Fluorescence-Based Quantification of Mitochondrial Damage in Human Airway Smooth Muscle Cells

Abstract

Mitochondrial quality control is essential for maintaining cellular homeostasis by balancing the removal of damaged mitochondria (mitophagy) with the generation of new mitochondria (mitochondrial biogenesis). A key feature of mitochondrial damage is loss of mitochondrial membrane potential (ΔΨ_m), which initiates mitophagy, enabling effective mitochondrial clearance. Although an array of tools exists to assess mitochondrial damage (depolarization), many rely on acute, non-physiological depolarization or provide semiquantitative measures of mitochondrial damage, limiting their ability to resolve intact versus damaged mitochondria within heterogeneous mitochondrial networks. Therefore, in the present study we developed and validated an imaging-based assay to quantify intact mitochondria in human airway smooth muscle (hASM) cells using dual-fluorescence labeling. This approach combines a ΔΨ_m-dependent (MitoTracker Red FM) dye with a ΔΨ_m-independent label (CellLight Mitochondria-GFP). Dual-labeled mitochondria in untreated hASM cells exhibited ~10% non-overlap between the two fluorescence signals, indicating presence of damaged (depolarized) mitochondria in homeostatic conditions. Dose- and time-dependent treatment with the mitochondrial uncoupler FCCP induced loss of membrane potential, confirmed by TMRM, and resulted in a marked reduction in fluorescence overlap, volume of intact mitochondria and increased mitochondrial fragmentation. Complementary analysis using the redox-sensitive reporter pMitoTimer was performed, where a shift in fluorescence signal from green to red is indicative of increased mitochondrial oxidative stress and rate of mitochondrial turnover. Together, these findings validate the dual-labeling strategy as a quantitative method to distinguish intact from damaged mitochondria in situ and as a useful tool for studying mitochondrial quality control, potentially translatable to various cell and disease models.

Graphical Abstract

New and Noteworthy

We introduce an imaging-based approach to quantitatively distinguish intact from damaged mitochondria within heterogeneous mitochondrial networks using fluorescent labels that exhibit different sensitivities to mitochondrial membrane potential. By combining membrane potential-independent CellLight GFP label with membrane potential-dependent MitoTracker Red, this method sensitively quantifies basal and stress-induced mitochondrial damage in hASM cells. This assay provides a practical and interpretable metric of mitochondrial integrity that complements existing methods that measure mitochondrial membrane potential and oxidative stress.

Introduction

Mitochondrial quality control encompasses the coordinated cellular processes responsible for maintaining a functional and metabolically efficient mitochondrial population. Failure of mitochondrial quality control results in increased accumulation of damaged mitochondria that can have deleterious effects on cellular metabolism, increase reactive oxygen species (ROS) production, and ultimately trigger apoptosis (1, 2). In order to maintain cellular homeostasis, it is essential that identification and elimination of dysfunctional mitochondria must be balanced by replacement with newer, more efficient mitochondria (3–5).

Mitophagy, the selective degradation of mitochondria via autophagy, is a central mechanism in mitochondrial quality control (4, 6). During oxidative phosphorylation, mitochondria constitutively generate ROS and the lack of adequate ROS scavenging can subject mitochondria to oxidative stress, ultimately resulting in mitochondrial damage. As a result of damage, mitochondria undergo membrane depolarization, a key signal for initiating mitophagy (7, 8). Depolarization activates ubiquitin-dependent pathways or ubiquitin-independent (receptor-mediated) pathways, resulting in the engulfment of damaged mitochondria in autophagosomes. Subsequent delivery to lysosomes facilitates degradation and recycling of mitochondrial components, which is important for mitochondrial turnover and the maintenance of physiological conditions (6, 9, 10).

Mitochondrial depolarization can be experimentally induced by various cytotoxic mitochondrial uncouplers and damage-inducing agents to explore mitophagy. Current methods to visualize mitochondrial turnover or quantify mitochondrial depolarization (damage) in vitro include potentiometric dyes to measure changes in mitochondrial membrane potential (ΔΨ_m), fluorescence-based assays to measure mitochondrial ROS and measurement of cytochrome c or mitochondrial DNA release from damaged mitochondria (7, 11–15). Although these techniques provide valuable physiological insight, they have some limitations. First, they often rely on non-physiological, acute mitochondrial depolarization, making it difficult to model the more subtle endogenous processes that regulate mitochondrial quality control under homeostatic conditions. Second, many of these assays are semiquantitative, have limited dynamic range or provide incomplete information regarding the extent of intact versus damaged mitochondria. Consequently, there remains a need for methods that more accurately quantify mitochondrial damage by distinguishing functional from non-functional organelles within a heterogeneous mitochondrial network.

To address these challenges, we implemented a dual-fluorescence mitochondrial labeling assay using two commercially available dyes: CellLight Mitochondria-Green Fluorescent Protein (GFP) and MitoTracker^® Red Fei Mao (FM). CellLight Mitochondria-GFP, is a BacMam expression fusion construct encoding GFP fused to the leader sequence of E1α pyruvate dehydrogenase, packaged in a baculovirus, providing specific targeting to the mitochondria and labeling in a ΔΨ_m-independent manner (Figure 1A) (16–18). MitoTracker Red FM is a carbocyanine-based dye that is readily sequestered within the mitochondria in a strictly ΔΨ_m-dependent manner (19). MitoTracker Red provides robust labeling of functional (polarized) mitochondria, however its dependence on ΔΨ_m restricts its ability to label depolarized mitochondria (20–22). Previously, in untreated human airway smooth muscle (hASM) cells, we observed that mitochondrial volume measured with CellLight GFP labeling was greater than that measured with MitoTracker Red, suggesting that a subset of mitochondria had decreased ΔΨ_m and therefore failing to take up MitoTracker Red (23).

Figure 1.

(A) Principle of CellLight Labeling. CellLight^™ Mitochondria-GFP, BacMam 2.0, is a fusion construct of the Leader sequence of E1 alpha pyruvate dehydrogenase (PDH) and emGFP packaged within the insect baculovirus and targets mitochondria. During transduction, the baculovirus particles enter the cell by endocytosis, which allows the translocation of the baculoviral DNA to the nucleus where the GFP-fusion transgene is expressed resulting in PDH expressing GFP. (B) Conceptual Framework. hASM cells were first transduced with CellLight^™ Mitochondria-GFP (CellLight GFP) and subsequently, the hASM cells were treated with FCCP or left untreated. Following treatment, the CellLight GFP labeled hASM cells were loaded with MitoTracker^™ Red FM (MitoTracker Red) and imaged. We hypothesize that, in untreated (control) hASM cells, mitochondria exhibit increased overlap of CellLight GFP and MitoTracker Red signals, indicating the presence of predominantly functional mitochondria with intact mitochondrial membrane potential (ΔΨ_m). However, on inducing graded mitochondrial depolarization with FCCP, the loss of ΔΨ_m (mitochondrial damage) results in decreased uptake of MitoTracker Red thereby reducing MitoTracker Red signal while CellLight GFP signal persists. GFP: Green Fluorescent Protein; hASM: Human Airway Smooth Muscle; PDH: Pyruvate Dehydrogenase; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; ΔΨ_m: Mitochondrial Membrane Potential

In the present study, using hASM cells, we describe an imaging-based assay to quantify intact versus damaged mitochondria and illustrate its utility in monitoring mitochondrial quality control. We hypothesize that, under homeostatic conditions, mitochondria exhibit increased fluorescence overlap from both CellLight GFP and MitoTracker Red signals, reflecting a predominantly functional mitochondrial network with intact ΔΨ_m. However, upon loss of ΔΨ_m (mitochondrial damage), MitoTracker Red uptake is selectively reduced while CellLight GFP signal persists (Figure 1B). The resulting loss of fluorescence overlap can be quantified in real-time as a direct index of mitochondrial damage, providing a straightforward approach to assess mitochondrial integrity.

Materials and Methods

Ethics Statement

The study was reviewed by Mayo Clinic’s Institutional Review Board (IRB #16–009655) and considered to be of minimal risk and therefore exempt from further review. The reasons for this include, a written informed consent was obtained from all potential patients during pre-surgical evaluation (in the Pulmonary, Oncology, or Thoracic Surgery Clinics) in a non-threatening environment. Patient history was provided at the time of tissue acquisition to indicate patient sex, demographics, pulmonary disease status, pulmonary function testing, imaging, co-morbidities, and medications. Patient samples were numbered but none of the patient identifiers were stored or attached to the samples therefore preventing retrospective patient identification. From a large set of patient donors (ranging in age from 59 to 70 years), a total of six (3 female and 3 male) patients were selected for this study after the exclusion of patients with chronic lung disease, asthma, any other respiratory diseases, and any recent history of smoking (Table 1).

Table 1.

Bronchiolar Tissue Sample Donor demographics.

Patient	Sex	Age (years)	Respiratory Disease	Smoking Status
1	F	59	None	Non-Smoker
2	F	61	None	Non-Smoker
3	F	70	None	Non-Smoker
4	M	60	None	Non-Smoker
5	M	69	None	Non-Smoker
6	M	70	None	Non-Smoker

Dissociation of Cells from Bronchiolar Tissue

During surgery, bronchiolar tissue samples containing third to sixth-generation bronchi were obtained and evaluated by Clinical Pathology. From all bronchiolar tissue samples, from regions of the tissue that were assessed to be normal, the smooth muscle layer was dissected. The cells in the smooth muscle layer were dissociated by enzymatic digestion using papain and collagenase with ovomucoid/albumin separation as per the manufacturer’s instructions (Worthington Biochemical, Lakewood, NJ) and as described previously (24–26). Cells were maintained in phenol red-free DMEM/F-12 medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS) and 100 U/mL of penicillin/streptomycin, at 37°C (5% CO₂ - 95% air, pH 7.4). To ensure the use of cells within their optimal phenotypic and metabolic characteristics, cells from 1–3 passages were used. Prior to treatments, the cells were serum-deprived, by replacing the growth medium with DMEM/F-12 medium lacking FBS, for 24 h. The dissociated bronchiolar cells were phenotyped prior to experimental use (Figure 2A).

Figure 2.

(A) Experimental design. Bronchiolar tissue samples were collected during lung resection surgery from anonymized female and male patients without any history of smoking or respiratory diseases. The smooth muscle layer was dissected from the third to sixth order bronchi, and cells were dissociated. The hASM phenotype was confirmed by their immunoreactivity to α-SMA and hASM cells from six patients were serum deprived for 24 h and randomly assigned into two experimental groups: 1) FCCP-treated and 2) time-matched untreated control. (B) Confirmation of hASM phenotype. Representative maximum intensity Z-projection image of dissociated cells shows the phenotype of hASM cells based on their immunoreactivity to α-SMA expression and larger size (scale bar = 50 μm). hASM: Human Airway Smooth Muscle; α-SMA: α-Smooth Muscle Actin; DAPI: 4′,6-Diamidino-2-Phenylindole, Dihydrochloride

Immunocytochemical Labeling and Phenotyping Dissociated hASM Cells

The hASM phenotype of the dissociated cells from each bronchiolar tissue sample was confirmed by immunocytochemical analysis of α-smooth muscle actin (α-SMA) expression as previously described (23, 27–30). Briefly, dissociated cells were plated at a density of ~10,000 cells/well in a Nunc^™ Lab-Tek^™ II Chamber 8-well multichamber slide (Thermo Fisher Scientific, Rockford, IL), incubated till cell adherence was achieved and serum-deprived. Cells were fixed with 4% paraformaldehyde in 1X phosphate-buffered saline (PBS) for 10 min at room temperature and washed with 1X PBS. To prevent non-specific antibody binding, cells were incubated in a blocking buffer (10% normal donkey serum, 0.2% triton X-100 and 1X PBS) for 1 h at room temperature (RT). Cells were then incubated overnight at 4°C with anti-α-SMA antibody (1:500 dilution, rabbit polyclonal; Abcam Cat# ab5694, RRID: AB_2223021; Boston, MA) in a diluent solution (2.5% normal donkey serum, 0.25% sodium azide, 0.2% triton X-100, 1X PBS). This was followed by incubation for 1 h at RT in a species specific biotin-conjugated secondary antibody (1:400, diluted in diluent solution; Jackson ImmunoResearch Labs Cat# 711-067-003, RRID: AB_2340595; West Grove, PA), followed by Streptavidin Alexa Fluor 568 (1:200 in PBS; Thermo Fisher Scientific Cat# S-11226, RRID: AB_2315774; Carlsbad, CA) for 30 min at RT. The cells were mounted using Fluoro-Gel II mounting medium containing 4′,6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI; Electron Microscopy Sciences, Hatfield, PA). The dissociated cells were imaged to distinguish α-SMA expressing hASM cells using a Nikon ECLIPSE Ti inverted microscope system (RRID: SCR_021242) with a ×60/1.4 NA oil-immersion objective and the percentage of hASM cells was determined as a fraction of the total dissociated cells (determined from DAPI labeled nuclei; Figure 2B).

Experimental Design

FCCP (Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone; 10 mg, Sigma Aldrich St. Louis, MO) was reconstituted in Dimethyl Sulfoxide (DMSO) to a stock concentration of 10 μM, as per manufactures instructions. hASM cells isolated from each patient were serum-deprived for 24 h and split into two treatment groups: 1) FCCP-treated, 2) Time-matched untreated control (Figure 2A). Prior to experimentation, the effect of vehicle (DMSO) in inducing mitophagic responses was first examined and confirmed to be comparable to untreated control. To optimize the technique proposed to quantify mitochondrial damage, various FCCP concentrations (0.5 μM, 1.0 μM and 2.5 μM) and exposure durations (0.5 h, 1.0 h, 1.5 h, 3.0 h and 6 h) were used in the study. The maximum duration and concentration of exposure to FCCP was restricted to avoid irreversible mitochondrial damage and/or cell death.

Measurement of Mitochondrial Membrane Potential

hASM cells were plated at a density of ~15,000 cells/well into 8 well Ibidi μ-slide plates (Ibidi GmbH, Gewerbehof Gräfelfing, Germany), incubated to allow for cell adherence and subsequently serum deprived. After treatment, hASM cells were loaded with 50 nM tetramethylrhodamine methyl ester (TMRM; Invitrogen, Carlsbad, CA; excitation/emission wavelength: 548/574 nm) in serum-free media, pre-warmed to 37°C, for 30 min followed by extensive washing with 1X Hanks’ Balanced Salt Solution (HBSS; Invitrogen, Carlsbad, CA). A low concentration of TMRM was used to ensure that the dye itself does not influence the change in membrane potential. TMRM labeled mitochondria in hASM cells were imaged using the Nikon Eclipse A1 laser scanning confocal microscope (RRID: SCR_020317) with a ×60/1.4 NA oil-immersion objective at 12-bit resolution into a 1,024×1,024-pixel array. The dynamic range for imaging was set by first scanning a region containing no TMRM fluorescence signal and then a second region of interest containing maximum TMRM fluorescence. Time lapse 2D (xyt) imaging was performed, and images were acquired every 10 s for 6 min. At the 3 min mark, FCCP (1.25 μM) was added to induce maximum depolarization of the mitochondrial membrane potential. The images acquired were analyzed using NIS-Elements software (version 5.20.02; RRID: SCR_014329; Nikon Instruments Inc., Melville, NY). The change in mitochondrial membrane potential was reflected by the difference in TMRM fluorescence intensity between maximum depolarization (t=6 min) and baseline (t=0 min) (31, 32). The change in TMRM fluorescence across cells was represented as percent change in mean fluorescence intensity (MFI). Multiple hASM cells (identified by size and elongated shape) were visualized within a single microscopic field. Due to the large size of hASM cells, the borders of some cells overlapped, and only those hASM cells whose borders were not overlapping were included. Typically, this selection process results in the analysis of 2 to 3 hASM cells per field. Based on an a priori power analysis of variance in TMRM measurements in untreated hASM cells, n=10 individual hASM cells per patient (6 patients) were included in the analysis.

Labeling of Mitochondria in hASM Cells and Confocal Imaging

hASM cells were plated at a density of ~15,000 cells/well into 8-well Ibidi μ-slide plates (Ibidi GmbH, Gewerbehof Gräfelfing, Germany) and incubated to allow cell adherence. Adhered hASM cells were transduced with CellLight^™ Mitochondria-GFP, Bac-Mam 2.0 (Invitrogen, Carlsbad, CA, USA) to label mitochondria, as per the manufacturer’s instructions. The hASM cells were incubated in CellLight GFP containing growth media for 12 h at 37°C. Subsequently, the hASM cells were washed with 1X HBSS and serum-deprived for 24 h. Although the efficiency of transduction was relatively low (~60%), GFP protein could be observed in mitochondria. Following treatment, CellLight GFP transfected hASM cells were loaded with 200 nM MitoTracker^™ Red FM (Invitrogen, Carlsbad, CA; excitation/emission wavelength: 581/644 nm), in serum-free media, pre-warmed to 37°C, for 15 min followed by extensive washing with 1X HBSS. Labelled mitochondria in hASM cells were imaged using the same confocal microscope system used for TMRM measurements as described above. The dynamic range for imaging was set by first scanning a region containing no fluorescence signal and then a second region containing maximum fluorescence signal. A series of 0.5 μm optical slices were acquired for each hASM cell to obtain a Z-stack (xyz) using the NIS-Elements software (version 5.20.02; RRID: SCR_014329; Nikon Instruments Inc., Melville, NY). Image acquisition was performed using resonant scanning system on the Nikon NIS confocal platform to minimize photobleaching and phototoxicity. Resonant scanning operates at high mirror frequencies with ultra-short pixel dwell times (tens of nanoseconds to ~0.1 μs), resulting in rapid volumetric acquisition (30–45 s per z-stack). This configuration substantially reduces exposure-induced artifacts while preserving signal-to-noise ratio. Cells were imaged only at the defined experimental time points and were not subjected to continuous laser exposure across the 6 h treatment window. We confirmed that no progressive loss of fluorescence intensity was observed within individual acquisition windows under the selected imaging parameters, thereby minimizing photobleaching and ensuring signal stability during acquisition. Multiple hASM cells (identified by size and elongated shape) were visualized within a single microscopic field. Similar hASM cell selection criteria was used as discussed above.

Image Deconvolution and Measurement of Change in Fluorescent Intensity

The 3D images of hASM cells with CellLight GFP and MitoTracker Red labeled mitochondria were deconvolved using the automatic deconvolution algorithm on NIS-Elements software (Modified Richardson Lucy method; Point Scan Confocal modality; Nikon Instruments Inc.; version 5.20.02; RRID: SCR_014329) (29, 33–35). Deconvolution enhances the signal-to-noise ratio in the images, thereby improving contrast and edge detection. The voxel dimensions of each deconvolved optical slice were 0.207 × 0.207 × 0.5 μm. After deconvolution, the boundaries of each hASM cell were delineated and region of interest (ROI) was defined in ImageJ-Fiji software (https://imagej.nih.gov/ij/; version 1.53t, RRID: SCR_002285; NIH, Public Domain). The fluorescence signal for each label (CellLight GFP or MitoTracker Red) was identified by thresholding and the number of voxels containing fluorescence was determined. For image analysis, MFI for each label was measured and the MFI for MitoTracker Red was divided by the MFI for CellLight GFP to generate a ratiometric analysis of change in fluorescent intensity between the two labels.

Quantification of Fluorescence Overlap

The 3D images of hASM cells containing CellLight GFP and MitoTracker Red labeled mitochondria were used to quantify colocalization of MitoTracker Red with CellLight GFP using EzColocalization plugin (https://sites.imagej.net/EzColocalization/plugins/) available on ImageJ-Fiji software, as described previously (29, 36–38). The 3D images of transfected hASM were deconvolved, delineated and processed using the NIS-Elements software (version 5.20.02; RRID: SCR_014329; Nikon Instruments Inc., Melville, NY) and ImageJ-Fiji software (https://imagej.nih.gov/ij/; version 1.53t, RRID: SCR_002285; NIH, Public Domain) as described above. In the EzColocalization plugin, the channels (“Reporters”) were defined and the images from each channel were aligned and thresholded. The range of pixels to be identified was defined and watershed segmentation was performed to filter out any unwanted signals. Localization of signals were then visualized and the metric matrices option was used to identify the percentage of pixels that fall above the threshold of both channels. The colocalization metrics were defined and the Manders’ Coefficients M1 (MitoTracker Red colocalization with CellLight GFP) and M2 (CellLight GFP colocalization with MitoTracker Red) were analyzed as a measure of colocalization. Based on an a priori power analysis of variance in Manders’ Coefficient M1 measurements in untreated hASM cells, n=10 individual hASM cells per patient (6 patients) were included in the analysis.

Measurement of Mitochondrial Volume Density and Quantify Intact Mitochondrial Volume

Mitochondrial volume density was measured using the mitochondrial analyzer plugin (https://github.com/AhsenChaudhry/Mitochondria-Analyzer) available on ImageJ (32, 39, 40). This plugin incorporates adaptive thresholding and object-based segmentation optimized for densely distributed mitochondrial networks and enables robust volumetric quantification in images with heterogeneous contrast and spatial complexity. Although alternative approaches such as MiNA (Mitochondrial Network Analysis) provide valuable insight into network topology assessment (e.g., branch length and junction density), the present study is focused on 3D volumetric assessment of dual-signal overlap. Therefore, for this application, the adaptive threshold-based analysis of the mitochondrial analyzer plugin provides greater robustness across heterogeneous conditions and was therefore selected.

The deconvolved Z-stacks were processed for background correction and ridge filter detection. The labeled mitochondria within hASM cells were identified by thresholding to create a binary image and then skeletonized for morphometric analysis. Using the thresholded binary image, total mitochondrial volume from each fluorescent channel was measured by the mitochondrial analyzer plugin where the number of voxels containing fluorescently labeled mitochondria (CellLight GFP or MitoTracker Red) within a cell was determined. Mitochondrial volume density was calculated as the ratio of mitochondrial volume within the cell to the total volume of the defined ROI (mean cross-sectional area of the cell from all optical slices × total number of slices × 0.5 μm) and represented as a percentage of the volume of the cell (23, 41, 42). The volume of intact mitochondria was quantified by thresholding to the overlap between the two fluorescent labels and quantifying mitochondrial volume as described above. Based on an a priori power analysis of variance in mitochondrial volume density measurements in untreated hASM cells, n=10 individual hASM cells per patient (6 patients) were included in the analysis.

Mitochondrial Complexity Index (MCI)

From the 3D images of CellLight GFP transduced and MitoTracker Red labeled hASM cells, mitochondrial morphological parameters, including mitochondrial volume and mitochondrial surface area, were measured using the mitochondrial analyzer plugin as described above. The extent of mitochondrial fragmentation was assessed by calculating the mitochondrial complexity index (MCI) using the following equation (41–44):

$$MCI=\frac{S{A}^3}{16{\pi }^2{V}^2}$$

Where SA is the total mitochondrial surface area measured within the cell and V is the total volume occupied by the mitochondria within the cell. More fragmented mitochondria exhibit lower surface area, higher volume and thus, lower MCI. Based on an a priori power analysis of variance in MCI in untreated hASM cells, n=10 individual hASM cells per patient (6 patients) were included in the analysis.

Transfection of pMitoTimer Plasmid

pMitoTimer was a gift from Zhen Yan (Addgene plasmid # 52659; http://n2t.net/addgene:52659; RRID: Addgene_52659) (45). The plasmid DNA construct was sequenced, and transfection efficiency was validated prior to experimentation. hASM cells were plated at a density of ~15,000 cells/well into 8-well Ibidi μ-slide plates (Ibidi GmbH, Gewerbehof Gräfelfing, Germany) and incubated to allow cell adherence. Adhered hASM cells were transfected with the pMitoTimer plasmid using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). For the transfection, 100 ng of plasmid DNA was mixed with Lipofectamine 3000 and P3000 in Opti-MEM reduced serum media (Invitrogen, Carlsbad, CA, USA), prepared as per manufacturer’s instructions. Following incubation for 5 min at RT, the transfection mixture was added onto the cells and incubated at 37°C for 24 h. Post-transfection, the hASM cells were serum-deprived for 24 h. After treatment, transfected hASM cells were imaged in 3D using the same confocal microscope system used for TMRM measurements as described above. The excitation/emission wavelengths were set for the green florescent protein (GFP, excitation/emission wavelength: 488/518 nm) and red fluorescent protein (RFP, excitation/emission wavelength: 543/572 nm) expressed by pMitoTimer. The efficiency of transfection was ~70% and both GFP and RFP signals were observed in mitochondria. The 3D images of transfected hASM were deconvolved, delineated and processed using the NIS-Elements software (version 5.20.02; RRID: SCR_014329; Nikon Instruments Inc., Melville, NY) and ImageJ-Fiji software (https://imagej.nih.gov/ij/; version 1.53t, RRID: SCR_002285; NIH, Public Domain) as described above. Within delineated hASM cells, MFI for each channel was measured and the MFI for RFP was divided by the MFI for GFP to generate a ratiometric analysis of protein oxidation and mitochondrial turnover. Similar hASM cell selection criteria was used as discussed above. Based on an a priori power analysis of variance in MFI measurements in untreated hASM cells, n=10 individual hASM cells per patient (6 patients) were included in the analysis.

Statistical Analysis

Sex is an important biological variable, therefore hASM cells for all experiments were dissociated from bronchial tissue samples obtained from both female and male patients. However, the study was not powered to detect sex differences in the major outcome variables, and sex was considered only as a random variable in the statistical analysis. All experiments were conducted using the same six patients to maintain consistency and reduce variability. Importantly, hASM cells from the same patient and passage (1–3 passages) were randomly distributed into treatment groups, and all groups were processed in the same experimental conditions to exclude any batch effects. Therefore, hASM cells from each patient served as their own control. The basal outcome measures varied across patient samples; however, patient-to-patient comparison was beyond the scope of the present study. Sample size (n) represents the number of bronchial samples (patients) and the n used for each experiment is provided in the figure legends. The minimum number of patients and cells per patient used was determined by a power analysis of primary outcome measures. The expected effect size was calculated with an a priori biologically relevant difference of 20% and equal variance, with sample size estimated using d=1.4, α=0.05, and β=0.8 (10 cells/patient; 6 patients - 3 females and 3 males). The experiments were designed such that the comparisons were made within a patient. Statistical analysis was performed using GraphPad Prism 10 (version: 10.6.1, RRID:SCR_000306; San Diego, CA). Normal distribution of data was confirmed using the Shapiro-Wilk test. For all experiments, all analyses were performed in a blinded manner and mixed-effects analysis with Sidak’s multiple comparison test was used. Statistical significance was concluded if P<0.05, indicated by * in the figures, and the absolute P values are stated.

Results

Phenotype of dissociated hASM cells

The cells dissociated from the dissected smooth muscle layer of third to sixth-generation bronchi were assessed for their immunoreactive to α-SMA (Figure 2A). Consistent with our previous studies (23, 42, 43), a majority of the dissociated cells (90%−95%) displayed immunoreactivity to α-SMA, indicating that they were hASM cells (Figure 2B). These hASM cells were also larger and displayed a characteristic elongated shape that distinguished them from other cells. In all experiments involving imaging-based analysis, the distinct morphological features of the cells were used to identify hASM cells.

Characterization of CellLight GFP and MitoTracker Red labeling in hASM cells

To validate that the pattern of CellLight GFP labeling was restricted to mitochondria, untreated (control) hASM cells were first labeled with CellLight GFP and following 24 h serum-deprivation the same cells were labeled with MitoTracker Red. The presence of off-target effects and time-dependent changes in overlap of the two fluorescent labels was examined (Figure 3A). MFI was quantified for both CellLight GFP and MitoTracker Red signals in control hASM cells across time up to 6 h post-MitoTracker Red labeling. Individual assessment of CellLight GFP signal showed no significant decrease in with time, with a 7.2% decrease compared to 0 h (Figure 3B). However, MitoTracker Red signal showed a significant decrease at 6 h (P=0.043), with a 15.4% decrease compared to 0 h. Ratiometric analysis of MitoTracker Red fluorescence relative to CellLight GFP showed a decrease at 6 h (P=0.047), compared to 0 h (Figure 3C). The extent of colocalization of MitoTracker Red with CellLight GFP was quantified using the Manders’ colocalization coefficient (Figure 3D). Across time, no significant change in the overlap coefficient was observed indicating maximum overlap between the two labels.

Figure 3. CellLight GFP fluorescence is not significantly disrupted with time.

(A) Representative maximum intensity Z projection images show untreated hASM cells transduced with CellLight GFP and labeled with MitoTracker Red to visualize mitochondria. Overlay of the two fluorescence labels confirm that the pattern of CellLight GFP labeling was restricted to mitochondria, similar to MitoTracker Red (Scale bar = 50 μm). (B) Using 3-D reconstructed Z-stack images of labeled hASM cells, the MFI of CellLight GFP and MitoTracker Red were quantified across different time points using ImageJ. MFI from each label was normalized to that observed at 0 h with CellLight GFP. (C) Ratiometric analysis of MitoTracker Red to CellLight GFP fluorescence intensity significantly decreased at 6 h (*P=0.047) compared to 0 h. (D) Colocalization of MitoTracker Red to CellLight GFP was measured by Manders’ overlap coefficient using ImageJ. The overlap coefficient did not significantly change with time. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red and orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per time point (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between 0 h and other time points. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; 3-D: 3-Dimensional; MFI: Mean Fluorescence Intensity; CL-GFP: CellLight GFP; MT-Red: MitoTracker Red

Intact mitochondrial volume decreases with time in untreated hASM cells

Mitochondrial volume density was quantified for both CellLight GFP and MitoTracker Red labeled mitochondria in untreated (control) hASM cells across time up to 6 h post-MitoTracker Red labeling. Individual assessment of CellLight GFP showed no significant decrease in mitochondrial volume density across time, with a maximum decrease of 6.8% observed at 6 h compared to 0 h (Figure 4A). Whereas, with MitoTracker Red, mitochondrial volume density significantly decreased by 15.7% at 6 h (P=0.031), compared to 0 h. Intact (polarized) mitochondrial volume was quantified by thresholding to the overlap between CellLight GFP and MitoTracker Red. Intact mitochondrial volume decreased with time, with a significant decrease of 8.5% observed at 6 h (P=0.036) compared to 0 h (Figure 4B). Using the same images, MCI was measured as a 3D metric of mitochondrial morphological complexity, which factors both mitochondrial surface area and total mitochondrial volume. Individual assessments of both fluorescent labels showed that MCI decreased across time (Figure 4C). However, MCI for MitoTracker Red was marginally higher across time compared to that observed with CellLight GFP, indicating that intact (polarized) mitochondrial morphology is more complex, and therefore, more filamentous. This is further reflected by the MCI of intact (polarized) mitochondria, quantified by thresholding to the overlap between CellLight GFP and MitoTracker Red, versus the MCI of damaged (depolarized) mitochondria, quantified by thresholding to the non-overlap between CellLight GFP and MitoTracker Red (Figure 4D).

Figure 4. Volume of intact mitochondria decreases marginally with time.

(A) Using the 3-D reconstructed Z-stack images of labeled hASM cells, mitochondrial volume density was measured in ImageJ. Mitochondrial volume density did not significantly change with time with CellLight GFP compared to 0 h, whereas with MitoTracker Red mitochondrial volume density significantly decreased at 6 h (*P=0.029) compared to 0 h. (B) Intact mitochondrial volume was measured by thresholding the overlap between CellLight GFP and MitoTracker Red, followed by measurement of mitochondrial volume and normalized to 0 h. Intact mitochondrial volume significantly decreased at 6 h (*P=0.036) compared to 0 h. (C) MCI was measured for the individual labels using ImageJ. MCI decreased with time with both labels, with MCI in MitoTracker Red labeled hASM cells being higher when compared to CellLight GFP. (D) MCI of intact mitochondria was measured by thresholding to the overlap between CellLight GFP and MitoTracker Red and MCI of damaged mitochondria was measured by thresholding to the regions of non-overlap between CellLight GFP and MitoTracker Red. MCI of intact mitochondria was significantly higher compared to that in damaged mitochondria. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red, orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red (intact mitochondria) and grey symbols represent measurements from regions with non-overlap between CellLight GFP and MitoTracker Red (damaged mitochondria). Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per time point (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between 0 h and other time points. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; 3-D: 3-Dimensional; MCI: Mitochondrial Complexity Index

pMitoTimer fluorescence shifts with time in untreated hASM cells

To assess mitochondrial turnover, hASM cells were transfected with pMitoTimer plasmid, a DsRed1-E5 reporter expressing plasmid fused to the mitochondrial targeting sequence of human cytochrome c oxidase subunit VIII (COX8) gene (Figure 5A). pMitoTimer encodes a newly synthesized mitochondrion with GFP, which shifts irreversibly to RFP when oxidized (Figure 5B). Changes in MFI of GFP and RFP was quantified in untreated (control) hASM cells across time up to 6 h post-transfection (Figure 5C). With time, GFP signal decreased marginally with a concomitant increase in RFP signal (Figure 5D). Over time, MFI of GFP decreased from 2% at 0.5 h to 7% at 6 h, whereas MFI of RFP increased from 1% at 0.5 h to 4% at 6 h. Ratiometric analysis of RFP relative to GFP showed a marginal shift of mitochondrial fluorescence toward RFP, with 1% shift observed at 0.5 h compared to 9% shift observed at 6 h (Figure 5E).

Figure 5. MitoTimer fluorescence spectrum shifts with time in untreated hASM cells.

(A) Diagrammatic representation shows plasmid DNA containing the pMitoTimer construct. (B) When transfected, pMitoTimer is transiently expressed as GFP in newly synthesized mitochondria, which shifts to RFP when mitochondria are subjected to oxidative stress. (C) Representative maximum intensity Z-projection images show pMitoTimer expressing hASM cells (scale bar = 50 μm). (D) Using 3-D reconstructed Z-stack images of the pMitoTimer expressing hASM cells, MFI of the individual fluorescent channels were quantified by ImageJ at different time points and normalized to their respective MFI at 0 h. (E) pMitoTimer RFP to GFP ratio was quantified to assess mitochondrial damage as a result of mitochondrial turnover, which increased with time. Green symbols represent measurements from GFP, red symbols represent measurements from RFP and orange symbols represent measurements from the overlap between GFP and RFP. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per time point (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between GFP and RFP as well as between 0 h and other time points. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; RFP: Red Fluorescent Protein; 3-D: 3-Dimensional; MFI: Mean Fluorescence Intensity

Dose-dependent effects of FCCP on CellLight GFP and MitoTracker Red labeling

To determine the use of CellLight GFP and MitoTracker Red to determine mitochondrial damage, hASM cells were treated with different concentrations of FCCP for 6 h (Figure 6A). Change in mitochondria membrane potential was first assessed in hASM cells treated with FCCP treatment using a well-established membrane potential indicator TMRM (Figure 6B). FCCP concentration as low as 0.5 μM resulted in the significant loss of TMRM fluorescence signal from the mitochondria compared to untreated hASM cells (P=0.0008). The decrease in TMRM fluorescence was more significant with increasing concentrations of FCCP, with 1 μM FCCP showing a 44% loss in MFI (P=0.0004). Having confirmed the loss of membrane potential with TMRM, in separate experiments, CellLight GFP transfected hASM cells were treated with different concentrations of FCCP and labeled with MitoTracker Red post-treatment (Figure 1B). MFI was quantified for both CellLight GFP and MitoTracker Red signals across the different FCCP concentration (Figure 6C). Individual assessments of the fluorescent labels showed significant decrease in fluorescence across the different FCCP concentrations compared to untreated hASM cells. However, MitoTracker Red signal decreased significantly more compared to CellLight GFP across all FCCP concentrations (0.5 μM, P=0.012; 1.0 μM, P=0.0063; 2.5 μM, P=0.0005). Lower concentration of FCCP (0.5 μM) yielded 50% disruption of MitoTracker Red labeling (P=0.007), with the largest disruption (~95%) observed at 2.5 μM FCCP when compared to untreated hASM cells (P<0.0001; Figure 6D). Ratiometric analysis of MitoTracker Red fluorescence relative to CellLight GFP showed a significant decrease across all FCCP concentrations compared to untreated hASM cells (Figure 6E). The extent of colocalization of MitoTracker Red with CellLight GFP after FCCP treatment was quantified using the Manders’ colocalization coefficient (Figure 6F). With increasing concentrations of FCCP the overlap coefficient significantly decreased and the largest decrease of ~96% was observed at 2.5 μM FCCP (P<0.0001).

Figure 6. MitoTracker Red fluorescence is disrupted with increasing FCCP concentrations.

(A) Impact of increasing mitochondrial depolarization on CellLight GFP and MitoTracker Red labeling was assessed by treating hASM cells with different concentrations of FCCP (0.5, 1, 2.5 and 5 μM) for 6 h. (B) Following FCCP treatment, changes in ΔΨ_m was assessed using time-lapse images of TMRM labeled hASM cells. The difference in TMRM MFI between complete depolarization and at the start of imaging (baseline) was assessed and normalized to that at 2.5 μM FCCP. Red symbols represent TMRM MFI. (C) Representative maximum intensity Z projection images show hASM cells transduced with CellLight GFP and labeled with MitoTracker Red to visualize mitochondria (Scale bar = 50 μm). (D) Using 3-D reconstructed Z-stack images of labeled hASM cells, MFI of the individual labels were quantified across different FCCP concentrations using ImageJ and normalized to that observed in untreated (control, 0 μM) hASM cells. (E) Ratiometric quantification of MitoTracker Red to CellLight GFP fluorescence intensities significantly decreased across all FCCP concentrations, compared to control. (F) Colocalization of MitoTracker Red to CellLight GFP was measured by Manders’ overlap coefficient using ImageJ, which decreased significantly with increasing concentrations of FCCP. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red and orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per treatment group (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between FCCP treated and control (0 μM) hASM cells. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; 3-D: 3-Dimensional; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; ΔΨ_m: Mitochondrial Membrane Potential; TMRM: Tetramethylrhodamine, Methyl Ester; MFI: Mean Fluorescence Intensity; CL-GFP: CellLight GFP; MT-Red: MitoTracker Red

FCCP promotes dose-dependent decrease in intact mitochondrial volume

Mitochondrial volume density was quantified for both CellLight GFP and MitoTracker Red labeled mitochondria across the different FCCP concentration (Figure 7A). Individual assessments of the fluorescent labels showed significant increase in mitochondrial volume density across all FCCP concentrations, with a peak observed at 0.5 μM FCCP concentration compared to untreated hASM cells. Additionally, across all FCCP concentrations the increase was more significant with CellLight GFP compared to MitoTracker Red. However, intact mitochondrial volume decreased significantly across all FCCP concentrations compared to untreated hASM cells (Figure 7B). The hASM cells treated with 1 μM FCCP contained 47% intact mitochondrial volume (P=0.0064) compared to 68% (P=0.0018) and 12% (P<0.0001) observed at 0.5 μM and 2.5 μM, respectively. MCI quantified for the individual fluorescent probes significantly decreased across all FCCP concentrations (P<0.0001; Figure 7C). However, MCI for MitoTracker Red was marginally higher compared to that observed with CellLight GFP. MCI of intact (polarized) mitochondria decreased significantly, with significance observed starting at 0.5 μM FCCP when compared to untreated hASM cells (P<0.0001; Figure 7D). Furthermore, the gap in MCI values between intact and damaged mitochondria decreased with increasing concentration of FCCP.

Figure 7. Intact mitochondrial volume decreases with increasing FCCP concentrations.

(A) Using the 3-D reconstructed Z-stack images of labeled hASM cells, mitochondrial volume density was measured across the different FCCP concentrations using ImageJ. (B) Intact mitochondrial volume was measured by thresholding to the overlap between CellLight GFP and MitoTracker Red, followed by measurement of mitochondrial volume and normalized to untreated (control, 0 μM). Intact mitochondrial volume significantly decreased across all FCCP concentrations when compared to control. (C) MCI was measured for the individual labels using ImageJ. MCI decreased with increasing FCCP concentration with both labels compared to control, with MCI in MitoTracker Red labeled hASM cells being higher when compared to CellLight GFP. (D) MCI of intact mitochondria was measured by thresholding the overlap between CellLight GFP and MitoTracker Red and MCI of damaged mitochondria was measured by thresholding the regions of non-overlap between CellLight GFP and MitoTracker Red. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red, orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red (intact mitochondria) and grey symbols represent measurements from regions with non-overlap between CellLight GFP and MitoTracker Red (damaged mitochondria). Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per treatment group (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between FCCP treated and control (0 μM) hASM cells. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; 3-D: 3-Dimensional; MCI: Mitochondrial Complexity Index

FCCP mediates dose-dependent change in pMitoTimer fluorescence

Changes in mitochondrial turnover as a result of oxidative stress was determined by transfection of hASM cells with pMitoTimer and quantifying the change in GFP and RFP signals in response to increasing concentrations of FCCP for 6 h (Figure 8A). Treatment with increasing concentrations of FCCP significantly decreased GFP with a concomitant increase in RFP signal, with significance observed starting at 0.5 μM FCCP when compared to untreated hASM cells (P=0.027; Figure 8B). Ratiometric analysis of RFP relative to GFP showed a significant shift in fluorescence toward RFP, with a 54% shift at 1 μM FCCP (P=0.00046) and a 68% shift at 2.5 μM FCCP (P<0.0001) when compared to untreated hASM cells (Figure 8C).

Figure 8. MitoTimer fluorescence spectrum shifts with increasing FCCP concentrations.

(A) Representative maximum intensity Z-projection images show pMitoTimer transfected hASM cell treated with different concentrations of FCCP for 6 h (scale bar = 50 μm). (B) Using 3-D reconstructed Z-stack images of pMitoTimer expressing hASM cells, MFI of individual fluorescent channels were quantified following FCCP treatment using ImageJ and normalized to their respective untreated (control, 0 μM). (C) Ratiometric quantification of RFP to GFP fluorescence intensities significantly increases with increasing FCCP concentration. Green symbols represent measurements from GFP, red symbols represent measurements from RFP and orange symbols represent measurements from the overlap between GFP and RFP. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per treatment group (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between GFP and RFP as well as between FCCP treated and control (0 μM) hASM cells. hASM: Human Airway Smooth Muscle; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; GFP: Green Fluorescent Protein; RFP: Red Fluorescent Protein; 3-D: 3-Dimensional; MFI: Mean Fluorescence Intensity

Time-dependent effects of FCCP on CellLight GFP and MitoTracker Red labeling

To determine time-dependent changes in fluorescence signal of CellLight GFP and MitoTracker Red with FCCP, hASM cells were treated with 1 μM FCCP from 0 h to 6 h (Figure 9A). In FCCP-treated hASM cells, change in mitochondrial membrane potential with time was assessed using TMRM (Figure 9B). FCCP resulted in a significant loss of TMRM fluorescence as early as 0.5 h post-treatment (P=0.022) compared to 0 h and decreased more significantly with time. Following this, CellLight transfected hASM cells were treated with FCCP and labeled with MitoTracker Red post-treatment (Figure 9C). Individual assessments of the fluorescent labels showed significant decrease in fluorescence signal of both probes across time compared to 0 h (Figure 9D). The disruption of CellLight GFP labeling ranged from 10% at 0.5 h (P=0.003) to 40% at 6 h (P=0.001), whereas the disruption of MitoTracker Red labeling ranged from 40% at 0.5 h (P=0.0013) to 93% at 6 h (P=0.0007). Ratiometric analysis of MitoTracker Red fluorescence relative to CellLight GFP showed a significant decrease starting at 1 h (P=0.0073) up to 6 h (P<0.0001), compared to 0 h (Figure 9E). The extent of colocalization of MitoTracker Red with CellLight GFP after FCCP treatment was quantified using the Manders’ colocalization coefficient (Figure 9F). The overlap coefficient decreased significantly with time post-FCCP treatment, with the largest decrease of ~85% was observed after 6 h FCCP treatment compared to 0 h (P<0.0001).

Figure 9. FCCP increasingly disrupts MitoTracker-Red fluorescence with time.

(A) Impact of time-dependent mitochondrial depolarization on CellLight GFP and MitoTracker Red labeling was assessed by treating hASM cells with 1 μM FCCP for 0.5, 1, 1.5, 3 and 6 h. (B) Following FCCP treatment, changes in ΔΨ_m was assessed using time-lapse images of TMRM labeled hASM cells. The difference in TMRM MFI between complete depolarization and at the start of imaging (baseline) was assessed and normalized to that at 6 h post-FCCP treatment. Red symbols represent TMRM MFI. (C) Representative maximum intensity Z projection images show hASM cells transduced with CellLight GFP and labeled with MitoTracker Red to visualize mitochondria (Scale bar = 50 μm). (D) Using 3-D reconstructed Z-stack images of labeled hASM cells, MFI of the individual labels were quantified across time post-FCCP treatment using ImageJ and normalized to that observed in untreated (control, 0 h) hASM cells. (E) Ratiometric quantification of MitoTracker Red to CellLight GFP fluorescence intensities significantly decreased across time, compared to 0 h. (F) Colocalization of MitoTracker Red to CellLight GFP, measured by Manders’ overlap coefficient using ImageJ, decreased significantly with time after FCCP treatment compared to 0 h. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red and orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per time point (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between FCCP treated and control (0 h) hASM cells. hASM: Human Airway Smooth Muscle; GFP: Green Fluorescent Protein; 3-D: 3-Dimensional; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; ΔΨ_m: Mitochondrial Membrane Potential; TMRM: Tetramethylrhodamine, Methyl Ester; MFI: Mean Fluorescence Intensity; CL-GFP: CellLight GFP; MT-Red: MitoTracker Red

FCCP decreases intact mitochondrial volume with time

Mitochondrial volume density was quantified for both CellLight GFP and MitoTracker Red signals post-FCCP treatment. Individual assessments of the fluorescent labels showed significant increase in mitochondrial volume density with CellLight labeling as early as 1.5 h post-FCCP treatment, compared to 0 h (P=0.0027; Figure 10A). Additionally, the increase was more significant with CellLight GFP labeling compared to MitoTracker Red starting at 1.5 h post-treatment (1.5 h, P=0.035; 3 h, P=0.018; 6 h, P=0.0075). Intact mitochondrial volume was significantly decreased over time compared to 0 h, with 94% intact mitochondrial volume observed at 0.5 h and 50% intact mitochondrial volume observed at 0.5 h (P=0.0057; Figure 10B). MCI quantified for the individual fluorescent labels significantly decreased with time (Figure 10C). However, MCI for MitoTracker Red was marginally higher compared to that observed with CellLight GFP from 0.5 h to 1.5 h. MCI of intact (polarized) mitochondria decreased significantly, with significance observed starting as early as 0.5 h post-treatment when compared to 0 h (P<0.0037; Figure 10D). Furthermore, the gap in MCI values between intact and damaged mitochondria decreased with time.

Figure 10. FCCP increases mitochondrial volume density with time.

(A) Using the 3-D reconstructed Z-stack images of each label across different time points, mitochondrial volume density was measured in ImageJ. (B) Intact mitochondrial volume was measured by thresholding to the overlap between CellLight GFP and MitoTracker Red, followed by measurement of mitochondrial volume in ImageJ and normalization to that at 0 h. Intact mitochondrial volume significantly decreased across time, compared to 0 h. (C) MCI was measured for the individual labels using ImageJ. MCI decreased with time post-FCCP treatment with both labels compared to control, with MCI in MitoTracker Red labeled hASM cells being higher when compared to CellLight GFP. (D) MCI of intact mitochondria was measured by thresholding to the overlap between CellLight GFP and MitoTracker Red and MCI of damaged mitochondria was measured by thresholding the regions of non-overlap between CellLight GFP and MitoTracker Red. Green symbols represent measurements from CellLight GFP, red symbols represent measurements from MitoTracker Red, orange symbols represent measurements from the overlap between CellLight GFP and MitoTracker Red (intact mitochondria) and grey symbols represent measurements from regions with non-overlap between CellLight GFP and MitoTracker Red (damaged mitochondria). Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per time point (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between CellLight GFP and MitoTracker Red as well as between FCCP treated and control (0 h) hASM cells. hASM: Human Airway Smooth Muscle; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; GFP: Green Fluorescent Protein; RFP: Red Fluorescent Protein; 3-D: 3-Dimensional; MFI: Mean Fluorescence Intensity

FCCP mediates time-dependent change in pMitoTimer fluorescence

FCCP mediated mitochondrial damage as a result of oxidative stress was determined by transfection of hASM cells with pMitoTimer and quantifying the time-dependent changes in GFP and RFP signals in response to 1 μM FCCP (Figure 11A). With time, FCCP significantly decreased GFP signal and increased RFP signal (Figure 11B). Across time, the GFP signal decreased from 70% at 0.5 h to 13% at 6 h (P=0.0045), whereas the RFP signal increased from 8% at 0.5 h to 87% at 6 h (P=0.0010). Ratiometric analysis of RFP relative to GFP showed a significant shift in fluorescence toward RFP, with significance observed as early as 1 h post-treatment (P=0.018; Figure 11C). Compared to 0 h, a 54% shift towards RFP was observed at 6 h post-treatment.

Figure 11. FCCP shifts MitoTimer fluorescence spectrum with time.

(A) Representative maximum intensity Z-projection images show pMitoTimer transfected hASM cells exposed to 1 μM FCCP across different time points (scale bar = 50 μm). (B) Using 3-D reconstructed Z-stack images of the pMitoTimer expressing hASM cells, MFI of individual fluorescent channels were quantified across different time points post-FCCP treatment using ImageJ and normalized to their respective untreated (control, 0 h). (C) Ratiometric quantification of RFP to GFP fluorescence intensities significantly increased with time post-FCCP treatment. Green symbols represent measurements from GFP, red symbols represent measurements from RFP and orange symbols represent measurements from the overlap between GFP and RFP. Each symbol represents results from one bronchial sample (patient), with solid symbols representing hASM cells from female patients and open symbols representing hASM cells from male patients. Data are presented as means ± SD using a nonlinear regression model (curve fit analysis) from six patients. Statistical analyses were based on measurements from n=60 hASM cells per treatment group (n=10 cells per patient; 6 patients) and mixed-effects analysis with Sidák’s multiple comparison test was used to compare measurements between GFP and RFP as well as between FCCP treated and control (0 h) hASM cells. hASM: Human Airway Smooth Muscle; FCCP: Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone; GFP: Green Fluorescent Protein; RFP: Red Fluorescent Protein; 3-D: 3-Dimensional; MFI: Mean Fluorescence Intensity

Discussion

In the present study, we validated a robust technique that utilizes simultaneous confocal imaging of dual labeled mitochondria to quantify the volume of intact mitochondria in hASM cells. By combining a membrane potential-independent mitochondrial label (CellLight Mitochondria-GFP) with a membrane potential-dependent label (MitoTracker Red FM), this novel method allows us to distinguish intact (polarized) from damaged (depolarized) mitochondria under both homeostatic conditions and in response to graded mitochondrial stress. The key findings of the present study are as follows: (1) In untreated hASM cells, with increasing time, mitochondrial volume density measured using CellLight GFP exceeds that measured using MitoTracker Red, indicating mitochondrial clearance under homeostatic conditions; (2) Dose- and time-dependent treatment of FCCP demonstrates differential stability of the two probes, with MitoTracker Red fluorescence declining more rapidly than CellLight GFP; (3) The assay sensitively detects decrease in intact mitochondrial volume in homeostatic conditions as well as in response to FCCP mediated mitochondrial depolarization; (4) The assay accurately quantifies the morphology of intact mitochondria (more filamentous) versus damaged mitochondria (more fragmented). Our findings establish this assay as a complementary tool for assessing mitochondrial quality and provide new insight into the dynamics of mitochondrial depolarization and damage in hASM cells.

Labeling characteristic of CellLight GFP and MitoTracker Red

Loss of mitochondrial membrane potential (ΔΨ_m) can occur in response to a wide range of stressors and accumulation of such depolarized (damaged) mitochondria can disrupt cellular homeostasis (4, 7, 8). Visualization and quantification of mitochondrial depolarization is therefore critical for understanding its many patho-physiological consequences. Over the years, multiple strategies have been developed to assess changes in ΔΨ_m, among which fluorescent potentiometric dyes such as TMRM are widely used (13, 14, 31, 32). While informative, ΔΨ_m measurement using TMRM is influenced by several factors, including plasma membrane potential, photobleaching, and limited sensitivity to minute ΔΨ_m changes within mitochondrial subpopulations (15, 46, 47). Therefore, appropriate controls and complementary assays are required to ensure accurate quantitative interpretation of ΔΨ_m.

In the present study, we used 3D confocal imaging of fluorescent probes which exhibit markedly different sensitivities to ΔΨ_m to distinguish intact (polarized) mitochondria from damaged (depolarized) mitochondria. One such probe is MitoTracker, a useful tool commonly used in evaluating mitochondrial morphology and volume, and have been extensively validated across multiple cell types, including hASM cells (20, 23, 32, 34, 35, 48–52). Previously, we demonstrated that mitochondrial volume density obtained using MitoTracker labeling and 3D confocal imaging are comparable to those derived from electron microscopy (EM) (35). However, compared to EM, MitoTracker-based imaging offers the advantages of higher sampling size and measurements from metabolically active mitochondria. MitoTracker dye accumulation is strictly ΔΨ_m dependent and therefore fails to label mitochondria that have undergone depolarization (20, 22). Therefore, a decrease in MitoTracker signal likely reflects loss of polarized, functional mitochondria.

Alternatively, CellLight Mitochondria-GFP labels mitochondria independently of ΔΨ_m (53, 54). CellLight GFP is a commercial BacMam vector which are baculoviruses engineered to carry GFP, fused to the leader sequence of E1α pyruvate dehydrogenase, into living mammalian cells enabling transient mitochondrial labeling (16–18). The specificity of mitochondrial labeling by CellLight GFP has been extensively validated by colocalization with MitoTracker dyes and EM (55). Although transduction efficiency can vary across cell types, CellLight GFP has been widely used for live-cell and super-resolution imaging, illustrating its compatibility for dynamic mitochondrial studies in primary or immortalized cells (18, 53, 54). Therefore, by combining CellLight GFP with MitoTracker Red labeling, we are able to quantify changes in mitochondrial volume as a measure of mitochondrial damage.

Comparison of CellLight GFP and MitoTracker Red labeling in physiological conditions

Building on the differential labeling characteristics of CellLight GFP and MitoTracker Red, we applied the dual-labeling strategy to assess mitochondrial integrity and off-target effects in hASM cells in untreated conditions. By thresholding for the fluorescence overlap between the two probes and quantifying mitochondrial volume, we were able to derive a quantitative index of intact, membrane-polarized mitochondria. Consistent with previous observations, mitochondrial volume measured using CellLight GFP was greater than that measured using MitoTracker Red in untreated hASM cells across time (23). These findings indicate that a fraction of the mitochondrial volume fails to uptake MitoTracker despite being structurally present, suggesting reduced ΔΨ_m in a subset of the mitochondrial population even under homeostatic (untreated) conditions.

Assessment of mitochondrial turnover in the absence of overt stress has been given comparatively limited attention. Defining what constitutes “healthy” mitochondrial population is inherently challenging, as mitochondria are highly dynamic organelles that continuously undergo fission and fusion to remodel their network architecture to adapt their density and function in response to fluctuating metabolic demands (56–58). Past investigations on mitochondrial health were primarily focused on pathological abnormalities of mitochondria, emphasizing gross alterations in mitochondrial morphology or disruption of inner and outer mitochondrial membranes (8, 59–62). While informative, these approaches often do not directly quantify the volume of intact mitochondria within a heterogeneous network. Our findings demonstrate the existence of a basal mitochondrial population with reduced ΔΨ_m, consistent with ongoing homeostatic turnover of damaged mitochondria. Importantly, this subpopulation would likely be underestimated or entirely overlooked when using ΔΨ_m-dependent dyes alone, highlighting a key limitation of conventional approaches. By integrating ΔΨ_m-dependent with ΔΨ_m-independent mitochondrial labels, our approach provides enhanced sensitivity for detecting subtle mitochondrial damage under physiological conditions. Importantly, these observations further support the concept that mitochondrial turnover is a continuous, tightly regulated homeostatic process that operates even in the absence of exogenous stressors.

Optimization of FCCP concentration and duration of exposure

Having established that the combined use of CellLight GFP and MitoTracker Red can resolve basal ΔΨ_m heterogeneity under physiological conditions, we optimized this strategy under conditions of defined mitochondrial stress. To achieve this, hASM cells were exposed to FCCP, a well-established mitochondrial uncoupler extensively utilized to model mitochondrial stress by inducing loss in ΔΨ_m (63, 64). Across a range of concentrations, FCCP-induced loss of ΔΨ_m, as confirmed independently by TMRM, resulted in a marked reduction in MitoTracker Red signal and its overlap with CellLight GFP. Notably, even low concentrations of FCCP (0.5 μM) resulted in a substantial decrease in MitoTracker Red signal, indicating that partial depolarization is sufficient to disrupt MitoTracker Red uptake. Time-course experiments further revealed differential stability of the two probes, with MitoTracker Red signal declining more rapidly than CellLight GFP. This divergence resulted in a dose- and time-dependent reduction in intact mitochondrial volume. Although some loss of MitoTracker signal may reflect dye redistribution or photobleaching, the relative preservation of CellLight GFP labeling supports the interpretation that reduction in fluorescence overlap primarily reflect changes in ΔΨ_m rather than loss of mitochondrial volume. This is reflected by the increased mitochondrial volume density observed in individual assessments of the two dyes. The graded loss of fluorescence overlap and intact mitochondrial volume across FCCP concentrations and exposure durations underscores the utility of this approach for quantitatively assessing the severity of mitochondrial damage. These findings further highlight the importance of ratiometric and colocalization-based analyses, particularly in longitudinal imaging experiments.

Interestingly, we also observed a progressive decline in CellLight GFP signal with increasing FCCP concentration (70±10%) and duration of exposure (21.3±15%), with a similar decrease in mitochondrial volume density. A similar, albeit less pronounced, decline in CellLight GFP signal (7.2%) and mitochondrial volume density was observed in untreated hASM cells over time. This decline of CellLight GFP signal likely reflects mitochondrial clearance as a result of activation of mitophagy and argues against non-specific loss of fluorescence and reflects physiological mitochondrial clearance (4–6). While pH-sensitive mitophagy reporters are commonly used to assess mitochondrial delivery to lysosomes and overall mitophagic flux, they are unable to monitor the rate of clearance of specific mitochondrial subpopulations (12, 15, 22, 65–67). In contrast, this assay also enables the simultaneous quantification of the rate of mitochondrial clearance, providing a more nuanced assessment of mitochondrial quality control.

Use of pMitoTimer to Assess Mitochondrial Turnover

To further evaluate that the observed loss of fluorescence overlap between CellLight GFP and MitoTracker Red reflects true mitochondrial damage (depolarization) rather than dye-specific artifacts, we complemented our imaging-based analysis with the a redox-sensitive mitochondrial reporter. The pMitoTimer reporter was developed from a fluorescent reporter gene, pTimer, which encodes DsRed1-E5 that targets mitochondria through the mitochondria targeting sequence at the N-terminus of its coding region and fluoresces as GFP when newly synthesized and irreversibly shifts to RFP upon oxidation (45). A key advantage of using pMitoTimer is that the shift from green to red fluorescence is independent of pH, ionic strength, and protein concentration, but is highly sensitive to oxidative stress driven by excessive ROS production, enabling direct assessment of mitochondrial protein oxidation and turnover rate (68–70). Under physiological conditions, the balanced synthesis and degradation of mitochondrial proteins yields a relatively stable red and green fluorescence signal. Therefore, ratiometric analysis of red/green fluorescence signal corrects for variations in expression level and provides a robust index of mitochondrial turnover rate than comparison of individual fluorescence intensities (70).

In the present study, there was a shift in the pMitoTimer signal in untreated hASM cells resulting in decreased red/green ratio over time, albeit not significant, indicative of the presence of basal mitochondrial turnover. FCCP-mediated loss of ΔΨ_m is associated with an increase in ROS formation and oxidative stress due to impaired electron transport chain (1, 7, 8, 71). As a result, FCCP treatment induced a significant dose- and time-dependent shift from green to red fluorescence accompanied by an accumulation of red puncta, indicative of increased rate of mitochondrial turnover. Importantly, the correlation between reduced uptake of MitoTracker Red, decreased intact mitochondrial volume and shift in pMitoTimer fluorescence strongly supports the fact that FCCP induces progressive mitochondrial damage rather than probe-specific signal loss. These findings demonstrate that mitochondrial depolarization and oxidative injury occur in parallel and can be quantified using complementary imaging-based metrics. When integrated with the dual-labeling approach described above, pMitoTimer provides additional validation of mitochondrial turnover, strengthening the overall assessment of mitochondrial quality control in hASM cells.

Mitochondrial Fission Separates Damaged Mitochondrial Segments

Mitochondria are dynamic organelles that are continuously remodeled through coordinated cycles of fusion and fission (fragmentation) (21, 57, 58). Through mitophagy, damaged mitochondria are selectively removed to maintain a healthy mitochondrial network (3, 10, 57). In hASM cells, the mitochondrial network comprises both elongated/filamentous mitochondrial networks as well as fragmented/globular mitochondria, contributing to variability in mitochondrial morphology metrics such as the mitochondrial complexity index (MCI) (34, 41, 42, 59). Under physiological conditions, the overall morphology of intact mitochondria remained relatively unchanged over time, with predominantly filamentous networks. However, analysis of individual probes revealed that MitoTracker Red labeled mitochondria exhibited higher MCI values compared with CellLight GFP labeled mitochondria, indicating that polarized (intact) mitochondria tend to maintain a more filamentous morphology, whereas CellLight GFP labels a more morphologically heterogeneous mitochondrial population. Following FCCP exposure, a pronounced dose- and time-dependent shift toward mitochondrial fragmentation was observed, reflected by reduced MCI in both labeling approaches. Consistent with observations in untreated hASM cells, MitoTracker Red labeled mitochondria had relatively higher MCI values at lower FCCP concentrations and shorter exposure durations, suggesting that early ΔΨ_m loss mediated fragmentation occurs within localized regions of the mitochondrial network prior to widespread fragmentation and structural remodeling. Mitochondrial damage is frequently accompanied by increased fission to segregate damaged mitochondrial segments that are selectively eliminated by mitophagy (4, 6, 57). Our findings support this framework, indicating that mitochondrial depolarization detected by dual-fluorescence imaging is strongly associated with progressive fragmentation and segregation of damaged mitochondrial segments prior to clearance.

Summary and Conclusion

In this study, we present a straightforward and interpretable approach for assessing mitochondrial integrity that complements established assays. By integrating ΔΨ_m-dependent and ΔΨ_m-independent mitochondrial labels, this method enables us to differentiate between intact and damaged mitochondria within a heterogeneous mitochondrial population. Future studies applying this strategy under pathophysiological conditions, including inflammation, metabolic stress and aging-related insults, is necessary to fully define its utility in understanding endogenous mitochondrial turnover mechanisms. In hASM cells, mitochondrial dysfunction has been implicated in various airway diseases, wherein this approach can be valuable in elucidating mitochondrial turnover. By enabling accurate quantification of intact versus damaged mitochondria in situ, this approach addresses an important methodological gap in mitochondrial biology and provides a framework for advancing our understanding of mitochondrial damage and rate of turnover.

Nomenclature

hASM

human Airway Smooth Muscle

FCCP

Carbonyl cyanide-p-rifluoromethoxyphenylhydrazone

ROS

Reactive Oxygen Species

α-SMA

α-Smooth Muscle Actin

DAPI

Diamidino-2-Phenylindole, Dihydrochloride

TMRM

Tetramethylrhodamine, Methyl Ester

ΔΨ_m

Mitochondrial Membrane Potential

MCI

Mitochondrial Complexity Index

MFI

Mean Fluorescence Intensity

GFP

Green Fluorescent Protein

RFP

Red Fluorescent Protein

AU

Arbitrary Units

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.