The human colorectal cancer cell line Caco-2 (ATCC, Manassas, VA, USA) and murine skeletal myoblasts C2C12 (Riken Bioresource Center, Tsukuba, Japan) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Nissui, Tokyo, Japan) containing 10% heat-inactivated fetal bovine serum (FBS, Capricorn Scientific GmbH, Ebsdorfergrund, Germany) at 37 °C in 5% CO2. C2C12 cells (2.0 × 105 cells/mL in a 6-well dish) were cultured for 48 h and the medium was replaced with DMEM containing 2% horse serum (HS) (Thermo Fisher Scientific, Waltham, MA, USA) to induce differentiation. After another 24 h, the medium was replaced with DMEM containing 2% HS. The medium was replaced every 2 days and used for experiments 10 days after the induction of differentiation.

Substrain OS-1010 of Fructobacillus fructosus, which produces high levels of NMN and NAD+, was provided by Osaka Soda Co. Ltd. (Osaka, Japan). The OS-1010 was heated at 65 °C for 30 min and subsequently spray-dried. This dead bacterial powder contained 2.7 µmol of NMN and 14.2 μmol of NAD per gram of dried bacterial weight. The supernatant of Caco-2 cells treated with F. fructosus OS-1010 was prepared by centrifugation at 3000×g at 10 min.

RNA was prepared from cells using the High Pure RNA Isolation Kit (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s protocol. RT-qPCR was performed using the GoTaq 1-Step RT-PCR System (Promega, Madison, WI, USA) and Thermal Cycler Dice Real Time System TP-800 (Takara, Shiga, Japan). The samples were analyzed in triplicate. The PCR primer sequences used were as follows: mouse β-actin forward primer 5′-TGGCACCCAGCACAATGAA-3′ and reverse primer 5′-CTAAGTCATAGTCCGCCTAGAAGCA-3′; mouse Sirt1 forward primer 5′-GCAGACGTGGTAATGTCCAAACAG-3′ and reverse primer 5′-ACATCTTGGCAGTATTTGTGGTGAA-3′; mouse Tfam forward primer 5′-CATTTATCTATCTGAAAGCTTCC-3′ and reverse primer 5′-CTCTTCCCAAGACTTCATTTC-3′; mouse brain-derived growth factor (Bdnf) forward primer 5′-GTCAAGTTGGGAGCCTGAAATAGTG-3′ and reverse primer 5′-AGGATGCTGGTCCAAGTGGTG-3′; and mouse MyoD forward primer 5′-ATGAGGCCTTCGAGACGCTC-3′ and reverse primer 5′-CAGAGCCTGCAGACCTTCGA-3′. β-actin was used as a housekeeping gene. Samples were normalized and analyzed by the ΔΔCT method (Ishibashi et al. 2023).

Cells were stained with 250 nM MitoTracker Red CMXRos (Thermo Fischer Scientific) at 37 °C for 30 min and subsequently with 200 nM MitoTracker Green FM (Thermo Fischer Scientific) at 37 °C for 30 min. Finally, cells were stained with Hoechst 33,342 (Dojindo, Kumamoto, Japan) at 37 °C for 30 min. Stained cells were analyzed using an IN Cell Analyzer 2200 (Cytiva, Tokyo, Japan) to quantitatively determine the number, area, and activity of mitochondria, and the images were analyzed using the IN Cell Investigator high-content image analysis software (Cytiva) (Inotsuka et al. 2021).

Firstly, Caco-2 cells (1.4 × 105 cells/mL) were cultured in DMEM containing 10% exosome-depleted FBS (System Bioscience, Mountain View, CA, USA) and 100 µg/mL OS-1010 for 24 h. After culture, the MagCapture Exosome Isolation Kit PS ver. 2 (FUJIFILM Wako Pure Chemical Corp.) was used to isolate exosomes from the medium of Caco-2 cells, according to the manufacturer’s instructions. The amount of exosomes used for each experiment was prepared as a protein equivalent and measured using a MicroBCA Protein Assay Kit (Thermo Fisher Scientific Inc.) (Sugihara et al. 2019; Ogawa et al. 2021).

The expression profiles of miRNAs in the exosomes were evaluated by microarray analysis using a 3D-Gene Human miRNA Oligo chip (Toray, Kanagawa, Japan). MiRNA preparation and subsequent operations were performed by Kamakura Techno-Sciences, Inc. (Kanagawa, Japan). After global normalization of the miRNA expression levels, we calculated the ratios of each miRNA for comparison between the control and experimental samples. We then established criteria for the regulated miRNA: (upregulated miRNA) ratio ≥ 1.5-fold (Bolstad et al. 2003). The miRNA target genes were predicted using TargetScan (https://www.targetscan.org/vert_80/, accessed February 1, 2023). We then used the tools and data provided by Database for Annotation, Visualization, and Integrated Discovery (DAVID, http://david.abcc.ncifcrf.gov, accessed on February 20, 2023) to identify significantly enriched pathways (Huang et al. 2009a, 2009b; Inotsuka et al. 2021; Ishibashi et al. 2023).

All experiments were repeated at least three times, and representative data are shown. The results are shown as the mean ± standard error. Multiple comparisons between groups were performed using a one-way ANOVA with Tukey’s post-hoc test. Statistical significance was defined as p < 0.05 when compared to the control (*p < 0.05; **p < 0.01; ***p < 0.001).