Mitochondrial adaptations to inactivity in diaphragm muscle fibers

Abstract

Type I and IIa diaphragm muscle (DIAm) fibers comprise slow and fast fatigue-resistant motor units that are recruited to accomplish breathing and thus have a high duty cycle. In contrast, type IIx/IIb fibers comprise more fatigable fast motor units that are infrequently recruited for airway protective and straining behaviors. We hypothesize that mitochondrial structure and function in type I and IIa DIAm fibers adapt in response to inactivity imposed by spinal cord hemisection at C₂ (C₂SH). At 14 days after C₂SH, the effect of inactivity on mitochondrial structure and function was assessed in DIAm fibers. Mitochondria in DIAm fibers were labeled using MitoTracker Green (Thermo Fisher Scientific), imaged in three-dimensions (3-D) by fluorescence confocal microscopy, and images were analyzed for mitochondrial volume density (MVD) and complexity. DIAm homogenate from either side was assessed for PGC1α, Parkin, MFN2, and DRP1 using Western blot. In alternate serial sections of the same DIAm fibers, the maximum velocity of the succinate dehydrogenase reaction (SDH_max) was determined using a quantitative histochemical technique. In all groups and both sides of the DIAm, type I and IIa DIAm fibers exhibited higher MVD, with more filamentous mitochondria and had higher SDH_max normalized to both fiber volume and mitochondrial volume compared with type IIx/IIb Diam fibers. In the inactive right side of the DIAm, mitochondria became fragmented and MVD decreased in all fiber types compared with the intact side and sham controls, consistent with the observed reduction in PGC1α and increased Parkin and DRP1 expression. In the inactive side of the DIAm, the reduction in SDH_max was found only for type I and IIa fibers. These results show that there are intrinsic fiber-type-dependent differences in the structure and function of mitochondria in DIAm fibers. Following C₂SH-induced inactivity, mitochondrial structure (MVD and fragmentation) and function (SDH_max) were altered, indicating that inactivity influences all DIAm fiber types, but inactivity disproportionately affected SDH_max in the more intrinsically active type I and IIa fibers.

NEW & NOTEWORTHY Two weeks of diaphragm (DIAm) inactivity imposed by C2SH caused reduced mitochondrial volume density, mitochondrial fragmentation, and a concomitant reduction of SDH_max in type I and IIa DIAm fibers on the lesioned side. Type I and IIa DIAm fibers were far more sensitive to inactivation than type IIx/IIb fibers, which exhibited little pathology. Our results indicate that mitochondria in DIAm fibers are plastic in response to varying levels of activity.

INTRODUCTION

The diaphragm muscle (DIAm) is unique to mammals and the primary muscle involved in breathing. Neural control of the DIAm involves the orderly activation of motor units, comprising a phrenic motor neuron and the DIAm fibers it innervates (1, 2). Motor units are classified based on their mechanical and fatigue properties (3) and their fiber-type composition (4–6). Slow, fatigue-resistant DIAm motor units (type S) comprise type I fibers expressing the slow myosin heavy chain (MyHC_Slow) isoform. Fast, fatigue-resistant (type FR) DIAm motor units comprise type IIa fibers expressing MyHC_2A. More fatigable fast (type FInt and FF) DIAm motor units comprise type IIx/IIb fibers expressing MyHC_2X and MyHC_2B isoforms in varying proportions. Type S and FR DIAm motor units are recruited for ventilation, with a duty cycle of ∼40% (1, 2, 4, 5) whereas FInt and FF motor units are far less active and recruited infrequently for expulsive airway clearance/straining behaviors (1, 2, 7). Thus, there are marked differences in the activity and energetic demands of type I and IIa compared with type IIx/IIb DIAm fibers (8, 9).

Mitochondria produce ATP to meet the energetic demands of muscle fibers based on their activation. Mitochondrial volume density (MVD) varies across fiber types, matching differences in activation history rather than differences in ATP consumption once activated (6, 8). For example, type I and IIa DIAm fibers in adult rats have greater MVD and are recruited more frequently than type IIx/IIb fibers (6, 10), whereas type IIx/IIb fibers have the greatest ATP consumption rates during activation and are recruited infrequently (6, 8). The density and structure of mitochondria adapt in response to altered energy demands (11–14). Under conditions of increased energy demand, mitochondrial biogenesis occurs in muscle fibers to increase MVD (15–17). In addition, differences in mitochondrial morphology may also underlie functional differences (11–14), with mitochondrial fragmentation associated with decreased mitochondrial respiratory capacity (18–20).

Respiratory capacity of muscle fibers is often assessed by respirometry, detecting changes in O₂ consumption rates (OCR) in response to specific inhibitors of the electron transport chain (ETC) and the H⁺ ionophore carbonyl cyanide 4-trifluoromethoxyphenylhydrazone (FCCP). The maximum respiratory capacity is determined based on changes in OCR. Although respirometry claims to assess maximum respiratory capacity of single muscle fibers, in practice, measurements are conducted on small bundles of permeabilized muscle fibers, obfuscating fiber-type-specific information. Previously, we developed a quantitative histochemical technique to measure the maximum velocity of the succinate dehydrogenase reaction (SDH_max) in single muscle fibers (21, 22). Succinate dehydrogenase is a key enzyme in the tricarboxylic acid (TCA) cycle, located in the inner mitochondrial membrane, as well as complex II of the ETC. Thus, SDH_max provides estimates of the maximum respiratory capacity of single muscle fibers. Previously, we reported that the SDH_max of type I and IIa DIAm fibers is higher than that of type IIx/IIb fibers (10, 18, 21, 23). Importantly, we found that after 2 wk of DIAm inactivity imposed by C₂ spinal cord hemisection (C₂SH), there was an ∼25% reduction in SDH_max in type I and IIa DIAm fibers, whereas type IIx/IIb fibers were unaffected (24).

Recently, we reported that differences in SDH_max among DIAm fibers reflect both differences in MVD as well as differences in the respiratory capacity per mitochondrial volume (intrinsic respiratory capacity) (10, 18). In alternate serial sections, we determined the SDH_max, MVD, and morphology (labeled via MitoTracker Green). Using this approach, we hypothesized that after 2 wk of inactivity imposed by C₂SH, SDH_max, MVD, and the intrinsic mitochondrial respiratory capacity are selectively reduced in type I and IIa DIAm fibers. We further hypothesized that after 2 wk of C₂SH, mitochondrial morphology will become more fragmented in type I and IIa DIAm fibers, underlying the reduced intrinsic mitochondrial respiratory capacity.

MATERIALS AND METHODS

Ethics Approval

All procedures were performed in accordance with the American Physiological Society’s Animal Care Guidelines and the National Institutes of Health (NIH) guide for the use and care of laboratory animals. These procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Mayo Clinic.

Experimental Animals

A total of 14 adult male Sprague–Dawley (300–370 g) were obtained from Envigo (Indianapolis, IN) for use in the study. Eight rats were assigned to a C₂SH group, anticipating that based on previous studies, 25% of the animals would be excluded due to >10% recovery of inspiratory DIAm electromyogram (EMG) activity during the 2-wk period (25, 26). Male rats were used to compare recovery rates in past studies that included only males (25, 26). In addition, there are no sex-based differences in DIAm force, fatigue, fiber-type-specific cross-sectional areas, MVD, and SDH_max in rats (10, 27–32). Animals were acclimated for at least 1 wk before surgery and maintained on an alternating 12:12-h light-dark cycle with ad libitum access to fresh water and rat chow. For all surgical procedures, anesthesia was administered via intraperitoneal injections of xylazine (10 mg/kg) and ketamine (100 mg/kg). The sham control animals underwent the same surgical procedures (DIAm electrode insertion, laminectomy), excluding hemisection of the spinal cord.

Cervical Spinal Cord Hemisection Surgery

The surgical methods for C₂SH have been previously described in detail (33, 34). Briefly, after performing a dorsal laminectomy, the C₂ spinal cord was cut using a surgical microknife, beginning anterior to the dorsal root entry zone fissure and proceeding ventrally taking care to preserve the dorsal funiculus. As in previous studies, DIAm EMG activity was monitored during surgery to ensure that rhythmic inspiratory activity on the right (ipsilateral) side of the DIAm disappeared during C₂SH surgery. In addition, animals were included in the C₂SH group only if inspiratory DIAm EMG activity remained absent [<10% of presurgical root mean square (RMS) amplitude] at 3, 7, and 14 days postsurgery. As in previous studies, spontaneous recovery of inspiratory DIAm EMG activity after C₂SH was observed in two of the eight animals, and these animals were excluded from further analysis. Immediately after surgery, rats were administered subcutaneous carprofen (5 mg/kg) and buprenorphine (0.1 mg/kg). Fourteen days following C₂SH or sham surgery, the rats were anesthetized by intraperitoneal injection of ketamine (80 mg/kg) and xylazine (20 mg/kg) and subsequently euthanized via exsanguination.

Diaphragm EMG Activity

A pair of fine wire electrodes (∼2 mm recording area) were implanted bilaterally into the midcostal region of the DIAm to record chronic DIAm EMG activity, as previously described (35, 36). The EMG electrode wires were tunneled subcutaneously to the dorsum of the rat for repeated recordings. Inspiratory DIAm EMG activity was recorded 3 days before C₂SH or sham surgeries, and these recordings were used as a baseline for subsequent analyses of DIAm EMG activity after C₂SH and sham surgeries (25, 26). We monitored the inspiratory-related activity throughout the 14-day post-C₂SH period (i.e., at 3, 7, and 14 days). Amplitude of inspiratory DIAm EMG activity was quantified as the root mean square (RMS; Fig. 1).

Figure 1.

A: representative recordings of right and left DIAm EMG activity from a rat that underwent C₂SH surgery (right side) and a sham control animal that underwent all surgical procedures except C₂SH. Recordings were obtained immediately before laminectomy (presurgery), immediately after C₂SH or laminectomy in sham controls (postsurgery), 3 days after surgery, and 14 days after surgery. B: root mean square (RMS) amplitude of DIAm EMG activity was calculated for each animal and normalized to the presurgery average inspiratory RMS EMG. The scatterplot shows average normalized inspiratory RMS EMG for the left (sham: white; C₂SH: blue) and right (sham: tan; C₂SH: red) sides. Inspiratory RMS EMG activities from the left and right sides were averaged for 20 consecutive breaths, and each rat is represented by a different symbol. Importantly, inspiratory DIAm RMS EMG on the right (lesioned) side disappears (<10% presurgery) in the C₂SH group postsurgery and does not recover over the next 14 days (*P < 0.05). DIAm, diaphragm muscle; EMG, electromyogram; C₂SH, spinal cord hemisection at C₂.

Tissue Preparation

DIAm strips (∼2 mm wide; 2–3/rat) were dissected from the right midcostal region. The DIAm strips were stretched upto 150% of resting length, to approximate L_o and fresh-frozen on cork in melting isopentane cooled by liquid nitrogen, in a manner identical to past reports (24, 31, 37). Transverse serial sections of DIAm fibers were cut at 10 µm thickness using a cryostat (Reichert Jung Frigocut 2800 Cryostat, Reichert Microscope Services, Depew, NY), maintained at −30°C.

Mitochondrial Imaging and Morphometry

To label mitochondria in type-identified DIAm fibers, alternate 10-μm-thick serial sections were placed in a solution containing 1.5-µL MitoTracker Green (300 nM; Thermo Fisher Scientific) in 5 mL of PBS for 30 min. The sections were then washed three times with PBS for 10 min each wash and a coverslip was placed for imaging.

Imaging of mitochondrial morphology and volume was accomplished using a ×40 oil-immersion objective (NA 1.3) on an Olympus FV2000 laser confocal microscope capable of sequential multicolor fluorescence imaging. Images were captured at 16-bit resolution. The 3-D high-resolution confocal imaging techniques used to reconstruct mitochondrial structure have been previously reported in detail (10, 14, 18, 38). The empirically calculated point spread function for the ×40 objective was used to set a Z-axis step size of 0.5 µm (39, 40). Based on these calibrations, the voxel dimensions were 0.207 µm x-axis, 0.207 µm y-axis, and 0.5 µm Z-axis resulting in 0.021 µm³ voxels. A region of interest (ROI) without MitoTracker Green signal was imaged to set the black level. Another ROI was delineated reflecting the most intense MitoTracker Green fluorescence, and the gain was adjusted to prevent saturation and maximize the dynamic range. After image acquisition, a blind deconvolution algorithm (Point Scan Confocal, 3 iterations; NIS-Elements; Nikon Instruments Inc., RRID:SCR_014329) was applied for each 0.5 µm optical slice in the Z-stack as previously described (41) to improve the spatial resolution. This deconvolution algorithm increased the spatial resolution twofold to ∼125 nm. The deconvolved images were then further processed for background correction, ridge filter detection, skeletonization, and thresholding with ImageJ (imagej.nih.gov/ij/, Fiji, RRID:SCR_002285), as previously described (14, 42). The gray level of MitoTracker Green fluorescence was thresholded to unambiguously identify voxels containing mitochondria using ImageJ/Fiji software and used to create binary images (10, 14, 43). These binarized images were then reconstructed in 3-D using ImageJ and NIS Elements (10) .

After reconstruction of binarized images in the Z-stack, the number of positive voxels within a muscle fiber boundary was used to determine fiber volume. The number of MitoTracker Green containing binarized voxels within each muscle fiber was used to determine mitochondrial volume. MVD was the ratio of mitochondrial volume to the total muscle fiber volume (Fig. 2).

Figure 2.

A: representative photomicrographs of mitochondria labeled using MitoTracker Green in an alternate serial section (10 µm thick) of DIAm fibers (other alternate sections of the same fibers used for MyHC fiber-type classification and SDH_max measurements). MitoTracker Green fluorescence was visualized using an Olympus FV2000 laser scanning confocal microscope. B: fluorescence intensity was thresholded to produce binary images of the MitoTracker Green labeled mitochondria. C: a series of 20 optical slices (each 0.5 µm) were imaged for each DIAm fiber, and the images of labeled mitochondria were reconstructed in 3-D for morphometric analyses. The 3-D Z-projection images of MitoTracker Green-labeled mitochondria in a type I, IIa, and IIx/IIb fiber are shown. DIAm, diaphragm muscle; MyHC, myosin heavy chain; SDH_max, maximum velocity of the succinate dehydrogenase reaction; 3-D, three-dimensional.

Mitochondrial morphology was measured using the ImageJ/Fiji Mitochondria Analyzer plug-in. Mitochondrial complexity index (MCI) was used as a morphological measurement (44). Mitochondrial complexity index (MCI) was calculated using the following equation:

$$\text{MCI}=\frac{S{A}^3}{16{\Pi }^2{V}^2},$$

where SA is total mitochondrial surface area within a muscle fiber and V is the total mitochondrial volume within the muscle fiber.

Detection and Quantification of Proteins Using Western Blot

Expressions of Parkin, peroxisome proliferator-activated receptor-γ coactivator 1-α (PGC1α), Mitofusin2 (MFN2), and dynamin-related protein 1 (DRP1) proteins in the DIAm were evaluated by Western blot. A segment of the midcostal DIAm (∼10–15 mg) was homogenized, and total protein lysate was extracted in 1× Cell Lysis Buffer (Cell Signaling Technology) supplemented with protease inhibitor cocktail (Millipore Sigma) and phosphatase inhibitors (PhosSTOP, Millipore Sigma) to inhibit the degradation of proteins. Protein concentration was quantified using detergent compatible (DC) protein assay (Bio-Rad) that uses a colorimetric method, similar to the Lowry method. For each blot, 60–80 µg of total protein samples were denatured in 1× Laemmli sample buffer (Bio-Rad) with β-mercaptoethanol at 100°C for 3 min, separated by stain-free SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad) using the Trans-Blot Turbo system (Bio-Rad). Target-specific primary antibodies were used to detect each of the proteins at a dilution of 1:1,000. The target-specific primary antibodies were Parkin (Santa Cruz Biotechnology, Cat. No. sc-32282), PGC1α (Novus Biologicals, Cat. No. NBP1-04676), MFN2 (Abcam, Cat. No. ab50843), and DRP1 (Abcam, Cat. No. ab56788). Horseradish peroxidase-conjugated secondary antibodies were used at a dilution of 1:7,500. Bands were visualized with chemiluminescent substrate SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific). Band intensity was quantified using Image Lab software (Bio-Rad) and normalized to total protein loaded in the lane. Each animal served as their own control, and protein expression on the right side of the DIAm was compared with the left side in the same gel.

Quantification of Mitochondrial DNA Copy Number

Total DNA was extracted from DIAm tissues (∼10–15 mg) using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) and quantified using nanodrop. Mitochondrial DNA copy number was quantified using the Rat Mitochondrial DNA Copy Number Assay Kit (Detroit R&D). In brief, ∼10–20 ng of total DNA was used for real-time PCR following the manufacturer’s instructions using one set of nuclear and another set of mitochondrial DNA-specific primers. Cycle threshold (Ct) values for both nuclear and mitochondrial copy numbers were calculated using the following formula:

$$\text{Mitochondrial}\hspace{0.17em}\text{DNA}\hspace{0.17em}\text{copy}\hspace{0.17em}\text{number = }\Delta Ct=Ct(\text{mitochondria})-Ct(\text{nucleus}).$$

DNA copy number was presented as 2^−ΔCt and compared between the left and right sides of the DIAm for each rat. Each animal served as their own control, and the mitochondrial DNA copy number on the right side of the DIAm was compared with the left side of each animal.

Maximum Velocity of Succinate Dehydrogenase Reaction

The quantitative histochemical technique for measuring SDH_max in single muscle fibers has been previously described in detail (21, 22). Briefly, alternate serial sections of DIAm fibers were cut at 6 µm and incubated in a solution containing 80 mM succinate (maximum substrate level determined in preliminary studies), 1.5 mM nitro blue tetrazolium (NBT—the reaction indicator), 5 mM EDTA, 0.2 mM mPMS, and 0.1 mM azide in 0.1 M phosphate buffer (pH = 7.6). In previous studies, we determined that with succinate concentrations >80 mM, the SDH reaction in muscle fibers is not substrate limited (10, 21, 22). The reduction of NBT to its diformazan (NBT_dfz) is used as the SDH reaction indicator. NBT_dfz precipitation was quantified using an inverted light microscope (Olympus IX71, Olympus America, Melville, NY) with a ×40 objective. An interference filter (570 nm) was used to limit the spectral range of the light source to the maximum absorption wavelength of NBT_dfz. Optical density (OD) calibration was performed by imaging a series of known gray levels at 16-bit resolution. The dynamic range during imaging was maximally adjusted while avoiding saturation. An image of nonspecific activity was acquired in one section of muscle fibers incubated in a 200-µL solution that did not contain succinate. Thereafter, another section of muscle fibers was incubated in a solution containing 80 mM succinate, and images were repeatedly acquired every 15 s as the SDH reaction proceeded in a 1,024 × 1,024 pixel array using a digital camera (Hamamatsu ORCA Flash 4.0, Model C11440). The images were then processed using MetaMorph software (Molecular Devices LLC., Sunnyvale, CA). The SDH_max was calculated using the Beer–Lambert–Bouguer law:

$$\frac{d[{\text{NBT}}_{\text{dfz}}]([\text{fumarate}])}{dt}=\frac{d\text{OD}/dt}{kl},$$

where the change in NBT_dfz concentration over time (d[NBT_dfz]/dt) within DIAm fibers was determined by the change in average optical density (dOD/dt) within the fiber boundary, with k being the molar extinction coefficient of NBT_dfz (26,478 mol/cm) and l the path length of light absorbance (6 µm). The average SDH_max within the volume of an individual DIAm fiber (fiber cross-sectional area times section thickness) was expressed as mmol fumarate/L fiber/min. SDH_max was also normalized for MVD of individual DIAm fibers and expressed as mmol fumarate/L mitochondria/min.

In a variety of respirometry approaches, basal, reserve, and maximum respiratory capacities (i.e., OCR) are determined using a stress test that includes inhibition of complex I and III of the ETC (45, 46) and uncoupling of O₂ consumption from the proton (H⁺) gradient using the H⁺ ionophore FCCP (47, 48). To verify that SDH_max indeed reflects the maximum respiratory capacity of DIAm fibers, a similar stress test was used. The velocity of the SDH reaction (with and without 80 mM succinate across a 10-min period) was determined in three alternate serial sections (6 µm each) of DIAm fibers that were 1) untreated (normal SDH solution), 2) treated with FCCP (1 mM to disrupt the proton gradient), and 3) treated with rotenone (1 mM) to inhibit complex I of the ETC and antimycin A (1 mM to inhibit complex III of the ETC). Consistent with previous studies (10), we found that the rate of the SDH reaction in all DIAm fibers was unaffected by FCCP treatment indicating that SDH_max reflects maximum respiratory capacity (maximum OCR). Following treatment with a combination of rotenone and antimycin A, the velocity of the SDH reaction was markedly slowed in all DIAm fibers.

Diaphragm Muscle Fiber-Type Classification

Immunoreactivity for different MyHC isoforms was used to classify different DIAm fiber types, as previously described (4, 9, 32). Briefly, DIAm sections were fixed in 4% paraformaldehyde for 10 min and rinsed with 0.1 mM PBS before incubation at 4°C in solutions containing primary antibodies for MyHC_Slow [NBP2 (IgG), 1:25 dilution; Novus], MyHC_2A [N1551 (IgM), 1:1 dilution; Novus], and MyHC_2X [NBP1 (IgG), 1:10 dilution; Novus]. Immunoreactivity for MyHC_Slow and MyHC_2A isoforms was determined in the same section with Alexa594 and Cy3 conjugated secondary antibodies, respectively, at a 1:200 dilution. In a second alternate section, MyHC_2X isoform immunoreactivity was determined using a Cy3-conjugated secondary antibody (1:200 dilution). A third alternate section of the same DIAm fibers was incubated with the BF35 antibody that reacts with all MyHC isoforms except MyHC_2X (i.e., all-but-2x MyHC antibody). Immunoreactivity for the BF35 antibody was determined using a Cy3-conjugated secondary antibody (1:200 dilution). Based on the pattern of immunoreactivity, DIAm fibers were classified as type I (immunoreactivity for MyHC_Slow), type IIa (immunoreactivity for MyHC_2A), and type IIx/IIb (immunoreactivity for MyHC_2X with some fibers displaying varying immunoreactivity for the all-but-2x MyHC antibody) (4, 9, 32). These antibodies have been extensively validated in single fiber histochemical and biochemical studies (9, 23, 49–53). After imaging, the proportions of different DIAm fiber types and their cross-sectional areas were determined using morphometric tools in ImageJ (ImageJ, US National Institutes of Health, Bethesda, MD, https://imagej.nih.gov/ij/, 1997–2018). Based on laminin immunoreactivity delineating the muscle fiber sarcolemma (Fig. 3A), cross-sectional areas of the type identified DIAm fibers were measured (n = 15 fibers/type per side/animal with 6 animals/group).

Figure 3.

A: representative, triple-labeled confocal microscope images of cross sections of DIAm fibers immunoreacted to primary antibodies for MyHC_slow (red), MyHC_2A (green) isoforms, and MyHC_2X unlabeled in this image. Also included is immunoreactivity to a primary antibody for laminin (blue) to clearly define muscle fiber borders. B: scatterplots showing the mean cross-sectional areas of type I, IIa, and IIx/IIb fibers from the left (sham: white; C₂SH: blue) and right (sham: tan; C₂SH: red) sides of the DIAm (n = 15 fibers/type/rat; n = 6 rats/group; each rat is represented by a different symbol). Note that type I and IIa DIAm fiber cross-sectional areas were equivalent on both left and right sides but significantly smaller than those of type IIx/IIb fibers. There were no changes in DIAm fiber cross-sectional areas at 14 days after surgery in both C₂SH and sham groups. DIAm, diaphragm muscle; MyHC_slow, slow myosin heavy chain; C₂SH, spinal cord hemisection at C₂.

Based on immunoreactivity for the anti-MyHC_Slow and anti-MyHC_2A antibodies, the classification of type I and IIa fibers, respectively, in the rat DIAm was unambiguous (Fig. 3A). The classification of fibers expressing the MyHC_2X isoform in the rat DIAm was also unambiguous. However, we were unable to validate an anti-MyHC_2X antibody, and without this, it was not possible to clearly distinguish fibers expressing any amount of the MyHC_2B isoform. For this reason, DIAm fibers expressing the MyHC_2X isoform were classified as type IIx/IIb, although most were likely type II_X fibers. Fiber-type proportions within the midcostal region of the DIAm were assessed from a sample of 100 fibers per animal (n = 6 animals/group).

Muscle Fiber Sampling and Statistical Analyses

To ensure unbiased sampling, we implemented a stereological method in which image fields were selected at lower magnification (×20 objective) to represent the entirety of the DIAm from the thoracic surface at the top left corner of the image field and moving in a diagonal direction until the abdominal surface was reached. At lower magnification, 100 muscle fibers were stereologically sampled to determine fiber-type proportions. For mitochondrial imaging and measurements of SDH_max, the magnification was increased to ×40 within these image fields and only fibers with borders fully contained within the field of view were sampled. This sampling method allowed for ∼8 fibers per higher magnification image field to be sampled. Images were sampled in this way until 15 DIAm fibers per fiber type for each animal per side were obtained. In previous studies, we found that 15 DIAm fibers per fiber type per animal is sufficient to detect differences in SDH_max of >20% across fiber types (α = 0.05, β = 0.8) (10, 23, 32). Outliers were classified as data points that were greater than twice the standard deviation from the mean. These outliers were excluded from analysis of this data.

Prism 8 was used for all statistical analyses (GraphPad, Carlsbad, CA). The data sets were assessed for normality with D’Agostino and Pearson tests. A two-way or three-way ANOVA was performed on the measured variables (e.g., fiber cross-sectional area, MVD, and MCI) with the following grouping factors: fiber type, side (left vs. right) and surgery (C₂SH vs. sham). When appropriate within-group differences were evaluated using Bonferroni post hoc tests. Cluster analyses were performed on z-scores using an R-based software package (Bluesky Statistics, Chicago, IL), with data points being ascribed to a single cluster (K-means clustering) and centroids reported. Data were reported as the mean ± 95% confidence interval of the mean. Statistical significance was set at P < 0.05.

RESULTS

Diaphragm EMG Activity in C₂SH and Sham Rats

As expected, C₂SH resulted in the absence of DIAm EMG activity on the ipsilateral side across the entire 14-day period in 6 of the 8 animals (P < 0.0001; Fig. 1A). Per a priori inclusion criteria, the two animals in which inspiratory activity recovered >10% presurgery were excluded from further analysis. In sham control rats, inspiratory DIAm activity was comparable on both left and right sides and unaffected by surgery (Fig. 1B). At 14 days postsurgery, inspiratory RMS EMG activity on the left side of DIAm in both C₂SH and sham rats was comparable (Fig. 1B). There was no significant increase in inspiratory EMG activity in the left side of the DIAm at the time of C₂SH surgery, at both 3 and 14 days post-C₂SH.

Ventilatory Patterns in C₂SH and Sham Rats

In anesthetized rats, respiratory rate was ∼60 breaths/min and unaffected after both C₂SH and sham surgeries (n = 4–6 animals/group). Respiratory rate remained at ∼60 breaths/min in both groups and across all time points (Table 1). However, compared with inspiratory duration before surgery and inspiratory duration in sham controls, inspiratory duration on day 0 (the day of surgery) was significantly prolonged immediately after C₂SH surgery (n = 6 animals/group; P < 0.0006, Table 1). This prolongation of inspiratory duration disappeared by 3 days after C₂SH surgery (P < 0.0006; Table 1). As a result of the prolonged inspiratory duration, inspiratory duty cycle (inspiratory duration divided by total respiratory cycle duration) was significantly increased immediately after C₂SH at day 0 (n = 6 animals/group, P = 0.0006). The C₂SH-induced prolongation of inspiratory duty cycle disappeared 3 days post-C₂SH (P = 0.0004; Table 1). Overall, the impact of transient C₂SH-induced changes in inspiratory duration and duty cycle was restricted to the left side of the DIAm and indicated a brief initial compensatory increase in activity of the intact left side of the DIAm, compared with both the right side in C₂SH rats and sham control animals.

Table 1.

Respiratory parameters

Parameter	Presurgery (Day 0)	Immediately Postsurgery (Day 0)	3 Days Postsurgery	14 Days Postsurgery	ANOVA
Respiratory rate, beats/min	C2SH: 60 ± 24 Sham: 52 ± 8	C2SH: 62 ± 8 Sham: 56 ± 10	C2SH: 61 ± 11 Sham: 53 ± 16	C2SH: 73 ± 17 Sham: 58 ± 11	Time: P = 0.29 Surgery: P = 0.13 Interaction: P = 0.80
Inspiratory duration, s	C2SH: 0.32 ± 0.06 Sham: 0.30 ± 0.02	C2SH: 0.56 ± 0.07* Sham: 0.32 ± 0.03	C2SH: 0.31 ± 0.12 Sham: 0.35 ± 0.06	C2SH: 0.34 ± 0.06 Sham: 0.32 ± 0.02	Time: P = 0.0006 Surgery: P = 0.01 Interaction: P = 0.002
Duty cycle, %	C2SH: 31 ± 9 Sham: 27 ± 2	C2SH: 60 ± 10* Sham: 30 ± 6	C2SH: 37 ± 7 Sham: 31 ± 12	C2SH: 39 ± 7 Sham: 32 ± 6	Time: P = 0.0006 Surgery: P = 0.0004 Interaction: P = 0.002

All data are represented as means ± 95% confidence interval. *P < 0.05 compared with all other groups and timepoints. Two-way ANOVA with Bonferroni post hoc tests. n = 6 for C₂SH animals, n = 4 for sham controls.

Diaphragm Fiber-Type Proportions and Cross-Sectional Areas

Fiber-type proportions were comparable on both left and right sides of the DIAm and were similar between C₂SH animals and sham controls. The proportions of type I and IIa fibers in the DIAm were ∼35% each with ∼30% type IIx/IIb. These fiber-type proportions are consistent with past reports in adult Sprague–Dawley rats (23), including those with a C₂SH lesion (33, 54).

Fiber cross-sectional areas varied across DIAm fiber types (P < 0.0001) but not between C₂SH animals and sham controls, nor between the left and right sides within C₂SH (Fig. 3B). As previously observed, cross-sectional areas of type IIx/IIb fibers were approximately fourfold larger than either type I or type IIa fibers in both C₂SH animals and sham controls (P < 0.0001; Fig. 3B). There were no differences in cross-sectional areas of type I fibers in either C₂SH animals or sham controls [sham left: 678 ± 118 µm²; sham right: 674 ± 190 µm²; C₂SH left: 563 ± 53 µm²; C₂SH right (inactive): 719 ± 118 µm²; all combinations of Bonferroni post tests P > 0.10]. There were no differences in cross-sectional areas of type IIa fibers in either C₂SH animals or sham controls [sham left: 636 ± 88 µm²; sham right: 668 ± 112 µm²; C₂SH left: 581 ± 93 µm²; C₂SH right (inactive): 708 ± 123 µm²; all combinations of Bonferroni posttests P > 0.42]. There were no differences in cross-sectional areas of type IIx/IIb fibers in either C₂SH animals or sham controls [sham left: 2,638 ± 123 µm²; sham right: 2,594 ± 40 µm²; C₂SH left: 2,491 ± 38 µm²; C₂SH right (inactive): 2,525 ± 127 µm²; all combinations of Bonferroni posttests P > 0.71]. These results are consistent with previous observations from our laboratory that reported no change in DIAm fiber cross-sectional areas at 2 wk after C₂SH (24, 33). However, in one previous study, we did note a minor decrease in cross-sectional area of type IIx/IIb fibers at 6 wk after C₂SH (54), but fiber length was not controlled at the time of freezing the DIAm in this study.

Mitochondrial Volume Density in Diaphragm Fibers

Using MitoTracker Green labeling and confocal imaging, MVD was measured in type-identified fibers on both sides of the DIAm in C₂SH animals and sham controls (n = 15 fibers/type/side/animal in each of six animals). MVD was dependent on fiber type (P < 0.0001), surgical group (P < 0.0001), and side (P < 0.0001), with an interaction between surgical group and side (P < 0.0001; Fig. 4A). In both experimental groups and both sides of the DIAm, the overall MVD in I and IIa fibers was higher than in type IIx/IIb fibers (P < 0.0001; Fig. 4A). In type I and IIa fibers, MVD was reduced by ∼40% 14 days after C₂SH compared with the left side (P < 0.0001) and sham controls (P < 0.0001; Fig. 4A). There was no difference in MVD in type I and IIa DIAm fibers in the sham controls, regardless of side. There was no difference between type I and IIa DIAm fibers on the left side of the C₂SH animals and sham controls. In type IIx/IIb fibers, MVD was reduced by ∼35% 14 days after C₂SH compared with the left side (P < 0.0001) and sham controls (P < 0.0001; Fig. 4A). There was no difference in MVD in type IIx/IIb fibers in the sham rats between the two sides (Fig. 4A). There was also no difference in MVD in type IIx/IIb fibers between the left side of the C₂SH animals and sham controls (Fig. 4A). The MVD in the right side of the C₂SH animals was decreased by ∼35%–40% in all fiber types (Fig. 4A).

Figure 4.

A: scatterplots showing that mitochondrial volume density (MVD; mitochondrial volume normalized to fiber volume) of type I and IIa DIAm fibers is greater than type IIx/IIb fibers (*P < 0.05). Fourteen days after C₂SH, MVD was reduced in all DIAm fibers on the right, inactive side (*P < 0.05), with the reduction was proportionately greater in type I and IIa fibers. Type IIx/IIb DIAm fibers from all experimental groups exhibited lower MVDs than type I and IIa DIAm fibers (#P < 0.05). Sham surgery did not affect MVD in DIAm fibers. B: scatterplots showing that the mitochondrial complexity index (MCI) was higher in type I and IIa fibers compared with type IIx/IIb fibers (*P < 0.05). Fourteen days after C₂SH, MCI was reduced in all DIAm fibers on the right, inactive side (*P < 0.05), with the reduction being proportionally greater in type I and IIa fibers. Type IIx/IIb DIAm fibers from all experimental groups exhibited lower MCI than type I and IIa DIAm fibers (#P < 0.05). Sham surgery did not affect the mean MCI in DIAm fibers. DIAm, diaphragm muscle; C₂SH, spinal cord hemisection at C₂.

Mitochondrial Fragmentation in Diaphragm Fibers

Based on three-dimensional (3-D) reconstruction of confocal optical sections, the extent of mitochondrial fragmentation was measured by calculating MCI in type identified DIAm fibers in C₂SH animals and sham controls (n = 15 fibers/type/side/animal with 6 animals/group). The MCI was dependent on fiber type (P < 0.0001), surgical group (P < 0.0001), and side (P < 0.0001), with interactions between side and surgical group (P < 0.0001) and fiber type, surgical group, and side (P < 0.0001; Fig. 4B). The MCI in type I and IIa DIAm fibers was higher than IIx/IIb fibers in both sham and C₂SH groups indicating more filamentous mitochondria (P < 0.0001). In both type I and IIa DIAm fibers, the MCI was reduced by ∼80% 14 days after C₂SH compared with the left side (P < 0.0001) and sham controls (P < 0.0001; Fig. 4B). There was no difference in the MCI in type I and IIa DIAm fibers between the left and right sides of the DIAm in sham controls. There was no difference in the MCI in type I and IIa DIAm fibers between the sham group and left side of C₂SH animals (Fig. 4B). In type IIx/IIb DIAm fibers, the MCI was reduced by ∼65% 14 days after C₂SH compared with the left side (P < 0.0001) and sham controls (P < 0.0001; Fig. 4B). There was no difference in the MCI in type IIx/IIb DIAm fibers between the left and right sides of sham controls. There was also no difference in the MCI in type IIx/IIb DIAm fibers between the sham controls and the left side in C₂SH animals (Fig. 4B).

Mitochondrial Biogenesis and Mitophagy-Related Marker Expression in Diaphragm Fibers

Based on Western blot analysis (Fig. 5A), Parkin expression in the right side of the DIAm in C₂SH animals was increased compared with the left side (P = 0.003; n = 4/group; Fig. 5B). There was no difference in the Parkin expression between the left and right sides of the DIAm in sham controls (Fig. 5B). PGC1α expression was decreased on the right side of the C₂SH animals as compared with the left side (P = 0.007; n = 4/group; Fig. 5C). No significant difference was observed in the PGC1α expression between the left and right sides of the DIAm in sham controls (Fig. 5C). Consistent with this finding, mitochondrial DNA copy number was significantly reduced in the right side of the DIAm in C₂SH animals as compared with the left side (P < 0.0001; n = 4/group; Fig. 5D). There was no difference in mitochondrial DNA copy number between the left and right sides of the DIAm in sham controls (Fig. 5D). Expression of MFN2 was similar between left and right sides of the DIAm in both C₂SH and sham animals (Fig. 6, A and B). In contrast, expression of DRP1 increased in the right side of the C₂SH animals compared with the left side (P < 0.0001; n = 4/group; Fig. 6C). There was no significant difference in expression DRP1 observed between the left and right side of DIAm in sham controls. These results are consistent with mitochondrial fragmentation in DIAm fibers of C₂SH animals (Fig. 4B).

Figure 5.

A: representative Western blots showing protein expression for Parkin and PGC1α from DIAm samples from left and right sides in both sham controls and C₂SH animals. B: scatterplots showing that Parkin expression is increased in the right side of the DIAm in C₂SH animals compared with the left side (*P < 0.05). There was no difference in Parkin expression between left and right sides of the DIAm in sham controls. C: scatterplots showing that PGC1α expression was decreased in the right side of C₂SH animals as compared with the left side (*P < 0.05). No significant difference was observed in the PGC1α expression between left and right sides of the DIAm in sham controls. D: scatterplots showing that mitochondrial DNA copy number was significantly reduced in the right side of the DIAm in C₂SH animals as compared with the left side (*P < 0.05). There was no difference in mitochondrial DNA copy number between left and right sides of the DIAm in sham controls. DIAm, diaphragm muscle; PGC1α, peroxisome proliferator-activated receptor-γ coactivator 1-α; C₂SH, spinal cord hemisection at C₂.

Figure 6.

A: representative Western blots showing protein expression for MFN2 and DRP1 in samples from left and right sides of the DIAm in both sham controls and C₂SH animals. B: scatterplots showing that expression of MFN2 was similar between left and right sides of the DIAm in both C₂SH and sham animals. C: scatterplots showing that the expression of DRP1 increased in the right side of the DIAm in C₂SH animals compared with the left side (*P < 0.05). There was no significant difference in expression DRP1 observed between left and right sides of DIAm in sham controls. DIAm, diaphragm muscle; DRP1, dynamin-related protein 1; MFN2, mitofusin2; C₂SH, spinal cord hemisection at C₂.

SDH_max per Fiber Volume in Diaphragm Fibers

Using a quantitative histochemical procedure, SDH_max per fiber volume was measured in type-identified fibers on both sides of the DIAm in C₂SH animals and sham controls (n = 15 fibers/type/side/animal with 6 animals/group). The SDH_max per fiber volume of DIAm fibers was dependent on fiber type (P < 0.0001), surgical group (P < 0.0001), and side (P < 0.0001), with interactions between side and surgical group (P < 0.0001) and fiber type, surgical group, and side (P < 0.0001; Fig. 7A). Overall, the SDH_max per fiber volume of type I and IIa DIAm fibers was significantly higher than type IIx/IIb fibers, consistent with their recruitment during breathing. In type I and IIa DIAm fibers, the SDH_max per fiber volume was reduced by ∼30% 14 days after C₂SH compared with the left side (P < 0.0001) and sham controls (P < 0.0001; Fig. 7A). There was no difference in the SDH_max per fiber volume in type I and IIa DIAm fibers between the left and right sides in sham controls. There was no difference in the SDH_max per fiber volume in type I and IIa DIAm fibers between the sham controls and the left side in C₂SH animals (Fig. 7A). There was no difference in the SDH_max per fiber volume in type IIx/IIb DIAm fibers between the left and right sides in both C₂SH and sham control animals, indicating that C₂SH had no effect on SDH_max per fiber volume in type IIx/IIb DIAm fibers (Fig. 7A).

Figure 7.

A: SDH_max in type I, IIa, and IIx/IIb fibers for the left (C₂SH: blue; sham: white) and right (C₂SH: red; sham: tan) sides of the DIAm at 14 days postsurgery. SDH_max per fiber volume was significantly reduced in type I and IIa DIAm fibers in the right (injured) side of the DIAm compared with the left side (*P < 0.05). No differences were observed in SDH_max per fiber volume between the right to left sides of the DIAm in the sham controls. SDH_max per fiber volume was greater in type I and IIa DIAm fibers regardless of group or side of the DIAm. B: cluster analysis of the relationship between MVD and SDH_max per fiber volume showed three distinct groups, with clear separation of type I and IIa DIAm fibers from type IIx/IIb fibers on both sides of the DIAm and in both C₂SH and sham control animals. DIAm, diaphragm muscle; MVD, mitochondrial volume density; SDH_max, maximum velocity of the succinate dehydrogenase reaction; C₂SH, spinal cord hemisection at C₂.

In sham controls and on the left side of the C₂SH animals, the relationship between z-scores of SDH_max per fiber volume and MVD revealed two distinct clusters of DIAm fibers. In one cluster, type I and IIa fibers displayed higher SDH_max per fiber volume and higher MVD (centroid: x = 0.98, y = 0.95) compared with a second cluster of lower SDH_max per fiber volume and lower MVD in type IIx/IIb fibers (centroid: x = −1.19, y = −1.29; Fig. 7B). In the right side of the DIAm in C₂SH animals, a third cluster was evident due to the reduction of SDH_max per fiber volume and MVD in type I and IIa fibers (centroid: x = −0.31, y = −0.07; Fig. 7B).

SDH_max per Mitochondrial Volume in Diaphragm Fibers

Using a quantitative histochemical procedure, SDH_max per mitochondrial volume was measured in type identified DIAm fibers on both sides of the DIAm in C₂SH animals and sham controls (n = 15 fibers/type/side/animal for both groups). The SDH_max per mitochondrial volume was dependent on fiber type (P < 0.0001), surgical group (P < 0.0001), and side (P < 0.0001), with interactions between side and surgical group (P < 0.0001) and fiber type, surgical group, and side (P < 0.0001; Fig. 8A). Overall, the SDH_max per mitochondrial volume of type I and IIa fibers was significantly higher than type IIx/IIb fibers, consistent with their recruitment in breathing. In type I and IIa DIAm fibers, the SDH_max per mitochondrial volume was reduced by ∼40% 14 days after C₂SH compared with the left side of the DIAm (P < 0.0001) and sham controls (P < 0.0001; Fig. 8A). There was no difference in the SDH_max per mitochondrial volume in type I and IIa DIAm fibers between the sham controls and the left side in C₂SH animals (Fig. 8A). There was no difference in the SDH_max per mitochondrial volume in type I DIAm fibers between the sham controls and the left side in C₂SH animals (Fig. 8A). In type IIx/IIb fibers, SDH_max per mitochondrial volume did not change when comparing the left and right of the sham controls and did not change after C₂SH (Fig. 8A). In both experimental groups and both sides of the DIAm, the SDH_max per mitochondrial volume in IIx/IIb fibers was lower than that in type I and type IIa fibers (P < 0.0001; Fig. 8A). There was no difference in the SDH_max per mitochondrial volume in type IIx/IIb DIAm fibers between left and right sides in both sham controls and C₂SH animals, indicating that C₂SH had no effect on the SDH_max per mitochondrial volume in type IIx/IIb DIAm fibers (Fig. 8A).

Figure 8.

A: SDH_max in type I, IIa, and IIx/IIb fibers for the left (sham: white; C₂SH: blue) and right (sham: tan; C₂SH: red) sides of the DIAm at 14 days postsurgery. SDH_max per mitochondrial volume significantly declined in type I and IIa DIAm fibers in the right (injured) side of the DIAm compared with the left side (*P < 0.05). No differences were observed in SDH_max per mitochondrial volume from the right to left sides of the DIAm in the sham controls. SDH_max per mitochondrial volume was greater in type I and IIa DIAm fibers compared with type IIx/IIb fibers regardless of group or side of the DIAm. B: cluster analysis of the relationship between MCI and SDH_max per mitochondrial volume showed four distinct groups, with clear seperation of type I and IIa DIAm fibers from the right (lesioned) side of the DIAm in C₂SH animals distinct from the type I and IIa fibers from the left (intact) side. A similar separation was observed in the distribution of MCI and SDH_max per mitochondrial volume in type IIx/IIb fibers from the right (lesioned) side of the DIAm in C₂SH animals compared with sham controls. DIAm, diaphragm muscle; MCI, mitochondrial complexity index; SDH_max, maximum velocity of the succinate dehydrogenase reaction; C₂SH, spinal cord hemisection at C₂.

In sham controls and on the left side of the C₂SH animals, the relationships between z-scores of SDH_max per mitochondrial volume and MCI revealed four distinct clusters of DIAm fibers (Fig. 8B). In type I and IIa fibers, one cluster was formed by fibers from the C₂SH right (lesioned) side (centroid: x = −1.21, y = −0.16) and another from type I and IIa fibers of all other groups (centroid: x = 0.99, y = 0.98; Fig. 7B). These clusters were distinct from two separate clusters of type IIx/IIb fibers, similarly, differentiated between the C₂SH right (lesioned) side (centroid: x = −1.31, y = −1.28) and IIx/IIb fibers from all other groups (centroid: x = −0.37, y = −1.26; Fig. 8B).

DISCUSSION

The novel findings of the present study indicate that 2 wk of DIAm inactivity imposed by C₂SH is associated with a decrease in MVD and mitochondrial fragmentation in all fiber types but is more pronounced in type I and IIa fibers. C₂SH-induced DIAm inactivity also reduced maximum respiratory capacity (SDH_max) in type I and IIa fibers. The reduction in SDH_max was apparent whether normalized to fiber volume or MVD, indicating robust changes in overall mitochondrial function. The MVD, MCI, and SDH_max were higher in type I and IIa fibers compared with type IIx/IIb, consistent with our previous work (10, 18). This is consistent with the constant rhythmic activity of type I and IIa fibers with their recruitment to support breathing whereas type IIx/IIb fibers are less frequently activated for expulsive behaviors. In support, C₂SH-induced inactivity predominantly affected MVD and SDH_max in the type I and IIa DIAm fibers.

C₂SH Unilaterally Disrupts Inspiratory-Related Diaphragm Muscle Activity

The descending glutamatergic synaptic inputs responsible for inspiratory drive to phrenic motor neurons and consequently DIAm motor unit recruitment and muscle activation are primarily ipsilateral (7, 55, 56). The ipsilateral excitatory input to phrenic motor neurons is predominantly disrupted by C₂SH resulting in inactivation of descending inspiratory drive. However, we and others have shown that spontaneous recovery of inspiratory-related DIAm activity can occur over time after C₂SH (33, 54, 57–62), likely due to the strengthening of latent contralateral descending inputs to phrenic motor neurons, as well as enhanced strength of local or ascending interneuronal circuitry (63–66). Consistent with this neuroplasticity, we found that inspiratory-related DIAm EMG activity recovered in two of eight animals after C₂SH. Importantly, these animals were not included in our analysis of DIAm inactivity. In agreement with the relatively selective effect of C₂SH on inspiratory-related DIAm activation, we recently reported that C₂SH predominantly disrupts glutamatergic synaptic input to smaller phrenic motor neurons (67), which comprise type S or FR motor units (type I and IIa fibers, respectively) that are recruited for ventilatory behaviors (1, 26, 67–70). On the intact contralateral side of the DIAm, there was a transient increase in inspiratory-related DIAm activity, as well as an increase in inspiratory duration and duty cycle immediately following C₂SH. Notably, these changes in inspiratory duration and duty cycle were transient, returning to sham levels by 3 days postsurgery. The transient increase in inspiratory activity of the contralateral side of the DIAm did not affect mitochondrial structure and function.

No Effect of Inactivity on DIAm Fiber-Type Proportions or Cross-Sectional Areas

Similar to previous studies, we found that 2 wk of inactivity did not affect the relative proportions of DIAm fibers or fiber CSAs (33, 54). In a previous study, we did find a small, but significant decrease in the CSA of type IIx/IIb DIAm fibers following 6 wk after C₂SH (54).

Effect of Inactivity on Mitochondrial Structure in the DIAm

In preliminary studies, we compared MitoTracker Green and MitoTracker Red labeling of mitochondria in DIAm fibers and found no difference in MVD or MCI measurements. Thus, MitoTracker Green was used to be consistent with our prior studies (10, 18). The MVD of DIAm fibers, measured by confocal imaging in the present study, was consistent with our previous reports using the same technique (10, 18) as well as with results from our laboratory using EM (6). We found that type I and IIa DIAm fibers have greater MVD than type IIx/IIb fibers, similar to past observations (10, 18), with no differences between the right and left sides of the DIAm in sham controls. Following C₂SH, the MVD of DIAm fibers on the left (active) side was similar to comparable DIAm fiber types in the sham controls. Following 2 wk of inactivity imposed by C₂SH, MVD was reduced across all fiber types on the right (inactive) side compared with the intact left side. The relative decrease in MVD in the inactive right side of the DIAm was similar across all fiber types (∼40%).

We investigated the molecular mechanisms underlying mitochondrial remodeling by measuring protein expression of Parkin (mitophagy) and PGC1α (mitochondrial biogenesis) and mitochondrial DNA copy number (mitochondrial biogenesis). Following 2 wk of inactivity, Parkin expression increased on right side of the DIAm, whereas PGC1 α and mitochondrial DNA copy number decreased compared with the intact active left side of the DIAm. These results indicate that the reduction in MVD in DIAm fibers induced by inactivity results from both increased mitophagy and decreased mitochondrial biogenesis.

Mitochondria in type I and IIa DIAm fibers were more filamentous (i.e., higher MCIs) compared with type IIx/IIb fibers. Following 2 wk of inactivity imposed by C₂SH, mitochondria became more fragmented (i.e., reduced MCIs) across all DIAm fiber types. However, the effect of C₂SH on the MCI was disproportionately greater in type I and IIa fibers compared with type IIx/IIb fibers, indicating a greater impact of inactivity on type I and IIa DIAm fibers. Consistent with the C₂SH-induced mitochondrial fragmentation, DRP1 protein expression was higher on the right side of DIAm compared with the left side, with unchanged MFN2 expression. Importantly, mitochondrial morphology may affect the respiratory capacity of different cell types (38, 44). In human airway smooth muscle cells, maximum mitochondrial OCR was higher in more filamentous mitochondria (38). In addition, highly active tissues such as DIAm have mitochondria that are more filamentous (43, 71), consistent with the results of the present study. Filamentous mitochondria have a higher surface-to-volume ratio (MCI), which allows for a greater inner mitochondrial membrane surface area. Therefore, a greater inner mitochondrial membrane surface area would allow for increased O₂ consumption and oxidative phosphorylation.

Effect of Inactivity on Mitochondrial Respiratory Capacity in DIAm Fibers

We measured the respiratory capacity of individual DIAm fibers by determining the maximum velocity of the SDH reaction (21, 22). Succinate dehydrogenase, located in the inner mitochondrial matrix, is a key enzyme of the TCA cycle as well as complex II of the ETC. Measurements of SDH_max when normalized for mitochondrial volume provide an assessment of the maximum respiratory capacity of mitochondria. To ensure that SDH_max truly reflected maximum respiratory capacity, we validated the rate of reaction by using FCCP to disrupt the proton gradient, and second by treatment with rotenone or antimycin A to inhibit complex I or III of the ETC, respectively (10). Our method is an improvement on similar approaches used in respirometry systems to measure maximum OCR (45, 46), as they can only assess suspended cells or tissue, not single cells. Other studies have used respirometry to measure mitochondrial OCR in muscle fiber bundles, but not single fibers (72–74). In addition, mitochondrial OCR has not been normalized for MVD but instead uses estimations of total cell number, total protein content, mitochondrial protein, or DNA concentration (72–74). A major advantage of our technique is that SDH_max and MVD can be measured in the same fibers. Importantly, evidence suggests that the respiratory capacity of mitochondria is affected by volume density and morphological changes (10, 19, 20).

Consistent with past reports in the rat DIAm (10, 18, 23, 24, 32, 75), SDH_max per fiber volume was found to be greater in type I and IIa DIAm fibers compared with IIx/IIb. The higher SDH_max per fiber volume of type I and IIa DIAm fibers is consistent with their increased levels of activity related to breathing (1, 2, 7) and their higher MVDs (6, 10, 18). The lower SDH_max per fiber volume of type IIx/IIb DIAm fibers is consistent with their infrequent activation during expulsive behaviors (1, 2, 7) and their lower MVDs (6, 10, 18). Following 2 wk of inactivity imposed by C₂SH, SDH_max per fiber volume in type I and IIa DIAm was significantly reduced compared with similar fiber types in the left, intact side of the DIAm in the same animals as well as type I and IIa DIAm fibers in the sham controls. There were no significant differences in SDH_max per fiber volume in type IIx/IIb fibers between the right inactive and left active sides in the C₂SH animals or between C₂SH and sham controls. In the present study, there was a relationship between SDH_max per fiber volume and mitochondrial volume density. In sham control animals and on the left (intact) side of the DIAm in C₂SH animals, there were two distinct populations of DIAm fibers with type I and IIa fibers forming one group with higher SDH_max per fiber volume and higher MVDs compared with type IIx/IIb fibers forming the second group. This relationship was impacted by C₂SH with the type I and IIa fibers on the right inactive side of the DIAm having lower SDH_max per fiber volume and MVDs than their active counterparts. Type I and IIa fibers in the right (inactive) side of the DIAm still maintained a higher SDH_max per fiber volume and mitochondrial volume than type IIx/IIb fibers. These results indicate that SDH_max and MVD are interrelated and dependent on activity, especially in the more active type I and IIa DIAm fibers.

Similar to a previous study (10), we found that SDH_max per mitochondrial volume was significantly higher in type I and IIa fibers as compared with type IIx/IIb fibers. In the present study, we found that following 2 wk of inactivity imposed by C₂SH, SDH_max per mitochondrial volume in type I and IIa DIAm was significantly reduced in the inactive right side of the DIAm compared with the left active side in the same animals. The SDH_max per mitochondrial volume of type I and IIa fibers was also lower in the right inactive side of the DIAm when compared with the same fiber type in sham controls. Importantly, our results indicate a relationship between the SDH_max per mitochondrial volume and mitochondrial morphology (MCI), with two distinct groups showing type I and IIa DIAm fibers comprising one group with higher SDH_max per mitochondrial volume and higher MCI, whereas type IIx/IIb fibers comprised a second group with lower SDH_max per mitochondrial volume and lower MCI. Following 2 wk of inactivity imposed by C₂SH, two distinct clusters were still apparent but with the SDH_max per mitochondrial volume and MCI reduced in type I and IIa fibers in the inactive right side of the DIAm.

Clinical Significance and Conclusions

The results of the present study are applicable to mechanically ventilated patients, where DIAm activity is reduced. Mechanical ventilation is often lifesaving and necessary, but the prolonged inactivity of the DIAm during mechanical ventilation can lead to long-term problems, including an inability to wean from the ventilator. Decreased mitochondrial respiratory capacity in type I and IIa fibers that support breathing could underlie this failure. Type I and IIa fibers are highly active, and per the results of the present study, are more likely to be acutely affected by inactivity, such as mechanical ventilation.

Overall, 14 days of DIAm inactivity imposed by C₂SH caused reduced MVD, mitochondrial fragmentation, and a concomitant reduction of SDH_max in type I and IIa DIAm fibers on the lesioned side. Type I and IIa DIAm fibers were far more sensitive to inactivation than type IIx/IIb fibers, which exhibited little functional deficits. Our novel assessment of mitochondrial morphology and function within the same DIAm fibers indicates remarkable plasticity in response to varying levels of activity and suggests that the effects of 14-day C₂SH are primarily on the phrenic motor neurons that innervate type I or IIa DIAm fibers.

GRANTS

This work was supported by the National Institutes of Health Grants R01-AG044615 (to G.C.S. and C.B.M.), R01-HL146114 (to G.C.S. and C.B.M.), and T32-HL105355 (to A.D.B.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.D.B., M.J.F., C.B.M., and G.C.S. conceived and designed research; A.D.B., M.J.F., L.A.D., D.D., C.B.M., and G.C.S. performed experiments; A.D.B., M.J.F., L.A.D., D.D., C.B.M., and G.C.S. analyzed data; A.D.B., M.J.F., L.A.D., D.D., C.B.M., and G.C.S. interpreted results of experiments; A.D.B., M.J.F., D.D., C.B.M., and G.C.S. prepared figures; A.D.B., M.J.F., C.B.M., and G.C.S. drafted manuscript; A.D.B., M.J.F., L.A.D., D.D., C.B.M., and G.C.S. edited and revised manuscript; A.D.B., M.J.F., L.A.D., D.D., C.B.M., and G.C.S. approved final version of manuscript.