Fischer 344 rats (6 weeks old, male; body weight, 150±20 g; CLEA Japan) were randomly divided into five groups and treated for 12 weeks as follows (Fig. 1A; n=10 in each group): i) With a choline-sufficient amino acid-defined (CSAA) diet (Research Diets, Inc.) with lactose hydrate (FUJIFILM, Wako Pure Chemical Corporation) as the vehicle; ii) with a choline-deficient l-amino acid-defined (CDAA) diet (Research Diets Inc.) with the vehicle; iii) with the CDAA diet with rifaximin (ASKA Pharmaceutical Co. Ltd.; 100 mg/kg); iv) with the CDAA diet and L-carnitine (Otsuka Pharmaceutical Co. Ltd.; 200 mg/kg); and v) with the CDAA diet with rifaximin and L-carnitine, as previously described (23,24). All drugs were administered by intragastric gavage once a day. The rats were housed in plastic cages (2 rats/cage) in a pathogen-free room and were provided with free access to their diet and drinking water, and were kept under controlled, stable ambient conditions (23±3°C/12-h light/dark cycle with 50±20% humidity). At the end of the experiment, all rats underwent the following procedures: Euthanasia by an intraperitoneal injection of pentobarbital sodium (200 mg/kg), the opening of the abdominal cavity, blood collection via puncture of the aorta and harvesting of the liver, ileum and gastrocnemius muscle for histological and molecular evaluation. The anesthetized rats were then decapitated for assuring death. The present study was approved by the Animal Ethics Committee of Nara Medical University (approval no. 12764), and all protocols were performed in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.

L6 rat skeletal muscle myoblasts (cat. no. JCRB9081, Japanese Collection of Research Bioresources Cell Bank) were cultured and differentiated into myotubes as previously described (25,26). Briefly, the cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Nacalai Tesque, Inc.) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) and antibiotics (1% penicillin and streptomycin) at 37°C in a 5% CO₂ air environment. To induce differentiation into myocytes, the cells were further cultured in DMEM containing 2% horse serum for 8 days. The cells were supplemented with fresh medium every 48 h and were used at the stage of myotubes (60-70%) (25,27,28). Myogenic differentiation into myotubes was confirmed by using a AE2000-1080M microscope (Shimadzu Corporation) based on the morphological alignment, elongation and fusion (data not shown). Mycoplasma testing was performed using a MycoProbe^® Mycoplasma Detection kit (R&D Systems, Inc.) according to the manufacturer's protocol. The differentiated myotubes were stimulated with LPS (O55:B5; MilliporeSigma) or tumor necrosis factor-α (TNF-α) at various concentrations and treated with LPS (1.0 μg/ml) or TNF-α (20 ng/ml) and L-carnitine (5 mM) and/or rifaximin (10 μM) for 48 h.

According to the clinical criteria for sarcopenia assessment, the PMI (cross-sectional area/height²) was assessed on a single computed tomography (CT) slice (image) at the level of the L3 pedicle using Slice-O-Matic (Tomovision) (29). All rats underwent an abdominal CT scan before and at 4, 8, and 12 weeks after the start of CDAA feeding and/or rifaximin/L-carnitine treatment using CosmoScan FX (Rigaku Corporation) as previously described (26). The forelimb grip strength of the experimental rats was simultaneously measured using a grip strength meter (MK-380Si; Muromachi Kikai, Co., Ltd.) as previously described (26). During the grip strength test, the rats were allowed to use their front paws to grab a horizontal bar mounted on the gauge, and the tail was slowly pulled back. The peak tension was recorded at the time the mouse released the grip on the bar. Measurements were repeated three times, and the mean of three measurements was recorded.

The serum ALT and albumin concentrations in the rats were measured using a Rat Alanine Aminotransferase ELISA kit (cat. no. ab285264, Abcam) and Rat Albumin ELISA kit (cat. no. ab108789, Abcam), respectively. All samples were processed and assayed according to the manufacturer's protocol.

The serum IGF-1 concentrations in the rats were measured using a Mouse/Rat IGF-I/IGF-1 Quantikine ELISA kit (cat. no. MG100, R&D Systems, Inc.). All samples were processed and assayed according to the manufacturer's protocol.

The concentrations of ammonia in plasma and muscle (100 mg of rat gastrocnemius muscle tissue homogenate) were measured using the Ammonia Assay kit (cat. no. ab83360, Abcam), according to the manufacturer's protocol.

The intrahepatic TG concentrations in 100 mg frozen liver tissue per mouse were measured using the Triglyceride-Glo™ Assay (Promega Corporation), according to the manufacturer's instructions.

The levels of myostatin in 100 mg rat gastrocnemius muscle tissue homogenate were measured using a myostatin ELISA kit (cat. no. DGDF80, R&D Systems, Inc.) following the manufacturer's instructions.

A 4-kDa fluorescein isothiocyanate (FITC)-dextran (MilliporeSigma) solution was used for the intestinal permeability measurement, as previously described (30). Another set of rats (n=5) in each group was applied to evaluate in vivo intestinal permeability. Briefly, 6 h after initiating fasting conditions (only fasted from food), FITC-dextran was gently administered via oral gavage to the rats at 40 mg/kg, 200 μl body weight. Blood was collected from the portal vein at 1 h after the FITC-dextran administration. To evaluate the degree of gut permeability, blood was analyzed by the fluorescence measurement of the concentration of FITC-labeled dextran at an excitation wavelength of 490 nm and an emission wavelength of 520 nm using a NanoDrop 3300 fluorospectrometer (Thermo Fisher Scientific, Inc.).

The liver, ileum and gastrocnemius specimens were fixed in 10% formalin, incubated overnight at room temperature and embedded in paraffin. Sections of 5-μm thickness were stained with hematoxylin and eosin (H&E) and Sirius Red (performed at Narabyouri Research Co., Nara, Japan). For immunohistochemical staining, the liver tissue sections were blocked with 10% goat serum (Abcam) for 30 min following deparaffinization and antigen retrieval, and then incubated overnight at 4°C with mouse-monoclonal CD68 antibody (1:100; cat. no. GTX41868, GeneTex, Inc.). The sections were washed three times with phosphate-buffered saline and subsequently incubated with a goat anti-mouse IgG (H+L) HRP-conjugated secondary antibody (1:2,000; cat. no. 62-6520, Thermo Fisher Scientific, Inc.) for 30 min at room temperature. The slides were developed with DAB until the signal clearly appeared, and the nuclei were stained with hematoxylin for 5 min at room temperature, and images were obtained using a BX53 microscope (Olympus Corporation).

For immunofluorescence, the ileum sections were deparaffinized and rehydrated and blocked in a similar manner to immunohistochemical staining, and then rabbit-polyclonal Zonula occludens-1 (ZO-1; 1:100; cat. no. 61-7300, Invitrogen; Thermo Fisher Scientific, Inc.) and rabbit-polyclonal Occludin (1:100; cat. no. 71-1500, Invitrogen; Thermo Fisher Scientific, Inc.) were used as primary antibodies. Following overnight incubation at 4°C, the immunofluorescence detection of the primary antibodies was performed using goat anti-rabbit IgG (H+L) Alexa Fluor-conjugated secondary antibodies (1:200; cat. nos. R37116 and A-21207, Thermo Fisher Scientific, Inc.) for 1 h at room temperature. The sections were mounted on Vectashield mounting medium with 4′,6-diamidino-2-phenylindole Fluoromount-G mounting medium for fluorescent nucleic acid staining (Vector Laboratories, Inc.) and images were captured using a BZ-X700 microscope (Keyence Corporation). Semi-quantitative analysis was performed for five fields per section in high-power fields at ×400 magnification using ImageJ software version 64 (National Institutes of Health).

Total RNA was isolated from the liver and muscle tissues and cultured rat L6 myocyte cells. The RNeasy Mini kit (Qiagen GmbH) was used for the liver tissues and L6 cells, and the RNeasy Fibrous Tissue Mini kit (Qiagen GmbH) was used for the muscle tissues. The RNA samples were then treated with DNase in order to remove DNA contamination with TURBO DNA-free™ DNase (Invitrogen; Thermo Fisher Scientific, Inc.) and total RNA (2 μg) was reverse transcribed into complementary DNA (cDNA) using the High-Capacity RNA-to-cDNA kit (Applied Biosystems; Thermo Fisher Scientific, Inc.) by applying the following three stages: 37°C for 15 min, 85°C for 5 sec and then cooling at 4°C. qPCR was performed using the primer pairs listed in Table SI, and SYBR™-Green PCR Master Mix (Applied Biosystems; Thermo Fisher Scientific, Inc.), and an Applied Biosystems StepOnePlus™ Real-Time PCR^® system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The thermocycling conditions were as follows: 95°C for 5 min, followed by 40 cycles of 95°C for 10 sec, 60°C for 20 sec and then 72°C for 20 sec; melting at 95°C for 5 sec, 65°C for 60 sec and 97°C for 1 sec and cooling at 40°C for 10 sec. Relative expression was normalized to Gapdh expression, and estimated using the 2^-ΔΔCq method, and presented as the fold change relative to the control (31).

Whole cell lysate proteins were extracted from the intestinal and muscle tissues, and cultured rat L6 myocyte cells using tissue-protein extraction reagent (T-PER) supplemented with proteinase and phosphatase inhibitors (all from Thermo Fisher Scientific, Inc.). The protein concentration was measured using a protein assay (Bio-Rad Laboratories, Inc.). In total, 50 μg whole cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (NuPAGE™ 4-12%, Bis-Tris; Thermo Fisher Scientific, Inc.) and transferred to Invitrolon polyvinylidene difluoride membranes (Thermo Fisher Scientific, Inc.), which were subsequently blocked for 1 h with 5% bovine serum albumin (Thermo Fisher Scientific, Inc.) in Tris-buffered saline supplemented with Tween-20 (Cell Signaling Technology, Inc.). The membranes were then incubated overnight at 4°C with antibodies against rabbit-polyclonal ZO-1 (1:500; cat. no. 61-7300) and rabbit-polyclonal Occludin (1:125; cat. no. 71-1500) (from Invitrogen; Thermo Fisher Scientific, Inc.), p70S6K (1:1,000, cat. no. 9202), phosphorylated (p-) p70S6K (Thr389; 1:1,000, cat. no. 9205), nuclear factor-κB (NF-κB) p65 (1:1,000, cat. no. 8242), p-NF-κB p65 (Ser536; 1:1,000, cat. no. 3033), GAPDH (1:1,000, cat. no. 2118) and actin (1:1,000, cat. no. 4967) (from Cell Signaling Technology, Inc.). The membranes were washed and incubated at room temperature for 1 h with Amersham ECL horseradish peroxidase-conjugated immunoglobulin G F(ab)2 fragment antibody (1:5,000 dilution; cat. no. NA931, GE Healthcare; Cytiva) and developed using Clarity Western enhanced chemiluminescence substrate (Bio-Rad Laboratories, Inc.). Immunoblotting bands were densitometrically analyzed using ImageJ 64-bit Java 1.8.0 software (National Institutes of Health).

Total DNA was obtained from the gastrocnemius muscle tissue using a DNA Extractor^® TIS kit (FUJIFILM, Wako Pure Chemical Corporation). The mtDNA copy number was assessed using RT-qPCR as described in above according to mtDNA (Rnr2)/nDNA (Gapdh). The primer sequences used were as follows: RNR2 forward, 5′-AGC TAT TAA TGG TTC GTT TGT-3′ and reverse, 5′-AGG AGG CTC CAT TTC TCT TGT-3′; and nuclear-encoded GAPDH forward, 5′-GGA AAG ACA GGT GTT TTG CA-3′ and reverse, 5′-AGG TCA GAG TGA GCA GGA CA-3′. The PCR amplification process was as follows: One cycle at 95°C for 10 min, 40 cycles at 95°C for 15 sec and 60°C for 1 min. Both mtDNA and nDNA threshold cycle (CT) average values were obtained, and the mtDNA content was calculated relative to nDNA as mtDNA/nDNA=2^{(CTnDNA-CTmtDNA)}.

The oxygen consumption rate (OCR) was measured using the XFe96 extracellular flux analyzer (Agilent Technologies, Inc.) as previously described (32). Briefly, the L6 cells were seeded at 8,000 cells per well in a 96-well tissue culture plate 24 h before running the flux analyzer. To assess the mitochondrial respiration rate, the cells were incubated in XF base medium (Agilent Technologies, Inc.) supplemented with 10 mM glucose, 1 mM pyruvate and 2 mM glutamine for 1.5 h at 37°C. Subsequently, the OCR was measured following the addition of 1 μM oligomycin to inhibit ATP synthesis from oxidative phosphorylation, 1.5 μM carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) to uncouple the mitochondrial membrane that stimulates respiration, and a mixture containing 0.5 μM each of antimycin A and rotenone (A+R) to inhibit complex I and III that terminates mitochondrial oxidative phosphorylation. After the seahorse measurements were completed, total cellular content was measured using a CyQUANT Cell Proliferation Assay kit (Thermo Fisher Scientific, Inc.), and the OCR values were normalized to the cellular mitochondrial content (number of cells x mtDNA/nDNA). The basal OCR was calculated as follows: [OCR_(initial)-OCR_{(A + R)}]. The maximum OCR was computed as follows: [OCR_(FCCP)-OCR_{(A + R)}].

Tetramethylrhodamine methyl ester (TMRM; FUJIFILM, Wako Pure Chemical Corporation) were used to assess the mitochondrial membrane potential. The L6 cells were stained by addition of the dye to the culture medium at 100 nM for 30 min at 37°C. Subsequently, the cells were stained with the Hoechst 33342 nuclear dye (Thermo Fisher Scientific, Inc.) at a 1:500 dilution in PBS at room temperature for 10 min, washed with PBS, and visualized using a BZ-X700 (Keyence Corporation). The intensity of the fluorescence signal was quantified using ImageJ software version 64 (National Institutes of Health).

Statistical analyses were performed using Prism, version 9.1.2 (GraphPad Software, Inc.). Data were analyzed using one-way ANOVA followed by Tukey's test as a post hoc test. Values are presented as the mean ± standard deviation. A value of P<0.05 was considered to indicate a statistically significant difference.

The experimental design of the study is illustrated in Fig. 1A. The CDAA-fed rats exhibited a significant decrease in body weight and hepatomegaly as compared with the CSAA-fed rats; however, neither rifaximin nor L-carnitine inhibited these physical impairments (Fig. 1B and C). Serological analysis revealed elevated alanine aminotransferase (ALT) levels, as well as a decreased albumin level in the CDAA-fed rats; neither rifaximin nor L-carnitine significantly affected the ALT levels; however, L-carnitine suppressed the progression of hypoalbuminemia (Fig. 1D and E). Rifaximin also mildly improved CDAA-induced hyperammonemia (Fig. 1F). The histological assessment demonstrated that hepatic lipid accumulation in the CDAA-fed rats was suppressed by L-carnitine, but not by rifaximin (Fig. 1G). Concomitantly, treatment with L-carnitine decreased the hepatic TG levels, with the decreased expression of the lipogenesis-related gene, sterol regulatory element binding transcription factor 1 (Srebf1), and the increased expression of carnitine palmitoyltransferase 1A (Cpt1a) (Fig. 1H and I). Moreover, Sirius Red staining revealed that hepatic fibrosis progression was prevented by rifaximin, which was consistent with the reduced expression mRNA levels of hepatic profibrotic markers [i.e., actin alpha 2, smooth muscle (Acta2) and collagen type i alpha 1 chain (Col1a1)]; however, these events were not significantly mediated by L-carnitine (Fig. 1G-K).

Based on the suppression of liver fibrosis following rifaximin treatment, the present study then examined the pro-inflammatory status and the hepatic LPS/TLR4 pathway. The results revealed an extensive hepatic infiltration of CD68-positive macrophages in the CDAA-fed rats, which was attenuated by treatment with rifaximin (Fig. 2A and B). Corresponding with the upregulated expression of hepatic LPS-binding protein (Lbp), the mRNA levels of Tlr4 and its coreceptor, Cd14, were increased in the CDAA-fed rats. Notably, treatment with rifaximin inhibited these effects (Fig. 2C and D). In this context, rifaximin treatment significantly reduced the hepatic mRNA levels of pro-inflammatory cytokines, including Tnfa Il1b and Il6 (Fig. 2E). On the other hand, treatment with L-carnitine alone did not affect the inflammatory response (Fig. 2A-E).

Subsequently, the present study assessed the changes in the status of IGF-1, which is a potent myotrophic factor. Both the hepatic and serum concentrations of IGF-1 were decreased in the CDAA-fed rats, compared with the CSAA-fed rats (Fig. 2F and G). It was found that L-carnitine suppressed the CDAA-induced decrease in IGF-1 production (Fig. 2F and G). Of note, combination treatment with L-carnitine and rifaximin effectively enhanced the suppressive effects of L-carnitine on the reduction of the liver and serum IGF-1 levels in the CDAA-fed rats by reducing the LPS overload, which was recognized to suppress the hepatic production of IGF-1 (Fig. 2F and G).

To elucidate the functional mechanisms of the rifaximin-mediated prevention of the hepatic overload by endogenous LPS, intestinal permeability was then evaluated. Immunofluorescence staining revealed that the areas that were immuno-positive for ZO-1 and Occludin, the representative markers of tight junction proteins (TJPs), were profoundly less in number in the CDAA-fed group than in the CSAA-fed group; notably, this deprivation of TJPs was effectively restored by treatment with rifaximin (Fig. 3A-C). The results of western blot analysis supported these findings; rifaximin replenished the TJP protein levels (Fig. 3D and E). In addition, the leakage of plasma FITC-dextran was augmented by >2.5-fold by the CDAA diet and was inversely proportional with the deprivation of TJPs (Fig. 3F). Treatment with rifaximin led to a marked decreased in the leakage of FITC-dextran in the CDAA-fed rats (Fig. 3F). Treatment with L-carnitine alone did not affect the impairment of intestinal barrier function (Fig. 3).

Considering the marked weight loss of the rats, it was hypothesized that the CDAA-fed rats would undergo skeletal muscle atrophy. After 12 weeks of feeding, the body length of the rats was significantly shortened in the CDAA-fed group, and rifaximin or L-carnitine treatment did not reverse this change (Fig. 4A and B). After 4 weeks of feeding, the value of PMI, determined as the cross-sectional area/height², significantly decreased in the CDAA-fed group (Fig. 4C and D). Notably, rifaximin and L-carnitine suppressed the CDAA-induced decrease in PMI after 8 weeks of treatment, which was enhanced after 12 weeks of combined treatment (Fig. 4C-E).

The present study further investigated the efficacy of both agents on forelimb grip strength. The CDAA diet significantly weakened forelimb grip strength after 4 weeks (Fig. 4F and G). In accordance with the effects observed on muscle atrophy, monotherapy with rifaximin and L-carnitine prevented the CDAA-induced decrease in forelimb grip strength, and the combination of these agents augmented these preventive effects after 12 weeks (Fig. 4F and G).

To elucidate the effects of rifaximin and L-carnitine on skeletal muscle protein metabolism, the changes in protein synthesis and degradation in the gastrocnemius muscle tissues were further analyzed. As illustrated in Fig. 5A-C, the histological assessment of gastrocnemius muscle fiber revealed a marked decrease in fiber length and density in the CDAA-fed group. Treatment with either rifaximin or L-carnitine efficiently ameliorated muscle fiber atrophy, and the combination of both agents potentiated these effects (Fig. 5A-C). The phosphoinositide 3-kinase/Akt/mammalian target of rapamycin (mTOR) pathway is a pivotal process in muscle protein synthesis. In CDAA-fed group, the phosphorylation of intramuscular p70S6K (Thr389), a hallmark of activation by mTOR, was significantly diminished as compared to CSAA-fed group (Fig. 5D and E). This CDAA-induced decrease in p70S6K phosphorylation was suppressed by rifaximin and L-carnitine treatment (Fig. 5D and E).

The present study then evaluated the muscle levels of ammonia and myostatin, which is a myokine that inhibits muscle cell growth. In the CDAA-fed rats, the intramuscular ammonia levels were elevated along with changes in serum ammonia levels; this effect was attenuated by rifaximin treatment (Fig. 5F). Consistently, the marked increase in the myostatin levels in the CDAA-fed rats was suppressed by rifaximin treatment (Fig. 5G). Furthermore, in the CDAA-fed rats, there was an upregulation in the levels of the inflammatory mediators, Tnfa, Il1b and NLR family pyrin domain containing 3 (Nlrp3), following NF-κB activation in the skeletal muscle, and rifaximin potently suppressed the activation of these pro-inflammatory pathways (Fig. 5H-J). These effects were not observed in the muscles of the L-carnitine-treated rats (Fig. 5F-J).

Based on the fact that mitochondrial function dictates muscle fiber homeostasis, the present study investigated mitochondrial biogenesis in the gastrocnemius muscle tissue. The CDAA-fed rats manifested a prominent decrease in the intramuscular levels of mitochondrial biogenesis-related genes, peroxisome proliferator-activated receptor γ coactivator-1α (PGC-1α; Ppargc1a) and mitochondrial transcription factor A (TFAM; Tfam) (Fig. 5K). Of note, these decreases were efficiently restored by treatment with either rifaximin or L-carnitine; this was augmented by the combination of both agents (Fig. 5K).

Following the changes in these regulatory factors, the intramuscular mRNA expression levels of Atrogin-1 and muscle RING-finger protein-1 (MuRF-1), which are the pivotal markers of the ubiquitin-proteasome system (UPS), were upregulated in the CDAA-fed group; treatment with rifaximin and L-carnitine attenuated these effects (Fig. 5L).

In the model in the present study, it was found that L-carnitine supplementation attenuated the impairment of mitochondrial biogenesis and the progression of UPS in the course of muscle degradation. Therefore, the present study then examined whether L-carnitine could directly affect differentiated L6 rat myotubes in vitro (Fig. 6A). As shown in Fig. S1A and B, LPS stimulation decreased the mRNA levels of Ppargc1a and Tfam in a concentration-dependent manner, and conversely increased those of Atrogin-1 and MuRF-1 in rat myotubes. Of note, L-carnitine suppressed the LPS-induced decrease in the levels of mitochondrial biogenesis-related markers, and consequently inhibited the UPS stimulation by LPS (Fig. 6B and C). To further evaluate mitochondrial biogenesis, the cellular mitochondrial content was measured by quantifying mtDNA over nuclear DNA (mtDNA/nDNA). The results revealed a a marked reduction in mtDNA/nDNA in the LPS-stimulated L6 myotubes, and treatment with L-carnitine restored the LPS-stimulated reduction in mtDNA/nDNA (Fig. 6D). Moreover, mitochondrial membrane potential, an indicator of mitochondrial function, was assessed by the fluorescence intensity of TMRM that accumulates in the mitochondria. As shown in Fig. 6E and F, mitochondrial membrane potential was decreased in the LPS-stimulated L6 myotubes, and this decrease was suppressed by treatment with L-carnitine.

Subsequently, the cellular effects of both agents on mitochondrial respiration were examined using seahorse analyses to measure the OCR of L6 myotubes stimulated with LPS. The OCR values were normalized to the cellular mitochondrial content and the basal and maximal respiration rates were calculated. The results revealed that stimulation with LPS downregulated both the basal OCR and maximal respiratory capacity in L6 myotubes, which were restored by treatment with L-carnitine (Fig. 6G and H).

Inflammatory cytokines have been reported to impair intramuscular mitochondrial biogenesis. Thus, the present study also examined the effects of both agents on the TNF-α-induced impairment of mitochondrial biogenesis. TNF-α administration exacerbated the impairment of mitochondrial biogenesis, mitochondrial membrane potential and respiratory capacity and promoted the UPS in the rat myotubes in L6 myotubes (Figs. S1C and D, and S2C-F). Notably, L-carnitine significantly prevented this sequence of muscle degradation induced by TNF-α stimulation (Fig. S2A-F). By contrast, rifaximin did not directly affect myotubes. These findings suggest that L-carnitine reversed the LPS- or TNF-α-induced impairment of mitochondrial biogenesis in myotubes.

Taken together, the results demonstrated that rifaximin enhanced the inhibitory effects of L-carnitine supplementation on skeletal muscle atrophy in CDAA-fed cirrhotic rats via multifunctional mechanisms based on the modulation of the gut-liver-muscle axis (Fig. 7).