Modulation of muscle anabolism and gut microbiota by Polyporus umbellatus extract attenuates dexamethasone-induced muscle atrophy

Abstract

To investigate the pharmacological mechanisms by extract of Polyporus umbellatus (PU) protects against dexamethasone (DEX)-induced muscle atrophy, focusing on its direct effects on muscle cell signaling, mitochondrial function, oxidative stress, and its indirect influence via gut microbiota modulation. In vitro, DEX-treated C2C12 myotubes were used to assess PU’s effects on cell viability, myotube morphology, myogenic/atrophy gene expression, Akt/mTOR/FoxO3a signaling pathways, mitochondrial function, and oxidative stress. In vivo, a DEX-induced muscle atrophy mouse model was employed to evaluate the efficacy of orally administered PU and L. gasseri (ATCC 19992) alone on muscle mass, strength, exercise performance, and gene expression. Gut microbiota composition was analyzed via 16 S rRNA sequencing, with predicted microbial enzyme functions and correlations to muscle parameters examined. In vitro, PU significantly attenuated DEX-induced C2C12 myotube atrophy, activated Akt/mTOR signaling, inhibited FoxO3a signaling, mitigated oxidative stress, and enhanced mitochondrial function. In vivo, PU dose-dependently improved grip strength, muscle mass, and exercise performance in DEX-treated mice, concurrently upregulating myogenic and mitochondrial biogenesis genes. PU treatment significantly modulated gut microbial diversity and composition, notably increasing L. gasseri abundance. Oral administration L. gasseri recapitulated PU’s protective effects on muscle phenotype, gene expression, and gut microbiota modulation. L. gasseri levels and predicted microbial D-lactate dehydrogenase activity correlated positively with muscle health. However, bioactivity-guided fractionation of PU did not identify a single predominant active compound. In conclusion, PU protects against glucocorticoid-induced muscle atrophy through a dual mechanism involving direct muscle-protective actions and beneficial modulation of the gut microbiota, partly mediated by enrichment and direct effects of L. gasseri.

Introduction

Skeletal muscle, constituting approximately 40% of total body mass, is not merely a structural entity but also a critical metabolic organ essential for locomotion, protein storage, and overall energy homeostasis [1]. Its integrity is vital, yet various conditions, including aging, chronic diseases, and notably, prolonged glucocorticoid therapy, can disrupt muscle homeostasis, leading to atrophy—a debilitating loss of muscle mass and strength [2, 3]. Glucocorticoid-induced muscle atrophy, frequently encountered with drugs like dexamethasone (DEX), presents a significant clinical challenge due to the limited availability of effective and safe therapeutic interventions [4, 5]. While DEX effectively models rapid muscle wasting by suppressing anabolic pathways like IGF-1/Akt/mTOR and activating catabolic systems such as the ubiquitin-proteasome pathway [4, 5], there remains an urgent need for novel protective strategies.

Polyporus umbellatus (Pers.) Fries is a medicinal mushroom that has been used in traditional practices for various health-promoting purposes. Its reported activities, such as immunomodulatory and antioxidant effects [6, 7, 8, 9, 10, 11, 12–13], suggest potential for broader pharmacological applications. However, its specific role and mechanistic action in counteracting muscle atrophy, particularly glucocorticoid-induced wasting, remain largely underexplored.

Emerging research highlights a crucial bidirectional communication pathway between the gut and skeletal muscle, termed the “gut-muscle axis” [14, 15]. The gut microbiota, a complex ecosystem within the host, influences muscle physiology through various mechanisms, including the production of metabolites like short-chain fatty acids (SCFAs) and the modulation of systemic inflammation [16]. Dysbiosis, an imbalance in this microbial community, is increasingly associated with muscle wasting conditions [17], suggesting that interventions targeting the gut microbiome could offer novel therapeutic avenues for muscle atrophy. Given the known polysaccharide content of Polyporus umbellatus, which can act as prebiotics [6, 7–8], and its reported general health-promoting effects, we hypothesized that an extract might exert muscle-protective effects, in part, by modulating the gut microbiota.

This study aimed to pharmacologically investigate the potential of an aqueous extract of Polyporus umbellatus (PU) to mitigate DEX-induced muscle atrophy. We hypothesized that PU acts through a multifaceted, dual mechanism involving: (1) direct effects on muscle cells, including the modulation of key anabolic (e.g., Akt/mTOR) and catabolic (e.g., FoxO3a) signaling pathways, enhancement of mitochondrial function, and reduction of oxidative stress; and (2) indirect effects mediated by the gut microbiome, potentially through the enrichment of beneficial bacteria such as Lactobacillus gasseri. Using in vitro C2C12 myotube models and an in vivo DEX-induced muscle atrophy mouse model, including experiments with direct L. gasseri administration, we sought to elucidate these mechanisms. Our findings provide novel insights into the pharmacological actions of this mushroom extract, highlighting its potential as a natural product-derived therapeutic agent for muscle wasting conditions.

Materials and methods

Plant material and Polyporus umbellatus extract (PU) Preparation and characterization

Dried fruiting bodies of Polyporus umbellatus (Pers.) Fries were purchased from a commercial vendor (Samhong, Gyeonggi-do, Korea) and authenticated by Prof. Dae-Sik Jang (Kyung Hee University, Seoul, Korea). A voucher specimen (accession number KIST-2023-PU) was deposited at the Korea Institute of Science and Technology. The dried fruiting bodies were ground into a coarse powder using a mechanical grinder. For aqueous extraction, the powder was extracted with distilled water at room temperature. The combined aqueous extracts were filtered through filter paper and the combined filtrate was concentrated. The resulting P. umbellatus extract (PU) was characterized by HPLC-DAD fingerprinting and stored at -20 °C until further use.

Isolation and identification of PU-derived compounds

PU (100 g) was suspended in distilled water (1 L) and sequentially partitioned three times with equal volumes of ethyl acetate (EtOAc; Samchun Pure Chemical Co., Ltd., Seoul, Republic of Korea) and n-butanol (n-BuOH; Samchun Pure Chemical Co., Ltd.). The solvents were evaporated under reduced pressure to yield the EtOAc fraction (PUEA; 1 g) and the n-BuOH fraction (PUBu; 18 g).

The PUBu fraction was subjected to silica gel column chromatography (63–200 μm particle size, Kieselgel 60, Cat# 107734, Merck, Darmstadt, Germany; column dimensions 10 × 60 cm) using a stepwise gradient elution of dichloromethane (DCM; Samchun Pure Chemical Co., Ltd.) and methanol (MeOH; Fisher Scientific, Pittsburgh, PA, USA) (100:0 to 0:100, v/v), yielding 35 primary fractions (PUBu-1 to PUBu-35) based on TLC analysis (Silica gel 60 F254 plates, Cat# 105554, Merck; visualized with 20% H2SO4 spray and heating at 110 °C for 5 min).

Further purification of selected primary fractions using repeated column chromatography (Silica gel 60, Sephadex LH-20 [GE Healthcare, Chicago, IL, USA]) and preparative medium-pressure liquid chromatography (MPLC; CombiFlash Rf+, Teledyne ISCO, Lincoln, NE, USA; C18 column) led to the isolation of eight known compounds: ergosterol (1; 5.3 mg) from PUBu-2 [22], N-(2′-hydroxytetracosanoyl)-1,3,4-trihydroxy-2-octadecanone (2; 5.6 mg) from PUBu-12 [13], uracil (3; 3.1 mg) from PUBu-32 [23], 4-hydroxybenzoic acid (4; 2.6 mg) from PUBu-8 [24], polyporusterone B (5; 3.9 mg) from PUBu-13 [25, 26], polyporusterone A (6; 2.3 mg) from PUBu-13 [25, 26], 3,4-dihydroxybenzaldehyde (7; 1.3 mg) from PUBu-6 [27], and protocatechuic acid (8; 3.2 mg) from PUBu-8 [28] (Structures in Supporting Information Fig. S1).

Compound identification was confirmed by comparing their spectroscopic data ¹H and ¹³C NMR, HR-ESI-MS) with published literature values. ¹H (400 MHz) and ¹³C (100 MHz) NMR spectra were recorded on a Bruker Avance DRX-400 spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) in appropriate deuterated solvents (e.g., CDCl3, MeOD; Cambridge Isotope Laboratories, Inc., Tewksbury, MA, USA) with chemical shifts reported in δ (ppm) relative to the residual solvent peaks. High-resolution electrospray ionization mass spectrometry (HR-ESI-MS) data were obtained in positive and/or negative ion mode using a Waters Synapt G2-Si mass spectrometer (Waters Corporation, Milford, MA, USA). Detailed spectroscopic data for each isolated compound (1–8) are provided in Supporting Information (Figures S4-S11).

High-performance liquid chromatography (HPLC-DAD) fingerprinting of PU extract

HPLC analysis for quality control and fingerprinting of the PU extract was performed using an Agilent 1260 Infinity II LC System (Agilent Technologies, Santa Clara, CA, USA) equipped with a G7111A quaternary pump, G7129A autosampler, G7116A column thermostat, and G7117A diode array detector (DAD). Separation was achieved on a YMC Triart C18 column (4.6 × 250 mm, 3 μm particle size; Cat# TA12S03-2546WT, YMC Co., Ltd., Kyoto, Japan) maintained at 30 °C. The mobile phase consisted of 0.1% (v/v) formic acid (Sigma-Aldrich) in HPLC-grade water (A) and 0.1% (v/v) formic acid in acetonitrile (HPLC grade, Fisher Scientific) (B). The gradient elution profile was: 0–5 min, 10% B; 5–40 min, linear gradient from 10 to 90% B; 40–45 min, 90% B; 45–46 min, linear gradient from 90 to 10% B; 46–50 min, 10% B. The flow rate was 1.0 mL/min, and the injection volume was 10 µL. Detection was performed by DAD scanning from 200 to 400 nm, with chromatograms monitored at 254 nm. Samples were prepared by dissolving 10 mg of PU extract in 1 mL HPLC-grade MeOH (Fisher Scientific), filtering through a 0.45 μm PTFE syringe filter (Cat# 6784 − 1304, Whatman), and injecting aliquots for analysis. The presence of compounds 3–8 in the extract was confirmed by comparing retention times and UV spectra with those of the isolated standards (Supporting Information Figure S1A). Compounds 1 (ergosterol) and 2 were not clearly identifiable in the HPLC chromatogram of the crude extract under these specific conditions (254 nm detection), likely due to low abundance or suboptimal UV absorbance at this wavelength.

C2C12 cell culture and differentiation

Murine C2C12 myoblasts (CRL-1772, passages 5–10; ATCC, Manassas, VA, USA) were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Cat# SH30243.01, Cytiva, Marlborough, MA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Cat# 16000044, Thermo Fisher Scientific, Waltham, MA, USA) and 1% penicillin/streptomycin (P/S; Cat# SV30010, Cytiva) at 37 °C in a humidified atmosphere containing 5% CO₂ (CO2 Incubator, Thermo Fisher Scientific). Upon reaching 70–80% confluence, differentiation into myotubes was induced by switching to differentiation medium (DMEM containing 2% horse serum [Cat# 16050122, Gibco, Thermo Fisher Scientific] and 1% P/S). The differentiation medium was replaced every 48 h for 6 days.

In vitro treatments with PU

For in vitro experiments, differentiated C2C12 myotubes were treated with varying concentrations of PU (10, 20, and 40 µg/mL; dissolved in sterile distilled water, filtered through a 0.22 μm filter, and freshly prepared) or vehicle (sterile distilled water). Dexamethasone (DEX; Cat# D4902, Sigma-Aldrich, St. Louis, MO, USA), dissolved in DMSO (Cat# D2650, Sigma-Aldrich), was added to the medium 2 h after PU pre-treatment, at a final concentration of 100 µM for viability, ROS, GSH, mitochondrial activity, and ATP assays, or 50 µM for Western blot experiments, for a total incubation period of 24 h with DEX. The final concentration of DMSO in the medium was maintained below 0.1% (v/v). Control cells received vehicle treatments equivalent to the highest concentrations of PU vehicle and DMSO used.

Cytotoxicity assay

Cell viability was assessed using the EZ-Cytox Cell Viability Assay Kit (Cat# EZ-1000, DoGenBio, Seoul, Republic of Korea). C2C12 myotubes (passage < 10) were seeded in 96-well plates (1 × 10⁴ cells/well) and differentiated as described above. Cells were treated with PU (10, 20, and 40 µg/mL) and/or DEX (100 µM) for 24 h. Following treatment, EZ-Cytox reagent was added according to the manufacturer’s instructions, and plates were incubated for 2 h at 37 °C. Absorbance at 450 nm was measured using a microplate reader (Multiskan SkyHigh, Thermo Fisher Scientific, Roskilde, Denmark).

Intracellular glutathione (GSH) level measurement

Reduced glutathione (GSH) levels were measured using the GSH/GSSG-Glo™ Assay (Cat# V6611, Promega, Madison, WI, USA). Differentiated C2C12 myotubes were treated with PU (10, 20, and 40 µg/mL) and/or DEX (100 µM) for 24 h. Cell lysis and GSH quantification were performed according to the manufacturer’s protocol. Luminescence was measured using a GloMax^® Navigator Microplate Luminometer (Promega).

Mitochondrial activity measurement

Mitochondrial mass/content was assessed using MitoTracker™ Green FM (Cat# M7514, Invitrogen, Thermo Fisher Scientific). Differentiated C2C12 myotubes (passage < 10) in 96-well plates (1 × 10⁴/well) were treated with PU (10, 20, and 40 µg/mL) and/or DEX (100 µM) for 24 h. Cells were then incubated with 100 nM MitoTracker Green FM in serum-free DMEM for 30 min at 37 °C. After washing with PBS, fluorescence intensity was measured using a microplate reader (Multiskan SkyHigh, Thermo Fisher Scientific) at excitation/emission wavelengths of 490/516 nm.

Cellular ATP quantification

Intracellular ATP levels were determined using the CellTiter-Glo^® 2.0 Assay kit (Cat# G9241, Promega). Differentiated C2C12 myotubes (passage < 10) were treated with PU (10, 20, and 40 µg/mL) and/or DEX (100 µM) for 24 h. Following treatment, the assay was performed directly in the wells according to the manufacturer’s instructions. Luminescence was measured using a GloMax^® Navigator Microplate Luminometer (Promega).

Immunofluorescence staining and myotube diameter analysis

Differentiated C2C12 myotubes (passage < 10) were fixed with 4% formaldehyde in phosphate-buffered saline (PBS) for 15 min at room temperature and permeabilized with 0.25% Triton X-100 in PBS for 10 min. After blocking with 5% bovine serum albumin (BSA) in PBS for 1 h, cells were incubated overnight at 4 °C with a primary antibody against myosin heavy chain (MHC; MF20, Developmental Studies Hybridoma Bank, Iowa City, IA, USA) diluted 1:100 in 1% BSA in PBS. Cells were then washed with PBS and incubated with an Alexa Fluor 488-conjugated secondary antibody (1:100, Invitrogen, Thermo Fisher Scientific) and Hoechst 33,342 (1:1000, Invitrogen) for nuclear staining for 1 h at room temperature. Images were captured using an Operetta High Content Imaging System (PerkinElmer Inc., Waltham, MA, USA) and myotube diameters were quantifiedusing Operetta analysis software.

Intracellular ROS measurement

Intracellular reactive oxygen species (ROS) levels were measured using the 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA) assay (Cat# D6883, Sigma-Aldrich). Differentiated C2C12 myotubes (passage < 10) were treated with PU (10, 20, and 40 µg/mL) and/or DEX (100 µM) for 24 h. Cells were then washed with PBS and incubated with 20 µM DCFH-DA in serum-free DMEM for 30 min at 37 °C. After washing again with PBS, fluorescence intensity was measured immediately using a microplate reader (Multiskan SkyHigh Microplate Spectrophotometer) at excitation/emission wavelengths of 485/530 nm.

RNA extraction and quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Total RNA was isolated from C2C12 myotubes and flash-frozen tibialis muscle tissue using the HybridR RNA isolation kit (GeneAll, Seoul, Korea) according to the manufacturer’s instructions. RNA quality and quantity were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific). cDNA was synthesized from 1 µg of total RNA using the Reverse Transcription Premix (ELPIS-Biotech, Daejeon, Korea). Quantitative real-time PCR (qRT-PCR) was performed on a LightCycler 480 Real-Time PCR System (Roche, Basel, Switzerland) using the PowerUpTM SYBRTM Green Master Mix (Thermo Fisher Scientific). Reaction conditions were: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min, followed by a melt curve analysis. Relative gene expression was calculated by the 2-ΔΔCt method with β-actin as the housekeeping gene for normalization. All primer sequences (synthesized by Bioneer, Daejeon, Republic of Korea) are listed in Table 1.

Table 1. qRT-PCR mouse primer sequences (5’-3’)

Species	Target gene	Direction	Primer sequence (5’- 3’)	Gene ID
Mouse	Fbxo32	Forward	AACCCTTGGGCTTTGGGTTT	NM_026346.3
		Reverse	GGACTTAAGCCCGTGCAGAT
	Myod1	Forward	CATAGACTTGACAGGCCCCG	NM_010866.2
		Reverse	CGGGTCCAGGTCCTCAAAAA
	Myf6	Forward	ACAGATCGTCGGAAAGCAGC	NM_008657.3
		Reverse	CACTCCGCAGAATCTCCACC
	Myog	Forward	AGCTATCCGGTTCCAAAGCC	NM_031189.2
		Reverse	GCACAGGAGACCTTGGTCAG
	Myh2	Forward	AGCGAAGAGTAAGGCTGTCC	NM_001039545.2
		Reverse	AGGCGCATGACCAAAGGTT
	Ppargc1a	Forward	GTTGCCTGCATGAGTGTGTG	NM_008904.3
		Reverse	CACATGTCCCAAGCCATCCA
	Ucp3	Forward	GTTTTGCGGACCTCCTCACT	NM_009464.3
		Reverse	CTCTGTGCGCACCATAGTCA
	Tomm20	Forward	TGTGCGGTGTGTTGTCTGTT	NM_024214.2
		Reverse	TAAGTGCCCAGAGCACAGGA
	Nrf1	Forward	CCCGTGTTCCTTTGTGGTGA	NM_001410231.1
		Reverse	ATTCCATGCTCTGCTGCTGG
	Tfam	Forward	GGGAATGTGGAGCGTGCTAA	NM_009360.4
		Reverse	TGATAGACGAGGGGATGCGA
	Gpx1	Forward	AGTCCACCGTGTATGCCTTC	NM_001329527.1
		Reverse	CCTCAGAGAGACGCGACATT
	Sod1	Forward	GGGAAGCATGGCGATGAAAG	NM_011434.2
		Reverse	GCCTTCTGCTCGAAGTGGAT
	Actb	Forward	CATTGCTGACAGGATGCAGAAGG	NM_007393.5
		Reverse	TGCTGGAAGGTGGACAGTGAGG

Western blot analysis

C2C12 myotubes, seeded in 6-well plates (1.6 × 10⁵ cells/well), were lysed on ice for 30 min in ice-cold RIPA buffer (Cat# 89900, ThermoFisher Scientific) supplemented with phenylmethylsulfonyl fluoride and sodium orthovanadate as protease and phosphatase inhibitors, respectively. Following centrifugation, protein concentrations were determined using the Pierce™ BCA Protein Assay Kit (Cat# 23225, Thermo Fisher Scientific). Equal amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 10% gels and transferred to polyvinylidene difluoride membranes (Bio-Rad, Hercules, CA, USA). Membranes were blocked with 5% BSA (GenDEPOT, Katy, TX, USA) in Tris-buffered saline with 0.1% Tween 20 for 1 h at room temperature to prevent non-specific binding. Subsequently, the membranes were incubated overnight at 4 °C with primary antibodies diluted in blocking buffer (a detailed list of dilutions and sources is provided in Table 2). After washing with TBST, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse secondary antibodies (1:5000 or 1:1000 dilution, respectively; Santa Cruz Biotechnology, Dallas, TX, USA) for 2 h at room temperature. Following additional washes with TBST, protein bands were visualized using an enhanced chemiluminescence reagent (ThermoFisher Scientific) and imaged on a LAS-4000 (Fujifilm, Minato, Japan). Densitometric quantification was performed using ImageJ software (v1.53, NIH, Bethesda, MD, USA). Target prottein expression was normalized to the GAPDH expression detected on the same membrane [29] (Table 2).

Table 2. List of Antibodies

Antibody	Brand name	Dilution
Phospho-p70 S6 Kinase (Thr389)	CST #9205	1:1000
p70 S6 Kinase	CST #2708	1:1000
Phospho-4E-BP1 (Thr37/46)	CST #2855	1:1000
4E-BP1	CST #9644	1:1000
Phospho-Akt (Ser473)	CST #9271	1:1000
Akt	CST #9272	1:1000
GAPDH	CST #2118	1:1000
Phospho-FoxO3a (Ser253)	CST #9466	1:1000
FoxO3a	CST #12829	1:1000
MurF1	sc-398608	1:1000
Atrogin-1	sc-166806	1:1000

Lactobacillus gasseri culture and growth rate assessment

Lactobacillus gasseri ATCC 19,992, ATCC 9857, and ATCC 4963 were obtained from the American Type Culture Collection (Manassas, VA, USA). Strains were cultured in De Man–Rogosa–Sharpe (MRS) broth (Cat# MB-D1043, KisanBio Co., Ltd., Seoul, Republic of Korea) at 37 °C under anaerobic conditions in a Vinyl Anaerobic Chamber (Coy Laboratory Products, Grass Lake, MI, USA).

To assess the effect of PU on bacterial growth, overnight cultures of each L. gasseri strain were diluted 1:1000 in fresh MRS broth, with or without sterile-filtered PU extract (final concentrations: 10, 50, 100, 200 µg/mL). Cultures (200 µL/well) were grown in triplicate in 96-well microplates (Corning Inc., Corning, NY, USA) at 37 °C under anaerobic conditions with continuous shaking (200 rpm) in a microplate reader equipped with an incubator (Synergy H1, BioTek Instruments, Winooski, VT, USA). Optical density (OD) at 600 nm was measured every 30 min for 24 h. Maximum specific growth rates were calculated from the steepest slope of the ln(OD₆₀₀) versus time plot during the exponential growth phase (typically between OD₆₀₀ values of 0.05 and 0.3). L. gasseri ATCC 19,992, which exhibited the most significant growth enhancement with PU, was selected for in vivo studies. For oral gavage, L. gasseri ATCC 19,992 was grown overnight in MRS broth, harvested by centrifugation (5000 x g, 10 min, 4 °C), washed twice with anaerobic PBS (pH 7.4), and resuspended in an anaerobic 0.5% CMC solution to a final concentration of 1 × 10¹⁰ CFU/mL immediately before administration.

Animal studies

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of the Korea Institute of Science and Technology (Approval No.: KIST-2021-091002) and conducted in strict accordance with the ARRIVE guidelines and the NIH Guide for the Care and Use of Laboratory Animals. The sample size for in vivo experiments (n = 8–10 mice per group) was determined based on power calculations (G*Power 3.1 software) using data from our preliminary studies and previously published literature on DEX-induced muscle atrophy models [18, 19–20], aiming for a power of > 0.8 and an alpha of 0.05 to detect a biologically significant difference of approximately 20–25% in primary outcome measures such as grip strength.

Male C57BL/6J mice (8 weeks old, weighing 20–25 g) were obtained from Orient Bio Inc. (Seongnam, Republic of Korea) and housed under specific pathogen-free (SPF) conditions (5 mice per cage) with a 12-hour light/dark cycle (lights on at 7:00 AM), controlled temperature (22 ± 2 °C), and humidity (50 ± 10%). Mice were acclimatized for one week with ad libitum access to standard chow diet (AIN-76 A, Cat# TD170481, Envigo, Indianapolis, IN, USA) and autoclaved water.

PU experiment

Mice (n = 60) were randomly assigned to six groups (n = 10 per group) using a computer-based random number generator: (1) Control (Vehicle: 0.5% CMC orally + saline i.p.), (2) DEX (Vehicle orally + DEX 25 mg/kg/day i.p.), (3) DEX + Oxymetholone (Oxy; 50 mg/kg/day orally + DEX i.p.), (4) DEX + PU (10 mg/kg/day orally + DEX i.p.), (5) DEX + PU (50 mg/kg/day orally + DEX i.p.), and (6) DEX + PU (100 mg/kg/day orally + DEX i.p.). PU and oxymetholone (Oxy; Cat# A8393, Sigma-Aldrich; positive control) were suspended/dissolved in 0.5% (w/v) carboxymethyl cellulose sodium salt (CMC; Cat# C4888, Sigma-Aldrich) in sterile water and administered daily via oral gavage (10 mL/kg volume) for four consecutive weeks. During the final two weeks of the study (days 15–28), mice in groups 2–6 received daily intraperitoneal (i.p.) injections of DEX (25 mg/kg/day; dissolved in sterile 0.9% saline) at approximately 9:00 AM.

L. gasseri experiment

A separate cohort of male C57BL/6J mice (8 weeks old, n = 30) was used. Mice were randomly assigned to three groups (n = 10 per group): (1) Control (Vehicle: 0.5% CMC orally + saline i.p.), (2) DEX (Vehicle orally + DEX 25 mg/kg/day i.p.), and (3) DEX + L. gasseri (LG) (L. gasseri ATCC 19992, 1 × 10⁹ CFU/mouse/day in 0.5% CMC orally + DEX i.p.). Treatments were administered for four weeks, with DEX injections during the final two weeks, as described for the PU experiment.

Outcome assessments and sample collection

Body weight was recorded weekly. Grip strength and exercise performance tests were conducted 24–48 h before euthanasia. At the end of the four-week treatment period (day 29), following a 12-hour fast (water available), mice were euthanized by intraperitoneal injection of zoletil and rumpun followed by cervical dislocation. Blood was collected via cardiac puncture for serum preparation. The gastrocnemius, soleus, tibialis anterior, and quadriceps muscles were rapidly dissected, weighed, with portions snap-frozen in liquid nitrogen and stored at -80 °C for molecular analyses, and other portions fixed for histology. Cecal contents were collected and immediately snap-frozen for microbiome analysis.

Grip strength and exercise performance test

Grip strength was assessed using a grip force meter (Bioseb, Pinellas Park, FL, USA) following a standardized protocol. The mice were positioned on a grid and carefully encouraged to grasp. After gentle caudal traction, the maximum force recorded before the mice lost their grip across three trials was used. Blinding of the investigators to the treatment groups ensured objective measurement. Exercise performance was evaluated on a Touchscreen Treadmill (Panlab, Harvard Apparatus, Holliston, MA, USA). After acclimation to the apparatus and the use of a mild shock to discourage resting, the mice gradually ran at increasing speeds (starting at 10 m/min) until exhaustion (defined as hanging on the shock bar for over 10 s). The final distance, time, and speed were recorded.

Skeletal muscle histology

Gastrocnemius muscles were fixed in 10% neutral buffered formalin (Sigma-Aldrich) for 24 h, processed through graded ethanol and xylene, and embedded in paraffin wax. Cross-Sect. (4 μm thick) were cut, mounted on slides, and stained with hematoxylin and eosin (H&E; Sigma-Aldrich). Digital images of entire muscle cross-sections were captured using a slide scanner (Axio Scan.Z1, Zeiss, Oberkochen, Germany) at 200x objective magnification. Muscle fiber cross-sectional area (CSA) was quantified from at least three non-overlapping, randomly selected fields per animal (representing > 500 fibers per animal) using ImageJ software (v1.53, NIH).

Gut Microbiome analysis by 16 S rRNA gene sequencing

Cecal contents were collected at the time of sacrifice and immediately frozen at -80 °C. Genomic DNA was extracted from the cecal samples using the QIAamp^® Fast DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The V3-V4 hypervariable region of the bacterial 16 S rRNA gene was amplified using barcoded primers (forward: 5’-CCTACGGGNGGCWGCAG-3’; reverse: 5’-GACTACHVGGGTATCTAATCC-3’). PCR products were purified using AMPure XP beads (Beckman Coulter, Brea, CA, USA) and quantified using a Qubit 4 Fluorometer (Invitrogen). Raw sequencing data were acquired on a MiSeq platform (Illumina) with paired-end 2 × 300 bp reads. An average sequencing depth of 26,665 ± 12,280 reads per sample (mean ± SD) was achieved after quality filtering.

Raw sequencing data were analyzed using the QIIME 2 pipeline (version 2022.8) [21]. Demultiplexed paired-end reads were imported, and primers were removed using cutadapt via q2-cutadapt. Reads were then quality filtered, denoised, merged, and chimeric sequences were removed using the DADA2 algorithm (via q2-dada2) [22]. Based on quality plots, forward reads were truncated at 290 bp and reverse reads at 214 bp; reads with more than 2 expected errors (maxEE) for either the forward or reverse reads were discarded, and reads shorter than 150 bp after truncation were removed.

Taxonomy was assigned to the resulting amplicon sequence variants (ASVs) using a pre-trained Naive Bayes classifier (q2-feature-classifier classify-sklearn) based on the SILVA 138 SSU Ref NR 99 (99% identity OTUs) reference database [23], trimmed to the V3-V4 region amplified by the 341 F/805R primers. ASVs assigned as mitochondria or chloroplast were filtered out. For diversity analyses, the ASV table was rarefied to 5,000 sequences per sample using q2-feature-table rarefy to normalize for differences in sequencing depth (samples with fewer sequences were excluded from diversity analyses).

Alpha diversity metrics (Shannon index, Observed ASVs, and Faith’s Phylogenetic Diversity [PD]) and beta diversity metrics (Bray-Curtis dissimilarity and unweighted UniFrac distance) were calculated using q2-diversity [24]. Principal coordinate analysis (PCoA) was used to visualize beta diversity. Differential abundance analysis of taxa between groups was determined using ANCOM2.1 [25]. Functional profiles of the gut microbiome (predicted enzyme abundances based on EC numbers) were inferred using PICRUSt2 (v2.5.1) [26] with default settings, using the ASV table prior to rarefaction.

Statistical analysis

Data analysis was conducted using R software (v.4.1.2) [R Core Team, 2021]. All values are presented as mean ± standard error of the mean (SEM). Differences between groups were analyzed by one-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) post-hoc test for pairwise comparisons. Differences in gut microbiota composition were assessed using permutational multivariate analysis of variance (PERMANOVA) with 999 permutations on Bray-Curtis and unweighted UniFrac distances. Correlations between muscle parameters and microbial taxa were assessed using Spearman’s rank correlation. P-values for correlation analyses were adjusted for multiple comparisons using the Benjamini-Hochberg False Discovery Rate (FDR) procedure. A p-value (or adjusted p-value/FDR q-value) < 0.05 was considered statistically significant.

Results

PU attenuates DEX-Induced atrophy and modulates myogenic regulatory factors and key signaling pathways in C2C12 myotubes

To investigate the direct cellular effects of PU on glucocorticoid-induced muscle atrophy, C2C12 myotubes were treated with DEX. As anticipated, DEX treatment (100 µM for 24 h, unless otherwise stated) significantly reduced C2C12 myotube viability. Co-treatment with PU (10, 20, and 40 µg/mL) dose-dependently reversed this DEX-induced cytotoxicity. Interestingly, we observed a slight, though not statistically significant, decrease in cell viability in the myotubes treated with 40 µg/mL of PU alone (Con + PU40 group) compared to the control group. This may suggest that at high concentrations, in the absence of a potent stressor like dexamethasone, the complex mixture of compounds within the extract could exert mild cytostatic or metabolic-stress effects. (Fig. 1A). Immunofluorescence staining for myosin heavy chain (MHC) revealed that DEX markedly impaired myotube formation and reduced myotube diameter, indicative of atrophy (Fig. 1B). Importantly, PU treatment significantly ameliorated these DEX-induced atrophic changes, evidenced by increased myotube diameter and the presence of larger, multinucleated myotubes, particularly at the 40 µg/mL concentration (Fig. 1B, C).

[See PDF for image]

Fig. 1

Polyporus umbellatus (Pers.) water extract (PU) protects C2C12 myotubes against dexamethasone-induced atrophy and modulates key anabolic and catabolic signaling pathways. (A) Cell viability of C2C12 myotubes treated with DEX (100 µM) and/or different concentrations of PU (10, 20, or 40 µg/mL) for 24 h, assessed by EZ-Cytox assay. (B, C) epresentative immunofluorescence images of C2C12 myotubes stained for myosin heavy chain (MHC, green) and nuclei (Hoechst, blue). (D-I) mRNA expression levels of atrophy-related genes (Fbxo32, Myh2) and myogenic regulatory factors (Myog, Myod1, Myf5, Myf6) in C2C12 myotubes treated with DEX and/or PU, determined by qRT-PCR. β-actin was used as an internal control. (J) Representative Western blots and quantification of phosphorylated Akt (p-Akt), total Akt, phosphorylated p70-S6K1 (p-p70-S6K1), total p70-S6K1, phosphorylated 4E-BP1 (p-4E-BP1), and total 4E-BP1 in C2C12 myotubes treated with DEX (50 µM) and/or PU. (K) Representative Western blots and quantification of MurF1, phosphorylated FoxO3a (p-FoxO3a Ser253), total FoxO3a, and Atrogin-1 in C2C12 myotubes treated with DEX and/or PU. GAPDH was used as a loading control for Western blot analyses. All data are presented as mean ± SEM from at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test

Mechanistically, PU counteracted the catabolic effects of DEX. DEX treatment significantly upregulated the mRNA expression of Fbxo32 (Atrogin-1), a key E3 ubiquitin ligase, while downregulating Myh2, which encodes a major structural muscle protein (Fig. 1D, E). Concomitantly, DEX suppressed the expression of myogenic regulatory factors (MRFs) Myod1, Myog, Myf5, and Myf6 (Fig. 1F-I). PU treatment significantly reversed these DEX-induced alterations, downregulating Fbxo32 and upregulating Myh2 and the MRFs (Fig. 1D-I).

Furthermore, Western blot analysis was employed to examine PU’s impact on key signaling pathways. DEX treatment (50 µM) markedly reduced the phosphorylation of Akt, p70-S6K1 (a downstream target of mTOR), and 4E-BP1 (another mTOR target), indicating suppression of the anabolic Akt/mTOR pathway (Fig. 1J). PU treatment, particularly at 40 µg/mL, markedly increased the phosphorylation of Akt, p70-S6K1, and 4E-BP1 in the presence of DEX, although levels did not fully return to those observed in the control group. Conversely, DEX treatment increased the levels of phosphorylated FoxO3a and the protein expression of the atrophy-related E3 ubiquitin ligase MurF1 (Fig. 1K). PU treatment, especially at 40 µg/mL, decreased the levels of phosphorylated FoxO3a and attenuated the DEX-induced upregulation of MurF1 (Fig. 1K). These findings suggest that PU directly protects C2C12 myotubes by activating anabolic Akt/mTOR signaling and inhibiting catabolic FoxO3a activation and downstream ubiquitin ligase expression.

PU mitigates DEX-Induced oxidative stress and improves mitochondrial function in C2C12 myotubes

Given that oxidative stress and mitochondrial dysfunction contribute to muscle atrophy [27], we investigated PU’s effects on these parameters. DEX treatment significantly increased intracellular ROS production and decreased GSH levels in C2C12 myotubes (Fig. 2A, B). DEX also impaired mitochondrial function, reducing mitochondrial content (MitoTracker Green staining) and cellular ATP production (Fig. 2C, D). PU co-treatment effectively mitigated these detrimental effects, suppressing ROS, restoring GSH, and enhancing mitochondrial content and ATP levels (Fig. 2A-D). Furthermore, PU significantly upregulated the mRNA expression of antioxidant enzymes Sod1 and Cat (Fig. 2E, F) and Ppargc1a (PGC-1α), a master regulator of mitochondrial biogenesis (Fig. 2G). A trend towards increased expression of Ucp3 and Tomm20, markers of mitochondrial function, was also observed, although not reaching statistical significance (Fig. 2H, I).

[See PDF for image]

Fig. 2

PU protects muscle cells from Dex-induced oxidative stress and mitochondrial dysfunction. (A) Intracellular reactive oxygen species (ROS) levels, measured using the DCFH-DA assay. (B) Intracellular reduced glutathione (GSH) levels. (C) Mitochondrial content, assessed by MitoTracker Green staining. (D) Cellular ATP content, quantified using a luciferase-based assay. (E, F) mRNA expression levels of antioxidant enzyme genes Sod1 and Cat, determined by qRT-PCR. (G-I) mRNA expression levels of genes involved in mitochondrial biogenesis and function: Ppargc1a, Ucp3, and Tomm20. β-actin was used as an internal control. All data are presented as mean ± SEM from at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test

Preliminary phytochemical analysis of PU identified eight known compounds (Supporting Information Fig. S1). However, bioactivity-guided fractionation and subsequent in vitro testing of these isolated compounds did not reveal a single constituent responsible for the potent anti-atrophic effects of the crude PU extract (Supporting Information Fig. S2), suggesting potential synergistic or additive interactions among multiple components.

PU ameliorates DEX-Induced muscle atrophyand improves physical performance in vivo

To validate the in vitro findings, PU was evaluated in a mouse model of DEX-induced muscle atrophy (experimental design, Fig. 3A). While DEX treatment (25 mg/kg/day, i.p., for 14 days) did not significantly affect overall body weight during the treatment period (Fig. 3B), it markedly reduced forelimb grip strength (Fig. 3C). Oral PU administration (10, 50, and 100 mg/kg/day) dose-dependently mitigated this decline, with the 100 mg/kg dose demonstrating efficacy comparable or superior to the positive control, Oxy (Fig. 3C). Qualitative observations indicated that DEX-treated mice exhibited signs consistent with muscle wasting, such as reduced muscle bulk, effects that appeared to be mitigated in the PU-treated groups.

[See PDF for image]

Fig. 3

PU ameliorates DEX-induced muscle atrophy in mice. (A) Experimental design of the in vivo study. Mice received daily oral gavage of vehicle, PU (10, 50, or 100 mg/kg/day), or oxymetholone (Oxy, 50 mg/kg/day) for four weeks. During the final two weeks, all groups except the control group received daily intraperitoneal injections of dexamethasone (DEX, 25 mg/kg/day). (B) Body weight of mice. (C) Forelimb grip strength measured at the end of the study. (D) Representative hematoxylin and eosin (H&E)-stained cross-sections of gastrocnemius muscle and quantification of muscle fiber cross-sectional area (CSA). (E-H) Muscle mass relative to total body weight for (E) quadriceps, (F) gastrocnemius, (G) tibialis anterior (TA), and (H) soleus muscles. (I-L) mRNA expression levels of atrophy-related gene Fbxo32 and myogenic regulatory factors Myog, Myod1, and Myf5 in TA muscle, determined by qRT-PCR. β-actin was used as an internal gene control. Data represent the mean ± SEM with eight to ten biological replicates. Data are presented as mean ± SEM (n = 8–10 mice per group). *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test. Con, Control

Histological analysis of gastrocnemius muscle revealed that DEX treatment significantly reduced muscle fiber CSA (Fig. 3D). PU treatment, at all tested doses, significantly attenuated this reduction in fiber CSA (Fig. 3D). Consistent with this, DEX significantly reduced the mass of hindlimb muscles (quadriceps, gastrocnemius, tibialis, and soleus) (Fig. 3E-H). PU treatment dose-dependently counteracted these reductions, with notable improvements in gastrocnemius and tibialis anterior mass (Fig. 3F, G). At the molecular level in the tibialis muscle, PU treatment reversed the DEX-induced downregulation of MRFs (Myod1, Myog, Myf5) and upregulation of Fbxo32 (Fig. 3I-L), mirroring the in vitro findings.

Furthermore, DEX administration significantly impaired exercise capacity, reducing maximal running speed, time to exhaustion, and total distance covered (Fig. 4A-C). PU treatment, particularly at 100 mg/kg/day, significantly improved these exercise parameters (Fig. 4A-C). This functional improvement was associated with enhanced mitochondrial biogenesis markers in tibialis muscle; PU significantly upregulated Ppargc1a and its downstream target Tfam, with a trend towards increased Nrf1 expression (Fig. 4D-F).

[See PDF for image]

Fig. 4

PU improves exercise capacity and promotes mitochondrial biogenesis in skeletal muscle of DEX-treated mice. (A) Maximal running speed. (B) Time to exhaustion. (C) Total distance run during a treadmill exercise test. (D-F) mRNA expression levels of genes involved in mitochondrial biogenesis (Pparg1a, Nrf1, and Tfam) in TA muscle, determined by qRT-PCR. Data are presented as mean ± SEM (n = 8–10 mice per group). *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test

PU modulates gut microbiota composition and diversity in DEX-Treated mice

Given the emerging role of the gut-muscle axis, we investigated PU’s impact on the gut microbiome. 16 S rRNA gene sequencing of cecal contents revealed that DEX treatment significantly reduced alpha diversity (Shannon index, observed ASVs, Pielou’s evenness) compared to control mice (Fig. 5A-C). PU treatment, particularly at 100 mg/kg/day, restored these alpha diversity indices (Fig. 5A-C). Beta diversity analysis (PCoA of unweighted UniFrac distances) showed distinct clustering, indicating that PU induced shifts in the overall microbial community structure (Fig. 5D).

[See PDF for image]

Fig. 5

PU modulates gut microbial diversity and composition in DEX-treated mice. (A-C) Alpha diversity indices: (A) Shannon diversity index, (B) Observed ASVs (Amplicon Sequence Variants), (C) Pielou’s evenness. (D) Beta diversity analysis using Principal Coordinate Analysis (PCoA) of Bray-Curtis distances. (E, F) Relative abundance of bacterial taxa at the (E) phylum and (F) order levels. (G-I) Relative abundance of specific genera: Clostridium sensu stricto 1 (G), Bilophila (H), and Lactobacillus gasseri (I). Data are presented as mean ± SEM (n = 8–10 mice per group). *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test for (A-C) and (G-I); PERMANOVA for (D)

At the phylum level, PU treatment progressively mitigated the DEX-induced decrease in Firmicutes abundance (Fig. 5E) and a similar trend was observed for the order Lachnospirales (Fig. 5F). At the genus level, PU (100 mg/kg) reversed the DEX-induced increase in Clostridium sensu stricto 1 (Fig. 5G) and significantly enriched Bilophila (Fig. 5H). Notably, while DEX reduced the abundance of Lactobacillus gasseri, PU treatment at 10 and 50 mg/kg/day increased its relative abundance (Fig. 5I). In vitro, PU supplementation (10–200 µg/mL) promoted the growth of L. gasseri ATCC 19,992 in a concentration-dependent manner (Supporting Information Fig. S3A), suggesting a potential prebiotic effect of PU on this strain.

Gut microbial signatures correlate with muscle protection by PU

Correlation analyses linked gut microbiota alterations to muscle health. Overall gut microbiota beta diversity (Bray-Curtis) significantly associated with tibialis and gastrocnemius muscle mass and grip strength (Adonis analysis, Fig. 6A). Spearman correlation revealed positive associations between the relative abundance of L. gasseri and Bilophila with both muscle mass and grip strength (Fig. 6B-H). PICRUSt2-based functional prediction and ANCOM analysis identified nine microbial enzymes that were significantly altered by PU treatment compared to the DEX group (Table S2). Notably, D-lactate dehydrogenase (cytochrome; EC 1.1.2.4) abundance showed a significant positive correlation with grip strength and muscle mass (Fig. 7A). Furthermore, L. gasseri abundance positively correlated with several of these beneficial microbial enzymes, including D-lactate dehydrogenase (Fig. 7B), the levels of which were increased in all PU-treated groups (Fig. 7C-F).

[See PDF for image]

Fig. 6

Correlation analysis between gut microbial composition and muscle atrophy-related markers in DEX-treated mice. (A) Association between gut microbial beta diversity (Bray-Curtis distances) and muscle atrophy-related traits, assessed by Adonis analysis. P-values are displayed within the bars. (B) Heatmap depicting Spearman’s rank correlation coefficients between the relative abundance of identified microbial genera and markers associated with muscle atrophy. Taxonomic classifications (class and phylum) are indicated by color coding. P-values were adjusted for multiple comparisons using the Benjamini-Hochberg (BH) false discovery rate (FDR) method. (C-H) Scatter plots showing Spearman’s rank correlations between the relative abundance of Lactobacillus gasseri (C-E) and Bilophila (F-H) and specific muscle function/mass markers: (C) grip strength, (D) quadriceps muscle mass, (E) gastrocnemius muscle mass, (F) grip strength, (G) gastrocnemius muscle mass, and (H) soleus muscle mass. The blue line represents the fitted correlation, and the shaded area represents the 95% confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001, based on adjusted p-values

[See PDF for image]

Fig. 7

Relative abundance of candidate microbial enzymes and correlation with muscle atrophy-related traits. (A) Heatmap depicting Spearman’s rank correlation coefficients between the relative abundance of nine microbial enzymes (predicted by PICRUSt2) and muscle atrophy-related traits that were significantly affected by PU treatment. P-values were adjusted for multiple comparisons using the BH FDR method. (B) Scatter plots showing positive correlations between L. gasseri abundance and the abundance of four candidate microbial enzymes (EC 1.1.2.4, EC 2.1.1.186, EC 3.6.3.20, and EC 6.6.1.2). The blue line represents the fitted correlation, and the shaded area represents the 95% confidence interval. (C-F) Relative abundance of specific candidate microbial enzymes: (C) D-lactate dehydrogenase (cytochrome) (EC 1.1.2.4), (D) 23 S rRNA (cytidine2498-2’-O)-methyltransferase (EC 2.1.1.186), (E) cobaltochelatase (EC 6.6.1.2), and (F) glycerol-3-phosphate transport ATPase (EC 3.6.3.20). Data are presented as mean ± SEM (n = 8–10 mice per group). *P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant, compared to the DEX-treated group, unless otherwise indicated; one-way ANOVA followed by Tukey’s post-hoc test

Direct administration of L. gasseri ATCC 19,992 partially mimics pu’s protective effects against DEX-Induced muscle atrophy

To directly assess the contribution of L. gasseri to muscle protection, DEX-treated mice were orally administered L. gasseri ATCC 19,992 (experimental design, Fig. 8A). While L. gasseri did not reverse DEX-induced body weight decline (Fig. 8B), it significantly restored grip strength and the mass of all four measured hindlimb muscles (quadriceps, gastrocnemius, tibialis anterior, and soleus) (Fig. 8C, D). Histologically, L. gasseri administration effectively reversed DEX-induced reductions in gastrocnemius muscle fiber CSA (Fig. 8E). Mechanistically, in the tibialis muscle, L. gasseri treatment downregulated Trim63 expression and upregulated the structural protein (Myh2) and MRFs (Myog, Myod1, Myf5, Myf6) (Fig. 8F-K).

[See PDF for image]

Fig. 8

Effect of L. gasseri on muscle atrophy and exercise performance in DEX-treated mice. (A) Animal study design. (B-D) body weight (B), grip strength normalized to body weight (C), and muscle weight relative to body weight for quadriceps, gastrocnemius, tibialis, and soleus muscles (D). (E) Hematoxylin and eosin (H & E) staining of the gastrocnemius muscle and quantification of the average muscle fiber cross-sectional area. (F-K) mRNA expression of proteolysis and myogenesis-related markers. (L) Exercise performance including maximal speed capacity, time to exhaustion, and distance to exhaustion. (M-O) mRNA expression of mitochondrial biogenesis markers. Data represent the mean ± SEM with eight to ten biological replicates. The black dots on the bar graph represent the data values for each mouse on the x-axis. Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test (*P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant)

Furthermore, L. gasseri supplementation significantly improved exercise performance, restoring maximal speed, exhaustion time, and distance run to levels comparable with or even exceeding the control group (Fig. 8L). This was associated with the upregulated expression of mitochondrial biogenesis markers (Nrf1, Ucp3, Tomm20) in the tibialis muscle (Fig. 8M-O).

L. gasseri administration induces specific gut microbiota alterations

16 S rRNA sequencing of cecal samples from L. gasseri-treated mice revealed significant alterations in the overall gut microbial community beta diversity compared to the DEX group, although alpha diversity metrics and phylum/class level abundances were not significantly different (Fig. 9A-C). This suggests targeted rather than broad taxonomic shifts. At the genus level, L. gasseri administration led to a significant enrichment of Weissella koreensis and Akkermansia (Fig. 9D), both of which positively correlated with grip strength (Fig. 9E). These findings indicate that orally administered L. gasseri can modulate the gut ecosystem in a manner that supports muscle health, providing direct evidence for its role in the gut-muscle axis during DEX-induced atrophy.

[See PDF for image]

Fig. 9

Effect of L. gasseri gut microbial diversity and bacterial taxa. (A, B) Relative abundance of bacterial taxa in the Phylum (A) and Class (B) level. The X-axis is the group, and the Y-axis is the proportion of relative abundance of bacteria; different colors represent different bacteria taxa. (B) Alpha diversity indices, Shannon diversity, Observed ASV, and Faith’s Phylogenetic diversity. (C) Unweighted unifrac and Bray Curtis beta diversity. (D) Relative abundance of specific genus levels including Weissella koreensis and Akkermansia. (E) Spearman’s rank correlation coefficients between grip strength and the relative abundance of Weissella koreensis or Akkermansia. The blue line represents the first-order correlation function. The gray area represents the confidence interval. Data represent the mean ± SEM with eight to ten biological replicates. The black dots on the bar graph represent the data values for each mouse on the x-axis. Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test (*P < 0.05, **P < 0.01, ***P < 0.005, ns: not significant)

Discussion

Glucocorticoid-induced muscle atrophy remains a significant clinical challenge, driving the search for novel, safe therapeutic interventions [1, 2, 3–4]. This study provides a comprehensive pharmacological investigation of PU, a medicinal mushroom, for its potential to counteract DEX-induced muscle wasting. P. umbellatus has a history of use in traditional systems for various ailments, and its reported biological activities, such as alleviating systemic stressors like inflammation and oxidative damage [6], align with factors relevant to muscle health. Our in vitro data strongly support PU’s direct muscle-protective actions. PU treatment attenuated DEX-induced C2C12 myotube atrophy by downregulating the key E3 ubiquitin ligase Fbxo32 and upregulating Myh2 and crucial myogenic regulatory factors (MRFs) [9, 10, 45, 46]. Mechanistically, Western blot analyses confirmed that PU beneficially shifts intracellular signaling by stimulating the pro-anabolic Akt/mTOR cascade (evidenced by increased phosphorylation of Akt, p70-S6K1, and 4E-BP1) and concurrently inhibiting the catabolic FoxO3a pathway and its downstream target MurF1 [28]. Furthermore, PU effectively countered DEX-induced mitochondrial dysfunction and oxidative stress in vitro by reducing ROS, restoring GSH, and upregulating Ppargc1a (PGC-1α) and the antioxidant enzyme genes (Sod1, Cat) [29, 30, 31, 32, 33, 34, 35–36].

These robust in vitro benefits translated effectively in vivo. Oral PU administration to DEX-treated mice resulted in dose-dependent improvements in grip strength, muscle mass, and exercise performance, with an efficacy at the highest dose comparable with or superior to oxymetholone, suggesting potential as a natural product-derived alternative with a potentially favorable safety profile [37, 38]. These functional gains were supported by an increased muscle fiber CSA and the favorable modulation of MRF and Fbxo32 expression in muscle tissue.

A critical and novel dimension of our study is the elucidation of PU’s significant impact on the gut-muscle axis. PU treatment reversed DEX-induced gut dysbiosis, restoring alpha diversity and, importantly, enriching L. gasseri, a probiotic with known muscle-protective potential [39, 40, 41–42] [60–63]. Our in vitro data further suggest a prebiotic-like effect of PU on L. gasseri growth. The pivotal role of this bacterium was substantiated by our in vivo experiments where direct oral administration of L. gasseri ATCC 19,992 alone partially recapitulated the muscle-protective effects of PU, including improvements in muscle mass, function, and molecular markers of myogenesis and mitochondrial biogenesis. This provides direct evidence for L. gasseri’s contribution to PU’s therapeutic profile. Interestingly, L. gasseri administration also modulated the broader gut microbiota, enriching other beneficial taxa like Akkermansia muciniphila [43], which correlated with improved muscle function, suggesting a cascade of positive microbial interactions. Our correlation analyses also linked these PU- and L. gasseri-induced microbial shifts, and predicted increases in specific microbial enzymes like D-lactate dehydrogenase (cytochrome), with improved muscle parameters, hinting at functional metabolic contributions from the altered microbiome.

While L. gasseri emerges as a key player, the comprehensive benefits of PU likely stem from a broader synergistic action of its multiple constituents, a common characteristic observed in many natural product extracts. Our bioactivity-guided fractionation efforts did not pinpoint a single ubiquitously active compound, reinforcing this notion. The dual action of PU—directly on muscle cells and indirectly via the gut microbiome—presents a compelling therapeutic paradigm, where direct cellular protection is complemented by sustained systemic benefits derived from a healthier gut ecosystem.

Limitations and strengths of the study

This study has several limitations that should be acknowledged. First, while our bioactivity-guided fractionation isolated eight known compounds from PU, none individually replicated the full protective effects of the crude extract in vitro, suggesting that PU’s efficacy likely stems from synergistic interactions or yet unisolated active components; further phytochemical investigation is warranted. Second, our findings are based on an acute DEX-induced muscle atrophy model in mice, and future studies are needed to validate PU’s efficacy in other muscle wasting models (e.g., sarcopenia, cachexia) and to assess long-term safety before clinical translation. Third, while we demonstrated L. gasseri’s contribution and correlated microbial changes with muscle health, definitive proof of the gut microbiota’s causal role in PU’s full effect requires future gold-standard experiments like fecal microbiota transplantation (FMT) or studies in germ-free/antibiotic-treated mice. Comprehensive metabolomic analyses are also needed to identify specific microbial metabolites mediating the gut-muscle communication. Fourth, while key signaling pathways and atrophy markers were validated at the protein level in vitro, extensive protein-level confirmation for all in vivo qPCR targets (e.g., MyHC isoforms, Pax7) was not conducted in this study.

Despite these limitations, this study possesses significant strengths. It provides a robust pharmacological investigation of PU, a medicinal mushroom with a history of traditional use, bridging this background with modern scientific validation. A key strength is the elucidation of a novel dual mechanism involving both direct muscle-protective effects (via Akt/mTOR activation, FoxO3a inhibition, mitochondrial enhancement, and oxidative stress reduction) and significant, beneficial modulation of the gut microbiome. The direct in vivo testing of L. gasseri administration provides strong support for its role in PU’s efficacy. The comprehensive approach, utilizing well-established in vitro and in vivo models with detailed molecular, functional, and advanced gut microbiome analyses, offers a multi-layered understanding of PU’s therapeutic potential. The consistency of findings across platforms lends considerable weight to our conclusions, highlighting PU as a promising candidate for developing standardized botanical drugs or functional foods for muscle health.

This study provides compelling pharmacological evidence that PU effectively mitigates DEX-induced muscle atrophy through a sophisticated dual mechanism. PU directly enhances muscle anabolism by activating Akt/mTOR signaling and inhibiting FoxO3a, improves mitochondrial function, and reduces oxidative stress in muscle cells. Concurrently, PU beneficially remodels the gut microbiota, increasing diversity and notably enriching L. gasseri. Our novel findings demonstrate that direct administration of L. gasseri partially mimics PU’s muscle-protective effects, underscoring the significant contribution of the gut-muscle axis to PU’s therapeutic action. Although a single active compound was not identified, the comprehensive benefits of PU likely arise from a synergistic interplay of its multiple constituents, a characteristic often observed with natural product extracts. In summary, this research elucidates multifaceted pharmacological actions of PU, highlighting its potential for development as a standardized botanical therapeutic or functional food ingredient to combat muscle wasting conditions, leveraging both its direct muscle-enhancing properties and its capacity to foster a gut microbiome conducive to muscle health.

Acknowledgements

This research was supported by Korea Institute of Science and Technology intramural research grants (2E32621 and 2Z07251).

Author contributions

Nguyen Bao Ngoc: Methodology, Formal analysis, Data curation, Visualization, Writing – original draft. Subeen Kim: Methodology, Formal analysis, Data curation, Writing – original draft. Hye-Young Youn: Methodology, Formal analysis, Data curation, Writing – review & editing. Huitae Min: Methodology, Formal analysis, Data curation. Tam Thi Le: Methodology, Formal analysis, Writing – chemistry part. Mauliasari Intan Rizki: Methodology, Formal analysis. Kwang Hyun Cha: Methodology, Formal analysis. Dae Won Kim: Writing – review & editing. Young Tae Park: Methodology, Formal analysis. Sang Hoon Jung: Supervision, Funding acquisition. Myungsuk Kim: Supervision, Funding acquisition, Conceptualization, Methodology, Formal analysis, Visualization, Writing – review & editing.

Funding

This research was supported by Korea Institute of Science and Technology intramural research grants (2E32621).

Data availability

The data used in this article, including raw chemical data and results from in vitro/in vivo experiments, are publicly available on Figshare at https://doi.org/10.6084/m9.figshare.27020872.

Declarations

Competing interests

The authors declare that they have no conflicts of interest in this work. Regarding the work submitted, we declare that we do not have any commercial or associative interests that would constitute a conflict of interest.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Evans, WJ. Skeletal muscle loss: cachexia, sarcopenia, and inactivity. Am J Clin Nutr; 2010; 91, 4 pp. 1123S-1127S.1:CAS:528:DC%2BC3cXjvVCkur0%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20164314]

2. Takemoto, Y; Fukada, S-I. Molecular mechanism maintaining muscle satellite cells and the roles in sarcopenia. Clin Calcium; 2017; 27, 3 pp. 339-344. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28232647]

3. Lo, JH et al. Sarcopenia: current treatments and new regenerative therapeutic approaches. J Orthop Translat; 2020; 23, pp. 38-52. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32489859][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7256062]

4. Oray, M et al. Long-term side effects of glucocorticoids. Expert Opin Drug Saf; 2016; 15, 4 pp. 457-465.1:CAS:528:DC%2BC28Xitlynt78%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26789102]

5. Wang, BY et al. Is dexamethasone-induced muscle atrophy an alternative model for naturally aged sarcopenia model?. J Orthop Translat; 2023; 39, pp. 12-20. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36605620]

6. Zhao, Y-Y. Traditional uses, phytochemistry, pharmacology, pharmacokinetics and quality control of Polyporus umbellatus (Pers.) fries: a review. J Ethnopharmacol; 2013; 149, 1 pp. 35-48.1:CAS:528:DC%2BC3sXhtV2rtb%2FO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23811047]

7. Kawashima, N et al. Choreito, a formula from Japanese traditional medicine (Kampo medicine), for massive hemorrhagic cystitis and clot retention in a pediatric patient with refractory acute lymphoblastic leukemia. Phytomedicine; 2012; 19, 12 pp. 1143-1146. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22819300]

8. Wang, J-H et al. Antifibrotic effects of CGX, a traditional herbal formula, and its mechanisms in rats. J Ethnopharmacol; 2010; 127, 2 pp. 534-542. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19833189]

9. Zhou, W-W; Lin, W-H; Guo, S-X. Two new polyporusterones isolated from the sclerotia of Polyporus umbellatus. Chem Pharm Bull; 2007; 55, 8 pp. 1148-1150.1:CAS:528:DC%2BD1cXlvF2itbY%3D

10. He, P-f et al. Structure Elucidation and antioxidant activity of a novel polysaccharide from Polyporus umbellatus sclerotia. Int J Biol Macromol; 2016; 82, pp. 411-417.1:CAS:528:DC%2BC2MXhslKgtL3M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26476244]

11. Peng, K et al. Polyporus umbellatus polysaccharides ameliorates carbon tetrachloride-induced hepatic injury in mice. Afr J Pharm Pharmacol; 2012; 6, 37 pp. 2686-2691.1:CAS:528:DC%2BC38XhsFOrsLjM

12. Li, X; Xu, W; Chen, J. Polysaccharide purified from Polyporus umbellatus (Per) Fr induces the activation and maturation of murine bone-derived dendritic cells via toll-like receptor 4. Cell Immunol; 2010; 265, 1 pp. 50-56.1:CAS:528:DC%2BC3cXhtV2iu73K [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20673883]

13. Jia, W et al. Homogeneous polyporus polysaccharide inhibits bladder cancer by polarizing macrophages to M1 subtype in tumor microenvironment. BMC Complement Med Ther; 2021; 21, 1 150.1:CAS:528:DC%2BB3MXht1KhtrfF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34034714][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8152148]

14. Bindels, LB; Delzenne, NM. Muscle wasting: the gut microbiota as a new therapeutic target?. Int J Biochem Cell Biol; 2013; 45, 10 pp. 2186-2190.1:CAS:528:DC%2BC3sXhtFKlu7fI [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23831839]

15. Canfora, EE; Jocken, JW; Blaak, EE. Short-chain fatty acids in control of body weight and insulin sensitivity. Nat Rev Endocrinol; 2015; 11, 10 pp. 577-591.1:CAS:528:DC%2BC2MXhtlGgtL3E [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26260141]

16. Frampton, J et al. Short-chain fatty acids as potential regulators of skeletal muscle metabolism and function. Nat Metab; 2020; 2, 9 pp. 840-848.1:CAS:528:DC%2BB3cXlvFKhtbo%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32694821]

17. Liu, C et al. Understanding the gut microbiota and sarcopenia: a systematic review. J Cachexia Sarcopenia Muscle; 2021; 12, 6 pp. 1393-1407. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34523250][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8718038]

18. Ko H et al (2025) Poncirus trifoliata extract and its active coumarins alleviate Dexamethasone-Induced skeletal muscle atrophy by regulating protein synthesis, mitochondrial biogenesis, and gut microbiota. Phytother Res

19. Nguyen, NB et al. Cornflower extract and its active components alleviate Dexamethasone-Induced muscle wasting by targeting cannabinoid receptors and modulating gut microbiota. Nutrients; 2024; 16, 8 1130.1:CAS:528:DC%2BB2cXpvVKrtrk%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38674820][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11054969]

20. Ko, H et al. Identification of components from Aralia elata and their effects on muscle health and gut microbiota. J Funct Foods; 2024; 121, 106384.1:CAS:528:DC%2BB2cXhslCjsL3O

21. Bolyen, E et al. Reproducible, interactive, scalable and extensible Microbiome data science using QIIME 2. Nat Biotechnol; 2019; 37, 8 pp. 852-857.1:CAS:528:DC%2BC1MXhsVeksr%2FO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31341288][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7015180]

22. Callahan, BJ et al. DADA2: High-resolution sample inference from illumina amplicon data. Nat Methods; 2016; 13, 7 pp. 581-583.1:CAS:528:DC%2BC28XosVWitb4%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27214047][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4927377]

23. Quast, C et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res; 2012; 41, D1 pp. D590-D596. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23193283][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3531112]

24. McMurdie, PJ; Holmes, S. Phyloseq: an R package for reproducible interactive analysis and graphics of Microbiome census data. PLoS ONE; 2013; 8, 4 e61217.1:CAS:528:DC%2BC3sXntVWht7w%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23630581][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3632530]

25. Kaul, A et al. Analysis of Microbiome data in the presence of excess zeros. Front Microbiol; 2017; 8, 2114. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29163406][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5682008]

26. Caspi, R et al. The metacyc database of metabolic pathways and enzymes and the biocyc collection of pathway/genome databases. Nucleic Acids Res; 2014; 42, D1 pp. D459-D471.1:CAS:528:DC%2BC2cXos12h [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24225315]

27. Kim YI et al (2022) Antioxidant Activity of Valeriana fauriei Protects against Dexamethasone-Induced Muscle Atrophy. Oxid Med Cell Longev, 2022: p. 3645431

28. Schiaffino, S et al. Mechanisms regulating skeletal muscle growth and atrophy. FEBS J; 2013; 280, 17 pp. 4294-4314.1:CAS:528:DC%2BC3sXht1ymu7nK [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23517348]

29. Schakman, O et al. Glucocorticoid-induced skeletal muscle atrophy. Int J Biochem Cell Biol; 2013; 45, 10 pp. 2163-2172.1:CAS:528:DC%2BC3sXht1Sjur3K [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23806868]

30. Romanello, V; Sandri, M. Mitochondrial biogenesis and fragmentation as regulators of protein degradation in striated muscles. J Mol Cell Cardiol; 2013; 55, pp. 64-72.1:CAS:528:DC%2BC38Xht12jsb3L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22902321]

31. Schiaffino, S; Sandri, M; Murgia, M. Activity-dependent signaling pathways controlling muscle diversity and plasticity. Physiology; 2007; 22, 4 pp. 269-278.1:CAS:528:DC%2BD2sXhtVentrbJ [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17699880]

32. Yeon, M et al. Gomisin G improves muscle strength by enhancing mitochondrial biogenesis and function in disuse muscle atrophic mice. Biomed Pharmacother; 2022; 153, 113406.1:CAS:528:DC%2BB38Xit1yjt7fF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36076532]

33. Hees, JT; Harbauer, AB. Metabolic regulation of mitochondrial protein biogenesis from a neuronal perspective. Biomolecules; 2022; 12, 11 1595.1:CAS:528:DC%2BB38XivVynurrF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36358945][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9687362]

34. Leone, TC et al. PGC-1α deficiency causes multi-system energy metabolic derangements: muscle dysfunction, abnormal weight control and hepatic steatosis. PLoS Biol; 2005; 3, 4 e101. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15760270][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1064854]

35. Ábrigo, J et al. Role of oxidative stress as key regulator of muscle wasting during cachexia. Oxidative Med Cell Longev; 2018; 2018, 1 2063179.

36. Ceci, R et al. Moringa oleifera leaf extract protects C2C12 myotubes against H2O2-induced oxidative stress. Antioxidants; 2022; 11, 8 1435.1:CAS:528:DC%2BB38Xit1Kns73L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35892637][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9330721]

37. Wang, Y et al. Protective effects of Lepidium draba L. leaves extract on testis histopathology, oxidative stress indicators, serum reproductive hormones and inflammatory signalling in oxymetholone-treated rat. Andrologia; 2021; 53, 11 e14239.1:CAS:528:DC%2BB3MXitVOhtbzE [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34520070]

38. Feng, J et al. Renoprotective and hepatoprotective activity of Lepidium draba L. extracts on oxymetholone-induced oxidative stress in rat. J Food Biochem; 2022; 46, 9 e14250.1:CAS:528:DC%2BB38XhvVWks7vN [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35633194]

39. Lee, J et al. Lactobacillus rhamnosus JY02 ameliorates sarcopenia by anti-atrophic effects in a dexamethasone-induced cellular and murine model. J Microbiol Biotechnol; 2023; 33, 7 915.1:CAS:528:DC%2BB3sXhsleis7bP [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36998149][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10394339]

40. Lee, K et al. Lactobacillus plantarum HY7715 ameliorates sarcopenia by improving skeletal muscle mass and function in aged balb/c mice. Int J Mol Sci; 2021; 22, 18 10023.1:CAS:528:DC%2BB3MXitlGnt7fO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34576187][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8466743]

41. Chen, Y-M et al. Lactobacillus plantarum TWK10 supplementation improves exercise performance and increases muscle mass in mice. Nutrients; 2016; 8, 4 205. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27070637][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4848674]

42. Chen, L-H et al. Lactobacillus paracasei PS23 decelerated age-related muscle loss by ensuring mitochondrial function in SAMP8 mice. Aging; 2019; 11, 2 756.1:CAS:528:DC%2BC1MXitlagtLfF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30696799][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6366975]

43. Naito, Y; Uchiyama, K; Takagi, T. A next-generation beneficial microbe: Akkermansia muciniphila. J Clin Biochem Nutr; 2018; 63, 1 pp. 33-35.1:CAS:528:DC%2BC1MXisF2jt7Y%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30087541][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6064808]