A GC-MS Based Metabolic Profiling of Probiotic Lactic Acid Bacteria Isolated from Traditional Food Products

A GC-MS based metabolic profiling was carried out to study metabolic differences of lactic acid bacteria isolated from different food sources. Metabolic fingerprinting is a non-targeted procedure where all detectable peaks are considered to establish sample classification. A total of 40 compounds were identified as major metabolites contributing to the difference among five different probiotic lactic acid bacteria. Some of the metabolites identified in this study have been reported as a defrosting agent, antioxidant, flavour agent, antimicrobial, natural food additive, anti-inflammatory, anti-sleep disorder agent and anti-cancer agents. These results suggest that GC-MS based metabolomic analysis is a useful tool to facilitate future investigations into the characterization of probiotic lactic acid bacteria.


INTRODUCTION
Lactobacillus and Pediococcus is a grampositive facultative anaerobic bacterium found widely in fermented food products and can be investigated for the metabolites present in them by GC-MS based metabolic profiling.
Metabolic fingerprinting is explained as the semi-quantitative investigation of extracellular (exo-metabolome) and intracellular (endometabolome) metabolites, respectively (Villas-Boas et al., 2005). These metabolic profiles GC-MS has consistently been the most favoured analytical technique for the analysis of metabolites present in distinct biological samples. Gas chromatography and mass spectrophotometry (GC-MS) present the high chromatographic resolution ascribed to the high sensitivity and specificity of mass spectrophotometry (Villas-Boas et al., 2005). In comparison with other techniques, GC-MS can produce a comparably high reproducibility, high resolution, high-quality sensitivity, and good-throughput analysis, which can be used for analyzing the metabolic products, inclusive of carbohydrates, fatty acids, organic acids, and amino acids (Park et al., 2016). Derivatization is necessary before the investigation by GC-MS as most of the metabolites are non-volatile (Schummer et al., 2009). Nowadays, most adaptable derivatization technique is silylation, which has the ability to derivatize compounds having polar functional groups by mixing TMS (Trimethylsilyl) reagent to develop TMS Compounds (Nordström A, 2004). However, for the study of different biological metabolic fingerprinting, N, O-bis (trimethylsilyl) trifluoroacetamide with 1% trimethylchlorosilane is generally used as silylation reagent and derivatization method with GC-MS (Alreshidi et al., 2015;Li et al., 2013). For the study of metabolites in the microbial sample, derivatization based GC-MS technique can be applied to identify the intracellular as well as extracellular metabolites of Lactobacillus and Pediococcus species.
Liquid injection GC-MS provides an inexpensive and simple option for the analysis of metabolites produced by probiotic lactic acid bacteria. As the concentration of metabolites varies widely in different probiotic lactic acid bacteria, it is required to develop the analytical plan that would permit concurrent quantification of them in a single run utilizing comparably a small amount of sample along with minimum sample preparation. GC-MS has high sensitivity and hence can be used for the investigation of less common samples that might only be available in minute quantity. In this study, direct-injection GC-MS methodology was used for the profiling of different metabolites, which has the high specificity of mass spectrometry along with the high reproducibility and high resolution of gas chromatography.
The results of this study provide reference data for interpreting the differences in the metabolite profiles of different probiotic lactic acid bacteria isolated from different food sources.

Collection of food samples
Food items viz. Dosa batter, jalebi batter, maida dough, sauerkraut and soymilk were used in this experiment. Dosa batter, Jalebi batter, Soymilk were collected from a local market in New Delhi, India. Maida dough and Sauerkraut were prepared at home. The food samples were taken in a sterilized bag and stored at -4° until use.

Strain isolation
1 g or 1 ml of food sample were added into 9 ml of 0.85% (w/v) normal saline. After homogenization, serial dilutions were prepared upto 10 -9 with 0.85% (w/v) normal saline and 0.1 ml decimal of appropriate dilutions were plated onto de Man, Rogosa, Sharpe (MRS) agar medium (Himedia, India) (de Man et al., 1960). The agar plates were incubated at 35° for 24 h under anaerobiosis. Morphologically different colonies were picked and re-streaked onto MRS agar plates up to purity. Glycerol stocks of strains were preserved at -20°.

Metabolic fingerprinting
Metabolites were extracted by following the method given by Coucheney et al. (2008) with minor modifications.

Sampling for metabolic fingerprinting
Microbial suspensions (20 ml) of isolate (grown in MRS broth for 24 h at 35°C) were disrupted using pulsed, high-frequency sound waves (>20kHz) for five cycles (30 sec. run; one min break) to extract intracellular metabolites. Suspensions procured after sonication were centrifuged for 10 min at 10,000 rpm in order to segregate the extra as well as intracellular metabolites from the cells.

Extraction for metabolic fingerprinting
The supernatant was removed and the metabolites were extracted with a Methanol: Water: Chloroform mixture (2: 0.8: 1): 2.5 ml of cold chloroform and 5 ml of cold methanol (-20°C) and the phases were allowed to separate. Metabolic fingerprints were assessed in the aqueous phase after freeze-drying while in the organic phase, the dried chemical extracts obtained after complete evaporation of the solvent was used.

Generation of metabolic fingerprints and data processing Derivatization of dried samples
The derivatization method given by Mastrangelo et al. (2015) and Park et al. (2019) was used in this study for the evaluation of metabolic fingerprinting. The freeze-dried samples and dried chemical extracts were methoxymated using methoxyamine in pyridine solution and trimethylsilylated by BSTFA (N,O-Bis(trimethylsilyl) trifluoroacetamide) with 1 % chlorotrimethylsilane (TMCS).

GC-MS analysis
Derivatives from Lactobacillus and Pediococcus strains were analysed using Thermo Trace 1300GC coupled with Thermo TSQ 800 Triple Quadrupole MS (Pogacic et al., 2015). Sixmicrolitre aliquots were injected in splitless mode. The samples were warmed to 65°C for 15 min and the metabolites were extracted before being adsorbed on the trap at 35°C. The trap was heated at 250°C for 0.1 min, leading to desorption of the metabolites. Metabolites were then separated on TG 5MS column (30m X 0.25mm, 0.25µm thickness) with column makeup of 5% diphenyl; 95% dimethyl polysiloxane. The temperature of the oven was initially 50°C, maintained for 3 min. The temperature was increased at 15°C/min to 220°C. The mass spectrometer was operated in the scan mode within a mass range of m/z 30 to 700. The quadruple mass spectrophotometry parameters were set to the conditions: ion source temperature of 230°C, split flow of 40 ml/min, carrier flow of 1.5 ml/ min, injector temperature and MS transfer line temperature of 250°C. Ionization was done by electronic impact at 70 eV. All samples were analysed in the same GC-MS and injected in a randomized order over the GC-MS run. Blank samples (boiled deionized water) were injected after every sample to verify the instrumental carryover.

Data pre-processing and data analysis
The GC-MS raw data files were converted to netCDF format. Further, the raw Hexadecane,2,6,11,15 tetramethyl Hydroperoxide, heptyl ■ 21 Hydroperoxide, 1-methylhexyl ■ 22 Hydroperoxide, pentyl Propanoic acid, 2-hydroxy-, methylester data were processed to time-and mass-aligned chromatographic peak areas with the XCalibur 2.2SP1 with Foundation 2.0SP1. Metabolites were identified by comparison of mass spectra and retention times with those of authentic standards from NIST 2.0 Mass Spectral Library (Mastrangelo et al., 2015).

Metabolic fingerprinting
The exo-and endo-metabolites of probiotic isolates in culture supernatant provides a window for explaining the comprehensive nature of metabolites. The metabolite profiling provides information about the nutritional as well as of toxic effects, if any of metabolites on the interactive effects of the dietary components on the health of the gut of the host.
All five bacterial strains secreted different metabolites. The metabolites, their retention time, peak area, area % and peak height are represented in Table 3 and the GC-MS chromatograms are shown in Figure 1.
Total scans for L. plantarum DB-2 was 5709. Total run time was 19.42 min (5.00 -24.42 min) (Figure 1a). The maximum peak area of 19659374.61 at the peak height of 7235909.56 was observed with 85.29% area covered at retention time 10.53 min (Table 3a, Figure 1b).
L. fermentum J-1 has total of 5713 scans. Total run time was 19.43 min (5.00 -24.43 min) (Figure 1c). The maximum peak area of 863376582.00 at the peak height of 157335492.57 was observed with 54.03% area covered at retention time 11.37 min (Table 3b, Figure 1d).  L. plantarum DB-2, P. acidilactici M-3 and P. pentosaceus SM-2. A metabolic compound -Tetradecanoic acid was present in L. fermentum J-1, P. acidilactici M-3 and P. pentosaceus SM-2 ( Table 2)  Metabolite -Squalene was synthesized in the culture supernatant of P. acidilactici M-3 and P. pentosaceus SM-2 ( Table 2). Squalene is a triterpine compound, originally found in shark liver oil and has a high therapeutic potential. Squalene is a therapeutic agent having anti-cancerous, anti-tumour, chemo-preventive, antioxidant, and sunscreen properties along with anti-microbial activity (Ezhilan and Neelamegam, 2012). In industry, it is scarcely produced by plant sources along with the shark liver oil. Microbial production of Squalene is an attractive alternative to label the issue with an objective to increase its productivity and purity by using biotechnological interventions. Although, a detailed study is required for the commercialization of the production of Squalene from native probiotic bacteria. 9 metabolites out of 40 were produced by only one selected isolate viz. Acetic acid, anhydride with formic acid; 4-amino-1-butanol; 2-Ethoxyethylamine; Ethylamine; Formamide; Hydroperoxide, heptyl; Hydroperoxide, 1-methylhexyl; 4-Penten-2-ol and Propanoic acid, 2-hydroxy-, methylester. Compounds -Acetic acid, anhydride with formic acid; 2-Ethoxyethylamine; Ethylamine; Formamide were produced by L. plantarum DB-2. Metabolite -4-amino-1-butanol was found to be synthesized by L. plantarum SK-3. Metabolite -Hydroperoxide, heptyl was present in P. acidilactici M-3. Compounds -Hydroperoxide, 1-methylhexyl and Propanoic acid, 2-hydroxy-, methylester were secreted by L. fermentum J1. Metabolite -4-Penten-2-ol was produced by P. pentosaceus SM-2 ( Table 2). Probiotics has a vital role in well-being and health beyond basic nutrition. Probiotics have potential health benefits such as antimicrobial activity, antimutagenic, anticarcinogenic activities, constipation, regulation of immune function, improving gastrointestinal health, reducing lactose intolerance, and allergenic diseases such as food allergy, etc. Due to these health benefits, probiotic bacteria and their metabolites are gaining importance in pharmaceutical preparation, food products, medicines and dietary supplements (Da Cruz et al., 2009;Settanni and Moschetti, 2010;Fernandez et al., 2013). Production of different primary and secondary metabolites by the probiotic bacteria, employs their biological effects directly or by modifying the immune system. The metabolites of lactic acid bacteria provide therapeutic benefits in preventing and curing various diseases and also responsible for the development of flavour, aroma and texture in food products (Chen et al., 2014) have been remarked with various functional as well as therapeutic properties viz. food additives, flavouring agents, anti-inflammatory, anti-cancerous, potential against skin disorders, cholesterol-lowering and anti-diarrheal properties and endorse their huge potential in food and pharmaceutical industries.
Many researchers have observed similar metabolic products in many lactic acid bacteria influencing human metabolism when incorporated into the system as probiotics or the food fermented by the respective bacteria.  Phenol, 2, 4-bis (1, dimethylethyl)) after GC-MS in symbiotic cow milk beverage fermented with L. kefiranofaciens, Candida kefir and Saccharomyces boulardii. The fermented cow milk symbiotic beverage had the greater anti-diarrhoeal effect as compared to other milk beverages, thereby indicating the role of antibacterial metabolites produced by the probiotic bacteria. Padmavathi bis (1,1-dimethylethyl) during ripening period (180 days) of cheese fermented with L. casei LC2W and found these metabolites help in accelerating the ripening process.
The reliability of the technique in investigating the effect of metabolic compounds on colonic metabolic signature has been recently