Sustained NFκB inhibition improves insulin sensitivity but is detrimental to muscle health

Summary Older adults universally suffer from sarcopenia and approximately 60–70% are diabetic or prediabetic. Nonetheless, the mechanisms underlying these aging‐related metabolic disorders are unknown. NFκB has been implicated in the pathogenesis of several aging‐related pathologies including sarcopenia and type 2 diabetes and has been proposed as a target against them. NFκB also is thought to mediate muscle wasting seen with disuse, denervation, and some systemic diseases (e.g., cancer, sepsis). We tested the hypothesis that lifelong inhibition of the classical NFκB pathway would protect against aging‐related sarcopenia and insulin resistance. Aged mice with muscle‐specific overexpression of a super‐repressor IκBα mutant (MISR) were protected from insulin resistance. However, MISR mice were not protected from sarcopenia; to the contrary, these mice had decreases in muscle mass and strength compared to wild‐type mice. In MISR mice, NFκB suppression also led to an increase in proteasome activity and alterations in several genes and pathways involved in muscle growth and atrophy (e.g., myostatin). We conclude that the mechanism behind aging‐induced sarcopenia is NFκB independent and differs from muscle wasting due to pathologic conditions. Our findings also indicate that, while suppressing NFκB improves insulin sensitivity in aged mice, this transcription factor is important for normal muscle mass maintenance and its sustained inhibition is detrimental to muscle function.


Introduction
Aging is characterized by a deterioration in glucose homeostasis and the progressive loss of muscle mass and function. However, the molecular basis for insulin resistance and sarcopenia of aging is unknown. The transcription factor NFjB is a key regulator of inflammatory responses. Accumulating evidence suggests that the NFjB pathway may be involved in the aging process and in the pathogenesis of various aging-related pathologies. For example, upregulation of the NFjB pathway in aged animals and older humans has been demonstrated in multiple tissues such as skeletal muscle, liver, kidney, heart, and gastric mucosa (Helenius et al., 1996;Korhonen et al., 1997;Walter & Sierra, 1998;Xiao & Majumdar, 2000;Ghosh et al., 2015). In line with these findings, a global gene expression microarray analysis of various human and mouse tissues determined that the NFjB motif was the motif most strongly associated with aging (Adler et al., 2007).
NFjB is upregulated in muscle of both aged rodents (Phillips & Leeuwenburgh, 2005) and human subjects (Buford et al., 2010;Thalacker-Mercer et al., 2010;Ghosh et al., 2015), and the NFjB pathway has been implicated in the pathogenesis of insulin resistance and type 2 diabetes (Yuan et al., 2001). Therefore, the NFjB pathway may be involved in the glucose metabolism abnormalities that occur during aging. NFjB also is thought to play a key role in the muscle wasting seen with cancer, muscle disuse, and denervation (Cai et al., 2004;Judge et al., 2007;Van Gammeren et al., 2009;Reed et al., 2011). Thus, NFjB may also mediate the muscle loss characteristic of the aging process. Despite the evidence indicating that NFjB may be involved in the glucose metabolism abnormalities and muscle wasting seen with certain pathologic conditions (e.g., type 2 diabetes, cancer, denervation, and disuse), it is not known whether NFjB is a link between the pathogenesis of insulin resistance and the muscle atrophy of aging.
In this study, we tested the hypothesis that muscle-specific NFjB inactivation would protect against aging-induced insulin resistance and sarcopenia. If positive, these findings would demonstrate a role for NFjB on the pathophysiology of these metabolic disorders that frequently affect older adults. Such results also would provide proof-of-concept that interventions that target NFjB could be effective in the prevention and treatment of aging-related metabolic conditions.

Animal model
We studied C57BL/6 mice with muscle-specific overexpression of an IjBa super-repressor mutant (MISR) and wild-type (WT) littermates. Generation of these mice and genotyping procedures were described previously (Cai et al., 2004). This is a well-characterized model for studies of suppressed NFjB activity, which carries an IjBa transgene (S32A/S36A) that functions as a potent dominant negative inhibitor of the classical NFjB pathway, because mutating serines 32 and 36 for alanine prevents phosphorylation and dissociation of IjBa from NFjB p50/65. Expression of the transgene in skeletal muscle is driven by a muscle creatine kinase (MCK) promoter. Mice were genotyped by PCR, and overexpression of IjBa (nfkbia) gene in muscle was confirmed by RNA sequencing (Fig. S1, Supporting information). Mice were housed in an animal room maintained at 23°C with a 12-h light/12-h dark cycle and were provided standard laboratory chow and water ad libitum. We studied male and female mice of different ages; the ages and sexes of the mice are described in each experiment.

Muscle-specific NFjB inhibition protects against aginginduced insulin resistance
Intraperitoneal glucose tolerance testing (GTT) was performed in male and female mice of various ages (1-3 months, 12-13 months, and 20-22 months). Blood glucose levels during the GTT were similar in male WT and MISR in all age groups (Fig. S2). In 1-to 3-and 20-to 22-monthold female MISR mice, plasma glucose was modestly reduced compared with WT. A hyperinsulinemic euglycemic clamp procedure also was performed because it is more precise and sensitive than the GTT assay for quantitation of insulin-mediated glucose disposal in the periphery (which mainly reflects muscle glucose disposal). Clamp studies were performed in 3-and 18-month-old male mice. Both the glucose infusion rate (GIR) required to maintain euglycemia (Fig. 1A,B) and the M value ( Fig. 1C) were lower in 18-month-old WT mice than in 3-month-old WT mice by 40%. However, 18-month-old MISR mice had increases in both GIR and M compared with aged-matched WT mice by 29%, indicating improved insulin sensitivity. Consistent with this finding, Akt phosphorylation in muscle tissue harvested at the end of the insulin clamp was higher in 18-month-old MISR mice vs. age-matched WT mice (Fig. 1D). Endogenous glucose production was fully suppressed in all groups during insulin infusion (not shown). Blood glucose and plasma insulin concentrations during the clamp were similar between groups (Fig. S3). In the basal (noninsulin stimulated) state, Akt phosphorylation was not different between WT and MISR mice in any age group (Fig. S4).
Indirect calorimetry was performed in 3-, 13-, and 35-month-old male mice to assess the effects of aging and NFjB suppression on whole-body substrate utilization and energy expenditure. In both genotypes, aging led to modest reductions in oxygen consumption, carbon dioxide production, and resting metabolic rate (Fig. S5). The respiratory quotient also decreased in WT and MISR mice with aging, suggestive of more reliance on lipid oxidation as age advanced. There were no significant differences between genotypes in any of these metabolic parameters. Spontaneous activity during light and dark phases also tended to decrease with age in both genotypes, and aged MISR mice seemed to be less active in the dark phase compared to aged WT mice, although this difference did not reach statistical significance (Fig. S5). Overall, aging led to a subtle hypometabolic state in both genotypes, and NFjB suppression in muscle did not notably affect whole-body substrate utilization.

Altered content of lipid metabolites and carnitine in muscle from aged and MISR mice
Muscle content of ceramide, acylcarnitine, and diacylglycerol species was measured in mice from different age groups (3, 13, and 35 months old) as aging is associated with increased intramyocellular lipid accumulation (Petersen et al., 2003) and these lipid metabolites have been implicated in the pathogenesis of insulin resistance (Schooneman et al., 2013). In both genotypes, the muscle content of various ceramide species (C24, C24:1, and C26:1), total major ceramide (sum of C18, C18:1, C22, C24, and C24:1), and sphingolipids (Sph and dhSph) increased with aging ( Fig. 2A). Elevated content of these metabolites in aged mice was associated with increased mRNA levels of the sptlc1 gene that encodes  For panels E and F, n = 3-6 per group. †Genotype effect P < 0.05 by two-way ANOVA; §Age effect P < 0.05 by two-way ANOVA; *P < 0.05 by Tukey's post hoc test. All data are means AE SE. one of the two subunits of serine palmitoyltransferase (SPT) (Fig. 2B). This enzyme catalyzes the first and rate-limiting step in the de novo synthesis of sphingolipids. MISR mice of all age groups had lower C18, C18:1, and total major ceramide content, compared to WT ( Fig. 2A). The muscle content of C14, C16, C20, C20:1, C20:4, C22, C22:1, and C26 ceramides did not differ based on genotype or age group (data not shown).

Sustained inactivation of NFjB is detrimental to muscle health
The effect of NFjB suppression on muscle mass was first assessed in 3-, 13-, and 35-month-old male mice. Total and whole-body lean mass, measured by quantitative magnetic resonance (QMR), increased in both genotypes from 3 to 13 months ( Fig. 3A,B). As the mice aged further (35 months old), total mass and whole-body lean mass decreased in both genotypes. Consistent with these findings, wet weights of gastrocnemius, quadriceps, tibialis anterior, and soleus muscles declined with age progression in WT and MISR mice ( Fig. 3C-F). We hypothesized that NFjB inactivation would protect against aging-induced sarcopenia. However, against our hypothesis, whole-body lean mass was lower in MISR compared with WT mice (Fig. 3B). Moreover, the magnitude of the aging-induced decline in whole-body lean mass (Fig. 3B) and in the weights of some muscles (gastrocnemius and soleus) ( Fig. 3C-F) was higher in MISR than WT mice. As the reduction in muscle mass with NFjB inactivation was unexpected, we aged a second cohort of male and female mice and measured lean mass by QMR. In line with the findings from the first cohort, whole-body lean mass was lower in male MISR mice (Fig. S7A). In addition, reduced lean mass also was seen in female MISR mice (Fig. S7B).
Sarcopenia associated with aging manifests as decreases in muscle mass and neuromuscular function, but they do not necessarily change to the same degree over time. Thus, we tested the effects of NFjB inactivation on neuromuscular function by measuring grip strength, contractile force, and nerve conduction velocity in hindlimb (primarily gastrocnemius) muscles from male mice. Consistent with the muscle mass measurements, grip strength (combined forelimb and hindlimb) decreased with advancing age in both genotypes, and grip strength was 22% lower in aged (35-month-old) MISR compared with age-matched WT animals (Fig. 3G). Hindlimb force generation increased in both genotypes as mice matured from 3-6 to 12-18 months of age (Fig. 3H). Muscles of 3-to 6-month-old MISR mice generated 20% less force than age-matched WT mice and muscles of 12-to 18-month-old MISR mice generated 22% less force than age-matched WT mice. Sciatic nerve conduction velocity was assessed in 3-and 18-month-old mice. Taking together all groups, aging leads to a significant reduction in conduction velocity (Fig. 3I). Compared with 3-month-old mice from the same genotype, conduction velocity was 18% lower in 18-month-old WT mice and 23% lower in 18-month-old MISR mice. Sciatic nerve conduction velocity was not significantly different between WT and MISR mice.

NFjB inhibition alters the expression of several genes involved in muscle growth and atrophy
As NFjB modulates the expression of hundreds of genes, we conducted whole-genome RNA sequencing to facilitate the detection of alterations in gene expression that could explain reduced muscle size and strength in MISR mice. The total number of genes differentially regulated in WT vs. MISR was 659 (P-value <0.05); 228 genes changed in 3-to 6-, 197 in 12-to 18-, and 424 in 33-to 36-month-old mice (Figs 4A, and S8A for differentially expressed genes selected with adjusted P-value <0.05). Figure 4B shows the top signaling pathways involved in muscle growth/ development differentially expressed during aging in both WT and MISR mice and pathways differentially expressed between genotypes. A heatmap with all pathways/functional gene sets is shown in Fig. S8B. In addition to analyzing changes in gene sets, we focused on specific genes involved in muscle growth and atrophy. NFjB suppression resulted in significantly elevated expression of mstn (myostatin) (Fig. 4C), a potent inhibitor of muscle growth. NFjB inhibition also led to higher expression of insulin-like growth factor binding protein (IGFBP)5, which gene product suppresses muscle growth (Salih et al., 2004) (Fig. 4D). Finally, MISR mice had lower expression of fibroblast growth factor binding protein (Fgfbp)1, which plays a key role in neuromuscular junction (NMJ) formation and function (Williams et al., 2009) (Fig. 4E).
Altered expression of genes that modulate muscle progenitor cell migration, differentiation, and fusion NFjB suppression resulted in higher expression of myod1 (MyoD) (Fig. 4F), a master regulator of myogenesis that promotes muscle cell differentiation. NFjB suppression also led to changes in expression of genes that encode for proteins within the noncanonical NFjB pathway, which promotes skeletal muscle cell differentiation and fusion (Enwere et al., 2012). Changes included higher mRNA levels of map3k14 (NFjBinducing kinase; NIK) ( Fig. 4G) and Tbkbp1 (Fig. 4H), and lower mRNA levels of Traf2 (Fig. 4I) and Traf3 (Fig. 4J). Lastly, MISR mice of all age groups had lower mRNA levels of c-met (Fig. 4K), a gene that encodes MET, a tyrosine kinase receptor essential for muscle progenitor cell delamination and migration.

In vitro NFjB inhibition accelerates muscle cell differentiation
Adenoviral-mediated overexpression of an IjBa super-repressor (SR) mutant (S32A/S36A) in C2C12 myoblasts recapitulated the in vivo changes in expression of some genes and proteins involved in muscle differentiation, growth, and atrophy, including enhanced myostatin gene expression (Fig. 5A). NFjB suppression with the IjB-SRa mutant also led to increased NIK (Fig. 5B) and lower TRAF2 (Fig. 5C) protein levels, reflecting increased flux through the noncanonical NFjB pathway. Inhibition of NFjB in C2C12 myoblasts promoted early differentiation, shown by increased expression of fast myosin (SKM) heavy chain at earlier time points (Fig. 5D). C2C12 cell proliferation was not affected by the SR mutant (Fig. S9).

Elevated proteasome activity in MISR mice
As proteasomes are the primary sites for protein degradation in muscle, we measured activity of 26S (the most abundant proteasome) and 20S (its proteolytic core) in muscle from WT and MISR mice. The activities of 26S and 20S were increased in MISR mice, and these differences were more pronounced in young (3-to 6-month-old) and middle-aged (12-to 18-month-old) mice (Fig. 6A,B). In parallel, the total content of proteasomal protein represented by the a4 subunit of the 20S proteasome core also was higher in muscle from MISR mice (Fig. 6C). In 3-to 6-and 12-to 18-month-old MISR mice, elevated proteasome activity was associated with a trend (P = 0.07) for higher expression of psme4 (Fig. 6D), which encodes proteasome activator (PA)200. Transcript levels of the E3 ubiquitin ligases fbxo32 (atrogin) and trim63 (MuRF1) were not significantly different between age and genotype groups (Fig. 6E,F). Gene expression levels of several proteasomal   Table S2). (C-K) mRNA levels of mstn, igfbp5, fgfbp1, myod1, map3k14 (NIK), tbkbp1, traf2, traf3, and c-met in quadriceps muscle from WT (white) and MISR (black) mice. n = 5-6 per group. Data analyzed by two-way ANOVA; *P < 0.05 by Tukey's post hoc test. All data are means AE SE.
Role of NFjB in aged muscle, N. Zhang et al. subunits increased with aging, but there was no difference between WT and MISR mice (Fig. S10). In addition to the proteasome, the autophagylysosome pathway is another major regulator of muscle protein degradation. We found that the expression of several genes in the autophagy-lysosome pathway-Foxo3, Ctsl, Lamp2, map1lc3a (LC3A), map1lc3b (LC3B), Atg3, and Atg7-were increased with aging, whereas map1lc3a (LC3A) expression was higher in MISR mice (Fig. S11A-G). Consistent with the increase in autophagy genes, aging (in WT and MISR) resulted in higher LC3-II/LC3-I ratio, which is another marker suggestive of increased autophagy (Fig. S11H).

Discussion
Multiple lines of evidence indicate that NFjB is activated during aging, and suggest that inhibition of this pathway may be effective in slowing down aging and age-related diseases. For example, Adler et al. identified NFjB as a candidate driver of aging-related transcriptional changes in multiple human and mouse tissues (Adler et al., 2007) and showed that genetic blockade of NFjB protected the skin from the effects of aging (Adler et al., 2008). Zhang et al. reported that hypothalamic-specific overexpression of the super-repressor IjBa mutant in mice had global anti-aging effects, including prolonging lifespan, improving cognition, and increasing dermal thickness, bone density, muscle size, exercise endurance, and tendon resistance, whereas overexpression of constitutively active IKKb had opposite effects (Zhang et al., 2013). Thus, we predicted that suppression of NFjB in muscle would be protective against aging-induced insulin resistance and sarcopenia, two common age-related disorders. However, while NFjB inhibition improved insulin resistance in aged mice, it did not slow down aging-related muscle loss and had a detrimental effect on muscle mass and strength.
The sarcopenic phenotype in MISR mice was unexpected because sustained stimulation of the classical IKK-NFjB pathway through robust (10-15 fold) overexpression of IKKb leads to severe muscle atrophy (Cai et al., 2004). In addition, NFjB is thought to protect muscle from various atrophy-inducing insults. Cai et al. reported that muscle from MISR mice did not become atrophic when subjected to denervation and cancer (Cai et al., 2004). Protection from denervation, unloading, immobilization, endotoxin administration, and cryoinjury also have been reported using different models of suppressed NFjB signaling (Hunter & Kandarian, 2004;Mourkioti et al., 2006;Judge et al., 2007;Van Gammeren et al., 2009;Haegens et al., 2012;Okazaki et al., 2014;Oh et al., 2016). The resistance conferred by NFjB axis suppression to these atrophy-inducing stimuli indicates that NFjB has opposing effects (both negative and positive) on muscle development and health, depending on the level of flux (i.e., dosage) through the pathway. Our findings also indicate that the mechanism behind the muscle loss seen with aging differs from muscle wasting in other pathologic conditions (cancer, denervation, disuse, etc.). In line with this notion, aging did not upregulate the expression of atrogin and MuRF1, ubiquitin ligases implicated in the atrophy caused by denervation and cancer (Bodine et al., 2001;Cai et al., 2004;Sacheck et al., 2007). While the basis for aging-induced sarcopenia is not fully understood, our results point toward alterations in the autophagy-lysosome pathway, which is another key cellular mechanism involved in protein degradation (Wohlgemuth et al., 2010). Our global transcriptomics approach to facilitate identification of potential molecular changes associated with decreased muscle mass revealed that MISR mice had altered expression of many genes that impact muscle (Summarized in Fig. S12), indicating that the atrophy is multifactorial. MISR mice had elevated levels of myostatin, a member of the TGFb protein family that is a potent inhibitor of muscle growth (McPherron et al., 1997). MISR mice also had elevations in igfbp5 and reductions in c-met gene expression. IGFBP5 is a ubiquitously expressed IGF binding protein with high affinity for IGF1, a growth factor essential for muscle mass attrition. Transgenic mice overexpressing IGFBP5 suffer from growth retardation and muscle atrophy, due to suppressed IGF1 activity (Salih et al., 2004). MET (encoded by c-met) is a tyrosine kinase receptor that interacts with its ligand hepatocyte growth factor (also known as scatter factor) and produced by nonsomitic mesodermal cells to delineate the migratory route of muscle progenitor cells. The importance of MET in muscle development is underscored by findings that mouse embryos lacking functional MET have underdeveloped skeletal muscle tissue in the limbs (Bladt et al., 1995). MISR mice of all ages also had a significant reduction in the mRNA level of fgfbp1, which plays a critical role in cell differentiation and migration by binding to fibroblast growth factors and potentiating their biological effect (Williams et al., 2009). FGFBP1 also promotes normal development and function of the NMJ (Williams et al., 2009).
MISR mice had enhanced expression of MyoD, a myogenic factor that is a key driver of muscle differentiation. The increase in MyoD is in line with work from Guttridge et al. that have reported that NFjB downregulates MyoD expression (Guttridge et al., 2000). MISR mice also displayed changes in genes that encode for proteins within the noncanonical NFjB pathway, including increases in map3k14 (NIK) and Tbkbp1, and decreases in TRAF2 and TRAF3. NIK is a central component of the noncanonical NFjB pathway that integrates signals from TNF receptor family members. NIK is subjected to a fate control mechanism, such that the basal level of NIK is kept low by the TRAF proteins. Therefore, a decrease in TRAF2 and TRAF3, as seen in MISR mice, would be consistent with increased flux via this pathway. The noncanonical NFjB pathway promotes fusion (Enwere et al., 2012;Hindi et al., 2013), is elevated in muscle cells from older human subjects (Urban et al., 2014), and has been associated with muscle wasting (Li et al., 2009). Moreover, the findings from the cell culture experiments indicate that NFjB suppression accelerates myoblast differentiation in vitro. Thus, it is possible that increased expression of MyoD and enhanced flux via the noncanonical NFjB pathway may lead to premature differentiation and fusion in vivo. While gene expression levels of myosin heavy chains were not different between genotypes (Fig. S13), a more detailed analysis of muscle cell differentiation across the lifespan in MISR mice in vivo will be required to firmly establish whether premature fusion contributed to the decrease in muscle mass. Lastly, NFjB is thought to control satellite cell function and their dysfunction during development could have resulted in reduced muscle mass. Yet, this possibility seems less likely because myogenic colony formation efficiency and satellite cell frequency are not affected in young MISR mice (Oh et al., 2016).
In addition to the altered expression of several genes involved in muscle development, growth, and atrophy, muscle from MISR mice had elevated proteasome activity. One mechanism behind the increase in activity is the elevation in total content of protein of the proteasome catalytic core, present in all enzymatically active proteasome assemblies. Another potential mechanism behind the increase in proteasome activity in MISR mice is enhanced expression of psme4 (PA200), which binds to proteasomes to stimulate the hydrolysis of peptides (Ustrell et al., 2002). It is also possible that NFjB suppression leads to changes in the abundance of distinct forms of the proteasomes, each with specific activities and functions, from the 26S super-assemblies to the free 20S cores.
The insulin clamping experiments uncovered insulin resistance in aged WT mice that was not readily apparent upon glucose tolerance testing. Increased intramyocellular ceramide content likely contributed to reduced insulin action in aged mice, as ceramides directly inhibit Akt. The elevation in muscle ceramide content was accompanied by an agemediated increase in the level of one of the two subunits of SPT, the rate-limiting enzyme that controls de novo synthesis of sphingolipids. Moreover, NFjB suppression led to decreases in ceramide and acylcarnitine content, which likely contributed to the improvements in insulin action. Aged MISR mice also had lower level of the acylcarnitine precursor carnitine, likely as a result of lower expression of aldh9a1 (TMABA-DH), a key carnitine biosynthetic enzyme. As variations in muscle fiber types also can affect glucose metabolism and insulin action, we compared fiber-type distribution between groups. WT and MISR mice did not have significant differences in fiber-type distribution, whereas aging led to a small increase in the proportion of IIa/x vs. IIb fibers (Fig. S14).
Another unexpected finding in this study was the dissociation between muscle mass and insulin action in MISR mice. That is, MISR mice had improved insulin sensitivity despite having lower muscle mass. Insulin action/sensitivity modulates muscle mass through mTOR-regulated protein synthesis. Conversely, sarcopenia is thought to promote insulin resistance (reduced insulin-mediated whole-body glucose disposal) because muscle tissue is the main site responsible for glucose disposal, and sarcopenia promotes physical inactivity. Improved insulin sensitivity in MISR mice may have been a compensatory mechanism in response to the negative effects of NFjB suppression on muscle mass. It is also possible that insulin-mediated glucose disposal and muscle mass control are not inextricably linked.
NFjB suppression resulted in altered expression of multiple genes and this finding was anticipated because this transcription factor is at the nexus of numerous signaling pathways. Changes in gene expression in MISR mice are likely caused by direct and indirect mechanisms. For example, the 5 0 -regulatory region of myostatin has the consensus sequence of the NFjB binding site (Ma et al., 2001). Thus, based upon our results, NFjB appears to function as a transcriptional repressor of the myostatin gene. NFjB also has been reported to suppress fgfbp1 promoter activity (Rosli et al., 2013). In addition to these documented reports of transcriptional regulation by NFjB, the promoter regions of several genes altered in MISR mice have predicted NFjB binding motifs, including slc22a5, igfbp5, map3k14 (NIK), tbkbp1, traf2, and foxo3. Besides direct transcriptional regulation by NFjB, gene expression changes in MISR mice also could be mediated via indirect mechanisms. For example, MISR mice had increased expression levels of the transcription factor myod1 and the promoter regions of some genes altered in MISR mice have predicted myod1 binding sites, including slc22a5, traf2, map1lc3a, and map1lc3b. The c-met promoter region is regulated by peroxisome proliferator-activated receptor (PPAR)c (Kitamura et al., 1999), a transcription factor/nuclear receptor that can interact directly with NFjB (Chung et al., 2000). The aldh9a1 (TMABA-DH) (Wen et al., 2012) promoter also is modulated by PPARa, which cross talks with NFjB (Delerive et al., 1999). Various other genes altered in MISR mice have predicted PPARc (slc22a5,myod1,igfbp5,map3k14,tbkbp1,traf2,foxo3,lamp2, and map1lc3a and map1lc3b) and PPARa (slc22a5, myod1, igfbp5, map3k14, and tbkbp1) binding motifs.
In conclusion, sustained NFjB suppression confers some beneficial effects on glucose homeostasis (i.e., insulin sensitivity), likely due to reduced levels of ceramides and acylcarnitines in muscle. However, our findings indicate that NFjB plays an important role in muscle development and maintenance throughout the lifespan. Moreover, we demonstrate that sustained (lifelong) NFjB suppression has harmful effects on muscle mediated by increased proteasome activity and altered expression of several genes involved in muscle growth and atrophy. Future research could test whether a more subtle intervention to block this pathway (e.g., pharmacological), applied at a different time in life, or for less time (e.g., inducible genetic approach), can improve muscle function without impairing its growth.

Experimental procedures MISR mice
Generation of these mice and genotyping procedures were described previously (Cai et al., 2004). All procedures were approved by the Institutional Animal Care and Use Committee at the University of Texas Health Science Center at San Antonio (UTHSCSA).

Intraperitoneal glucose tolerance test
Mice were injected intraperitoneally with dextrose (1.5 g kg À1 ). Glucose concentrations were measured in tail blood at times 0, 30, 60, 90, and 120 min using an automatic glucose meter (Roche Diagnostics, Indianapolis, IN, USA). Food was removed 8-10 h before the experiments.

Hyperinsulinemic, euglycemic clamp studies
Under anesthesia with inhaled isoflurane, a catheter was inserted into the right atrium of the heart through the jugular vein. Three to five days later, a 90-min euglycemic hyperinsulinemic (18 mU kg À1 min À1 ) clamp with tritiated glucose was performed in fasting mice, as described previously (Liang et al., 2009). At the end of the insulin clamp, quadriceps muscles were removed and snap-frozen for insulin signaling assays. Plasma insulin concentrations were measured using an ELISA kit (Crystal Chem Inc, Downers Grove, IL, USA). The M value was calculated as the GIR minus the glucose space.

Muscle mass measurement
For muscle tissue harvesting, mice were anesthetized (80 mg kg À1 ketamine and 10 mg kg À1 xylazine, i.p.) and euthanized by cervical dislocation. Individual muscles were carefully dissected and weighed to 0.1 mg accuracy.

Grip strength
The combined strength of forelimbs and hindlimbs was measured using a Grip Strength Meter with mesh grid pull bar (Columbus Instruments 1027 CSM, Columbus, OH) specifically designed for mice. Mice were allowed to grasp the pull bar with forelimbs and hindlimbs and were then gradually pulled backward in a horizontal plane until they lost their grip. Maximum grip strength is recorded on the device. Mice are not trained prior to testing. The highest value from five consecutive trials is designated the mouse's maximum grip strength.

Contractile force
Maximal isometric torque of the posterior crural muscles (gastrocnemius, soleus, plantaris) was measured in vivo in 3-to 6-and 12-to 18-month-old male mice. Mice were anesthetized using inhaled isoflurane. Mice rested in a prone position on a heated (37°C) platform. The knee was fixed with clamps with the foot attached to a footplate located on the shaft of a servomotor that measures changes in force (model 1300A; Aurora Scientific, Aurora, ON, Canada). Two stainless steel subdermal electrodes were inserted subcutaneously in the body of the posterior crural muscles. Peak isometric torque was measured in response to voltage delivered by an electrical stimulator (model 701C; Aurora Scientific) at increasing frequencies (20,40,60,80,100,and 150 Hz).

Proteasome activity
Muscle tissues were homogenized in ice-cold proteasome homogenization buffer (50 mM Tris-HCl, pH7.5, 0.05% NP-40, 0.25 mM sucrose, 1 mM DTT, 2 mM ATP, 5 mM MgCl 2 ), followed by centrifugation at 10,000 g for 20 min at 4°C. Supernatant was transferred to a new tube, and protein concentrations were determined; 40 lg of proteins per sample per assay was used in all assays, which were performed in triplicates and repeated at least once. Chymotrypsin-like proteasome activity was measured in reaction buffer which contained 40 lM of substrate Suc-LLVY-AMC (EMD, Gibbstown, NJ, USA) in 50 mM Tris-HCL, pH 7.5, 0.25 mM sucrose, 1 mM DTT, and 5 mM MgCl 2 . The proteasome inhibitor MG-132 at 1 lM (Sigma-Aldrich, St. Louis, MO) was included in a separate reaction to monitor nonspecific background reactions. For 20S chymotrypsin-like activity assay, a final concentration of 0.03% SDS was included in the reaction buffer. For 26S activity assay, 46 lM ATP was added into the reaction mixture without SDS. The reaction was monitored for 120 min at 37°C by measuring the fluorescence intensity in a 96-well plate reader at the following wave lengths: excitation = 380 nm, emission = 440 nm. Proteasome activity at the 60-min time point was calculated by subtracting nonspecific background values obtained in the presence of inhibitors from those without inhibitors.

C2C12 cell culture experiments
Cell culture experiments were carried out in 6-well plates with growth media (DMEM with 10% FBS and 1% penicillin/streptomycin) seeded with 2 9 10 4 cells well À1 . C2C12 myoblasts at 70% confluence were transduced for 12 h with 10 8 IU well À1 Ad-IjBa-SR or Ad-GFP (Vector Biolabs, Philadelphia, PA, USA). We have confirmed the ability of this adenoviral vector to inhibit NFjB signaling in cultured muscle cells (Reyna et al., 2008). To induce differentiation, the media was then changed to DMEM with 2% FBS. For measurement of fast myosin skeletal heavy chain over time, cells were differentiated with 2% HS. Cells were harvested 3 days later with lysis buffer (20 mM Tris, pH 7.5, 5 mM EDTA, 10 mM Na 3 PO 4 , 100 mM NaF, 2 mM Na 3 VO 4 , 1% Nonidet P-40, 10 lM leupeptin, 3 mM benzamidine, 10 lg mL À1 aprotinin, and 1 mM phenylmethylsulfonyl fluoride) or TRIzol for protein content and mRNA expression measurements, respectively. Myoblast proliferative capacity was assessed using a CellTiter 96 AQueous One Solution Proliferation Assay (Promega, Sunnyvale, CA, USA). Cells were seeded at 2000 cells well À1 and transduced in 2% FBS differentiation media with Ad-10 7 IU well À1 IjBa-SR or Ad-GFP for 12 h before being switched to growth media for measurements of cell proliferation over time.

Statistical analysis
All data are presented as the mean AE SE. Comparisons of means between groups were done by two-way ANOVA followed by the Tukey's test or unpaired two-tailed t-test, as specified in the figure legends. GraphPad Prism 7 (GraphPad Software, La Jolla, CA) was used for statistical analyses.

Other methods
The Supplemental Information section provides detailed descriptions of the indirect calorimetry and spontaneous activity measurements, nerve conduction studies, metabolomics, real-time PCR, Western blotting, fiber typing, capillary electrophoresis, RNA sequencing procedures, and gene expression data analyses.

Supporting Information
Additional Supporting Information may be found online in the supporting information tab for this article.              Table S1 (a) Differentally expressed genes (DEGs) comparing MISR/WT at 3-6 months of age. (b) Differentally expressed genes (DEGs) comparing MISR/ WT at 12-18 months of age. (c) Differentally expressed gene (DEGs) comparing MISR/WT at 33-36 month of age.

Table S2
Gene lists corresponding to Gene Ontology and IPA pathways/gene sets differentially regulated by aging and NFkB suppression (from Figure 4B).
Data S1 Procedures and Materials.