SESN2 prevents the slow-to-fast myofiber shift in denervated atrophy via AMPK/PGC-1α pathway

Sestrin2 (SESN2), a stress-inducible protein, has been reported to protect against denervated muscle atrophy through unfolded protein response and mitophagy, while its role in myofiber type transition remains unknown. A mouse sciatic nerve transection model was created to evaluate denervated muscle atrophy. Myofiber type transition was confirmed by western blot, fluorescence staining, ATP quantification, and metabolic enzyme activity analysis. Adeno-associated virus (AAV) was adopted to achieve SESN2 knockdown and overexpression in gastrocnemius. AMPK/PGC-1α signal was detected by western blot and activated with 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR). C2C12 myotubes with rotenone treatment were adopted for in vitro experiments. SESN2 was found to be upregulated in denervated skeletal muscles and rotenone-treated C2C12 cells. Knockdown of SESN2 aggravated muscle atrophy and accelerated myofiber type transition from slow-twitch to fast-twitch. Moreover, AMPK/PGC-1α signaling was proven to be activated by SESN2 after denervation, which further induced the expression of hypoxia-inducible factor HIF2α. Exogenous activation of AMPK/PGC-1α signaling could counteract the addition of slow-to-fast myofiber shift caused by SESN2 knockdown and lead to the retainment of muscle mass after denervation. Collectively, the present study indicates that SESN2 prevents myofiber type transition from slow-twitch to fast-twitch and preserves muscle mass in denervated atrophy via AMPK/PGC-1α signaling. These findings contribute to a better understanding of the pathogenesis of muscle atrophy and provide novel insights into the role of SESN2 in myofiber type transition.

surgery on the right hind legs as delineated previously [16]. The mice were then randomly divided into four groups (six mice per group): sham operation group (control); denervation group; denervation + AAV-shSESN2 (adeno-associated virus, U6-MCS-CAG-EGFP; Genochem, China) group and denervation + AAV-SESN2 group (CMV-betaGlobin-MCS-SV40 PolyA, Genochem). The sequence of shSESN2 primers was 5′-GCG TCT TTG GCA TCA GAT ACG-3′. AAV9 injections of gastrocnemius (GAS) were completed 3 weeks in advance to construct SESN2 knockout or overexpression models. In particular, 10 µl of virus (1.0 × 10 12 viral genomes (vg)/ml) were injected into each point around the gastrocnemius. To prevent the backflow of viral particles, the syringe was left in place for an extra 5 min after injection. Four to six injections were conducted on every limb. As for preparation of AMPK activation models, AICAR (an AMPK activator; MedChemExpress, USA, HY-13417) was administered to the mice via an intraperitoneal injection at a dose of 500 mg/kg once a day for 7 days prior to operation. An additional 7 days of AICAR treatment was performed until muscles were collected. Mice were euthanized by cervical dislocation at the indicated time, and GAS, extensor digitorum longus (EDL), and soleus (SOL) were removed, weighed, and frozen for the next experiments. For cervical dislocation, a trained person grasped the skin on the back of the neck by the thumb and forefinger and immediately pulled on the base of the tail in an opposite upward direction from the head. The immediate dislocation of the spinal column from the brain ensured death within a few seconds.
All animal experiments were completed as per the guidelines of the Chinese National Institutes of Health and obtained the approval of the Ethical Committee on Animal Experiments (Huazhong University of Science and Technology, no. 2021-S2789).

Muscle mass measurements, H&E staining, and fiber diameter quantification
Muscles (GAS, EDL, and SOL) of the denervated and control sides were collected and weighed at the proper temporal points. The wet weight ratio (muscular weight of the denervated side divided by the weight of the control side) was utilized to assess muscle atrophy.
For hematoxylin and eosin (H&E) staining, muscle samples were subjected to fixation in paraformaldehyde (4%) for 24 h, and afterwards subjected to dehydration and paraffin embedment. The 4-μm cross-cutting muscle sections were acquired before H&E staining (Bioyear, PRC) as per the instructions.
For quantitative analysis of fiber diameters, muscular transection was stained for wheat germ agglutinin (WGA) to visualize the fibers. ImageJ program was leveraged to quantify the minimum Feret's diameter of myofibers stochastically selected within each group.
In short, the protein was abstracted via RIPA lysis buffering solution to which 1% protease suppressor was added (Roche, USA, 11206893001). The same quantity of protein (10-50 μg) was isolated via 10% SDS-PAGE, and afterwards moved onto nitrocellulose films (Merck Millipore, USA, Z358657). The films were subjected to blockade with 5% w/v BSA prior to cultivation with the first antibodies under 4 °C overnight. Subsequently, the films were cultivated with second antibodies conjugated with antirabbit or antimouse IgG (Abcam) for 60 min under ambient temperature and visualized via the Immobilon ECL matrix tool (Merck Millipore, WBKLS0050).

Immunofluorescence
Paraffin-embedded GAS underwent cross-sectional cuts, which were made into 4-μm sections as previously stated [16]. Following antigen retrieval, permeabilization, and goat serum blocking, primary antibody (mouse anti-slow MHC antibody and rabbit-antifast MHC antibody; Abcam) incubation was implemented at 4 °C. Next, samples were stained for 1 h using secondary Alexa Fluor 488-conjugated goat anti-mouse and Alexa Fluor 647-conjugated goat anti-rabbit IgG antibodies (Abcam, ab150113 and ab150079), followed by 5 min DAPI/PI (Sigma, USA, D9542) staining. The samples were then imaged with a fluorescence microscope. Similar processes were oriented with C2C12 cells.

ATP level
ATP contents in mouse GAS were identified via ATP analysis tool (Beyotime, China, S0026). Briefly, GAS tissue was homogenized in ATP assay lysis buffering solution, and the supernatant was harvested via centrifugation at 12,000g for 300 s under 4 °C and subjected to quantification via BCA analysis. Subsequently, 100 μl ATP identification reagent was supplemented into 100 μl supernatant, and the firefly luciferase (FFL) activity was identified and studied via luminescent spectrometric analysis (EnSpire, USA). The ATP content was normalized to cell protein level and presented as proportion (%) in contrast to the controls. Similar procedures were performed in C2C12 cells.

Metabolic enzyme activity analysis
The activities of LDH and SDH were identified via commercially available tools bought from Nanjing Jiancheng Biological Engineering Institution (China, A020-1-2 and A022-1-1) as per the supplier's specification. Specific enzyme activity was presented as U/mg protein.

Cell culture and transfection
C2C12 cells were obtained from iCell Bioscience Inc. and cultivated in DMEM (Gibco, USA) including 10% FBS (Gibco), 100 units/mL penicillin, and 100 μg/mL streptomycin solution (Sangon Biotech, China) in 5% CO 2 at 37 ℃ in a cell culture incubator. Cells were differentiated into myotubes posterior to the cultivation with differentiation medium (2% horse serum in DMEM) for 1 week as delineated previously [11]. For genetic knockout, three short interference RNAs for murine SESN2 (siSESN2) and negative controls (siNC) were synthesized by RiboBio Company (PRC). The transfection process was completed as per the specification of Lipofectamine 2000 (Invitrogen, USA, 11668019). Cells were collected at 48 h posterior to the transfection process for the subsequent assays. siRNAs with maximal suppressive effect were utilized for assays later on. The siRNA sequences are presented in Additional file 1: Table S1.

Detection of mitochondrial ROS level
Identification of mitochondrial ROS (mtROS) was completed via the fluorogenic stain MitoSOX Red (Thermo Fisher, USA, M36008), which was targeted to mitochondria in alive cells. In short, 5 mM stock solution and 5 μM work solution of MitoSOX were produced as per the supplier's specification. Cells were cultivated in work solution for 600 s without light under 37 °C and afterwards cleaned three times in a warm buffering solution. The pictures of alive cells were collected via a confocal laser scan microscope. Finally, more than 30 cells in every group were chosen stochastically to produce the eventual outcomes.

Statistical analysis
Statistical analysis of three independent experiments with at least three technical repetitions was conducted using GraphPad Prism 9.0.0 (GraphPad, USA), and the statistical data were expressed as mean ± standard deviation. Differences were evaluated by oneway or two-way analysis of variance (ANOVA) with Tukey post-hoc test. P-value less than 0.05 was considered statistically significant.

Denervation led to skeletal muscle atrophy and slow-to-fast myofiber shift
To reveal the development of denervation-induced muscle atrophy, a murine sciatic nerve cross-section model was constructed and diverse types of skeletal muscle were studied. As described previously [15], muscular mass of GAS (mixed-type muscle), extensor digitorum longus (EDL; fast-twitch muscles involving type IIb and IIx myofibers), and soleus (SOL; slow-twitch muscles involving type I and IIa myofibers) reduced quickly upon denervation. In this process, muscle atrophy exhibited a representative two-phase characteristic with rapid weight loss (approximately 47%) in 14 days at the beginning and afterwards a slighter reduction (approximately 14%) in the following 14 days (Fig. 1A). As a vital structure protein in myofiber, MHC was identified via western blot and unveiled a similar change as reported previously (Fig. 1B). Later, H&E staining and WGA immunofluorescence were performed to quantify fiber diameter, revealing that denervation drove the myofibers to decrease with a progressively smaller diameter (Fig. 1C, D). Overall, skeletal muscles, irrespective of muscle type, atrophied quickly upon denervation.
Muscle atrophy is usually accompanied by myofiber type transition. As seen in Fig. 1E, a significant decrease of slow MHC, MHC IIa, myoglobin, and troponin I-ss was observed over time after denervation. In contrast, fast MHC and MHC IIb, meanwhile, increased dramatically. Immunofluorescence analysis indicated similar results, showing decreased percentage of slow oxidative fibers and increased percentage of fast glycolytic fibers (Fig. 1F). As slow oxidative fibers contain more mitochondria than fast glycolytic fibers, such changes further led to reduced mitochondria number and ATP content (Fig. 1E, G, H). In addition, the activity of SDH and LDH, which were used as markers of oxidative and glycolytic capacities, was further detected. Obviously, denervation led to Fig. 1 Denervation led to skeletal muscle atrophy and slow-to-fast myofiber shift. A Wet weight ratio (the weight of the operational side divided by the contralateral side) of GAS, EDL, and SOL at indicated timepoints post-denervation. B Western blot analysis of the dynamic changes of MHC protein expression after denervation. C, D Analysis of gastrocnemius fibers diameter by H&E and immunofluorescence staining. Scale bar, 50 μm. E Western blot analysis of proteins related to muscle fiber type. F Immunofluorescence analysis of the fiber type composition in denervated GAS. Scale bar, 50 μm. G Microstructure of GAS. Red arrows indicate mitochondria. Scale bar, 2 μm. H, I Relative ATP content and metabolic enzyme activity (LDH and SDH) in GAS at different timepoints post-denervation. Data presented as mean ± standard deviation (SD). n = 6. *P < 0.05 versus control (week 0) attenuated SDH activity and enhanced LDH activity (Fig. 1I). Overall, denervation promoted myofiber type transition from slow-twitch to fast-twitch.

SESN2 protected against muscle atrophy and myofiber type transition after denervation
In this section, we explore the role of SESN2 in denervated muscle atrophy and myofiber type transition. Firstly, evident accumulation of SESN2 protein was identified in the 14 days at the beginning after denervation, before a progress decrease in the following 14 days (Fig. 2A). Given that the progression of muscle atrophy occurred mainly in the first 14 days in a coincidental manner, we speculated that SESN2 might participate in the process of muscle atrophy. Subsequently, SESN2 knockdown and overexpression models were produced to explore the participation of SESN2 in the muscle atrophy of denervated GAS. The expression of SESN2 was verified via western blot (Fig. 2B). GAS specimens were harvested 14 days after denervation. MHC expression and muscle mass evaluation unveiled that muscle atrophy was remarkably deteriorated posterior to SESN2 knockdown and was mildly reversed via SESN2 overexpression (not significant) (Fig. 2B, C). Analysis of fiber size revealed that the diameter of fibers decreased in denervated GAS and deteriorated by AAV-shSESN2 (Fig. 2D). Finally, the effect of SESN2 on myofiber type transition was explored. As seen in Fig. 2E, F, knockdown of SESN2 significantly accelerated the reduction of slow oxidative fiber and addition of fast glycolytic fiber. Overexpression of SESN2 in denervated GAS exerted no obvious effect on myofiber type transition. ATP content and enzyme activity analysis revealed a trend consistent with the above data (Fig. 2G, H). Together, the above results indicate that SESN2 protected against muscle atrophy and the slow-to-fast myofiber type shift upon denervation.

Rotenone induced slow-to-fast fiber type shift in C2C12 myotubes
To replicate the denervated condition in vitro, C2C12 cells were treated with a concentration gradient of rotenone (a mitochondrial electron transport chain complex I inhibitor). As expected, obvious accumulation of mtROS was observed in C2C12 cells treated with 100 nM or more rotenone (Fig. 3A). SESN2 was also induced by rotenone treatment (Fig. 3C). More importantly, rotenone treatment induced a significant reduction of slow MHC-positive myotubes and addition of fast MHC-positive myotubes (Fig. 3B). Analysis of protein expression, ATP content, and enzyme activity also revealed a consistent trend (Fig. 3C-E). Briefly, these data indicate that rotenone treatment could induce the slow-to-fast fiber type shift in C2C12 myotubes, which was similar to the in vivo denervation model. In the following experiments, 100 nM rotenone was adopted.

Knockdown of SESN2 accelerated slow-to-fast fiber type shift in rotenone-treated C2C12 myotubes
To further confirm the role of SESN2 in myofiber type transition in vitro, knockdown of SESN2 was achieved in rotenone-treated C2C12 cells. The inhibitory effect of siS-ESN2 was confirmed by western blot (Fig. 4A). Immunofluorescence analysis revealed that siSESN2 treatment significantly accelerated the rotenone-induced reduction of slow MHC-positive myotubes and addition of fast MHC-positive myotubes (Fig. 4B). The results of western blot indicated that knockdown of SESN2 further suppressed the expressions of slow oxidative fiber-related proteins (slow MHC, MHC IIa, myoglobin, and troponin I-ss) and enhanced the expression of fast glycolytic fiber-related proteins (fast MHC, MHC IIb) (Fig. 4C). ATP content and enzyme activity detection displayed similar changes (Fig. 4D, E). These data prove that SESN2 was required to prevent slowto-fast fiber type shift in rotenone-treated C2C12 myotubes. and MHC. C Appearance and wet weight ratio of GAS in different groups. D Quantification of muscle fiber diameter by immunofluorescence staining of WGA. Scale bar, 50 μm. E Western blot analysis of proteins related to muscle fiber type. F Immunofluorescence analysis of fiber type composition in denervated GAS. Scale bar, 50 μm. G, H Relative ATP content and metabolic enzyme activities (LDH and SDH) in GAS. Data presented as mean ± SD. n = 6. *P < 0.05 versus control (week 0 or sham operation group). # P < 0.05 versus denervation group. Den denervation, Con control

SESN2 prevented slow-to-fast shift in rotenone-treated C2C12 myotubes via AMPK/PGC-1α
pathway Given the vital role of AMPK/PGC-1α signaling in myofiber type transition, we wonder whether this signal was working downstream SESN2. As seen in Fig. 5A, rotenone treatment activated and induced expression of PGC-1α and HIF-2α, while knockdown of SESN2 reversed the above variation, confirming the regulatory effect of SESN2 on AMPK/PGC-1α signaling. To further clarify the role of AMPK/PGC-1α in myofiber type transition, C2C12 myotubes were pretreated with AICAR, an AMPK activator, before rotenone. It was seen that AICAR successfully activated AMPK and downstream PGC-1α and HIF-2α. Besides, AICAR treatment reversed the effect of rotenone, showing increased slow oxidative fiber-related proteins (slow MHC, MHC IIa, myoglobin, and troponin I-ss) and decreased fast glycolytic fiber-related proteins (fast MHC, MHC IIb) (Fig. 5B). Consistently, AICAR treatment reversed the rotenone-induced reduction of slow MHC-positive myotubes and addition of fast MHC-positive myotubes (Fig. 5C). This trend was further confirmed in the detection of ATP content and enzyme activity (Fig. 5D, E). To sum up, AMPK/PGC-1α signaling mediated the inhibitory effect of SESN2 on myofiber type transition.

SESN2 activated AMPK/PGC-1α signaling in denervated muscle
To investigate the relationship between SESN2 and AMPK/PGC-1α signaling in denervated atrophy, dynamic monitoring of protein content was performed. Similar to SESN2, AMPK was activated in the first 2 weeks after denervation, while the expression of PGC-1α and downstream HIF-2α reached a peak 1 week post-denervation and fell down rapidly afterwards (Fig. 6A). As AMPK/PGC-1α signaling is well documented in myofiber type transition, we speculated that this variation might be an endogenous protective response in the early stage of denervated muscle atrophy. Subsequently, GAS with SESN2 knockdown was isolated 1 week post-denervation for western blot. As seen in Fig. 6B, knockdown of SESN2 suppressed the activation of AMPK/PGC-1α in the early stage of denervation. Together, the above data prove that the AMPK/PGC-1α signaling was activated by SESN2 in denervated muscle.

AMPK/PGC-1α signaling prevented slow-to-fast myofiber shift and protected against muscle atrophy upon denervation
Here we wonder whether AMPK/PGC-1α signaling played a role in myofiber type transition. Western blot analysis of GAS isolated 1 week post-denervation indicated that knockdown of SESN2 significantly accelerated the decrease of slow oxidative fiberrelated proteins (slow MHC, MHC IIa, myoglobin, and troponin I-ss) and increase of fast glycolytic fiber-related proteins (fast MHC, MHC IIb) induced by denervation, while the AICAR treatment made up for the lack of SESN2 (Fig. 7A). Consistently, knockdown of SESN2 led to a further reduction of slow oxidative fiber and addition of fast glycolytic fiber compared with the denervation + shNC group, while AICAR treatment reversed this trend induced by SESN2 knockdown (Fig. 7B). Detection of ATP content and enzyme activity also consolidated the effect of AICAR (Fig. 7C, D). It needs to be emphasized that the analysis of fiber type, ATP content, and enzyme activity was performed in GAS 2 weeks post-denervation given that muscle atrophy happens more Scale bar, 50 μm. D, E Detection of relative ATP content and metabolic enzyme activity (LDH and SDH) in C2C12 myotubes. Data presented as mean ± SD. *P < 0.05 versus control or siNC group. # P < 0.05 versus rotenone treatment group. Rot rotenone, Con control slowly than signal change. In short, these results demonstrate that AMPK/PGC-1α signaling attenuated the slow-to-fast fiber shift in denervated muscle.
In addition to the role of AMPK/PGC-1α signaling in myofiber type transition, its effect on muscle atrophy was further clarified. It was obvious that knockdown of SESN2 further aggravated the denervation-induced weight loss, MHC degradation, and fiber diameter reduction. Interestingly, the addition of AICAR reversed all these effects and achieved a similar degree of atrophy to that of the denervation + shNC group (Fig. 7E-G). The above results indicate that AMPK/PGC-1α signaling protected against muscle atrophy upon denervation.

Discussion
According to previous research [17,18], SESN2 is constantly triggered upon various stresses, and exhibits cytoprotection features via regulating the synthetic process of proteins, autophagic activity, and apoptosis. In our previous research, ERS and mitochondrial dysfunction were found in denervated muscle, while the endogenous expression of SESN2 in the early stage of denervation mediated unfolding protein response and mitophagy and ultimately alleviated muscle atrophy [15]. Given that myofiber type transition is a main feature of skeletal muscle atrophy, we wonder whether SESN2 was involved therein. As expected, the present study indicated that SESN2 was induced soon after denervation and subsequently activated the AMPK/PGC-1α signal, which was shown to prevent the slow-to-fast myofiber type shift in denervated muscle. This finding provides new insight into the function of SESN2 and extends our understanding of the process of myofiber type transition post-denervation.
In the past few years, myofiber type transition was observed and studied mainly in sports medicine, in which the PGC-1α signal was identified as the key regulator therein [19,20]. As a versatile transcription co-activator, PGC-1α participates in various cell activities, including adaptive calorigenesis, fatty acid oxidation, gluconeogenesis, and mitochondrial biogenesis [21]. In skeletal muscle, PGC-1α is highly expressed in slow oxidation fibers rather than fast glycolytic  AMPK/PGC-1α signaling prevented slow-to-fast myofiber shift and protected against muscle atrophy upon denervation. A Western blot analysis confirmed the role of AMPK/PGC-1α signaling in myofiber type transition in denervated GAS. B Immunofluorescence analysis of fiber type composition in denervated GAS. Scale bar, 50 μm. C, D Relative ATP content and metabolic enzyme activities (LDH and SDH) in GAS with different treatment. E Appearance and wet weight ratio of GAS in different groups. F Western blot analysis confirmed the expressions of SESN2 and MHC. G Quantification of muscle fibers diameter by immunofluorescence staining of WGA. Scale bar, 50 μm. Data presented as mean ± SD. n = 6. *P < 0.05 versus denervation + shNC group. # P < 0.05 versus denervation + shSESN2 group. Den denervation fibers [22]. Numerous studies have confirmed a close association of muscular contraction activity with elevated expressing level of PGC-1α [23,24]. Endurance exercise triggers the expression of PGC-1α mRNA and protein in rats and humans. Moreover, overexpression of PGC-1α can reinforce mitochondria biogenesis and facilitate the fast-to-slow myofiber transition in cultivated myoblasts and transgenic mice, which improves exercise performance [25]. However, the mechanism underlying myofiber fiber transition in denervated muscle was rarely reported. In the present research, a new role of PGC-1α was identified, where PGC-1α acted as an endogenous protective molecular against the slow-to-fast fiber shift and ultimately delayed the process of denervated muscle atrophy.
HIF-2α, as a member of the the Per-ARNT-Sim-bHLH family of transcription factors that regulate the cellular response to hypoxic conditions [26,27], was also detected in the present study. According to previous research, HIF-2α acts as a key mediator of PGC-1α-dependent myofiber type transition, a process that is partially muscle selective [22]. In our research, a similar expression pattern of HIF-2α and PGC-1α was observed in denervated GAS and C2C12 myotubes. Moreover, indirect upregulation of PGC-1α and HIF-2α prevented the slow-to-fast myofiber type shift and protected against muscle atrophy. Consistent with previous research, we speculate that PGC-1α regulation of myofiber type transition in denervated muscle is HIF-2α dependent.
In the dynamic monitoring of AMPK/PGC-1α signaling, we noticed a mismatch in the expression of AMPK and downstream PGC-1α and HIF-2α, as described in Fig. 6A. Given that the SESN2-AMPK pathway has been widely reported to activate autophagy and suppress protein synthesis [28][29][30], we speculate that the regulation of SESN2-AMPK pathway on PGC-1α expression is biphasic, with a promotion effect at early stage and a suppression effect afterward. Of course, other mechanisms might also be involved in PGC-1α regulation therein and worthy of further research.
Considering that different muscles contain different fibers and may respond differently to denervation [31], it will be difficult to apply this new regulatory mechanism to muscles other than GAS. As reported in previous research, type I myofibers of rat soleus underwent remarkable atrophy upon denervation, while type I myofibers in EDL maintained basically the same size in the first 14 days posterior to sciatic nerve section. A similar diversity was identified in type IIa myofibers, which underwent evident atrophy in soleus but merely minor atrophy in EDL. In rat diaphragm, denervation was similarly discovered to induce atrophy of type IIx and IIb myofibers, and showed no variation in type IIa and hypertrophy of type I myofibers [32]. However, the present study was meaningful in that it identified potential molecular pathological mechanisms of myofiber type transition upon denervation, which were previously unknown and out of the spotlight.
Here we noticed that overexpression of SESN2 made no difference to myofiber type transition and muscle atrophy upon denervation. As stated before, we speculate that the induction of SESN2 in denervated gastrocnemius is an adaptive response that critically mediates skeletal muscle adaptation to denervation by preventing the slow-to-fast myofiber shift, and is not enough to counteract the denervation-induced muscle atrophy.
As the atrophy of skeletal muscle upon denervation is a complicated process, no suitable in vitro model has yet been designed capable of replicating such a denervated state. Given that mitochondrial dysfunction and ROS overproduction were considered as important