Organic acids (OAs) were well known as the final products in the metabolic process, which therefore supply information for altered biochemical metabolism from various biological samples including microorganisms.1-3 In particular, lactic acid in industrial chemistry fields was reported as important metabolite in fermentation step.4-9 In our previous report,10 phenyllactic acid was produced in high abundance with lactic acid and acetic acid. Phenyllactic acid is a novel antimicrobial compound active against Gram-positive and Gram-negative bacteria.11,12 Phenyllactic acid can be converted from phenylpyruvic acid by NADH dependent lactic acid dehydrogenase (LDH, EC 1.1.1.2.7).14 Therefore, the analysis of OAs in culture media and cells of microorganisms is important in the monitoring and screening of microbial availability. In our previous reports, OA profiling analyses by gas chromatography (GC) and GC-mass spectrometry (GC-MS) were found to be useful for the comparative analysis between control and experimental groups.9,10,14,15

Gram positive Lactobacillus pentosus K34 (L. pentosus K34) is facultative anaerobic and hetero-fermentative as rod-shaped lactic acid bacterium with rounded ends, which may occur singly, in pairs or short chains. It is nonmotile and grows at 10 ℃ and 40 ℃ but not at 45 ℃.16 In contrast, Pediococcus lolli PL24 (P. lolli PL24) is homo-fermentative Gram-positive lactic acid bacterium. It is coccus-shaped, nonspore- forming, nonmotile, and occurs in pairs or tetrads.17 L. pentosus K34 and P. lolli PL24 have two different characteristics. First, P. lolli PL24 has a thermo-tolerance and grows at high temperature of 47 ℃, while L. pentosus K34 cannot grow at high temperature of 47 ℃. Thermotolerant lactic acid bacteria has a beneficial probiotic advantage, for example, maintaining cell viability during manufacturing processes such as heat drying.18 Second, P. lolli PL24 has been known to be obligately homofermentative, while L. pentosus has been known to be facultatively hetero-fermenative. Normally, the metabolic pathway of homo lactic fermentation uses the Embden- Meyerhoff-Parnas (EMP) pathway and the metabolic pathway of hetero lactic fermentation uses the pentose phosphate pathway. Thus, in this study, our previous OA profiling method9,14,15 using methoxime/tert-butyldimethylsilyl (MO/TBDMS) derivatives in combination with GCMS was applied to culture media of L. pentosus K34 and P. lolli PL24 for the monitoring of altered OA metabolism.

L. pentosus K34 with the strong inhibitory activity against gastro-intestinal pathogenic bacteria was isolated from the small intestine of Korean native chicken10,11 and P. lolli PL24 with the inhibitory activity against gastrointestinal pathogenic bacteria as well as thermo-tolerance at 47 ℃ was newly isolated from the waste of milk processing (n = 3). Each strain was activated at 37 ℃ for 2 days on de Man, Rogosa and Sharpe (MRS) (BD, USA) agar plate (n = 3). Then one loopful of each grown single colony was suspended into 1 mL 0.85% NaCl solution. Using syringe, 300 μL of each cell suspension was inoculated into 30 mL glass vial sealed by rubber and aluminum cap containing 20 mL MRS broth and incubated at 37 ℃ for 3 days without shaking. After cultivation, the culture broth was centrifuged at 10,000 g for 10 min. The supernatant was taken to the fresh new conical tube (SPL, Korea) and stored at ？70 ℃ deep freezer (Ilsin, Korea).

OA standards including 3,4-dimethoxybenzoic acid as internal standard (IS), triethylamine (TEA) methoxyamine hydrochloride were purchased from Sigma-Aldrich. N-Methyl- N-(tert-butyldimethylsilyl)trifluoroacetamid (MTBSTFA) + 1% tert-butyldimethylchlorosilane was provided from Thermo Scientific (Bellefonte, PA, USA). Toluene, diethyl ether, ethyl acetate, and sodium chloride (pesticide grade) were obtained from Kanto Chemical (Chuo-ku, Tokyo, Japan). Sulfuric acid was purchased from Samchun pure chemical Co. Ltd. (Pyeongtaek, Gyeonggido, South Korea). All other chemicals were of analytical reagent grade.

Derivatized samples were analyzed using an Agilent 6890N gas chromatograph interfaced to an Agilent 5975B mass-selective detector (70 eV, electron ionization source). The mass spectra were scanned in the mass range of 50- 650 u at a rate of 0.99 scans/s. The temperatures of the injector, interface, and ion source were 260, 300, and 230 ℃, respectively. An Ultra-2, cross-linked capillary column coated with 5% phenyl-95% methylpolysiloxane bonded phase (25 m × 0.20 mm I.D., 0.11 mm film thickness, Agilent Technologies, Santa Clara, CA, USA) was used for all analyses. Helium was used as the carrier gas at a flow rate of 0.5 mL/min in the constant flow mode. Samples (1 μL) were introduced in split-injection mode (10:1), and the oven temperature was set initially at 100 ℃ (2 min), then increased to 250 ℃ at rate of 5 ℃/min and finally programmed to 300 ℃ at rate of 20 ℃/min (5 min).

Control media (MRS), and culture media from L. pentosus K34 and P. lolli PL24 were used for experiments (n = 3). Aliquots of culture media (20 μL) containing IS (5 μg) were spiked to distilled water (1 mL) and reacted with methylhydroxylamine hydrochloride (1 mg) in alkaline condition at 60 ℃ for 30 min for conversion into MO derivative. The reaction mixture was then acidified to pH < 2 with 10% sulfuric acid solution, saturated with sodium chloride, and extracted with diethyl ether (4 mL) followed by ethyl acetate (2 mL). After addition of TEA (5 μL), the combined extracts were evaporated under a gentle stream of nitrogen (40 ℃) to dryness. Toluene (20 μL) as the solvent and MTBSTFA (20 μL) as the silylation reagent were added to the residue, and the mixture was heated at 60 ℃ for 30 min to form MO/TBDMS derivatives prior to analysis by GC-MS.

The mean peak area ratios to IS of OAs confirmed in the control media and culture media from L. pentosus K34 and P. lolli PL24 were normalized to the corresponding mean of those in the L. pentosus K34. Then normalized levels of 12 OAs were plotted with lines radiating for star symbol plotting using Microsoft Excel (Microsoft, Redmond, WA), as described in previous report.19

GC-MS profiles from the control media (A), L. pentosus K34 (B), and P. lolli PL24 (C) are showed in Figure 1. Seven OAs were screened in the control media (A), while 12 and 10 OAs were identified in L. pentosus K34 (B) and P. lolli PL24 (C), respectively (see also Table 1). In control media, lactic acid (No.2) was most abundant, followed by oleic acid (No.9), pyruvic acid (No.1), glycolic acid (No.3), succinic acid (No.7), and citric acid (No.12). Lactic acid (No.2) in L. pentosus K34 was most abundant, followed by oleic acid (No.9), succinic acid (No.7), pyruvic acid (No.1) and phenyllactic acid (No.8). In contrast, in P. lolli PL24, lactic acid (No.2) was most abundant, followed by oleic acid (No.9), pyruvic acid (No.1), phenyllactic acid (No.8), and succinic acid (No.7). The overall OA levels of L. pentosus K34 were higher when compared with those of the P. lolli PL24. This finding may suggest that OAs produced by P. lolli PL24 was smaller than those of L. pentosus K34. In particular, P. loll PL24 mainly produced lactic acid (No.2) and phenyllactic acid (No.8) which are assumed to be produced from pyruvic acid and phenylpyruvic acid by LDH via EMP pathway. The difference in the overall OA levels is due to the different metabolic pathway of each

strain. P. lolli PL24 with homo lactic fermentation mainly produced lactic acid from pyruvic acid by LDH, while L. pentosus K34 with hetero lactic fermentation produced

the various OAs including lactic acid throughout the different metabolic pathway. Typically, hydroxyl acids such as No.4, No.5, and No.6 produced by L. pentosus K34 might be produced by central metabolism and/or amino acid metabolism. Specifically, α-keto acids were produced by biosynthetic enzymes involved in the central metabolism and/ or also by aminotransferases from amino acids. Subsequently, hydroxyl acids were produced by hydroxyl acid dehydrogenases from α-keto acids.20

The OA levels in the media of control and P. lolli PL24 were normalized to the corresponding mean OA levels in the media of L. pentosus K34 (Table 1), which proved very informative by expressing the alteration (ranging from 0.00 to 8.20) of the OA values. In control media, the levels of glycolic acid (No.3) and citric acid (No.12) were highly elevated compared to those of L. pentosus K34 and P. lolli PL24. This is because these two OAs are used for growth of L. pentosus K34 and P. lolli PL24. In P. lolli PL24, the abundances of phenyllactic acid (No.8), 4-hydroxyphenyllactic acid (No.10), indole-3-lactic acid (No. 11), and citric acid (No.12) in the P. lolli PL24 were much larger, while the abundances of 2-hydroxyisovaleric acid (No.4), 2- hydroxyisocaproic acid (No.5), 2-hydroxy-3-methylvaleric acid (No.6), and succinic acid (No.7) were smaller than those of the L. pentosus K34. The star symbol plots drawn based on these values displays distorted patterns (Figure 2) for the control media (Figure 2A) and the P. lolli PL24 (Figure 2B) when compared to the pattern of the L. pentosus K34. Thus, this result may explain different OA metabolism and fermentation process of L. pentosus K34 and P. lolli PL24.

OA metabolic profiles in cell growth media from L. pentosus K34 as hetero fermentative and P. lolli PL24 as homo fermentative showed different patterns. And the star patterns were readily distinguishable and characteristic of each strain. This may explain for usefulness of OA profiling method using GC-MS combined with star pattern recognition for monitoring of altered and characteristic OA metabolism in various microorganisms.