Lp and Bb suppressed dexamethasone-induced MuRF1 expression in C2C12 cells

First, we screened MuRF1 expression-suppressing probiotics in the lactic acid bacteria collection isolated from human fecal bacteria using dexamethasone-treated C2C12 cells (Fig. 1). Dexamethasone treatment significantly induced MuRF1 expression. However, of tested bifidobacteria and lactobacilli, P61 and P62 significantly suppressed dexamethasone-induced MuRF1 expression. They also suppressed dexamethasone-induced MAFbx/atrogin-1 expression. They furthermore significantly suppressed LPS-induced tumor necrosis factor (TNF)-α and interleukin (IL)-6 expression and NF-κB activation in C2C12 cells (Fig. 1, Supplement Figure S1). When P61 and P62 were mixed (4:1, 1:1, and 1:4), the [1:4] mix most potently suppressed IL-6 expression in LPS-stimulated C2C12 cells (Supplement Figure S2). P61 and P62 were Bifidobacterium bifidum and Lactobacillus paracasei, respectively, based on the results of Gram staining, API kit assay, and whole genome analyses. The genome sequences of P61 and P62 showed the highest phylogenetic similarity to B. bifidum JCM1255 (98.8%) and L. paracasei subsp. paracasei JCM 8130 (98.6%), respectively, using OrthoANI (Supplement Tables S2).

Fig. 1

Effects of Bb and Lp on dexamethasone-induced MuRF1 and Atrogin-1/MAFbx expression and LPS-induced TNF-α and IL-6 expression and NF-κB activation in C2C12 cells. Effects on MuRF1 (a) and Atrogin-1/MAFbx expression (b). Effects on TNF-α (c) and IL-6 expression (d) and NF-κB activation (e). C2C12 cells (1 × 105 cells/mL) were treated with dexamethasone (10 µM) or LPS (100 ng/mL) in the absence or presence of Bb, Lp, or LB (1 × 104 colon-forming units [CFUs] /mL and Cr [5 mM]). Data are indicated as mean ± SD (n = 4). #p < 0.05 vs. NC. *p < 0.05 vs. group treated with vehicle with dexamethasone or LPS

Effects of Lp and Bb on the skeletal muscle strength and mass in aged mice

To confirm whether Lp and Bb could alleviate sarcopenia, we examined their effects on the skeletal muscle mass and strength in aged mice (Fig. 2). The bodyweight of aged mice slowly increased for 8 weeks (Supplement Figure S3). There was no significant difference in food intake and bodyweight gain between aged mice treated with Lp, Bb, LB (their [4:1] mix), creatine (Cr), and vehicle (Vh). First, we examined their effects on the skeletal muscle strength in a treadmill exhaustion test and grip strength tests. The treadmill distance was significantly shorter in aged mice than that in young mice. However, oral administration of LB most potently increased the treadmill distance and running time in aged mice, followed by Lp, Bb, and Cr. In an all-limb test, Lp, Bb, and LB increased grip strength 1.14-, 1.15-. and 1.16-fold, respectively, compared with those treated with vehicle. Lp, Bb, and LB increased grip strength and treadmill distance and running time more potently than Cr.

Fig. 2

Effects of Bb, Lp, and LB on the skeletal muscle strength and mass in aged mice. Effects on treadmill running distance (a) and time (b) and grip strength (c). Effects on the weights of gastrocnemius (GA, d), soleus (SOL, e), quadriceps femoris (QD, f), extensor digitorum longus (EDL, g), tibialis anterior (TA, h), and total muscles (i). Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). #p < 0.05 vs. Yg. *p < 0.05 vs. Vh/Ag

Next, we investigated the effects of Lp, Bb, and LB on muscle weight. The gastrocnemius (GA), soleus (SOL), quadriceps femoris (QD), extensor digitorum longus (EDL), and tibialis anterior (TA) muscle weight of aged mice were lower than those of young mice. However, oral administration of Lp, Bb, or LB significantly increased the weight of GA, SOL, QD, EDL, and TA muscles. Of these, LB most potently increased their weight: it increased the weight of GA, SOL, QD, EDL, and TA to 109.1%, 132.4%, 107.3%, 115.4%, and 106.1% of aged mice.

Effects of Lp and Bb on the AKT signal activation in the skeletal muscle of aged mice

To understand the action mechanism of probiotics on muscle weight and strength, we examined the effects of Lp, Bb, and LB on protein synthesis-involved AKT and mTOR activation in the GA muscle in aged mice (Fig. 3, Supplement Figure S4). The activation of AKT and mTOR was lower in the GA of aged mice than in those of young mice. Oral administration of Lp, Bb, or LB increased AKT and mTOR activation in aged mice.

Fig. 3

Effects of Bb, Lp, and LB on the AKT and mTOR activation in the skeletal muscle of aged mice. Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). #p < 0.05 vs. Yg. *p < 0.05 vs. Vh/Ag

Effects of Lp and Bb on the myogenesis-related gene expression in the skeletal muscle of aged mice

FOXO3a and NF-κB are transcription factors that regulates the transcription of mitogenesis-related genes in the muscle [22, 23]. The activation of these factors suppresses myoblast-differentiating myogenesis genes including MyHC and myogenin (MyoG) and induces muscle-degrading atrogenes including MuRF1 and MAFbx. We also found that FOXO3a and NF-κB were potently activated in the GA muscle of aged mice compared to that of young mice (Fig. 4). Therefore, we examined the effects of Lp, Bb, and LB on the myogenesis gene expression in the GA muscle of aged mice (Fig. 4, Supplement Figure S5). Oral administration of Lp, Bb, or LB significantly suppressed FOXO3a and NF-κB activation in aged mice. Furthermore, they decreased MuRF1, MAFbx, and p16 expression, assessed by immunoblotting. In the immunofluorescence staining, they also decreased MuRF+ and NF-κB+CD11c+ cell populations. Furthermore, they suppressed TNF-α and IL-6 expression. However, Cr did not affect their expression.

Fig. 4

Effects of Bb, Lp, and LB on the myogenesis gene expression in the GA muscle of aged mice. a Effects on MuRF-1, MAFbx, p16, p-FOXO3a, FOXO3a, p-p65, p65, and β-actin expression, assessed by immunoblotting. b Effects on NF-κB+/CD11c+ and MuRF1+ cell population, assessed by immunofluorescence staining. Effects on MuRF-1 (c), MAFbx/Atrogin-1 (d), TNF-α (e), and IL-6 expression (f), assessed by qPCR. Effects on PGC1α, MyHC, and β-action expression (g, immunoblotted), GA muscle cell size (h, H&E-stained), and MyHC-positive cell population (i, stained with immunofluorescence-stained). Effects on MyHC (j), MyHC 2 A (k), MyHC 2X (l), MyHC 2B (m), MyoG (n), PGC1a (o), SIRT1 (p), and mtDNA (q) expression (n), assessed by qPCR. Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). #p < 0.05 vs. Yg. *p < 0.05 vs. Vh/Ag

Oral administration of Lp, Bb, or LB increased MyHC and PGC1α expression, assessed by immunoblotting. They also increased muscle cell size and MyHC-positive cell population. In a qPCR analysis, they decreased MuRF1 and MAFbx expression, while MyHC and MyoG expression increased. Of MyHCs, MyHC2A expression was significantly increased by treatment with Lp or LB, while Bb did not affect its expression. However, Lp, Bb, and LB all increased the expression of MyHC2X, MyH2B, and MyHC more potently than Cr.

Next, we investigated the effects of Lp, Bb, and LB on mitochondrial gene expression-involved PGC1α, SIRT1, and mtDNA expression levels in the GA muscle. PGC1α and SIRT1 expression was lower in aged mice than in young mice. However, oral administration of Lp, Bb, or LB significantly increased PGC1α and SIRT1 expression in aged mice, assessed by qPCR analysis.

Effects of Lp and Bb on the cognitive function in aged mice

Aging induces cognitive decline with systemic inflammation including neuroinflammation [24]. We found that cognitive impairment-like behaviors and TNF-α and IL-6 expression were higher in the hippocampus of aged mice than in those of young mice (Fig. 5, Supplement Figure S6). However, oral administration of Lp, Bb, or LB alleviated cognitive impairment-like behavior in the Y-maze test. They also decreased TNF-α and IL-6 expression, FOXO3a activation, and NF-κB+Iba1+ cell population, while IL-10 and BDNF expression and BDNF+NeuN+ cell population increased in the hippocampus. Ageing increases LPS and corticosterone (cortisol) in the blood and LPS levels in the feces [25]. The bacterial endotoxin induces cognitive impairment with systemic inflammation [26]. We also found that LPS and corticosterone levels were higher in the blood of aged mice than in that of young mice. However, oral administration of Lp, Bb, or LB decreased blood LPS and corticosterone levels in aged mice.

Fig. 5

Effects of Bb, Lp, and LB on the cognitive function in aged mice. a Effects on spontaneous alternation in the Y-maze task. Effects on hippocampal BDNF (b), TNF-α (c), IL-6 (d), IL-1β (e), and IL-10 (f) expression, TNF-α to IL-10 expression ratio (g), and IL-6 to IL-10 expression ratio (h), assessed by qPCR. Effects on BDNF+NeuN+ (i) and NF-κB+Iba1+ cell populations (j). (k) Effects on p-FOXO3a and FOXO3a expression, assessed by immunoblotting. Effects on blood endotoxin (l) and corticosterone levels (m). Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). #p < 0.05 vs. Yg. *p < 0.05 vs. Vh/Ag

Effects of Bb, Lp, and LB on the gut inflammation and microbiota composition in aged mice

Ageing increases systemic inflammation including gut inflammation, which induces gut dysbiosis [27]. We found that the expression of inflammatory markers TNF-α, IL-1β, IL-6, and myeloperoxidase was higher in the colon of aged mice than in that of young mice (Fig. 6). Oral administration of Lp, Bb, or LB decreased TNF-α, IL-1β, and IL-6 expression, TNF-α to IL-10 expression ratio, and NF-κB+CD11c+ cell population in aged mice, while IL-10 expression increased. However, Cr did not affect the expression of inflammatory markers in aged mice.

Fig. 6

Effects of Bb, Lp, and LB on the gut inflammation in aged mice. Effects on TNF-α (a), IL-6 (b), IL-10 (c), IL-1β (d), and myeloperoxidase (MPO) expression (e), and TNF-α to IL-10 expression ratio (f), and IL-6 to IL-10 expression ratio (g), assessed by qPCR. h Effects on NF-κB+CD11c+ cell population. Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). #p < 0.05 vs. Yg. *p < 0.05 vs. Vh/Ag

The gut microbiota composition of aged mice was significantly different to that of young mice (Fig. 7). Although the α-diversity was not significantly different between aged and young mice, the β-diversity significantly different. Oral administration of Lp, Bb, or LB decreased Deferribacteres population in aged mice. They increased Prevotellaceae, Akkermansiaceae, and Bacteroidaceae populations and decreased Odoribacteraceae, Deferribacteraceae, Coriobacteriaceae, and Acholeplasmataceae populations at the family level.

Fig. 7

Effects of Bb, Lp, and LB on the gut microbiota composition in aged mice. Effect on the microbiota composition at the phylum (a) or family levels (b). Effects on OTUs (α-diversity, c) and β-diversity (principal coordinate analysis [PCoA] plot based on BrayCurtis) (d). The relationship between gut microbiota and total muscle weight (e), GA muscle weight (f), TNF-α to IL-10 expression ration (g), IL-6 to IL-10 expression ratio (h), MuRF1 expression (i), or MyHC expression (j) in the GA, assessed by Spearman coefficient test. Bb, Lp, and LB (1 × 109 CFU/mouse/day) and Cr (75 mg/kg) were orally gavaged once a day (6 days in one week) for 8 weeks. Data are indicated as mean ± SD (n = 6). *p < 0.05 vs. Vh/Ag

Of these bacteria, Akkermansiaceae and Bacteroidaceae populations, which were increased in aged mice, were positively correlated with total muscle weight, while Lachnospiraceae, Deferribacteraceae, and Selenomonadaceae populations were negatively correlated. GA muscle weight was positively correlated with the populations of Akkermansiaceae and Bacteroidaceae, while Odoribacteraceae, Deferribacteraceae, and Acholeplasmataceae populations were negatively correlated. MuRF1 expression was positively correlated with Odoribacteraceae, AC160630_f, and Acholeplasmataceae populations, while Akkermansiaceae and Bacteroidaceae populations were negatively correlated.

MyHC expression was positively correlated with Saccharimonas_f, SC160630_f, and Lactobacillaceae populations, while Deferribacteriaceae and Veillonellaceae populations were negatively correlated. TNF-α to IL-10 expression ratio was positively correlated with Streptococcaceae, Coriobacteriaceae, Veillonellaceae, and Muribacteriaceae populations, while Saccharimonas_f and Lactobacillaceae populations were negatively correlated. IL-6 to IL-10 expression ratio was positively correlated with Deferribacteraceae and Acholeplasmataceae populations, while FR888536_f, Clostridiales_us, Bacteroidaceae, and Akkermansiaceae populations were negatively correlated.