Identification of Antioxidants Produced by Lactobacillus plantarum

Biosci. Biotechnol. Biochem., 77 (6), 1299–1302, 2013

Note

Identification of Antioxidants Produced by Lactobacillus plantarum Yoko SUZUKI,1 Misa K OSAKA,1 Kazutoshi S HINDO,1 Toshiyuki K AWASUMI,1 Hiromi K IMOTO-N IRA,2 and Chise S UZUKI2; y 1

Department of Food and Nutrition, Faculty of Human Sciences and Design, Japan Women’s University, 2-8-1 Mejiro-dai, Bunkyo-ku, Tokyo 112-8681, Japan 2 NARO Institute of Livestock and Grassland Science, 2 Ikenodai, Tsukuba, Ibaraki 301-0901, Japan Received December 26, 2012; Accepted March 1, 2013; Online Publication, June 7, 2013 [doi:10.1271/bbb.121006]

We identified two compounds that demonstrated 2,2diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity from cultures of Lactobacillus plantarum. Spectroscopic analyses proved these compounds to be L-3-(4hydroxyphenyl) lactic acid (HPLA) and L-indole-3-lactic acid (ILA). The respective EC50 values for HPLA and ILA were 36:6  4:3 mM and 13:4  1:0 mM. Key words:

3-(4-hydroxyphenyl)lactic acid; indole-3lactic acid; radical scavenger; antioxidant; Lactobacillus plantarum

Oxidation is an essential process for many living organisms, but it produces free radicals that damage such important biomolecules as lipids, proteins and DNA. The accumulation of oxidative damage is thought as one of the primary determinants of aging. In addition to endogenous antioxidants, an intake of antioxidants as supplements or foods may help prevent the human body from cellular damage and diseases mediated by free radicals. The utilization of lactic acid bacteria (LAB) exerting antioxidative activity is a fascinating subject of study in this respect. Although some reports have documented the antioxidative effects of LAB,1–3) there are no reports identifying the antioxidants produced by LAB. The aim of this study was to isolate and identify the antioxidants in cultures of LAB. The LAB strains producing 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging metabolite(s) were first screened by using a colony assay.4) Briefly, LAB colonies on an MRS plate (Lactobacilli MRS broth, Becton Dickinson, Sparks, MD, USA, containing 1.6% agar) were replica plated on a sterilized filter paper. The filter paper was then placed on a new MRS plate and anaerobically cultured overnight. The filter paper was taken out and sprayed with a DPPH solution (80 mg/mL in ethanol). Those strains showing a white-on-purple spot were regarded as antioxidant-producing strains. Among the 109 strains tested, 37 were selected as antioxidant-producing strains. To analyze the bacterial metabolite(s), 37 strains were cultured in an MRS medium at 37
C for 24 h. The antioxidants were extractable with ethyl acetate (EtOAc) in a preliminary test, so the culture filtrates were extracted with EtOAc. The EtOAc extract of the MRS

medium itself contained various antioxidants, as judged by its DPPH radical scavenging activity. To detect the LAB-specific antioxidants, the reverse-phase high-performance liquid chromatographic (HPLC) profiles of the EtOAc extract of the culture filtrate (pH 4.0) and that of the MRS medium (pH 4.0 adjusted with 1% lactic acid) were compared (the HPLC conditions are shown in Fig. 1A). The peaks that were observed in some culture extracts and not detected in the MRS medium were collected and examined for their DPPH radical scavenging activity. Compounds, 1 (22 min), 2 (25.6 min) and 3 (26.3 min) were typically produced by the strains of Lactobacillus plantarum and Lactobacillus paraplantarum, whereas only compounds 1 and 3 showed DPPH radical scavenging activity (Fig. 1A). As already mentioned, compounds 1 and 3 were strain-specific antioxidants, but not the main antioxidants in the culture filtrate. To identify the antioxidants, L. plantarum F51 (the highest produced strain) was anaerobically cultured in 1 L of the MRS medium at 37
C for 24 h, and the resulting cells were removed by centrifugation. The culture filtrate (1 L, pH 4.0) was extracted by EtOAc, and the extract was evaporated to dryness. Compounds 1 and 3 were purified by preparative reversed-phase HPLC, using a Capcell Pak C18 AQ column (10 i.d.  250 mm) with a flow rate of 2.4 mL/min. The eluted peaks at 22 min and 26.3 min were collected to respectively yield pure compounds 1 (2.5 mg) and 3 (1.1 mg). High-resolution atmospheric pressure chemical ionization mass spectrometry (HRAPCI-MS (), JMST100LP (Jeol)) of 1 showed the ðM  HÞ peak at m=z 181.05011 (calcd. for C9 H9 O4 , 181.0509), and the molecular formula of 1 was determined to be C9 H10 O4 . The 1 H-NMR spectrum (Avance 400 (Bruker BioSpin)) of 1 in CD3 OD showed two sp2 methine signals (
6.69 and
7.07, each 2H equivalent) which were coupled with each other (J ¼ 8:5 Hz). This proved the presence of a para substituted benzene structure in 1. The 1 H-NMR spectrum also showed one sp3 methine (
4.26, H-20 ) and one sp3 methylene (
2.82 and
2.99, H-10 ) which were coupled with each other, indicating the presence of a partial structure of –CH–CH2 – in 1. These observations strongly suggested that 1 was closely

y To whom correspondence should be addressed. Fax: +81-29-838-8606; E-mail: csuzuki@affrc.go.jp Abbreviations: LAB, lactic acid bacteria; DPPH, 2,2-diphenyl-1-picrylhydrazyl; HPLA, 3-(4-hydroxyphenyl)lactic acid; ILA, indole-3-lactic acid; PLA, 3-phenyllactic acid; HPPA, 4-(3-hydroxyphenyl)pyruvic acid; PPA, 3-phenylpyruvic acid; IPA, indole-3-pyruvic acid; HRAPCI-MS, highresolution atmospheric pressure chemical ionization mass spectrometry; TFA, trifluoroacetic acid; EtOAc, ethyl acetate

1300

Y. SUZUKI et al.

A

B

Fig. 1. HPLC Profile of the EtOAc Extract of Lactobacillus plantarum F51 (A) and Structures of Compounds 1, 2 and 3 (B). An LC10AD HPLC system with an SPD-M10Avp photodiode array detector (Shimadzu, Japan) and a Capcell Pak C18 AQ column (4.0 i.d.  250 mm) was used for the analysis at a flow rate of 0.8 mL/min. The mobile phase used was 5 min with 100% solvent A (water containing 0.1% trifluoroacetic acid (TFA)) and then with a linear gradient from 100% solvent A to 100% solvent B (80% methanol containing 0.1% TFA) typically in 25 min. The chemical structures of compounds 1 as L-3-(4-hydroxyphenyl)lactic acid, 2 as L-phenyllactic acid, and 3 as L-indole-3-lactic acid are presented (B).

related to tyrosine, although the molecular formula of 1 (C9 H10 O4 ) was different from that of tyrosine (C9 H11 NO3 ). The 13 C-NMR spectrum of 1 also proved 1 to be similar to tyrosine, whereas the chemical shift of methine (
73.0, C-20 ) of 1 was different from that of tyrosine (
53.0). Considering the molecular formula of 1 and the foregoing observations, 1 was speculated to be 3-(4-hydroxyphenyl)lactic acid (HPLA). The reported 13 C-NMR spectral data5) for HPLA showed good agreement with those of 1. Finally, 1 was identified to be HPLA by co-chromatography (HPLC) with an authentic sample (Fig. 1B). 1 H-NMR of 1 (CD3 OD)
: 2.80 (dd, J ¼ 7:8, 13.7 Hz, H-10 b), 2.99 (dd, J ¼ 4:5, 13.7 Hz, H-10 a), 4.26 (dd, J ¼ 4:5, 7.8 Hz, H-20 ), 6.69 (d, J ¼ 8:5 Hz, H-3 and H-5), 7.07 (d, J ¼ 8:5 Hz, H-2 and H-4). 13 C-NMR of 1 (CD3 OD)
: 40.8 (C-10 ), 73.0 (C-20 ), 116.0 (C-3 and C-5), 129.5 (C-1), 131.5 (C-2 and C-6), 177.1 (C-4), 177.3 (C-30 ). The HRAPCI-MS () data for 3 showed the ðM  HÞ peak at m=z 190.06291 (calcd. for C11 H11 O3 , 190.06297), and the molecular formula of 3 was determined to be C11 H12 O3 . The 1 H- and 13 C-NMR data for 3 showed it to be closely related to tryptophan (molecular formula of C11 H13 NO2 ), although the 13 C-chemical shift of methine (
70.6, C-20 ) of 3 was different from that of tryptophan (
54.6). Considering the related differences between 1 and tyrosine, 3 was proposed to be indol-3lactic acid (ILA, Fig. 1A). Compound 3 was confirmed to be ILA by co-chromatography (HPLC) with an authentic sample (Fig. 1B). To our knowledge, the DPPH radical scavenging activity of HPLA and ILA has not previously been demonstrated. 1 H-NMR of 3 (CD3 OD)
: 3.10 (dd, J ¼ 7:1, 14.9 Hz, H-10 b), 3.30 (m, H-10 a), 4.42 (m, H-20 ), 6.99 (dd, J ¼ 7:3, 7.8 Hz, H-5), 7.07 (dd, J ¼ 7:3, 7.8 Hz, H-6), 7.12 (s, H-2), 7.31

Table 1. The Absolute Configuration of Compounds 1–3 in a Chiral HPLC Columna Compound b

L-HPLA

D-HPLA

HPLA from F51 (1) L-PLAc D-PLA

PLA from F51 (2) L-ILAc D-ILA

ILA from F51 (3)

Retention time (min) 11.5 14.5 11.5 15.2 14.6 15.3 20.4 18.0 20.4

a

The solvent used for compounds 1 and 3 was n-hexane:2-propanol:TFA ¼ 80:20:1, and that for compound 2 was n-hexane:2-propanol:TFA ¼ 90:10:0.1. b (S)-3-(4-Hydroxyphenyl)-2-hydroxypropanoic acid was purchased from AstaTech (Bristol, PA, USA). c L-PLA and L-ILA were purified by HPLC from an L-lactate dehydrogenase reaction mixture containing 0.4 mM NADH, and either 2 mM phenylpyruvic acid or 2 mM indolepyruvic acid.16)

(d, J ¼ 7:8 Hz, H-7), 7.58 (d, J ¼ 7:8 Hz, H-4). 13 CNMR of 3 (CD3 OD)
: 31.5 (C-10 ), 70.6 (C-20 ), 111.0 (C-3), 112.1 (C-7), 119.4 (C-4), 119.6 (C-5), 122.2 (C-6), 124.6 (C-2), 128.0 (C-3a), 138.1 (C-7a), 174.7 (C-30 ). Compound 2 (Fig. 1) did not show any DPPH radical scavenging activity when an aliquot contained in the 10-mL culture filtrate was assayed. It was identified as 3-phenyllactic acid (PLA) by the MS, NMR, and cochromatography (HPLC) data with an authentic sample (Fig. 1B). 1 H-NMR of 2 (CD3 OD)
: 2.82 (dd, J ¼ 7:8, 13.3 Hz, H-10 b), 3.03 (dd, J ¼ 4:3, 13.3 Hz, H-10 a), 4.36 (dd, J ¼ 4:3, 7.8 Hz, H-20 ), 7.27 (dd, J ¼ 7:6, 7.6 Hz, H-4), 7.29 (d, J ¼ 7:6 Hz, H-2 and H-6), 7.40 (dd, J ¼ 7:6, 7.6 Hz, H-3 and H-5). 13 C-NMR of 2 (CD3 OD)

Antioxidants Produced by L. plantarum Table 2. Production of HPLA, ILA and PLA by LAB

a b

1301 a

Species

No. of strainsb

HPLA (mM)

ILA (mM)

PLA (mM)

Lactobacillus casei/rhamnosus Lactobacillus coryniformis Lactobacillus paraplantarum Lactobacillus plantarum Lactobacillus curvatus Lactobacillus sakei Lactococcus lactis Leuconostoc citreum

6(6) 3(2) 5(5) 39(36) 10(2) 15(5) 13(3) 3(1)

14:4  14:2 2:5  4:3 72:8  40:4 89:2  63:1 0:6  2:0 1:4  3:1 2:3  4:9 0:9  1:6

5:5  7:5 0:4  0:7 20:7  20:2 17:7  13:0 0:3  0:6 0:8  1:0 0:5  0:7 0:3  0:4

93:0  78:8 0:0  0:0 230:3  174:3 233:9  175:6 3:9  8:3 5:9  10:7 10:5  18:1 4:0  6:9

Leuconostoc mesenteroides

15(10)

2:1  3:8

0:7  1:3

19:3  37:2

Concentration (mM) of each compound in the culture filtrate is indicated. Data are presented as mean  standard deviation. Number of analyzed strains is indicated. Numbers in parentheses indicate the number of antioxidant positive strains in the colony assay.


: 40.3 (C-10 ), 73.4 (C-20 ), 125.7 (C-4), 127.8 (C-2 and C-6), 128.9 (C-3 and C-5), 136.0 (C-1), 177.7 (C-30 ). The absolute configurations of compounds 1–3 obtained from L. plantarum F51 were analyzed by using a chiral HPLC column (Daicel OD-H, 10 i.d.  250 mm), with a flow rate 3.0 mL/min, detected at 254 nm. All the compounds were determined to be L-isomers. Details of the analytical conditions and results are shown in Table 1. We evaluated the productivity of HPLA, PLA and ILA for the LAB strains by analyzing the EtOAc extracts of the culture filtrates of 109 strains belonging to 9 LAB species by reverse-phase HPLC. The average amounts of these compounds produced by the strains of each species are summarized in Table 2. HPLA, PLA and ILA were detected in all the strains of L. plantarum and L. paraplantarum. In the other species analyzed, certain strains, but not all, produced low levels of these compounds. The production of hydroxy acids from aromatic amino acids may therefore be highly specific to L. plantarum and L. paraplantarum. HPLA, ILA and PLA were found as metabolites of LAB by a transamination reaction of the aromatic amino acids followed and subsequent dehydration of the -keto acids.6–8) Certain strains of L. plantarum from sourdough have been reported to produce HPLA and PLA8) and reported to be antifungal compounds.8,9) An aromatic aminotransferase has been shown to be active with tyrosine, phenylalanine, tryptophan, methionine and leucine.10,11) Enzymatic conversion by certain dehydrogenases12) and chemical oxidation steps13) may be involved in the subsequent conversion of -keto acids to hydroxy acids. According to a genomic analysis, the strains of L. plantarum have various dehydrogenases, including three hydroxyisocaproate dehydrogenases (hicD1, hicD2 and hicD3), and several (1 to 4) hydroxyacid dehydrogenases.14) Which dehydrogenases are responsible for the high production of hydroxy acids in the strains of L. plantarum remains to be investigated. To evaluate the DPPH radical scavenging activity of aromatic amino acids and their metabolites, the concentration required to scavenge 50% of the free radicals in the reaction mixture within 20 min (EC50 ) was determined. When 100 mM was used, the L-aromatic amino acids and DL-PLA did not show any DPPH radical scavenging activity. The respective EC50 values for DL-HPLA and DL-ILA were 36:6  4:3 mM (6.7 mg/mL) and 13:4  1:0 mM (2.7 mg/mL), whereas that for the EtOAc extract of the culture filtrate was 0.71 mg/mL.

HPLA and ILA were therefore strain-specific antioxidants, but not the main ones. The EC50 value for LHPLA (purchased from Astatech) was identical to that for DL-HPLA (purchased from Sigma), suggesting there was no difference in the effect of L-HPLA and DL-HPLA on the DPPH radical scavenging activity. The respective EC50 values for the -keto acids, 4-(3hydroxyphenyl)pyruvic acid (HPPA), 3-phenylpyruvic acid (PPA) and indol-3-pyruvic acid (IPA) were 0:19  0:01 mM, 0:16  0:03 mM, and 0:14  0:01 mM. The EC50 value for 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox) was 0:12  0:07 mM, which is comparable to those of the -keto acids. Recent studies have indicated that the keto-carbonyl groups provided strong antioxidative activity, and thus the -keto acids have the ability to scavenge H2 O2 radicals.15) Furthermore, HPPA is converted into homogentisate in photosynthetic organisms, this being the first aromatic precursor of tocopherols (vitamin E). HPPA, PPA and IPA may be useful compounds as antioxidants because of their high radical scavenging activity. The protective effect of hydroxy acids and keto acids on ultraviolet B-derived oxidative stress in keratinocytes is under investigation.17) The strains showing a white-on-purple spot in the first screening were not always HPLA- and ILA-producing strains. A thin-layer chromatographic analysis suggested that other compounds with radical scavenging activity were present in the EtOAc extract of the culture filtrate (data not shown). It would be interesting to identify these compounds and to expand the utilization of LAB exerting its antioxidative activity.

Acknowledgment This study was supported in part by a grant from The National Agriculture and Food Research Organization.

References 1) 2) 3) 4) 5)

Terahara M, Nishide S, and Kaneko T, Biosci. Biotechnol. Biochem., 64, 1868–1873 (2000). Li S, Zhao Y, Zhang L, Zhang X, Huang L, Li D, Niu C, Yang Z, and Wang Q, Food Chem., 135, 1914–1919 (2012). Terahara M, Kurama S, and Takemoto N, Biosci. Biotechnol. Biochem., 65, 1864–1868 (2001). Takao T, Kitatani F, Watanabe N, Yagi A, and Sakata K, Biosci. Biotechnol. Biochem., 58, 1780–1783 (1994). Ishida K, Okita Y, Matsuda H, Okino T, and Murakami M, Tetrahedron, 55, 10971–10988 (1999).

1302 6) 7) 8)

9) 10) 11) 12)

Y. SUZUKI et al. Gummalla S and Broadbent JR, J. Dairy Sci., 84, 1011–1019 (2001). Dallagnol AM, Catalan CA, Mercado MI, Font de Valdez G, and Rollan GC, J. Appl. Microbiol., 111, 1447–1455 (2011). Lavermicocca P, Valerio F, Evidente A, Lazzaroni S, Corsetti A, and Gobbetti M, Appl. Environ. Microbiol., 66, 4084–4090 (2000). Valerio F, Lavermicocca P, Pascale M, and Visconti A, FEMS Microbiol. Lett., 233, 289–295 (2004). Yvon M, Thirouin S, Rijnen L, Fromentier D, and Gripon JC, Appl. Environ. Microbiol., 63, 414–419 (1997). Rijnen L, Bonneau S, and Yvon M, Appl. Environ. Microbiol., 65, 4873–4880 (1999). Jia J, Mu W, Zhang T, and Jiang B, Appl. Biochem. Biotechnol.,

13) 14)

15) 16)

17)

162, 242–251 (2010). Nierop Groot MN and de Bont JAM, Appl. Environ. Microbiol., 64, 3009–3013 (1998). Siezen RJ, Francke C, Renckens B, Boekhorst J, Wels M, Kleerebezem M, and van Hijum SA, J. Bacteriol., 194, 195–196 (2012). Liu J, Litt L, Segal MR, Kelly MJ, Pelton JG, and Kim M, Int. J. Mol. Sci., 12, 6469–6501 (2011). Rossmann MG, Liljas A, Branden C-I, and Banaszak LJ, ‘‘The Enzyme’’ Vol. 11, ed. Boyer PD, Academic Press, New York, p. 61 (1975). Aoki-Yoshida A, Ichida K, Aoki R, Kawasumi T, Suzuki C, and Takayama Y, Biosci. Biotechnol. Biochem., in press.