Isolation and characteristics of lactic acid bacteria isolated from ripe mulberries in Taiwan.

The objective of this study was to isolate, characterize, and identify lactic acid bacteria (LAB) from ripe mulberries collected in Taiwan. Ripe mulberry samples were collected at five mulberry farms, located in different counties of Taiwan. Eighty-eight acid-producing cultures were isolated from these samples, and isolates were divided into classes first by phenotype, then into groups by restriction fragment length polymorphism (RFLP) analysis and sequencing of 16S ribosomal DNA (rDNA). Phenotypic and biochemical characteristics led to identification of four bacterial groups (A to D). Weissella cibaria was the most abundant type of LAB distributed in four mulberry farms, and Lactobacillus plantarum was the most abundant LAB found in the remaining farm. Ten W. cibaria and one Lactococcus lactis subsp. lactis isolate produced bacteriocins against the indicator strain Lactobacillus sakei JCM 1157(T). These results suggest that various LAB are distributed in ripe mulberries and W. cibaria was the most abundant LAB found in this study.

INTRODUCTION

Mulberry (Morus australis) has been widely cultivated in China and southeastern Asia for thousands of years. The leaves of mulberry trees represent indispensable food for silkworms and the ripe fruits are edible and have been widely used in juices, wines and jams (23). In Taiwan, mulberry trees are widespread and the ripe fruits are always harvested in early April. The ripe fruit is sweet with a very mild flavor and is therefore popular in Taiwan. Although mulberries are very popular, they have not been studied in detail.

Isolation of lactic acid bacteria (LAB) from fruits and vegetables have frequently been reported (2,3,7,14,18). However, studies on LAB associated with ripe mulberries remain scarce.

Bacteriocins produced by LAB have attracted special interest as potential alternative safe commercial food preservatives. LAB have been used as food and feed preservatives for centuries, and bacteriocin-producing LAB could replace chemical preservatives for the prevention of bacterial spoilage and the outgrowth of pathogenic bacteria in food products (6).

The objectives of this study were to isolate LAB from ripe mulberries, identify the isolates at the species level, and screen them for antibacterial activity.

MATERIALS AND METHODS

Sampling

Samples for LAB isolation were collected from five mulberry farms (M1 to M5), which were respectively located in the following five counties in Taiwan: Tainan, Chiayi, Taichung, Miaoli and Taoyuan. The farms in these counties are listed from south to north and are 40–70 km apart. Three separate ripe fruit samples were randomly collected at each mulberry farm. The samples were taken aseptically and packaged into clean bags, then stored at 4°C. Samples were analyzed within 24 h of acquisition at the mulberry farms.

Isolation of LAB

MRS (de Man, Rogosa, and Sharpe)-agar plates were used for the isolation of LAB. To distinguish acid-producing bacteria from other bacteria, 1% CaCO3 was added to the MRS-agar plates. Each mulberry sample was crushed and mixed with 0.75% NaCl solution. Dilutions of the mixed solution (10- to 1000-fold) were spread directly onto the surface of MRS-agar plates. Samples were incubated under anaerobic conditions (Mitsubishi AnaeroPak System, Pack-Anaero, Mitsubishi Gas Chemicals, Tokyo, Japan) at 30°C for 3 to 5 days. Colonies of acid-producing bacteria, identified by a clear zone around each colony, were randomly selected from MRS-agar plates and purified by replating on MRS-agar plates. Colonies were reselected and initially Gram stained and tested for production of catalase. Only Gram-positive, catalase-negative strains were selected. The selected strains were stored at –80°C in 10% skim milk broth.

Identification of isolates

Restriction fragment length polymorphism (RFLP) analysis and sequence analysis of 16S ribosomal DNA (rDNA) were used to classify and identify the bacterial isolates. DNA was isolated by using the methods described by Kozaki et al. (12). Polymerase chain reaction (PCR) was carried out using a Takara Ex Taq gene amplification PCR kit (Takara Bio, Shiga, Japan) to prepare reaction mixtures and performed on a Gene Amp PCR System 9700 (PerkinElmer Corp., Boston, MA, USA) following the methods described by Chen et al. (4). RFLP analysis of 16S rDNA was carried out with the methods described by Chen et al. (4). In this study, three restriction enzymes, AccII (CG/CG), HaeIII (GG/CC) (16), and AluI (AG/CT) (22) were used for grouping.

For sequence analysis of 16S rDNA, the PCR products were purified with a Clean/Gel Extraction Kit (BioKit, Miaoli, Taiwan) and then sequenced with the following primer: 5´-CTGCTGCCTCCCGTAG-3´ (27F). DNA sequencing was performed using an ABI 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). Partial sequencing, approx. 1000 bp ahead was performed and the sequence manually aligned with the software Genetyx-Win version 5.1. Sequence homologies were examined by comparing the sequences obtained with those in the DNA Data Bank of Japan (DDBJ; http://www.ddbj.nig.ac.jp/) using FASTA. During sequence comparison, an initial threshold of 99% homology was required when comparing the raw sequencing result, while 100% homology was obligated for the sequencing data with high signal to noise ratios.

Detection of antibacterial activity

The agar-well diffusion method as described by Onda et al. (15) was used to detect and determine the antibacterial activities of isolates. Lactobacillus sakei JCM 1157T, Listeria monocytogenes ATCC 19111 and Helicobacter pylori ATCC 43504T were used as the indicator strains. L. monocytogenes was propagated in PBN broth (pH 7.3) with the following components: 0.5% peptone, 0.3% beef extract, and 0.8% NaCl. H. pylori was propagated in a Tryptic soy agar (Difco, Sparks, MD, USA) with 5% defibrinated sheep blood added. To exclude the effect of lactic acid, antibacterial activity was later confirmed by pH adjustment (21).

Effect of enzymes on antimicrobial activity

The effect of various enzymes on antimicrobial activity was determined by incubating 500 μL of the neutralized and filter-sterilized cell-free supernatant with 20 μL of the following enzyme solutions at a final concentration of 3 mg/mL respectively: proteinase K (pH 7.0), trypsin (pH 7.0) and catalase (pH 7.0). Enzymes were purchased from Sigma Chemicals, St. Louis, MO, USA. After 5 h of incubation at 37 °C, antimicrobial activity was determined by using agar-well diffusion assay. Untreated samples were used as control and L. sakei JCM 1157T was used as the indicator strain.

RESULTS

A total of 88 acid-producing bacterial strains were isolated from the samples collected in mulberry farms of Taiwan. Of these, 13 strains were isolated from farm M1, 18 from farm M2, 20 from farm M3, 18 from farm M4, and 19 from farm M5 (Table 1). The 88 isolates were classified into four groups (A to D) based on cell morphology and the results of 16S rDNA RFLP analysis (Fig. 1). Of these isolated short-rod strains, 54 were placed in group A, 14 in group B, and 18 in group D, because digestion of their DNA with AccII, HaeIII, and AluI gave the same RFLP patterns. The remaining coccal isolates were placed into group C.

Table 1

Isolates from ripe mulberries

Mulberry farm	Strain No.	Specie	Mulberry farm	Strain No.	Specie	Mulberry farm	Strain No.	Specie
M1	H041703	W. cibaria		H042046	L. pseudomesenteroides		I042210	W. cibaria
	H041706	W. cibaria	M3	I041701	L. plantarum		I042211	W. cibaria
	H041710	W. cibaria		I041702	L. plantarum		I042212	W. cibaria
	H041711	W. cibaria		I041703	L. plantarum		I042213	W. cibaria
	H041713	W. cibaria		I041704	W. cibaria		I042214	W. cibaria
	H041715	W. cibaria		I041705	W. cibaria		I042215	W. cibaria
	H041716	W. cibaria		I041706	L. plantarum		I042217	W. cibaria
	H041717	W. cibaria		I041707	L. plantarum		I042218	W. cibaria
	H041718	W. cibaria		I041708	L. plantarum		I042219	W. cibaria
	H041719	W. cibaria		I041709	L. plantarum	M5	I042501	L. pseudomesenteroides
	H041720	W. cibaria		I041710	L. plantarum		I042502	W. cibaria
	H041721	W. cibaria		I041711	L. plantarum		I042503	W. cibaria
	H041722	W. cibaria		I041712	L. plantarum		I042504	W. cibaria
M2	H042001	W. cibaria		I041713	L. plantarum		I042505	W. cibaria
	H042002	W. cibaria		I041714	L. plantarum		I042506	L. pseudomesenteroides
	H042003	W. cibaria		I041715	L. plantarum		I042507	L. pseudomesenteroides
	H042004	L. pseudomesenteroides		I041716	L. plantarum		I042508	W. cibaria
	H042005	L. pseudomesenteroides		I041717	L. plantarum		I042509	W. cibaria
	H042006	W. cibaria		I041718	L. plantarum		I042510	W. cibaria
	H042007	L. pseudomesenteroides		I041719	L. plantarum		I042511	L. lactis subsp. lactis
	H042008	W. cibaria		I041720	L. plantarum		I042512	L. pseudomesenteroides
	H042009	W. cibaria	M4	I042201	W. cibaria		I042513	L. lactis subsp. lactis
	H042010	W. cibaria		I042202	L. pseudomesenteroides		I042514	W. cibaria
	H042011	L. pseudomesenteroides		I042203	W. cibaria		I042516	W. cibaria
	H042014	L. pseudomesenteroides		I042204	W. cibaria		I042517	W. cibaria
	H042015	W. cibaria		I042205	W. cibaria		I042518	W. cibaria
	H042016	W. cibaria		I042206	W. cibaria		I042519	W. cibaria
	H042017	L. pseudomesenteroides		I042207	W. cibaria		I042520	W. cibaria
	H042018	W. cibaria		I042208	L. pseudomesenteroides
	H042021	L. pseudomesenteroides		I042209	W. cibaria

Figure 1

16S rDNA RFLP patterns of AccII, HaeIII, and AluI digests from groups A to D. Lane M, size marker; A, AccII digestion patterns; H, HaeIII digestion patterns; U, AluI digestion patterns.

To identify the isolates, representative strains were randomly selected from each group, and 16S rDNA sequencing analysis was carried out. The results identified group A isolates as W. cibaria, group B as Leuconostoc pseudomesenteroides, group C as Lactococcus lactis subsp. lactis, and group D as Lactobacillus plantarum. The sequences determined in this study have been deposited in the DDBJ database with sequential accession numbers AB510744 to AB510757.

When assessing inhibitory activity against the indicator strain L. sakei JCM 1157T, clear inhibition zones were observed for 10 W. cibaria and 1 L. lactis subsp. lactis isolates. However, only the same L. lactis subsp. lactis strain, I042513, showed inhibitory activity against L. monocytogenes ATCC 19111 and none of the isolates showed inhibitory activity against H. pylori ATCC 43504T (Table 2).

Table 2

Inhibition spectra and effect of enzymes on the bacteriocin produced by lactic acid bacteria from ripe mulberries.

Isolated strain	Indicator microorganisms			Enzyme treatments
	L. sakei	L. monocytogenes	H. pylori	Diameter of the zone of inhibition (mm) [αβ]
	JCM 1157T	ATCC 19111	ATCC 43504T	Control	Proteinase K	Trypsin	Catalase
W. cibaria
H041703	+	-	-	11	N	N	11
H041706	+	-	-	11	N	N	11
H041710	+	-	-	11	N	N	11
H041713	+	-	-	11	N	N	11
H041715	+	-	-	10	N	N	10
H041718	+	-	-	11	N	N	11
H041719	+	-	-	11	N	N	11
H041720	+	-	-	11	N	N	11
H041721	+	-	-	11	N	N	11
H041722	+	-	-	11	N	N	11
L. lactis subsp. lactis
I042513	+	+	-	17	N	17	17

Wells (8 mm diameter) were filled with 100 μL of culture supernatant.

L. sakei JCM 1157T was used as the indicator strain.

The effects of enzymes on the antimicrobial activity of the isolated substance from 10 W. cibaria and one L. lactis subsp.

lactis strains are shown in Table 2. All the antimicrobial

substances were inactivated by proteinase K but not affected by treatment with catalase. However, trypsin inactivated the antimicrobial substance from 10 W. cibaria but showed no effect on the antimicrobial substance from L. lactis subsp. lactis strain. Protease sensitivity assays demonstrated that the antimicrobial substance produced by the strain listed in Table 2 can be described as bacteriocin since its inhibitory activity was completely eliminated by treatment with enzyme proteinase K (13).

DISCUSSION

LAB cultures were isolated from ripe fruits of all five mulberry farm samples, with differences in abundance and diversity evident among the farms (Table 1). Although the mulberry farms are 40 to 70 km apart from each other, the same dominant species, W. cibaria, was found in all five farms.

In addition, W. cibaria was also the most abundant species of LAB found in the mulberry farms, except for farm M3. Instead of W. cibaria, L. plantarum was the most abundant species of LAB found in farm M3. L. pseudomesenteroides was found only at farms M2, M4 and M5, while L. lactis subsp. lactis was only found at farm M5.

Species such as W. cibaria, L. plantarum and L. lactis subsp. lactis have been frequently found in environments associated with plants (8,9,11,18). In a previous study, LAB associated with wine grapes was studied by Bae et al. (2). Differently from ripe mulberries, L. lindneri was the most abundant species of LAB found from wine grapes. In the case of tomato fruits, L. plantarum was the most abundant species of LAB found (8). Differences in LAB distribution, including not only species but also number, were observed among these fruits.

In this study, we focused not only on the distribution of LAB in mulberry farms, but also on the bacteriocin-producing ability of the isolates. A total of 11 strains, including 10 W. cibaria and 1 L. lactis subsp. lactis, showed inhibitory activities against the indicator strain L. sakei JCM 1157T (Table 2). However, only the same strain of L. lactis subsp. lactis showed inhibitory activity against L. monocytogenes ATCC 19111. Bacteriocins from L. lactis subsp. lactis (1,5,10,15) have been frequently studied. However, studies on bacteriocins from W. cibaria remain scarce (17).

The sensitivity of the substances to proteinase K or trypsin revealed their proteinaceous nature. In addition, no effect was observed with the catalase treatment, suggesting that the active agent is not involved in the H2O2 antimicrobial mechanism. The results obtained in this study were similar to the previous studies (13,17).

Furthermore, it is interesting that 10 bacteriocin-producing W. cibaria strains were all distributed in farm M1. It is difficult to interpret why W. cibaria strains found in other mulberry farms did not produce bacteriocin, and why some W. cibaria produced bacteriocin but some did not. In the previous study of Trias et al. (18), W. cibaria strain TM128 effectively decreased the infection level of fungi. A similar result was also reported by Valerio et al. (19). It is therefore suspected that these bacteriocin-producing W. cibaria may protect mulberries from other microorganisms. However, more studies are necessary to test this hypothesis.

Besides W. cibaria strains, L. lactis subsp. lactis I042513 was the other bacteriocin-producing LAB strain found in this study. An additional experiment was performed to find more information about this bacteriocin. Nisin-specific PCR primers, NISL: 5’-CGAGCATAATAAACGGC-3’ and NISR: 5’-GGATAGTATCCATGTCTGAAC-3’, were used for PCR amplification following the methods and conditions described by Villani et al. (20). The expected amplified band, located at 320 bp, was obtained. The PCR product was sequenced, and the DNA sequence was analyzed and found to encode a peptide containing an amino acid sequence with 100% coincidence to nisin Z (data not shown). However, more detailed protein analyses are necessary to support this result.

In conclusion, the results of this study suggest that various LAB are distributed in ripe mulberries. W. cibaria was the most abundant LAB in four mulberry farms, while L. plantarum was the most abundant LAB in a fifth mulberry farm. Many LAB isolated from ripe mulberries were found to produce bacteriocins. Future studies in our laboratory will characterize and identify the bacteriocins, and we anticipate that the bacteriocins from these LAB may be useful as biopreservatives. The authors hope that the results of this study will offer useful information for the improvement of mulberry cultivation.