Limits...
Evaluation and characterization of bacterial metabolic dynamics with a novel profiling technique, real-time metabolotyping.

Fukuda S, Nakanishi Y, Chikayama E, Ohno H, Hino T, Kikuchi J - PLoS ONE (2009)

Bottom Line: Notably, intestinal bacteria maintain homeostatic balance in mammals via multiple dynamic biochemical reactions to produce several metabolites from undigested food, and those metabolites exert various effects on mammalian cells in a time-dependent manner.We identified trans11, cis13 conjugated linoleic acid as the intermediate of linolenic acid hydrogenation by B. fibrisolvens, based on the results of (13)C-labeling RT-MT experiments.RT-MT is useful for the characterization of beneficial bacterium that shows potential for use as probiotic by producing bioactive compounds.

View Article: PubMed Central - PubMed

Affiliation: RIKEN Research Center for Allergy and Immunology, Yokohama, Japan.

ABSTRACT

Background: Environmental processes in ecosystems are dynamically altered by several metabolic responses in microorganisms, including intracellular sensing and pumping, battle for survival, and supply of or competition for nutrients. Notably, intestinal bacteria maintain homeostatic balance in mammals via multiple dynamic biochemical reactions to produce several metabolites from undigested food, and those metabolites exert various effects on mammalian cells in a time-dependent manner. We have established a method for the analysis of bacterial metabolic dynamics in real time and used it in combination with statistical NMR procedures.

Methodology/principal findings: We developed a novel method called real-time metabolotyping (RT-MT), which performs sequential (1)H-NMR profiling and two-dimensional (2D) (1)H, (13)C-HSQC (heteronuclear single quantum coherence) profiling during bacterial growth in an NMR tube. The profiles were evaluated with such statistical methods as Z-score analysis, principal components analysis, and time series of statistical TOtal Correlation SpectroScopY (TOCSY). In addition, using 2D (1)H, (13)C-HSQC with the stable isotope labeling technique, we observed the metabolic kinetics of specific biochemical reactions based on time-dependent 2D kinetic profiles. Using these methods, we clarified the pathway for linolenic acid hydrogenation by a gastrointestinal bacterium, Butyrivibrio fibrisolvens. We identified trans11, cis13 conjugated linoleic acid as the intermediate of linolenic acid hydrogenation by B. fibrisolvens, based on the results of (13)C-labeling RT-MT experiments. In addition, we showed that the biohydrogenation of polyunsaturated fatty acids serves as a defense mechanism against their toxic effects.

Conclusions: RT-MT is useful for the characterization of beneficial bacterium that shows potential for use as probiotic by producing bioactive compounds.

Show MeSH

Related in: MedlinePlus

Principal components analysis (A) and loading plot analysis (B) of three strains of B. fibrisolvens.MDT-10 (red), MDT-5 (blue), and A38 (purple) are shown. Data under conditions of LA addition (open squares), LNA addition (close squares), or no addition (close circles) were calculated. Data are representative of three independent experiments. (A) Hourly 1H-NMR profiling data of three strains of B. fibrisolvens incubated with or without LA or LNA were analyzed. Contributions of PC1 and PC2 were 77.0% and 11.6%, respectively. All 0 h samples were assembled at the left side of Figure 3A, indicating that metabolic conditions before bacterial growth were similar. However, after 10 h, the samples were located in the right side of Figure 3A, suggesting that PC1 means bacterial growth and PC2 means metabolic differences among B. fibrisolvens strains. (B) Blue peaks and words indicate PC1 contribution and those in red indicate PC2 contribution. The 4.2–5.2 ppm region was omitted to eliminate the effects of imperfect water suppression.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2654759&req=5

pone-0004893-g003: Principal components analysis (A) and loading plot analysis (B) of three strains of B. fibrisolvens.MDT-10 (red), MDT-5 (blue), and A38 (purple) are shown. Data under conditions of LA addition (open squares), LNA addition (close squares), or no addition (close circles) were calculated. Data are representative of three independent experiments. (A) Hourly 1H-NMR profiling data of three strains of B. fibrisolvens incubated with or without LA or LNA were analyzed. Contributions of PC1 and PC2 were 77.0% and 11.6%, respectively. All 0 h samples were assembled at the left side of Figure 3A, indicating that metabolic conditions before bacterial growth were similar. However, after 10 h, the samples were located in the right side of Figure 3A, suggesting that PC1 means bacterial growth and PC2 means metabolic differences among B. fibrisolvens strains. (B) Blue peaks and words indicate PC1 contribution and those in red indicate PC2 contribution. The 4.2–5.2 ppm region was omitted to eliminate the effects of imperfect water suppression.

Mentions: Different from Z-score analysis, it is possible to detect characteristic metabolites by PCA. Interestingly, the differences in PCA profiles were more remarkable among the different strains than among the different medium conditions (Fig. 3-A). The PCA profiles of the three strains revealed that the metabolic characteristics of MDT-10 and MDT-5 were similar, while those of A38 were quite different. We calculated the first two components that contained principal components 1 and 2 contributing to 77.0% and 11.6% of the data variance, respectively. The strains were well separated primarily in PC2 and appeared to form two groups of A38 and of MDT-10 and MDT-5. In contrast, PC1 was dominated by the time course, which is common to the three strains. The spectral regions that contributed most to the difference between the two groups are shown in the corresponding loading plot (Fig. 3-B). Metabolites that exerted the greatest influence on this separation were lactate and glucose in PC1 and butyrate, acetate, and formate in PC2 (Fig. 3-B). As the three strains grew, they consumed glucose and produced lactate; however, butyrate, acetate, and formate production was different between the two groups. It is well known that B. fibrisolvens strains could be classified into Type I (high butyrate and formate production) or Type II (high lactate production) based on their fermentation products but not their 16S rDNA sequences [63]–[65]. It has been reported that these differences may be the basis for genetic differences related to the butyrate production pathway [65]. Therefore, this analytical method could be used to characterize bacterial fermentation patterns derived from functional genetic information.


Evaluation and characterization of bacterial metabolic dynamics with a novel profiling technique, real-time metabolotyping.

Fukuda S, Nakanishi Y, Chikayama E, Ohno H, Hino T, Kikuchi J - PLoS ONE (2009)

Principal components analysis (A) and loading plot analysis (B) of three strains of B. fibrisolvens.MDT-10 (red), MDT-5 (blue), and A38 (purple) are shown. Data under conditions of LA addition (open squares), LNA addition (close squares), or no addition (close circles) were calculated. Data are representative of three independent experiments. (A) Hourly 1H-NMR profiling data of three strains of B. fibrisolvens incubated with or without LA or LNA were analyzed. Contributions of PC1 and PC2 were 77.0% and 11.6%, respectively. All 0 h samples were assembled at the left side of Figure 3A, indicating that metabolic conditions before bacterial growth were similar. However, after 10 h, the samples were located in the right side of Figure 3A, suggesting that PC1 means bacterial growth and PC2 means metabolic differences among B. fibrisolvens strains. (B) Blue peaks and words indicate PC1 contribution and those in red indicate PC2 contribution. The 4.2–5.2 ppm region was omitted to eliminate the effects of imperfect water suppression.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2654759&req=5

pone-0004893-g003: Principal components analysis (A) and loading plot analysis (B) of three strains of B. fibrisolvens.MDT-10 (red), MDT-5 (blue), and A38 (purple) are shown. Data under conditions of LA addition (open squares), LNA addition (close squares), or no addition (close circles) were calculated. Data are representative of three independent experiments. (A) Hourly 1H-NMR profiling data of three strains of B. fibrisolvens incubated with or without LA or LNA were analyzed. Contributions of PC1 and PC2 were 77.0% and 11.6%, respectively. All 0 h samples were assembled at the left side of Figure 3A, indicating that metabolic conditions before bacterial growth were similar. However, after 10 h, the samples were located in the right side of Figure 3A, suggesting that PC1 means bacterial growth and PC2 means metabolic differences among B. fibrisolvens strains. (B) Blue peaks and words indicate PC1 contribution and those in red indicate PC2 contribution. The 4.2–5.2 ppm region was omitted to eliminate the effects of imperfect water suppression.
Mentions: Different from Z-score analysis, it is possible to detect characteristic metabolites by PCA. Interestingly, the differences in PCA profiles were more remarkable among the different strains than among the different medium conditions (Fig. 3-A). The PCA profiles of the three strains revealed that the metabolic characteristics of MDT-10 and MDT-5 were similar, while those of A38 were quite different. We calculated the first two components that contained principal components 1 and 2 contributing to 77.0% and 11.6% of the data variance, respectively. The strains were well separated primarily in PC2 and appeared to form two groups of A38 and of MDT-10 and MDT-5. In contrast, PC1 was dominated by the time course, which is common to the three strains. The spectral regions that contributed most to the difference between the two groups are shown in the corresponding loading plot (Fig. 3-B). Metabolites that exerted the greatest influence on this separation were lactate and glucose in PC1 and butyrate, acetate, and formate in PC2 (Fig. 3-B). As the three strains grew, they consumed glucose and produced lactate; however, butyrate, acetate, and formate production was different between the two groups. It is well known that B. fibrisolvens strains could be classified into Type I (high butyrate and formate production) or Type II (high lactate production) based on their fermentation products but not their 16S rDNA sequences [63]–[65]. It has been reported that these differences may be the basis for genetic differences related to the butyrate production pathway [65]. Therefore, this analytical method could be used to characterize bacterial fermentation patterns derived from functional genetic information.

Bottom Line: Notably, intestinal bacteria maintain homeostatic balance in mammals via multiple dynamic biochemical reactions to produce several metabolites from undigested food, and those metabolites exert various effects on mammalian cells in a time-dependent manner.We identified trans11, cis13 conjugated linoleic acid as the intermediate of linolenic acid hydrogenation by B. fibrisolvens, based on the results of (13)C-labeling RT-MT experiments.RT-MT is useful for the characterization of beneficial bacterium that shows potential for use as probiotic by producing bioactive compounds.

View Article: PubMed Central - PubMed

Affiliation: RIKEN Research Center for Allergy and Immunology, Yokohama, Japan.

ABSTRACT

Background: Environmental processes in ecosystems are dynamically altered by several metabolic responses in microorganisms, including intracellular sensing and pumping, battle for survival, and supply of or competition for nutrients. Notably, intestinal bacteria maintain homeostatic balance in mammals via multiple dynamic biochemical reactions to produce several metabolites from undigested food, and those metabolites exert various effects on mammalian cells in a time-dependent manner. We have established a method for the analysis of bacterial metabolic dynamics in real time and used it in combination with statistical NMR procedures.

Methodology/principal findings: We developed a novel method called real-time metabolotyping (RT-MT), which performs sequential (1)H-NMR profiling and two-dimensional (2D) (1)H, (13)C-HSQC (heteronuclear single quantum coherence) profiling during bacterial growth in an NMR tube. The profiles were evaluated with such statistical methods as Z-score analysis, principal components analysis, and time series of statistical TOtal Correlation SpectroScopY (TOCSY). In addition, using 2D (1)H, (13)C-HSQC with the stable isotope labeling technique, we observed the metabolic kinetics of specific biochemical reactions based on time-dependent 2D kinetic profiles. Using these methods, we clarified the pathway for linolenic acid hydrogenation by a gastrointestinal bacterium, Butyrivibrio fibrisolvens. We identified trans11, cis13 conjugated linoleic acid as the intermediate of linolenic acid hydrogenation by B. fibrisolvens, based on the results of (13)C-labeling RT-MT experiments. In addition, we showed that the biohydrogenation of polyunsaturated fatty acids serves as a defense mechanism against their toxic effects.

Conclusions: RT-MT is useful for the characterization of beneficial bacterium that shows potential for use as probiotic by producing bioactive compounds.

Show MeSH
Related in: MedlinePlus