Acetate Does Not Affect Palmitate Oxidation and AMPK Phosphorylation in Human Primary Skeletal Muscle Cells

Our recent in vivo human studies showed that colonic administration of sodium acetate (SA) resulted in increased circulating acetate levels, which was accompanied by increments in whole-body fat oxidation in overweight-obese men. Since skeletal muscle has a major role in whole-body fat oxidation, we aimed to investigate effects of SA on fat oxidation and underlying mechanisms in human primary skeletal muscle cells (HSkMC). We investigated the dose (0–5 mmol/L) and time (1, 4, 20, and 24 h) effect of SA on complete and incomplete endogenous and exogenous oxidation of 14C-labeled palmitate in HSkMC derived from a lean insulin sensitive male donor. Both physiological (0.1 and 0.25 mmol/L) and supraphysiological (0.5, 1 and 5 mmol/L) concentrations of SA neither increased endogenous nor exogenous fat oxidation over time in HSkMC. In addition, no effect of SA was observed on Thr172-AMPKα phosphorylation. In conclusion, our previously observed in vivo effects of SA on whole-body fat oxidation in men may not be explained via direct effects on HSkMC fat oxidation. Nevertheless, SA-mediated effects on whole-body fat oxidation may be triggered by other mechanisms including gut-derived hormones or may occur in other metabolically active tissues.


INTRODUCTION
In the last decades obesity has reached pandemic proportions worldwide (1), indicating the necessity to take action and prevent the development of obesity-related comorbidities. Obesity is characterized by an imbalance in energy intake and expenditure, with the excess energy stored in adipose tissue depots. It is increasingly clear that not adipose mass per se but rather adipose tissue dysfunction plays a central role in the observed metabolic derangements (1). A limited buffering capacity may result in systemic lipid overflow and an increased lipid supply to non-adipose tissues. Consequently, this may, together with an impaired oxidative capacity cause ectopic fat deposition in important metabolically active tissues (i.e. skeletal muscle, liver), which may link to the development of insulin resistance (2).
Of note, evidence is accumulating that energy and substrate metabolism and insulin sensitivity may be modulated by gut microbially derived metabolites, thus bringing attention to the cross-talk between gut-microbiota-derived metabolites and host metabolic health (3,4). In the colon, fermentation of indigestible food compounds by the resident microbiota plays a pivotal role in the production of short-chain fatty acids (SCFA), including acetate, propionate, and butyrate (3,5,6). Of these SCFA, acetate is the most abundant in the colon and has potential beneficial effects on energy expenditure and substrate metabolism (7). Circulating levels of acetate can reach physiological concentrations that range between 5 up to an average of 450 µmol/L, predominantly after prebiotic consumption (4,(8)(9)(10)(11)(12)(13)(14)(15). Subsequently, gut-derived acetate seems to have a broad metabolic role, since the SCFA-sensing G protein-coupled receptors (GPR, 41/43) are expressed in various tissues such as the adipose tissue (16), skeletal muscle, liver (17), and pancreatic beta cells (18,19). Circulating acetate has been shown to have antilipolytic effects in adipose tissue in in vitro (20)(21)(22), animal (20,23), and human studies (24)(25)(26)(27), thus potentially reducing lipid overflow, improving lipid buffering capacity and impacting peripheral insulin sensitivity. Other potential metabolic effects may include the induction of satiety mediated via the stimulation of glucagon-like peptide 1 (GLP-1) and peptide YY (PYY) secretion (28,29).
In addition to satiety, SCFA effects on intestinal gluconeogenesis and increased skeletal muscle fat oxidation have been observed at least in rodents (3). Furthermore, an in vitro study in L6 myotubes showed that acetate acutely (following 2 min) increased AMP-activated protein kinase (AMPK) phosphorylation via increased adenosine monophosphate (AMP)/adenosine triphosphate (ATP) ratio (30). Moreover, intragastric injection of acetate in mice showed a rapid increase in muscle AMP/ATP ratio and AMPK phosphorylation (31).
In addition to effects on skeletal muscle metabolism, acetate mediated upregulation of genes involved in fat oxidation (uncoupling protein 2, peroxisome proliferator-activated receptor alpha, and carnitine palmitoyl transferase 1) and increased AMPK phosphorylation in murine liver (32)(33)(34).
Although no actual substrate oxidation was measured, these studies suggest that acetate may increase liver and skeletal muscle oxidative capacity in an AMPK-mediated fashion, at least in rodent models. Our recently published human intervention study in overweight/obese normoglycemic men showed that acute distal, not proximal, colonic acetate infusions (180 mmol/L) increased fasting fat oxidation by around 25%, which may be possibly mediated via increased circulating acetate levels (7). Together, these data suggest that the observed in vivo effects on whole-body fat oxidation in overweight/obese humans may be partly mediated via a direct effect of circulating acetate on skeletal muscle fat oxidation. Therefore, we hypothesized that SA (sodium acetate) increases fat oxidation in a time-and dosedependent manner in differentiated human primary muscle cells (HSkMC).
Fat oxidation was investigated by measuring SA effects on both endogenous (intracellular) and exogenous (extracellular) 14 C-palmitate oxidation. Additionally, from a mechanistic perspective, we investigated SA-mediated effects on AMPK phosphorylation in HSkMC.

Cell Culture
HSkMC were cultivated in proliferation medium under 5% CO 2 at 37°C. For differentiation towards functional myotubes, medium was changed to differentiation medium when reaching 80% confluence, as described above. At day 8-11 of differentiation viable myotubes were used for fat oxidation experiments.

Cytotoxicity Assay
A fluorometric lactate dehydrogenase (LDH) assay was used for estimating the number of nonviable cells, as marker of cytotoxicity (#G7890, Promega). Differentiated cells were cultured without (0 mmol/L), or with physiological (0.1 mmol/L) and supraphysiological (1 and 5 mmol/L) concentrations of SA during 4 and 24 h ( Figure S1). Fluorescence was measured using a Spectramax plate reader (Molecular Devices, CA, USA). Percentage of cytotoxicity was calculated in comparison to Triton X-100 (100% lysis positive control) with the same incubation times.

Fat Oxidation
For measurement of in vitro fat oxidation cells were incubated with 14 C-labeled palmitate (250 µCi/ml; PerkinElmer, Boston, MA, USA) and non-labeled palmitate. Palmitate was coupled to fatty acid free BSA (7.5%) together with HEPES (100 mmol/L) and carnitine (40 mmol/L).The label solution contained 14 Clabeled palmitate (1 mCi/ml), 100 mM non-labeled (cold) palmitate, 0.25% BSA, 12.5 mM HEPES, and 1 mM L-Carnitine. After SA incubation, complete ( 14 CO 2 ) and incomplete ( 14 ASM, acid soluble metabolites) oxidation products were measured, as previously described (36). 14 ASM include acetyl-CoA, acetyl carnitines, ketone bodies, and TCA cycle intermediaries. Briefly, following incubation, medium was transferred to a custom-made Teflon 24-well CO 2 trapping plate that was sealed, 70% perchloric acid was injected (Hamilton syringe, 1705N) through a silicon layer in the lid directly into the media. This moved the CO 2 through a tunnel to an adjacent well where it was trapped in 1 N NaOH. After overnight trapping, complete and incomplete oxidation products were measured by Scintillation counting (using a Tri-Carb 2910 TR liquid scintillation analyzer, Perkin Elmer). For measuring endogenous fat oxidation, cells were pre-incubated for 24 h with 14 C-labeled palmitate (250 µCi/ml; PerkinElmer, Boston, MA, USA) without carnitine and washed prior to SA incubations.
Finally, after washing with TBST, last time with TBS, bands were visualized using ECL substrate (Super Signal West Femto, #34095, Thermo Scientific) according to the protocol supplied by the manufacturer using the Bio-Rad ChemiDoc imaging system (Biorad).

Statistical Analysis
Values are expressed as mean and standard deviation. Significance was determined using the nonparametric Mann Whitney U-test when comparing two groups or the Kruskal-Wallis H-test, for multiple comparisons. In case of significant Kruskal-Wallis, Dunn's post hoc test was performed.
Statistics were performed using the GraphPad Prism 5.0a software package (GraphPad Software, San Diego, CA, USA) and a P < 0.05 (two-sided P-value) was considered statistically significant.

Time-Dependent Effect of SA on Exogenous and Endogenous Palmitate Oxidation
First of all, both acute (4 h) as well as chronic (24 h) incubation with physiological (0.1 mmol/L) and supraphysiological concentrations (1 and 5 mmol/L) of SA did not affect LDH activity, indicating no cytotoxic effect of SA in HSkMC (see Figure S1). Secondly, we investigated whether SA has a timedependent effect on complete or incomplete exogenous/ endogenous 14 C-palmitate oxidation in differentiated HSkMC. Briefly, cells were incubated with a physiological concentration of SA (0.1 mmol/L) or without (0 mmol/L) for up to 24 h. As expected, and shown in Figure 1, an increase in fat oxidation was observed over time. However, both complete ( Figure 1A) and incomplete ( Figure 1B) exogenous fat oxidation were not affected when comparing SA to control treated cells after acute (1-4 h) and chronic (20-24 h) exposure. Next, to investigate the effect of SA on endogenous fat oxidation cells were first preincubated with 14 C-palmitate for 24 h and after removal of the label, cells were incubated with SA (0.1 mmol/L) or without (0 mmol/L) for up to 24 h. Although we found a significant increase over time, no differences were observed between SA and control treated cells in complete ( Figure 1C) and incomplete ( Figure 1D) endogenous fat oxidation at each timepoint.

Dose-Dependent Effect of SA on Exogenous or Endogenous Fat Oxidation
Subsequently, we investigated the dose-effect of SA on exogenous fat oxidation in differentiated HSkMC. Cells were incubated without (0 mmol/L), or with physiological (0.1, 0.25, 0.5 mmol/L) and supraphysiological (1 and 5 mmol/L) concentrations of SA for up to 20 h. However, as shown in Figure 2, SA did neither increase complete (Figure 2A) nor incomplete ( Figure 2B) exogenous fat oxidation as compared to control treated cells.
To investigate dose-dependent effect of SA on endogenous fat oxidation cells were first pre-incubated with 14 C-palmitate for 24 h and after removal of the label, cells were treated without (0 mmol/L), or with physiological (0.1, 0.25, 0.5 mmol/L) and supraphysiological (1 and 5 mmol/L) concentrations of SA for up to 20 h. In line with the exogenous oxidation data, we did not observe a dose-dependent increase in complete ( Figure 2C) palmitate oxidation as compared to control treated cells. However, we found a significant decrease in the incomplete oxidation ( Figure 2D), but only following incubation with the highest dose (5 mmol/L) of SA (P < 0.05).

Effect of SA on AMPK Phosphorylation
In line with our oxidation data, we did not observe a time (2 min, 10 min, and 24 h) or dose (0.1 and 0.5 mmol/L) dependent effect on total or phosphorylated Thr 172 -AMPKa content in HSkMC ( Figure 3). Of interest, Thr 172 -AMPKa phosphorylation was increased in our HSkMC model following 4-and 6-h incubations with the major AMPK activator AICAR (1 mmol/L) ( Figure S3). Nevertheless, this AICAR-mediated increase in AMPK phosphorylation was not accompanied by an increased complete nor incomplete exogenous/endogenous fat oxidation using AICAR (1 mmol/L) ( Figure S2). In contrast, treatment of HSkMC using an inhibitor (etomoxir, 100 µmol/L) showed marked changes in complete and incomplete palmitate oxidation, indicating that our HSkMC is dynamic and fully functional ( Figure S2).

DISCUSSION
Recently we showed that distal (not proximal) colonic infusions of acetate and SCFA mixtures (rich in acetate) increased circulating acetate levels (120 min) in overweight/obese humans (7,37). Of note, we found a positive relationship between increments in acetate levels and fasting whole-body fat oxidation. Therefore, we hypothesized that circulating acetate may directly affect fat oxidation in human skeletal muscle contributing to the observed increments in in vivo whole-body fat oxidation. To our knowledge, this is the first study that investigates the direct (dose/time) effects of SA on fat oxidation (endo-exogenous) and whether effects are dependent on AMPK phosphorylation in a HSkMC model. However, our data showed no time-or dose-dependent increase of SA on either endo-or exogenous palmitate oxidation in HSkMC. In line, AMPK phosphorylation was not affected by acute nor chronic SA treatment.
In vivo colonic infusions (37, 38) showed an acute effect on fasting whole-body fat oxidation in overweight-obese men with acetate concentrations reaching average plasma acetate levels of 40 µmol/L. However, our in vitro experiments showed no dose-or time effect of SA, with physiological concentrations (0-0.5 mmol/ L), reached after in vivo dietary fiber fermentation (10)(11)(12)(13), and no effect with supraphysiological concentrations (1-5 mmol/L). Although the actual interstitial/intracellular concentration of acetate in the skeletal muscle remains unknown, it has been suggested in rats (39) and humans (40) that acetate uptake in skeletal muscle is proportional to circulating levels. In support, using labeled 13 C SA in healthy lean individuals has shown 13 CO 2 breath recovery rates between 40-80%, which may indicate acetate uptake and utilization/oxidation in the TCA cycle of metabolically relevant organs (i.e. liver, muscle) (41)(42)(43). Of note, in our in vitro experiments SA was provided as a single bolus, and acetate may have been rapidly absorbed, and used into the tricarboxylic acid cycle (TCA) and saturate connecting metabolic pathways (44). In support, acute intravenous infusions of acetate (2.5 mmol per min for 1 h) in humans did not increase energy expenditure, which was partly explained by the fact that acetate might replace long chain fatty acids as preferred oxidation fuel in TCA-connected metabolic pathways (45). Of interest, a slight time-dependent decrease in endogenous (incomplete) oxidation was observed ( Figure 2D), which reached significance following incubation with the highest dose of SA (5 mmol/L) compared to control cells, possibly indicating preferential use of SA as oxidation fuel. Future studies using labeled SA are needed in order to investigate the metabolic fate of acetate in human skeletal muscle. Moreover, whether metabolic effects of acetate in human skeletal muscle are  14 C-palmitate exogenous oxidation (5 independent experiments) was measured following 20h incubation without (0 mmol/l) or with (0.1-5 mmol/l) SA. In addition, following 24h pre-incubation with 14 C-palmitate, (C) complete ( 14 CO 2 ) and (D) incomplete (ASM) 14 C-palmitate endogenous oxidation (4 independent experiments) was measured following 20h incubation without (0 mmol/l) or with (0.1-5 mmol/l) SA. Data expressed as mean and standard deviation and expressed as relative to control treated cells. P value corresponds to Kruskal-Wallis test. Post-hoc test significance is compared to control (0 mmol/l) indicated as asterisk (*) when P < 0.05. FIGURE 3 | Time and dose-dependent effects of SA on total and pAMPK in HSkMC. Cells were incubated without (0 mmol/l) and with (0.1 and 0.5 mmol/l) SA for 2 or 10 min and up to 24 h. After incubation, cells were lysed with RIPA buffer supplemented with protease and phosphatase inhibitors and samples were subjected to western analysis for total AMPKa (62kD band), phosphorylated AMPK (Thr 172 -AMPKa) and GAPDH (37kD band) was used as a loading control.
GPR-dependent remains unclear given the inconclusive expression of GPR 41 in our HSkMC model ( Figure S5) and therefore warrant further investigation of other GPRs (e.g. GPR43).
As mentioned above, our design (one single bolus of SA), may not mimic a continuous delivery of gut-derived acetate to the skeletal muscle. Acetate colonic absorption seems to be concentration dependent (46) and to gradually increase plasma levels, thus reaching skeletal muscle in a more continuous manner in vivo. Nevertheless, our doses were based on human prebiotic interventions that report circulating acetate concentrations up to 450 µmol/L (9)(10)(11)(12)(13)(14). In support of a continuous supply, a 4-week supplementation of a high amount of fermentable (type 2) resistant starch (30 g/d vs rapidly digestible starch 20 g) increased acetate uptake in skeletal muscle (and adipose tissue) in healthy subjects (47). Although no fat oxidation was measured, the intervention improved peripheral insulin sensitivity (euglycemic hyperinsulinemic clamp) (47), thus potentially indicating that a continuous supply of gut-derived acetate into the muscle (instead of a single bolus) is needed to induce major metabolic effects.
Based on previous animal data, acetate (oral injection) acutely increases phosphorylation of Thr 172 -AMPKa, at least in rat skeletal muscle (in vivo) and in vitro using L6 myotubes (30,31). The putative mechanism of acetate in the regulation of skeletal muscle fat oxidation, is through rapid (acute) catabolic conversion of acetate to acetyl CoA (mediated by acetyl Co A synthase), increasing the AMP/ATP ratio which subsequently increases AMPK phosphorylation. We replicated the exact same incubation time/dose as in above mentioned animal studies; however, we could not corroborate previous results. Importantly, our model is fully functional as observed by pronounced effects of Etomoxir. Nevertheless, our model may not be the ideal system to investigate (acute) AMPK-dependent effects on fat oxidation (See Figure S2) and/or effects may be donor-specific (48,49). Of note, AICAR administration in humans did not increase AMPK phosphorylation in skeletal muscle but increased hepatic fatty acid oxidation and lowered hepatic glucose production (50). Furthermore, AMPK-dependent fat oxidation might be limited in HSkMC and/or an AMPK-dependent effect on glucose uptake/oxidation may be preferentially activated (51,52) as well as muscle fiber type-dependent activation of AMPK (53) in HSkMC. Although, acetate has shown AMPK-dependent fat oxidation in rodents, acetate-mediated effect on AMPK activation in skeletal muscle may be species-specific (54).
Lastly, circulating acetate may modulate other metabolically active tissues in humans that may explain the in vivo increments in whole-body fat oxidation. First, SA may increase liver/adipose tissue fat oxidative capacity, since in vitro and rodent studies have shown increments in phosphorylated AMPK and total AMPK (31,32). In addition, acetate turnover rate is the highest in the liver of mammals as compared to other tissues (i.e. muscle and adipose tissues) indicating an important regulatory role of the liver in acetate metabolism (55). Second, other gut-derived metabolites such as butyrate may increase mitochondrial function and expression of fatty acid oxidative genes as observed in mice skeletal muscle (56,57). Kinetic studies demonstrated that around 24% of colonic acetate is converted via bacterial cross-feeding into butyrate in metabolically healthy adults (58).
Finally, SCFA might affect the secretion of gut-derived hormones such as PYY and GLP-1, which have been associated with increased whole-body fat oxidation and energy expenditure in humans (59,60). In line, data from our group and others reported increments of circulating PYY levels following colonic SCFA infusion in humans (37, 61). In agreement, a recent study reported that human skeletal muscle and muscle progenitor cells express PYY and its receptors, suggesting that our observed increase in whole-body fat oxidation might be partly explained via PYY-mediated effects on muscle fat oxidation (62). Our model showed no expression of the PYY receptor NPY2R during proliferation and differentiation (data not shown).
In addition, other gut-derived peptides warrant further investigation with respect to their respective roles on skeletal muscle fat oxidation under different metabolic status. For instance, ghrelin and gastric inhibitory polypeptide (GIP) may affect fat oxidation and thereby a protective role in lipid-induced skeletal muscle insulin resistance (63,64). Moreover, fibroblast growth factor (FGF) 15/19 secreted from small intestine should be further investigated regarding their involvement in human skeletal muscle fat oxidation (65).
In conclusion, our data showed no time-or dose-dependent increase of SA on endogenous or exogenous palmitate oxidation as well as no increments in AMPK phosphorylation following acute/chronic SA treatment in our human primary muscle cell model HSkMC derived from a lean insulin sensitive male donor. However, we cannot exclude that our previously reported in vivo effect of colonic acetate administration on fat oxidation in overweight individuals might be partly explained by the effect of other gut-derived metabolites and their signaling pathways (i.e. via PYY) on muscle fat oxidation or by direct effects of SCFA (i.e. butyrate) on other tissues (i.e. liver and adipose tissue). Furthermore, fat oxidation might be donor-dependent and/or species-specific.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Medical Ethical Committee Jessa Hospital, Hasselt, and Hasselt University, Belgium. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
MG, JJ, EC, and EB were responsible for the study concept and design, analysis and interpretation of the data, and critical revision of the manuscript for important intellectual content. MG, YE, and NH generated data. MG acquired all data, completed statistical analysis, and drafted the manuscript. EB obtained funding and supervised the study. All authors contributed to the article and approved the submitted version.

FUNDING
The study was supported by a Kootstra Talent grant from Maastricht University Medical Centre + (Applicant: EB, Candidate: EC) and MG's salary was paid by a grant from Consejo Nacional de Ciencia y Tecnologıá (CONACYT). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ACKNOWLEDGMENTS
We are thankful for all volunteers that participate in our studies.