Different Impacts of Heat-Killed and Viable Lactiplantibacillus plantarum TWK10 on Exercise Performance, Fatigue, Body Composition, and Gut Microbiota in Humans

Lactiplantibacillus plantarum TWK10, a probiotic strain, has been demonstrated to improve exercise performance, regulate body composition, and ameliorate age-related declines. Here, we performed a comparative analysis of viable and heat-killed TWK10 in the regulation of exercise performance, body composition, and gut microbiota in humans. Healthy adults (n = 53) were randomly divided into three groups: Control, TWK10 (viable TWK10, 3 × 1011 colony forming units/day), and TWK10-hk (heat-killed TWK10, 3 × 1011 cells/day) groups. After six-week administration, both the TWK10 and TWK10-hk groups had significantly improved exercise performance and fatigue-associated features and reduced exercise-induced inflammation, compared with controls. Viable TWK10 significantly promoted improved body composition, by increasing muscle mass proportion and reducing fat mass. Gut microbiota analysis demonstrated significantly increasing trends in the relative abundances of Akkermansiaceae and Prevotellaceae in subjects receiving viable TWK10. Predictive metagenomic profiling revealed that heat-killed TWK10 administration significantly enhanced the signaling pathways involved in amino acid metabolisms, while glutathione metabolism, and ubiquinone and other terpenoid-quinone biosynthesis pathways were enriched by viable TWK10. In conclusion, viable and heat-killed TWK10 had similar effects in improving exercise performance and attenuating exercise-induced inflammatory responses as probiotics and postbiotics, respectively. Viable TWK10 was also highly effective in regulating body composition. The differences in efficacy between viable and heat-killed TWK10 may be due to differential impacts in shaping gut microbiota.


Introduction
Probiotics are defined as "live microorganisms that, when administered in adequate amounts, confer a health benefit on the host" [1]. There is accumulating evidence supporting the health benefits of probiotic consumption, including improvement of gastrointestinal function, modulation of immune and mental functions, and reduction in the risks of genital infection, cardiovascular disease, and other metabolic disorders [2][3][4]. Generally, probiotics improve the health status of hosts by enhancing intestinal barrier integrity, regulating the immune system, improving gut microbial composition, competitive exclusion of pathogens, reducing the intestinal pH, and increasing short-chain fatty acids (SCFAs), mucus, and bacteriocin production [5,6].
Emerging evidence supports the role of probiotics supplementation in improving the health status and exercise performance of elite athletes and the general population [7][8][9][10][11][12]. Data are presented as mean ± SD. Statistical differences among groups were analyzed by one-way ANOVA with Tukey post-hoc test.

Experimental Design
This was a double-blinded and placebo-controlled trial, with a 2-week wash-out period and a 6-week intervention period. Eligible subjects were equally assigned to three groups (8-9 male and 9 female subjects in each group), including Control (placebo, 3 capsules/day), TWK10 (viable TWK10, 3 × 10 11 CFU/day), and TWK10-hk (heat-killed TWK10, 3 × 10 11 cells/day), based on individual exercise capacity, determined from the basal value of maximal oxygen consumption (VO 2max ). Maximal oxygen consumption and exercise performance were evaluated using a treadmill (Pulsar, h/p/cosmos, Nussdorf-Traunstein, Germany) and an auto respiratory analyzer Vmax 29c (Sensor Medics, Yorba Linda, CA, USA). Running speed on the treadmill started at 7.2 km/h and increased by 1.8 km/h every 2 min until volitional fatigue, according to the Bruce protocol [44]. Oxygen consumption was considered maximum when the respiratory exchange ratio (volume ratio of carbon dioxide produced to oxygen consumed; VCO 2 /VO 2 ) was >1.10 and the maximum heart rate was achieved (maximum heart rate = 220 − age). VO 2max was used as a reference to adjust individual appropriate exercise intensity for physiological adaptation (60% VO 2max ) and exhaustive endurance performance (85% VO 2max ) tests. Adjustment of exercise intensity was calculated according to a previously described formula [11]. Subjects were required to avoid any strenuous physical activity for three days before VO 2max assessment and exercise tests. Endurance performance was assessed with a warm-up stage for 5 min, followed by an exercise test on the treadmill at 85% VO 2max workload. Oxygen consumption, heart rate, and Borg's rating of perceived exertion scale were monitored every 5 min during submaximal endurance exercise, to determine achievement of exhaustion. Sustained exercise duration was recorded as the endurance index.
During the 6-week experimental period, subjects were required to take one capsule three times daily after meals, and to maintain their regular lifestyles. Information of caloric intake before and after administration were recorded as reference values. Physiological adaptation effects were determined before and after administration. Fresh fecal samples were collected for gut microbiota analysis and SCFA measurement on the last day of the 2-week wash-out period and after 6 weeks of administration.

Body Composition
Body composition was measured by applying the multi-frequency principle with a bioelectrical impedance analyzer (BIA) on the InBody 770 (In-body, Seoul, Korea). This device takes 30 impedance measurements with frequencies of 1, 5, 50, 260, 500, and 1000 kHz for approximately 60 s. Before testing, age, sex, and height were entered for each subject. Subjects cleaned their hands and feet before contacting the electrodes and were then requested to stand on the center of electrodes and grasp the hand electrodes with their arms held so that there was no contact between the arms and the torso. The position was held for the duration of the test. Subjects fasted for at least 8 h prior to the tests.

DNA Extraction
Freshly collected fecal samples were washed three times with phosphate-buffered saline and centrifuged at 14,000× g for 5 min for extraction of bacterial genomic DNA. Fecal pellets were resuspended in 180 µL TE buffer containing lysozyme (final conc. 10 mg/mL), a suspension of glass beads (300 mg, 0.1 mm in diameter; Biospec, Bartesville, OK, USA) was added, and samples homogenized for 30 s using a FastPrep 24 homogenizer (MP Biomedicals, USA), to ensure complete disruption of cell walls and release of DNA molecules into the solution. Bacterial genomic DNA was then extracted using a Genomic DNA Mini Kit (Geneaid, Taipei, Taiwan), according to the manufacturer's instructions. DNA concentrations were determined by spectrophotometry using a BioDrop instrument (Biochrom, Biochrom Ltd., Cambridge, UK). DNA samples were stored at −20 • C until further processing.

SCFA Levels in Feces
Freshly collected fecal samples were mixed with 70% ethanol solution at a ratio of 1 mg of fecal sample: 10 µL 70% ethanol, and then homogenized with appropriate amounts of glass beads (1.0 mm in diameter; Biospec Products) by vortexing at 3000 rpm for 10 min. Homogenized samples were centrifuged at 14,000× g for 10 min, and the supernatants were collected for fatty acid derivatization, according to a previously described method [53]. Derivatized supernatants were filtered using a 0.22-µm polycarbonate syringe filter (Millipore, St. Charles, MO, USA). SCFAs were separated and quantified using high-performance liquid chromatography (HITACHI, Tokyo, Japan) on a C18 HTec column (NUCLEODUR, Macherey-Nagel, Düren, Germany), with column temperature 40 • C, flow rate 1 mL/min, and detection wavelength 400 nm.

Statistical Analysis
Data are expressed as mean ± SD. Statistical analysis was performed using GraphPad Prism 8.1.1 (GraphPad Software, San Diego, CA). Statistical differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. Differences between before and after administration were analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. Differences in the changes between before and after administration in each subject among the three groups were analyzed by Kruskal-Wallis test with Dunn post-hoc test. Spearman's correlation coefficient was used for analyses of correlations between gut microbial abundances and exercise-associated phenotypic features. P < 0.05 was considered statistically significant.

Both Viable and Heat-Killed TWK10 Improved Exercise Endurance Performance and Physical Adaptation
In this study, we assessed the effects of viable and heat-killed TWK10 on exercise endurance performance, which was evaluated by a time-to-exhaustion test with an 85% VO 2max workload. There were no significant differences in the basic demographic characteristics or baseline exhaustion time values of subjects among the three groups (Control, TWK10, and TWK-hk) ( Table 1). At week 0, there were no significant differences in mean exhaustion time among the three groups; however, after six weeks of administration, the mean exhaustion times in the TWK10 (17.55 ± 3.98 min; P < 0.001) and TWK10-hk (16.72 ± 5.91 min; P < 0.01) groups were significantly higher than those in the Control group (12.23 ± 2.08 min). Within-group comparisons demonstrated that, after 6 weeks of treatment, mean exhaustion times in the TWK10 and TWK10-hk groups were significantly increased by 1.38-fold (P < 0.001) and 1.33-fold (P < 0.001), respectively, whereas no significant difference was observed in the Control group. These results show that both viable and heat-killed TWK10 improved exercise endurance performance, while there was no significant difference in the level of improvement between the TWK10 and TWK10-hk groups ( Figure 1A). baseline exhaustion time values of subjects among the three groups (Control, TWK10, and TWK-hk) ( Table 1). At week 0, there were no significant differences in mean exhaustion time among the three groups; however, after six weeks of administration, the mean exhaustion times in the TWK10 (17.55 ± 3.98 min; P < 0.001) and TWK10-hk (16.72 ± 5.91 min; P < 0.01) groups were significantly higher than those in the Control group (12.23 ± 2.08 min). Withingroup comparisons demonstrated that, after 6 weeks of treatment, mean exhaustion times in the TWK10 and TWK10-hk groups were significantly increased by 1.38-fold (P < 0.001) and 1.33-fold (P < 0.001), respectively, whereas no significant difference was observed in the Control group. These results show that both viable and heat-killed TWK10 improved exercise endurance performance, while there was no significant difference in the level of improvement between the TWK10 and TWK10-hk groups ( Figure 1A).
Microorganisms 2022, 10, x FOR PEER REVIEW 7 of 24 Figure 1. Effects of TWK10 on exercise endurance performance and fatigue-associated blood indicators. (A) Endurance performance was evaluated under 85% VO2max exercise intensity before and after TWK10 administration. Statistical differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. The significance of differences between parameters before and after administration was analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. ** P < 0.01, *** P < 0.001. During fixed intensity and period exercise tests, blood samples were collected for (B) lactate, (C) ammonia, (D) glucose, and (E) CK measurements at the indicated time points after TWK10 administration. Data are presented as mean ± SD. Statistical differences among groups were analyzed by one-way ANOVA with Tukey post-hoc test. Different letters (a, b, c) indicate significant differences among groups at P < 0.05.
Next, we assessed fatigue-related indices and found that there were no significant differences in concentrations of lactate, ammonia, glucose, and creatine kinase in plasma from all three groups during and after exercise challenges just before administration (data not shown), whereas the concentrations of lactate and ammonia in plasma from all the three groups immediately after 6-week of administration were increased during exercise challenge, reaching peak levels after 30 min of exercise stimulation (time point: E30), and Effects of TWK10 on exercise endurance performance and fatigue-associated blood indicators. (A) Endurance performance was evaluated under 85% VO 2max exercise intensity before and after TWK10 administration. Statistical differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. The significance of differences between parameters before and after administration was analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. ** P < 0.01, *** P < 0.001. During fixed intensity and period exercise tests, blood samples were collected for (B) lactate, (C) ammonia, (D) glucose, and (E) CK measurements at the indicated time points after TWK10 administration. Data are presented as mean ± SD. Statistical differences among groups were analyzed by one-way ANOVA with Tukey post-hoc test. Different letters (a, b, c) indicate significant differences among groups at P < 0.05. Next, we assessed fatigue-related indices and found that there were no significant differences in concentrations of lactate, ammonia, glucose, and creatine kinase in plasma from all three groups during and after exercise challenges just before administration (data not shown), whereas the concentrations of lactate and ammonia in plasma from all the three groups immediately after 6-week of administration were increased during exercise challenge, reaching peak levels after 30 min of exercise stimulation (time point: E30), and gradually decreased to the basal level during the recovery phase. In the TWK10 and TWK10hk groups, lactate and ammonia levels in the exercise phase [for lactose (time points: E10, E15, and E30); for ammonia (time points: E15 and E30)] and in the recovery phase [for lactate (time points: R20 and R40); for ammonia (time point: R20)] were significantly lower than those in the Control group ( Figure 1B,C). Lactate concentrations in the TWK10 group in the exercise phase at time points E15 and E30 were significantly lower (P < 0.05) than those in the TWK10-hk group ( Figure 1B). Further, plasma glucose concentrations in the TWK10 group at time points E15, E30, and R20 were significantly higher (P < 0.05) than those in both the TWK10-hk and Control groups. After exercise stimulation, the elevated glucose level was gradually reduced and returned to basal level at resting state ( Figure 1D). For creatine kinase (CK), no significant differences among the three groups were observed during the exercise test ( Figure 1E).

Heat-Kill TWK10 Was More Effective on Reducing Exercise-Induced Inflammatory Response
The neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) serve as markers of systemic inflammatory response in humans [54,55]. Therefore, to evaluate the effect of TWK10 on the exercise-induced systemic inflammatory response, NLR and PLR were examined. CBC was determined from blood samples collected at time point R90 (90 min in recovery state after 30-min exercise stimulation at 60% VO 2max ) and NLR and PLR were calculated. Before administration, no significant differences in PLR and NLR values were observed among the three groups. Following administration of heat-killed TWK10 for 6 weeks, NLR and PLR in the TWK10-hk group were significantly decreased (P < 0.05 and P < 0.01, respectively), whereas, in subjects who received viable TWK10 for 6 weeks, no significant alteration in NLR was observed; however, a significant reduction in PLR l (P < 0.05) was detected ( Figure 2). killed TWK10 for 6 weeks, NLR and PLR in the TWK10-hk group were significantly decreased (P < 0.05 and P < 0.01, respectively), whereas, in subjects who received viable TWK10 for 6 weeks, no significant alteration in NLR was observed; however, a significant reduction in PLR l (P < 0.05) was detected ( Figure 2). Data are presented as mean ± SD. The significance of differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. Differences before and after administration were analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. * P < 0.05, ** P < 0.01.

Viable TWK10 Was More Effective on the Modulation of Muscle Weight and Body Fat Mass
To assess the impact of viable and heat-killed TWK10 on modulation of body composition, muscle weight, body fat mass, and body mass index (BMI) of subjects before and after administration were measured using an InBody 770 Body Composition Analyzer. Considering individual within-group differences, we calculated the changes after six weeks of treatment, compared with before treatment, for each subject, then compared Data are presented as mean ± SD. The significance of differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. Differences before and after administration were analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. * P < 0.05, ** P < 0.01.

Viable TWK10 Was More Effective on the Modulation of Muscle Weight and Body Fat Mass
To assess the impact of viable and heat-killed TWK10 on modulation of body composition, muscle weight, body fat mass, and body mass index (BMI) of subjects before and after administration were measured using an InBody 770 Body Composition Analyzer. Considering individual within-group differences, we calculated the changes after six weeks of treatment, compared with before treatment, for each subject, then compared these changes among the three groups. After six weeks of administration, there were no significant differences in mean muscle weight among the Control, TWK10, and TWK10-hk groups; however, a significant increase (p < 0.001) in muscle weight was observed in the TWK10 group following treatment. Next, we compared differences in muscle weight before and after treatment in each subject among the three groups. A significant increase of mean muscle weight was observed in the TWK10 group after six weeks of treatment, relative to that in the Control group (p < 0.001). Further, the mean fat mass (%) of subjects in the TWK10 group after administration was significantly lower (p < 0.001) than that before administration, and a significant difference (p < 0.001) in the decrease in mean fat mass after six weeks of administration was observed in the TWK10 group relative to the Control group. No significant difference in the change in BMI was observed among the three groups (Table 2). The changes in muscle weight, body fat, and BMI of the indicated groups were calculated as differences before and after administration. Data are presented as mean ± SD. Statistical differences among groups were analyzed by two-way repeated-measures ANOVA with Tukey post-hoc test. Statistical differences between values before and after administration were analyzed by two-way repeated-measures ANOVA with Bonferroni post-hoc test. *** P < 0.001. Significant differences in the changes between before and after administration in each subject among the three groups were analyzed by Kruskal-Wallis test with Dunn post-hoc test. Different letters (a, b, c) indicate significant differences among groups (p < 0.05).

Viable and Heat-Killed TWK10 May Trigger Distinct Gut Microbial Community Changes
To further understand the impact of TWK10 on gut microbial composition, we analyzed the characteristics of bacterial community composition by high-throughput sequencing of the V3-V4 region of the 16S ribosomal RNA gene. To determine how the overall profile of microbial composition was modulated by administration of viable or heat-killed TWK10, alpha-and beta-diversity indices were analyzed. Median Shannon index values for α-diversity in each group ranged from 3.2 to 3.5, and no significant differences were observed between values before and after administration in each group, nor among the three groups.
The median observed ASVs ranged from 130 to 170, and no significant differences were observed between before and after administration in each group, nor among the three groups ( Figure 3A). To evaluate differences in microbial community structures in each group before and after administration among the three groups, we generated NMDS ordination plots of unweighted and weighted Unifrac distances for all six experimental groups. There were no significant differences between before and after administration, nor among the three groups. After six weeks of administration, the β-diversity profiles of fecal microbial composition in subjects showed a strong trend toward significant dif-ference between the TWK10 and TWK10-hk groups (PERMANOVA, P = 0.063), based on NMDS analysis with unweighted UniFrac metric ( Figure 3B). Box plots of unweighted UniFrac distances showed that gut communities in the TWK10 group after administration were significantly different from those of the TWK10-hk group after administration (p < 0.0001) ( Figure 3C). β-diversity profiles in the TWK10 group after administration showed a marginally significant difference (p = 0.072, by unweighted UniFrac distance) to those in the Control group after administration ( Figure 3C), while those in the TWK10-hk group differed significantly from those in the Control group (PERMANOVA, P = 0.036, by NMDS with unweighted UniFrac metric; P = 0.072 by unweighted UniFrac distance) ( Figure 3B,C). β-diversity profiles in the TWK10-hk group differed significantly different from those before administration (p < 0.0001, by weighted UniFrac distance; P = 0.031, by unweighted UniFrac distance), whereas those in the TWK10 group did not differ significantly before and after administration ( Figure 3C).   The overall microbiota structures of subjects in the three groups before and after administration at the phylum level are presented in Figure 3D. The predominant phyla detected in subjects were Firmicutes, Bacteroidota, Actinobacteriota, Proteobacteria, and Verrucomicrobiota. After six weeks of administration, a weak increasing trend in the relative abundance of Verrucomicrobiota in the TWK10 group (P = 0.114), and significant increases in the relative abundances of Proteobacteria in the Control (P = 0.011) and TWK10-hk (P = 0.030) groups were observed, relative to those before administration ( Figure S1A). At the family level, the top 10 predominant bacteria in fecal samples from subjects were Lachnospiraceae, Bacteroidaceae, Ruminococcaceae, Bifidobacteriaceae, Prevotellaceae, Selenomonadaceae, Coriobacteriaceae, Veillonellaceae, Acidaminococcaceae, and Streptococcaceae ( Figure 3E). After 6 weeks of administration in the TWK10 group, significant decreasing trends in the abundance of Ruminococcaceae (P = 0.109) and Eggerthellaceae (P = 0.053), a significant increase in the abundance of Lactobacillaceae (P = 0.039), and significant increasing trends in the relative abundances of Prevotellaceae (P = 0.109) and Akkermansiaceae (P = 0.114) were observed, relative to those before administration. In the TWK10-hk group after 6 weeks of administration, a significant increasing trend (P = 0.072) in the abundance of Enterobacteriaceae, a significant decrease (P = 0.030) in the abundance of Peptostreptococcaceae, a significant decreasing trend (P = 0.107) in the abundance of Lachnospireaceae, and an increasing trend (P = 0.148) in the abundance of Atopobiaceae were observed as compared with before administration ( Figure S1B). In the TWK10 group after six weeks of administration, genus-level analysis revealed a significant decrease in Lachnospira abundance (P = 0.022), a decreasing trend in Faecalibacterium abundance (P = 0.077), and a significant increasing trend in Akkermansia abundance (P = 0.114), compared with before administration. Meanwhile, after six weeks of administration in the TWK10-hk group, we detected significant increases in the abundances of Lactococcus (P = 0.046) and Escherichia-Shigella (P = 0.038), decreasing trends in the abundances of Roseburia (P = 0.110) and Lachnospira (P = 0.064), and a marginal increase in the abundance of Oscillibacter (P = 0.184), relative to before administration ( Figure S2).

Viable and Heat-Killed TWK10 Showed Different Imapcts on Gut Microbial Co-Occurrence Networks
As intra-microbiota bacterial interactions play roles in shaping the gut microbiota community [56], we further investigated alterations in bacteria-bacteria interactions after viable or heat-killed TWK10 administration using bacterial co-occurrence networks analysis. Among the ASVs obtained in the study, 16S rRNA read counts were >50 in 233 and 258 ASVs before administration in the TWK10 and TWK10-hk groups, respectively. SparCC correlation coefficients >|0.6| and P < 0.05 were considered indicative of a connection between bacterial groups. Gut bacteria co-occurred more frequently before administration than after administration in both the TWK10 and TWK10-hk groups. Specifically, in the TWK10 and TWK10-hk groups, 37 and 30 nodes, respectively, were observed before administration, whereas 28 and 25 respective nodes were observed after administration. These decreases reflected dramatic changes in bacterial network structures in response to the administration of viable TWK10 ( Figure 4A,B). In the TWK10 group before administration, Lachnospiraceae was positively correlated with Butyricicoccaceae, Carnobacteriaceae, and Sutterellaceae, whereas it was negatively correlated with Anaerovoracaceae and Christensenellaceae; Ruminococcaceae was positively correlated with Bacteroidaceae, Butyricicoccaceae, Carnobacteriaceae, and Desulfovibrionaceae; and Bifidobacteriaceae was positively correlated with Veillonellaceae ( Figure 4A). Meanwhile, in the TWK10 group after administration, Lachnospiraceae was positively correlated with Actinomycetaceae, Butyricicoccaceae, Carnobacteriaceae, and Streptococcaceae; and Bifidobacteriaceae was positive correlated with Bacteroidaceae and Enterobacteriaceae, and negatively correlated with Eubacterium coprostanoligenes group, Marinifilaceae, and Oscillospiraceae. In addition, a firm network among the families, Anaerovoracaceae, Christensenellaceae, Coriobacteriales incertae sedis, E. coprostanoligenes group, Marinifilaceae, Oscillospiraceae, and Rikenellaceae was observed ( Figure 4B). In the TWK10-hk group before administration, Lachnospiraceae was positively correlated with Butyricicoccaceae, whereas it was negatively correlated with Christensenellaceae, Coriobacteriales incertae sedis, and Oxalobacteraceae; a positively correlated network among Carnobacteriaceae, Eggerthellaceae, and Streptococcaceae was observed; and Prevotellaceae was negatively correlated with Bacteroidaceae, Bifidobacteriaceae, and Monoglobaceae ( Figure 4C). However, heat-killed TWK10 administration altered the topology of the co-occurrence network, with the number of connections among gut bacteria reduced. Specifically, Lachnospiraceae was positively correlated with Eggerthellaceae and Sutterellaceae, and negatively correlated with Christensenellaceae; Ruminococcaceae was positively correlated with Bifidobacteriaceae and Butyricicoccaceae, and negatively correlated with Erysipelotrichaceae; and Anaerovoracaceae, Desulfovibrionaceae, Marinifilaceae, and Rikenellaceae formed a positively correlated network ( Figure 4D).

Viable and Heat-Killed TWK10 Showed Differenet Impacts on Predicted Gut Microbial Community Functional Profiles
Functional profiles of bacterial communities of the three groups (Control, TWK10, and TWK10-hk) were predicted using PICRUSt2. LEfSe analysis was then performed to explore level 3 KEGG pathways with significant differences in abundance before and after administration. Butanoate metabolism, glutathione metabolism, and ubiquinone and other terpenoid-quinone biosynthesis pathways were over-represented (α = 0.1, LDA score > 2.0) following TWK10 administration ( Figure 5A). In the TWK10-hk group, RNA polymerase, pentose phosphate pathway, sulfur relay system, galactose metabolism, beta_lactam resistance, and starch and sucrose metabolism were over-represented in the microbiota of subjects before administration, whereas phenylalanine metabolism, tryptophan metabolism, butanoate metabolism, and taurine and hypo-taurine metabolism were significantly en-riched (α = 0.1, LDA score > 2.0) following administration ( Figure 5B). These results demonstrate that viable and heat-killed TWK10 induced different alterations in the overall predicted functional features of gut microbiota.
Carnobacteriaceae, Eggerthellaceae, and Streptococcaceae was observed; and Prevotellaceae was negatively correlated with Bacteroidaceae, Bifidobacteriaceae, and Monoglobaceae ( Figure 4C). However, heat-killed TWK10 administration altered the topology of the co-occurrence network, with the number of connections among gut bacteria reduced. Specifically, Lachnospiraceae was positively correlated with Eggerthellaceae and Sutterellaceae, and negatively correlated with Christensenellaceae; Ruminococcaceae was positively correlated with Bifidobacteriaceae and Butyricicoccaceae, and negatively correlated with Erysipelotrichaceae; and Anaerovoracaceae, Desulfovibrionaceae, Marinifilaceae, and Rikenellaceae formed a positively correlated network ( Figure 4D).  . Co-occurrence network analysis of the gut microbiota. Bacterial networks were generated using SparCC correlation coefficients, based on relative abundances at the family level: (A) In the TWK10 group before administration (TWK10_before), there were 37 nodes, and correlation coefficients ranged from |0.60| to |0.92|; (B) In the TWK10 group after administration (TWK10_after), there were 28 nodes, and correlation coefficients ranged from |0.60| to |0.74|; (C) In the TWK10-hk group before administration (TWK10-hk_before), there were 30 nodes, and the correlation coefficients ranged from |0.61| to |1.00|; (D) In the TWK10-hk group after administration (TWK10-hk_after), there were 25 nodes, and correlation coefficients range from |0.60| to |0.85|. Nodes represent bacteria families; grey and orange color edges represent negative and positive correlation coefficients, respectively. The size and the degree of the green color of nodes in the network represent the relative abundance of each taxon in each group. polymerase, pentose phosphate pathway, sulfur relay system, galactose metabolism, beta_lactam resistance, and starch and sucrose metabolism were over-represented in the microbiota of subjects before administration, whereas phenylalanine metabolism, tryptophan metabolism, butanoate metabolism, and taurine and hypo-taurine metabolism were significantly enriched (α = 0.1, LDA score > 2.0) following administration ( Figure 5B). These results demonstrate that viable and heat-killed TWK10 induced different alterations in the overall predicted functional features of gut microbiota.

Both Viable and Heat-Killed TWK10 Increased Gut SCFA Levels
Acetate, propionate, and butyrate are the main SCFAs metabolized by gut microorganisms. Therefore, we collected fecal samples from subjects before and after administration to examine whether TWK10 could influence SCFA production. As shown in Table 3, subjects receiving both viable and heat-killed TWK10 had significantly higher acetate concentrations in feces relative to before administration (P < 0.05). Further, a significantly increasing trend in propionate was detected in feces of subjects treated with heat-killed TWK10 (P = 0.0857), and in butyrate in feces of subjects treated with viable TWK10 (P = 0.0744). In the Control group, no significant alterations of any of the three tested SCFAs were observed after administration. Data are presented as mean ± SD. The significance of differences among groups were analyzed by Kruskal-Wallis test with Dunn's post-hoc test. Differences between values before and after administration were analyzed by Mann-Whitney U test. * P < 0.05, § P = 0.0857, # P = 0.0744.

Correlation between Gut Microbial Composition and TWK10-Mediated Health Benefits
Correlations between the relative abundances of gut bacterial families and TWK10mediated phenotypic features related to host health benefits were assessed by Spearman's correlation analysis. For exercise endurance performance, the effects on exhaustion time mediated by administration of heat-killed TWK10 were significantly and positively correlated with the Veillonellaceae population, and significantly negatively correlated with E. coprostanoligenes group, Erysipelatoclostridiaceae, and Lachnospiraceae populations. In subjects receiving viable TWK10, exhaustion time was significantly positively correlated with the Coriobacteriaceae population. The correlation coefficients between Veillonellaceae and the exhaustion time in subjects administered viable and heat-killed TWK10 were strengthened from −0.35 to 0.06, and from 0.51 to 0.65, respectively, following administration. Changes in body fat mass (%) mediated by heat-killed TWK10 administration were significantly positively correlated with the presence of Erysipelatoclostridiaceae and Erysipelotrichaceae, and significantly negatively correlated with the Veillonellaceae population. In subjects administered viable TWK10, the abundances of Bacteroidaceae, Oscillospiraceae, Rikenellaceae, and Ruminococcaceae were negatively correlated with muscle weight. In pro-inflammatory responses detected following the administration of heat-killed TWK10, a strong positive correlation was observed between the abundance of Enterobacteriaceae and PLR, whereas Akkermansiaceae abundance was strongly negatively correlated with NLR ( Figure 6).

Discussion
Viable TWK10 has been previously elucidated to have potential probiotic effects in enhancing exercise performance, increasing muscle mass and strength, improving body conformation, and ameliorating age-associated cognitive decline and impairments in mice and humans [11,12,[39][40][41]; however, the impacts of heat-killed TWK10 on health promotion have yet to be confirmed. Therefore, in the current study, we applied the treadmill method to evaluate exercise performance before and after administration of viable or heat-

Discussion
Viable TWK10 has been previously elucidated to have potential probiotic effects in enhancing exercise performance, increasing muscle mass and strength, improving body conformation, and ameliorating age-associated cognitive decline and impairments in mice and humans [11,12,[39][40][41]; however, the impacts of heat-killed TWK10 on health promotion have yet to be confirmed. Therefore, in the current study, we applied the treadmill method to evaluate exercise performance before and after administration of viable or heat-killed TWK10 by assessing changes in individual VO 2max . Six weeks of administration of both viable and heat-killed TWK10 significantly improved exercise performance, consistent with our previous findings in humans, where administration of viable TWK10 significantly increased exercise performance relative to the placebo group [11,12].
Blood glucose and glycogen stored in muscle are established as the major energy sources during exercise. Exercise can spike blood glucose, most commonly via the release of the stress hormone, adrenaline. Key actions of adrenaline include increasing heart rate and blood pressure, expanding the lung air passages, enlarging the pupils in the eye, redistributing blood to the muscles, and altering metabolism in response to acute stress [57]. Generally, exercise induces an increase in gluconeogenesis and raises blood glucose, to provide fuel for muscles in response to exercise demand. After exercising, the body adjusts the blood glucose by secreting insulin, to lower the excess sugar in circulation that is no longer needed by the muscles. In this study, we observed a significant elevation in circulating glucose during exercise in subjects who received viable TWK10, reflected in an improvement in exercise endurance performance ( Figure 1D). These results are consistent with our previous findings [11,12]. In contrast, no significant elevation of plasma glucose was observed in subjects who received the heat-killed form of TWK10; therefore, we speculate that viable TWK10 can accelerate cross-talk between the gut and brain, to efficiently trigger adrenaline release, boost energy supplies, and consequently prolong exercise duration. However, further investigations are needed to elucidate the role of TWK10 in adrenaline signaling and improved exercise performance.
In addition to the exercise-induced boost in blood glucose, significantly reduced lactate production in circulation was also observed in subjects who received viable TWK10 ( Figure 1B). Muscle contraction depends on the breakdown of adenosine triphosphate (ATP) and the concomitant release of free energy. Anaerobic glycolysis is the major metabolic pathway used when oxygen supply is limited during exercise, such as high-intensity, sustained, isometric muscle activity. Glycolysis produces pyruvate from glucose, which is then reduced to lactate-by-lactate dehydrogenase, without oxygen consumption [58]. In groups administered both viable and heat-killed TWK10, exercise endurance (time to exhaustion) was improved and less plasma lactate was produced during exercise, indicating that subjects receiving TWK10 were less likely to obtain energy for exercise demand through anaerobic glycolysis. Sufficient oxygen supply in exercising muscle promotes aerobic glucose breakdown, resulting in the conversion of pyruvate to acetyl-CoA, which is subsequently metabolized in the TCA cycle to produce ATP for exercise demand. Enhanced TCA cycle activity results in reduced lactate production. In KEGG analysis, we detected an increasing trend in TCA cycle enrichment (P = 0.069) in subjects receiving heat-killed TWK10, relative to the baseline value (data not shown). Another possible reason for the observation of reduced levels of lactate during exercise is an alteration of energy supply. Ketosis decreases muscle glycolysis and plasma lactate levels, while increasing intramuscular triacylglycerol oxidation during exercise, providing an alternative substrate for oxidative phosphorylation [59]. Further investigations are necessary to understand the effects of both viable and heat-killed TWK10 on energy metabolism during exercise.
There is good evidence that inactivated microbial cells exhibit health-promoting effects through immune system modulation, prevention of pathogenic infection, and reduction in oxidative stress [60][61][62]. Heat-killed Lactobacillus brevis SBC8803 downregulates the expression of pro-inflammatory cytokines and enhances intestinal barrier function under oxidative stress [63]. Likewise, heat-killed Bacillus coagulans GBI-30 promoted immune responses and modulated inflammatory cytokine expression in cell-base assays [64]. It is established that both viable or heat-killed bacteria can exhibit anti-inflammatory effects; however, which form of bacteria is more effective depends on the strain. In this study, subjects receiving both viable and heat-killed TWK10 showed significant reduction in PLR post-exercise challenge, whereas significantly reduced NLR was observed only in subjects who received heat-killed TWK10 (Figure 2), indicating that heat-killed TWK10 has the potential to reduce damage or increase recovery rate after exercise. Exerciseinduced production of muscle damage indicators and inflammatory biomarkers can lead to a temporary reduction in muscular force [65] and decreased physical performance [66]. These findings suggested that the improvement of exercise performance mediated by viable and heat-killed TWK10 may be achieved in a different manner. Therefore, further investigations are needed to further elucidate the differences in underlying mechanisms.
Total daily energy expenditure is determined by basal metabolic rate (BMR), foodinduced thermogenesis, and energy required for physical activity. Generally, the BMR accounts for 65-75% of total energy expenditure and is considered to be proportional to fat-free mass [67]. As Huang et al. previously reported [68], viable TWK10 upregulates BMR in gnotobiotic mice, accompanied by a reduction in fat mass, without altering dietary intake. In addition, the proteomic analysis demonstrated that peroxisomal acyl-coenzyme A oxidase 2 and very long-chain acyl-CoA synthetase were significantly upregulated in the liver of mice administered viable TWK10, indicating that lipid metabolism is enhanced by TWK10 [69]. In this study, viable TWK10 significantly promoted the development of body composition toward an improved configuration by increasing muscle mass proportion and reducing fat mass in humans (Table 2), consistent with our previous findings [12]. These results suggest that the increase in muscle mass mediated by viable TWK10 may contribute to the upregulation of energy expenditure while promoting lipid metabolism and further reducing body fat accumulation. In addition, there is evidence that gut microbiota are directly involved in regulating energy metabolism; thus, changes in the composition and abundance of gut bacteria may modify energy consumption and expenditure [70]. To better understand the mechanisms underlying the regulation of body composition by viable and heat-killed TWK10, further investigation of energy metabolism and gut microbiota is needed.
Although the relationships between diet, gut microbiota, host immunity, and host metabolism are becoming more evident [71][72][73], that between microbiota and exercise has not been fully investigated. Many studies have shown that the consumption of probiotics has the potential to positively modify gut microbiota community structures, which may be important to increase exercise performance in physical activity practitioners and athletes [74][75][76][77]; however, the potential mechanisms by which probiotic strains may modulate gut microbiota profiles to improve exercise performance remain unclear. Ecological diversity of microbiota is important for promoting health stability and exercise performance. Microbiota alpha-diversity has been linked to human health, with loss of diversity associated with several conditions, including autism, gastrointestinal diseases, and obesityassociated inflammatory characteristics [78]. In this study, no differences in alpha-diversity of the gut microbiota were observed before and after six weeks of administration in any of the three groups; however, the number of gut microbial taxa (mean observed ASVs) tended to be higher in the TWK10-hk group after administration than that in the Control group ( Figure 3A), strongly suggesting that the significant difference (P = 0.036) in beta-diversity, based on NMDS with unweighted UniFrac metric, in the TWK10-hk group after administration, relative to that in the Control group, was due to an increase in ASVs in the TWK10-hk group. Together with detection of a highly significant (P < 0.0001, by unweighted UniFrac) difference in beta-diversity distance in the TWK10-hk group after administration relative to the TWK10 group, these findings indicate that six-week administration of heat-killed TWK10 has a stronger effect in rearranging microbial community structures than viable TWK10 ( Figure 3B,C).
At the bacterial phylum level, no obvious alterations in predominant gut microbiota, except for Verrucomicrobiota and Proteobacteria, were detected following six-week administration of viable or heat-killed TWK10. Relative abundance of the phylum, Verrucomicrobiota, a mucin-degrading bacteria that resides in the mucus layer and represents 1%-4% of fecal microbiota in healthy humans [79], was weakly significantly (P = 0.114) increased by administration of viable TWK10. Further, the administration of heat-killed TWK10 led to a significant (P = 0.030) increase in the relative abundance of the phylum, Proteobacteria, which mainly comprised the family Enterobacteriaceae. Nevertheless, the gut microbiota in both the viable and heat-killed TWK10 administration groups were more complex and apparently divergent from the Control group at the bacterial family and genus levels, representing differences in microbial profiles and well-characterized structures. Verrucomicrobiota was mostly comprised of the single genus, Akkermansia (family Akkermansiaceae), which showed a weak significant (P = 0.114) increase in the TWK10 group after administration (Figures S1 and S2). Akkermansia muciniphila is a mucin-degrading bacteria and its abundance is inversely correlated with obesity and associated metabolic disorders [80,81]. The proportions of the genus, Akkermansia, in athletes are significantly higher in those with low BMI [76], which is generally considered a healthier metabolic profile [82]; however, although we detected a significant increase in the relative abundance of Akkermansia, no significant reduction in BMI was observed in the TWK10 group after administration. Further, we detected a weakly significant (P = 0.109) increase of Prevotellaceae, predominantly comprising the genus, Prevotella, following the administration of viable TWK10. Prevotella has been identified as more universal in populations with plant-rich diets, abundant in carbohydrates and fiber [83], and Prevotella and Akkermansia produce acetate via the Wood-Ljungdahl and acetyl-CoA pathways [84,85].
Based on gut microbial co-occurrence network analysis, we found that the number of bacteria-bacteria interactions decreased in response to the administration of viable and heat-killed TWK10 (Figure 4). Following the administration of viable TWK10, three butyrate-producing bacterial families, Lachnospiraceae, Butyricicoccaceae, and Actinomycetaceae, were strongly and positively correlated with one another. In addition, the families, Christensenellaceae, Eubacterium coprostanoligenes, and Oscillospiraceae, of the phylum Firmicutes, and Coriobacteriales incertae sedis, of the phylum Actinobacteria, which are regarded as potentially beneficial bacteria [86][87][88][89], showed a strong positive correlation with one another. Meanwhile, following administration of heat-killed TWK10, acetate-and butyrate-producing bacteria, such as Bifidobacteriaceae (acetate), Butyricicoccaceae (butyrate), and Ruminococcaceae (butyrate), showed strong positive correlations with one another. Further, two butyrate-producing and potentially probiotic families, Lachnospiraceae and Eggerthellaceae, were positively correlated with one another. These findings demonstrate differences in the influence on the co-occurrence of gut microbial components between viable and heat-killed TWK10; however, both viable and heat-killed TWK10 were able to rearrange gut microbial community structures. In addition, it has been confirmed that the administration of probiotics increased the production of SCFAs by modulation of the gut microbiota [41,90,91]. Alterations in gut microbiota can stimulate differential production of SCFAs (e.g., butyrate and acetate), which play important roles in the maintenance of gut and metabolic health [92,93].
Spearman's correlation analysis revealed that exercise-associated host phenotypic features were positively or negatively correlated with specific microbiota, and that those correlations differed between viable and heat-killed TWK10; that is, exhaustion time was positively correlated with Veillonellaceae in subjects administered heat-killed TWK10, and with Coriobacteriaceae in subjects administered viable TWK10. Relative abundance of Veillonellaceae, which produces propionate from lactate [94], was increased in athletes after running a marathon, and oral administration of the Veillonella strain significantly increased exhaustion time in mice [77]. In the current study, trends toward strengthened correlation between Veillonellaceae and exhaustion time were also observed in subjects administered both viable and heat-killed TWK10, relative to basal state. Body fat mass was positively correlated with the butyrate-producing bacteria, Erysipelatoclostridiaceae and Erysipelotrichaceae, in subjects administered heat-killed TWK10. Further, the significant reduction in NLR values in subjects administered heat-killed TWK10 was negatively correlated with a trend toward an increased relative abundance of Akkermansiaceae, which is regarded as an intestinal mucin-degrader [95] and effective in reducing inflammation [96]. Accordingly, to clarify the differences in effects on exercise performance between viable and heat-killed TWK10, it will be necessary to comprehensively analyze more samples to determine the effects of each TWK10 state (viable and heat-killed) on correlations between exercise-associated host phenotypic features and microbial structures.
Various studies have investigated the effects of the gut microbiome on exercise performance [77,97,98]. Exercise performance was improved in mice adapted with individual bacterial taxa, relative to their germ-free counterparts, indicating that increased microbial diversity has a beneficial effect on exercise. Recent studies have also shown that gut microbiota may be critical for skeletal muscle metabolism and host function [21,99]. Additionally, natural reseeding of the gut microbiota or infusion of acetate reversed the loss of endurance capacity and muscle contractile function in antibiotic-treated mice [21]. Further, the probiotics, Streptococcus thermophilus FP4 and Bifidobacterium breve BR03, attenuated performance decrements and muscle tension in the days following muscle-damaging exercise [100]. Nevertheless, while these studies showed that probiotics can modulate gut microbiota and improve exercise capacity, their effects on performance remain unclear.
To reveal the putative mechanisms underlying the probiotic effects of TWK10 in improving exercise performance, functional profiling was performed using PICRUSt2, based on gut microbial taxa. LEfSe analysis, based on differential functional abundances identified through KEGG pathway mapping, revealed that several metabolic pathways were significantly upregulated in subjects administered with viable or heat-killed TWK10 ( Figure 5). Butanoate metabolism was elevated in groups treated with both viable and heat-killed TWK10. Increased butanoate metabolism results in butyric acid formation, which is a major SCFA that reduces inflammation and promotes gut health [101][102][103][104]. The upregulation of pathways, such as glutathione metabolism, ubiquinone and other forms of terpenoid-quinone biosynthesis, suggests that viable TWK10 can increase antioxidant defense and detoxification reactions, which protect the body from oxidative stress and have benefits on exercise. Increased metabolism of the non-enzymatic antioxidants, glutathione and ubiquinone (coenzyme Q10), leads to increased resistance to exercise-induced oxidative challenges [105,106]. In addition, glutathione supplementation induces aerobic metabolism and improves an acidic environment in skeletal muscle, which in turn prevents exercise-induced fatigue [107]. Meanwhile, KEGG pathway analysis of the heat-killed TWK10-administered group indicated the elevation of different metabolic pathways, including phenylalanine metabolism, tryptophan metabolism, and taurine and hypo-taurine metabolism. Phenylalanine is an essential gluconeogenic amino acid as well as a gluconeogenic and ketogenic, which becomes trans-aminated into different intermediates of the gluconeogenic pathway [108]. Phenylalanine supplementation can increase plasma glucagon concentrations during exercise. Glucagon is a key hormone involved in fat catabolism during exercise, suggesting that increased phenylalanine metabolism can stimulate fat oxidation through glucagon secretion [109]. Tryptophan is an essential amino acid which is metabolized through the kynurenine pathway to generate a number of bioactive substances, thereby modulating health and disease states, ranging from intestinal conditions to inflammation and cancer progression [110,111]. Kynurenine and its metabolites can mediate the effects of exercise, mood, and neuronal excitability and, ultimately, communicate with microbiota. Taurine is an amino acid that can regulate the gut micro-ecology and has the potential to enhance gut-resistance to pathogenic bacteria [112]. The results obtained in our study demonstrate that administration of viable and heat-killed TWK10 can influence the gut microbiota, which has emerged as an important driving force in modulating metabolic activities, although the magnitude, features, and strength of viable and heat-killed TWK10 were inconsistent.
The gut microbiota regulates multiple functions related to host physical health, and mental health through the gut-brain axis which is the two-way communication pathway between the enteric and central nervous systems [113]. Through these pathways, produced SCFAs, bile acids, and tryptophan by the gut microbiota interact with enteroendocrine cells and activate the vagus nerve which serves a critical role in communication between the gut microbiota and the brain [114,115]. Metabolomics is an emerging technology that could simultaneously quantify multiple types of small molecules, such as amino acids, fatty acids, carbohydrates, or other products of cellular metabolic functions [116]. Metabolome response closely associates with physical activity and biological functions [117,118]. Therefore, to determine the metabolome associations with TWK10-mediated phenotypic features could provide greater insights into the understanding of the underlying the different mechanisms of viable and heat-killed TWK10.

Conclusions
In the present study, we present, for the first time, evidence of the effects of TWK10 from a clinical trial using both viable and heat-killed TWK10, demonstrating that both exert sufficient probiotic and postbiotic effects, respectively, in improving exercise performance and fatigue-associated features, and mitigate responses to exercise-induced inflammation. Further, we detected differences in the regulation of body composition and anti-inflammation responses between viable and heat-killed TWK10. These differences may be due to differential impacts in shaping gut microbiota. Further studies are needed to clarify the differences in the efficacy of the two states of TWK10 in promoting exercise performance in clinical trials with more subjects.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10112181/s1, Figure S1: Relative abundances of major gut microbiota among groups (at the phylum and family levels); Figure S2: Relative abundances of major gut microbiota among groups (at the genus level). Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.