Isolation of a lactic acid bacterium and yeast consortium from a fermented material of Ulva spp. (Chlorophyta)статья из журнала
Аннотация: Aims: Microbiota in a fermented culture of Ulva spp. was examined with the objective to characterize the type of fermentation and to obtain starter microbes for performing seaweed fermentation. Method and Results: Fermented Ulva spp. cultures which were obtained and transferred in a laboratory were examined for their microbiota. With phenotypic characterization and phylogenetic analysis based on rRNA gene nucleotide sequences, the predominant micro-organisms were identified as Lactobacillus brevis, Debaryomyces hanseni var. hansenii, and a Candida zeylanoides-related specimen, suggesting that the observed fermentation can be categorized to lactic acid and ethanol fermentation. Inoculating the individually cultured cell suspensions of the three kinds of micro-organisms with cellulase induced the fermentation in various kinds of seaweed. Conclusions: A microbial consortium composed of a lactic acid bacterium, L. brevis, and yeasts, D. hansenii and a C. zeylanoides-related specimen, were predominant in a fermented culture of Ulva spp. Lactic acid and ethanol fermentation could be induced in various kinds of seaweed by adding this microbial consortium along with cellulase. Significance and Impact of the Study: This is the first report of lactic acid and ethanol fermentation in seaweed, which is expected to provide a new material for food and dietary applications. Fermentation is important in the agriculture and livestock industries for preparing compost and silage diet, respectively. Fermentation is also an integral process used to produce a great variety of food items for our daily life. Fermented products produced by the food and food-related industries can be divided into two categories: those made from plant materials and those made from animal materials. Fermented products made from both plant and animal materials are common for terrestrial purposes. However, few products of aquatic origin have been subject to fermentation: only fish-fermented products such as fish sauce are available (Steinkraus 1993) and algae-fermented products have never been developed to our knowledge. This may be related to a lack of variations in algal fermentation skills: only methane fermentation with seaweed is well studied (Wise et al. 1979; Sivalingam 1982; Hansson 1983) and different types of algal fermentation such as organic acid fermentation which is able to provide food and dietary products are scarcely known. Production of organic acids or volatile fatty acids during seaweed fermentation has been reported in several studies (Hansson 1983 etc.). However, these products are mostly reported as intermediate products during the methane fermentation process and not regarded as the final products of fermentation. It is also known that organic acids can be produced from algal carbohydrates by activities of intestinal microbes of human (Michel et al. 1996) and marine invertebrates (Sawabe et al. 2003). But these are only observations of faecal microbes' activities to produce the organic acids and not regarded as established fermentation skills. Therefore, it can be said that a method to perform organic acid fermentation using seaweed as single carbon source has not been developed to our knowledge. Among the organic acid fermentations, lactic acid fermentation is probably the most common and important in food and food-related industries. However, lactic acid fermentation with seaweed is not known to date. For example, organic acids involved in the methane fermentation process are mostly acetic acid and formic acid, and not lactic acid (Kida 2004). Acetic acid and formic acid are also the major products in the faecal fermentation with algal carbohydrates, and only trace level of lactic acid is detected (Sawabe et al. 2003). Furthermore, microbes involved in this faecal production of lactic acid from algal carbohydrates are assigned to Vibrio strains (Sawabe et al. 2003) and not to typical strains known as lactic acid bacterium (LAB). Therefore, an observation of the predominance of LAB in a culture supplemented with seaweed as a single nutrient source is worth reporting. Our laboratory has been promoting research to convert seaweed fronds to detrital materials of one cell unit. This product is termed single cell detritus (SCD) and is expected be useful as a hatchery diet for fish in place of cultures of microalgae (Uchida et al. 1997a,b). During the research, we unexpectedly obtained a fermented material of seaweed with ester-like odour(s) by treating seaweed (Ulva spp.) fronds with cellulase-active enzymes and leaving the fronds for a long period at low temperature in our laboratory. Gas production from the Ulva culture was not so prominent, suggesting that some unfamiliar fermentation different from methane production is occurring. Furthermore, it was observed that transfer of a portion of the original fermented culture to a new Ulva culture resulted in a successful preparation of another fermented culture (laboratory observation). The present study started from primary trials of culture transfers of the fermented Ulva material without losing its ability to act as a fermentation seed. During these culture transfer trials, additions of three elements was noticed, i.e. (i) NaCl, (ii) cellulase, and (iii) seed culture, were promotional for fermentation. For the first of the present study, microbiota present in the three and four times transferred Ulva cultures was examined with an objective to characterize the type of fermentation. Furthermore, culture transfer was conducted in different conditions without one of the above three elements to understand better the suitable transfer conditions of the fermented culture. Further objective of the present study was to demonstrate a versatile seaweed fermentation method using the starter microbes obtained and characterized in this study. Ulva fronds. Fronds of Ulva spp. were collected at the beach of Uminokouen, Yokohama, and visually observed epiphytic organisms were eliminated by washing with tap water. Fresh fronds were used to prepare cultures 1 (the original fermented culture) and 2 (the second fermented culture transferred from 1), and freeze-stocked (−40°C) fronds (of the same lot harvested for preparing culture no. 1) were used to prepare cultures 3–5 (A–E). Transfer of fermented cultures. Details of the fermented Ulva cultures are given in Table 1 with their approximate culture compositions. To prepare culture 1, Ulva fronds were mixed with four times their weight of sterile (autoclaved) natural seawater (SSW) and fragmented to ca 5 mm square pieces with a mixer (MK-K75; National, Tokyo, Japan). The Ulva-suspension was then treated with enzyme mixtures containing cellulase (final concentration 0·22% w/w; ONOZUKA R-10; Yakult Honsha, Tokyo, Japan) and abalone acetone powder (a visceral enzyme product effective for algal degradation; Sigma Chemical Co., St Louis, MO, USA) solution for 48 h at 5°C. The abalone acetone powder solution was prepared by suspending the enzyme product at a concentration of 5% w/w against SSW and the supernatant collected by centrifuge (11 000 g, 5 min) was added to the frond suspension at a final concentration of 0·22% w/w. The enzyme-treated algal products (ca 50 g) were washed twice with three times the volume of SSW by decantation, suspended in 280 ml of SSW, and left in a sterile 500 ml volume polycarbonate bottle (Nalgene, Rochester, NY, USA) with a screw cap closed for 17 months at 2°C. To prepare culture 2, frond decomposition was conducted in a reaction mixture containing 20% w/w fronds, 0·8% cellulase and 8·6% w/w mannitol for 48 h at 5°C, followed by two washes with SSW and fermentation for 2 months at 5°C after transfer of culture 1 at 5% volume. To prepare culture 3, frond decomposition was conducted in a reaction mixture containing 83% fronds, 0·42% cellulase, 0·42% abalone acetone powder and 16% SSW for 93 h at 5°C, followed by two washes and fermentation for 7 days at 20°C after the transfer of culture 2 at 0·5% volume. The decomposition and fermentation steps were conducted simultaneously to prepare cultures 4 and 5. Culture 4 was prepared by incubating a reaction mixture containing 50% fronds, 1% cellulase, 5% NaCl, and 1% volume of culture 3 for 8 days at 20°C. Culture 5A was prepared by incubating a reaction mixture containing 50% fronds, 1% cellulase, 5% NaCl and 1% volume of culture 4 for 8 days at 20°C. Culture 5B was prepared without addition of cellulase, NaCl, and the seed culture. Culture 5C was prepared as for culture 5A but without the addition of the seed culture. Culture 5D was prepared without the addition of NaCl. Culture 5E was prepared without the addition of cellulase. All the fermented cultures were preserved at 5°C until transfer. Enumeration of micro-organisms contained in cultures was conducted by direct counting (Porter and Feig 1980), most probable number (MPN) or agar plate methods. For the direct counting method, micro-organisms including bacteria, yeast and fungi were stained with 4,6′-diamidino-2-phenylindole, collected on a 0·2 μm filter (Nucleopore PC MB; Whatman, Maidstone, UK) and counted with a fluorescent microscope. For the MPN method, five sets of a medium containing 2·5 g yeast extract, 50 g peptone, 25 g NaCl per litre of distilled water (pH 6·9) were used for culture. For the agar plate method, heterotrophic microbes (including bacteria, yeast and fungi) were counted on standard method agar (Nissui Pharmaceutical, Tokyo, Japan) supplemented with NaCl at 0, 2·5 and 5% w/v (abbreviated as SMA0, SMA2·5 and SMA5, respectively). LAB were counted on plate count agar with bromocresol purple (BCP; Nissui Pharmaceutical) supplemented with NaCl at 0 and 5% (abbreviated as BCP0 and BCP5, respectively). Yellow-coloured (i.e. acid-producing) colonies formed on the BCP0 and BCP5 plates were tentatively regarded as those of LAB. Yeast and fungi were counted on Sabouraud agar (Nissui Pharmaceutical) containing 50 mg l−1 each of streptomycin, kanamycin sulphate, penicillin G potassium and chloramphenicol with NaCl at 0 and 5% (abbreviated as SBR0+ and SBR5+, respectively). Marine microbes (including bacteria, yeast and fungi) were counted on Marine Agar 2216 (MA; Difco-Becton Dickinson and Co., Sparks, MD, USA). Only the largest CFU numbers were shown for the viable counts obtained from a medium containing NaCl at different levels. Tentative typing and numerical estimation of each yeast type were made based on visual observation of the colony shape formed on SBR5+ plates and occasional microscopic observation of the vegetative cells. Analysis of bacterial isolates. Twenty bacterial colonies were picked up at random from each of the BCP plates prepared for microbial counting of fermented cultures 4 and 5 (A, C–E), and purified. For culture 5B, 20 colonies were picked up from the MA plates prepared before culturing and another 20 colonies were picked up from the SMA2.5 plate prepared after culturing. The isolates were examined for the Gram stain, cell shape, motility, pigment, O/F, oxidase, catalase, agar degradation, gelatine degradation, growth rate, and gas production tests using common methods (Harrigan and McCance 1966; Baumann and Baumann 1981) and tentatively grouped to genus level according to the scheme of Shimidu (1985). To determine the nucleotide sequences of 16S rRNA genes (16S rDNA), the total DNA of the isolates was extracted using a commercial kit (GenTorukun; Takara, Tokyo, Japan) and used as a template for PCR amplification. The nearly full-length 16S rDNA was amplified using the primers 27F (5′-dAGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-dTACGGTTACCTTGTTACGACTT-3′) (Weisburg et al. 1991). The composition of the PCR mixture and the thermal profiles were the same as described in a previous paper (Maeda et al. 1999). The PCR conditions were as follows: 30 cycles of 93°C for 60 s (denaturation), 48°C for 45 s (annealing), 72°C for 90 s (extension) and a final 5-min extension step for 3′ A-overhangs of PCR products. Partial nucleotide sequences (Escherichia coli position 41–338) of the 16S rDNA were determined for all isolates using a Dye Terminator Cycle Sequencing Ready Reaction kit (Perkin Elmer, Boston, MA, USA) and the 27F primer, and were used for typing isolates. Productions of lactic acid and ethanol from glucose were measured by a commercial kit (F kit d-lactic acid/l-lactic acid, F kit ethanol; Roche Diagnostics, Basel, Switzerland) and gas production was visually observed. Growth temperature range of the microbes was tested using MRS medium (Merck Co., Darmstadt, Germany). Growth salt range was tested at 20°C using MRS medium. Acid production from sugars was examined using API50CHL (bioMerieux, Marcy l'Etoile, France). The base composition of DNA was measured by high-performance liquid chromatography (Ezaki et al. 1987). Nucleotide sequences of the nearly full-length 16S rDNA (E. coli position 41–1492) were determined only for the representative strains of B4101, B5201, B5401 and B5501. The nucleotide sequences obtained were entered for BLAST searching (Altschul et al. 1997) into the website of the EMBL–GenBank–DDBJ (http://www.ddbj.nig.ac.jp/), and a simple similarity value was calculated to the most neighbouring specimen. Analysis of yeast isolates. Yeast colonies formed on the SBR5 plates were grouped based on colony appearance and 10 colonies were isolated from each of cultures 4, 5A and 5C–E. Another 10 colonies were isolated at random from the colonies formed on each of the BCP0 plates from cultures 4, 5A and 5C–E. Total DNA of the yeast isolates was extracted using a commercial kit (GenTorukun; Takara, Otsu, Japan) and used as a template for PCR amplification. The nearly full-length 18S rDNA was amplified using primers P1 (5′-dAGAGTTTGATCCTGGCTCAG-3′) and NS8 (5′-dTACGGTTACCTTGTTACGACTT-3′) (Suzuki et al. 1999). The PCR conditions were as follows: 25 cycles of 94°C for 30 s (denaturation), 55°C for 30 s (annealing), 72°C for 2 min (extension), and a final 7-min extension step for 3′ A-overhangs of PCR products. First, the partial nucleotide sequence (Saccharomyces cervisiae position 1670–1850) of the 18S rDNA was determined for all isolates using the primer NS8 and used to type isolates. As for the representative strains of each type obtained from different cultures, nucleotide sequences of the nearly full-length 18S rDNA were determined. Phylogenetic analysis was performed using the ClustalW program (Thompson et al. 1994). Alignment gaps and unidentified base positions were not considered. Evolutionary distances were calculated according to Kimura's two-parameter model (Kimura 1980). The phylogenic tree was constructed using the neighbour-joining method (Saitou and Nei 1987) and a bootstrap analysis of 1000 replicates was performed. Phenotypic characterization was performed by common methods (Yarrow 1998). Morphological observation of cells was conducted on Bacto yeast morphology agar and Bacto YM broth (Difco-Becton Dickinson and Co.). Formation of ascospores was observed on acetate agar (McClay et al. 1959) and malt extract agar (Wickerham 1951). Fermentation of sugars was examined using Durham tubes and fermentation basal medium (Wickerham 1951). The nucleotide sequences described in this report have been deposited with DDBJ under accession numbers AB070606, AB070607, AB070608, AB070609, AB070610, AB070611, AB070854, AB070855 and AB070856 for isolates B4101, B5201, B5401, B5406, B5407, B5501, Y5201, Y5206 and Y5318, respectively. The nucleotide sequences used for a phylogenetic study of yeast had accession numbers J01353, AB013567, AB013590, AB013509, AB013553, AB013532 and AB013555 for S. cerevisiae, Debaryomyces hansenii var. fabri, D. hansenii var. hansenii, Candida zeylanoides, C. ralunensis, C. boleticola and C. krissii respectively. Seaweeds were freshly collected from coastal waters at Yokosuka, except for Undaria pinnatifida and Laminaria japonica. One sample of U. pinnatifida was a commercial dried product ('Wakamidori', passed through a 74 μm mesh; Riken Shokuhin, Tokyo, Japan). The other two samples of U. pinnatifida were fresh commercial products harvested in Iwate. The sample of L. japonica was a damp-dried product fragmented and passed through a 74 μm mesh. All fresh samples of seaweed were freeze-dried and fragmented by a mixer to pass through a 1 mm mesh. For Chondracanthus tenellus, Gracilaria vermiculophyra, C. teedii, Ishige okamurae, and Gelidium linoides, the frond tissue was so hard that it could not be fragmented sufficiently by the mixer alone, and further degradation was conducted with a mortar until the material passed through a 1-mm mesh. Seaweed materials (0·5 g) were suspended in 9 ml of 3·5% NaCl solution (autoclaved) containing 0·1 g of cellulase R-10 in a sterile 12-ml volume plastic tube. The cultures were inoculated with a microbial mixture and incubated for 7 days at 20°C with the screw cap tightly closed, and rotated moderately at 5 rpm (Rotator RT-550; Taitec, Tokyo, Japan). To prepare the microbial mixture, strains B5201, Y5201 and Y5206 were freshly grown in MRS broth (Merck Co.) for LAB and Bacto YM broth for yeast. The cultured cells were collected by centrifuge, washed twice, and suspended in a sterile 0·85% NaCl solution at the concentration O.D.660 nm = 1·0. A mixture of 50 μl each of the cell suspensions containing the strains B5201 (2·5 × 107 CFU), Y5201 (2·7 × 104 CFU), and Y5206 (6·0 × 105 CFU) was inoculated as a starter. The fermented cultures were tested by smell and evaluated as +, ±, or − if they contained favourable (ester-like), seaweed-originated, or unacceptable (rotten) odours, respectively. The supernatant of the cultured water was obtained by centrifuging for 10 min at 10,000 g and lactic acid and ethanol contents were determined using commercial kits (F kit d-lactic acid/l-lactic acid, F kit ethanol). Data were obtained from duplicate cultures and average values are shown. After incubation for 17 months at 2°C, culture 1 produced fruity or ester-like favourable odours. The culture was observed microscopically and found to contain a number of protoplasts of Ulva with rod bacteria, yeast and fungus-like cells, suggesting that a kind of fermentation had occurred in the bottle because of the actions of some micro-organisms. An aliquot of culture 1 was transferred to a new Ulva-culture and incubated for 1 week to obtain culture 2. Culture 2 was successfully fermented, which was judged from its fruity ester-like odour. The transfer was conducted at a 2–4-month interval from cultures 1 to 5 (A–E) without losing its ability for inducing fermentation. Culture 5A was prepared with the addition of NaCl (final concentration 5% w/w), cellulase (final concentration 1% w/w) and seed culture 4 (final concentration 1% w/w) and successfully fermented based on its fruity ester-like odours. In contrast, culture 5B, which was prepared without the addition of the three elements (i.e. NaCl, cellulase and seed culture), produced unacceptable odours and was judged rotten. Cultures 5 (C–E), which were prepared with one of the three elements lacking (Table 1), also successfully fermented, suggesting that a combination of at least two elements among the NaCl, cellulase and seed culture promoted the fermentation of Ulva. The fermented culture of Ulva could have been transferred for over 8 months without losing its ability to induce fermentation of a new Ulva culture, suggesting the existence of micro-organisms available as a starter for seaweed fermentation. The results of microbial counting are given in Table 2 for cultures 4 and 5 (A–E). For culture 5A, the total number of micro-organisms, including bacteria, yeast and fungi, was estimated as 6·6 ± 1·5 × 107 cells per litre by the direct counting method. The viable number of micro-organisms was estimated as 5·4 × 107 cells per litre by the MPN method. The maximum number of micro-organisms estimated by the plate counting method was 7·1 ± 4·1 × 107 CFU ml−1, which was counted as yellow-coloured colonies on the BCP0 plates. The counting results obtained from the three methods did not differ significantly, suggesting that the predominant micro-organisms in culture 5A were culturable on the BCP0 plate and probably grouped with LAB. Yeast and fungal colonies were also counted at 2·8 ± 1·5 × 107 CFU ml−1 on the SBR0+ plates, and most (>98%) of these colonies were observed to be composed of two types of yeast, based on the morphological characteristics of the colonies. The seed culture 4 was also found to contain LAB and yeast as the predominant micro-organism, suggesting that these kinds of micro-organisms are involved in the observed fermentation and have been transferred from culture to culture. The control culture 5B prepared without the three elements (NaCl, cellulase and seed) contained marine microbes at 1·0 ± 0·1 × 105 CFU ml−1 but not LAB and yeast (<103 CFU ml−1) initially (before incubation). After incubation, marine microbes predominated in the rotten culture 5B at 8·2 ± 0·4 × 108 CFU ml−1. This figure differed greatly from the result obtained from direct counting, 2·0 ± 0·8 × 109 cells per ml, compared with the fermented cultures. Culture 5C, which was prepared without the addition of seed culture 4, contained predominantly yeast cells at 1·2 ± 0·1 × 108 CFU ml−1, suggesting that the addition of the seed culture had contributed a supplementary source of LAB and that yeast cells were originally contained in the Ulva fronds. The microbial number counted for culture 5C on the MA plates was mostly (>80%) of yeast origin, which was determined by microscopic observation. Culture 5D, prepared without the addition of NaCl, contained predominantly LAB and yeast, which was similar to culture 5A, except that the number of LAB was slightly higher at 2·3 ± 0·1 × 108 CFU ml−1. Culture 5E, prepared without the addition of cellulase, contained a less fermented odour but both LAB and yeast were predominant at 6·8 ± 2·2 × 107 CFU ml−1 and 1·3 ± 0·3 × 107 CFU ml−1, respectively. Twenty bacterial colonies were isolated from each of the BCP0 plates of the fermented cultures for further characterization. For the rotten culture 5B, 20 colonies on the MA and SMA2·5 plates were isolated both before and after the incubation of the culture, because larger numbers of colonies were obtained with these agar plates than with the BCP0 plates. For culture 5C, only yeast colonies were observed on the BCP0 plates and bacterial colonies were not obtained. The results of the phenotypic characterization of the isolates are given in Table 3, with typing based on the partial nucleotide sequences of 16S rDNA. All 20 strains isolated from cultures 4, 5A and 5E were grouped to monotype. This bacterium had the following characteristics: Gram-positive, rod structure, nonmotile, nonpigmented, oxidase negative, catalase negative and showed fermentative utilization of glucose. It was assigned to the genus Lactobacillus. Simple similarity of the partial sequence of 16S rDNA (including two variable regions) was 100% among these strains which were designated Lactobacillus I. For the isolates from culture 5D, 17 strains were determined as Lactobacillus I, but the B5406 and B5409 strains and the B5407 strain were designated Lactobacillus II and III, respectively, based on the partial 16S rDNA nucleotide sequences. For culture 5B, 13 strains formed yellow-orange pigmented colonies before incubation and were grouped as Flavobacterium–Cytophaga. Another six strains and one strain were grouped to Acinetobacter–Moraxella and Alteromonas, respectively. For culture 5B after incubation, 12, four, three and one strains were grouped to Vibrio–Aeromonas, Alteromonas, Alcaligenes-Pseudomonas and Streptococcus–Leuconostoc, respectively. It was concluded that Lactobacillus I was predominant in the fermented cultures, while the control culture prepared without addition of NaCl, cellulase and seed culture resulted in spoilage and contained various kinds of saprophytic bacteria. The bacterial strain B5201, a representative strain of Lactobacillus I, was further characterized (Table 4). The strain B5201 was nonspore forming and nonmotile. Production of lactic acid from glucose was observed at a productivity of 50% on carbon basis with productions of ethanol and gas, suggesting that the style of the lactic acid fermentation is heterotype. The bacterium can grow at 15 and 40°C, but not at 45°C, and in a medium containing 2·5 and 5% NaCl. Acid production was observed from l-arabinose, esculin, fructose, galactose, gluconate, glucose, maltose, ribose and d-xylose. The GC content of DNA was 45% on a molecular basis. The sequence similarity of the 16S rDNA (1475 bp) was 99·8% against the closest neighbour of Lactobacillus brevis. Based on the above results, the strain B5201 was identified as L. brevis. The nearly full-length nucleotide sequences of the 16S rDNA were determined for strains B4101, B5401 and B5501 along with B5201, and no sequence difference was observed among these four strains grouped to Lactobacillus I. Strain B5406 (Lactobacillus II) was determined for a nucleotide sequence of 16S rDNA (1487 bp) and had the highest homology with L. casei (simple similarity value 99·9%). Strain B5407 (Lactobacillus III) was determined for a nucleotide sequence of 16S rDNA (1481 bp) and assigned to L. pentosus (simple similarity value 100%). The morphological characteristics of yeast colonies formed on the SBR0+ plates were similar, while those formed on the SBR5+ plates were well developed (or differentiated) and could be grouped into three types. Therefore, 10 colonies were isolated from each of the SBR5+ plates for the fermented cultures, 4 and 5 (A, C–E). No yeast colony could be obtained from the rotten culture 5B. The colonies obtained from cultures 4 and 5 (A, D, E) were tentatively grouped into two types: one type of colony was large and rough (designated type LR, Table 5) and the other type of colony was medium and smooth (type MS). The proportion of type LR was estimated for each culture: 97% for 4, 84% for 5A, 50% for 5D and 90% for 5E. For culture 5C, all colonies were similar to type LR, but 89% of colonies were slightly smaller and, therefore, were designated as type MLR (medium-large and rough). The typing of yeasts based on the morphological characteristics of colonies coincided with the typing based on the results of microscopic observation of cell shape and 18S rDNA partial nucleotide sequence (Table 5). Type LR yeast had a spherical to short-oval cell shape (2·7–5·4) × (2·7–6·0) μm, vegetative reproduction by multilateral budding and no pigment. The partial nucleotide sequence of 18S rDNA (S. cerevisiae position 1427–1738) showed 100% similarity among the 10 isolates of this group, which was tentatively designated as Debaryomyces I based on the result from the BLAST search. Type MLR yeast showed the same morphological characteristics as type LR yeast but the nucleotide sequence differed at one position among the partial 18S rDNA sequence. This yeast was designated Debaryomyces II. Type MS yeast had a short-oval to long-oval cell shape (1·2–3·2) × (4·3 × 10·0) μm, vegetative reproduction by multilateral budding and no pigment. Much longer cells occurred when the culture became old. The partial nucleotide sequence of 18S rDNA showed 100% similarity among the 10 isolates of this group, which was tentatively designated Candida I based on the result from the BLAST search. Yeast colonies were also observed on the BCP0 plate, which was prepared for counting the number of LAB. These yeast colonies were large and not yellow-coloured and, therefore, easily discriminated from LAB colonies. To examine the yeast flora of the fermented cultures more closely, 10 yeast colonies were isolated at random from each of the BCP0 plates of cultures 4 and 5 (A, D, E). For culture 5C, 10 yeast colonies were also isolated from the BCP0 plates. These colonies could be visually grouped into two types: a white colony that darkened after ageing (type WD) and a white colony that retained its colour with age (type W). Five colonies of each of these strains were isolated. The yeast isolates obtained from the BCP0 plates were composed of Debaryomyces I and Candida I for cultures 4 and 5 (A, D, E), and Debaryomyces I and Debaryomyces II for culture 5C, based on determination of the partial 18S rDNA sequences, which coincided with the results from the SBR5+ plate isolates. The yeast strains Y5201 [a representative of Debaryomyces I) and Y5206 (a representative of Candida I] were further characterized (Table 6). Strain Y5201 showed vegetative reproduction with multilateral budding, no pseudomycelium formation, but formation of round warty ascospores. Weak fermentation was observed from d-glucose and sucrose but not from d-galactose, maltose, lactose and raffinose. Strain Y5201 grew on l-arabinose, esculin, fructose, galactose, gluconate, glucose, maltose, ribose and d-xylose as a single carbon source. The yeast did not grow at 35°C. The sequence similarity of the 18S rDNA (1752 bp) was 99·8% against the closest neighbour D. hansenii var hansenii. Based on the above results and a phylogenic study (Fig. 1), strain Y5201 was assigned to D. hansenii var hansenii. Strain Y5206 showed vegetative reproduction with multilateral budding, formation of pseudomycelium but not of ascospores. Fermentation was not observed on d-glucose, sucrose, d-galactose, maltose, lactose and raffinose. Strain Y5206 grew on d-glucose, l-sorbose, d-glucosamine, α,α-trehalose, glycerol, d-sorbitol, d-mannitol, and 2-keto-d-gluconic acid as a single carbon source. The yeast did not grow at 35°C. The sequence similarity of the 18S rDNA (1750 bp) was 99·5% against the closest neighbour C. zeylanoides. Based on the above results and a phylogenic study, strain Y5206 was tentatively grouped to a C. zeylanoides-related specimen (Fig. 1). The sequence similarity of the 18S rDNA of strain Y5318, a representative of Debaryomyces II, was 99·5% (different at eight positions among 1752 bp) against D. hansenii var. fabri and 99·4% against D. hansenii var. hansenii. A phylogenetic study assigned strain Y5318 to the Debaryomyces cluster (Fig. 1). Neighbour-joining tree based on 18S r RNA gene sequence. The numerals represent the confidence level from 1000 replicate bootstrap samplings. Bar: 0·01 nucleotides substitution per site To test the possible use of the micro-organisms isolated from the Ulva culture as a starter for fermentation of different seaweeds, a cell suspension mixture containing strains B5201, Y5201 and Y5206 was added to suspensions of 22 kinds of seaweed along with cellulase. Reduction of pH values and production of favourable ester-like odours and gas were observed in many cultures (Table 7). Production of lactic acid and ethanol were also observed in the range of 0–1·14 g 100 ml−1 and 0·03–0·41 g 100 ml−1, respectively. Production of lactic acid was highest for cultures of vascular plant and Chlorophyta, second highest for cultures of Rhodophyta and comparatively low for cultures of Phaeophyta. However, production of lactic acid was exceptionally high for U. pinnatifida and L. japonica among the Phaeopyta seaweeds. Fermented material of Ulva spp. was initially obtained unexpectedly in our laboratory and was observed to be available as a seed culture for fermentation of Ulva. The objective of the present study was to obtain micro-organism(s) available as a starter for seaweed fermentation. Therefore, we simplified the culture conditions necessary before transfer of the seed culture as follows: (i) single use of cellulase was adopted instead of the combined use of cellulase and abalone acetone powder, (ii) freeze-stocked fronds were used instead of fresh fronds, and (iii) NaCl was added because it is well known that addition of NaCl helps to prevent the growth of saprophyte micro-organisms in conventional fermented food industries. For cultures 1–4, only the culture conditions that were used to prepare the seed cultures are given in Table 1; however, we tried many other culture conditions at each transfer step such as those used for cultures 5 (A–E) (unpublished observations). During the trial and error of transfer, we observed that the addition of three elements, NaCl, cellulase and seed culture, promoted Ulva fermentation. The result that LAB and yeast are predominant in the fermented materials implied that both lactic acid and ethanol fermentation are involved in the fermentation of Ulva. The cellulase is expected to act not only for decomposing seaweed but also for supplying glucose utilized as a substrate for lactic acid and ethanol fermentation. This idea is confirmed by the observation that glucose is produced and the accumulation of lactic acid follows during the culture of Ulva suspensions (unpublished observations). The Ulva fronds used in this study were not sterilized, and contained many indigenous bacteria such as the Flavobacterium–Cytophaga group, and saprophytic bacteria such as the Vibrio–Aeromonas and Alteromonas (including Pseudoalteromonas) groups, before and after the incubation, respectively (Table 3). However, the successfully fermented cultures contained a simple microbial flora, mostly composed of L. brevis, D. hansenii var. hansenii and C. zeylanoides-related species. Therefore, these three kinds of micro-organisms seem more suitably termed 'a consortium' rather than just 'a mixture', although any actual relationships among these micro-organisms are not demonstrated in the present study. The observation that two types of Debaryomyces yeast were predominant in culture 5C, which was prepared without inoculation of the seed culture, suggested a common existence of Debaryomyces-type yeast in the Ulva fronds and that the strain Y5201 originates from Ulva fronds. In contrast, no L. brevis strain was isolated from culture 5C, and the origin of L. brevis strain B5201 is unclear. However, the fact that LAB of different types were isolated from culture 5D (Table 2), suggests that LAB is distributed on Ulva fronds to some extent. Furthermore, L. brevis is known to grow in high saline cultures containing 6.5% NaCl (Miyao 1995). Considering these facts, it is probable that strain B5201 originates from the Ulva fronds. We have already developed a method to detect LAB and yeast at the species or strain level by PCR techniques, and a study to examine the involvement of each micro-organism of the consortium during fermentation is in progress (Uchida et al. 2004). Occasional predominance of L. brevis was also observed in U. pinnatifida cultures prepared without the addition of starter microbes in a serial study (Uchida et al. 2004). We have demonstrated that the microbial consortium acts as a starter for fermenting various kinds of seaweeds. Although the microbiota of the cultures after incubation was not examined in the present study, the predominance of the added microbes was confirmed as for the case of U. pinnatifida culture (Uchida et al. 2004). The production efficiency of lactic acid varied among the seaweeds. The high production efficiency of lactic acid and ethanol in the vascular plant and Chlorophyta is probably related to the high cellulose content of these sea grass and seaweed (Bobin-Dubigeon et al. 1997). Among the seaweeds tested in the present study, U. pinnatifida was remarkably easily degraded to one cell unit by the action of cellulase during incubation. This point is the focus of another study (Uchida and Murata 2002), and the optimum culture conditions have already been reported. This is the first report of a method to perform a lactic acid and ethanol fermentation in various kinds of seaweeds using cellulase for saccharizing and a LAB and yeast consortium as fermenting starter. We thank Dr Motofumi Suzuki for technical assistance in determining the 18S rDNA sequence and Kazuhiro Nagaura for nomenclature characterization of seaweed samples. This study was supported by a grant (the PIONEER Research Project Fund Grant PRPF1204) from the Research Council of the Ministry of Agriculture, Forestry and Fisheries, Japan.
Год издания: 2004
Авторы: Motoharu Uchida, Masakazu Murata
Издательство: Oxford University Press
Источник: Journal of Applied Microbiology
Ключевые слова: Seaweed-derived Bioactive Compounds, Marine and coastal plant biology, Probiotics and Fermented Foods
Другие ссылки: Journal of Applied Microbiology (PDF)
Journal of Applied Microbiology (HTML)
PubMed (HTML)
Journal of Applied Microbiology (HTML)
PubMed (HTML)
Открытый доступ: bronze
Том: 97
Выпуск: 6
Страницы: 1297–1310