Hypothalamic Sirt1 protects terminal Schwann cells and neuromuscular junctions from age‐related morphological changes

Summary Neuromuscular decline occurs with aging. The neuromuscular junction (NMJ), the interface between motor nerve and muscle, also undergoes age‐related changes. Aging effects on the NMJ components—motor nerve terminal, acetylcholine receptors (AChRs), and nonmyelinating terminal Schwann cells (tSCs)—have not been comprehensively evaluated. Sirtuins delay mammalian aging and increase longevity. Increased hypothalamic Sirt1 expression results in more youthful physiology, but the relationship between NMJ morphology and hypothalamic Sirt1 was previously unknown. In wild‐type mice, all NMJ components showed age‐associated morphological changes with ~80% of NMJs displaying abnormalities by 17 months of age. Aged mice with brain‐specific Sirt1 overexpression (BRASTO) had more youthful NMJ morphologic features compared to controls with increased tSC numbers, increased NMJ innervation, and increased numbers of normal AChRs. Sympathetic NMJ innervation was increased in BRASTO mice. In contrast, hypothalamic‐specific Sirt1 knockdown led to tSC abnormalities, decreased tSC numbers, and more denervated endplates compared to controls. Our data suggest that hypothalamic Sirt1 functions to protect NMJs in skeletal muscle from age‐related changes via sympathetic innervation.

motor nerve terminal, acetylcholine receptors (AChRs), and specialized glial cells termed perisynaptic Schwann cells or terminal Schwann cells (tSCs). Unlike the myelinating Schwann cells that surround motor nerve axons, tSCs are nonmyelinating glial cells that surround nerve terminals at the NMJ. tSCs regulate morphological stability and function of the NMJ and play a critical role in regulating the nerve-muscle interface with important roles in synaptic regeneration in homeostasis (Reddy, Koirala, Sugiura, Herrera & Ko, 2003), after injury (Kang, Tian & Thompson, 2003), and in aging (Griffin & Thompson, 2008).
Age-associated NMJ degeneration is well characterized in animals and humans (Balice- Gordon, 1997;Boaro, Soares & Konig, 1998;Deschenes, Roby, Eason & Harris, 2010;Jang & Van Remmen, 2010;Kawabuchi et al., 2001;Luff, 1998;McMullen & Andrade, 2009;Smith & Chapman, 1987;Steinbach, 1981;Wernig & Herrera, 1986). Aging results in progressive NMJ degradation causing a steady decline in muscle mass and strength, termed senile muscle atrophy, or sarcopenia. Interactions among tSCs, presynaptic nerve terminals, and postsynaptic endplates and muscle play critical roles in synaptic growth, maintenance, and survival. The exchange of trophic factors has been implicated in pre-and postsynaptic development as well as preserving neuronal and synaptic plasticity at the NMJ. In addition, factors such as mitochondrial dysfunction, oxidative stress, inflammation, changes in muscle fiber innervation, and motor unit mechanical properties probably contribute to NMJ degeneration. Interventions such as caloric restriction and exercise may positively affect the NMJ and attenuate the age-related progressive impairment in motor function (Valdez et al., 2010).
Recent studies have demonstrated that hypothalamic Sirt1 regulates mammalian aging and longevity in mice. Stereotactic injection of Sirt1-expressing lentiviruses into the dorsomedial hypothalamus (DMH) of aged adult wild-type (WT) mice ameliorates age-associated declines in physical activity and body temperature, supporting the notion that the hypothalamus, specifically the DMH, is one of the control centers for mammalian aging and longevity (Satoh et al., 2010(Satoh et al., , 2013. Mice with brain-specific overexpression of Sirt1 (BRASTO) show a significant delay in aging with extended lifespans in both males and females (Satoh et al., 2013). Specifically, skeletal muscle in BRASTO mice maintains youthful morphology and mitochondrial function during the aging process due to enhanced sympathetic nervous tone during the dark period. Although the mechanism by which the signal from the hypothalamus is specifically directed to skeletal muscle remains unknown, increased sympathetic tone may contribute to skeletal muscle and NMJ changes and increased longevity to delay the effects of aging.
Sympathetic innervation controls muscle metabolism, maintenance, and function of nerve-muscle contact (Roatta & Farina, 2009). There are a few studies reporting on direct innervation of skeletal muscle fibers by nonmyelinated, noradrenergic fibers (Barker & Saito, 1981;Lynch & Ryall, 2008), suggesting that sympathetic actions on skeletal muscle are at least partially mediated by neural mechanisms. It has been reported that sympathetic neurons coinnervate several targets in muscle, including blood vessels, motor neurons, muscle fibers, and NMJs (Khan et al., 2016;Rudolf et al., 2013). Moreover, cAMP/PKA-dependent signaling at the NMJ is important for synapse stabilization and metabolic control of AChR function (Li, Yi & Thompson, 2011).
This study examines the effects of aging on all components of the NMJ, including tSCs. In addition, we investigate the impact of differing levels of brain or hypothalamic Sirt1 expression on NMJ architecture in skeletal muscle and suggest a role for sympathetic innervation in mediating the systemic effects of central Sirt1 changes at the level of the NMJ.

| Body mass and sternomastoid (SM) mass decrease with age in WT mice
Peak average body weight occurred at 14 months of age (41.34 AE 4.9 g) and declined by 26% by 33 months of age (30.67 AE 3.0 g) in WT mice. Wet weight of the SM muscle in isolation paralleled total body mass with a 33% decline in mass between 14 and 33 months of age (0.021 AE 0.003 g and 0.014 AE 0.001 g, respectively; Table 1).

| Terminal Schwann cells
The Schwann cell marker S100 was present in tSCs at the NMJ, and these cells had ovoid-shaped nuclei and cytoplasmic processes that colocalized with AChRs. tSCs in young adult mice were present in almost 100% of NMJs and stayed in contact with the entire endplate (Figures 1 and 2a). In contrast, tSC numbers decreased with age ( Figure 1f). The percentage of NMJs with tSCs present decreased significantly (p < .05) from 100% at 9 months of age to 11.2 AE 1.9% at 33 months of age ( Figure 2a). In addition, tSC bodies exhibited irregular contour and variability in staining intensity with S100 Ab (Figure 2a). tSC processes were thinner with advancing age, but overall were less affected by mouse age than tSC bodies (data not shown).

| Presynaptic nerve terminal
NMJ innervation was evaluated with immunostaining with NF200 Ab and a-BTX to identify the presence of nerve terminals. For quantification, NMJs were divided into three groups: (i) fully innervated, (ii) partially innervated (less branching or fragmented), and (iii) not S100 BTX DAPI NF200 BTX DAPI S100 BTX DAPI innervated (denervated) (Figure 2b). Nerve terminals from young adults were distinct, branching, and colocalized with endplates (a-BTX staining) (Figure 1b). NMJ innervation significantly (p < .05) decreased with age beginning at 14 months of age with denervation of nearly 20% of NMJs ( Figure 2b). The percentage of fully denervated NMJs increased to over 35% at 17 months of age, but then decreased again until advanced age (33 months) ( Figure 2b). Aged mice demonstrated almost no nerve branching and frequently had no nerve terminal associated with the NMJ. Some nerve sprouting was observed only in 14-month-old mice, but by 33 months of age, nearly half of the assessed NMJs were denervated ( Figure 2b). Variability in nerve morphology was also seen, including enlarged and spherical nerve ends as well as thin axon terminals ( Figure 2b).

| Motor endplates (AChRs)
Age-associated changes on postsynaptic AChRs were characterized after immunostaining with a-BTX, and the number of AChR fragments per NMJ was evaluated (Figures 1 and 2c). Data are presented as percentage of NMJs that contain normal, partially fragmented, or fully fragmented AChRs. Most endplates from young mice (Figures 1 and   2c) formed normal, pretzel-shaped, synaptic gutters consisting of a continuous structure with less than five fragments. Endplates from aged mice formed a single cluster (often diffusive) or were more fragmented or granular (Figures 1 and 2c). At 14 months of age,~34% of NMJs showed partial or full fragmentation which increased drastically between 17 and 33 months of age (p < .05). By 33 months of age, only~12% of AChRs showed normal morphology.

NMJ morphology than WT mice
Each of the three main NMJ components (tSCs, nerve terminal, and AChRs) was compared in young (7-month-old) and aged (25-and 33month-old) BRASTO mice and age-matched WT controls qualitatively ( Figure 3) and quantitatively ( Figure 4). NMJ morphology in young (7month-old) BRASTO mice did not differ from characteristic, healthy NMJ morphology seen in age-matched controls (Figure 3a-d). When mice became old, more tSC processes and tSC bodies were observed at the NMJ in SM muscles of BRASTO mice compared to controls (Figures 3   and 4a). The percentage of NMJs with tSC bodies present significantly (p < .05) increased and was nearly double that of controls at 33 months of age (Figure 4a). Similarly, the percentage of fully innervated NMJs was significantly higher in the aged BRASTO SM muscles compared to controls at both ages (p < .05, Figure 4b). With respect to motor endplates, full fragmentation of the AChRs was significantly lower in BRASTO mice compared to their age-matched controls (p < .05, Figure 4c). Similarly, the percentage of normal motor endplates in BRASTO mice was approximately double that of their age-matched controls at both 25 and 33 months of age (p < .05, Figure 4c). In summary, the aged BRASTO mice showed a more youthful morphology of each of the three main NMJ components compared to WT controls.

| DMH-specific Sirt1 knockdown results in a more aged NMJ morphology
DMH and lateral hypothalamus (LH) neurons are responsible for phenotypes observed in aged BRASTO mice (Satoh et al., 2013). In particular, Sirt1 in the DMH plays a critical role in maintaining physical activity.
NMJ analyses were thus performed 3 weeks after stereotactic injections of shRNA-Sirt1 or shRNA-Ctr into the DMH of 3-month-old WT mice. The efficiency of the lentiviral knockdown was noted to be a 65% reduction in mRNA (data not shown) and similar to previous reports (Satoh et al., 2013). DMH-specific Sirt1 knockdown accelerated ageassociated NMJ morphologic changes compared to age-matched controls ( Figures 5 and 6). A variety of tSC morphological abnormalities ( Figure 5) not observed in normal, healthy NMJs in control mice (13.6 AE 0.74%, p < .05; Figure 6a) was also seen in the NMJs from DMH-specific Sirt1 knockdown mice. tSCs demonstrated atypically large and intensively stained cell bodies (Figure 5d), often observed outside the NMJ area (Figure 5e), and many were also lightly stained with S100 Ab with some processes (Figure 5f). In addition, the number of tSCs associated with each endplate was less in shRNA-Sirt1 mice ( (red), respectively. Graphs show the proportion of NMJs with the three classifications of endplate fragmentation (left graph) and the percentage of fully fragmented AChRs (right graph). S100 Ab (for tSCs, a); NF200 Ab (for neurofilaments, b); BTX = a-bungarotoxin (for AChRs, a-c); DAPI = nuclear staining (blue, a-c); Scale bar = 20 lm. Data AE SD; different letter indicates significant difference; p < .05 SNYDER-WARWICK ET AL.

| Sympathetic innervation is present at the NMJ and increased in BRASTO mice
We demonstrated that BRASTO mice have a more youthful NMJ  to be qualitatively consistent with higher mRNA expression levels of b2-AR and higher cAMP levels in aged BRASTO skeletal muscle, compared to those in age-matched control skeletal muscle (Satoh et al., 2013). NPY staining also showed similar colocalization to the NMJs (data not shown). Therefore, increased sympathetic NMJ innervation was observed with increased Sirt1 function in the hypothalamus.

| DISCUSSION
The aging process is associated with progressive loss of muscle mass, termed sarcopenia, which affects 13%-24% of humans under the age of 70 and 43%-60% of people over the age of 80 (Baumgartner et al., 1998). Neuromuscular dysfunction is a likely etiology of sarcopenia (Pannerec et al., 2016). Age-associated changes occur not only in muscle and nerve, but also at the interface between the two, the NMJ. Age-related morphologic NMJ changes have been previously  Kawabuchi et al., 2001;Luff, 1998;McMullen & Andrade, 2009;Smith & Chapman, 1987;Steinbach, 1981;Wernig & Herrera, 1986), and NMJ disruption has been identified as a driver of sarcopenia in a rodent model (Ibebunjo et al., 2013). The majority of those studies, however, described changes in nerve terminals or AChRs. In this study, we conducted comprehensive quantification of the age-associated changes of all three of the main NMJ components: tSCs, nerve terminals, and motor endplates. In addition, we demonstrated that hypothalamic Sirt1 signaling through sympathetic innervation is critical to counteract age-associated NMJ decline.
Given their roles in synaptic function and maintenance, tSC alterations could contribute to declines in muscle function with advancing age. With progressive age, we noted decreased tSC numbers at shRNA-Ctr shRNA-Sirt1 S100 BTX DAPI S100 NF200 DAPI shRNA-Ctr shRNA-Sirt1 shRNA-Sirt1 shRNA-Sirt1 shRNA-Ctr shRNA-Sirt1 in 3-month-old WT mice results in a more aged NMJ morphology compared to age-matched, control-injected mice (shRNA-Ctr, a, c, g). Representative low magnification images of whole mount sternomastoid muscle show tSC abnormalities (less staining intensity and fewer numbers) in shRNA-Sirt1 mice (b, e, f). Note also tSC anomalies (arrows): large and intensely stained tSC bodies (d), tSCs migrated outside of NMJ area (e), or lightly stained with few processes colocalized with AChRs (f) in shRNA-Sirt1 mice. Less severe changes are observed for nerve terminals (h, arrows), and almost no changes are seen in motor endplates (b, df). S100 Ab (for tSCs, green), NF200 Ab (for neurofilaments, green), BTX (for AChRs, red), and DAPI (nuclear staining, blue). Scale bar = 20 lm the WT NMJ, with abnormalities in tSC contour, staining, and cytoplasmic processes. Our tSC data corroborate with those of others (Boaro et al., 1998;Chai et al., 2011;Kawabuchi et al., 2001;Ludatscher, Silbermann, Gershon & Reznick, 1985). By 27 months of age, tSC degeneration was noted in 35% of NMJs in mouse gastrocnemius in contrast to the absence of tSC degenerative changes in 6month-old mice (Ludatscher et al., 1985). Additionally, tSCs have been noted to be thin, disorganized, and to only partially cover the endplates in advanced aged mice (Chai et al., 2011;Kawabuchi et al., 2001), similar to the tSC changes we noted in the DMH-specific Sirt1 knockdown mice in this study. The functional sequelae of loss of tSC structure and number are unknown. Age-related tSC loss, migration, and disorganization have been hypothesized to be related to myofiber denervation (Chai et al., 2011;Connor, Suzuki, Lee, Sewall & Heisey, 2002). Connor et al. (2002) noted extension of tSC cytoplasmic processes in 24-month-old rats in a manner similar to that seen after denervation (Connor et al., 2002), and Kawabuchi et al. (2001) noted similar findings in 24-month-old mice (Kawabuchi et al., 2001). Whether muscle denervation is a cause or effect of tSC changes, however, has not been conclusively demonstrated.
Nerve terminal morphology was also impacted by age in WT animals. Nerve terminals showed less branching with advancing age, and NMJ innervation declined with age. Similar morphological changes were noted in the DMH-specific Sirt1 knockdown mice. did not correlate with muscle force deficits. Similarly, we did not observe differences in muscle force between aged BRASTO and control mice and also between DMH-specific Sirt1 knockdown mice and controls. These findings may be the result of insufficient test sensitivity, insufficient time to assessment, or actual physiological compensation in the muscle. A compensatory mechanism fits with initial decline in muscle force that later improves to baseline levels. Similarly, others have reported no major age-dependent physiological differences, despite age-related morphological changes. The increased endplate fragmentation noted by Willadt et al. (2016) did not correlate with function; compound muscle action potential (CMAP) amplitude was stable with age (Willadt et al., 2016). Similarly, the absence of nerve terminals in 40% of NMJs in 34-month-old mice did not result in major physiological changes compared to young (8to 12-month-old) mice (Banker et al., 1983). Functional stability might result from compensatory mechanisms, such as motor unit enlargement or increased synaptic vesicle release (Robbins, 1992), or from incomplete NMJ affliction with age-related changes, as greater than 80% denervation is required to impart a functional deficit in muscle force (Gordon, Yang, Ayer, Stein & Tyreman, 1993). The aging synapse, therefore, is highly compensated rather than in a progressive state of deterioration from youth.
Importantly, we demonstrated that changes in central Sirt1 expression counteract age-associated NMJ decline in skeletal muscle.
Brain-specific overexpression of Sirt1 results in a more youthful NMJ phenotype compared to age-matched controls with respect to all three of the main NMJ components: tSCs, nerve terminals, and AChRs. Conversely, DMH-specific knockdown of Sirt1 resulted in more aged tSC and NMJ morphologies (Figure 8). Of the NMJ components, tSCs showed the biggest effect from Sirt1 knockdown within the DMH, suggesting that tSCs may be early harbingers of systemic or environmental change. This hypothesis requires further investigation. Interestingly, overexpression of Sirt1 in skeletal muscle does not delay metabolic age-related changes (White et al., 2013).
Together, these results implicate central Sirt1 expression within the hypothalamus, rather than local or peripheral Sirt1 expression, as a modifier of tSC and NMJ morphology.
We hypothesized that the sympathetic nervous system mediates the effects of central Sirt1 expression on NMJ morphology. Impaired sympathetic neural function is associated with several age-related systemic changes (Santulli & Iaccarino, 2013), and systemic physiological improvements seen in BRASTO mice result from sympathetic stimulation (Satoh et al., 2013). In addition, the sympathetic nervous system is present within skeletal muscle of young adult mice (Khan  Rudolf et al., 2013) and may modify muscle contraction, perfusion, and metabolism (Rudolf et al., 2013). Sympathetic innervation, as indicated by TH staining, was expressed in 96% of NMJs in the mouse EDL and 91% of those in the soleus (Khan et al., 2016).
This same study found that chemical sympathectomy resulted in decreased NMJ size, complexity, and electrophysiologic values that were rescued by sympathicomimetic agents. Another sympathetic receptor that mediates responses to catecholamines, the b2-AR, is found extensively throughout the body. While all three b-adrenergic receptor types are present in skeletal muscle, there is a 10-fold increased proportion of the b2 isoform in skeletal muscle (Santulli & Iaccarino, 2013). An age-related decline in responsiveness to b2-ARs has been proposed (Ford, Dachman, Blaschke & Hoffman, 1995), and b2 receptor agonists have reversed age-related muscle weakness and wasting in rats (Ryall, Plant, Gregorevic, Sillence & Lynch, 2004).
Because of the evidence supporting sympathetic innervation contributing to more youthful systemic physiology and NMJ morphology and function, we evaluated the presence of sympathetic innervation in our models via TH, b2-AR, and NPY staining. We showed colocalization of all of these markers with AChRs, confirming sympathetic innervation at the NMJ. Interestingly, quantification of sympathetic NMJ innervation showed a twofold increase in the NMJs (Rudolf et al., 2013) staining for TH and a threefold increase in NMJs staining for b2-AR in BRASTO mice compared to age-matched controls.
These data suggest that the mechanism linking the more youthful NMJ morphology associated with central Sirt1 overexpression is sympathetically mediated. Manipulation of sympathetic innervation and activity is required in future studies to elucidate this mechanism.
Our study demonstrates degenerative changes in the three main NMJ components: tSCs, nerve terminals, and AChRs with advancing age. We further illustrate brain-specific Sirt1 overexpression resulting in more youthful morphologies of all NMJ components, as well as more aged NMJ morphologies with DMH-specific Sirt1 knockdown.
The NMJ effects seen with manipulation of hypothalamic Sirt1 expression confirm the importance of Sirt1 in delaying age-related changes in skeletal muscle. The mechanisms linking central Sirt1 expression with muscle structure remain to be elucidated, but the data from this study suggest an association with sympathetic innervation.

| CONCLUSION S
Age-associated morphologic changes occur in all NMJ components, including tSCs, nerve terminals, and AChRs, with~80% of NMJs exhibiting synaptic abnormalities by 17 months of age. Brain-specific Sirt1 overexpression results in more youthful-appearing NMJs, and DMH-specific Sirt1 knockdown effects a more aged NMJ morphology. Differences in NMJ morphology due to Sirt1 central expression are associated with changes in sympathetic innervation at the NMJ.
We conclude that central Sirt1 expression protects against agerelated decline in skeletal muscle, mediated via the NMJ and sympathetic nervous system.

| Mice
The analyses were performed in wild-type (WT, C57BL/6J) and S100-GFP transgenic mice at 3,9,14,17,20,25,29, and 33 months of age. S100-GFP mice (generous gift from Dr. Susan Mackinnon, Washington University School of Medicine, St. Louis, MO) utilize the S100B promoter to drive GFP expression in all glial cells (Feng et al., 2000). We also used two types of Sirt1 mice: transgenic mice (7-, 16-, 20-, 25-and 33-month-old) where brain-specific Sirt1 was overexpressed (BRASTO mice (Satoh et al., 2010(Satoh et al., , 2013 and shRNA-Sirt1-injected mice (3-month-old) where Sirt1 was knocked down in the DMH using lentivirus and the stereotactic injection procedure described previously (Satoh et al., 2013)). The BRASTO mice with an HA-tagged Sirt1 transgene driven by the mouse prion promoter were backcrossed to C57BL/6J for 8 to 10 generations (Satoh et al., 2010). Mice in the aging cohorts were carefully inspected every day for all experiments, and both sexes were used for experiments. All Age (months) * * * F I G U R E 8 Sirt1 plays a protective role against age-associated NMJ morphological abnormalities in the sternomastoid muscle. Percentages of NMJ abnormalities (≥1 in any of the 3 NMJ components) are shown for WT, BRASTO, and DMH-specific Sirt1 knockdown (shRNA-Sirt1) mice at differing ages. Young adult mice (age 3 months) with DMH-specific Sirt1 knockdown (red bar) display a more aged NMJ morphology, with a significantly greater percentage of NMJ abnormalities than age-matched controls (shRNA-Ctr). BRASTO mice (ages 25 and 33 months, green bars) have a more youthful NMJ morphology, with significantly fewer NMJ abnormalities, than age-matched controls. Data AE SD; *p < .05; different letter indicates significant difference (for 3-, 25-and 33month-old mice) of NMJs compared to other muscles, thereby providing an excellent model for NMJ evaluation, was selected for our studies. SM muscles from both sides of the neck were dissected and postfixed overnight at 4°C. For immunofluorescence labeling of NMJ components, muscles were processed as whole mounts (Love & Thompson, 1998)

| NMJ analysis and quantification
NMJ analyses were performed on whole mounts and frozen sections of the SM muscles (Love & Thompson, 1998;Valdez et al., 2010). To facilitate whole mount imaging, the entire SM muscle was flat mounted in Vectashield mounting medium with a coverslip. Morphologic changes between young adult and aging mice were analyzed in three major NMJ components: tSCs (S100 Ab labeling), the presynaptic nerve terminals (NF200/SV2 Abs labeling), and postsynaptic AChRs on the motor endplate (a-BTX labeling). The whole mount muscle and muscle sections were imaged using an Axio Imager M2 fluorescent microscope (Zeiss, For quantitative analysis, muscle sections from three to nine animals from each age-group were processed. At least 15 sections and at least six or more macroscopic fields of view/section were analyzed per muscle totaling >100 NMJs per muscle. The sections selected for staining were from different parts of the muscle, but more sections were analyzed from middle portion of the muscle where the NMJ distributions were nearly identical at any sectioning level. All quantitative analyses were confirmed via blinded evaluators.
The three cellular components lying in close apposition at the NMJs-tSCs with their processes, motor nerve terminals, and AChRs -were evaluated. Only NMJs with motor endplates lying en face in the major plane of section were selected for quantification. Muscles where antibody staining was too faint to quantify due to poor antibody penetration were excluded from further analysis.
The percent of NMJs with tSCs (cell bodies) and the number of tSCs present per NMJ were determined in images labeled with S100 Ab, a-BTX, and DAPI (Love & Thompson, 1998 five or fewer fragments), partially fragmented (AChRs with more than five fragments or one irregular structure), or fully fragmented (AChRs with patchy or granular distribution). Most NMJs in young adult mice contain fewer than five AChR islands (Li et al., 2011).
The presence of sympathetic innervation at the NMJs was analyzed in the SM muscle sections from 20-month-old BRASTO (n = 4) and control (n = 3) mice stained with antibodies against b2-AR, NPY, and TH in combination with a-BTX and DAPI.

Sirt1 and shRNA-Ctr
For a loss-of-function approach, we used short hairpin RNA (shRNA) to knock down the endogenous expression of Sirt1 from the DMH.
shRNA-Sirt1 and control (shRNA-Ctr) lentivirus were prepared in the Hope Center Viral Vectors Core at Washington University School of Medicine as described previously (Satoh et al., 2013). Lentivirus was produced by cotransfecting HEK293T cells with vectors containing the firefly luciferase or Sirt1 shRNA via the calcium phosphate precipitation procedure (Li et al., 2010). Virus with titer of 3.6 to 4.3 9 10 8 TU/ml was used for stereotactic injection (see below).

| Stereotactic injection of lentivirus expressing
Sirt1 or firefly luciferase shRNA into the DMH WT mice (3-month-old; n = 4 mice/group/experiment; n = 2 experiments) were used for stereotactic injection of experimental (shRNA-Sirt1) or control (shRNA-Ctr) lentivirus. The injection was performed under aseptic technique at the Animal Surgery Core at Washington University School of Medicine according to the method previously described in detail (Satoh et al., 2013). Immediately after the injection, animals were allowed to recover in a temperature regulated incubator (32°C) until fully awake and then were transferred to an isolated animal room for 72 hr and maintained without disturbance. All injected mice had 3 weeks to fully recover before being used for tissue collec-

| Tetanic muscle force testing
Young adult WT, aged WT, DMH-specific Sirt1 knockdown mice, and BRASTO with their control mice were used for measuring the evoked compound action potential (cMAP) in the extensor digitorum longus (EDL) muscle upon electrical stimulation of the sciatic nerve as described previously in detail (Santosa et al., 2013). Animals were anesthetized and immobilized in an automated functional station (FASt System, Red Rock laboratories, St. Louis, MO, USA). Twitch contractions were used to estimate the optimal stimulus amplitude and optimal muscle length for isometric force production in the EDL muscle, with the distal muscle tendon fixed to a 5 Newton (N) load cell. The EDL, instead of SM, muscle was utilized as it has a long tendon, which is required to facilitate the fixation to the load cell.
Tetanic contractions were recorded at increasing frequencies of stimulation (5-200 Hz), allowing 2-min intervals between stimuli to prevent muscle fatigue. Maximum isometric tetanic force was automatically calculated from the resulting sets of recorded force traces.
Following assessment, animals were euthanized, the dissected EDL muscle was weighed, and the tetanic specific muscle force (N/cm 2 ) was calculated by dividing the absolute muscle force by the physiological muscle cross-sectional area. A total of 3-8 animals per group were assessed.

| Statistical analyses
All data are presented as the mean AE SD. Significance is shown as *p < .05. Quantifications were performed from at least three experimental groups in a blinded fashion unless otherwise noted. Statistical analyses were performed by two-tailed, unpaired Student's t tests, or one-way ANOVA followed by a Bonferroni post hoc test. All analyses were performed using Microsoft Excel, GRAPHPAD PRISM 6.00 (San Diego, CA, USA; http://www.graphpad.com), and SPSS (SPSS Inc., Chicago, IL, USA).

ACKNOWLEDG MENTS
The Hope Center Viral Vectors Core is supported by the Hope Center

CONFLI CT OF INTEREST
The authors have no competing financial interests.

AUTHOR CONTRI BUTIONS
ASW designed and supervised the study, performed data analysis, and wrote the manuscript. AS assisted with experiments and data analysis. KBS performed muscle force testing and assisted with