DHHC21 deficiency attenuates renal dysfunction during septic injury

Renal dysfunction is one of the most common complications of septic injury. One critical contributor to septic injury-induced renal dysfunction is renal vascular dysfunction. Protein palmitoylation serves as a novel regulator of vascular function. Here, we examined whether palmitoyl acyltransferase (PAT)-DHHC21 contributes to septic injury-induced renal dysfunction through regulating renal hemodynamics. Multispectral optoacoustic imaging showed that cecal ligation and puncture (CLP)-induced septic injury caused impaired renal excretion, which was improved in DHHC21 functional deficient (Zdhhc21dep/dep) mice. DHHC21 deficiency attenuated CLP-induced renal pathology, characterized by tissue structural damage and circulating injury markers. Importantly, DHHC21 loss-of-function led to better-preserved renal perfusion and oxygen saturation after CLP. The CLP-caused reduction in renal blood flow was also ameliorated in Zdhhc21dep/dep mice. Next, CLP promoted the palmitoylation of vascular α1-adrenergic receptor (α1AR) and the activation of its downstream effector ERK, which were blunted in Zdhhc21dep/dep mice. Vasoreactivity analysis revealed that renal arteries from Zdhhc21dep/dep mice displayed reduced constriction response to α1AR agonist phenylephrine compared to those from wild-type mice. Consistently, inhibiting PATs with 2-bromopalmitate caused a blunted vasoconstriction response to phenylephrine in small arteries isolated from human kidneys. Therefore, DHHC21 contributes to impaired renal perfusion and function during septic injury via promoting α1AR palmitoylation-associated vasoconstriction.

Septic injury often progresses into multiple organ dysfunction, serving as a leading cause of mortality in the intensive care units 1,2 . Kidneys are one of the most susceptible organs to septic injury. Patients with septic injury-induced renal dysfunction tend to have prolonged hospital-stays, worse outcomes, and significantly higher mortality 1 . The pathogenesis of septic injury-associated renal dysfunction is a complex and multifactorial process; one prevailing mechanism is renal vascular dysfunction [3][4][5] . Studies of patients and large animals with septic injury-induced renal dysfunction have shown significantly increased renal vascular resistance 6,7 , contributing to decreased renal blood flow (RBF) in septic injury [7][8][9] . This reduction of RBF has been confirmed in sepsis patients with kidney injury using different techniques including renal Doppler, thermodilution and phase contrast magnetic resonance imaging 10,11 . Regional ischemia and hypoxia caused by insufficient renal tissue perfusion can result in impaired glomerular filtration coupled with tubular epithelial cell death, leading to renal dysfunction [12][13][14] .
Mechanisms responsible for renal tissue hypoperfusion in septic injury have not yet been fully elucidated. Protein palmitoylation, the covalent attachment of a 16-carbon palmitic acid to the cysteine residue(s) of a protein, has recently been recognized as a novel regulator of vascular function 15,16 . The attached palmitate increases the hydrophobicity of the modified proteins in blood vessels and affects their localization, stability, and conformation, thereby regulating many aspects of vascular function at posttranslational level [17][18][19] . This reversible modification is catalyzed by a family of palmitoyl acyltransferases (PATs) containing the Asp-His-His-Cys (DHHC) motif 20,21 . So far, 23 PAT-DHHCs have been identified in humans 21 . Recent studies have reported the involvement of PAT-DHHCs in hemodynamic regulation, peripheral arterial disease, and inflammation-induced vascular injury 15,16,22 ; however, there is lack of information regarding the roles of PATs in regulating renal vascular function in septic injury. www.nature.com/scientificreports/ PAT-DHHC21, unlike the majority of other DHHCs, is expressed on the plasma membrane, making it possible for DHHC21 to regulate the activities of cell surface receptors in blood vessels 18,23 . Our previous findings indicate that DHHC21 functional deficiency demonstrates survival benefits in animals with septic injury 22 . Moreover, a recent study shows that DHHC21 regulates vascular tone through the palmitoylation of vascular α1-adrenergic receptor (α1AR), a seven-transmembrane G protein-coupled receptor (GPCR) that induces smooth muscle contraction 15 . Increased release of endogenous α1AR agonist norepinephrine has been reported in patients with septic injury, contributing to intrarenal vasoconstriction and tissue hypoperfusion 24 . Norepinephrine is also used as the first-line vasopressor to treat septic shock, which could further augment regional ischemia in kidneys 12,25 . Therefore, in this study, we utilized DHHC21-deficient (Zdhhc21 dep/dep ) mice and tested the hypothesis that DHHC21 contributes to renal dysfunction and tissue damage in septic injury via promoting α1AR palmitoylation-dependent renal vasoconstriction and tissue hypoperfusion. Our findings indicated that Zdhhc21 dep/dep mice were more resistant to septic injury-induced renal dysfunction and structural damage. DHHC21 loss-of-function improved renal perfusion and RBF through inhibiting palmitoylation of α1AR and its ability to mediate renal vasoconstriction. Thus, we suggest DHHC21 as a novel mediator of septic injuryinduced renal dysfunction via mechanisms involving α1AR palmitoylation and α1AR activation-associated renal tissue hypoperfusion.

Results
DHHC21 loss-of-function reduces renal dysfunction in septic injury. Zdhhc21 dep/dep mice were employed to investigate the role of PAT-DHHC21 in renal dysfunction during septic injury. DHHC21 functional deficiency in Zdhhc21 dep/dep mice is due to a 3 bp deletion in exon 7 of Zdhhc21 gene, resulting in the loss of a single amino acid phenylalanine (F) at position 233 26 . Detailed information about the genotype and phenotype of Zdhhc21 dep/dep mice is presented in Supplementary Figure S1. We compared renal excretion between WT and Zdhhc21 dep/dep mice using a novel and minimally invasive imaging technique, multispectral optoacoustic tomography (MSOT). MSOT relies on the photoacoustic effect and detects the ultrasound signals produced by the thermal expansion of tissues/molecules when excited by near-infrared (NIR, 680-950 nm) laser pulses 27 . NIR dye IRDye800CW, which is exclusively excreted by the kidneys 28,29 , was injected intravenously 24 h after sham procedure and CLP-induced septic injury. As shown in Fig. 1, real-time recording of IRDye800CW clearance showed that IRDye800CW first appeared at the renal cortex region, then moved towards the medulla-pelvis region before being rapidly excreted by the kidneys. The time interval between the peak intensity of the cortex and medulla-pelvis region (Tmax-delay) was determined based on the clearance kinetics of IRDye800CW as shown in Fig. 2a. Greater Tmax-delay indicates that IRDye800CW takes longer to transit from cortex to medulla-pelvis region, suggesting impaired renal excretion 30 . DHHC21 loss-of-function showed no significant effect on renal excretion in sham controls; however, septic injury-induced increase in Tmax-delay was greatly inhibited in Zdhhc21 dep/dep mice (Fig. 2b). DHHC21 functional deficiency attenuates septic injury-induced renal damage. We then performed Periodic Acid-Schiff staining to visualize kidney structural damage in WT and Zdhhc21 dep/dep mice after CLP. As shown in Fig. 3a, septic injury led to severe renal tissue damage, including glomerular abnormality, loss of brush border in proximal tubule, vacuolation of tubule epithelium, tubular cell detachment/necrosis, and neutrophil infiltration; however, these morphological alterations were attenuated in Zdhhc21 dep/dep mice. Quantitative analysis showed that septic injury resulted in significantly increased renal injury score, which was greatly inhibited by DHHC21 loss-of-function (Fig. 3b). The levels of circulating creatinine and blood urea nitrogen (BUN) were also measured as indicators of renal dysfunction. Consistent with the histology results, septic injury significantly up-regulated plasma levels of creatinine and BUN in WT mice; however, these changes were ameliorated in Zdhhc21 dep/dep mice (Fig. 3c,d).
We also compared the levels of circulating cytokines in WT and Zdhhc21 dep/dep mice after sham and septic injury. Cytokine array analysis revealed that CLP led to an increased production of cytokines such as interleukin-1 (IL-1), IL-6, and TNF-α in both WT and Zdhhc21 dep/dep mice; yet, there was no remarkable differences in cytokine profiles between the two mouse strains after CLP ( Supplementary Fig. S2). These data indicate that the protective effects of DHHC21 functional deficiency on kidney function during septic injury are not due to inhibiting the production of pro-inflammatory cytokines. DHHC21 functional deficiency prevents the reduction of kidney perfusion and oxygen saturation during septic injury. Given the fact that renal tissue hypoperfusion is a major cause of renal dysfunction during septic injury, we compared kidney perfusion between WT and Zdhhc21 dep/dep mice after septic injury using MSOT. Detecting intrinsic NIR light absorbing hemoglobin via MSOT has been reported to be a reliable technique to measure organ perfusion in many studies 31,32 . As shown in Fig. 4a,b, the intensity of total hemoglobin was significantly decreased in kidneys of WT mice after CLP compared to sham controls. This reduction was greatly attenuated by DHHC21 functional deficiency, suggesting better-preserved renal perfusion in Zdh-hc21 dep/dep mice after septic injury. Next, we differentiated oxygenated and deoxygenated hemoglobin based on their distinctive spectra and calculated the oxygen saturation of the renal tissue. After 24 h of CLP, WT mice displayed remarkably less oxygenated hemoglobin but greater deoxygenated hemoglobin in kidneys than shamoperated mice, whereas these responses were ameliorated in Zdhhc21 dep/dep mice (Fig. 4c). Consistently, quantitative data in Fig. 4d indicated that Zdhhc21 dep/dep mice were more resistant to septic injury-induced decrease in renal oxygen saturation. Our results indicated that the RBF in WT mice was greatly reduced after septic injury; this reduction was significantly blunted by DHHC21 functional deficiency (Fig. 5a,c). We also monitored the changes in mean arterial pressure (MAP) during septic injury. As shown in Fig. 5b,d, septic injury caused a drastic decrease in MAP in WT mice; yet, no significant difference in MAP was  DHHC21 functional deficiency decreases α1AR palmitoylation and inhibits α1AR-mediated signaling pathway activation and vasoconstriction. To investigate the mechanism by which DHHC21 regulates RBF and renal tissue perfusion, we determined the roles of DHHC21 in α1AR palmitoyla-  www.nature.com/scientificreports/ tion and function. Both DHHC21 and α1AR are localized in the plasma membrane, which allows DHHC21 to dynamically interact with α1AR. Moreover, α1AR has previously been reported to co-immunoprecipitate with DHHC21, indicating their direct binding interaction 15 . We compared the levels of α1AR palmitoylation in renal arteries of WT and Zdhhc21 dep/dep mice after CLP, using resin-assisted capture (RAC, Supplementary Fig. S3). As illustrated in Fig. 6a, septic injury resulted in a dramatic increase in α1AR palmitoylation in renal arteries of WT mice, which was significantly reduced in Zdhhc21 dep/dep mice. Furthermore, we assessed the phosphorylation of extracellular signal-regulated kinase (ERK), a downstream signaling event of α1AR activation, in renal arteries of WT and Zdhhc21 dep/dep mice after septic injury. The level of phosphorylated ERK was significantly increased by CLP in renal arteries of WT mice; this response was diminished by DHHC21 loss-of-function (Fig. 6b).
In an effort to further evaluate the functional impacts of DHHC21 on α1AR-mediated vasoconstriction, we isolated renal arteries from WT and Zdhhc21 dep/dep mice and challenged the vessels with phenylephrine, an α1AR agonist. Our wire myograph data showed no significant difference in the high potassium (60 mM)-induced vasoconstrictions between WT renal arteries and Zdhhc21 dep/dep renal arteries ( Supplementary Fig. S4a); yet renal arteries isolated from Zdhhc21 dep/dep mice displayed a significant reduction in phenylephrine-induced vasoconstriction compared to those of WT mice (Fig. 6c,d). These results suggest the critical role of DHHC21-mediated α1AR palmitoylation in α1AR signaling activation and α1AR-mediated vasoconstriction. www.nature.com/scientificreports/ Consistent findings were observed in small arteries isolated from viable human kidneys. Our immunofluorescence images showed evidence for co-localization of DHHC21 and α1AR in human renal arteries (Fig. 7a). PAT inhibitor 2-bromopalmitate (2-BP) was applied to block the function of DHHC21 in small arteries isolated from human kidneys. 2-BP treatment showed no effects on high potassium-induced vasoconstriction (Supplementary Fig. S4b); however, 2-BP-treated small arteries displayed blunted constriction responses to phenylephrine compared to vehicle-treated ones (Fig. 7b). Collectivity, the above findings demonstrated that DHHC21 contributes to impaired renal perfusion and function during septic injury via promoting α1AR palmitoylation and α1AR-mediated renal vasoconstriction.

Discussion
In the present work, we report DHHC21 as a novel regulator of renal perfusion and function during septic injury. Our new findings indicate: (1) Zdhhc21 dep/dep mice exhibit better renal function and less renal damage after septic injury compared to WT mice; (2) DHHC21 functional deficiency preserves renal tissue perfusion, oxygen saturation, and RBF in septic injury; (3) DHHC21-catalyzed α1AR palmitoylation is required for the activation of α1AR signaling pathway and α1AR-induced constriction of renal arteries. This study provides a new mechanistic insight into the regulation of renal perfusion and function during septic injury. We suggest that DHHC21 functional deficiency confers protective effects on kidney function in septic injury and that inhibition of DHHC21 may serve as a therapeutic strategy to combat renal dysfunction during septic injury.
Renal tissue hypoperfusion/hypoxia is one of the leading causes of renal dysfunction during septic injury [3][4][5] . We detected renal hypoperfusion in mice after CLP, evidenced by weakened MSOT signal intensity of total hemoglobin, decreased oxygen saturation in renal tissue, and reduced RBF after septic injury. Consistent with our findings, the reduction in RBF has been observed in sepsis patients with kidney injury as well as in large-animal models of septic injury 7,10,11 . It is worth noting that several studies report that renal dysfunction can develop in sepsis animals with unchanged or even increased RBF 33,34 . These discrepant findings may be attributed to different animal models and methods used to measure blood flow. For instance, while CLP-induced polymicrobial septic injury has been commonly used, some studies use intravenous injection of Escherichia coli 35 .
Protein palmitoylation has recently been reported as a novel regulator of protein function, involved in the pathogenesis and progression of many diseases. It was first identified as a new type of post-translational modification by Schmidt et al. 36,37 . The discovery of the DHHC family had since promoted the rapid expansion of this field 38,39 . The roles of palmitoylation and PATs have been reported in numerous physiological and pathological conditions, including lipid metabolism, cancer, cardiovascular diseases, and neurological disorders 23,40,41 . Although the involvement of palmitoylation in polycystic kidney disease and kidney cancer has been previously reported 23,42 , there have been very limited studies examining the functional role of PATs in kidney diseases, specifically in renal dysfunction during septic injury. To the best of our knowledge, we are the first to investigate the role of DHHC21 in renal dysfunction in septic injury. Our findings indicate that the inhibition of DHHC21 greatly rescues kidney function and preserves kidney structure in septic injury.
Our study utilizing Zdhhc21 dep/dep mice demonstrates that the beneficial effects of DHHC21 functional deficiency on kidney function is attributed to its ability to improve renal tissue perfusion during septic injury. Under basal conditions, Zdhhc21 dep/dep mice show no signs of salt/water imbalance or renal structural/functional damage 15 . Yet, DHHC21 loss-of-function suppresses the septic injury-induced reduction of RBF, renal perfusion, and renal oxygen saturation. Increased renal vascular resistance caused by excessive renal vasoconstriction is considered as a major reason for impaired renal tissue perfusion in septic injury 6,9,33 . Our results indicate the involvement of α1AR in mediating septic injury-induced renal tissue hypoperfusion, as evidenced by the increase in α1AR palmitoylation and the activation of α1AR signaling pathway in septic injury. However, DHHC21 functional deficiency inhibits α1AR palmitoylation and results in a blunted ability of α1AR to activate its downstream effector and to mediate phenylephrine-induced vasoconstriction in renal arteries.
The molecular mechanism underlying the regulation of α1AR function by DHHC21 still remains to be fully elucidated. The defect in α1AR function caused by DHHC21 loss-of-function may be attributed to the conformational change of α1AR. Many GPCRs rely on palmitoylation for their appropriate intracellular conformation. The attached palmitate in the C-terminal tails of GPCRs inserts into the plasma membrane to create the fourth intracellular loop, which is essential for the interaction of GPCRs with their partner proteins and the propagation of GPCR signals 17,43 . Previous study has indicated that the palmitoylation of α1bAR also occurs in its C-terminal region 44 . Therefore, the lack of DHHC21-mediated α1AR palmitoylation may alter the intracellular conformation of α1AR, thus blocking the propagation of α1AR signals. This can be supported by our results showing that the activation of ERK, a downstream signaling event of α1AR activation, is inhibited in Zdhhc21 dep/dep mice subjected to septic injury. However, the DHHC21-regulated topology of α1AR carboxyl terminal needs to be further confirmed using X-ray crystallography.
Another novel aspect of our study is the utilization of MSOT to evaluate kidney perfusion and function during septic injury. MSOT has been reported as a label-free method to measure the perfusion of different organs and a reliable imaging technique to determine kidney function 28,32,45 . Compared to Doppler flowmeter which does not allow visualization of microvasculature 45 , MSOT generates images with high spatial resolution at 150 μm while providing quantitative assessment of the perfusion status in the kidney microvasculature. The water-soluble IRDye800CW is a safe tracer to use for measuring renal clearance, and it does not induce toxicity at the dose as high as 20 mg/kg 46 . The two-phase movement of IRDye800CW we observed is consistent with previously published studies 29 . Our MSOT recordings indicate a greater Tmax-delay in sepsis kidneys, suggesting impaired renal clearance. A nephropathy study demonstrates that the impaired renal function detected by MSOT is significantly correlated with glomerular histological damage 30 . Our data, consistent with previous publications, suggest that MSOT is a minimally-invasive and reliable technique to monitor renal perfusion and function.  www.nature.com/scientificreports/ Currently, there are no DHHC21-specific inhibitors available. The most commonly used PAT inhibitor is 2-BP which has a broad effect on multiple DHHCs and also interfere with fatty acid metabolism 47 . Thus, developing small molecule inhibitors that specifically target DHHC21 would be a promising direction. It is also worth pointing out that α1AR is not the only substrate of DHHC21. Several other DHHC21 substrates have recently been identified, including platelet endothelial cell adhesion molecule (PECAM-1), estrogen receptor, caveolin-1, endothelial nitric oxide synthase (eNOS), Fyn, superoxide dismutase (SOD-1), and PLCβ1 22,41,48,49 . While beyond the scope of the present study, we cannot rule out the possibility that these substrates may also affect renal function during septic injury.
In conclusion, the present study demonstrates for the first time that DHHC21 plays a critical role in regulating renal perfusion during septic injury via mechanism involving α1AR palmitoylation and α1AR-mediated vasoconstriction. Furthermore, inhibition of DHHC21 exerts protective effects on renal function during septic injury.

Materials and methods
Reagents. All reagents are listed in Supplementary Table S1.
Animals. Zdhhc21 dep/dep mice and their wild-type control mice (B6C3Fe) were purchased from Jackson Laboratory. The genotype of Zdhhc21 dep/dep mice was confirmed by sequencing (Genewiz, Inc., NJ, USA). Primers used for sequencing were: AGC TGA CTG AAG GGC ACC (forward) and AAA ACC TGT AAC GCA TTT CCA (reverse) 22 . Animals were maintained under a 12/12-h light/dark cycle with free access to food and water. Mice (16-20 weeks) of both genders were used for this study. All animal experiments are approved by the University of South Florida Institutional Animal Care and Use Committee and conform to the NIH Guide for the Care and Use of Laboratory Animals. The study was carried out in compliance with the ARRIVE guidelines.
Cecal ligation and puncture. Mice were anesthetized with isoflurane (3% induction and 1% maintenance). A midline incision was made in the shaved abdominal region, and the cecum was exteriorized, tightly ligated at 5 mm below the ileocecal valve, and perforated twice with a 20-gauge needle distal to the point of ligation. One mm of feces was extruded from each puncture hole. The cecum was then repositioned, and the abdomen was closed in two layers. 37 °C Lactated Ringer solution was applied topically to prevent the cecum from drying. Mice were then given 37 °C 0.9% saline subcutaneously for fluid resuscitation. A preoperative 0.5-1.5 mg/kg dose of Buprenorphine Sustained-Release was given for analgesia. Sham mice were subjected to the same surgical procedure but without cecal ligation and puncture 50,51 . Multispectral optoacoustic tomography. Assessment of renal excretion of IRDye800CW. Mice were anesthetized via isoflurane and depilated around the torso region. Mice were imaged using iThera Medical MSOT Imaging System (Munich, Germany) at multiple wavelengths: 700, 730, 760, 775, 785, 800, 850 nm at a rate of 10 frames/s 29 . Mice were placed in a supine position in a holder and moved along the horizontal linear stage to position the kidneys on top of the fix-positioned concave transducer. After baseline recording, 60 nmol of IRDye800CW dissolved in 100 μl of 0.9% saline was injected intravenously over a period of 10 s. Images were reconstructed using the back-projection formula and processed with multispectral analysis. Regions of interest (ROIs) were drawn around the renal cortex and medulla/pelvis region, and the temporal changes of MSOT signal in ROIs were depicted.
Measurement of renal perfusion. Mice were prepared for MSOT imaging utilizing the same procedures mentioned above. Mouse respiration was closely monitored throughout the entire procedure. Images were reconstructed and the spectral unmixing was performed. ROIs were drawn around the entire right kidney region. Mean pixel intensities of oxygenated hemoglobin (HbO 2 ) and deoxygenated hemoglobin (Hb) were acquired. Total hemoglobin (HbT = HbO 2 + Hb) and oxygen saturation (S O2 = HbO 2 /HbT) were then calculated 52 .
Kidney histopathology. Kidneys were fixed, cut along the sagittal plane, and processed for paraffinembedding. Sections (5 μm) were de-paraffinized, rehydrated, and oxidized with periodic acid for 8 min at room temperature (RT), followed by incubation with Schiff 's solution for 25 min. Sections were then counterstained with hematoxylin. Images were captured using Keyence BZ-X710 (Itasca, IL, USA). Renal structural damage was evaluated based on the following features: glomerular abnormality, loss of brush border of proximal tubule, vacuolation, dilation of tubule epithelium, tubular cell detachment/necrosis, and neutrophil infiltration 53 . Each feature was graded on a scale of 0-5 based on the severity.

Measurement of creatinine and blood urea nitrogen.
Mouse blood was collected via cardiac puncture 24 h after septic injury. Plasma was generated by centrifuging the blood at 2500 g for 15 min at RT. Measurement of creatinine: plasma was deproteinized using 10-kDa Spin Column; the levels of creatinine in the filtrate were then measured using a Creatinine Assay Kit. The level of blood urea nitrogen (BUN) was determined using a Urea Nitrogen Colorimetric Detection Kit following manufacturer's instructions.

Measurement of RBF and MAP.
Mice were anesthetized using isoflurane. The blood pressure transducer was cannulated into the carotid artery. A midline incision was made in the abdominal area followed by a left transverse incision to expose the renal artery. RBF was measured using an ultrasound transit-time flowmeter (TS-420; Transonic Systems Inc., Ithaca, NY, USA). After placing the flow probe around the exposed renal artery, mice were allowed to be stabilized for at least 30 min. RBF and MAP were recorded simultaneously using Small arteries from human kidneys. Small arteries (~ 1000 μm in diameter) were dissected from intact viable human kidneys that were rejected for transplant surgery. Vessel segments (~ 2 mm long) were mounted onto the L-bars (250 μm in diameter) of wire myograph chamber. Similar equilibration and normalization procedures were performed on human arteries. Next, the arteries were incubated with vehicle control (0.02% DMSO in PSS) for 1 h. The vessels were then tested for viability with 60 mM KPSS and treated with increasing concentrations of phenylephrine (10 nM-30 μM). After washing off, the vessels were incubated with 2-BP (100 μM) for 1 h; the vessel viability was then tested again with 60 mM KPSS with 100 μM 2-BP. The small arteries with no or reduced responses to KPSS were discarded. The arteries were then challenged with the same concentrations of phenylephrine followed by acetylcholine (10 μM). The concentration-response curves of vehicle control-or 2-BP-treated arteries were constructed and compared.
Immunofluorescence. Human renal arteries were fixed in 10% formalin for 48 h, paraffin-embedded and sectioned. Slides were then deparaffinized, rehydrated and permeabilized with PBS containing 0.05% Triton X-100 22 . After blocking, slides were labelled with rabbit anti-DHHC21 and goat anti-α1AR antibodies (1:100) overnight at 4 °C. After washing, secondary antibody incubation was done with donkey anti-rabbit Alexa Fluor 488 and donkey anti-goat Alexa Fluor 568 (1:500). After mounting with ProLong Diamond mounting medium with DAPI, slides were imaged with Leica SP8 spectral inverted laser scanning confocal microscope.
Statistical analysis. Detailed information (e.g. the name of the test, post-hoc test, n value, alpha level, p value) for each statistical test is listed in Supplementary Table S2. All data meet the normal distribution assumption. Normality test was performed using Shapiro-Wilk Test with alpha level set at 0.05. Comparisons between two groups were analyzed using Student's t test (two-tailed), and three groups or more were compared via Oneway ANOVA with Tukey post-hoc analysis. The comparison for wire myograph curves was performed using Two-way ANOVA. p < 0.05 was used for statistical significance. All statistical analyses were performed using GraphPad Prism 7.0d (San Diego, CA, USA

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.