 knockout exhibits mitochondrial fragmentation and increased fission.
• Acute inhibition of TRPV4 increases non-mitochondrial respiration and does not induce fragmentation of the mitochondrial network.

What is the implication of the main finding?

• The constitutive absence of TRPV4 activity alters the morphology of the mitochondrial network without affecting the basal metabolic function of the organelle.

Abstract

Microglia perform surveillance and phagocytosis to maintain the homeostasis of the central nervous system (CNS). These processes are energetically demanding, and given the critical roles of mitochondria in providing ATP, the characteristics of the mitochondrial network can modulate microglial behavior. Although the Ca²⁺-permeable Transient Receptor Potential Vanilloid 4 (TRPV4) is known for regulating microglial morphology and migration, and it is implicated in mitochondrial calcium uptake, it is unknown whether TRPV4 affects the mitochondrial network in microglia. Our study provides evidence that TRPV4 plays a role in the integrity and complexity of the mitochondrial network in microglia. Quantification of the Mitochondrial Fragmentation and Complexity Index (MFCI) and increased pDrp1 (Ser616) showed a shift towards mitochondrial network fragmentation, and lowered complexity in Trpv4 knockout versus wild-type primary murine microglia in vitro. The distribution of mitochondria within microglia showed significant differences in density at 10–32 µm away from the nucleus. Furthermore, acute pharmacological TRPV4 inhibition with GSK2193874 did not induce significant mitochondria network fragmentation. Our findings establish TRPV4 as a regulator of mitochondrial dynamics and adaptive responses, highlighting its importance for maintaining homeostasis in microglia and the entire CNS.

1. Introduction

Microglia, the sentinels of the central nervous system, play a pivotal role in homeostasis, synaptic remodeling, and response to injury or disease [1,2,3]. They maintain CNS homeostasis by constantly monitoring the microenvironment and rapidly adapting to insults within the CNS and to systemic immune system signals that can pass the blood-brain barrier (BBB), underscoring their state plasticity [1,4]. As microglia functionally adapt to their environment, they undergo dynamic morphological changes within a spectrum of highly ramified to amoeboid appearances. Microglial motility and these morphological shifts depend on a complex cytoskeletal network involving actin and tubulin remodeling [5].

Central to the dynamic behavior of microglia is precise regulation of intracellular Ca²⁺ transients, which guide process motility, migration, and cellular activation [6]. The Transient Receptor Potential Vanilloid 4 (TRPV4) channel is among those implicated in this regulation. It is a widely expressed, non-selective cation channel, responsive to diverse physical and chemical cues, including mechanical stress, osmotic changes, heat, pH, and shear stress. TRPV4 regulates microglial migration in a temperature-dependent manner and is implicated in responses to osmotic and mechanical stress, often through calcium signaling and eicosanoid production [7]. Pharmacological inhibition of TRPV4 halts both the actin and tubulin cytoskeletal dynamism in microglia [8].

To sustain their constant surveillance and rapid responses, microglia require robust cellular machinery, with mitochondria playing a crucial role in providing the energy and metabolic regulation necessary for these functions [9,10,11]. Mitochondria maintain the cytosolic calcium levels by controlling ATP production, and by regulating Ca²⁺ uptake of the endoplasmic reticulum, through continuously balancing fission (division) and fusion (merging) [12,13,14,15]. This dynamic process is governed by a set of key proteins, with fission mediated by Dynamin-related protein 1 (Drp1) and its regulators (Mff, MiD49, and MiD51), and fusion involving Mitofusin 1 and 2 (Mfn1/2) on the outer membrane and optic atrophy 1 (Opa1) on the inner membrane [15]. While physiological mitochondrial calcium uptake stimulates energy production, calcium overload can cause reactive oxygen species (ROS) generation, ATP disruption, and organelle damage [16]. Disruption of mitochondrial fission and fusion processes can lead to a fragmented network composed of numerous small mitochondria, or conversely, to a hyperfused network of elongated, connected mitochondria [17]. This imbalance is correlated with a series of diseases, such as Charcot-Marie-Tooth disease and dominant optic atrophy [18]. A balanced dynamic process is needed to ensure adequate mitochondrial function but also to respond to the cell’s needs. This is carried out by adapting the network to nutrient availability and the metabolic state of the cell. Mitochondrial fission, also called fragmentation, is often associated with metabolic dysfunction and disease, as this morphological state dominates during stress events and cell death. However, a certain level of fragmentation is needed for mitochondrial motility, inheritance of mitochondrial DNA (mtDNA), and autophagic clearance of mitochondria, known as mitophagy [17,19]. On the other hand, mitochondrial fusion is mainly associated with cell survival. Fusion is known to be positively associated with increased oxidative phosphorylation (OXPHOS) activity and the regulation of mtDNA, and high OXPHOS activity leads to elongation of the mitochondrial network. Consequently, loss of mitochondrial fusion leads to impaired OXPHOS, a decrease in mtDNA, and ROS production [19,20].

The Mitochondrial Calcium Uniporter (MCU) is the principal transporter responsible for Ca²⁺ entry across the inner mitochondrial membrane [21,22]. Increased mitochondrial Ca²⁺ directly regulates the proteins controlling mitochondrial dynamics, with the fission protein Drp1 being particularly sensitive to local Ca²⁺ levels [23,24,25]. The concentration of cytosolic Ca²⁺ is tightly controlled by various plasma membrane channels, such as Transient Receptor Potential (TRP) channels TRPV1 and TRPC3 [26,27], whose activity leads to Ca²⁺ influx that, when relayed through the MCU, connects external signals to the internal mitochondria control [26,28,29]. The TRPV4 ion channel, known for its localization on the plasma membrane, is also expressed on a subpopulation of mitochondria. This mitochondrial localization of TRPV4 has been shown to regulate mitochondrial function in several cell lines, as well as in human umbilical vein endothelial cells (HUVECs), and T cells [16,30,31]. TRPV4 localizes to the organelle via a conserved targeting sequence that allows it to function as a mitochondrial calcium importer [32]. This TRPV4-mediated Ca²⁺ influx is involved in controlling mitochondrial dynamics, as changes in local calcium concentration can influence the translocation and assembly of Drp1 on the outer mitochondrial membrane [33]. Furthermore, TRPV4 directly interacts with the mitochondrial fusion regulators Mfn1/2 [30], establishing its role in linking external signals to the precise control of mitochondrial fission and fusion machinery.

Although TRPV4 is known to influence microglial motility on the one hand and mitochondrial biology on the other hand, its influence on the mitochondrial network in microglia is unknown. Mitochondria are vital for cellular energy and calcium homeostasis, and the recent discovery of TRPV4’s presence in mitochondria as a calcium importer suggests that TRPV4-mediated signaling might affect mitochondrial network organization in microglia as well. Therefore, this study aims to elucidate the contribution of TRPV4 in regulating the dynamical distribution of mitochondria in microglia.

2. Materials and Methods

2.1. Animals

All experiments were performed using microglial cells obtained from 21-day-old Cx3cr1^eGFP/+ × Trpv4^+/+ (Trpv4 wild-type; WT) and Cx3cr1^eGFP/+ × Trpv4^−/− (Trpv4 knockout; KO) C57BL/6J littermates. The Cx3cr1^eGFP/eGFP mice used for breeding the experimental animals were obtained from the European Mouse Mutant Archive (EMMA) Institute, under the approval of Steffen Jung from Weizmann Institute of Science [34], and the Trpv4 KO mice were generated as previously described [35] and acquired from the Laboratory of Ion Channel Research at KU Leuven. The experimental animals used express cytosolic eGFP under the Cx3cr1 (C-X3-C motif chemokine receptor 1) promoter, to allow microglial cell visualization. Animals were housed in the animal facility, with a 12 h light/dark cycle. Food and water were offered ad libitum. The procedures used followed the EU Directive 2012/63/EU law for animal testing and were approved by the local ethical committee (Ethical Commission for Animal Experimentation, UHasselt, Diepenbeek, Belgium).

2.2. Microglia Isolation and Culture

Primary microglia cultures were obtained from the cortices of postnatal day (P) 21 mice using the magnetic cell sorting activating system (MACS) as previously described [36]. Pups of both sexes were included in the cultures. In short, the mice were sacrificed, and the brain tissue was collected in a sterile manner. The cortices were isolated, then mechanically triturated and enzymatically digested with papain (17 U/mg, Thermo Fisher, Waltham, MA, USA) and DNase (10 mg/mL, Roche, Basel, Switzerland) for 30 min at 37 °C. After digestion, the cells were strained (70 µm, EASYstainer, Greiner Bio-One, Kremsmünster, Austria) and centrifuged to remove debris. To isolate the mononuclear cells from the suspension, a Percoll (Sigma-Aldrich, St. Louis, MO, USA) gradient centrifugation (30%, 70%) was performed. The monolayer was collected and rinsed, then incubated for 15 min with MACS buffer (0.5% FBS, 0.5% EDTA in PBS) and 10 µL of CD11b beads (130-049-601, Miltenyi Biotec, Bergisch Gladbach, Germany) to collect the microglia from the healthy brains. Excess beads were removed, and the remaining cells were strained through an MS column (130-042-201, Miltenyi Biotec) to separate the CD11b positive and negative fractions. The positive fraction was cultured for 7 days in medium supplemented with 10% fetal bovine serum (FBS, Gibco^TM, Grand Island, NY, USA), 10% horse serum (HS, Gibco^TM), and 1% P/S, referred to as 10:10:1 medium, in an incubator, at 37 °C, in a humidified atmosphere with 5% CO₂. All cell culture reagents, including the catalogue number and suppliers, are detailed in Supplementary Table S1.

The positive fraction of CD11b cells was seeded on coverslips in 24-well plates, and in glass bottom dishes (MatTek Life Sciences, Ashland, MA, USA) fitted with 2 well-inserts (ibidi GmbH, Gräfelfing, Germany), at a density of 350,000 cells/mL (in 500 µL, and 70 µL, respectively). The culture surfaces were preliminary coated with Poly-D-Lysine solution (PDL, 20 µg/mL in sterile PBS 1X, Gibco^TM) for 2 h, and collagen IV solution (2 µg/mL, Sigma-Aldrich) until the cells were ready to be seeded. After a week of culture in 10:10:1 medium, the medium was replaced with TIC medium (short for TGF-β, IL-34, and cholesterol), to ensure survival, branching, and a consistent gene expression profile [37,38]. This medium consists of DMEM F-12 (Gibco™) supplemented with insulin (5 µg/mL, Sigma-Aldrich), N-acetyl-Cysteine (5 µg/mL, Sigma-Aldrich), Apo Transferrin (100 µg/mL, Sigma-Aldrich), Heparan sulphate (1 µg/mL, Sigma-Aldrich), human TGF-β (2 ng/mL, Peprotech, Cranbury, NJ, USA), murine IL-34 (100 ng/mL, Bio-techne, Minneapolis, MN, USA), sodium-selenite (100 ng/mL, Sigma-Aldrich), cholesterol (1.5 µg/mL, Sigma-Aldrich), and L-glutamine (2 mM, Gibco^TM), reagent-specific details are listed in (Supplementary Table S1). The cultures were maintained for 5–7 days, after which experimental procedures were started.

2.3. Bone Marrow-Derived Macrophages Isolation and Culture

Bone marrow-derived macrophages (BMDMs) were isolated from 9-week-old female mice, two WT and three Trpv4 KO animals, as previously described [39,40]. Briefly, the hind legs were dissected, and the femur and tibia were cleaned of muscle tissue before being sterilized in 70% ethanol. The epiphyses were removed, and bone marrow was flushed with sterile PBS using a syringe with a 25G needle. The resulting cell suspension was centrifuged, and the pellet was resuspended in BMDM differentiation medium. This medium consisted of RPMI 1640 (Gibco^TM) supplemented with 10% FBS (Gibco^TM), 0.5% P/S, and 15% L929-conditioned medium (LCM, in-house) as a source of macrophage colony-stimulating factor (M-CSF).

After the bone marrow tissue was processed, the cell suspension was seeded in 100 mm Petri Dishes and cultured in complete medium (RPMI 10% FBS, 0.5% P/S, 15% LCM) at 37 °C, in a humidified atmosphere with 5% CO₂. On day 4 in vitro, half of the medium was replenished with fresh complete medium. One week after seeding, the cells differentiated into mature BMDMs, and they were detached via EDTA (1:50 in PBS, Sigma-Aldrich) incubation, followed by gentle scraping. The cells were collected, centrifuged, and resuspended in medium containing 5% LCM. Finally, BMDMs were seeded for metabolic assay experiments at a density of 210,000 cells/mL and allowed to adhere overnight prior to the initiation of the experiments.

2.4. Immunocytochemistry

Once the cells were in culture for one week, they were fixed using ice-cold 4% paraformaldehyde (PFA, Sigma-Aldrich) and sucrose (5%, Fisher BioReagents, Waltham, MA, USA) for 15 min at room temperature (RT). The fixing solution was rinsed out with PBS, then incubated with bovine serum albumin solution (5% BSA in PBS, Sigma-Aldrich) for 1 h at RT. To analyze mitochondrial morphology and dynamics, cells were either stained only against HSP60 (HB7863, 1:2000, HelloBio, Bristol, UK), used to stain the mitochondrial matrix, or in combination with either Mfn1 (13798-1-AP, 1:500, Proteintech, Rosemont, IL, USA) or pDrp1 (Ser616) (PA5-64821, 1:500, Invitrogen^TM, Carlsbad, CA, USA) (Supplementary Table S1). Additionally, for the mitochondrial dynamics proteins, a group of WT cells was pretreated with staurosporine (1 µM, MCE, Monmouth Junction, NJ, USA) for three hours prior to the start of the protocol to serve as a positive control of apoptosis-induced mitochondrial fragmentation.

Primary antibodies were incubated overnight at 4 °C on an orbital shaker. For visualization, either Alexa Fluor 647 goat anti-mouse (A-21235, 1:1000, Invitrogen^TM) or Alexa Fluor 568 goat anti-mouse (A-11031, 1:1000, Invitrogen^TM) antibodies were used for marking HSP60, and Alexa Fluor 647 donkey anti-rabbit (A-31573, 1:1000, Invitrogen^TM) was used for Mfn1 and pDrp1, respectively. The secondary antibodies were incubated for 1 h at RT on an orbital shaker. Sequentially, the nuclei were counterstained with DAPI (1:10,000, Invitrogen^TM) and the coverslips were mounted on microscope slides using Fluoromount-G (Invitrogen^TM) (Supplementary Table S1). Images from 10 regions of interest (N = 3 independent cultures containing WT and Trpv4 KO samples, 5 regions per technical replicate) were acquired using Plan-ApoChromat 20×/0.8 NA air and Plan-ApoChromat 63×/1.4 NA oil immersion objectives on the LSM900 confocal microscope (Zeiss, Oberkochen, Germany).

2.5. Mitochondrial Network Characteristics and Parameters

Mitochondrial morphology and network characteristics were quantified from the stacked confocal fluorescent images (0.14 µm interval over a 2.38–5.18 µm range; 17–37 slices) obtained using the Mitochondrial Analyzer plugin within FIJI (ImageJ, version 2.16.0/1.54p; Java 1.8.0_332) [41,42]. The images were processed by a standardized procedure that included background subtraction to remove uneven illumination, noise reduction, and contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE) to optimize signal detection. Mitochondrial objects were then segmented from the processed images using adaptive thresholding, which locally determines thresholds to separate individual mitochondria and network parameters (

Table 1. Mitochondria network parameters.

Parameter	Description	Represents
Morphological parameters
Count	Number of identified mitochondria per cell	Organelle abundance
Total volume	The sum of all mitochondrial volumes in an image	The space the network occupies inside the cytoplasm
Total surface area	The sum of all the surface areas of the mitochondria detected	Cumulative area occupied by mitochondria
Sphericity	1—spherical objects 0—irregular objects	Cells with values closer to 1 have more spherical, compact mitochondria; values closed to 0 indicate irregular, less compact mitochondria
Network parameters
Branches	Total number of branches in an image	Network complexity measure
Branch junctions	Number of intersections of two or more branches within the skeletonized mask
Brach endpoints	Total number of branch endings that do not connect with another branch	Number of terminal ends in the network or loose mitochondria

).

Mitochondria network morphology was quantitatively assessed using the Mitochondrial Fragmentation and Complexity Index (MFCI), which reflects mitochondrial dynamism. This index integrates two components: the degree of fragmentation and the extent of the network complexity. Fragmentation was quantified by the ratio of mitochondrial number to total mitochondrial volume, where an increase signifies a greater number of smaller mitochondria. Complexity was determined by the ratio of the number of branch junctions (network nodes) to the mitochondria number, with a higher value indicating greater branching and density of the mitochondrial network. These parameters were combined into the MFCI formula:

$$Mitochondria Fragmentation and Complexity Index (MFCI)={\displaystyle \frac{{(\mathrm{m}\mathrm{i}\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{h}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{r}\mathrm{i}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r})}^2}{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{m}\mathrm{i}\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{h}\mathrm{o}\mathrm{n}\mathrm{d}\mathrm{r}\mathrm{i}\mathrm{a}\mathrm{l} \mathrm{v}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{m}\mathrm{e} \times \# \mathrm{b}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{h} \mathrm{j}\mathrm{u}\mathrm{n}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}}}$$

A lower MFCI corresponds to a healthier, highly fused, and structurally complex mitochondrial network, while a higher MFCI signifies a shift towards fragmentation. The MFCI was calculated for each cell analyzed (N = 3, 26 cells selected per group) and checked for statistical significance. Genotype-specific differences were also investigated by subdividing both the WT and Trpv4 KO groups into three distinct cellular categories: branched (B), bipolar (bip), and lamellipodium with a trailing edge (LTE), based on their morphology to facilitate the comparison (WT, n_B = 10, n_bip= 9, n_LTE = 7, and Trpv4 KO, n_B = 9, n_bip = 8, n_LTE = 9). Mitochondrial parameters of interest were checked for statistical significance.

2.6. Mitochondrial Dynamics Proteins Fluorescence Analysis

Mitochondrial fusion protein Mfn1 and fission activation protein phosphorylated Dynamin protein 1 (pDrp1 at Ser616) were quantified from fluorescently labelled mitochondria of WT, WT treated with staurosporine, and Trpv4 KO fixed microglia. High-resolution confocal images (0.19 µm interval over a 4–7.5 µm range; 21–38 slices) were pre-processed and then analyzed using the Mitochondria Analyzer plugin available in FIJI (ImageJ) [41,42]. The cleaned images were thresholded to identify the individual mitochondria, and their parameters (volume, branches, and branch junctions) were measured. The intensities of either Mfn1 or pDrp1 were measured for each identified mitochondrion and then normalized by the volume of the individual mitochondrion, accounting for potential differences in the shape and size of mitochondria. These ratios were averaged from all mitochondria per cell for each experimental group (Mfn1: N = 2, WT n = 10 cells, WT + STS n = 9 cells, Trpv4 KO n = 10 cells; pDrp1: N = 2, WT n = 11 cells, WT + STS n = 7 cells, Trpv4 KO n = 5 cells) and tested for statistical significance.

2.7. Mitochondrial Dynamics Protein Extraction and Western Blot

Protein expression was assessed by SDS-PAGE and Western blotting in WT and Trpv4 KO BMDM cells. They were lysed in 250 µL RIPA buffer, supplemented with protease inhibitors (Roche) and phosphatase inhibitor (1:10, Targetmol, Boston, MA, USA). Protein concentration (N = 3 animals per genotype) was determined using the Pierce BCA Assay Kit (Thermo Fisher). Samples were prepared with equal amounts of protein: 0.7 µg protein for Mfn1 and 2.5 µg protein for pDrp1. Following protein separation on 12% SDS-PAGE gels, samples were transferred to PVDF membranes (VWR, Radnor, PA, USA), which were blocked in 2% BSA (Sigma-Aldrich) for one hour at RT. Primary antibodies—Mfn1 (13798-1-AP, 1:1000, Proteintech), pDrp1 (PA5-64821, 1:500, Invitrogen^TM), total Drp1 (12957-1-AP, 1:1000, Proteintech), β-actin (sc-47778, 1:2000, Santa Cruz, Dallas, TX, USA), GAPDH (sc-137179, 1:2000, Santa Cruz)—were incubated overnight at 4 °C. Subsequently, the HRP-conjugated anti-rabbit (1:2000, Biotium, Fremont, CA, USA) secondary antibody was incubated for one hour at RT. The signals were detected using ECL substrate (Bio-Rad Clarity, Hercules, CA, USA) and the Amersham Imager 680 (GE Healthcare, Cytiva, Marlborough, MA, USA), and the band intensity was measured using FIJI (ImageJ), then normalized to β-actin (for Mfn1) or total Drp1 (for pDrp1). Full Western blot membranes are provided in Supplementary Figure S1.

2.8. Mitochondrial Density Distribution Analysis

Confocal images were processed using FIJI (ImageJ) [42], and we performed a Sholl-like analysis to compare mitochondrial density distribution within the three most frequent microglial morphologies between WT (N = 3, 26 cells analyzed, n_B = 10, n_bip = 9, n_LTE = 7) and Trpv4 KO (N = 3, 26 cells analyzed, n_B = 9, n_bip= 8, n_LTE = 9) cells. Maximum intensity projection images were created of the mitochondria (HSP60 signal) and of the microglia (cytosolic eGFP signal). The parameters of confocal image acquisition were 0.14 µm interval over a 2.38–5.18 µm range, resulting in 17–37 slices. Concentric shells were generated around the center of the nucleus with a 2 µm step size, extending to the cell extremities. One cell at a time was analyzed by creating a region of interest (ROI) with the free-hand tool in Fiji, followed by the ‘Clear outside’ command to make sure no other source of intensity was measured. The eGFP image was binarized, and its integrated density (total cell area) was measured. The integrated density of the mitochondrial signal was calculated for each shell. To calculate the relative mitochondrial density within each radial shell, these values were corrected according to the formula:

$$Mitochondrial density={\displaystyle \frac{Shell value}{Total cell value}} \times 100$$

The distribution of these relative pixel percentages was plotted to determine where the mitochondrial density is highest and if the organelles extend all the way to the ends of the cell extremities. The data were compared first by genotype and then for each morphology or subtype individually.

2.9. Mitochondrial Live Imaging

To assess the effects of acute TRPV4 inhibition on an intact mitochondrial network, we analyzed the network parameters previously described on recordings from live cells that were administered GSK2193874, a specific TRPV4 blocker. After being cultured for a week in TIC medium, WT microglia were incubated with MitoTracker Deep Red FM (Invitrogen^TM, Thermo Fisher) prepared in DMEM F-12 at 10 pM/mL (1:10⁵ dilution) for 20 min in an incubator, at 37 °C, in a humidified atmosphere with 5% CO₂. After incubation, the medium was removed and replaced with OptiMEM (Gibco^TM). Live confocal time-lapse images were acquired using a C-ApoChromat 63×/1.2 NA water immersion objective on the LSM880 confocal microscope (Zeiss). During imaging, the focus was set on the microglial branches, and Z-stacks (0.53 µm interval over a 3.18–5.3 µm total range; 6–10 slices) were selected to encapsulate the entire thickness of the branch. To accurately assess mitochondrial network morphology and complexity, our analysis was restricted to mitochondria within stable—i.e., non-moving—cellular branches, as it is challenging to disentangle mitochondrial movement from cellular displacement in moving microglial branches during live recordings. The parameters used for the recording are: 30 cycles, every 35 s, for a total of 15 min of recording, both the cytosolic eGFP channel, as well as the mitochondria dye channel. After baseline measurements, a selective TRPV4 inhibitor, GSK2193874 (GSK21, 10 µM, Tocris Bioscience, Bristol, UK) [43] or vehicle control (0.01% DMSO in OptiMEM, Gibco^TM) was administered, and the branch was recorded again (Supplementary Table S1). Images were processed, and the branches were isolated in an ROI; the rest of the image was excluded from the analysis. The same Mitochondrial Analyzer plugin was used to assess whether there are statistically significant differences between the baseline and the inhibition of TRPV4 in the selected ROI. Six independent experiments were performed, where at least one stable branch per cell, and at least two cells per treatment group (Vehicle control or TRPV4 inhibition) were recorded (N = 6 independent cultures, min. 4 cells per experiment, 28 cells in total, 14 cells per group, from each cell a branch was analyzed); all obtained values were normalized using the last 5 min (12 cycles) of the baseline recording and compared to the values from the vehicle control experiment recordings that were done for each experiment. The values obtained were preliminarily normalized either by the volume of the branch (from the eGFP channel) or by the number of mitochondria identified. Afterwards, the time courses obtained were normalized to the stable phase (the last five minutes of the baseline recording) and represented graphically. To assess the effect of the administered treatment, the values from all the time points were averaged per parameter, the overall treatment effect, and the interaction between time and treatment were tested for statistical significance.

2.10. Seahorse Extracellular Flux Assay

To evaluate the metabolic profile and respiratory capacity of BMDMs, real-time measurements of the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were performed using a Seahorse XF Pro M analyzer (Agilent, Santa Clara, CA, USA). BMDMs for WT (N = 2 animals) and Trpv4 KO (N = 3 animals) mice were seeded in 96-well Seahorse XF Pro M cell culture plates at a density of 37,500 cells per well in differentiation medium. On the day of the assay, the culture medium was replaced with warm Seahorse XF medium (Agilent) supplemented with 2 mM glutamine (Gibco^TM), 1 mM sodium pyruvate (Gibco^TM), and 10 mM glucose (Gibco^TM). Cells were incubated in a non-CO₂ incubator at 37 °C for 60 min prior to the start of the assay to allow for temperature and pH equilibration.

MitoStress Test protocol [44] was employed to assess the metabolic impact of TRPV4 inhibition. The sensor cartridge was loaded with compounds as follows: port A—either the TRPV4-specific inhibitor GSK2193874 (final concentration 1 µM) or vehicle control (final concentration 0.01% DMSO); port B—Oligomycin (ATP synthase inhibitor; final concentration 1.5 µM); port C—carbonyl cyanide-4 (phenylhydrazone) piperidine (FCCP; uncoupling agent; final concentration 1.5 µM); port D—combination of Antimycin A and Rotenone (mitochondrial complex III and I inhibitors; final concentrations 2.5 µM and 1.25 µM), co-injected with Hoechst 33342 (Thermo Fisher, Waltham, MA, USA) (final dilution 1:2500) to facilitate nuclear staining for cell normalization (Supplementary Table S1).

Each measurement cycle consisted of three minutes of mixing followed by three minutes of measurement. Following the completion of the assay, cell counts were determined by imaging the center of each well on a Nikon Eclipse Ti2-E microscope (Nikon Corporation, Tokyo, Japan). To confirm culture integrity and equal cell distribution across genotypes, representative images of the nuclear staining were captured post-assay (Supplementary Figure S2). The cell counts were used to normalize OCR and ECAR values, and a quality control check was performed to remove replicates with measurement artifacts. For statistical analysis, technical replicates from all animals were pooled for each experimental group (n_WT = 9, n_WT+GSK21 = 9, n_KO = 23, n_KO+GSK21 = 10). The BMDM cultures used in this study were prepared according to standardized protocols previously characterized and published by our group and co-authors [44].

2.11. Statistical Analysis

For all statistical analyses, GraphPad Prism 10 (version 10.1.2) and R Studio (version 2025.09.1-401) were used. The mitochondria density distribution data were checked for outliers using the Grubb’s test. Normality of the data was checked using the Shapiro–Wilk test. The data obtained from the Mitochondria Analyzer plugin were tested for significance using the Mann–Whitney test, in the comparison of the two genotypes (N = 3, 26 cells per group), and using the Ordinary One-Way ANOVA, in the subtype comparisons, followed by a Tukey’s post-hoc test (N = 3, WT, n_B = 10, n_bip = 9, n_LTE = 7 and Trpv4 KO, n_B = 9, n_bip = 8, n_LTE = 9). Differences in average Mfn1 (WT N =10 cells, WT + STS N = 9 cells, Trpv4 KO N = 10 cells), and pDrp1 (Ser616) (N = 11 cells, WT + STS N = 7 cells, Trpv4 KO N = 5 cells), were tested for normal distribution then by One-way ANOVA with Tukey’s post-hoc test, and Kruskal–Wallis test with Dunn’s multiple comparison test, respectively. The protein expression data set (N = 3 animals per genotype) was tested for normal distribution and by Welch’s t-test for significance. When comparing the cell area, the branch length from the nucleus between WT and KO, a two-way ANOVA test was used, accompanied by the Tukey multiple comparisons test or the unpaired t-test. To analyze the mitochondrial density distribution and the effect of acute TRPV4 inhibition, R Studio was used to apply the Linear Mixed-Effects Model (LMM), paired with a Type III ANOVA analysis, to test for fixed (genotype) and random (radial distance) effects in the data, and to generate an overall p-value per group. For the MitoTracker acute inhibition experiment, the MFCI for every group was calculated and compared using the Two-Way ANOVA with Tukey’s multiple comparisons test to check for significance. The time-course and fold change between the groups (N = 6, 14 branches analyzed per treatment group) were tested by the Two-way repeated measures ANOVA. Where appropriate, Tukey’s or Dunn–Šidák post-hoc tests were used to compare all individual time points, and separate t-tests were used for the comparisons of the last 4 cycles after treatment. To statistically test the results of the extracellular flux assay, 3 to 11 technical replicates per group were used for the statistical testing (WT, N = 2; Trpv4 KO, N = 3). These values were tested using a Two-way ANOVA, followed by Tukey’s post-hoc test. Data are shown as mean ± standard error of the mean (SEM). All values of p < 0.5 are considered statistically significant.

3. Results

3.1. Trpv4 Knockout Induces Fragmentation and Decreased Complexity of the Mitochondrial Network

We examined mitochondrial morphology and network complexity in primary cultured microglial cells. The murine microglia were isolated from Cx3cr1^eGFP/+ × Trpv4^+/+ (Trpv4 wild-type; WT) and Cx3cr1^eGFP/+ × Trpv4^−/− (Trpv4 knockout; KO) littermates and identified by their cytosolic green fluorescent protein (eGFP) expression. Mitochondria were stained using an antibody against Heat Shock Protein 60 (HSP60), a marker for the mitochondrial matrix (Figure 1A). Colocalization of eGFP-expressing cytoplasm and HSP60-stained mitochondria, alongside DAPI nuclear counterstaining, allowed for the assessment of cellular morphology. The analysis of the mitochondrial network was performed using a mitochondria binary mask (Figure 1A), and visible differences can be observed by zooming in on the analyzed image (Figure 1H).

Quantitative analysis of mitochondrial parameters revealed significant alterations in both mitochondrial morphology and network complexity within Trpv4 KO microglia compared to WT controls. This was carried out using the Mitochondrial Fragmentation and Complexity Index (MFCI) (Figure 1B), which is higher when the fragmentation is increased, and the complexity of the network is decreased. Specifically, Trpv4 KO cells exhibited a significant increase in mitochondrial number (Figure 1C). Quantification of the mean surface area per mitochondrion (total mitochondrial surface area divided by the number of mitochondria) was significantly reduced in Trpv4 KO microglia (Figure 1D). This finding suggests that individual mitochondria are smaller in the case of TRPV4 deficiency. This was further supported by a significant decrease in the mean mitochondrial volume in Trpv4 KO cells (Figure 1E), which was calculated using the 3D information from the z-stack confocal images acquired (0.14 µm interval over a 2.30–5.18 µm range; 17–37 slices). Besides the decrease in individual mitochondrial dimensions, the overall mitochondrial network in Trpv4 KO microglia displayed reduced complexity due to the increase in fragmentation (Figure 1B,C). The significant decrease in the sphericity parameter reveals that the Trpv4 KO cells have a more irregular shape, and they are less compacted than the WT counterparts (Figure 1F). The increase in the number of branch endpoints within the mitochondrial network (Figure 1G) suggests that while individual mitochondria may be smaller and more numerous, they are more fragmented and irregular. Taken together, the analysis showed a shift towards a more fragmented and less complexly organized mitochondrial population in Trpv4 KO microglia.

3.2. Loss of TRPV4 Shifts Mitochondrial Dynamics Equilibrium Toward Fission

To investigate the molecular mechanisms underlying the altered mitochondrial morphology in the Trpv4 KO cells, we quantified the mitochondrial recruitment of proteins involved in the fission-fusion equilibrium: the fusion protein Mitofusin-1 (Mfn1) and the fission activation protein, phosphorylated Dynamin-related protein 1 (pDrp1 at Ser616) (Figure 2A). Analysis of high-resolution confocal images (0.19 µm interval over a 4–7.5 µm range; 21–38 slices) revealed a small but not significant decrease in mitochondrial Mfn1 levels in Trpv4 KO cells compared to WT (Figure 2B).

In contrast, loss of TRPV4 activity resulted in a striking increase in the accumulation of pDrp1 (Ser616) in the cytosol, as well as at the mitochondrial network. Quantification of mitochondria-associated pDrp1 fluorescence showed a significant increase in the TRPV4-deficient group compared to WT controls (p = 0.0009, Figure 2C). Interestingly, while the pro-apoptotic stimulus staurosporine (STS) induced a slight upward trend in pDrp1 recruitment, it did not reach statistical significance in our experimental model. These results indicate that the absence of TRPV4 activity shifts the mitochondrial dynamics equilibrium toward fission, by increasing the recruitment and activation of the fission protein Drp1, without significant alterations in the fusion capacity.

To validate the changes observed in mitochondrial dynamics proteins, we complemented our microglial analysis with biochemical studies in bone marrow-derived macrophages (BMDMs). BMDMs offer a robust, high-yield alternative for investigating fundamental myeloid pathways.

Using Western blotting in BMDMs lysates from WT and Trpv4 KO mice (N = 3 animals per genotype, Figure 3 and Figure S2), we quantified total cellular protein pools. This biochemical approach allowed the quantification of total cellular protein pools, independent of mitochondrial volume or localized recruitment. Consistent with the imaging data, Trpv4 KO BMDMs exhibited a significant increase in the ratio of active pDrp1 (Ser616) to total Drp1 compared to WT controls (p = 0.0456, Figure 3B).

However, while Mfn1 recruitment per mitochondrial volume did not reduce significantly in Trpv4 KO microglia (Figure 2B), Western blot analysis of BMDM lysates revealed a notable trend toward a reduction in total cellular Mfn1 abundance in the absence of TRPV4 (p = 0.0511, Figure 3A). These results suggest that while the mitochondrial occupancy is similar between Trpv4 KO and WT cells, the total cellular reservoir is strongly decreased in Trpv4 KO cells.

Our results suggest that the effects on mitochondrial markers align across both microglia and BMDMs, suggesting that TRPV4 plays a canonical role in regulating mitochondrial homeostasis in these myeloid populations.

3.3. TRPV4 Deficiency Increases the Spatial Distribution of Mitochondria in Microglia

To investigate the impact of TRPV4 on mitochondrial distribution in primary murine microglia, we performed a density distribution analysis on WT and Trpv4 KO microglia. Because microglia can adopt very distinct morphologies and the mitochondrial network will depend strongly on the cell shape, we subdivided the primary microglia across three distinct categories, based on cellular morphology: branched, bipolar, and lamellipodium with a trailing edge (LTE) (Figure 4A) [4]. For the mitochondrial distribution analysis, the nucleus was designated as the central point with concentric circles (shells) drawn at 2 µm intervals extending to the cellular extremities (Figure 4B), to assess the specific distribution of the mitochondria within the same subtype (exemplified for branched microglia in Figure 4C).

Initial analysis showed no significant differences in cell area (Figure 4D) or in the length of the branches, measured from the nucleus (Figure 4G), between WT and Trpv4 KO cells across all subtypes, which is in line with our previous findings [8]. Generally, mitochondria were predominantly situated in close proximity to the nucleus, forming a dense network, especially in the branched and LTE morphologies. Mitochondria in the cellular extremities were less dense and were spread along the length of the cell processes. Furthermore, mitochondrial distribution did not significantly differ among the distinct morphologies within either the WT (p = 0.978) or Trpv4 KO (p = 0.594) groups (Figure 4E,F).

However, a close to significant difference in mitochondrial distribution emerged when comparing WT and Trpv4 KO in branched microglia (p = 0.0545) (Figure 4H), but not in the bipolar (p = 0.25) and LTE morphology (p = 0.279) (Figure 4I,J). In the branched WT cells, mitochondrial density was peaking around the nucleus and sharply decreased towards the trailing edge. In contrast, branched Trpv4 KO cells exhibited a higher density of mitochondria around the nucleus, surpassing the level of the WT, and a gradual decrease in mitochondrial density from the nucleus to the extremities, suggesting a more spread-out distribution throughout the cell body compared to the WT cells. This is supported by the statistically significant differences observed between 10 and 32 µm away from the nucleus (Figure 4H). The bipolar cells have their mitochondria distributed in a very similar pattern, with no significant differences identified (Figure 4I). Even though the LTE Trpv4 KO cells do not show a high peak of mitochondria in the perinuclear regions, as the WT cells do, there are significantly more mitochondria distributed between 16 and 30 µm from the nucleus (Figure 4J). In conclusion, mitochondria in Trpv4 KO cells are more likely to have more mitochondria spread towards the branches (Figure 4H) and to localize further away from the nucleus than in WT cells, especially in the LTE morphology (Figure 4I).

3.4. Adaptive Mitochondrial Network Reorganization Is Lost in Trpv4-Deficient Microglial Morphologies

To explore how mitochondrial network parameters vary with microglial subtypes, we compared mitochondrial features within each genotype (WT and Trpv4 KO). Calculating the MFCI revealed significant differences between WT and Trpv4 KO in the branched and bipolar subtypes (Figure 5A), leaning the balance towards a high MFCI index for the knockout cells, highlighting fragmentation and loss of mitochondrial network complexity. Consistent across all three morphological categories, the number of mitochondria in each of the Trpv4 KO subtypes was higher than that of the WT counterpart (Figure 5B,D). Some differences were observed in the mean mitochondrial surface, with the branched and the LTE categories having smaller mitochondria in the Trpv4 KO versus the WT cells (Figure 5C). While many mitochondrial network complexity parameters did not significantly differ among the three morphologies, key indicators of mitochondrial fragmentation revealed significant changes in the WT group. The number of mitochondria (Figure 5D) was significantly higher in the bipolar morphology, as compared to the branched morphology. However, the mean surface area of a mitochondrion (Figure 5E) was significantly higher in the migratory LTE morphology, as compared to the bipolar one, indicating smaller mitochondria. Notably, no significant differences were observed in these parameters among the different morphologies in the Trpv4 KO group, revealing stable parameters across all subtypes (Figure 5D,E).

Taken together, these findings suggest that mitochondrial network organization in Trpv4 KO microglia is altered towards a fragmented, less complex state. In the WT cells, we observed distinct mitochondrial characteristics in the bipolar morphology, where more mitochondria were present (Figure 5D), but their size was smaller compared to the other morphologies, suggesting mitochondrial fragmentation, a result supported by the trend in the MFCI (Figure 5A,E). In contrast, Trpv4 KO microglia lack these changes within the subtypes, indicating that TRPV4 might be required for the coupling between morphological state and mitochondrial network dynamics.

3.5. Acute TRPV4 Inhibition Enhances Glycolytic Reserve Independently of Mitochondrial Network Remodeling

Mitochondria showed fragmentation and more widespread distribution in fixed primary microglia of mice in which Trpv4 is constitutively knocked out. To assess the acute effects of TRPV4 inhibition on mitochondrial network dynamics in live microglia, we performed time-lapse imaging on cells loaded with MitoTracker Deep Red FM. We recorded baseline mitochondrial behavior within stable cellular branches (Figure 6A,B), followed by the addition of GSK2193874 (GSK21, 10 µM), a selective TRPV4 antagonist. Because the antagonist is prepared in DMSO, a vehicle control test with DMSO (<0.01% in OptiMEM) was taken alongside each experiment to account for any potential solvent effects.

Analysis of the MFCI revealed no significant differences between the baseline and the TRPV4 inhibition (GSK21) treatment (Figure 6C). A non-significant trend towards a lower MFCI value was observed upon inhibition, hinting towards a shift to increased mitochondrial network complexity in the stable microglial branches. While the static MFCI (Figure 6C) showed no significant changes, we performed a more comprehensive, time-resolved analysis (Figure 6D–G) to capture the dynamic nature of the mitochondrial network.

Quantitative analysis revealed consistent, but non-significant, increases across several measured mitochondrial network parameters following acute GSK21 treatment. Specifically, the number of mitochondria normalized to the branch volume remained stable and showed no significant difference compared to vehicle control (interaction between time and the treatment, p = 0.911) (Figure 6D). The mean surface area per mitochondrion (Figure 6E) appeared to increase after GSK21 administration as well, albeit not significant over time (p = 0.295), compared to the vehicle control. The mean volume per mitochondria (Figure 6F) shows a trend (p = 0.068) towards increased frequency of distended mitochondria. Meanwhile, the mitochondrial network complexity parameter, namely the mean branch junctions (number of branch junctions divided by number of mitochondria) (Figure 6G) shows a slightly elevated number of branching points. Taken together with the apparent increase in mean surface area (Figure 6E) at the same timepoints, these data hint towards elongation or slight enlargement of the mitochondrial network with increased complexity. Comparison of the mitochondria network parameters in the last 4 cycles after TRPV4 inhibition (Figure 6D–G) revealed no significant differences, even after 10 min post-treatment, indicating that the fragmented mitochondria morphology seen in the Trpv4 KO cells was not induced by the treatment.

To confirm these findings, the average fold change in all the timepoints was calculated and compared between the groups. The fold change in mitochondrial number reveals that the baseline recordings are quite stable, where the vehicle application and the TRPV4 inhibition show increased variation (Figure 6H). Mean surface area fold increases non-significantly after TRPV4 inhibition (GSK21) (Figure 6I). The fold change in the mean volume per mitochondrion, when averaged per cell, did not show any significant differences from the other groups (Figure 6J). Mean number of branch junctions fold change increased slightly in the TRPV4 inhibition group above the means of the other groups, without reaching significance (Figure 6K).

The trending increases in the mean volume and in the number of branch junctions per mitochondria suggest that acute TRPV4 inhibition by GSK21 did not induce the fragmented mitochondrial morphology that was observed in the chronic Trpv4 knockout model, at least in the time frame that is feasible in live measurements. This indicates a critical difference in how the mitochondrial network responds to the immediate absence of TRPV4 activity versus its long-term, constitutive loss.

To investigate if these morphological observations correlated with metabolic shifts, we utilized WT and Trpv4 KO BMDMs as a high-throughput model and performed real-time extracellular flux analysis (Seahorse MitoStress Test). Building on the validation shown in Figure 3, which identified BMDMs as a representative myeloid model for studying TRPV4-dependent mitochondrial function, we utilized this high-yield system to measure the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) (Figure 7A,B).

Baseline mitochondrial parameters, including basal respiration, ATP-linked energy production, and proton leak, were comparable across all genotypes and treatment groups, suggesting that the mitochondrial oxidative phosphorylation machinery remains intact despite the loss of TRPV4 calcium influx (Figure 7C–F). Post-assay imaging of nuclear density confirmed that these metabolic profiles were not influenced by differential cell adherence or loss between WT and Trpv4 KO groups (Supplementary Figure S2), even following the transition to serum-free Seahorse medium and exposure to mitochondrial inhibitors (Oligomycin, FCCP, AA/Rot).

However, in contrast to the initial baseline observations, calculated ATP-linked respiration showed a significant increase specifically in Trpv4 KO cells following application of GSK21 (1 µm) (p = 0.0031, Figure 7E), an effect not observed in the WT group.

Application of the TRPV4-specific inhibitor GSK21 (1 µM; a lower concentration of GSK21 was used to circumvent metabolic toxicity) induced distinct metabolic shifts as shown by extracellular flux assay, and revealed a significant increase in the non-mitochondrial respiration in both WT (p = 0.0024) and Trpv4 KO (p = 0.0461) cells (Figure 7D). Notably, acute treatment with GSK21 significantly enhanced the glycolytic reserve specifically in Trpv4 KO cells (p = 0.0002), while WT cells showed no significant change, despite the absence of TRPV4 activity (Figure 7G and Figure S3). This increased reserve glycolytic capacity due to GSK21 in Trpv4 KO cells suggests an aspecific effect of the treatment, or a compensatory shift toward glycolytic metabolism when oxidative phosphorylation is challenged.

Overall, while acute GSK21 application did not induce mitochondrial network fragmentation (Figure 6D,K) as observed with Trpv4 KO (Figure 1C and Figure 2C), it significantly modulated metabolic reserve and ATP-linked respiration in the knockout model. These results suggest that the Trpv4 KO model may undergo chronic compensatory metabolic adaptations that are not established during acute pharmacological inhibition of TRPV4.

4. Discussion

Microglia exhibit dynamism and are fundamental for maintaining the homeostasis of the central nervous system. Their diverse functions, encompassing morphology, migration, and metabolic regulation, are closely connected to the organization, morphology, and spatial distribution of their mitochondrial networks [45,46]. Our study provides compelling evidence that the polymodal thermo-mechanosensitive ion channel TRPV4 plays a critical and complex role in orchestrating these mitochondrial characteristics in primary murine microglia. This regulatory function is likely bidirectional: while TRPV4 signaling influences mitochondrial morphology, the capacity of the mitochondria to fuel localized processes is essential for maintaining their adaptive cellular architecture and dynamics. These findings expand upon the established functions of TRPV4 in regulating microglial morphology and migration, including the crucial transition from ramified surveillant to amoeboid states [8]. This suggests that mitochondrial dynamics are an important, TRPV4-dependent component of the microglial response to environmental cues.

Our quantitative analysis of the mitochondrial network demonstrated alterations in Trpv4 KO microglia, primarily indicating a shift towards enhanced mitochondrial fragmentation by enhanced fission activity. While global network parameters (e.g., branch length, number of branches) were largely preserved, Trpv4 KO cells consistently displayed a greater number of individual mitochondria, indicating an increase in mitochondrial fragmentation overall (Figure 1C) and across all three morphologies (Figure 5B). Interestingly, this mitochondrial phenotype persists despite the fact that Trpv4 KO microglial morphology remains largely indistinguishable from WT in vitro. However, the physiological relevance of these mitochondrial shifts could be highlighted by findings in acute brain slices, where TRPV4 deficiency leads to decreased process complexity and surveillance [8]. In our model, this fragmentation was characterized by a significantly lower mean surface area per mitochondrion (total surface area of the HSP60 or MitoTracker divided by the number of mitochondria) (Figure 1D and Figure 5C,E) and a higher number of branch endpoints (Figure 1G), indicating mitochondria network fragmentation (Figure 1B). These findings are consistent with an imbalance in mitochondrial network dynamics, favoring increased fission [47]. Quantification of mitochondria fusion protein Mfn1 and fission activator pDrp1 (Ser616) showed that the absence of TRPV4 activity shifts the mitochondria dynamics equilibrium towards fission. The levels of mitochondrial fusion protein Mfn1 were consistent across genotypes (Figure 2B) compared to the cellular Mfn1 levels, which decreased in the Trpv4 KO cells (Figure 3A), suggesting that fusion is unaffected. In contrast, the recruitment of fission activator pDrp1 (Figure 2C and Figure 3B) was enhanced.

Fragmented mitochondria are often associated with reduced metabolic efficiency, increased ROS, and altered calcium handling, potentially contributing to a less homeostatic or even inflammatory microglial morphology, impacting their roles in neuroinflammation and neurodegeneration [48,49]. However, our real-time extracellular flux analysis revealed that neither the constitutive loss of TRPV4 nor its acute pharmacological inhibition significantly impaired baseline mitochondrial respiration (Figure 7C). This suggests that the core OXPHOS machinery remains functional despite the structural remodeling of the network. This decoupling of morphology from metabolism is consistent with emerging evidence that Drp1-dependent mitochondrial fission is not crucial in metabolic reprogramming of microglia [50]. While basal metabolism remained stable, acute application of the TRVP4 antagonist significantly enhanced the glycolytic reserve (p = 0.0002) and ATP-linked respiration (p = 0.0031) specifically in the Trpv4 KO cells (Figure 7E,G). Unlike the WT cells, which showed more modest shifts, the KO cells appear to be metabolically primed—maintaining homeostasis under basal conditions but exhibiting an exaggerated compensatory reliance on glycolytic and ATP-linked reserves when the system is acutely challenged. These data, from a limited number of samples, suggest that TRPV4 activity may modulate metabolic flexibility by allowing the cells to adapt their energetic strategy under stress, rather than controlling the basal metabolic rate.

The observation of these metabolic shifts in BMDMs, coupled with our findings that mitochondrial dynamics proteins (Mfn1 and pDrp1) shift in identical directions in both microglia and BMDMs (Figure 3), suggests that TRPV4-mediated mitochondrial regulation is a canonical mechanism shared across myeloid populations. This shared biology justifies the use of the high-yield BMDM model to uncover the bioenergetic consequences of TRPV4 dysregulation that likely exist in microglia as well.

Beyond the gross genotype differences, the mitochondria density distribution revealed distinct, morphology-dependent patterns in mitochondrial distribution, revealing distinct patterns between genotypes (Figure 4E,F). Notably, in the lamellipodium with a trailing edge (LTE) subtype, WT cells displayed higher mitochondrial density near the nucleus, with a sharp decrease towards the dynamic lamellipodium and trailing edge cell extremities (Figure 4J). In contrast, Trpv4 KO LTE cells exhibited a more widespread mitochondrial distribution with a higher density extending further away from the nucleus towards the trailing edge (Figure 4J). A similar pattern was observed in Trpv4 KO cells with a branched morphology, which displayed increased mitochondrial density both in the perinuclear region and throughout the cellular processes compared to WT counterparts (Figure 4H). These findings are in line with previous research from Pietramale et al. [46], which showed that more mitochondria were located at the cell center and decreased towards the processes. Additionally, they found that mitochondria are often absent in the branches of surveilling microglia. While localized energetic demands for extensive cytoskeletal reorganization are highest at the leading lamellipodium, the observed WT clustering might be necessary to support perinuclear processes such as signal transduction, internal Ca²⁺ store replenishing, and rapid protein synthesis required for cell functions [18,51]. The more dispersed distribution of mitochondria in the absence of TRPV4 suggests a loss of this precise spatial regulation. This should also be checked by future studies in vivo, since the complexity and surveillance of Trpv4 KO microglia are decreased [8]. Furthermore, while our study highlights the intrinsic effects of Trpv4 KO, indirect effects on from other brain cells on microglial mitochondrial biology cannot be excluded in a global knockout model.

However, our live-imaging experiments following acute pharmacological blockade of TRPV4 (Figure 6) provide a crucial temporal contrast to these chronic observations. The focus was on the mitochondrial branches to better separate the individual mitochondria. MitoTracker Deep Red FM, a live-cell mitochondrial dye, was used to visualize healthy functional mitochondria (Figure 6A,B). This is a refinement of the HSP60 staining, which detects both healthy and dysfunctional mitochondria that lost their membrane potential, but still retain the protein due to incomplete degradation [52,53]. While the genetic KO model suggests that a lack of TRPV4 promotes mitochondria dispersion into the cell processes, our results showed that acute pharmacological inhibition of TRPV4 with GSK2193874 did not induce mitochondrial fragmentation or immediate translocation of the network within the 15-min recording window (Figure 6C,D,H). This discrepancy suggests that the widespread mitochondrial distribution in Trpv4 KO cells is likely not a direct effect of the lack of activity on mitochondrial movement, but rather a consequence of chronic adaptive remodeling. It is possible that the immediate blockade led to a non-significant increase in the mean surface area and volume per mitochondrion, as well as to an apparent increase in the number of mitochondria and branch junctions per mitochondrion (Figure 6D–G), as an immediate cellular response to loss of TRPV4 activity. Meanwhile, the Trpv4 KO cells suffer from a chronic loss of TRPV4-regulated Ca²⁺ levels (Figure S3B,C), which is not achieved during our acute live-imaging experiment due to cell stability constraints. These acute inhibition recordings suggest a shift towards a rapidly distended, more complex, and hyperfused morphology. Mitochondria typically transition to a hyperfused state in response to acute stress or bioenergetic crisis, as fusion is a pro-survival mechanism that attempts to share resources and maximize ATP output via oxidative phosphorylation (OXPHOS), making the network more efficient and distributing energy over larger distances [19,54]. Our finding that acute application of GSK2193874 significantly increases ATP-linked respiration in Trpv4 KO cells (Figure 7E) supports the presence of a metabolic stress response. This aligns with the concept known as Stress-induced Mitochondrial Hyperfusion (SiMH) [54], where the network attempts to maximize ATP output via OXPHOS efficiency during an acute bioenergetic crisis. The fact that this surge is most prominent in the KO model suggests that the chronic lack of TRPV4 pre-sensitizes the cells to this stress response. Specifically, the sudden lack of TRPV4 activity might initiate this compensatory hyperfusion, suggesting that basal TRPV4 activity is normally required to maintain the steady balance of the network, preventing it from tipping into an elongated, stress-response state.

The mechanism for this dynamic control is highlighted by the direct molecular connection between TRPV4 and mitochondrial dynamics. TRPV4 is known to directly interact with the fusion proteins Mfn1/2 [30]. Functionally, the channel serves as a critical entry point for extracellular Ca²⁺, delivering the signal to the MFN proteins, which are Ca²⁺ sensitive and concentrated at ER-mitochondria contact sites (MAMs) [55]. Our results show that mitochondrial Mfn1 levels remain stable among our tested groups (Figure 2B). However, the total cellular pool of Mfn1 is reduced in Trpv4 KO cells compared to the WT (Figure 3A), potentially due to increased protein turnover or diminished Mfn1 stability. Moreover, Drp1 is heavily regulated by Ca²⁺, primarily through post-translational phosphorylation and dephosphorylation. Calcium/calmodulin-dependent protein kinase II (CaMKII) phosphorylates Drp1 at Ser616, which promotes its translocation from the cytosol to the mitochondrial outer membrane, triggering mitochondrial fragmentation [23,24,25]. In our experiments, there was an increase in pDrp1 (Ser616) in two myeloid cell types, microglia and BMDMs (Figure 2C and Figure 3B). These data suggest that TRPV4 activity is involved in maintaining mitochondrial homeostasis by balancing the levels of fusion proteins and the recruitment of fission mediators.

Our findings show that TRPV4 deficiency leads to a more dispersed mitochondrial network. Mitochondrial motility is regulated by cytosolic calcium, most likely through adaptor proteins such as Miro1 [56,57,58,59,60], and its dysregulation caused by a loss of TRPV4 activity can impair this control. We found more mitochondria in the distal cellular compartments, a finding that can manifest as either an inability to efficiently retract mitochondria back to the perinuclear region or an overall increase in their movement. We discuss putative and non-mutually exclusive mechanisms to explain this. The first suggests that altered cytosolic calcium levels impair the ability of mitochondria to move back towards the nucleus. This might be because TRPV4 inhibition immobilizes microtubules [8], which can potentially affect post-translational modifications of tubulin, making efficient mitochondrial attachment to microtubules difficult [57,59,61,62,63]. This cascade of changes, along with impaired activity of motor proteins like dynein [64], disrupts the retrograde mitochondrial transport [65,66]. The second mechanism proposes that the lack of TRPV4 activity leads to hypermobility throughout the cell, possibly due to a calcium-dependent shift in the balance of motor protein activity favoring kinesin, and promoting a preference for mitochondrial fission over fusion [14,15,17,47] (Figure 2C and Figure 3B). Furthermore, if TRPV4 serves as a physical anchor between the plasma membrane and the sub-membrane cytoskeleton, its genetic lack of function would disrupt mitochondrial trafficking pathways more severely than a blockade of its pore. This would explain why the widespread mitochondrial distribution (Figure 4H–J) is a hallmark of the KO but not the acute inhibitor group. In either case, our data suggest that TRPV4 finely tunes the dynamic regulation of mitochondrial localization in microglia, and its dysregulation contributes to the observed fragmentation or altered distribution of the mitochondrial network.

5. Conclusions

Our work indicates TRPV4 as a key regulator of microglial mitochondrial distribution, network morphology, and adaptive dynamics. The observed mitochondrial fragmentation with enhanced pDrp1 (Ser616) fission activation (Figure 1, Figure 2, and Figure 3), altered spatial distribution (Figure 4), and lack of morphology-dependent adaptability in Trpv4 KO microglia (Figure 5) points to a fundamental role for TRPV4 in maintaining mitochondrial homeostasis. The rapid, unstable yet non-significant alterations in network parameters observed upon acute TRPV4 inhibition (Figure 6), coupled with this metabolic surge, might suggest its contribution to immediate, adaptive cellular responses. Future studies should aim to characterize the interactions between TRPV4-mediated calcium signaling, cytoskeletal dynamics, and mitochondrial fission-fusion proteins in regulating microglial mitochondrial health and function. Better understanding these TRPV4-dependent pathways is crucial for developing potential treatments for neurological disorders linked to mitochondrial dysfunction.