Using common taxonomic methods such as 16S rDNA sequencing and physiological and biochemical analysis to identify members of the Lactobacillus casei group (LCG) is time-consuming, expensive and inaccurate. In this study, we applied matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) to rapid discriminate LCG strains by creating an analytical in-house database (IHDB) and to develop a classification model for subspecies-level differentiation based on MS biomarkers using ClinProTools bioinformatics software (Bruker Daltonics, Billerica, MA). Genotypic methods (housekeeping gene sequencing and species-specific PCR) were also established to validate the MALDI-TOF MS platform. A total of 48 LCG reference strains were correctly identified at the species level (mean score: 2.45 ± 0.1) by using MALDI-TOF MS with an IHDB and had high score values, which was in accordance with results from mutL gene sequencing and specific PCR-based methods. However, one strain that was identified as L. casei had a relatively low score value (2.02 ± 0.02), lower sequence similarities (mutL: 90.4%), and failed to amplify a species-specific amplicon; it may therefore represent an undescribed novel species. In addition, after implementation of the classification model (based on 2 biomarker peaks: m/z 4,930 and 5,303), L. paracasei strains could be clearly and easily differentiated to the subspecies level. Afterward, 7 LCG-related isolates from different probiotic samples were analyzed and accurately identified. Our data demonstrate the high-resolution performance of MALDI-TOF MS for fast and accurate demarcation of LCG strains when used with an IHDB coupled to ClinProTools; this methodology can serve as an alternative for quality control of probiotic products.

Trimethylamine oxide (TMAO) originates from trimethylamine (TMA), which is oxidized in the liver by hepatic flavin-containing monooxygenases (FMO3). TMA is produced by its dietary precursors such as choline, carnitine, and phosphatidylcholine by gut microbiota. TMAO attracts attention, identified as a novel and independent risk factor for promoting obesity, atherosclerosis and cardiovascular disease (CVD), chronic kidney disease (CKD), insulin tolerance, and colon cancer. Probiotics have been considered as live microorganisms, providing benefits to their host when they are given in sufficient quantities and administered continuously. The objective of this study is to suggest a method to select potential probiotic strains to reduce the serum concentration of TMAO in mice fed with choline. In this work, we chose three lactobacilli with strong adherence capability, and fed multistrain formula (MF) to the mice challenged with choline. On days 7, 14, and day 28, it was found that the MF-containing L. amylovorus LAM1345, Lpb. plantarum LP1145, and Lim. fermentum LF33 showed a significant reduction in serum TMAO and TMA levels. For the single strains, LP1145 reduced TMAO on days 14 and 28, and strain LAM1345 reduced TMAO significantly on days 7 and day 14. For strain LF1143 from strain LF33, it showed no significant effect on TMAO and TMA. Thus, MF showed the best effect, which may be due to the additive and synergetic effect and the contribution of strain LP1145 and LAM1345. Finally, for the LAM1345 and LP1145 strains, we used molecular identification and typing methods to assure that these two strains are unique strains. The methods used for LAM 1345 were leader peptidase A (lepA) gene analysis and phylogenetic analysis, while for strain LP 1145and other strains of Lpb. plantarum subsp. plantarum sequences were compared using the whole-genome multilocus sequence typing (wgMLST) method.

Two Gram-stain-negative, straight rods, non-motile, asporogenous, catalase-negative and obligately anaerobic butyrate-producing strains, HLW78^T and CYL33, were isolated from faecal samples of two healthy Taiwanese adults. Phylogenetic analyses of 16S rRNA and DNA mismatch repair protein MutL (mutL) gene sequences revealed that these two novel strains belonged to the genus Faecalibacterium. On the basis of 16S rRNA and mutL gene sequence similarities, the type strains Faecalibacterium butyricigenerans AF52-21^T(98.3–98.1 % and 79.0–79.5 % similarity), Faecalibacterium duncaniae A2-165^T(97.8–97.9 % and 70.9–80.1 %), Faecalibacterium hattorii APC922/41-1^T(97.1–97.3 % and 80.3–80.5 %), Faecalibacterium longum CM04-06^T(97.8–98.0% and 78.3 %) and Faecalibacterium prausnitzii ATCC 27768^T(97.3–97.4 % and 82.7–82.9 %) were the closest neighbours to the novel strains HLW78^T and CYL33. Strains HLW78^T and CYL33 had 99.4 % both the 16S rRNA and mutL gene sequence similarities, 97.9 % average nucleotide identity (ANI), 96.3 % average amino acid identity (AAI), and 80.5 % digital DNA–DNA hybridization (dDDH) values, indicating that these two strains are members of the same species. Phylogenomic tree analysis indicated that strains HLW78^T and CYL33 formed an independent robust cluster together with F. prausnitzii ATCC 27768^T. The ANI, AAI and dDDH values between strain HLW78^T and its closest neighbours were below the species delineation thresholds of 77.6–85.1 %, 71.4–85.2 % and 28.3–30.9 %, respectively. The two novel strains could be differentiated from the type strains of their closest Faecalibacterium species based on their cellular fatty acid compositions, which contained C_18 : 1  ω7c and lacked C_15 : 0 and C_17 : 1  ω6c, respectively. Phenotypic, chemotaxonomic and genotypic test results demonstrated that the two novel strains HLW78^T and CYL33 represented a single, novel species within the genus Faecalibacterium, for which the name Faecalibacterium taiwanense sp. nov. is proposed. The type strain is HLW78^T (=BCRC 81397^T=NBRC 116372^T).

Human gut microbiome has an essential role in human health and disease. Although the major dominant microbiota within individuals have been reported, the change of gut microbiome caused by external factors, such as antibiotic use and bowel cleansing, remains unclear. We conducted this study to investigate the change of gut microbiome in overweight male adults after bowel preparation, where none of the participants had been diagnosed with any systemic diseases. A total of 20 overweight, male Taiwanese adults were recruited, and all participants were omnivorous. The participants provided fecal samples and blood samples at three time points: prior to bowel preparation, 7 days after colonoscopy, and 28 days after colonoscopy. The microbiota composition in fecal samples was analyzed using 16S ribosome RNA gene amplicon sequencing. Our results demonstrated that the relative abundance of the most dominant bacteria hardly changed from prior to bowel preparation to 28 days after colonoscopy. Using the ratio of Prevotella to the sum of Prevotella and Bacteroides in the fecal samples at baseline, the participants were separated into two groups. The fecal samples of the Type 1 group was Bacteroides-dominant, and that of the Type 2 group was Prevotella-dominant with a noticeable presence Bacteroides. Bulleidia appears more in the Type 1 fecal samples, while Akkermensia appears more in the Type 2 fecal samples. Of each type, the gut microbial diversity differed slightly among the three collection times. Additionally, the Type 2 fecal microbiota was temporarily susceptible to bowel cleansing. Predictive functional analysis of microbial community reveals that their activities for the mineral absorption metabolism and arachidonic acid metabolism differed significantly between the two types. Depending on their fecal type, the variance of triglycerides and C-reactive protein also differed between the two types of participants. Depending upon the fecal type, the microbial diversity and the predictive functional modules of microbial community differed significantly after bowel preparation. In addition, blood biochemical markers presented somewhat associated with fecal type. Therefore, our results might provide some insights as to how knowledge of the microbial community could be used to promote health through personalized clinical treatment.