The SGLT2 inhibitor empagliflozin promotes increased fatty acid oxidation in skeletal muscle cells
Stanislava Stevanovic, Parmeshwar B. Katare, Hilde Mari Volledal, Hege G. Bakke, Klemen Dolinar, Sergej Pirkmajer, D. Margriet Ouwens, G. Hege Thoresen, Eili T. Kase, Arild C. Rustan
TL;DR
This study shows that empagliflozin, an SGLT2 inhibitor, changes energy metabolism in muscle cells by increasing fatty acid and leucine breakdown while reducing glucose and acetoacetate use.
Contribution
The study reveals a novel metabolic effect of empagliflozin on skeletal muscle cells, specifically promoting fatty acid oxidation.
Findings
Empagliflozin increased fatty acid and leucine catabolism in human myotubes.
The drug reduced glucose and acetoacetate oxidation and glycolysis in muscle cells.
AMPK and ACC phosphorylation was increased by empagliflozin in human myotubes.
Abstract
In this study we investigated the potential for the sodium-glucose cotransporter 2 (SGLT2) inhibitor empagliflozin (EMPA) to modify energy metabolism in human primary skeletal muscle cells and mouse C2C12 skeletal muscle cells. The results showed that treatment of human myotubes with EMPA for 96 h decreased oxidation of exogenously added glucose and acetoacetate measured as CO2 production, whereas CO2 production from exogenously added fatty acids and leucine was increased compared to control cells. Uptake of acetoacetate by the cells was decreased by EMPA. Moreover, there were no EMPA-induced changes in glucose, fatty acid or leucine uptake by human myotubes, neither was lactate concentration in cell culture medium changed after exposure to EMPA. Treatment with EMPA increased phosphorylation of AMP-activated protein kinase (AMPK) and acetyl-CoA carboxylase (ACC) in human myotubes, while…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8- —https://doi.org/10.13039/501100005366Universitetet i Oslo
- —https://doi.org/10.13039/501100009707Diabetesforbundet
- —Freia Chokoladefabriks’s Medical Foundation
- —Marie Skłodowska-Curie grant
- —Anders Jahre’s foundation
- —University of Oslo (incl Oslo University Hospital)
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPancreatic function and diabetes · Diabetes Treatment and Management · Metabolism, Diabetes, and Cancer
Introduction
Sodium-glucose cotransporter 2 inhibitors (SGLT2i), also called gliflozins, are a class of drugs that lower blood glucose levels by preventing glucose reabsorption in the kidneys, resulting in increased urinary glucose excretion (Scheen 2014). In addition to glucose-lowering effects, SGLT2i reduce the risk of cardiovascular events (Cowie and Fisher 2020; Zinman et al. 2015). While the renal effects of SGLT2 inhibition are well characterized, emerging evidence suggests other metabolic effects in non-renal tissues (Dabravolski et al. 2022; Szekeres et al. 2021).
SGLT2 is mainly expressed in kidneys (Chen et al. 2010; Hiraizumi et al. 2024). Effects of SGLT2i treatment on skeletal muscle and other tissues may therefore occur via alternative mechanisms or by indirect effects, like calorie restrictions, most likely explained by the urinary glucose loss induced by these glucose transport inhibitors (Op den Kamp et al. 2022).
Skeletal muscle, which plays a crucial role in systemic glucose disposal and lipid metabolism, remains an underexplored target for SGLT2i. Some studies have investigated the effects of SGLT2i on skeletal muscle energy metabolism in vivo, however it is not known whether effects seen in vivo are direct effects on skeletal muscle or whole-body metabolic effects. In individuals with type 2 diabetes (T2D), insulin resistance in e.g. skeletal muscle is important for development of the disease (Silva Rosa et al. 2020). One study found that treatment with the SGLT2i empagliflozin improved insulin sensitivity in skeletal muscle of individuals with short duration of T2D (Goto et al. 2020). Treatment for 2 weeks with another SGLT2i, dapagliflozin, improved muscle insulin sensitivity while enhancing endogenous glucose production in men with T2D (Merovci et al. 2014). A study in mice also showed that SGLT2i ameliorated insulin resistance by increasing glucose uptake in skeletal muscle (Obata et al. 2016). Moreover, dapagliflozin treatment of subjects with T2D for 5 weeks caused adaptive changes in skeletal muscle substrate metabolism in vivo favoring metabolism of fatty acids and ketone bodies, reducing amino acid levels and glycolytic flux without changes in mitochondrial integrity and function (Op den Kamp et al. 2022). Few studies with SGLT2i have been done with muscle cells in vitro, however in one study with mouse C2C12 skeletal muscle cells, the SGLT2i canagliflozin reduced mitochondrial function and glycolytic metabolism (VanDerStad et al. 2024).
Taken together, previous literature showed, mainly in subjects with T2D, that long-term treatment with SGLT2i could influence energy metabolism in skeletal muscle, but whether SGLT2i also has direct effects on skeletal muscle remains mainly unknown. Hence, to address potential direct influence on skeletal muscle energy and amino acid metabolism, we wanted to examine the effect of SGLT2i on human skeletal muscle cells in vitro. We have focused on the ability of empagliflozin (EMPA) to modify energy metabolism in primary human skeletal muscle cells (HSMC) established from healthy subjects. In some experiments C2C12 mice myotubes were also used to address specific mechanistic issues.
Materials
Dulbecco’s Phosphate Buffer Saline (DPBS with/without Mg^2+^and Ca^2+^), Dulbecco’s Modified Eagle’s Medium (DMEM) with GlutaMAX™ high and low glucose, heat-inactivated foetal bovine serum (FBS), penicillin–streptomycin (10,000 IE/mL), human epidermal growth factor (hEGF), amphotericin B, trypsin–EDTA, Restore™ PLUS Western Blot stripping buffer, Super Signal™ West Femto Maximum Sensitivity substrate, Pierce™ BCA Protein Assay Kit, Power SYBR® Green PCR primers, MicroAmp® Optical Adhesive Film, MicroAmp® Optical 96-well Reaction Plate, primers for TaqMan PCR and High-Capacity cDNA Reverse Transcription Kit, were all from ThermoFisher Scientific (Waltham, MA, US). Insulin (Actrapid®) was from NovoNordisk (Bagsvaerd, Denmark). Bovine serum albumin (BSA), 2-mercaptoethanol, L-carnitine, D-glucose, oleic acid (OA, 18:1, n-9), 4-(2-hydroxyethyl)−1-piperazine-ethanesulfonic acid (HEPES), dimethyl sulfoxide (DMSO), dexamethasone, protease and phosphates II inhibitor cocktail, gentamicin, Tris–HCl, EDTA, EGTA, Triton X-100, bis-2-(5-phenylacetamido-1,3,4-thiadiazol-2-yl)ethyl sulfide (BPTES), etomoxir, UK5099, oligomycin A, BAM15, rotenone and antimycin A were obtained from Sigma-Aldrich (St. Louis, MO, US). [1-^14^C]oleic acid (OA, 59.0 mCi/mmol), D-[^14^C(U)]glucose (3.0 mCi/mmol), and L-[^14^C(U)]leucine (59.0 mCi/mmol) were from Perkin Elmer NEN® (Boston, MA, US).
[3-^14^C]acetoacetic acid (50–60 mCi/mmol) was from American Radiolabeled Chemicals (St. Louis, MA, US), while 96-well and 6-well Corning® CellBIND tissue culture plates were from Corning (Schiphol-Rijk, the Netherlands). QIAshredder and RNeasy Mini kit were from QIAGEN (Venlo, the Netherlands). ScintiPlate®−96 TC, UniFilter®−96 GF/B microplates, TopSeal®-A transparent film, Ultima Gold™ XR, and 96-well Isoplate® were obtained from PerkinElmer (Shelton, CT, US). XF24e and XF96e microplates were from Agilent Technologies, Inc. (Santa Clara, CA, US). Bio-Rad Protein Assay Dye Reagent Concentrate, Bio-Rad Precision Plus Protein™ Dual Color standard, bromophenol blue, Trans-Blot® Turbo™ Mini-size Transfer, Clarity Western ECL substrates, Tris/glycine/SDS buffer were from Bio-Rad (Copenhagen, Denmark). Antibodies against total Akt (#9272, 1:1000) and phosphorylated Akt (#9271S, 1:1000), total ACC (#3676, 1:1000) and phosphorylated ACC (#3661, 1:1000), and total AMPKα (#2532, 1:1000) and phosphorylated AMPKα (#2535, 1:1000) were from Cell Signalling Technology® Inc. (Beverly, MA, US). β-actin antibody (#A5441, 1:5000) was from Sigma-Aldrich (St. Louis, MO, US). EMPA was from Selleckchem.com (Houston, TX, US). Pierce™ BCA™ protein assay was from Thermo Fisher Scientific (Waltham, MA, US).
Ethics statement
Human skeletal muscle biopsies were obtained after informed written consent and approval by the Regional Committee for Medical and Health Research Ethics South-East, Oslo, Norway (reference number: REK11959). The study was conducted in accordance with the guidelines of the Declaration of Helsinki. All data are pseudonymous, and donor identities are unknown to the authors.
Methods
Cell culture
Human skeletal muscle cells from five donors, 48.6 ± 3.2 (mean ± SEM) years and body mass index of 26.2 ± 1.7 kg/m^2^ were used*.* Not all donors were used in all experiments, in some experiments cell cultures with pooled donors were used (2–3 donors). Human satellite cells were isolated from muscle biopsy samples from musculus vastus lateralis as previously described (Lund et al. 2018). In brief, satellite cells were collected from muscle biopsies, cleaned from fibroblasts, and grown to three passages. Myoblasts were cultured and proliferated in DMEM-GlutaMAX (5.5 mM glucose) supplemented with FBS (10%), streptomycin (25 µg/mL), penicillin (25 IU), gentamicin (50 ng/mL), HEPES (25 mM), BSA (0.05%), dexamethasone (0.39 µg/mL) and hEGF (10 µg/mL). Myoblasts were differentiated into myotubes when 80–90% confluent by changing the medium to DMEM-GlutaMAX (5.5 mM glucose) supplemented with FBS (2%), streptomycin (25 µg/mL), penicillin (25 IU), amphotericin B (1.25 µg/mL), gentamicin (50 ng/mL), HEPES (25 mM), and insulin (25 pM). The cells were cultured at 37 °C in a humidified atmosphere that contained 5% CO_2_. The medium was changed every 2–3 days. The cells were differentiated for 7 days. On the third day of differentiation, cells were treated with different concentrations of EMPA or 0.1% DMSO (control) for 96 h (Andreadou et al. 2017). EMPA did not seem to affect the differentiation of the cells, as no effects of EMPA on mRNA expression levels of the differentiation markers myosin heavy chain (MYH)2, MYH7 or solute carrier family 2 member 4 (SLC2A4) were found (Supplementary Fig. 1).
Mouse C2C12 myoblasts (American Type Culture Collection) were cultured in DMEM high glucose (25 mM) (PAA Laboratories) supplemented with 10% FBS (Invitrogen), 50 IU/mL penicillin, 50 μg/mL streptomycin and 2 mM glutamine. Myotube differentiation was induced by switching to DMEM containing 3% horse serum, and during last 96 h before harvesting the cells were cultured in in 1:1 DMEM low glucose (5.5 mM) and F10 medium containing 2% horse serum with 500 nM EMPA or 0.1% DMSO (control). Cells were used for assays after 7 days of differentiation.
Substrate oxidation assay
Human myoblasts (7000 cells/well) were seeded in 96-well CellBIND® microplates where they were proliferated and differentiated into myotubes. On day 3 of differentiation, cells were treated either with EMPA (10—1000 nM) or with 0.1% DMSO (control) for 96 h. After 7 days of differentiation, myotubes were washed twice with PBS. A substrate oxidation assay was used to measure substrate uptake and oxidation as previously described (Wensaas et al. 2007). Myotubes were given D-[^14^C]glucose (0.5 µCi/mL, 200 µM), [^14^C]oleic acid (0.5 µCi/mL, 100 µM) or [^14^C]acetoacetate (0.5 µCi/mL, 100 µM) substrate for 4 h. The glucose substrate was prepared in DPBS (with Mg^2+^ and Ca^2+^) that was supplemented with HEPES (10 mM) and BSA (10 µM), while the oleic acid substrate was added in DPBS containing HEPES (10 mM), BSA (40 µM), and L-carnitine (1 mM). Acetoacetate substrate was made in DPBS containing HEPES (10 mM). After incubation, the amount of ^14^CO_2_ produced from cellular respiration was trapped in the 96-well UniFilter® microplate added sodium hydroxide as CO_2_ absorber and cell-associated (CA) radioactivity that remained in the cells were measured by 2450 MicroBeta^2^ liquid scintillation counter (PerkinElmer). The concentration of protein in each culture well was measured using the Bradford protein assay, measured with VICTOR™X_4_ multilabel plate reader (PerkinElmer) to relate the ^14^CO_2_ and CA data to cellular protein content. The total cellular uptake of the substrate in each culture well was calculated as the sum of ^14^CO_2_ and CA.
Scintillation proximity assay
Radiolabeled substrates taken up and accumulated by adherent cells concentrate close to the scintillator embedded in the plastic bottom of each well (ScintiPlate®−96 TC, PerkinElmer) and provide a stronger signal than the radiolabel substrates dissolved in the medium above cells. Measurements of fatty acids accumulated in the myotubes by scintillation proximity assay (SPA) were performed in DMEM without phenol red with [^14^C]oleic acid (0.5 μCi/mL, 100 μM) and monitored for 0, 2, 4, 6, 8, 12 and 24 h during the incubation.
Leucine metabolism and incorporation into protein
Cells were cultured in 96-well tissue culture plates. After 3 days of differentiation, the cells were incubated with EMPA for 96 h. On day 7 of differentiation, the cells were incubated in DPBS with 10 mM HEPES with L-[^14^C]leucine (0.5 μCi/mL, 100 μM) for 4 h. After incubation, the amount of ^14^CO_2_ was trapped in the 96-well UniFilter® microplate and cell-associated (CA) radioactivity that remained in the cells measured by 2450 MicroBeta^2^ liquid scintillation counter (PerkinElmer).
Measurement of acid-soluble metabolites
Measurement of acid-soluble metabolites (ASM) reflects incomplete fatty acid β-oxidation and consists mostly of tricarboxylic acid cycle metabolites. ASM measurement was performed with the use of a method modified from Skrede et al. (Skrede et al. 1994). Myotubes were treated with 100 µM oleic acid for 4 h during trapping. From the radiolabelled incubation medium, 100 µL was transferred to a new Eppendorf tube and precipitated with 300 µL cold HClO_4_ (1 M) and 30 µL BSA (6%). Eppendorf tube was then centrifuged at 10 000 rpm/10 min/4 °C and 200 µL of the supernatant was counted by liquid scintillation (Packard Tri-Carb 1900 TR, PerkinElmer).
Lactate measurement in cell medium
Cell culture media was collected on day 7 of differentiation after 48 h of being exposed to the cells before substrate oxidation measurements were conducted. Lactate level in the media was measured in 15 µL of the cell media with the use of XPER L1 Lactate monitoring system (TaiDoc Technology Corporation).
Extracellular flux analysis
Mouse C2C12 cells were seeded in XF24e or XF96e microplates, 2.5 × 10^5^/well. EMPA (500 nM) was added for the last 96 h before the assay. After differentiation for 7 days, cells were washed twice with Seahorse XF DMEM medium containing 5 mM HEPES (Agilent Technologies) supplemented with 5.5 mM glucose, 2 mM glutamine and 1 mM sodium pyruvate. After 1 h incubation in a CO_2_-free incubator at 37 °C, cells were placed in the XF24e or XF96e extracellular flux analyzer (Agilent Technologies) for real time cell metabolic analysis as described (Görigk et al. 2022). At baseline, the cellular oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) were recorded three times at 6 min intervals. This was followed by injection with fuel substrate inhibitors to induce a predominant utilization of glucose, fatty acids and glutamine respectively. To elicit this, the first injection was either with (i) assay medium (vehicle), or with (ii) 3 μmol/L BPTES and 40 μmol/L etomoxir to induce the predominant use of glucose, (iii) 3 μmol/L BPTES and 2 μmol/L UK5099 to induce the predominant use of fatty acids, or (iv) 2 μmol/L UK5099 and 40 μmol/L etomoxir to induce the predominant use of glutamine, respectively. For the Mito Stress Test, 1.0 μmol/L oligomycin A, 1 μmol/L BAM15, and 0.5 μmol/L rotenone/antimycin A together with 2 µM Hoechst 33,342 (Thermo Fisher Scientific) were injected. After each injection, the changes in OCR and ECAR were recorded three times at 6 min intervals. The compounds used in the assays were dissolved in DMSO and stored as aliquots at −20 °C before use. After the assay, fluorescently labeled nuclei were quantitated using the Biotek Cytation 5 scanner. The extracellular flux data were normalized to fluorescently labeled nuclei and analyzed using WAVE software (version 2.6.0.31, Agilent Technologies). ECAR values were converted to proton efflux rates (PER). The baseline-adjusted OCR and PER values were used to calculate determinants of ATP production rate, mitochondrial function and glycolysis as described elsewhere (Selig et al. 2019; Lund et al. 2019).
RNA isolation and analysis of gene expression by qPCR
Myotubes were cultured in 25 cm^2^ flasks and differentiated for 7 days. On day 3 of differentiation, myotubes were treated with 0.1% DMSO (control) or EMPA (500 nM) for 96 h. RNeasy Mini Kit was used for the extraction of total RNA from myotubes according to the manufacturer`s protocol (Qiagen, Hilden, Germany). The quality and concentration of RNA were measured by a microvolume spectrometer Nanodrop ND-1000 (Thermo Fisher Scientific). The RNA was reversely transcribed to cDNA with a High-Capacity cDNA Reverse Transcription and TaqMan Reverse Transcription Reagents with the use of Perkin Elmer Thermal Cycle 9600 (25 °C for 10 min, 37 °C for 80 min, and 85 °C for 5 min). Primer Express® (Thermo Fisher Scientific) was used for designing of gene-specific primers (Table 1), and gene expression was determined by qPCR using StepOnePlus™ Real-Time PCR System (Thermo Fisher Scientific) with SYBR®Green Master-mix. All experiments were run for 44 cycles (95 °C for 15 s followed by 60 °C for 60 s). Expression levels of the genes were normalized to the housekeeping gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase). The housekeeping gene RPLP0 (ribosomal protein lateral stalk subunit P0) was also analysed and there were no differences between normalizing with GAPDH and RPLP0 (data not presented). Table 1. Forward and reverse primers used for PCRGeneForwardReverse**PDK4TTTCCAGAACCAACCAATTCACATGCCCGCATTGCATTCTTAMYH2AAGGTCGGCAATGAGTATGTCACAACCATCCACAGGAACATCTTCMYH7CTCTGCACAGGGAAAATCTGAACCCCTGGACTTTGTCTCATTGLUT4ACCCTGGTCCTTGCTGTGTTACCCCAATGTTGTACCCAAACTACACBCCA TCT TCG ACG TCC TGA ATA CTCTG TTT AAC ACA TAG GCG ATG TAA GCPPARATCC ACC TGC AGA GCA ACC ACCG GAG GTC TGC CAT TTT TCD36AGT CAC TGC GAC ATG ATT AAT GGTCTG CAA TAC CTG GCT TTT CTC AAUCP3AGG ACC TTT GCC CAA CAT CAT GAGT CCA GCA GCT TCT CCT TGA GCPT1BGAG GCC TCA ATG ACC AGA ATGGTG GAC TCG CTG GTA CAG GAACPT2ATC GTG CCC ACC ATG CACAC TGA GGT ATC TCC TAA TGG TGT CTTRPLP0CCATTCTATCATCAACGGGTACAAAGCAAGTGGGAAGGTGTAATCCGAPDHTGCACCACCACCTGCTTAGCGGCATGGACTGTGGTCATGAG
Western immunoblotting
Myotubes were cultured in 6-well plates and differentiated for 7 days. During the last 96 h of differentiation, myotubes were treated with 500 nM EMPA. For measurement of Akt: Before harvesting, some wells were treated with insulin (100 nM) for 15 min. Samples were harvested in Laemmli buffer (0.5 M Tris HCl, 10% sodium dodecyl sulphate (SDS), 20% glycerol, 10% 2-mercaptoethanol, 5% bromophenol blue). Proteins from lysates were electrophoretically separated by SDS-PAGE (Bio-Rad 4–20% Mini Protean®TGX™precast gels with Tris/glycine buffer) and transferred to nitrocellulose membranes. Primary antibodies were incubated overnight. Bio-Rad ImmunStar™ WesternC™-kit was used for the visualization of immunoreactive bands, Bio-Rad Chemidoc™XRS + system for their detection, and Image Lab (version 6.0) software for the quantification. For measurement of AMPK, ACC and β-actin: Myotubes were washed with ice-cold PBS and lysed in Laemmli buffer (62.5 mM Tris–HCl (pH 6.8), 10% glycerol, 2% SDS, 5% (v/v) 2-mercaptoethanol, 0.002% bromophenol blue) as described (Dolinar et al. 2018). Total proteins were quantified with Pierce 660 nm Protein Assay Reagent (Thermo Fisher Scientific). 10 μg of proteins and protein molecular weight markers (Amersham ECL Full-Range Rainbow Molecular Weight Markers RPN800E (Cytiva)) were separated by their molecular weight with electrophoresis in 4–12% Criterion XT Bis–Tris polyacrylamide protein gel (Bio-Rad) in XT MES electrophoresis buffer (Bio-Rad) and transferred to PVDF membrane (Immobilon-P membrane, Merck) with wet electrotransfer in transfer buffer (31.3 mM tris base, 240 mM glycine, 10% (v/v) methanol, 0.01% SDS). After the transfer, membranes were stained with Ponceau S (0.1% in 5% (v/v) acetic acid) to evaluate uniformity of protein loading and transfer. Membranes were then destained in tris-buffered saline with Tween 20 (TBST: 20 mM tris (pH 7.5, adjusted with HCl), 150 mM NaCl, 0.02% (v/v) Tween 20) and blocked with 5% dry skimmed milk in TBST for 1 h at room temperature. Following the blocking, membranes were incubated with primary antibodies in primary antibody buffer (20 mM tris (pH 7.5, adjusted with HCl), 150 mM NaCl, 0.1% BSA and 0.1% sodium azide) overnight at 4 °C and then with goat anti-rabbit or anti-mouse IgG-HRP conjugate (BioRad), diluted 1:10.000 in TBST with 5% dry skimmed milk for 1 h at room temperature. Finally, membranes were incubated with Immobilon Crescendo Western HRP substrate (Merck) and then the signal was captured with FUSION FX6 (Vilber). Membranes were first probed for phosphoproteins, then stripped of antibodies in stripping buffer (62.5 mM tris (pH 6.8, adjusted with HCl), 2% SDS, 0.7% (v/v) 2-mercaptoethanol), re-blocked and probed for total proteins as described above. Bands were analysed using Quantity One 1-D Analysis Software (version 4.6.9; Bio-Rad).
Presentation of data and statistics
The data are presented as mean ± SEM. Data were analysed using a student unpaired t-test (GraphPad Prism 9.3.1) or by ANOVA (IBM SPSS Statistics 30.0.0.0). Statistical significance was considered at p < 0.05. The total numbers of biological replicates (cell cultures) are indicated in figure legends. For human cells, each type of experiment was performed on myotubes from at least three donors.
Results
Effect of empagliflozin on fuel metabolism in human myotubes
To study the effect of EMPA on glucose metabolism, human myotubes were treated with different concentrations of EMPA for 96 h. Treatment with 500 nM and 1000 nM EMPA decreased glucose oxidation in myotubes (Fig. 1A), while glucose uptake was not affected by 500 nM EMPA (Fig. 1B). Moreover, lactate content of cell culture medium did not differ between EMPA-treated and untreated cells (Fig. 1C).Fig. 1. Effect of empagliflozin on glucose metabolism in human myotubes. Cultured human myotubes were treated with empagliflozin (EMPA) for the last 96 h of the differentiation period. Thereafter, the cells were incubated with 200 µM [^14^C]glucose for 4 h. (A) Oxidation of glucose (CO_2_ production) (B) uptake (CO_2_ + CA) of glucose, and (C) lactate concentration in cell culture medium (from the last 48 h before harvesting). Data are presented as mean ± SEM of (A-B) 20–40 biological replicates and (C) 12 biological replicates. Absolute values for untreated control for glucose oxidation 4.7 ± 0.5 nmol/mg protein, glucose uptake 16.4 ± 1.0 nmol/mg protein and lactate content in cell medium 16.9 ± 2.2 nmol/mg protein. *Statistically significant versus control (p < 0.05, unpaired t-test). CA: cell-associated
To determine effects of EMPA treatment on fatty acid oxidation in human myotubes, both CO_2_ production and acid-soluble metabolites (ASM, FA β-oxidation) were measured after treatment with EMPA (500 nM for 96 h). An increased CO_2_ production from exogenously added oleic acid (OA) in cells treated with EMPA was observed (Fig. 2A), while β-oxidation of OA (ASM) was reduced (Fig. 2B). The ratio between CO_2_ and ASM was markedly increased after treatment with EMPA (Fig. 2C), indicating that EMPA increased the efficiency for oxidation of exogenously added fatty acids in myotubes. On the other hand, uptake of OA by myotubes after 4 h was not changed after treatment with EMPA (Fig. 2D, Supplementary Fig. 2). We also performed scintillation proximity assay (SPA) to study real time fatty acid accumulation during a longer period (24 h) after exposure to EMPA, and it was observed that EMPA increased OA accumulation at all time points examined (Fig. 2E).Fig. 2. Effect of empagliflozin on fatty acid metabolism in human myotubes. Cultured human myotubes were treated with 500 nM of EMPA for the last 96 h of differentiation. In a)-d) the cells were incubated with 100 µM [^14^C]oleic acid for 4 h. (A) CO_2_ production from oleic acid (OA), (B) fatty acid β-oxidation (ASMs) of OA, (C) the ratio between CO_2_ production and β-oxidation (CO_2_/ASM), and (D) uptake of OA (sum of CO_2_ + CA). In (E) the cells were incubated with 100 µM [^14^C]oleic acid together with EMPA during the last 24 h of the 96-h incubation, and accumulation of [^14^C]oleic acid was measured by scintillation proximity assay. Data are presented as mean ± SEM of (A,** D)** 20–40, or (B,** C)** 12, and 40 biological replicates, (E). Absolute values for untreated control for OA oxidation 4.0 ± 0.4 nmol/mg protein, ASM 12.7 ± 1.5 nmol/mg and OA uptake 60.3 ± 6.0 nmol/mg protein. *Statistically significant versus control (figures A-C) (p < 0.05, unpaired t-test), and (p < 0.05, one-way ANOVA repeated measurements versus control) for figure E. EMPA: empagliflozin; ASM: acid soluble metabolites; CA: cell-associated
We next examined if ketone body metabolism was modified after exposure of myotubes to EMPA. As shown in Fig. 3, both uptake and oxidation of acetoacetate were reduced by EMPA treatment.Fig. 3. Effect of empagliflozin on metabolism of acetoacetate in human myotubes. Cultured human myotubes were treated with 500 nM of EMPA for the last 96 h of differentiation. Thereafter, the cells were incubated with 100 µM [^14^C]acetoacetate for 4 h, and (A) CO_2_ production from acetoacetate and (B) uptake (CO_2_ + CA) of acetoacetate were determined. Data are presented as mean ± SEM of 30–32 biological replicates. Data are normalized to untreated control. Absolute values for control for acetoacetate oxidation 17.0 ± 0.5 nmol/mg protein and acetoacetate uptake 23.5 ± 0.4 nmol/mg protein. *Statistically significant versus control (p < 0.05, unpaired t-test). EMPA: empagliflozin; CA: cell-associated
Measurement of uptake and oxidation of leucine were performed to explore if amino acid metabolism was changed after exposure of human myotubes to EMPA As shown in Fig. 4A, oxidation of leucine was increased, while there was no difference between control and EMPA-treated cells for cellular uptake of leucine (Fig. 4B). Moreover, as a measurement of protein synthesis, leucine accumulation in the cells was determined. However, there was no effect on leucine accumulation after treatment with EMPA (Supplementary Fig. 3).Fig. 4. Effect of empagliflozin on metabolism of leucine in human myotubes. Cultured human myotubes were treated with 500 nM of EMPA for the last 96 h of differentiation. Thereafter, the cells were incubated with 100 µM L-[^14^C]leucine for 4 h, and (A) CO_2_ production from leucine and (B) uptake (CO_2_ + CA) of leucine were determined. Data are presented as mean ± SEM of 8 biological replicates. Data are normalized to untreated control. Absolute values for control for leucine oxidation 3.1 ± 0.6 nmol/mg protein and leucine uptake 32.0 ± 3.3 nmol/mg protein. *Statistically significant versus control (p < 0.05, unpaired t-test). EMPA: empagliflozin; CA: cell-associated
Effects of empagliflozin on signaling via AMPK in human myotubes
To study whether AMP-activated protein kinase (AMPK) could be modified after treatment with EMPA for 96 h, we measured phosphorylation of AMPK and ACC (acetyl-CoA carboxylase) in human myotubes. There were significant increases in phosphorylation of both AMPK (pAMPK/total AMPK) and ACC (pACC/total ACC) for EMPA-treated myotubes (Fig. 5).Fig. 5. Effect of empagliflozin on phosphorylation of AMPK and ACC in human myotubes. Cultured human myotubes were treated with 500 nM of EMPA for the last 96 h of differentiation. (A) Representative immunoblotting images of EMPA effects on pACC, ACC, pAMPK, AMPK and β-actin. Relative phosphorylation levels of AMPK (ratio pAMPK/AMPK) (B) and ACC (ratio pACC/ACC) (C). Values are presented as mean ± SEM from 6 biological replicates. *Statistically significant versus control (p < 0.05, unpaired t-test). C = control; E = empagliflozin
Effect of empagliflozin on mRNA expression of metabolic genes in human myotubes
To further explore the metabolic changes after treatment with EMPA (500 nM for 96 h) on human myotubes, expression levels of selected genes involved in energy metabolism and mitochondrial function; uncoupling protein 3 (UCP3), peroxisome proliferator-activated receptor alpha (PPARA) and pyruvate dehydrogenase kinase isoenzyme 4 (PDK4), as well as some genes involved in fatty acid metabolism; fatty acid translocase (CD36), carnitine palmitoyltransferase 1B and 2 (CPT1B and CPT2) and acetyl-CoA carboxylase beta (ACACB), were quantified (Supplementary Fig. 4). No differences in gene expression levels of these genes were found in EMPA-treated cells compared to control cells.
Effect of empagliflozin on mitochondrial and glycolytic function in mice C2C12 myotubes
To further determine effects of 96 h of EMPA treatment on cellular energy metabolism and mitochondrial function based on findings from studies with labeled exogenous substrates, we performed Mito Stress Test with Seahorse XF24e and XF96e analyzer and C2C12 myotubes. In the absence of fuel inhibitors, basal respiration was reduced after treatment with EMPA (Fig. 6A). We next examined whether changes in mitochondrial function after EMPA treatment were dependent on the nature of available fuel substrates. Combinations of chemical inhibitors were used to force a preferred use of glucose, fatty acids or glutamine. UK5099 was used to block the mitochondrial pyruvate carrier and subsequently entry of pyruvate into the TCA cycle, etomoxir was used to inhibit translocation of long-chain fatty acids into mitochondria and conversion of glutamate into α-ketoglutarate was blocked by BPTES (Görigk et al. 2022). Thus, combination of UK5099 and BPTES resulted in a preferred use of endogenous fatty acids, while combination of etomoxir and BPTES resulted in a preferred use of glucose as energy substrate. Combination of UK5099 and etomoxir resulted in a preferred use of glutamine (Supplementary Fig. 5). Under conditions promoting use of fatty acids both maximal respiration and spare respiratory capacity were increased after treatment with EMPA (Figs. 6B, C). Proton leak was increased under conditions favoring use of glutamine after exposure to EMPA (Fig. 6D). Total ATP production rate was also determined without and with fuel inhibitors, showing an increase after EMPA treatment for FA as substrate (Fig. 6E), and a reduction under no fuel restriction and with glutamine as substrate.Fig. 6. Effects of empagliflozin on mitochondrial function and total ATP production in C2C12 myotubes. C2C12 myotubes were grown in 24- or 96-well Seahorse tissue culture plates and treated with 500 nM of EMPA for the last 96 h of differentiation, before measurements of oxygen consumption rate (OCR) with the Seahorse XF24e or XF96e analyzer as described in the Methods section. (A) Basal respiration with all fuels, (B) maximal respiration, (C) spare respiratory capacity, (D) proton leak calculated from OCR, (E) total ATP production. Data are presented as mean ± SEM of 12–23 biological replicates. Glucose: addition of etomoxir and BPTES to enforce use of glucose, fatty acid: addition of UK5099 and BPTES to enforce use of fatty acids, glutamine: addition of UK5099 and etomoxir to enforce use of glutamine. *Statistically significant versus control (p < 0.05, unpaired t-test). EMPA, empagliflozin; OCR, oxygen consumption rate
During the Seahorse Mito Stress Test, determinants of glycolysis in C2C12 cells were calculated from the ECAR values under conditions of no fuel restriction. Treatment with EMPA for 96 h decreased basal glycolysis, basal proton efflux rate, compensatory glycolysis (glycolysis after inhibition of electron transport system, ETS), and spare glycolytic capacity (Figs. 7A-D).Fig. 7. Effects of empagliflozin on glycolytic function in C2C12 myotubes. C2C12 myotubes were grown in 24- or 96-well Seahorse tissue culture plates and treated with 500 nM of EMPA for the last 96 h of differentiation, before measurement of extracellular acidification rates (ECAR) with the Seahorse XF24e or XF96e analyzer and calculation of proton efflux rate (PER) from ECAR. (A) Basal glycolysis, (B) basal PER, (C) compensatory glycolysis, and (D) spare glycolytic capacity calculated from PER. Data are presented as mean ± SEM of 17–23 biological replicates. *Statistically significant versus control (p < 0.05, unpaired t-test). EMPA: empagliflozin
Effect of empagliflozin on insulin response in human myotubes
To explore effects of EMPA on insulin-dependent signaling in human myotubes, we measured pAkt and total Akt in human myotubes before and after stimulation with insulin (100 nM). There was no difference in basal and insulin-induced phosphorylation (pAkt/total Akt ratio) between the control cells and cells treated with EMPA (500 nM for 96 h) (Fig. 8).Fig. 8. Effect of empagliflozin on insulin-induced phosphorylation of Akt in human myotubes. Cultured human myotubes were treated with 500 nM of EMPA for the last 96 h of differentiation. In some culture wells 100 nM of insulin was added for 15 min. (A) Representative immunoblotting image of insulin and EMPA effects on pAkt and Akt. (B) Relative phosphorylation levels of Akt (ratio pAkt/Akt). Values are presented as mean ± SEM from 6 biological replicates normalised to control without insulin. *Statistically significant versus control; #statistically significant versus EMPA (p < 0.05, 2-way ANOVA). C = control; E, EMPA = empagliflozin; I = insulin
Discussion
In this study, we present data showing that prolonged treatment with the SGLT2 inhibitor empagliflozin (EMPA) can modify energy metabolism in primary skeletal muscle cells. In human skeletal muscle cells, it was found that EMPA enhanced metabolism of exogenously added fatty acids, promoting increased CO_2_ production, as well as amino acid oxidation, whereas glucose oxidation was reduced. The observed effects on fatty acid metabolism could be explained by activation of AMP-activated protein kinase (AMPK) and inhibition of acetyl-CoA carboxylase (ACC) but not by changes in gene expression of certain important metabolic genes. Mechanistic studies with mouse C2C12 skeletal muscle cells showed reduced basal respiration and glycolysis and increased utilization of endogenous fatty acids after EMPA treatment.
We observed an increased CO_2_ production of oleic acid in cells treated with EMPA, while incomplete OA oxidation (ASM) was reduced. The CO_2_/ASM-ratio is a marker of efficient lipid oxidation in that more CO_2_ is produced compared to ASM and suggests enhanced mitochondrial preference for fatty acid catabolism for energy production. Similarly, increased mitochondrial fatty acid and ketone body metabolism in human skeletal muscle biopsies was found after treatment with SGLT2i in vivo (Op den Kamp et al. 2022). In diabetic mice, metabolomic analysis of heart muscle tissue after EMPA treatment pointed at regulation of fatty acid oxidation (Zhang et al. 2022). Moreover, SGLT2i have also been shown to ameliorate mitochondrial dysfunction and dynamics, enhance fatty acid oxidation, change bioenergetics and ion homeostasis, and reduce formation of mitochondrial reactive oxygen species in heart muscle (Dabravolski et al. 2022). A review on SGLT2i focusing on several aspects of lipid metabolism described that SGLT2i may regulate key molecules in lipid accumulation (lipid storage) and transportation (lipoproteins) and thereby affect oxidation of fatty acids. Notably, SGLT2i was described to shift substrate utilization from carbohydrates to lipids and ketone bodies (Szekeres et al. 2021). Thus, substrate competition experiments (metabolic switching in vitro) is a possible approach for further studies with EMPA on energy metabolism in human myotubes (Thoresen et al. 2011).
Mitochondrial function studies in C2C12 cells after treatment with EMPA also showed increased fatty acid oxidation as well as enhanced total ATP production from endogenous fatty acids under conditions promoting use of fatty acids as energy source. Despite different experimental settings to investigate metabolism of exogenous vs. endogenous fatty acids, the increased fatty acid oxidation measured by respirometry (OCR) is in accordance with our data with human myotubes with oleic acid provided exogenously to the cells.
Similarly, we also observed decreased glucose oxidation in human myotubes after exposure to EMPA as well as decreased basal OCR and glycolytic function in C2C12 cells. It has previously been observed that the SGLT2i canagliflozin reduced mitochondrial function (OCR) and glycolytic metabolism of glucose in C2C12 cells (VanDerStad et al. 2024). Also, in other cell types SGLT2i have been shown to affect mitochondrial function (Anastasio et al. 2024; Cinquegrani et al. 2022; Zugner et al. 2022). In human umbilical vein endothelial cells (HUVEC), effects of various SGLT2i on mitochondrial function, glucose uptake and metabolic pathways were examined. Different effects of the inhibitors tested was demonstrated in these cells but some of the compounds decreased glucose metabolism and OCR (Zugner et al. 2022). In HCT 116 colorectal cancer cells SGLT2i decreased basal and maximal respiration (Anastasio et al. 2024), and in human myeloid angiogenic cells basal OCR and glycolysis were reduced by EMPA (Cinquegrani et al. 2022).
There is an increased risk for diabetic ketoacidosis with use of SGLT2i (Douros et al. 2020). The mechanism for this effect is not well described and could be related to decreased metabolism of ketones in skeletal muscle. In our study we observed decreased uptake and oxidation of acetoacetate after EMPA treatment. On the other hand, in vivo experiments with skeletal muscle biopsies, liver samples and heart mitochondria have shown an increased capacity to oxidize ketone bodies after treatment with SGLT2i (Dabravolski et al. 2022; Op den Kamp, et al. 2022; Gao et al. 2024).
Leucine is a branched-chain amino acid (BCAA), and it has been shown in an experimental model of heart failure that EMPA caused a shift in fuel metabolism away from glucose towards BCAA as well as other fuel substrates (fatty acids and ketones) (Santos-Gallego et al. 2019). In accordance with this, we observed increased oxidation of leucine after exposure to EMPA.
We did not observe EMPA-induced changes in fatty acid uptake during 4 h, whereas real time fatty acid accumulation during a longer period revealed increased fatty acid accumulation. In line with this, treatment with dapagliflozin in patients with T2D increased skeletal muscle fatty acid uptake (Latva-Rasku et al. 2024) and intramyocellular lipid content (Op den Kamp et al. 2022). Studies in diet-induced obese (DIO) mice, however, have observed decreased lipid accumulation in accordance with an increased mitochondrial biogenesis and protected mitochondrial function (Gao et al. 2024). In the study with C2C12 cells and canagliflozin treatment it was also observed decreased intramyocellular lipid content (VanDerStad et al. 2024). Thus, there might be species differences with regards to lipid accumulation after SGLT2i exposure.
It has been shown that SGLT2i can affect the AMPK pathway under various physiological and pathophysiological conditions (Safaie et al. 2024), as well as improve impaired oxidative muscle function in diabetic mice by AMPK activation (Nakamura et al. 2023). Moreover, SGLT2i induced metabolic reprogramming in mice by stimulation of AMPK (Kogot-Levin et al. 2023). Thus, AMPK might be a possible regulator for changes in energy metabolism in presence of SGLT2i in skeletal muscle cells. An increased AMPK phosphorylation was observed in EMPA-treated myotubes. In addition, ACC, a target for AMPK and a major controller of the rate of fatty acid oxidation within cells also showed an increased phosphorylation.
It has been observed that SGLT2i may counteract insulin resistance in skeletal muscle. In some studies, SGLT2i has been shown to increase insulin sensitivity in skeletal muscle (Goto et al. 2020; Merovci et al. 2014; Obata et al. 2016). Also in a study with dapagliflozin in diet-induced obese mice and C2C12 myotubes, it was observed that insulin sensitivity measured with various glucose disposal-related factors, including Akt, was enhanced in obese mice through SIRT1 activation (Gao et al. 2024). In this study we did not observe that the insulin response on pAkt was enhanced in human myotubes after exposure to EMPA. In a study with T2D subjects and dapagliflozin peripheral insulin sensitivity was not changed compared to control but tended to correlate with increased intramyocellular lipid accumulation (Op den Kamp et al. 2022; Op Kamp et al. 2021). To look further into the effect of EMPA on insulin action in human myotubes more studies with lower insulin concentrations must be performed.
Expression of skeletal muscle differentiation markers such as the myosin-heavy chains 2 (MYH2) and 7 (MYH7), and maturation factors related to metabolic muscle function like the insulin-regulated facilitative glucose transporter, solute carrier family 26 member 4 (SLC26A4) was not changed after treatment with EMPA. In addition, leucine accumulation was not affected. Together, these data suggest that EMPA did not modulate muscle cell differentiation and protein synthesis. With respect to amino acid metabolism and protein synthesis there is an ongoing discussion regarding the impact of SGLT2i on muscle mass and function. Some human studies have shown detrimental effects and others have shown neutral or beneficial effects of SGLT2i on skeletal muscle (Afsar and Afsar 2023). In subjects with T2D, SGLT2i treatment for 52 weeks was associated with weight loss without loss of skeletal muscle mass (Volpe et al. 2024). Although leucine uptake by human myotubes was not changed, we found that oxidation was increased by EMPA.
Expression of genes related to neither glucose nor fatty acid metabolism was altered after treatment with EMPA. Moreover, in C2C12 muscle cells no changes in gene expression of selected genes involved in glucose and lipid metabolism after canagliflozin treatment was observed (VanDerStad et al. 2024). Although not directly comparable, in vivo experiments on skeletal muscle from diet-induced obese mice and subjects with T2D showed that dapagliflozin enhanced expression of genes important for mitochondrial biogenesis and function, and fatty acid catabolism (Op den Kamp et al. 2022; Gao et al. 2024).
Conclusion
Direct exposure of EMPA to human skeletal muscle cells caused changes in energy metabolism like enhanced fatty acid and leucine catabolism, decreased oxidative metabolism of glucose and acetoacetate, and reduced glycolysis. Altered energy metabolism were not reflected by changes in gene expression but rather by activation of AMPK, suggesting changes in substrate utilization may be responsible for the observed metabolic effects. The long-term effects of skeletal muscle adaptions induced by SGLT2i in cardio-metabolic diseases needs to be further studied.
Supplementary Information
Below is the link to the electronic supplementary material.ESM1(DOCX 537 KB)