Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You and Your Team.

Learn More →

Complementary NAD+ replacement strategies fail to functionally protect dystrophin-deficient muscle

Complementary NAD+ replacement strategies fail to functionally protect dystrophin-deficient muscle Background: Duchenne muscular dystrophy (DMD) is a progressive muscle wasting disorder stemming from a loss of functional dystrophin. Current therapeutic options for DMD are limited, as small molecule modalities remain largely unable to decrease the incidence or mitigate the consequences of repetitive mechanical insults to the muscle during eccentric contractions (ECCs). Methods: Using a metabolomics-based approach, we observed distinct and transient molecular phenotypes in muscles of dystrophin-deficient MDX mice subjected to ECCs. Among the most chronically depleted metabolites was nicotinamide adenine dinucleotide (NAD), an essential metabolic cofactor suggested to protect muscle from structural and metabolic degeneration over time. We tested whether the MDX muscle NAD pool can be expanded for therapeutic benefit using two complementary small molecule strategies: provision of a biosynthetic precursor, nicotinamide riboside, or specific inhibition of the NAD-degrading ADP-ribosyl cyclase, CD38. Results: Administering a novel, potent, and orally available CD38 antagonist to MDX mice successfully reverted a majority of the muscle metabolome toward the wildtype state, with a pronounced impact on intermediates of the pentose phosphate pathway, while supplementing nicotinamide riboside did not significantly affect the molecular phenotype of the muscle. However, neither strategy sustainably increased the bulk tissue NAD pool, lessened muscle damage markers, nor improved maximal hindlimb strength following repeated rounds of eccentric challenge and recovery. (Continued on next page) * Correspondence: eugene.l.stewart@gsk.com Computational Sciences, Molecular Design, GlaxoSmithKline R&D, Collegeville, PA, USA Full list of author information is available at the end of the article © The Author(s). 2020 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://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Frederick et al. Skeletal Muscle (2020) 10:30 Page 2 of 14 (Continued from previous page) Conclusions: In the absence of dystrophin, eccentric injury contributes to chronic intramuscular NAD depletion with broad pleiotropic effects on the molecular phenotype of the tissue. These molecular consequences can be more effectively overcome by inhibiting the enzymatic activity of CD38 than by supplementing nicotinamide riboside. However, we found no evidence that either small molecule strategy is sufficient to restore muscle contractile function or confer protection from eccentric injury, undermining the modulation of NAD metabolism as a therapeutic approach for DMD. Keywords: MDX, NAD+, CD38, NR, Eccentric, Injury, Metabolomics, Therapeutics Background the buffer of isolated MDX muscles was reported to im- Dystrophinopathies are a class of diseases manifesting pri- prove fatigue resistance ex vivo [6]. In dystrophic mice har- marily in skeletal muscle and caused by a variety of muta- boring mutations in the FKRP gene, aberrant glycosylation tions in the 2.4 Mb dystrophin gene which render the of the DGC component, alpha-dystroglycan, can be par- dystrophin protein inactive. Dystrophin is a central force- tially overcome by supplementing the drinking water with transducing element of skeletal muscle, connecting the 5% ribitol, the substrate of the mutated enzyme [7]. These actin cytoskeleton to the extracellular matrix via the studies suggest that specific aspects of the pathophysiology dystrophin-glycoprotein complex (DGC). In patients with of muscular dystrophies arise from metabolic bottlenecks Duchenne muscular dystrophy (DMD), the absence of that may be bypassed using exogenous small molecules. functional dystrophin leads to limb muscle weakness, Nicotinamide adenine dinucleotide (NAD or NAD) is an followed by gradual muscle atrophy, cardiomyopathy, and essential metabolic co-factor, which has been directly impli- premature death. Dystrophin-deficient muscle is especially cated in the maintenance of muscle mass and function dur- susceptible to damage following eccentric contractions, in ing sarcopenia and other dystrophic states [8–11]. NAD has which the muscle lengthens while generating opposing also been found to be depleted in the muscle of MDX mice force, as occurs during the act of sitting, descending stairs, [10, 12], suggesting potential for therapeutic intervention. and other activities of daily life [1]. Damaged muscle fibers At present, two complementary strategies exist for manipu- then undergo repeated cycles of clearance by the immune lating tissue NAD pools. The first, and most commonly system and replacement by newly differentiated progeni- studied strategy, is to enhance NAD production by supple- tor cells, in a process that spans weeks. Though progress menting biosynthetic precursors, such as nicotinamide ribo- has been made in addressing the primary defects in dys- side (NR) or nicotinamide mononucleotide, to cells with a trophin via exon skipping or gene-replacement therapies functional NAD salvage pathway [13]. The second strategy [2], means of mitigating the effects of eccentric injuries to is to inhibit the enzymatic consumption of NAD by several the muscle of DMD patients have primarily been limited classes of enzymes, including sirtuins, poly-ADP-ribose to treatment with palliative anti-inflammatory drugs. polymerases (PARPs), and ADP-ribosyl cyclases (ARCs). The ability of muscle to harness chemical energy through CD38, the most widely expressed ARC, has been therapeut- aerobic and anaerobic respiration is inherently linked to its ically targeted for indications ranging from multiple mye- physical structure. Accordingly, analytical techniques such loma to neurodegeneration [14]. Our group has previously as NMR and mass spectrometry can detect chemical bio- shown that synthetic small molecule antagonists of CD38 markers of specific muscular dystrophies, which correlate are capable of acutely increasing NAD in muscle and liver with the cause and severity of the disease, and can aid in tissue [15, 16], and others have reported that chronically identifying points of therapeutic intervention. For example, dosing one of these thiazoloquinolin(on)es, known in the lit- it has long been appreciated that the muscle of adult erature as compound 78c, mitigated structural remodeling dystrophin-deficient MDX mice contains lower levels of and functional decline in the muscle of aged mice [11]. Des- energy-storing and redox-active metabolites, such as pite these preliminary findings, the extent to which NAD phospho-creatine, ATP, and beta-hydroxybutyrate [3], yet depletion is a pathologically relevant or therapeutically tract- high levels of taurine and non-polar amino acids [4]. More able feature of dystrophinopathies remains unclear. recently, distinct metabolomes were also identified in popu- Recent reports have specifically investigated the poten- lations of the MDX muscle-resident cells involved in regen- tial for NAD supplementation to counteract the pathology eration, including satellite cells and adipose progenitors [5]. of DMD [10, 17]. Here, we have examined this topic fur- Attempts to preserve muscle function by restoring specific ther by characterizing the global metabolome of MDX metabolic intermediates or co-factors in dystrophic muscle muscle following eccentric challenge and identifying dis- have shown hints of efficacy. For example, supplementing tinct stages of metabolic crisis and repair on the biochem- the TCA cycle intermediate, alpha-ketoglutarate (aKG), to ical level. We found evidence that NAD is depleted both Frederick et al. Skeletal Muscle (2020) 10:30 Page 3 of 14 Fig. 1 The metabolome of dystrophin-deficient muscle acutely responds to eccentric challenge. a Principal component analysis of the global metabolic profiles of MDX gastrocnemius muscles before and after eccentric contractions. Unchallenged WT muscles were included as a negative control. Arrows indicate the early chronological progression of injured groups and ellipses indicate 95% confidence intervals. b Heat map of metabolites with significantly different ion abundances in MDX muscle compared to WT (left) and sequential post-injury time points in MDX mice compared to uninjured controls (right). c Volcano plot of biochemical pathways significantly altered in MDX muscle at baseline. Significantly altered metabolites were used to calculate pathway impact and highly affected pathways are indicated. d Fold change of pathway impact scores during the acute (2 h) injury stage of MDX muscle compared to MDX baseline. e Volcano plot of metabolites significantly altered in MDX muscle at baseline. Example components of the NAD metabolome (red), glycolytic intermediates (blue), and polyamine pathway intermediates (green) are indicated. Dashed line indicates the significance threshold of non-adjusted p < 0.05. f Time course box and whisker plots of selected components of the NAD metabolome, glycolytic intermediates (g), and polyamine pathway (h) over 14 days post-challenge. N = 6 mice per time genotype and time point. Whiskers represent minimum and maximum values. Significance was determined by one-way ANOVA with Tukey’s post hoc test (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant; all p values non-adjusted) versus unchallenged WT (red) or MDX (blue) samples acutely and chronically in the muscle of MDX mice with metabolites or lipids. Principal component analysis of the wide-ranging metabolic consequences. We further metabolic profiles demonstrated a clear clustering of sam- attempted to reverse the NAD depletion using a novel, ples by time and genotype (Fig. 1a). Furthermore, the highly potent imidazoquinoline inhibitor of CD38 and MDX samples showed a well-defined chronological pro- compared its efficacy to nicotinamide riboside during re- gression following damage. The clustering of samples ob- peat bouts of eccentric challenge and recovery. Our results served 2 h post-damage showed distinct separation from may guide the development of NAD-targeting therapeu- the other time points, suggesting a period of acute crisis in tics for muscle diseases. the muscles. This was the only time at which the majority of significantly altered metabolites appeared depleted Results when compared to unchallenged controls (Fig. 1b), Dystrophin deficiency alters the muscle NAD metabolome potentially indicating rapid degradation, release from and energy producing pathways the tissue, or a synthetic bottleneck. The pattern of the To assess the biochemical adaptations of muscle to dys- post-injury sample groups from days 2-7 reflected a trophin deficiency, we analyzed the gastrocnemius mus- transition from the acute response to a repair and cles of MDX mice before and after eccentric challenge recovery stage. By post-injury day 14, the samples had a using an untargeted metabolic profiling platform. The chemical phenotype closely resembling that of the un- platform utilized a combination of GC/MS and LC/MS in injured MDX muscles (Fig. 1a-b). positive and negative ion mode to identify 762 chemical Comparing the metabolic profiles of MDX mice to entities, of which 552 were annotated as either polar wildtype (WT) mice at baseline, we identified several Frederick et al. Skeletal Muscle (2020) 10:30 Page 4 of 14 biochemical pathways with a disproportionate impact on the level of GAPDH, resulting in a characteristic buildup the muscle. Among the highest confidence pathways (p of intermediates in the pentose phosphate pathway [8, < 10^ −5 and pathway impact> 0.5) were those relating 18]. Consistent with this model, we observed a signifi- to the biosynthesis and metabolism of amino acids, such cant increase in ion counts for glucose-6-phosphate as arginine, phenylalanine, tyrosine, and tryptophan, as (G6P), dihydroxyacetone phosphate (DHAP), ribose-5- well as that of nicotinate and nicotinamide metabolism phosphate, and a positive trend in sedoheptulose-7- (Fig. 1c). Comparing the pathway impact scores from phosphate (S7P) in MDX muscle (Fig. 1e, g). Addition- the acutely injured (2 h) to uninjured MDX muscle, we ally, we noted an increase in several poly-cationic species were surprised to find glycolysis as the most responsive of polyamines, known to be derived from arginine (Fig. pathway (Fig. 1d). Additional pathways relating to energy 1e, h). Collectively, this pattern indicates a metabolic production or storage, including those of fatty acid syn- shift in MDX muscle at a steady state, partially stem- thesis, pyruvate metabolism, and the TCA cycle, were ming from the loss of the metabolic co-factor, NAD. also injury responsive. The robust response of the global metabolome at 2 h Upon examining the specific metabolites altered at post-eccentric challenge led us to investigate NAD-related baseline in the MDX muscle, we found a strikingly lower metabolites at this and subsequent time points. As far as the abundance of NAD than almost any other metabolite platform could resolve, NAD itself did not appear to re- (Fig. 1e, f). Consistent with the pathway analysis, several spond to the challenge (Fig. 1f). However, Nam was acutely other nicotinamide-containing metabolites and glycolytic depleted by more than one-third after 2 h, presumably intermediates also showed highly variable abundance. restricting any residual activity of the NAD salvage pathway, Primary NAD deficiency in mouse muscle and cultured and gradually returned to baseline after 4 days. Nam myotubes has been shown to restrict glycolytic flux at homeostasis was further altered by a doubling in the levels Fig. 2 In vivo characterization of a novel synthetic CD38 inhibitor. a Inhibition of the base-exchange activity of recombinant human (rh) or mouse (rm) CD38 protein using the established CD38 inhibitor 78C and the novel inhibitor GSK978A. Sigmoidal dose-response curves were fitted to replicate experiments (n = 12-15). Molecular structures and calculated pIC50 values are indicated. b Stability of 500 nM GSK978A in liver microsomal fractions from selected preclinical species over time. Half-lives (t ) were calculated from linear regressions. N = 3 experiments per 0.5 species. c Pharmacokinetics of GSK978A in mouse blood after an oral dose of 10 mg/kg. The dashed line indicates approximate IC50 of 12 ng/mL. N = 3 mice. d Intratissue compound exposures 2 h after a range of oral doses of GSK978A. N = 4 mice per dose. WAT, white adipose tissue. e Tissue-specific pharmacodynamics measured as NAD content relative to an internal standard (IS) 2 h after 0-30 mg/kg oral doses of GSK978A. Significance was determined by one-way ANOVA with Tukey’s post hoc test (*p < 0.05, ****p < 0.0001 compared to vehicle controls). f Tissue- specific pharmacodynamics as a function of exposure measured as normalized change in NAD content 2 h after 0-3 mg/kg doses of GSK978A. Lines indicate sigmoidal best-fit regressions approaching the indicated Bmax limits. N = 4 mice per dose. Mice were WT obese males for subpanels c-f. Error bars represent SEM Frederick et al. Skeletal Muscle (2020) 10:30 Page 5 of 14 of 1-me-Nam in the 2 days following injury, which only nor- GSK978A was more soluble and outperformed 78c in a malized after 14 days. Evidence of a further constriction in chromosomal stability test of genotoxicity, indicating im- glycolysis also emerged post-injury: levels of DHAP, imme- proved suitability for long-term administration (data not diately upstream of GAPDH, acutely increased in an oppos- shown). GSK978A was also predicted to have low intrinsic ing pattern to that of lactate, a glycolytic end-product (Fig. clearance in several small animal preclinical species, includ- 1g). The most striking indicator of the repair phase was the ing mice and rats, but not larger cynomolgus monkeys (Fig. appearance of polyamines, including spermine, spermidine, 2b). As literature suggested a role for CD38 in the preven- putrescine, and N-acetylputrescine, which were elevated in tion of diet-induced obesity [21], the drug metabolism and the days following injury (Fig. 1h and Supplemental Table pharmacokinetic characterization of the quinoline series 1). This class of biomolecules serves as a general marker of was originally performed in obese WT mice. To confirm cellular proliferation and is required for both myocyte differ- the slow clearance kinetics in vivo, we administered a single entiation and alternative macrophage activation [19, 20]. intermediate oral dose of 10 mg/kg and found the com- Consistently, in the case of NAD-related metabolites, glyco- pound still detectable in the blood after 24 h (Fig. 2c). We lytic intermediates, and polyamines, eccentric injury ampli- next performed a pharmacokinetic analysis of tissues sam- fied the disparities between MDX and WT muscle. pled 2 h after oral doses from 1-30 mg/kg. At these doses, the compound was identified within the liver, gastrocne- A novel synthetic CD38 antagonist increases NAD in mius, adipose, and brain tissues at exposures that well multiple tissues exceeded the IC50 of ~ 12 ng/mL (Fig. 2d). Accordingly, the Our group previously reported a series of novel chemical NAD content was found to be significantly elevated in the entities (NCEs), which potently inhibit the constitutive liver, muscle, and brain when normalized to an internal ana- NAD-degrading enzyme, CD38. Related screening efforts lytical standard (Fig. 2e). Despite high exposure, the NAD yielded the imidazoquinoline dubbed GSK978A, which ex- recovery from adipose was low and NAD changes were not hibited tenfold higher potency against mouse recombinant significant. However, in most tissues, peak NAD elevation CD38 than the human enzyme (Fig. 2a). This potency is of at least 30% was achieved at a dose of 3 mg/kg (Fig. 2f). orders of magnitude greater than that of natural products, To assess the potential pharmacodynamics of GSK978A such as quercetin, and approximates that of 78c, the best- in an eccentric challenge model, an acute study was per- studied synthetic CD38 inhibitor to-date [11, 14]. Yet formed in MDX mice following 5 days of dosing at 3 mg/ Fig. 3 Acute pharmacodynamics of GSK978A and NR in MDX hindlimbs following eccentric challenge. a Schematic of acute study design in MDX mice over 5 days (D1-D5) showing timing of eccentric contractions (ECC), assessments of contractility (CON), termination and tissue harvest (TERM), and compound administration. b Comparison of maximal tetanic force produced in the same limb at baseline and following eccentric challenge (ECC). Contractility (CON) was assessed over the 5-day protocol (D1-D5). c Masses of gastrocnemius muscles harvested from challenged (right, R) and contralateral (left, L) limbs at the study conclusion. d Total and (e) relative NAD content of gastrocnemius muscles harvested from challenged and contralateral limbs. f Total and (g) relative polyamine content of gastrocnemius muscles harvested from challenged and contralateral limbs. h Normalization of total polyamine to NAD contents in challenged limbs compared to the contralateral side. N = 8 MDX mice per group. Error bars represent SEM. Significance was determined by one-way ANOVA with Tukey’s post hoc test (**p <0.01, ***p < 0.001, ****p < 0.0001; ns, not significant or p value indicated) Frederick et al. Skeletal Muscle (2020) 10:30 Page 6 of 14 kg. Dietary NR, which has been suggested to improve the the physiology of MDX mice during three distinct stages: performance of MDX muscle [10], was included as a com- growth, recovery from an eccentric challenge, and recov- parator. Within 24 h of eccentric challenge, tetanic strength ery from a repeated challenge (Fig. 4a). We reasoned that was lessened by > 50% in all treatment groups, despite pres- this design would model the efficacy requirements of boys ervation of mass in the largest affected gastrocnemius mus- diagnosed with DMD. Beginning at 7-9 weeks of age, dur- cles (Fig. 3a-c). Interestingly, the challenged muscles also ing a period of rapid growth and peak muscle necrosis showed NAD depletion compared to the contralateral side, [22], MDX mice were randomized by hindlimb contractil- indicating that the muscle NAD pool does acutely respond ity and body weight, with a WT group included as a posi- to lengthening contractions (Fig. 3d). Mice treated with tive control for recovery. Over the course of 20 weeks, GSK978A, but not NR, showed a trend toward protection hindlimb weakness and increased CK release persisted in from this effect, though it could not be attributed to specific the MDX mice compared to WT controls (Fig. 4b-c). The NAD elevation in either limb (Fig. 3e). As a biomarker of pattern of hindlimb strength and serum creatine kinase muscle repair, total muscle polyamines showed clear eleva- (CK) release generally trended downward as MDX ani- tion in the injured limbs with a trend toward protection by mals reached maturity, but remained unchanged in both GSK978A, especially when polyamines were normalized to compound-treated groups, compared to the vehicle- NAD content (Fig. 3f-h). These results suggested that a lon- treated controls. MDX mice also accumulated lean mass ger treatment regimen might be necessary to provide func- steadily over the course of the study, reflecting character- tional improvements to MDX mice. istic hypertrophy, in a manner that was treatment- independent (Fig. 4d). At the study conclusion, gastrocne- Chronic NAD repletion does not provide functional mius muscles from the challenged MDX limbs were found protection from repetitive eccentric challenges to be ~ 15% less massive than the contralateral side in all We next designed a long-term study with chronic ad- treatment groups, reflecting an inability to fully regenerate ministration of GSK978A or NR to longitudinally assess injured fibers that was not observed in the WT. Fig. 4 Chronic pharmacodynamics of GSK978A and NR in MDX hindlimbs following eccentric challenge. a Schematic of chronic study design in MDX and WT mice over 20 weeks (weeks 1-20) showing timing of eccentric contractions (ECC), assessments of contractility (CON), termination and tissue harvest (TERM), and compound administration. GSK978A was dosed daily and contractility was assessed at varying intervals. b Maximal tetanic force production and (c) serum creatine kinase activity assessed at baseline, midpoint, and study conclusion. Arrows indicate timing of eccentric challenges. Significance was determined by repeated measure two-way ANOVA with Tukey’s post hoc test relative to MDX vehicle controls (**p < 0.01, ****p < 0.0001; ns, not significant compared to MDX vehicle controls). d Body composition assessed by qNMR at baseline, week 8, and week 18 of treatment. e Mass and (f) mass ratio of injured and contralateral gastrocnemius muscles collected after 20 weeks of treatment. N = 9-12 mice per group. Mice are MDX unless otherwise noted. Error bars represent SEM. Significance was determined by one-way ANOVA with Tukey’s post hoc test (****p < 0.001, **p < 0.01; ns, not significant) Frederick et al. Skeletal Muscle (2020) 10:30 Page 7 of 14 Surprisingly, contralateral muscles tended to be largest in the basis for our rationale. To address this question, we mice treated with GSK978A (Fig. 4e-f), which maybea analyzed the uninjured gastrocnemius muscles from the consequence of altered gait mechanics to favor the contra- chronically treated mice using a second untargeted lateral side, as it was not reflected in total lean mass. metabolomics platform, which utilized LC/MS to pro- Hindlimb strength was also serially assessed following vide broader coverage of the negatively charged metabo- eccentric challenges beginning at week 10 of treatment, to lome, including organic acids. This platform detected determine whether treated groups were protected from in- 3415 putatively annotated polar metabolites, based on jury or recovered faster. Compared to WT controls, MDX mass. When comparing the metabolites significantly al- mice showed approximately double the functional deficit tered between vehicle-treated MDX and WT groups, a within 1 day of eccentric challenge, despite similarly reversing pattern emerged in the GSK978A-treated mus- shaped tetani, but neither parameter was affected by cles, which was not observed in the NR-treated MDX GSK978A nor NR (Fig. 5a-b). Furthermore, MDX hin- animals (Fig. 6a). This anti-correlation was confirmed dlimbs showed highly similar recovery kinetics between with robust significance (R = −0.56, p < 1E^ −100) only the first and second challenges, while WT controls in the GSK978A treatment group (Fig. 6b-c). rebounded faster after the second bout (Fig. 5c-d). The ab- Suspecting that such a dramatic reversion effect may have sence of a protective repeated-bout effect in MDX hin- resulted from restoration of a pleiotropic co-factor, like dlimbs may reflect the fact that adult dystrophin-deficient NAD, we again performed a pathway analysis. Surprisingly, muscles are preconditioned to such cycles of damage and we found the most significant enrichment in only two path- repair by activities of normal living. ways: purine metabolism and the pentose phosphate path- way (Fig. 6d). Since the untargeted platform was limited in CD38 inhibition significantly reverts the MDX muscle its ability to address the central question of whether posi- metabolome to the WT state tively charged NAD was specifically restored, we instead Given the lack of physiological protection conferred utilized an enzymatic NAD assay on the same tissue sam- upon MDX muscle by NAD-modulating compounds, we ples and found only statistically non-significant elevations suspected that the treatments were simply ineffective at compared to MDX vehicle controls (Fig. 6e). Though the correcting the basal metabolic imbalance that formed NR-treated MDX muscles still contained significantly less Fig. 5 Time course of hindlimb contractile function following consecutive eccentric challenges. a Longitudinal tetanic force produced by the hindlimbs beginning at week 10 of treatment and including repeat bouts of eccentric challenge (arrows) and recovery. b High-resolution traces of tetani immediately before and 24 h after the first eccentric challenge. c Three-week longitudinal hindlimb force production normalized to pre- challenge levels after the first and (d) second successive eccentric challenges. N = 9-12 mice per group. Mice are MDX unless otherwise noted. Error bars represent SEM. Significance was determined by repeated measure two-way ANOVA with Tukey’s post hoc test (*p < 0.05, ***p < 0.001, ****p < 0.0001; ns, not significant compared to MDX vehicle controls) Frederick et al. Skeletal Muscle (2020) 10:30 Page 8 of 14 Fig. 6 Metabolomic signatures of NAD replacement strategies in MDX muscle. a Heatmap of ion counts for polar metabolites reaching significance (adjusted p < 0.05) in MDX muscle compared to WT controls (left column) aligned with the same metabolites identified in GSK978A- treated (center column) and NR-treated (right column) muscles after 20 weeks of treatment. b Correlation of the changes in ion abundance induced by GSK978A and (c) NR supplementation in MDX muscle after 20 weeks compared to changes driven by genotype alone. Colored points indicate metabolite ions significantly altered in at least one of the comparisons and the dashed line indicates the least-squares regression of these points. R indicates the correlation coefficient of the significant (adjusted p < 0.05) ions and p indicates the significance level of the associated Pearson’s correlation. d Volcano plot of biochemical pathways significantly altered in MDX muscle by GSK978A. Significantly altered metabolites were used to calculate pathway impact and highly affected pathways are indicated. e Total NAD content of gastrocnemius muscles harvested from contralateral limbs. Significance was determined by one-way ANOVA with Tukey’s post hoc test (**p < 0.01; ns, not significant). f Box plots of ion abundances for representative members of glycolysis (G6P, F6BP, G3P), pentose phosphate pathway (R5P), and TCA cycle (CA, SA). N = 9-12 mice per group. Whiskers represent minimum and maximum values. Significance was determined by one-way ANOVA with Tukey’s post hoc test (p values adjusted for multiple testing) NAD than WT muscle, the intermediate GSK978A group multiple metabolic pathways that is not recapitulated by was statistically indistinguishable from vehicle-treated supplementing a NAD precursor. groups of either genotype. Though sustained NAD elevations were not observed, Discussion we found several metabolic reversions consistent with Advancements in high-throughput discovery metabolo- transient restoration of NAD-dependent processes. For mics have led to improved biomarker detection and novel example, NAADH, the neutral reduced form of a sus- therapeutic strategies for many diseases, yet determining pected biomarker of NAD repletion [23] or overload, whether a complex molecular signature is a cause or con- was detected only in GSK978A-treated muscles (Supple- sequence of pathology is highly context-dependent [24]. mental Table 1). Importantly, ion counts for compo- Given its central role in maintaining ATP generation via nents of the proximal glycolytic pathway, including both glycolysis and oxidative phosphorylation, NAD is a those for glucose-6-phosphate and fructose-1,6,-bispho- co-factor well-positioned to influence muscle mass and sphate, were largely normalized by GSK978A, as were performance, and its synthesis and degradation have been the GAPDH substrate, glyceraldehyde-3-phosphate, and suggested to be dysregulated in the absence of dystrophin the pentose phosphate intermediate, ribose-5-phosphate [10, 12]. Our study confirms that dystrophin-deficient (Fig. 6f). We also observed discrepant influence on TCA muscle maintains a diminished NAD pool, even in the ab- cycle intermediates and a subtle net effect of increased sence of deliberate physiological challenges. Our observa- ATP and decreased phosphocreatine in this group (Sup- tion that NAD can be further depleted within hours of an plemental Table 1), potentially indicating elevated eccentric challenge suggests that the repetitive nature of substrate-level phosphorylation. Collectively, these global these lengthening contractions during activities of normal metabolomic profiles of dystrophin-deficient muscle living may collectively drive chronic NAD depletion in highlight a restorative effect of CD38 inhibition on diseased tissue, providing insight into the etiology of Frederick et al. Skeletal Muscle (2020) 10:30 Page 9 of 14 DMD. Though the degree to which wildtype dystrophin This is especially true when considering NAD synthesis acutely mitigates, this process remains unresolved, and consumption fluxes, which vary widely between associated therapeutic indications would be limited to mouse tissues [30]. Indeed, the turnover of NAD in less life-threatening conditions, such as exercise recov- mouse muscle was recently found to be the slowest of ery or muscular trauma. any tissue tested [30], and the degradative activity of The mechanism of acute muscle NAD depletion likely other enzymes, such as PARPs, may predominate [31]. It reflects an imbalance in production and consumption is also possible that GSK978A primarily influences the fluxes. One model suggests that calcium dysregulation, global metabolic profile of muscle via ancillary effects on stemming from microtears in the sarcolemma, leads to a calcium homeostasis, which is known to be dysregulated burst of genotoxic reactive oxygen species and hyperacti- in the absence of dystrophin [26], or infiltrating immune vation of NAD-consuming PARPs [25, 26]. However, cells. Nonetheless, our conclusion that CD38 inhibition cleavage of NAD by PARPs would be expected to liber- did not functionally protect dystrophin-deficient muscle ate nicotinamide, and our data clearly indicate the op- is in line with that of Spaulding et al., who found that posite pattern. Rather, our finding that the methylated long-term administration of the CD38-inhibiting flavon- waste product, 1-me-Nam, is more abundant in the days oid, quercetin, failed to protect isolated MDX muscles following injury, suggests that the removal of nicotina- from contraction-induced injury [32]. mide equivalents from the cytosol by the enzyme nico- Though the muscle exposure of GSK978A was com- tinamide N-methyltransferase (NNMT) may effectively parable to that of other tissues, the 30% NAD elevation limit the re-synthesis of NAD from nicotinamide via the that we observed 2 h after dosing was modest by com- NAD salvage pathway. The regulation of NNMT activity parison to the liver, brain, and adipose, which more than is still poorly understood [27], and bulk tissue analysis is doubled NAD content over the same period (Fig. 2). We unable to resolve whether infiltrating cell types are re- also failed to detect significant changes in the muscle sponsible for the effect, but a consistent pattern of in- NAD pool following acute or chronic treatment. This creased NNMT expression has been previously reported may be an indication of several factors. First, because in muscle biopsies from patients with a variety of dys- muscle makes up a large percentage of body mass, trophic conditions [10]. Furthermore, since 1-me-Nam muscle-targeting drugs must have high volumes of dis- is prone to urinary secretion, it may be a useful indicator tribution. Limited solubility or excessive albumin bind- of efficacy for oligonucleotide-based therapeutics, such ing could effectively limit the interaction of quinolones, as those being tested in the FKRP mutant model of like GSK978A, with their intended target. Second, the limb-girdle muscular dystrophy [28]. We also found ele- accuracy and variability of NAD quantitation is highly vations in MDX muscle of several positive biomarkers dependent on extraction conditions and analytical tech- previously identified in models of primary muscle NAD niques. Our reliance on multiple mass spectrometry- depletion [8, 18], including DHAP and S7P. Consistently, based and enzymatic assays made it challenging to re- the large-scale metabolic imbalance secondary to NAD producibly measure subtle shifts in the NAD pool. depletion appeared to be amplified in MDX muscle by a Third, there may exist a biological upper limit to the bottleneck in glycolysis. The near-complete reversal of steady-state NAD content of muscle, as suggested by this imbalance by a small molecule CD38 antagonist earlier transgenic models [33, 34]. The reversibility of provided compelling evidence that depletion of one or the NMN adenyl transferase enzymes may effectively more of the pathway's cofactors is largely responsible for limit the expansion of the NAD pool in a tissue-specific the distinctive metabolomic fingerprint. manner. Lastly, it is possible that CD38 is not a major CD38 is a uniquely complex pharmacological target consumer of NAD in muscle, or that the enzyme expres- due to an unusual array of enzymatic activities and sion is downregulated during pathology. Such transcrip- modes of regulation. Additionally, the ability of the tional compensation has been observed previously in CD38 extracellular domain to function as a cell surface DMD muscle [10]. Nonetheless, a global assessment of ligand for CD31 represents a signaling mechanism that the treated tissues was largely consistent with a transient may be more effectively disrupted with monoclonal anti- or compartmentalized restoration of NAD-dependent bodies than small molecules and may contribute to the pathways. This restoration did not manifest in the form phenotype of CD38 knockout mice [21]. As an enzyme, of nicotinamide-containing metabolites, as predicted, CD38 can convert not only NAD but also NADP and but rather in a more stable impact on pentose phosphate nicotinic acid into calcium-mobilizing second messen- pathway intermediates. The observed impact on purine gers, such as cADPR and NAADP, in a manner metabolism is likely to be directly linked via dependent on both membrane topology and local pH normalization of ribose-5-phosphate, the pentose [29]. Thus, despite broad exposure, the specific pharma- phosphate-derived nucleotide precursor (Fig. 6d, f). codynamic effects of GSK978A might vary by cell type. Interestingly, purine metabolism was identified in our Frederick et al. Skeletal Muscle (2020) 10:30 Page 10 of 14 initial characterization of MDX muscle, but was not im- Methods plicated in the acute response to eccentric injury (Fig. Animal care and use 1c, d). Thus, GSK978A may be more effective at restor- Male C57BL/10ScSn-Dmd<mdx>/J (MDX) and C57BL/ ing chronic metabolic imbalances, rather than buffering 10ScSn/J (WT) aged 7-9 weeks were individually housed acute challenges. with ad libitum access to regular chow and water during a A central finding of our work is that specific antagon- 12: 12 h light: dark cycle under controlled temperature ism of CD38 is a more effective strategy than NR supple- and humidity. Pharmacokinetic studies were performed in mentation for restoring the metabolic imbalance of 5-month old C57BL6 mice fed a high fat diet (Research MDX muscle. The low micromolar IC50 of GSK978A Diets D12492). GSK978A was custom synthesized and achieves muscle NAD elevation similar to that of natural dissolved at 0.3 mg/mL in vehicle containing 0.5% hydro- products, such as NR [35], at less than 1% of the effect- xpropyl methyl cellulose and 0.1% polysorbate 80, pH 4. ive dose. The relative inability of NR to affect the MDX Ten milliliters per kilogram was administered daily in the muscle metabolome likely stems from its poor bioavail- morning by oral gavage. NR chloride was custom synthe- ability and short (< 3 min) half-life in the blood [8, 30], sized and dissolved in the drinking water at 12 mM, sterile which is consistent with the absence of pharmaco- filtered, and administered ad libitum in light protected dynamics observed in several clinical trials [36–38]. bottles, as described [8, 40]. All compounds were reformu- However, both NAD-modulating strategies employed in lated weekly. Body composition was monitored by quanti- our study failed to improve muscle function. We were tative NMR spectroscopy. All studies were conducted in largely unable to reproduce the results of the Auwerx accordance with the GSK Policy on the Care, Welfare and group, who observed a significant reduction of plasma Treatment of Laboratory Animals and were reviewed the creatine kinase and a nearly 50% protection from eccen- Institutional Animal Care and Use Committee either at tric challenge in the same strain of MDX mice treated GSK or by the ethical review process at the institution with NR for only 12 weeks [10]. The discrepancies may where the work was performed. derive from the fact that Ryu et al. assessed hindlimb torque around the knee joint instead of the ankle, and Compound screening administered an NR-triflate salt to mice, instead of the NCEs were tested for inhibition of CD38 transglycosida- NR chloride salt used in all neutraceutical formulations. tion or base exchange activity by colorimetric assay based Nonetheless, our finding that neither GSK978A nor NR on a published method [41] using recombinant mouse had any effect on the performance of MDX muscle over CD38 soluble domain protein purified from Pichia pas- time raises the question of whether biochemical imbal- toris. Briefly, 0.5 nM enzyme was incubated in buffer con- ance is pathologically relevant in the absence of a central taining 50 mM HEPES, pH 7.4, 1 mM CHAPS, 2 mM structural component like dystrophin. The favorable EDTA, 250 μM isonicotinaldehyde 2-pyridinylhydrazone, pharmacokinetics and brain penetrance of GSK978A 100 μM NAD, 1% DMSO, and 1-10,000 nM NCEs while suggest that the compound may have better efficacy in absorbance was monitored at 405 nm. Inhibitor potency certain neurodegenerative disorders, which feature NAD was calculated with the following equation: y = A+((B-A)/ depletion, such as Cockayne syndrome or xeroderma pig- (1 + (10^x/10^C)^D)), where A is the enzyme-free re- mentosa [39]. These and other indications for small mol- sponse, B is the inhibitor-free response, C is the log ecule NAD-modulators warrant further investigation. (IC50), and D is the hill slope. Global metabolomics following eccentric challenge Conclusion (external platform) In summary, MDX mice exhibit a chronic NAD deficit Samples were prepared using the automated MicroLab with broad effects on the biochemical phenotype of the STAR system (Hamilton Company, Franklin MA). Recov- hindlimb muscle. The distinct global metabolome of ery standards were added prior to the extraction process dystrophin-deficient muscle becomes acutely altered by for quality control purposes. Samples were lysed in ice- eccentric injury and can be partly restored by inhibition cold methanol and the resulting extract was divided into of CD38, though this intervention does not confer pro- four fractions: one each for analysis by reversed-phase tection against future injury. While primary NAD defi- UPLC-MS/MS with positive and negative ion mode elec- ciency may suffice to drive both muscle weakness and a trospray ionization, one for normal-phase UPLC-MS/MS transcriptional profile resembling dystrophy over time, platform, and one for analysis by GC-MS. Samples were our current data strongly suggest that the characteristic centrifuged at 13,000×g for 10 min and supernatants were muscle weakness of MDX mice cannot be overcome by dried under nitrogen. The MS system was a Thermo Sci- NAD replacement strategies alone and that such strat- entific Q-Exactive high resolution/accurate mass orbitrap egies would be unlikely to benefit patients with DMD. mass spectrometer operated at 35,000 mass resolution Frederick et al. Skeletal Muscle (2020) 10:30 Page 11 of 14 which was interfaced with a heated electrospray ionization ten runs using a standard solution of 16 organic acids. (HESI-II) source. Dried sample extracts were reconstituted Peak detection and global alignment of all scans was in solvents amenable to their respective method. One ali- performed using a custom metabolomics data processing quot was analyzed using acidic positive ion optimized con- pipeline. Detected ion m/z values and isotope distribu- ditions and another using basic negative ion optimized tions were matched against the human metabolome conditions in two independent injections using separate database [43] assuming [M-H] and [M-2H] species and 13 12 dedicated columns (Waters UPLC BEH C18-2.1 × 100 at most two C/ C exchanges to tentatively annotate mm, 1.7 μm). The extracts reconstituted in acidic condi- metabolites, with the method-inherent limitation of be- tions were gradient eluted using water and methanol con- ing unable to distinguish between isomers. taining 0.1% formic acid, while the basic extracts, which also used water/methanol, contained 6.5 mM ammonium Biochemical pathway analysis bicarbonate. A third aliquot was analyzed via negative Pathway analysis was performed on metabolites reaching ionization following elution from a HILIC column (Wa- an adjusted significance threshold of p < 0.05 for a given ters UPLC BEH Amide 2.1 × 150 mm, 1.7 μm) using a gra- comparison using the MetaboAnalyst 4.0. platform [44] dient consisting of water and acetonitrile with 10 mM and referencing the current KEGG pathway library for ammonium formate. The MS analyses alternated between mouse. Over-representation analysis was performed MS and data-dependent MS scans using dynamic exclu- using Fisher’s exact test and pathway topology analysis sion, and the scan range was from 80-1000 m/z. The sam- was performed using relative-betweeness centrality. ples designated for GC-MS analysis were derivatized under nitrogen using bistrimethyl-silyl-trifluoroacetamide Tissue pharmacokinetics and pharmacodynamics (BSTFA). The GC column was a 20 m × 0.18 mm ID, with To determine the stability of new chemical entities 5% phenyl; 95% dimethylsilicone phase. Samples were ana- in vitro, cryopreserved liver microsomes from several spe- lyzed on a Thermo-Finnigan Trace DSQ fast-scanning cies (Sekisui Zenotech, Japan) were thawed and diluted to single-quadrupole mass spectrometer using electron 0.9 mg/mL in 50 mM phosphate buffer, pH 7.4. NCEs in impact ionization at unit mass resolution. Raw data was DMSO were added at 0.5 μM to the microsome suspen- extracted, peak-identified, and quality control processed sion and pre-incubated for 5 min at 37 °C in a standard using Metabolon’s hardware and software. Peaks were cell culture incubator with shaking at 80 RPM. Clearance quantified using area-under-the-curve. Compounds reactions were started by the addition of 2 mM NADPH were identified by comparison to library entries of puri- and 5 mM MgCl cofactors, then 100 μL of microsome fied standards or recurrent unknown entities. Propri- suspension was removed from the reaction at designated etary visualization software was used to confirm the time points and mixed with 200 μLice-coldstopsolution consistency of peak identification among the samples. containing 80:20 methanol: acetonitrile containing 1% acetic acid. Microsome extracts were centrifuged at 10, Global metabolomics following chronic interventions 000×g for 15 min and supernatants were subjected to LC- (internal platform) MS/MS analysis (below). Metabolic stability expressed as Polar metabolites were extracted from frozen tissues fol- a percentage of the parent compound remaining over time lowing lysis in a fivefold excess of ice cold 70% ethanol was determined from the peak area ratios in order to cal- using a bead homogenizer. Tissue lysates were further culate the turnover rate constant, k, by linear regression diluted 1:20 in 70% ethanol, incubated at 75 °C for 3 and half-life according to the equation t = ln(2)/k.For 0.5 min, and centrifuged at 13,000×g for 10 min. Superna- assessing NCE distribution and pharmacodynamics tants were lyophilized, resuspended in 0.1 mL water, and in vivo, 10 μL of blood was harvested from the mouse tail subjected to flow injection mass spectrometry. Non- vein, mixed with 50 μL of water and 40 μL of acetonitrile. targeted mass spectrometry of polar metabolites was Samples were sonicated for 5 min, vortexed for 5 min, and performed as described [42]. Briefly, Q-exactive Plus centrifuged at 2000×g for 20 min. Supernatants were di- (Thermo Scientific) in profile mode with scan range 50- luted 1:5 in water and subjected to LC-MS/MS analysis. 1000 m/z was calibrated according to manufacturer pro- Tissues were bead homogenized for 2× 1 min in a fourfold tocols. Resolution was set to 70,000 at 200 m/z with excess of ice-cold 80% acetonitrile and centrifuged at 13, automatic gain control target of 3E6 ions, 3.0 kV spray 000×g for 20 min. Supernatants were diluted 1:10 in water voltage, 120 ms maximum injection time, and 60 s acqui- and subjected to LC-MS/MS analysis. As an internal sition time. Samples were injected in a randomized se- standard, 1.5 μmol of O-NAD was spiked into the tissue quence and analyzed in negative ion mode using a matrix. LC-MS/MS was performed on an Agilent 1290 In- mobile phase consisting of 60% isopropanol, 40% water, finity system using a mobile phase of methanol containing 1mM NH F, 10 nM taurocholic acid, 20 nM homotaur- 0.1% formic acid and a Varian Polaris amide-C18 column ine. Quality control was performed before each batch of coupled to a Sciex API 4000 mass spectrometer. NAD Frederick et al. Skeletal Muscle (2020) 10:30 Page 12 of 14 peaks were normalized to the internal standard and drug in a cycling mix containing 0.1% BSA, 2% ethanol, concentrations were determined using a standard curve 100 μg/ml alcohol dehydrogenase, 10 μg/ml diaphorase, generated in the tissue matrix. 20 μM resazurin, and 10 μM flavin mononucleotide in 100 mM phosphate buffer. Enzymatic cycling at room Hindlimb eccentric challenge and longitudinal temperature produced resorufin, the fluorescence of contractility which was monitored over time at ex/em 544/590 nm. Mice were anesthetized using isoflurane (3%/L O ) and Muscle total polyamines were measured using a fluoro- placed on a warming pad with their right hind limbs re- metric total polyamine assay kit (K475-100, Biovision) strained at the knee and foot affixed to a force trans- according to the manufacturer protocol. Briefly, frozen ducer with motor-arm (Aurora Scientific Instruments, muscles were ground under liquid nitrogen and a 100 Aurora, ON). Platinum sub-dermal electrodes were mg portion was further dounce homogenized in 0.5 mL inserted dorsally and ventrally to the femur to apply of ice-cold homogenization buffer. Lysates were centri- electrical field stimulation (2.5 mA at 25 V) to the sciatic fuged at 5000×g for 5 min and supernatants were further nerve and trigger contraction of the plantarflexor mus- filtered through 10 kD molecular weight cutoff spin col- cles of the lower limb. Muscles were stimulated isomet- umns. Extracts were assayed by fluorometric enzymatic rically at a single twitch (200 μs pulse) and tetanic (150 assay and compared to a standard curve. Hz at 200 μs pulse for 0.8 s) frequencies to assess longi- tudinal force production over the course of the study. Statistics Eccentric injury was induced by subjecting hindlimbs to Data were compiled and analyzed using Microsoft Excel a series of 40 lengthening contractile stimuli, consisting and graphed using Graphpad Prism. Statistical tests of a sub-tetanic stimulation of 100 Hz at 200 μs pulse for (Student’s 2-tailed t test, one-way ANOVA, repeated 0.4 s, while the motorized footplate applied an eccentric measure two-way ANOVA, Tukey’s post hoc test, and rotational torque. Animals were returned to holding en- least-square correlation analysis) were calculated using closures and isometric titanic force was monitored to as- Graphpad Prism with a significance threshold of p < sess force deficit and recovery. 0.05, as indicated. For metabolomics data, p values were adjusted for multiple hypothesis testing using either Acute pharmacodynamics following eccentric challenge Benjamini’s and Hochberg’s method [45] (external plat- For acute eccentric challenge studies, male C57BL/ form) or Storey’s and Tibshirani’s method [46] (internal 10ScSn-Dmd<mdx>/J aged 24-26 weeks were individually platform data), and principal component analysis was housed and treated for 5 days, as above. Baseline body performed on the first two of ten components using cus- weight and contractility were assessed 1 week before the tom R scripts. start of dosing and used for group randomization. On day four of treatment, right hindlimbs were subjected to the Supplementary information eccentric damage protocol 15 min after oral dosing. On Supplementary information accompanies this paper at https://doi.org/10. 1186/s13395-020-00249-y. day five, oral compounds were dosed 15 min before con- tractility measurement and 60 min before sacrifice. Mus- Additional file 1. Frederick et al Supplemental Table 1. cles from both limbs were snap frozen and stored at −80 °C before analysis. Abbreviations NCE: Novel chemical entity; TCA: Tricarboxylic acid; DHAP: Dihydroxyacetone Creatine kinase measurement phosphate; ECCs: Eccentric contractions; NAD+ or NAD: Nicotinamide Mice were anesthetized using 3% isoflurane and venous adenine dinucleotide; NADP: Nicotinamide adenine dinucleotide phosphate; NAADH: Nicotinic acid adenine dinucleotide (reduced); NAADP: Nicotinic whole blood was collected in a microcapillary from the acid adenine dinucleotide phosphate; DGC: Dystroglycan complex; retro-orbital sinus. Blood was allowed to clot at room PARP: Poly-ADP-ribose polymerase; ARC: ADP-ribosyl cyclase; aKG: Alpha- temperature for 30 min, then centrifuged at 10,000×g for ketoglutarate; NR: Nicotinamide riboside; NNMT: Nicotinamide N-methyl transferase; Nam: Nicotinamide; S7P: Seduheptulose-7-phosphate; 5 min. The resulting serum samples were diluted 1:3 in F6BP: Fructose-1,6-bisphosphate; CK: Creatine kinase water and subjected to automated enzymatic assay (Beckman Coulter, Brea CA). Acknowledgements We wish to thank C. Haffner for expertise in chemical synthesis, J. McNulty, K. Morasco, N. Milliken, and Metabolon, Inc. for providing technical expertise, NAD and polyamine measurement and A. Hinken and H. Feldser for continued guidance and helpful revisions NAD was extracted from frozen muscles and measured of the text. by enzymatic cycling assay, as described [34]. Briefly, 50 Authors’ contributions mg of muscle was extracted in 0.5 mL 0.6 M perchloric DF, JB, JU, DS, FP, and HK designed experiments. DF, AM, MS, JV, JB, FP, and acid and diluted 1:100 in 100 mM phosphate buffer, pH JU performed experiments. AN, ES, and DS analyzed and graphed in vitro 8. Samples and NAD standards were further diluted 1:20 molecule validation and metabolomics data. DF wrote the manuscript. DF, Frederick et al. Skeletal Muscle (2020) 10:30 Page 13 of 14 AN, ES, FP, and HK edited the manuscript. The authors read and approved pathophysiology in a mouse model of Duchenne muscular dystrophy. PLoS the final manuscript. Genet. 2014;10(7):e1004490. 13. Yoshino J, Baur JA, Imai S-I. NAD+ intermediates: the biology and therapeutic potential of NMN and NR. Cell Metab. 2018;27(3):513–28. Funding 14. Hogan KA, Chini CCS, Chini EN. The multi-faceted ecto-enzyme CD38: roles This work was funded in its entirety by GlaxoSmithKline, Inc. and facilitated in immunomodulation, cancer, aging, and metabolic diseases. Front by postdoctoral fellowships awarded to DWF and AN. Immunol. 2019;10:1187. 15. Becherer JD, Boros EE, Carpenter TY, Cowan DJ, Deaton DN, Haffner CD, Availability of data and materials et al. Discovery of 4-amino-8-quinoline carboxamides as novel, The metabolomics datasets analyzed in the current study are available in the submicromolar inhibitors of NAD-hydrolyzing enzyme CD38. J Med Chem. online version of the article. 2015;58(17):7021–56. The datasets analyzed in the current study are available from the 16. Haffner CD, Becherer JD, Boros EE, Cadilla R, Carpenter T, Cowan D, et al. corresponding author upon request. Discovery, synthesis, and biological evaluation of thiazoloquin(az)olin(on)es as potent CD38 inhibitors. J Med Chem. 2015;58(8):3548–71. Ethics approval and consent to participate 17. Zhang H, Ryu D, Wu Y, Gariani K, Wang X, Luan P, et al. NAD+ repletion Not applicable. improves mitochondrial and stem cell function and enhances life span in mice. Science. 2016;352(6292):1436–43. https://doi.org/10.1126/science. Consent for publication aaf2693. Not applicable. 18. Oakey LA, Fletcher RS, Elhassan YS, Cartwright DM, Doig CL, Garten A, et al. Metabolic tracing reveals novel adaptations to skeletal muscle cell energy production pathways in response to NAD+ depletion. Wellcome Open Res. Competing interests 2019;3:147. The authors declare that they have no competing interests. 19. Luchessi AD, Cambiaghi TD, Hirabara SM, Lambertucci RH, Silveira LR, Baptista IL, et al. Involvement of eukaryotic translation initiation factor 5A Author details (eIF5A) in skeletal muscle stem cell differentiation. J Cell Physiol. 2009;218(3): Muscle Metabolism Unit, GlaxoSmithKline R&D, Research Triangle Park, NC, 480–9. Collegeville, PA, USA. Cellzome, GlaxoSmithKline R&D, Heidelberg, Germany. 20. Puleston DJ, Buck MD, Klein Geltink RI, Kyle RL, Caputa G, O’Sullivan D, et al. Target Sciences, Computational Biology, GlaxoSmithKline R&D, Collegeville, Polyamines and eIF5A hypusination modulate mitochondrial respiration and PA, USA. Computational Sciences, Molecular Design, GlaxoSmithKline R&D, macrophage activation. Cell Metab. 2019;30(2):352–363.e8. Collegeville, PA, USA. 21. Barbosa MTP, Soares SM, Novak CM, Sinclair D, Levine JA, Aksoy P, et al. The enzyme CD38 (a NAD glycohydrolase, EC 3.2.2.5) is necessary for the Received: 8 May 2020 Accepted: 22 September 2020 development of diet-induced obesity. FASEB J Off Publ Fed Am Soc Exp Biol. 2007;21(13):3629–39. 22. Dangain J, Vrbova G. Muscle development in mdx mutant mice. Muscle References Nerve. 1984;7(9):700–4. 1. Petrof BJ, Shrager JB, Stedman HH, Kelly AM, Sweeney HL. Dystrophin 23. Trammell SAJ, Schmidt MS, Weidemann BJ, Redpath P, Jaksch F, Dellinger protects the sarcolemma from stresses developed during muscle RW, et al. Nicotinamide riboside is uniquely and orally bioavailable in mice contraction. Proc Natl Acad Sci U S A. 1993 Apr 15;90(8):3710–4. and humans. Nat Commun. 2016;7:12948. 2. Chamberlain JR, Chamberlain JS. Progress toward gene therapy for 24. Sévin DC, Kuehne A, Zamboni N, Sauer U. Biological insights through Duchenne muscular dystrophy. Mol Ther J Am Soc Gene Ther. 2017;25(5): nontargeted metabolomics. Curr Opin Biotechnol. 2015;34:1–8. 1125–31. 25. Pirinen E, Cantó C, Jo YS, Morato L, Zhang H, Menzies KJ, et al. 3. Dunn JF, Frostick S, Brown G, Radda GK. Energy status of cells lacking Pharmacological inhibition of poly(ADP-ribose) polymerases improves dystrophin: an in vivo/in vitro study of mdx mouse skeletal muscle. Biochim fitness and mitochondrial function in skeletal muscle. Cell Metab. 2014;19(6): Biophys Acta BBA - Mol Basis Dis. 1991;1096(2):115–20. 1034–41. 4. Griffin J, Sang E, Evens T, Davies K, Clarke K. Metabolic profiles of dystrophin 26. Whitehead NP, Yeung EW, Allen DG. Muscle damage in mdx (dystrophic) and utrophin expression in mouse models of Duchenne muscular mice: role of calcium and reactive oxygen species. Clin Exp Pharmacol dystrophy. FEBS Lett. 2002;530(1-3):109–16. Physiol. 2006;33(7):657–62. 5. Joseph J, Cho DS, Doles JD. Metabolomic analyses reveal extensive progenitor cell 27. Pissios P, Nicotinamide N. methyltransferase: more than a vitamin B3 deficiencies in a mouse model of Duchenne muscular dystrophy. Metabolites. clearance enzyme. Trends Endocrinol Metab. 2017;28(5):340–53. 2018;8(4). https://doi.org/10.3390/metabo8040061. PMID 30282911. 28. Vannoy CH, Leroy V, Broniowska K, Lu QL. Metabolomics analysis of skeletal 6. Lindsay A, Chamberlain CM, Witthuhn BA, Lowe DA, Ervasti JM. muscles from FKRP-deficient mice indicates improvement after gene Dystrophinopathy-associated dysfunction of Krebs cycle metabolism. Hum replacement therapy. Sci Rep. 2019;9(1) [cited 2019 Jul 15]. Available from: Mol Genet. 2019 Mar 15;28(6):942–51. http://www.nature.com/articles/s41598-019-46431-1. 7. Cataldi MP, Lu P, Blaeser A, Lu QL. Ribitol restores functionally glycosylated 29. Malavasi F, Deaglio S, Funaro A, Ferrero E, Horenstein AL, Ortolan E, et al. α-dystroglycan and improves muscle function in dystrophic FKRP-mutant Evolution and function of the ADP ribosyl cyclase/CD38 gene family in mice. Nat Commun. 2018;9(1):3448. physiology and pathology. Physiol Rev. 2008;88(3):841–86. 8. Frederick DW, Loro E, Liu L, Davila A, Chellappa K, Silverman IM, et al. Loss 30. LiuL, SuX,Quinn WJ,Hui S, Krukenberg K, FrederickDW, et al.Quantitative of NAD homeostasis leads to progressive and reversible degeneration of analysis of NAD synthesis-breakdown fluxes. Cell Metab. 2018;27(5):1067–1080.e5. skeletal muscle. Cell Metab. 2016;24(2):269–82. 31. Bai P, Cantó C, Oudart H, Brunyánszki A, Cen Y, Thomas C, et al. PARP-1 9. Gomes AP, Price NL, Ling AJY, Moslehi JJ, Montgomery MK, Rajman L, inhibition increases mitochondrial metabolism through SIRT1 activation. Cell et al. Declining NAD(+) induces a pseudohypoxic state disrupting Metab. 2011;13(4):461–8. nuclear-mitochondrial communication during aging. Cell. 2013 Dec 19; 32. Spaulding HR, Ballmann CG, Quindry JC, Selsby JT. Long-term quercetin 155(7):1624–38. dietary enrichment partially protects dystrophic skeletal muscle. Kumar A, 10. Ryu D, Zhang H, Ropelle ER, Sorrentino V, Mázala DAG, Mouchiroud L, et al. editor. PLoS One. 2016;11(12):e0168293. NAD+ repletion improves muscle function in muscular dystrophy and counters global PARylation. Sci Transl Med. 2016;8(361):361ra139. 33. Costford SR, Brouwers B, Hopf ME, Sparks LM, Dispagna M, Gomes AP, et al. 11. Tarragó MG, Chini CCS, Kanamori KS, Warner GM, Caride A, de Oliveira GC, Skeletal muscle overexpression of nicotinamide phosphoribosyl transferase et al. A potent and specific CD38 inhibitor ameliorates age-related in mice coupled with voluntary exercise augments exercise endurance. Mol metabolic dysfunction by reversing tissue NAD+ decline. Cell Metab. 2018; Metab. 2017; [cited 2017 Nov 20]; Available from: http://linkinghub.elsevier. 27(5):1081–1095.e10. com/retrieve/pii/S2212877817306816. 12. Chalkiadaki A, Igarashi M, Nasamu AS, Knezevic J, Guarente L. Muscle- 34. Frederick DW, Davis JG, Dávila A, Agarwal B, Michan S, Puchowicz MA, et al. specific SIRT1 gain-of-function increases slow-twitch fibers and ameliorates Increasing NAD synthesis in muscle via nicotinamide Frederick et al. Skeletal Muscle (2020) 10:30 Page 14 of 14 phosphoribosyltransferase is not sufficient to promote oxidative metabolism. J Biol Chem. 2015;290(3):1546–58. 35. Cantó C, Houtkooper RH, Pirinen E, Youn DY, Oosterveer MH, Cen Y, et al. The NAD(+) precursor nicotinamide riboside enhances oxidative metabolism and protects against high-fat diet-induced obesity. Cell Metab. 2012;15(6):838–47. 36. Dollerup OL, Christensen B, Svart M, Schmidt MS, Sulek K, Ringgaard S, et al. A randomized placebo-controlled clinical trial of nicotinamide riboside in obese men: safety, insulin-sensitivity, and lipid-mobilizing effects. Am J Clin Nutr. 2018;108(2):343–53. 37. Dollerup OL, Trammell SAJ, Hartmann B, Holst JJ, Christensen B, Møller N, et al. Effects of nicotinamide riboside on endocrine pancreatic function and incretin hormones in nondiabetic men with obesity. J Clin Endocrinol Metab. 2019;104(11):5703–14. 38. Dollerup OL, Chubanava S, Agerholm M, Søndergård SD, Altıntaş A, Møller AB, et al. Nicotinamide riboside does not alter mitochondrial respiration, content or morphology in skeletal muscle from obese and insulin resistant men. J Physiol. 2020;598(4):731–54. https://doi.org/10.1113/JP278752. 39. Lautrup S, Sinclair DA, Mattson MP, Fang EF. NAD+ in brain aging and neurodegenerative disorders. Cell Metab. 2019;30(4):630–55. 40. Fang EF, Scheibye-Knudsen M, Brace LE, Kassahun H, SenGupta T, Nilsen H, et al. Defective mitophagy in XPA via PARP-1 hyperactivation and NAD(+ )/SIRT1 reduction. Cell. 2014;157(4):882–96. 41. Preugschat F, Tomberlin GH, Porter DJT. The base exchange reaction of NAD+ glycohydrolase: identification of novel heterocyclic alternative substrates. Arch Biochem Biophys. 2008;479(2):114–20. 42. Fuhrer T, Heer D, Begemann B, Zamboni N. High-throughput, accurate mass metabolome profiling of cellular extracts by flow injection-time-of-flight mass spectrometry. Anal Chem. 2011;83(18):7074–80. 43. Wishart DS, Feunang YD, Marcu A, Guo AC, Liang K, Vázquez-Fresno R, et al. HMDB 4.0: the human metabolome database for 2018. Nucleic Acids Res. 2018;46(D1):D608–17. 44. Chong J, Wishart DS, Xia J. Using MetaboAnalyst 4.0 for comprehensive and integrative metabolomics data analysis. Curr Protoc Bioinformatics. 2019;68(1) [cited 2020 Jun 26]. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/cpbi.86. 45. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B Methodol. 1995;57(1):289–300. 46. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci U S A. 2003;100(16):9440–5. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Skeletal Muscle Springer Journals

Complementary NAD+ replacement strategies fail to functionally protect dystrophin-deficient muscle

Loading next page...
 
/lp/springer-journals/complementary-nad-replacement-strategies-fail-to-functionally-protect-Vl0IlpLX2y
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
eISSN
2044-5040
DOI
10.1186/s13395-020-00249-y
Publisher site
See Article on Publisher Site

Abstract

Background: Duchenne muscular dystrophy (DMD) is a progressive muscle wasting disorder stemming from a loss of functional dystrophin. Current therapeutic options for DMD are limited, as small molecule modalities remain largely unable to decrease the incidence or mitigate the consequences of repetitive mechanical insults to the muscle during eccentric contractions (ECCs). Methods: Using a metabolomics-based approach, we observed distinct and transient molecular phenotypes in muscles of dystrophin-deficient MDX mice subjected to ECCs. Among the most chronically depleted metabolites was nicotinamide adenine dinucleotide (NAD), an essential metabolic cofactor suggested to protect muscle from structural and metabolic degeneration over time. We tested whether the MDX muscle NAD pool can be expanded for therapeutic benefit using two complementary small molecule strategies: provision of a biosynthetic precursor, nicotinamide riboside, or specific inhibition of the NAD-degrading ADP-ribosyl cyclase, CD38. Results: Administering a novel, potent, and orally available CD38 antagonist to MDX mice successfully reverted a majority of the muscle metabolome toward the wildtype state, with a pronounced impact on intermediates of the pentose phosphate pathway, while supplementing nicotinamide riboside did not significantly affect the molecular phenotype of the muscle. However, neither strategy sustainably increased the bulk tissue NAD pool, lessened muscle damage markers, nor improved maximal hindlimb strength following repeated rounds of eccentric challenge and recovery. (Continued on next page) * Correspondence: eugene.l.stewart@gsk.com Computational Sciences, Molecular Design, GlaxoSmithKline R&D, Collegeville, PA, USA Full list of author information is available at the end of the article © The Author(s). 2020 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://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Frederick et al. Skeletal Muscle (2020) 10:30 Page 2 of 14 (Continued from previous page) Conclusions: In the absence of dystrophin, eccentric injury contributes to chronic intramuscular NAD depletion with broad pleiotropic effects on the molecular phenotype of the tissue. These molecular consequences can be more effectively overcome by inhibiting the enzymatic activity of CD38 than by supplementing nicotinamide riboside. However, we found no evidence that either small molecule strategy is sufficient to restore muscle contractile function or confer protection from eccentric injury, undermining the modulation of NAD metabolism as a therapeutic approach for DMD. Keywords: MDX, NAD+, CD38, NR, Eccentric, Injury, Metabolomics, Therapeutics Background the buffer of isolated MDX muscles was reported to im- Dystrophinopathies are a class of diseases manifesting pri- prove fatigue resistance ex vivo [6]. In dystrophic mice har- marily in skeletal muscle and caused by a variety of muta- boring mutations in the FKRP gene, aberrant glycosylation tions in the 2.4 Mb dystrophin gene which render the of the DGC component, alpha-dystroglycan, can be par- dystrophin protein inactive. Dystrophin is a central force- tially overcome by supplementing the drinking water with transducing element of skeletal muscle, connecting the 5% ribitol, the substrate of the mutated enzyme [7]. These actin cytoskeleton to the extracellular matrix via the studies suggest that specific aspects of the pathophysiology dystrophin-glycoprotein complex (DGC). In patients with of muscular dystrophies arise from metabolic bottlenecks Duchenne muscular dystrophy (DMD), the absence of that may be bypassed using exogenous small molecules. functional dystrophin leads to limb muscle weakness, Nicotinamide adenine dinucleotide (NAD or NAD) is an followed by gradual muscle atrophy, cardiomyopathy, and essential metabolic co-factor, which has been directly impli- premature death. Dystrophin-deficient muscle is especially cated in the maintenance of muscle mass and function dur- susceptible to damage following eccentric contractions, in ing sarcopenia and other dystrophic states [8–11]. NAD has which the muscle lengthens while generating opposing also been found to be depleted in the muscle of MDX mice force, as occurs during the act of sitting, descending stairs, [10, 12], suggesting potential for therapeutic intervention. and other activities of daily life [1]. Damaged muscle fibers At present, two complementary strategies exist for manipu- then undergo repeated cycles of clearance by the immune lating tissue NAD pools. The first, and most commonly system and replacement by newly differentiated progeni- studied strategy, is to enhance NAD production by supple- tor cells, in a process that spans weeks. Though progress menting biosynthetic precursors, such as nicotinamide ribo- has been made in addressing the primary defects in dys- side (NR) or nicotinamide mononucleotide, to cells with a trophin via exon skipping or gene-replacement therapies functional NAD salvage pathway [13]. The second strategy [2], means of mitigating the effects of eccentric injuries to is to inhibit the enzymatic consumption of NAD by several the muscle of DMD patients have primarily been limited classes of enzymes, including sirtuins, poly-ADP-ribose to treatment with palliative anti-inflammatory drugs. polymerases (PARPs), and ADP-ribosyl cyclases (ARCs). The ability of muscle to harness chemical energy through CD38, the most widely expressed ARC, has been therapeut- aerobic and anaerobic respiration is inherently linked to its ically targeted for indications ranging from multiple mye- physical structure. Accordingly, analytical techniques such loma to neurodegeneration [14]. Our group has previously as NMR and mass spectrometry can detect chemical bio- shown that synthetic small molecule antagonists of CD38 markers of specific muscular dystrophies, which correlate are capable of acutely increasing NAD in muscle and liver with the cause and severity of the disease, and can aid in tissue [15, 16], and others have reported that chronically identifying points of therapeutic intervention. For example, dosing one of these thiazoloquinolin(on)es, known in the lit- it has long been appreciated that the muscle of adult erature as compound 78c, mitigated structural remodeling dystrophin-deficient MDX mice contains lower levels of and functional decline in the muscle of aged mice [11]. Des- energy-storing and redox-active metabolites, such as pite these preliminary findings, the extent to which NAD phospho-creatine, ATP, and beta-hydroxybutyrate [3], yet depletion is a pathologically relevant or therapeutically tract- high levels of taurine and non-polar amino acids [4]. More able feature of dystrophinopathies remains unclear. recently, distinct metabolomes were also identified in popu- Recent reports have specifically investigated the poten- lations of the MDX muscle-resident cells involved in regen- tial for NAD supplementation to counteract the pathology eration, including satellite cells and adipose progenitors [5]. of DMD [10, 17]. Here, we have examined this topic fur- Attempts to preserve muscle function by restoring specific ther by characterizing the global metabolome of MDX metabolic intermediates or co-factors in dystrophic muscle muscle following eccentric challenge and identifying dis- have shown hints of efficacy. For example, supplementing tinct stages of metabolic crisis and repair on the biochem- the TCA cycle intermediate, alpha-ketoglutarate (aKG), to ical level. We found evidence that NAD is depleted both Frederick et al. Skeletal Muscle (2020) 10:30 Page 3 of 14 Fig. 1 The metabolome of dystrophin-deficient muscle acutely responds to eccentric challenge. a Principal component analysis of the global metabolic profiles of MDX gastrocnemius muscles before and after eccentric contractions. Unchallenged WT muscles were included as a negative control. Arrows indicate the early chronological progression of injured groups and ellipses indicate 95% confidence intervals. b Heat map of metabolites with significantly different ion abundances in MDX muscle compared to WT (left) and sequential post-injury time points in MDX mice compared to uninjured controls (right). c Volcano plot of biochemical pathways significantly altered in MDX muscle at baseline. Significantly altered metabolites were used to calculate pathway impact and highly affected pathways are indicated. d Fold change of pathway impact scores during the acute (2 h) injury stage of MDX muscle compared to MDX baseline. e Volcano plot of metabolites significantly altered in MDX muscle at baseline. Example components of the NAD metabolome (red), glycolytic intermediates (blue), and polyamine pathway intermediates (green) are indicated. Dashed line indicates the significance threshold of non-adjusted p < 0.05. f Time course box and whisker plots of selected components of the NAD metabolome, glycolytic intermediates (g), and polyamine pathway (h) over 14 days post-challenge. N = 6 mice per time genotype and time point. Whiskers represent minimum and maximum values. Significance was determined by one-way ANOVA with Tukey’s post hoc test (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant; all p values non-adjusted) versus unchallenged WT (red) or MDX (blue) samples acutely and chronically in the muscle of MDX mice with metabolites or lipids. Principal component analysis of the wide-ranging metabolic consequences. We further metabolic profiles demonstrated a clear clustering of sam- attempted to reverse the NAD depletion using a novel, ples by time and genotype (Fig. 1a). Furthermore, the highly potent imidazoquinoline inhibitor of CD38 and MDX samples showed a well-defined chronological pro- compared its efficacy to nicotinamide riboside during re- gression following damage. The clustering of samples ob- peat bouts of eccentric challenge and recovery. Our results served 2 h post-damage showed distinct separation from may guide the development of NAD-targeting therapeu- the other time points, suggesting a period of acute crisis in tics for muscle diseases. the muscles. This was the only time at which the majority of significantly altered metabolites appeared depleted Results when compared to unchallenged controls (Fig. 1b), Dystrophin deficiency alters the muscle NAD metabolome potentially indicating rapid degradation, release from and energy producing pathways the tissue, or a synthetic bottleneck. The pattern of the To assess the biochemical adaptations of muscle to dys- post-injury sample groups from days 2-7 reflected a trophin deficiency, we analyzed the gastrocnemius mus- transition from the acute response to a repair and cles of MDX mice before and after eccentric challenge recovery stage. By post-injury day 14, the samples had a using an untargeted metabolic profiling platform. The chemical phenotype closely resembling that of the un- platform utilized a combination of GC/MS and LC/MS in injured MDX muscles (Fig. 1a-b). positive and negative ion mode to identify 762 chemical Comparing the metabolic profiles of MDX mice to entities, of which 552 were annotated as either polar wildtype (WT) mice at baseline, we identified several Frederick et al. Skeletal Muscle (2020) 10:30 Page 4 of 14 biochemical pathways with a disproportionate impact on the level of GAPDH, resulting in a characteristic buildup the muscle. Among the highest confidence pathways (p of intermediates in the pentose phosphate pathway [8, < 10^ −5 and pathway impact> 0.5) were those relating 18]. Consistent with this model, we observed a signifi- to the biosynthesis and metabolism of amino acids, such cant increase in ion counts for glucose-6-phosphate as arginine, phenylalanine, tyrosine, and tryptophan, as (G6P), dihydroxyacetone phosphate (DHAP), ribose-5- well as that of nicotinate and nicotinamide metabolism phosphate, and a positive trend in sedoheptulose-7- (Fig. 1c). Comparing the pathway impact scores from phosphate (S7P) in MDX muscle (Fig. 1e, g). Addition- the acutely injured (2 h) to uninjured MDX muscle, we ally, we noted an increase in several poly-cationic species were surprised to find glycolysis as the most responsive of polyamines, known to be derived from arginine (Fig. pathway (Fig. 1d). Additional pathways relating to energy 1e, h). Collectively, this pattern indicates a metabolic production or storage, including those of fatty acid syn- shift in MDX muscle at a steady state, partially stem- thesis, pyruvate metabolism, and the TCA cycle, were ming from the loss of the metabolic co-factor, NAD. also injury responsive. The robust response of the global metabolome at 2 h Upon examining the specific metabolites altered at post-eccentric challenge led us to investigate NAD-related baseline in the MDX muscle, we found a strikingly lower metabolites at this and subsequent time points. As far as the abundance of NAD than almost any other metabolite platform could resolve, NAD itself did not appear to re- (Fig. 1e, f). Consistent with the pathway analysis, several spond to the challenge (Fig. 1f). However, Nam was acutely other nicotinamide-containing metabolites and glycolytic depleted by more than one-third after 2 h, presumably intermediates also showed highly variable abundance. restricting any residual activity of the NAD salvage pathway, Primary NAD deficiency in mouse muscle and cultured and gradually returned to baseline after 4 days. Nam myotubes has been shown to restrict glycolytic flux at homeostasis was further altered by a doubling in the levels Fig. 2 In vivo characterization of a novel synthetic CD38 inhibitor. a Inhibition of the base-exchange activity of recombinant human (rh) or mouse (rm) CD38 protein using the established CD38 inhibitor 78C and the novel inhibitor GSK978A. Sigmoidal dose-response curves were fitted to replicate experiments (n = 12-15). Molecular structures and calculated pIC50 values are indicated. b Stability of 500 nM GSK978A in liver microsomal fractions from selected preclinical species over time. Half-lives (t ) were calculated from linear regressions. N = 3 experiments per 0.5 species. c Pharmacokinetics of GSK978A in mouse blood after an oral dose of 10 mg/kg. The dashed line indicates approximate IC50 of 12 ng/mL. N = 3 mice. d Intratissue compound exposures 2 h after a range of oral doses of GSK978A. N = 4 mice per dose. WAT, white adipose tissue. e Tissue-specific pharmacodynamics measured as NAD content relative to an internal standard (IS) 2 h after 0-30 mg/kg oral doses of GSK978A. Significance was determined by one-way ANOVA with Tukey’s post hoc test (*p < 0.05, ****p < 0.0001 compared to vehicle controls). f Tissue- specific pharmacodynamics as a function of exposure measured as normalized change in NAD content 2 h after 0-3 mg/kg doses of GSK978A. Lines indicate sigmoidal best-fit regressions approaching the indicated Bmax limits. N = 4 mice per dose. Mice were WT obese males for subpanels c-f. Error bars represent SEM Frederick et al. Skeletal Muscle (2020) 10:30 Page 5 of 14 of 1-me-Nam in the 2 days following injury, which only nor- GSK978A was more soluble and outperformed 78c in a malized after 14 days. Evidence of a further constriction in chromosomal stability test of genotoxicity, indicating im- glycolysis also emerged post-injury: levels of DHAP, imme- proved suitability for long-term administration (data not diately upstream of GAPDH, acutely increased in an oppos- shown). GSK978A was also predicted to have low intrinsic ing pattern to that of lactate, a glycolytic end-product (Fig. clearance in several small animal preclinical species, includ- 1g). The most striking indicator of the repair phase was the ing mice and rats, but not larger cynomolgus monkeys (Fig. appearance of polyamines, including spermine, spermidine, 2b). As literature suggested a role for CD38 in the preven- putrescine, and N-acetylputrescine, which were elevated in tion of diet-induced obesity [21], the drug metabolism and the days following injury (Fig. 1h and Supplemental Table pharmacokinetic characterization of the quinoline series 1). This class of biomolecules serves as a general marker of was originally performed in obese WT mice. To confirm cellular proliferation and is required for both myocyte differ- the slow clearance kinetics in vivo, we administered a single entiation and alternative macrophage activation [19, 20]. intermediate oral dose of 10 mg/kg and found the com- Consistently, in the case of NAD-related metabolites, glyco- pound still detectable in the blood after 24 h (Fig. 2c). We lytic intermediates, and polyamines, eccentric injury ampli- next performed a pharmacokinetic analysis of tissues sam- fied the disparities between MDX and WT muscle. pled 2 h after oral doses from 1-30 mg/kg. At these doses, the compound was identified within the liver, gastrocne- A novel synthetic CD38 antagonist increases NAD in mius, adipose, and brain tissues at exposures that well multiple tissues exceeded the IC50 of ~ 12 ng/mL (Fig. 2d). Accordingly, the Our group previously reported a series of novel chemical NAD content was found to be significantly elevated in the entities (NCEs), which potently inhibit the constitutive liver, muscle, and brain when normalized to an internal ana- NAD-degrading enzyme, CD38. Related screening efforts lytical standard (Fig. 2e). Despite high exposure, the NAD yielded the imidazoquinoline dubbed GSK978A, which ex- recovery from adipose was low and NAD changes were not hibited tenfold higher potency against mouse recombinant significant. However, in most tissues, peak NAD elevation CD38 than the human enzyme (Fig. 2a). This potency is of at least 30% was achieved at a dose of 3 mg/kg (Fig. 2f). orders of magnitude greater than that of natural products, To assess the potential pharmacodynamics of GSK978A such as quercetin, and approximates that of 78c, the best- in an eccentric challenge model, an acute study was per- studied synthetic CD38 inhibitor to-date [11, 14]. Yet formed in MDX mice following 5 days of dosing at 3 mg/ Fig. 3 Acute pharmacodynamics of GSK978A and NR in MDX hindlimbs following eccentric challenge. a Schematic of acute study design in MDX mice over 5 days (D1-D5) showing timing of eccentric contractions (ECC), assessments of contractility (CON), termination and tissue harvest (TERM), and compound administration. b Comparison of maximal tetanic force produced in the same limb at baseline and following eccentric challenge (ECC). Contractility (CON) was assessed over the 5-day protocol (D1-D5). c Masses of gastrocnemius muscles harvested from challenged (right, R) and contralateral (left, L) limbs at the study conclusion. d Total and (e) relative NAD content of gastrocnemius muscles harvested from challenged and contralateral limbs. f Total and (g) relative polyamine content of gastrocnemius muscles harvested from challenged and contralateral limbs. h Normalization of total polyamine to NAD contents in challenged limbs compared to the contralateral side. N = 8 MDX mice per group. Error bars represent SEM. Significance was determined by one-way ANOVA with Tukey’s post hoc test (**p <0.01, ***p < 0.001, ****p < 0.0001; ns, not significant or p value indicated) Frederick et al. Skeletal Muscle (2020) 10:30 Page 6 of 14 kg. Dietary NR, which has been suggested to improve the the physiology of MDX mice during three distinct stages: performance of MDX muscle [10], was included as a com- growth, recovery from an eccentric challenge, and recov- parator. Within 24 h of eccentric challenge, tetanic strength ery from a repeated challenge (Fig. 4a). We reasoned that was lessened by > 50% in all treatment groups, despite pres- this design would model the efficacy requirements of boys ervation of mass in the largest affected gastrocnemius mus- diagnosed with DMD. Beginning at 7-9 weeks of age, dur- cles (Fig. 3a-c). Interestingly, the challenged muscles also ing a period of rapid growth and peak muscle necrosis showed NAD depletion compared to the contralateral side, [22], MDX mice were randomized by hindlimb contractil- indicating that the muscle NAD pool does acutely respond ity and body weight, with a WT group included as a posi- to lengthening contractions (Fig. 3d). Mice treated with tive control for recovery. Over the course of 20 weeks, GSK978A, but not NR, showed a trend toward protection hindlimb weakness and increased CK release persisted in from this effect, though it could not be attributed to specific the MDX mice compared to WT controls (Fig. 4b-c). The NAD elevation in either limb (Fig. 3e). As a biomarker of pattern of hindlimb strength and serum creatine kinase muscle repair, total muscle polyamines showed clear eleva- (CK) release generally trended downward as MDX ani- tion in the injured limbs with a trend toward protection by mals reached maturity, but remained unchanged in both GSK978A, especially when polyamines were normalized to compound-treated groups, compared to the vehicle- NAD content (Fig. 3f-h). These results suggested that a lon- treated controls. MDX mice also accumulated lean mass ger treatment regimen might be necessary to provide func- steadily over the course of the study, reflecting character- tional improvements to MDX mice. istic hypertrophy, in a manner that was treatment- independent (Fig. 4d). At the study conclusion, gastrocne- Chronic NAD repletion does not provide functional mius muscles from the challenged MDX limbs were found protection from repetitive eccentric challenges to be ~ 15% less massive than the contralateral side in all We next designed a long-term study with chronic ad- treatment groups, reflecting an inability to fully regenerate ministration of GSK978A or NR to longitudinally assess injured fibers that was not observed in the WT. Fig. 4 Chronic pharmacodynamics of GSK978A and NR in MDX hindlimbs following eccentric challenge. a Schematic of chronic study design in MDX and WT mice over 20 weeks (weeks 1-20) showing timing of eccentric contractions (ECC), assessments of contractility (CON), termination and tissue harvest (TERM), and compound administration. GSK978A was dosed daily and contractility was assessed at varying intervals. b Maximal tetanic force production and (c) serum creatine kinase activity assessed at baseline, midpoint, and study conclusion. Arrows indicate timing of eccentric challenges. Significance was determined by repeated measure two-way ANOVA with Tukey’s post hoc test relative to MDX vehicle controls (**p < 0.01, ****p < 0.0001; ns, not significant compared to MDX vehicle controls). d Body composition assessed by qNMR at baseline, week 8, and week 18 of treatment. e Mass and (f) mass ratio of injured and contralateral gastrocnemius muscles collected after 20 weeks of treatment. N = 9-12 mice per group. Mice are MDX unless otherwise noted. Error bars represent SEM. Significance was determined by one-way ANOVA with Tukey’s post hoc test (****p < 0.001, **p < 0.01; ns, not significant) Frederick et al. Skeletal Muscle (2020) 10:30 Page 7 of 14 Surprisingly, contralateral muscles tended to be largest in the basis for our rationale. To address this question, we mice treated with GSK978A (Fig. 4e-f), which maybea analyzed the uninjured gastrocnemius muscles from the consequence of altered gait mechanics to favor the contra- chronically treated mice using a second untargeted lateral side, as it was not reflected in total lean mass. metabolomics platform, which utilized LC/MS to pro- Hindlimb strength was also serially assessed following vide broader coverage of the negatively charged metabo- eccentric challenges beginning at week 10 of treatment, to lome, including organic acids. This platform detected determine whether treated groups were protected from in- 3415 putatively annotated polar metabolites, based on jury or recovered faster. Compared to WT controls, MDX mass. When comparing the metabolites significantly al- mice showed approximately double the functional deficit tered between vehicle-treated MDX and WT groups, a within 1 day of eccentric challenge, despite similarly reversing pattern emerged in the GSK978A-treated mus- shaped tetani, but neither parameter was affected by cles, which was not observed in the NR-treated MDX GSK978A nor NR (Fig. 5a-b). Furthermore, MDX hin- animals (Fig. 6a). This anti-correlation was confirmed dlimbs showed highly similar recovery kinetics between with robust significance (R = −0.56, p < 1E^ −100) only the first and second challenges, while WT controls in the GSK978A treatment group (Fig. 6b-c). rebounded faster after the second bout (Fig. 5c-d). The ab- Suspecting that such a dramatic reversion effect may have sence of a protective repeated-bout effect in MDX hin- resulted from restoration of a pleiotropic co-factor, like dlimbs may reflect the fact that adult dystrophin-deficient NAD, we again performed a pathway analysis. Surprisingly, muscles are preconditioned to such cycles of damage and we found the most significant enrichment in only two path- repair by activities of normal living. ways: purine metabolism and the pentose phosphate path- way (Fig. 6d). Since the untargeted platform was limited in CD38 inhibition significantly reverts the MDX muscle its ability to address the central question of whether posi- metabolome to the WT state tively charged NAD was specifically restored, we instead Given the lack of physiological protection conferred utilized an enzymatic NAD assay on the same tissue sam- upon MDX muscle by NAD-modulating compounds, we ples and found only statistically non-significant elevations suspected that the treatments were simply ineffective at compared to MDX vehicle controls (Fig. 6e). Though the correcting the basal metabolic imbalance that formed NR-treated MDX muscles still contained significantly less Fig. 5 Time course of hindlimb contractile function following consecutive eccentric challenges. a Longitudinal tetanic force produced by the hindlimbs beginning at week 10 of treatment and including repeat bouts of eccentric challenge (arrows) and recovery. b High-resolution traces of tetani immediately before and 24 h after the first eccentric challenge. c Three-week longitudinal hindlimb force production normalized to pre- challenge levels after the first and (d) second successive eccentric challenges. N = 9-12 mice per group. Mice are MDX unless otherwise noted. Error bars represent SEM. Significance was determined by repeated measure two-way ANOVA with Tukey’s post hoc test (*p < 0.05, ***p < 0.001, ****p < 0.0001; ns, not significant compared to MDX vehicle controls) Frederick et al. Skeletal Muscle (2020) 10:30 Page 8 of 14 Fig. 6 Metabolomic signatures of NAD replacement strategies in MDX muscle. a Heatmap of ion counts for polar metabolites reaching significance (adjusted p < 0.05) in MDX muscle compared to WT controls (left column) aligned with the same metabolites identified in GSK978A- treated (center column) and NR-treated (right column) muscles after 20 weeks of treatment. b Correlation of the changes in ion abundance induced by GSK978A and (c) NR supplementation in MDX muscle after 20 weeks compared to changes driven by genotype alone. Colored points indicate metabolite ions significantly altered in at least one of the comparisons and the dashed line indicates the least-squares regression of these points. R indicates the correlation coefficient of the significant (adjusted p < 0.05) ions and p indicates the significance level of the associated Pearson’s correlation. d Volcano plot of biochemical pathways significantly altered in MDX muscle by GSK978A. Significantly altered metabolites were used to calculate pathway impact and highly affected pathways are indicated. e Total NAD content of gastrocnemius muscles harvested from contralateral limbs. Significance was determined by one-way ANOVA with Tukey’s post hoc test (**p < 0.01; ns, not significant). f Box plots of ion abundances for representative members of glycolysis (G6P, F6BP, G3P), pentose phosphate pathway (R5P), and TCA cycle (CA, SA). N = 9-12 mice per group. Whiskers represent minimum and maximum values. Significance was determined by one-way ANOVA with Tukey’s post hoc test (p values adjusted for multiple testing) NAD than WT muscle, the intermediate GSK978A group multiple metabolic pathways that is not recapitulated by was statistically indistinguishable from vehicle-treated supplementing a NAD precursor. groups of either genotype. Though sustained NAD elevations were not observed, Discussion we found several metabolic reversions consistent with Advancements in high-throughput discovery metabolo- transient restoration of NAD-dependent processes. For mics have led to improved biomarker detection and novel example, NAADH, the neutral reduced form of a sus- therapeutic strategies for many diseases, yet determining pected biomarker of NAD repletion [23] or overload, whether a complex molecular signature is a cause or con- was detected only in GSK978A-treated muscles (Supple- sequence of pathology is highly context-dependent [24]. mental Table 1). Importantly, ion counts for compo- Given its central role in maintaining ATP generation via nents of the proximal glycolytic pathway, including both glycolysis and oxidative phosphorylation, NAD is a those for glucose-6-phosphate and fructose-1,6,-bispho- co-factor well-positioned to influence muscle mass and sphate, were largely normalized by GSK978A, as were performance, and its synthesis and degradation have been the GAPDH substrate, glyceraldehyde-3-phosphate, and suggested to be dysregulated in the absence of dystrophin the pentose phosphate intermediate, ribose-5-phosphate [10, 12]. Our study confirms that dystrophin-deficient (Fig. 6f). We also observed discrepant influence on TCA muscle maintains a diminished NAD pool, even in the ab- cycle intermediates and a subtle net effect of increased sence of deliberate physiological challenges. Our observa- ATP and decreased phosphocreatine in this group (Sup- tion that NAD can be further depleted within hours of an plemental Table 1), potentially indicating elevated eccentric challenge suggests that the repetitive nature of substrate-level phosphorylation. Collectively, these global these lengthening contractions during activities of normal metabolomic profiles of dystrophin-deficient muscle living may collectively drive chronic NAD depletion in highlight a restorative effect of CD38 inhibition on diseased tissue, providing insight into the etiology of Frederick et al. Skeletal Muscle (2020) 10:30 Page 9 of 14 DMD. Though the degree to which wildtype dystrophin This is especially true when considering NAD synthesis acutely mitigates, this process remains unresolved, and consumption fluxes, which vary widely between associated therapeutic indications would be limited to mouse tissues [30]. Indeed, the turnover of NAD in less life-threatening conditions, such as exercise recov- mouse muscle was recently found to be the slowest of ery or muscular trauma. any tissue tested [30], and the degradative activity of The mechanism of acute muscle NAD depletion likely other enzymes, such as PARPs, may predominate [31]. It reflects an imbalance in production and consumption is also possible that GSK978A primarily influences the fluxes. One model suggests that calcium dysregulation, global metabolic profile of muscle via ancillary effects on stemming from microtears in the sarcolemma, leads to a calcium homeostasis, which is known to be dysregulated burst of genotoxic reactive oxygen species and hyperacti- in the absence of dystrophin [26], or infiltrating immune vation of NAD-consuming PARPs [25, 26]. However, cells. Nonetheless, our conclusion that CD38 inhibition cleavage of NAD by PARPs would be expected to liber- did not functionally protect dystrophin-deficient muscle ate nicotinamide, and our data clearly indicate the op- is in line with that of Spaulding et al., who found that posite pattern. Rather, our finding that the methylated long-term administration of the CD38-inhibiting flavon- waste product, 1-me-Nam, is more abundant in the days oid, quercetin, failed to protect isolated MDX muscles following injury, suggests that the removal of nicotina- from contraction-induced injury [32]. mide equivalents from the cytosol by the enzyme nico- Though the muscle exposure of GSK978A was com- tinamide N-methyltransferase (NNMT) may effectively parable to that of other tissues, the 30% NAD elevation limit the re-synthesis of NAD from nicotinamide via the that we observed 2 h after dosing was modest by com- NAD salvage pathway. The regulation of NNMT activity parison to the liver, brain, and adipose, which more than is still poorly understood [27], and bulk tissue analysis is doubled NAD content over the same period (Fig. 2). We unable to resolve whether infiltrating cell types are re- also failed to detect significant changes in the muscle sponsible for the effect, but a consistent pattern of in- NAD pool following acute or chronic treatment. This creased NNMT expression has been previously reported may be an indication of several factors. First, because in muscle biopsies from patients with a variety of dys- muscle makes up a large percentage of body mass, trophic conditions [10]. Furthermore, since 1-me-Nam muscle-targeting drugs must have high volumes of dis- is prone to urinary secretion, it may be a useful indicator tribution. Limited solubility or excessive albumin bind- of efficacy for oligonucleotide-based therapeutics, such ing could effectively limit the interaction of quinolones, as those being tested in the FKRP mutant model of like GSK978A, with their intended target. Second, the limb-girdle muscular dystrophy [28]. We also found ele- accuracy and variability of NAD quantitation is highly vations in MDX muscle of several positive biomarkers dependent on extraction conditions and analytical tech- previously identified in models of primary muscle NAD niques. Our reliance on multiple mass spectrometry- depletion [8, 18], including DHAP and S7P. Consistently, based and enzymatic assays made it challenging to re- the large-scale metabolic imbalance secondary to NAD producibly measure subtle shifts in the NAD pool. depletion appeared to be amplified in MDX muscle by a Third, there may exist a biological upper limit to the bottleneck in glycolysis. The near-complete reversal of steady-state NAD content of muscle, as suggested by this imbalance by a small molecule CD38 antagonist earlier transgenic models [33, 34]. The reversibility of provided compelling evidence that depletion of one or the NMN adenyl transferase enzymes may effectively more of the pathway's cofactors is largely responsible for limit the expansion of the NAD pool in a tissue-specific the distinctive metabolomic fingerprint. manner. Lastly, it is possible that CD38 is not a major CD38 is a uniquely complex pharmacological target consumer of NAD in muscle, or that the enzyme expres- due to an unusual array of enzymatic activities and sion is downregulated during pathology. Such transcrip- modes of regulation. Additionally, the ability of the tional compensation has been observed previously in CD38 extracellular domain to function as a cell surface DMD muscle [10]. Nonetheless, a global assessment of ligand for CD31 represents a signaling mechanism that the treated tissues was largely consistent with a transient may be more effectively disrupted with monoclonal anti- or compartmentalized restoration of NAD-dependent bodies than small molecules and may contribute to the pathways. This restoration did not manifest in the form phenotype of CD38 knockout mice [21]. As an enzyme, of nicotinamide-containing metabolites, as predicted, CD38 can convert not only NAD but also NADP and but rather in a more stable impact on pentose phosphate nicotinic acid into calcium-mobilizing second messen- pathway intermediates. The observed impact on purine gers, such as cADPR and NAADP, in a manner metabolism is likely to be directly linked via dependent on both membrane topology and local pH normalization of ribose-5-phosphate, the pentose [29]. Thus, despite broad exposure, the specific pharma- phosphate-derived nucleotide precursor (Fig. 6d, f). codynamic effects of GSK978A might vary by cell type. Interestingly, purine metabolism was identified in our Frederick et al. Skeletal Muscle (2020) 10:30 Page 10 of 14 initial characterization of MDX muscle, but was not im- Methods plicated in the acute response to eccentric injury (Fig. Animal care and use 1c, d). Thus, GSK978A may be more effective at restor- Male C57BL/10ScSn-Dmd<mdx>/J (MDX) and C57BL/ ing chronic metabolic imbalances, rather than buffering 10ScSn/J (WT) aged 7-9 weeks were individually housed acute challenges. with ad libitum access to regular chow and water during a A central finding of our work is that specific antagon- 12: 12 h light: dark cycle under controlled temperature ism of CD38 is a more effective strategy than NR supple- and humidity. Pharmacokinetic studies were performed in mentation for restoring the metabolic imbalance of 5-month old C57BL6 mice fed a high fat diet (Research MDX muscle. The low micromolar IC50 of GSK978A Diets D12492). GSK978A was custom synthesized and achieves muscle NAD elevation similar to that of natural dissolved at 0.3 mg/mL in vehicle containing 0.5% hydro- products, such as NR [35], at less than 1% of the effect- xpropyl methyl cellulose and 0.1% polysorbate 80, pH 4. ive dose. The relative inability of NR to affect the MDX Ten milliliters per kilogram was administered daily in the muscle metabolome likely stems from its poor bioavail- morning by oral gavage. NR chloride was custom synthe- ability and short (< 3 min) half-life in the blood [8, 30], sized and dissolved in the drinking water at 12 mM, sterile which is consistent with the absence of pharmaco- filtered, and administered ad libitum in light protected dynamics observed in several clinical trials [36–38]. bottles, as described [8, 40]. All compounds were reformu- However, both NAD-modulating strategies employed in lated weekly. Body composition was monitored by quanti- our study failed to improve muscle function. We were tative NMR spectroscopy. All studies were conducted in largely unable to reproduce the results of the Auwerx accordance with the GSK Policy on the Care, Welfare and group, who observed a significant reduction of plasma Treatment of Laboratory Animals and were reviewed the creatine kinase and a nearly 50% protection from eccen- Institutional Animal Care and Use Committee either at tric challenge in the same strain of MDX mice treated GSK or by the ethical review process at the institution with NR for only 12 weeks [10]. The discrepancies may where the work was performed. derive from the fact that Ryu et al. assessed hindlimb torque around the knee joint instead of the ankle, and Compound screening administered an NR-triflate salt to mice, instead of the NCEs were tested for inhibition of CD38 transglycosida- NR chloride salt used in all neutraceutical formulations. tion or base exchange activity by colorimetric assay based Nonetheless, our finding that neither GSK978A nor NR on a published method [41] using recombinant mouse had any effect on the performance of MDX muscle over CD38 soluble domain protein purified from Pichia pas- time raises the question of whether biochemical imbal- toris. Briefly, 0.5 nM enzyme was incubated in buffer con- ance is pathologically relevant in the absence of a central taining 50 mM HEPES, pH 7.4, 1 mM CHAPS, 2 mM structural component like dystrophin. The favorable EDTA, 250 μM isonicotinaldehyde 2-pyridinylhydrazone, pharmacokinetics and brain penetrance of GSK978A 100 μM NAD, 1% DMSO, and 1-10,000 nM NCEs while suggest that the compound may have better efficacy in absorbance was monitored at 405 nm. Inhibitor potency certain neurodegenerative disorders, which feature NAD was calculated with the following equation: y = A+((B-A)/ depletion, such as Cockayne syndrome or xeroderma pig- (1 + (10^x/10^C)^D)), where A is the enzyme-free re- mentosa [39]. These and other indications for small mol- sponse, B is the inhibitor-free response, C is the log ecule NAD-modulators warrant further investigation. (IC50), and D is the hill slope. Global metabolomics following eccentric challenge Conclusion (external platform) In summary, MDX mice exhibit a chronic NAD deficit Samples were prepared using the automated MicroLab with broad effects on the biochemical phenotype of the STAR system (Hamilton Company, Franklin MA). Recov- hindlimb muscle. The distinct global metabolome of ery standards were added prior to the extraction process dystrophin-deficient muscle becomes acutely altered by for quality control purposes. Samples were lysed in ice- eccentric injury and can be partly restored by inhibition cold methanol and the resulting extract was divided into of CD38, though this intervention does not confer pro- four fractions: one each for analysis by reversed-phase tection against future injury. While primary NAD defi- UPLC-MS/MS with positive and negative ion mode elec- ciency may suffice to drive both muscle weakness and a trospray ionization, one for normal-phase UPLC-MS/MS transcriptional profile resembling dystrophy over time, platform, and one for analysis by GC-MS. Samples were our current data strongly suggest that the characteristic centrifuged at 13,000×g for 10 min and supernatants were muscle weakness of MDX mice cannot be overcome by dried under nitrogen. The MS system was a Thermo Sci- NAD replacement strategies alone and that such strat- entific Q-Exactive high resolution/accurate mass orbitrap egies would be unlikely to benefit patients with DMD. mass spectrometer operated at 35,000 mass resolution Frederick et al. Skeletal Muscle (2020) 10:30 Page 11 of 14 which was interfaced with a heated electrospray ionization ten runs using a standard solution of 16 organic acids. (HESI-II) source. Dried sample extracts were reconstituted Peak detection and global alignment of all scans was in solvents amenable to their respective method. One ali- performed using a custom metabolomics data processing quot was analyzed using acidic positive ion optimized con- pipeline. Detected ion m/z values and isotope distribu- ditions and another using basic negative ion optimized tions were matched against the human metabolome conditions in two independent injections using separate database [43] assuming [M-H] and [M-2H] species and 13 12 dedicated columns (Waters UPLC BEH C18-2.1 × 100 at most two C/ C exchanges to tentatively annotate mm, 1.7 μm). The extracts reconstituted in acidic condi- metabolites, with the method-inherent limitation of be- tions were gradient eluted using water and methanol con- ing unable to distinguish between isomers. taining 0.1% formic acid, while the basic extracts, which also used water/methanol, contained 6.5 mM ammonium Biochemical pathway analysis bicarbonate. A third aliquot was analyzed via negative Pathway analysis was performed on metabolites reaching ionization following elution from a HILIC column (Wa- an adjusted significance threshold of p < 0.05 for a given ters UPLC BEH Amide 2.1 × 150 mm, 1.7 μm) using a gra- comparison using the MetaboAnalyst 4.0. platform [44] dient consisting of water and acetonitrile with 10 mM and referencing the current KEGG pathway library for ammonium formate. The MS analyses alternated between mouse. Over-representation analysis was performed MS and data-dependent MS scans using dynamic exclu- using Fisher’s exact test and pathway topology analysis sion, and the scan range was from 80-1000 m/z. The sam- was performed using relative-betweeness centrality. ples designated for GC-MS analysis were derivatized under nitrogen using bistrimethyl-silyl-trifluoroacetamide Tissue pharmacokinetics and pharmacodynamics (BSTFA). The GC column was a 20 m × 0.18 mm ID, with To determine the stability of new chemical entities 5% phenyl; 95% dimethylsilicone phase. Samples were ana- in vitro, cryopreserved liver microsomes from several spe- lyzed on a Thermo-Finnigan Trace DSQ fast-scanning cies (Sekisui Zenotech, Japan) were thawed and diluted to single-quadrupole mass spectrometer using electron 0.9 mg/mL in 50 mM phosphate buffer, pH 7.4. NCEs in impact ionization at unit mass resolution. Raw data was DMSO were added at 0.5 μM to the microsome suspen- extracted, peak-identified, and quality control processed sion and pre-incubated for 5 min at 37 °C in a standard using Metabolon’s hardware and software. Peaks were cell culture incubator with shaking at 80 RPM. Clearance quantified using area-under-the-curve. Compounds reactions were started by the addition of 2 mM NADPH were identified by comparison to library entries of puri- and 5 mM MgCl cofactors, then 100 μL of microsome fied standards or recurrent unknown entities. Propri- suspension was removed from the reaction at designated etary visualization software was used to confirm the time points and mixed with 200 μLice-coldstopsolution consistency of peak identification among the samples. containing 80:20 methanol: acetonitrile containing 1% acetic acid. Microsome extracts were centrifuged at 10, Global metabolomics following chronic interventions 000×g for 15 min and supernatants were subjected to LC- (internal platform) MS/MS analysis (below). Metabolic stability expressed as Polar metabolites were extracted from frozen tissues fol- a percentage of the parent compound remaining over time lowing lysis in a fivefold excess of ice cold 70% ethanol was determined from the peak area ratios in order to cal- using a bead homogenizer. Tissue lysates were further culate the turnover rate constant, k, by linear regression diluted 1:20 in 70% ethanol, incubated at 75 °C for 3 and half-life according to the equation t = ln(2)/k.For 0.5 min, and centrifuged at 13,000×g for 10 min. Superna- assessing NCE distribution and pharmacodynamics tants were lyophilized, resuspended in 0.1 mL water, and in vivo, 10 μL of blood was harvested from the mouse tail subjected to flow injection mass spectrometry. Non- vein, mixed with 50 μL of water and 40 μL of acetonitrile. targeted mass spectrometry of polar metabolites was Samples were sonicated for 5 min, vortexed for 5 min, and performed as described [42]. Briefly, Q-exactive Plus centrifuged at 2000×g for 20 min. Supernatants were di- (Thermo Scientific) in profile mode with scan range 50- luted 1:5 in water and subjected to LC-MS/MS analysis. 1000 m/z was calibrated according to manufacturer pro- Tissues were bead homogenized for 2× 1 min in a fourfold tocols. Resolution was set to 70,000 at 200 m/z with excess of ice-cold 80% acetonitrile and centrifuged at 13, automatic gain control target of 3E6 ions, 3.0 kV spray 000×g for 20 min. Supernatants were diluted 1:10 in water voltage, 120 ms maximum injection time, and 60 s acqui- and subjected to LC-MS/MS analysis. As an internal sition time. Samples were injected in a randomized se- standard, 1.5 μmol of O-NAD was spiked into the tissue quence and analyzed in negative ion mode using a matrix. LC-MS/MS was performed on an Agilent 1290 In- mobile phase consisting of 60% isopropanol, 40% water, finity system using a mobile phase of methanol containing 1mM NH F, 10 nM taurocholic acid, 20 nM homotaur- 0.1% formic acid and a Varian Polaris amide-C18 column ine. Quality control was performed before each batch of coupled to a Sciex API 4000 mass spectrometer. NAD Frederick et al. Skeletal Muscle (2020) 10:30 Page 12 of 14 peaks were normalized to the internal standard and drug in a cycling mix containing 0.1% BSA, 2% ethanol, concentrations were determined using a standard curve 100 μg/ml alcohol dehydrogenase, 10 μg/ml diaphorase, generated in the tissue matrix. 20 μM resazurin, and 10 μM flavin mononucleotide in 100 mM phosphate buffer. Enzymatic cycling at room Hindlimb eccentric challenge and longitudinal temperature produced resorufin, the fluorescence of contractility which was monitored over time at ex/em 544/590 nm. Mice were anesthetized using isoflurane (3%/L O ) and Muscle total polyamines were measured using a fluoro- placed on a warming pad with their right hind limbs re- metric total polyamine assay kit (K475-100, Biovision) strained at the knee and foot affixed to a force trans- according to the manufacturer protocol. Briefly, frozen ducer with motor-arm (Aurora Scientific Instruments, muscles were ground under liquid nitrogen and a 100 Aurora, ON). Platinum sub-dermal electrodes were mg portion was further dounce homogenized in 0.5 mL inserted dorsally and ventrally to the femur to apply of ice-cold homogenization buffer. Lysates were centri- electrical field stimulation (2.5 mA at 25 V) to the sciatic fuged at 5000×g for 5 min and supernatants were further nerve and trigger contraction of the plantarflexor mus- filtered through 10 kD molecular weight cutoff spin col- cles of the lower limb. Muscles were stimulated isomet- umns. Extracts were assayed by fluorometric enzymatic rically at a single twitch (200 μs pulse) and tetanic (150 assay and compared to a standard curve. Hz at 200 μs pulse for 0.8 s) frequencies to assess longi- tudinal force production over the course of the study. Statistics Eccentric injury was induced by subjecting hindlimbs to Data were compiled and analyzed using Microsoft Excel a series of 40 lengthening contractile stimuli, consisting and graphed using Graphpad Prism. Statistical tests of a sub-tetanic stimulation of 100 Hz at 200 μs pulse for (Student’s 2-tailed t test, one-way ANOVA, repeated 0.4 s, while the motorized footplate applied an eccentric measure two-way ANOVA, Tukey’s post hoc test, and rotational torque. Animals were returned to holding en- least-square correlation analysis) were calculated using closures and isometric titanic force was monitored to as- Graphpad Prism with a significance threshold of p < sess force deficit and recovery. 0.05, as indicated. For metabolomics data, p values were adjusted for multiple hypothesis testing using either Acute pharmacodynamics following eccentric challenge Benjamini’s and Hochberg’s method [45] (external plat- For acute eccentric challenge studies, male C57BL/ form) or Storey’s and Tibshirani’s method [46] (internal 10ScSn-Dmd<mdx>/J aged 24-26 weeks were individually platform data), and principal component analysis was housed and treated for 5 days, as above. Baseline body performed on the first two of ten components using cus- weight and contractility were assessed 1 week before the tom R scripts. start of dosing and used for group randomization. On day four of treatment, right hindlimbs were subjected to the Supplementary information eccentric damage protocol 15 min after oral dosing. On Supplementary information accompanies this paper at https://doi.org/10. 1186/s13395-020-00249-y. day five, oral compounds were dosed 15 min before con- tractility measurement and 60 min before sacrifice. Mus- Additional file 1. Frederick et al Supplemental Table 1. cles from both limbs were snap frozen and stored at −80 °C before analysis. Abbreviations NCE: Novel chemical entity; TCA: Tricarboxylic acid; DHAP: Dihydroxyacetone Creatine kinase measurement phosphate; ECCs: Eccentric contractions; NAD+ or NAD: Nicotinamide Mice were anesthetized using 3% isoflurane and venous adenine dinucleotide; NADP: Nicotinamide adenine dinucleotide phosphate; NAADH: Nicotinic acid adenine dinucleotide (reduced); NAADP: Nicotinic whole blood was collected in a microcapillary from the acid adenine dinucleotide phosphate; DGC: Dystroglycan complex; retro-orbital sinus. Blood was allowed to clot at room PARP: Poly-ADP-ribose polymerase; ARC: ADP-ribosyl cyclase; aKG: Alpha- temperature for 30 min, then centrifuged at 10,000×g for ketoglutarate; NR: Nicotinamide riboside; NNMT: Nicotinamide N-methyl transferase; Nam: Nicotinamide; S7P: Seduheptulose-7-phosphate; 5 min. The resulting serum samples were diluted 1:3 in F6BP: Fructose-1,6-bisphosphate; CK: Creatine kinase water and subjected to automated enzymatic assay (Beckman Coulter, Brea CA). Acknowledgements We wish to thank C. Haffner for expertise in chemical synthesis, J. McNulty, K. Morasco, N. Milliken, and Metabolon, Inc. for providing technical expertise, NAD and polyamine measurement and A. Hinken and H. Feldser for continued guidance and helpful revisions NAD was extracted from frozen muscles and measured of the text. by enzymatic cycling assay, as described [34]. Briefly, 50 Authors’ contributions mg of muscle was extracted in 0.5 mL 0.6 M perchloric DF, JB, JU, DS, FP, and HK designed experiments. DF, AM, MS, JV, JB, FP, and acid and diluted 1:100 in 100 mM phosphate buffer, pH JU performed experiments. AN, ES, and DS analyzed and graphed in vitro 8. Samples and NAD standards were further diluted 1:20 molecule validation and metabolomics data. DF wrote the manuscript. DF, Frederick et al. Skeletal Muscle (2020) 10:30 Page 13 of 14 AN, ES, FP, and HK edited the manuscript. The authors read and approved pathophysiology in a mouse model of Duchenne muscular dystrophy. PLoS the final manuscript. Genet. 2014;10(7):e1004490. 13. Yoshino J, Baur JA, Imai S-I. NAD+ intermediates: the biology and therapeutic potential of NMN and NR. Cell Metab. 2018;27(3):513–28. Funding 14. Hogan KA, Chini CCS, Chini EN. The multi-faceted ecto-enzyme CD38: roles This work was funded in its entirety by GlaxoSmithKline, Inc. and facilitated in immunomodulation, cancer, aging, and metabolic diseases. Front by postdoctoral fellowships awarded to DWF and AN. Immunol. 2019;10:1187. 15. Becherer JD, Boros EE, Carpenter TY, Cowan DJ, Deaton DN, Haffner CD, Availability of data and materials et al. Discovery of 4-amino-8-quinoline carboxamides as novel, The metabolomics datasets analyzed in the current study are available in the submicromolar inhibitors of NAD-hydrolyzing enzyme CD38. J Med Chem. online version of the article. 2015;58(17):7021–56. The datasets analyzed in the current study are available from the 16. Haffner CD, Becherer JD, Boros EE, Cadilla R, Carpenter T, Cowan D, et al. corresponding author upon request. Discovery, synthesis, and biological evaluation of thiazoloquin(az)olin(on)es as potent CD38 inhibitors. J Med Chem. 2015;58(8):3548–71. Ethics approval and consent to participate 17. Zhang H, Ryu D, Wu Y, Gariani K, Wang X, Luan P, et al. NAD+ repletion Not applicable. improves mitochondrial and stem cell function and enhances life span in mice. Science. 2016;352(6292):1436–43. https://doi.org/10.1126/science. Consent for publication aaf2693. Not applicable. 18. Oakey LA, Fletcher RS, Elhassan YS, Cartwright DM, Doig CL, Garten A, et al. Metabolic tracing reveals novel adaptations to skeletal muscle cell energy production pathways in response to NAD+ depletion. Wellcome Open Res. Competing interests 2019;3:147. The authors declare that they have no competing interests. 19. Luchessi AD, Cambiaghi TD, Hirabara SM, Lambertucci RH, Silveira LR, Baptista IL, et al. Involvement of eukaryotic translation initiation factor 5A Author details (eIF5A) in skeletal muscle stem cell differentiation. J Cell Physiol. 2009;218(3): Muscle Metabolism Unit, GlaxoSmithKline R&D, Research Triangle Park, NC, 480–9. Collegeville, PA, USA. Cellzome, GlaxoSmithKline R&D, Heidelberg, Germany. 20. Puleston DJ, Buck MD, Klein Geltink RI, Kyle RL, Caputa G, O’Sullivan D, et al. Target Sciences, Computational Biology, GlaxoSmithKline R&D, Collegeville, Polyamines and eIF5A hypusination modulate mitochondrial respiration and PA, USA. Computational Sciences, Molecular Design, GlaxoSmithKline R&D, macrophage activation. Cell Metab. 2019;30(2):352–363.e8. Collegeville, PA, USA. 21. Barbosa MTP, Soares SM, Novak CM, Sinclair D, Levine JA, Aksoy P, et al. The enzyme CD38 (a NAD glycohydrolase, EC 3.2.2.5) is necessary for the Received: 8 May 2020 Accepted: 22 September 2020 development of diet-induced obesity. FASEB J Off Publ Fed Am Soc Exp Biol. 2007;21(13):3629–39. 22. Dangain J, Vrbova G. Muscle development in mdx mutant mice. Muscle References Nerve. 1984;7(9):700–4. 1. Petrof BJ, Shrager JB, Stedman HH, Kelly AM, Sweeney HL. Dystrophin 23. Trammell SAJ, Schmidt MS, Weidemann BJ, Redpath P, Jaksch F, Dellinger protects the sarcolemma from stresses developed during muscle RW, et al. Nicotinamide riboside is uniquely and orally bioavailable in mice contraction. Proc Natl Acad Sci U S A. 1993 Apr 15;90(8):3710–4. and humans. Nat Commun. 2016;7:12948. 2. Chamberlain JR, Chamberlain JS. Progress toward gene therapy for 24. Sévin DC, Kuehne A, Zamboni N, Sauer U. Biological insights through Duchenne muscular dystrophy. Mol Ther J Am Soc Gene Ther. 2017;25(5): nontargeted metabolomics. Curr Opin Biotechnol. 2015;34:1–8. 1125–31. 25. Pirinen E, Cantó C, Jo YS, Morato L, Zhang H, Menzies KJ, et al. 3. Dunn JF, Frostick S, Brown G, Radda GK. Energy status of cells lacking Pharmacological inhibition of poly(ADP-ribose) polymerases improves dystrophin: an in vivo/in vitro study of mdx mouse skeletal muscle. Biochim fitness and mitochondrial function in skeletal muscle. Cell Metab. 2014;19(6): Biophys Acta BBA - Mol Basis Dis. 1991;1096(2):115–20. 1034–41. 4. Griffin J, Sang E, Evens T, Davies K, Clarke K. Metabolic profiles of dystrophin 26. Whitehead NP, Yeung EW, Allen DG. Muscle damage in mdx (dystrophic) and utrophin expression in mouse models of Duchenne muscular mice: role of calcium and reactive oxygen species. Clin Exp Pharmacol dystrophy. FEBS Lett. 2002;530(1-3):109–16. Physiol. 2006;33(7):657–62. 5. Joseph J, Cho DS, Doles JD. Metabolomic analyses reveal extensive progenitor cell 27. Pissios P, Nicotinamide N. methyltransferase: more than a vitamin B3 deficiencies in a mouse model of Duchenne muscular dystrophy. Metabolites. clearance enzyme. Trends Endocrinol Metab. 2017;28(5):340–53. 2018;8(4). https://doi.org/10.3390/metabo8040061. PMID 30282911. 28. Vannoy CH, Leroy V, Broniowska K, Lu QL. Metabolomics analysis of skeletal 6. Lindsay A, Chamberlain CM, Witthuhn BA, Lowe DA, Ervasti JM. muscles from FKRP-deficient mice indicates improvement after gene Dystrophinopathy-associated dysfunction of Krebs cycle metabolism. Hum replacement therapy. Sci Rep. 2019;9(1) [cited 2019 Jul 15]. Available from: Mol Genet. 2019 Mar 15;28(6):942–51. http://www.nature.com/articles/s41598-019-46431-1. 7. Cataldi MP, Lu P, Blaeser A, Lu QL. Ribitol restores functionally glycosylated 29. Malavasi F, Deaglio S, Funaro A, Ferrero E, Horenstein AL, Ortolan E, et al. α-dystroglycan and improves muscle function in dystrophic FKRP-mutant Evolution and function of the ADP ribosyl cyclase/CD38 gene family in mice. Nat Commun. 2018;9(1):3448. physiology and pathology. Physiol Rev. 2008;88(3):841–86. 8. Frederick DW, Loro E, Liu L, Davila A, Chellappa K, Silverman IM, et al. Loss 30. LiuL, SuX,Quinn WJ,Hui S, Krukenberg K, FrederickDW, et al.Quantitative of NAD homeostasis leads to progressive and reversible degeneration of analysis of NAD synthesis-breakdown fluxes. Cell Metab. 2018;27(5):1067–1080.e5. skeletal muscle. Cell Metab. 2016;24(2):269–82. 31. Bai P, Cantó C, Oudart H, Brunyánszki A, Cen Y, Thomas C, et al. PARP-1 9. Gomes AP, Price NL, Ling AJY, Moslehi JJ, Montgomery MK, Rajman L, inhibition increases mitochondrial metabolism through SIRT1 activation. Cell et al. Declining NAD(+) induces a pseudohypoxic state disrupting Metab. 2011;13(4):461–8. nuclear-mitochondrial communication during aging. Cell. 2013 Dec 19; 32. Spaulding HR, Ballmann CG, Quindry JC, Selsby JT. Long-term quercetin 155(7):1624–38. dietary enrichment partially protects dystrophic skeletal muscle. Kumar A, 10. Ryu D, Zhang H, Ropelle ER, Sorrentino V, Mázala DAG, Mouchiroud L, et al. editor. PLoS One. 2016;11(12):e0168293. NAD+ repletion improves muscle function in muscular dystrophy and counters global PARylation. Sci Transl Med. 2016;8(361):361ra139. 33. Costford SR, Brouwers B, Hopf ME, Sparks LM, Dispagna M, Gomes AP, et al. 11. Tarragó MG, Chini CCS, Kanamori KS, Warner GM, Caride A, de Oliveira GC, Skeletal muscle overexpression of nicotinamide phosphoribosyl transferase et al. A potent and specific CD38 inhibitor ameliorates age-related in mice coupled with voluntary exercise augments exercise endurance. Mol metabolic dysfunction by reversing tissue NAD+ decline. Cell Metab. 2018; Metab. 2017; [cited 2017 Nov 20]; Available from: http://linkinghub.elsevier. 27(5):1081–1095.e10. com/retrieve/pii/S2212877817306816. 12. Chalkiadaki A, Igarashi M, Nasamu AS, Knezevic J, Guarente L. Muscle- 34. Frederick DW, Davis JG, Dávila A, Agarwal B, Michan S, Puchowicz MA, et al. specific SIRT1 gain-of-function increases slow-twitch fibers and ameliorates Increasing NAD synthesis in muscle via nicotinamide Frederick et al. Skeletal Muscle (2020) 10:30 Page 14 of 14 phosphoribosyltransferase is not sufficient to promote oxidative metabolism. J Biol Chem. 2015;290(3):1546–58. 35. Cantó C, Houtkooper RH, Pirinen E, Youn DY, Oosterveer MH, Cen Y, et al. The NAD(+) precursor nicotinamide riboside enhances oxidative metabolism and protects against high-fat diet-induced obesity. Cell Metab. 2012;15(6):838–47. 36. Dollerup OL, Christensen B, Svart M, Schmidt MS, Sulek K, Ringgaard S, et al. A randomized placebo-controlled clinical trial of nicotinamide riboside in obese men: safety, insulin-sensitivity, and lipid-mobilizing effects. Am J Clin Nutr. 2018;108(2):343–53. 37. Dollerup OL, Trammell SAJ, Hartmann B, Holst JJ, Christensen B, Møller N, et al. Effects of nicotinamide riboside on endocrine pancreatic function and incretin hormones in nondiabetic men with obesity. J Clin Endocrinol Metab. 2019;104(11):5703–14. 38. Dollerup OL, Chubanava S, Agerholm M, Søndergård SD, Altıntaş A, Møller AB, et al. Nicotinamide riboside does not alter mitochondrial respiration, content or morphology in skeletal muscle from obese and insulin resistant men. J Physiol. 2020;598(4):731–54. https://doi.org/10.1113/JP278752. 39. Lautrup S, Sinclair DA, Mattson MP, Fang EF. NAD+ in brain aging and neurodegenerative disorders. Cell Metab. 2019;30(4):630–55. 40. Fang EF, Scheibye-Knudsen M, Brace LE, Kassahun H, SenGupta T, Nilsen H, et al. Defective mitophagy in XPA via PARP-1 hyperactivation and NAD(+ )/SIRT1 reduction. Cell. 2014;157(4):882–96. 41. Preugschat F, Tomberlin GH, Porter DJT. The base exchange reaction of NAD+ glycohydrolase: identification of novel heterocyclic alternative substrates. Arch Biochem Biophys. 2008;479(2):114–20. 42. Fuhrer T, Heer D, Begemann B, Zamboni N. High-throughput, accurate mass metabolome profiling of cellular extracts by flow injection-time-of-flight mass spectrometry. Anal Chem. 2011;83(18):7074–80. 43. Wishart DS, Feunang YD, Marcu A, Guo AC, Liang K, Vázquez-Fresno R, et al. HMDB 4.0: the human metabolome database for 2018. Nucleic Acids Res. 2018;46(D1):D608–17. 44. Chong J, Wishart DS, Xia J. Using MetaboAnalyst 4.0 for comprehensive and integrative metabolomics data analysis. Curr Protoc Bioinformatics. 2019;68(1) [cited 2020 Jun 26]. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/cpbi.86. 45. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B Methodol. 1995;57(1):289–300. 46. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci U S A. 2003;100(16):9440–5. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Journal

Skeletal MuscleSpringer Journals

Published: Oct 22, 2020

There are no references for this article.