Lactobacillus is the largest genus within the group of lactic acid bacteria.

Based on the 16S rRNA sequences, lactobacilli are phylogenetically distributed in seven groups: Lb. buchneri (bu), Lb. casei (ca), Lb. delbrueckii (de), Lb. plantarum (pl), Lb. reuteri (re), Lb. sakei (sa), and Lb. salivarius (sl). Traditionally, the genus is split into three groups based on the carbohydrate fermentation pathways: (1) obligate homofermentative, (2) facultative heterofermentative, and (3) obligate heterofermentative lactobacilli. The two most obvious beneficial roles of lactobacilli are as starter cultures (to produce acid rapidly) and as probiotic cultures. Lactobacilli are one of the few contaminant bacteria that are able to grow in cheese after manufacture (as nonstarter lactic acid bacteria (NSLAB)). Lactobacilli as starters and as NSLAB may play different roles in the metabolism of lactose, lactate, and citrate, and in proteolysis and lipolysis, which are considered the primary events during cheese ripening. This article focuses on (1) the taxonomy of the genus Lactobacillus; (2) isolation, enumeration, and identification of Lactobacillus spp.; and (3) the principal role ofLactobacillus as starter, NSLAB, and probiotics in fermented milks and cheeses.

Lactobacillus brevis
From MicrobeWiki, the student-edited microbiology resource

A Microbial Biorealm page on the genus Lactobacillus brevis

Classification
"Lactobacillus brevis", Jeff Broadbent, Utah State University(DOE) (2)

Higher order taxa

Bacteria; Firmicutes; Bacilli; Lactobacillales; Lactobacillaceae; Lactobacillus NCBI

Species
Lactobacillus brevis NCBI

Description and significance

Lactobacillus brevis is a species of lactic acid bacteria, all of which are Gram-positive, non spore forming organisms whose main metabolic pathway involves fermenting hexose sugars to produce lactic acid.

L. brevis plays a large part in the fermentation of wine and cheese (8)

The significance of all lactic acid bacteria throughout history up until today is their metabolism which is used for the preservations of foods and beverages. Despite a simple metabolic pathway, lactic acid bacteria play a prominent part in the worlds food supply and production of some beers and wine. Along with other lactic acid-producing bacteria, L. brevis plays an integral role in the fermentation of certain foods such as sauerkraut and pickles and is likewise the most common cause of spoiled beer. In fact, in

Germany at one point, more than half of beer spoilage incidents were due to L. brevis alone.(9) L. brevis can be isolated from the food sources in inhabits, one example being kimchi. The importance of L. brevis sequenced genome, is that integral proteins and enzymes can be isolated and used in experiment such as observing strains of hop-resistant lactic acid bacteria in beer spoilage or isolating decarboxylase DNA that can be cloned and inserted in other chromosomal DNA to encode novel proteins for that strain.

Genome structure
Whole-genome sequencing was conducted at the U.S. Department of Energy for eight other species of Lactic acid bacteria. The method used was shotgun sequencing, where different parts of a genome are sequenced and then the pieces were assembled. The completed sequenced genome of L. brevis shows that this bacteria is circular and has one chromosome with a length of 2.3 Mbp containing 2185 proteins and 82 RNAs. Two plasmids are also sequenced, much smaller in length, and have 11 and 12 proteins.(3) The evolution of Lactobacillales reveals a lot about the heritage of species and the ancestral gene losses, and may suggest where duplications and the addition of new and unique genes occurred. "Phylogenetic analyses, comparison of gene content across the group, and reconstruction of ancestral gene sets indicate a combination of extensive gene loss and key gene acquisitions via horizontal gene transfer during the coevolution of lactic acid bacteria with their habitats."(3) One notable difference in L. brevis from the other species in this order, includes the loss of genes that carry out the biosynthesis of arginine and aromatic amino acids.

Cell structure and metabolism

Lactobacillus brevis and the Lactobacilli in general, are Gram-positive, rod-shaped, non spore forming bacteria arranged in singles. This implies a thick cell wall and an inner membrane in its rod shape. Although the sequenced genome of L. brevis is relatively small, the lactic acid bacteria can encode a multitude of different transporters to fulfill the needs of both prototrophic and auxotrophic strains.(3) L. brevisbacteria are aerobe bacteria, they are motile, and probiotic.(4) Fermentation is the one most common pathway in L. brevis and most other lactic acid bacteria. In this metabolic pathway, sugar (hexoses) is converted to lactic acid by the 6-phosphogluconate pathway, producing also CO2 and ethanol.(2) Energy (ATP) is generated along the way by non-oxidative substrate level phoshorylation. The final product of fermentation, lactic acid contributes to production of such foods as cheese, yogurt, fermented milks and others by inhibiting the growth of other organisms and lowering the pH of the products. Along the pathway, many metabolic reactions occur, one of them being the

breakdown of milk products for cheese production. L. brevisspecifically is used in industrial production to act as a starter culture for several types of beer and sourdough bread.(4)

Ecology
Lactobacillus brevis is mainly found in food. Environmentally, it can be found on dairy farms in raw milk, especially in bovine feces; however it rarely teams up with eukaryotes in a symbiotic way.(10) Mostly it is found in proximity with other lactic acid bacteria, in a variety of foods. Most commonly these bacteria live in the food where they perform their metabolic pathways. Some major users of lactic acid bacteria include wine, salami, cheese, sourdough bread, pickles, yogurt, cocoa, and coffee. L. brevis specifically accommodates production of sour food.(8) One of the most interesting interactions is that with the human intestinal flora. Lactobacillus, when ingested, has probiotic effects as it improves the immune system.(8) Also, currently there are studies and research conducted on L. brevis and other lactic acid bacteria for incorporation into bio fuels and other environmental friendly and efficient materials.(8)

Pathology
L. brevis is not a pathogen and there are no cases where it has caused or contributed to disease in humans.(4) In fact, it is the other way around. This bacteria is probiotic, and improves the human immune system.(8)

Application to Biotechnology
L. brevis is hop-resistant, a compound that causes beer spoilage. This strain of bacteria is easy to study and research because its optimal growth conditions are known; fastest growth is at 30 degrees Celsius and at a pH of 4-6. Only a few strains of L. brevis are not resistant to hop compounds.(9) Another source reveals that when DNA PCR analysis was conducted on Lactobacillus brevis strains found in breweries, a primer was isolated that can lead to distinguishing spoilage and non-spoilage strains of the lactic acid bacteria. This primer fragment was named hitA, and it can encode proteins that are responsible for a nonATP binding membrane transporter which has been seen in both prokaryotic and eukaryotic species. In a hop bitter environment, the membrane is disintegrated due to a sensitive pH gradient, and H+ is exchanged for divalent cations, Mn2+ being an example. The experiment concluded that the hitA gene could be responsible for making products that are hop bitter resistant due to the introduction of metal ions which in essence destroys the intact proton gradient.(11)

Current Research
1. "Proteomic approach for characterization of hop-inducible proteins in Lactobacillus brevis"

In order for lactic acid bacteria to cause beer spoilage during growth in beer, they must have a resistance to hops. A bacterial hop compound is also known as an ionophore which by definition is a small molecule whose main purpose is to transport ions through the lipid bilayer of a cells membrane. In the experiment, two strains of Lactobacillus brevis (TMW1.465 and variant TMW1.465A) were observed and the acid stress response and hop adaptation caused many detectable changes in the cells properties. To more clearly identify the cause of the changes in metabolism, membrane physiology, and cell wall composition, the scientists set up an experiment to identify related vital proteins despite the disadvantage of an incomplete genome sequence. Conclusively, hop resistance in the variant strain implied altered mechanisms that accounted for intracellular acidification, and changes in mechanisms that dealt with energy generation, genetic information, and enzyme functionality. Enzymes that are hop-regulated were also defined as being manganese or divalent cation dependent. Thus the ability to control the manganese level leads to the ability to tweak the metabolisms of applicable cells. One change found in metabolism in response to an environmental stress condition, was the shift into an energy-saving mode. Lastly this experiment demonstrated that bacteria in a hop stress environment were correlated with the depletion of proton motive force and with a limitation of the divalent cation, which as mentioned above is needed for hop-regulated enzymes. (5)

2. "Cloning, sequencing and expression of a novel glutamate decarboxylase gene from a newly isolated lactic acid bacterium" A new strain of Lactobacillus brevis containing 84.292 mg/L/h of GABA (-aminobutyric acid) was isolated from a Korean fermented food, Kimchi. Lactobacillus brevis OPK-3 was found to have a core fragment of glutamate decarboxylase (GAD) DNA and this fragment was then isolated, cloned, and amplified (by method of PCR) to end up with a full-length piece of core GAD DNA and enough data for a nucleotide sequence analysis. The scientists found the open reading frame (ORF) of the DNA contained 1401 bases and had the ability to encode 467 amino acids for protein assembly. The molecular weight was found to be 53.4 kDa. When compared to the amino acid sequence of other Lactic Acid Bacteria GAD DNA, Lactobacillus plantarum GAD, Lactococcus lactis GAD, and Listeria monocytogenes, the LbGAD ORF was 83%, 71%, and 60% identical, respectively. Further in the experiment, the LbGAD gene was introduced into anEscherichia coli strain (UT481), and when the transformed E. coli was tested, a 53.4 kDa protein was apparent, and the bacteria also had heightened GAD activity. (6)

3. "Characterization of microflora in homemade semi-hard white Zlatar cheese" In this experiment, researchers observed changes in lactic acid bacteria and in the chemical composition of Zlatar cheese, over the course of 60 days during which it was ripening. This cheese was produced

from only nonpasteurized cows milk without any bacterial aid. The results were very interesting; it was found that the while the number of lactobacilli was slowly increasing over the 60 days, the percentage of cocci first decreased, and then began increasing again at the 30 day mark, eventually catching up with lactobacilli. Also, it was found that out of all bacteria, 57 isolates showed antimicrobial activity and were decreasing with cheese age, until no more was detected in samples of 60 day old cheese. In doing replication - PCR analysis, Lactobacillus brevis was found to be one of the main species of bacteria present in addition to Lactobacillus paracasei, Lactococcus lactis, and Enterococcuus faecium. (7)

Appl Environ Microbiol. 2003 January; 69(1): 444451. doi: 10.1128/AEM.69.1.444-451.2003 PMCID: PMC152408

Phylogenetic Diversity of Lactic Acid Bacteria Associated with Paddy Rice Silage as Determined by 16S Ribosomal DNA Analysis
Sad Ennahar, Yimin Cai,* and Yasuhito Fujita
Author information Article notes Copyright and License information This article has been cited by other articles in PMC.

Go to:

ABSTRACT
A total of 161 low-G+C-content gram-positive bacteria isolated from whole-crop paddy rice silage were classified and subjected to phenotypic and genetic analyses. Based on morphological and biochemical characters, these presumptive lactic acid bacterium (LAB) isolates were divided into 10 groups that included members of the genera Enterococcus, Lactobacillus, Lactococcus, Leuconostoc, Pediococcus, and Weissella. Analysis of the 16S ribosomal DNA (rDNA) was used to confirm the presence of the predominant groups indicated by phenotypic analysis and to determine the phylogenetic affiliation of representative strains. The virtually complete 16S rRNA gene was PCR amplified and sequenced. The sequences from the various LAB isolates showed high degrees of similarity to those of the GenBank reference strains (between 98.7 and 99.8%). Phylogenetic trees based on the 16S rDNA sequence displayed high consistency, with nodes supported by high bootstrap values. With the exception of one species, the genetic data was in agreement with the phenotypic identification. The prevalent LAB, predominantly

homofermentative (66%), consisted of Lactobacillus plantarum (24%), Lactococcus lactis (22%), Leuconostoc pseudomesenteroides (20%), Pediococcus acidilactici (11%), Lactobacillus brevis (11%), Enterococcus faecalis (7%), Weissella kimchii (3%), and Pediococcus pentosaceus (2%). The present study, the first to fully document rice-associated LAB, showed a very diverse community of LAB with a relatively high number of species involved in the fermentation process of paddy rice silage. The comprehensive 16S rDNA-based approach to describing LAB community structure was valuable in revealing the large diversity of bacteria inhabiting paddy rice silage and enabling the future design of appropriate inoculants aimed at improving its fermentation quality. Using rice silage as an animal feed has proven economically viable, not only as a way of disposing of rice straw residues but also as a real alternative for feeding livestock in regions where rice is the main crop (13). As a result, in Japan and other rice-producing countries, rice is no longer grown exclusively for human consumption but increasingly as a valuable forage crop. Forage rice is in fact believed to be an ideal alternative crop, not only in helping farmers adjust grain rice production but also in preserving the soil, leading to long-term utilization of the paddy field. Yet a major drawback of forage rice is that it yields low-quality silage, due to poor digestibility of nutrients, mostly crude proteins (40). Several processes have been developed to improve the fermentation and nutritional value of whole-crop silage from forage paddy rice. Breeding programs are carried out, and newly developed rice varieties with increased yield and amount of digestible nutrients are being grown and tested. Also, harvesting, preparation, and storage techniques are constantly being improved. However, most of the processes used to date still rely on heavy chemical treatments with ammonia and sodium hydroxide (10, 20, 23, 40, 41) and were reported to reduce the palatability of silage to ruminants (23). Of the many factors that can affect silage fermentation, the type of microorganisms that dominate the process often dictates the final quality of the silage. For instance, homolactic fermentation by lactic acid bacteria (LAB) is more desirable than other types of fermentation because the theoretical recoveries of dry matter and energy are greatest. During this type of fermentation, LAB utilize water-soluble carbohydrates (WSC) to produce lactic acid, the primary acid responsible for decreasing the pH in silage. In contrast, other fermentations are less efficient. Natural populations of LAB on plant material are often low in number and heterofermentative. Thus, the concept of using a microbial inoculant to silage involves adding fast-growing homofermentative LAB in order to dominate the fermentation, thereby producing higher-quality silage. Some of the commonly used homofermentative LAB in silage inoculants include Lactobacillus plantarum, Lactobacillus acidophilus, Pediococcus acidilactici, andEnterococcus faecium. Commercially available microbial inoculants contain one or more of these bacteria that have been selected for their ability to dominate the fermentation. From a microbiological point of view, to our knowledge, no information is available on the microbial ecology of paddy rice silage (PRS), especially with regard to the indigenous LAB

and their effects during the fermentation process. If identified, the LAB inhabiting PRS could be evaluated for growth kinetics and ability to decrease pH during forage paddy rice fermentation. Selected strains could be examined for their competitiveness as PRS inoculants compared to commercial inoculants. Ultimately, efficacious inoculants could be identified that would lead to an optimization of the microbial action, preserve a high level of crop nutrients, allow chemical treatments to be reduced, and consequently produce a palatable PRS with a high intake potential. The present study set out to screen, isolate, and identify the LAB colonizing PRS at various stages of fermentation, taking particular interest in species that are likeliest to play an important role in the fermentation process. Isolates were identified biochemically, and selected representative strains were identified at the molecular level using 16S rDNA sequence analysis.
Go to:

MATERIALS AND METHODS

Silage samples and bacterial isolates.

Whole-crop rice (Oryza sativa var. Hamasari) grown in a local paddy field (Ohtawara, Tochigi, Japan) was harvested at ripe stage. Silage was prepared in a small-scale fermentation system using the method of Tanaka and Ohmomo (35) modified as previously described (3). A total of 40 silage samples were collected at days 2, 10, 30, and 50 of the ensiling process. The samples (10 g) were blended with 90 ml of sterilized distilled water. Serial dilutions were used for isolation of LAB by using plate count agar (Difco Laboratories, Detroit, Mich.) with bromocresol purple and GYP (glucose, yeast extract, and peptone) agar (17) and for isolation of other microorganisms using the corresponding media as previously described (4, 5). Each suspected LAB colony was purified twice by streaking on MRS agar. The pure cultures were grown on MRS agar at 30C for 24 h, before being transferred to nutrient broth (Difco) with 10% glycerol and stored as stock cultures at 80C for further analysis.
Morphological, physiological, and biochemical tests.

Morphological characteristics and Gram staining of LAB were examined after 24 h of incubation on MRS agar. Catalase activity and gas production from glucose were determined by the methods of Kozaki et al. (17). All PRS isolates assigned to a particular LAB genus or species were identified on API 50 CH strips (bioMrieux, Tokyo, Japan) by using carbohydrate assimilation and fermentation of 49 different compounds (and one control). These were incubated at 30C for 3 to 6 days. The organisms were identified by using the APILAB Plus software version 3.3.3 from bioMrieux and Analytab Products' computer database for comparison of assimilation and/or fermentation patterns.
DNA base composition and sequencing of the 16S ribosomal DNA (rDNA).

Cells grown for 8 h in MRS broth at 30C were used for DNA extraction and purification as described by Saito and Miura (25). DNA base composition was determined, following

enzymatic digestion of DNA to deoxyribonucleosides, by the method of Tamaoka and Komagata (34) using high-performance liquid chromatography. The equimolar mixture of four deoxyribonucleotides in a Yamasa GC Kit (Yamasa Shoyu Co., Ltd., Choshi, Japan) was used as the quantitative standard. Amplification of the 16S rRNA gene was carried out in a ThermalCycler (GeneAmp PCR System 9700; PE Applied Biosystems, Foster City, Calif.) using the PCR method described by Suzuki et al. (33) and reagents from a Takara Taq PCR Kit (Takara Shuzo Co., Ltd., Otsu, Japan). Sequencing was performed twice on both strands by the dideoxy method of Sanger et al. (28), using a PRISM BigDye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems) in combination with an Applied Biosystems model 310A automated sequencing system.
Sequence alignments and phylogenetic inference.

Sequence similarity searches were performed in the GenBank data library using the BLAST program. The sequence information was then imported into the CLUSTAL W software program (Hitachi Software Engineering Co.) for assembly and alignment. The 16S rDNA sequences of RO strains were compared to sequences from type LAB strains held in GenBank (Fig. (Fig.11 and and2).2). Nucleotide substitution rates were calculated (15), and phylogenetic trees were constructed by the neighbor-joining method (26). Bacillus subtilis NCDO 1769 was used as an outgroup organism. The topologies of trees were evaluated by bootstrap analysis of the sequence data with CLUSTAL W software based on 100 random resamplings (36).

FIG. 1.

Phylogenetic tree showing the relative positions of PRS isolates RO1, RO3, RO6, RO17, RO90, and RO95 as inferred by the neighbor-joining method of complete 16S rDNA sequences. Bootstrap values for a total of 100 replicates are shown at the nodes of the (more ...)

FIG. 2.

Phylogenetic tree showing the relative positions of PRS isolates RO5, RO7, RO66, and RO97 as inferred by the neighbor-joining method of complete 16S rDNA sequences. Bootstrap values for a total of 100 replicates are shown at the nodes of the tree. References (more ...)
Nucleotide sequence accession numbers.

The nucleotide sequences for the 16S rDNA described in this report were deposited with GenBank under accession no. AF515219, AF515220, AF515221,AF515222, AF515223, AF515224, AF515225, AF515227, AF515228, and AF515229 for the strains RO97, RO66, RO5, RO7, RO90, RO6, RO3, RO95, RO1, and RO17, respectively.
Go to:

RESULTS AND DISCUSSION

PRS is an increasingly important source of animal feed in Japan and other rice-producing countries. Yet, while LAB in several forage crops and silages have been more or less described (2, 4, 6, 19, 22, 24,32), a similar study has yet to be carried out on PRS. This provides a unique opportunity to study and document the composition of the microbial flora inhabiting PRS and its effects on the fermentation process of the crop. In the present investigation, 400 PRS isolates were screened, of which 161 isolates were considered to be LAB as determined by culture on MRS agar, Gram stain appearance, catalase test, and gas production from glucose. Fifty percent of the presumptive LAB were characterized further by sugar fermentation assays using API 50 CH strips. This led to the delineation of 10 groups of isolates, each one displaying a distinct carbohydrate fermentation pattern (Table (Table1).1). Groups A to E included cocci that did not produce gas from glucose, while group F was represented by heterofermentative cocci. Groups G through J included rods, with those in group G being homofermentative and the others being heterofermentative. The various groups presumably represented six different LAB genera: Lactobacillus,Lactococcus, Pediococcus, Weissella, Enterococcus, and Leuconostoc, which accounted for various percentages of the total. This was in part confirmed by DNA base composition analysis, which showed that most strains had low G+C content within the range of the corresponding genera (Table (Table2):2): RO6 and RO3 from groups A and B with, respectively, 38.0 and 36.9 mol% (Lactococcus range: 38 to 40 mol%); RO95 and RO17 from groups C and D with, respectively, 39.3 and 44.4 mol% (Pediococcusrange: 34 to 43 mol%); RO90 from group E with 39.2 mol% (Enterococcus range: 37 to 45 mol%); RO1 from group F with 40.8 mol% (Leuconostoc range: 38 to 44 mol%); RO7, RO66 and RO97 from groups G, I, and J with, respectively, 45.5, 48.0, and 46.1 mol% (Lactobacillus range: 32 to 53% mol%); and RO5 from group H with 48.5 mol% (Weissella range: 37 to 48% mol%).

TABLE 1.

Phenotypic characteristics of LAB isolated from PRSa

TABLE 2.

G+C content of LAB isolated from PRS In an effort to identify PRS LAB isolates at the species level, molecular phylogeny analysis was conducted and phylogenetic trees were constructed based on the 16S rDNA sequences from evolutionary distances by the neighbor-joining method. Following phylogenetic analysis, strains RO95 and RO17, respectively, representative of groups C and D, were placed in the cluster making up the genusPediococcus, which, with the exception of Pediococcus urinaeequi, was recovered in 100% of bootstrap analyses (Fig. (Fig.1).1). They both formed a distinct cluster together with P. acidilactici and Pediococcus pentosaceus, supported with a bootstrap value of 100%. While RO95 was monophyletic with P. pentosaceus with a bootstrap of 100%, RO17 grouped with P. acidilactici in a 99% bootstrap cluster. The group E strain RO90 was placed in the enterococcal cluster on the phylogenetic tree, with E. faecalisbeing the most closely related species in 100% of bootstrap analyses (Fig. (Fig.1).1). RO1, the representative strain of group F, was unambiguously enclosed in the Leuconostoc phylogenetic group, which is very well delimited with a bootstrap of 100%. The strain distinctly clustered with the species Leuconostoc pseudomesenteroides, as indicated by the high bootstrap value (96%). The representative strains of groups A (RO6) and B (RO3) were clearly identified as Lactococcus lactis strains by forming a very well-defined cluster (100% bootstrap) with this species (Fig. (Fig.1).1). Furthermore, both strains were ascribed to the subspecies lactis on the phylogenetic tree, with a bootstrap of 99% supporting monophyly.

Nevertheless, as reported by Stackebrandt and Goebel (31), the 16S rDNA sequence analysis method, while very good at identifying the organisms by genus and species, cannot differentiate strains at the subspecies level and is therefore not the appropriate method to measure intraspecies relationships. Despite clustering together on the phylogenetic tree, RO3 and RO6 consistently showed discrepancy with regard to L-arabinose utilization (Table (Table1)1) and their 16S rDNA sequences were different even though by only 7 bp. In the same way, strains RO66 and RO97 from groups I and J, which were both phylogenetically identified (100% bootstrap) as L. brevis (Fig. (Fig.2),2), were shown to display marked differences in four fermentation reactions (Table (Table1).1). The group G strain RO7 was clearly assigned to the genus Lactobacillus, since it grouped on the phylogenetic tree together with L. pentosus, L. plantarum,L. paraplantarum, and L. arizonensis in a 100% bootstrap cluster (Fig. (Fig.2).2). Furthermore, RO7 appeared to be equally linked (93% bootstrap) to both L. plantarum and L. pentosus and its 16S rDNA sequence showed a similarity of 99.7% to L. plantarum and 99.8% to L. pentosus. This is not surprising, given that the two species themselves have very similar 16S rDNA sequences that differ only by 2 bp. In fact, it is widely acknowledged that L. plantarum and L. pentosus belong to the same 16S rRNA phylogenetic group and could only be distinguished using phylogenetic analysis of sequences of the 16S-23S large spacer region (12) or partial sequences of the recA gene (37). In the present study, we referred to carbohydrate fermentation patterns that showed unambiguously that the pattern of strain RO7 was an exact match of that of the L. plantarum type strain and different from that of the L. pentosus type strain. In fact, it appeared that, while the L. plantarum type strain produced acid from melezitose, D-raffinose, and -methyl-D-mannoside, the L. pentosus type strain did not. On the other hand, as opposed to L. plantarum, L. pentosus could grow using glycerol or D-xylose as a carbon source (data not shown). All other isolates from group G were similarly identified as L. plantarum. Strain RO5, which is representative of group H, was placed within the Weissella phylogenetic group, which is supported by a 100% bootstrap value (Fig. (Fig.2).2). The two newly described species Weissella cibaria (1) and Weissella kimchii (7), were the nearest phylogenetic relatives to RO5, with similarities exceeding 99%. On the phylogenetic tree RO5 formed, however, a well-defined cluster with W. kimchii, while W. cibariaappeared to be a less related species. Phylogenetic trees based on the 16S rDNA sequence displayed high consistency regarding the relationships between the organisms included. All the nodes leading to RO strain clusters are supported by high bootstrap values and should be considered significant. Also, RO strains were designated to the correct species with close homology. In fact, BLAST search analyses using the whole 16S rDNA sequence resulted in identifications above the 99% level, although in two instances (RO1 and RO95) it was slightly lower. All homologies displayed were, however, well above 97%, which is considered to be the cutoff value indicating species identity (31). On the other hand, genetic characterization of PRS isolates correlated well with their biochemical classification, which was obtained at percentages of identification of more than 99% using the APILAB Plus computer program. Only strain RO1

(Leuconostoc pseudomesenteroides) was biochemically misidentified as Leuconostoc mesenteroides, and strain RO5 (W. kimchii) could not be identified biochemically to the species level. Following biochemical and phylogenetic analyses, PRS isolates fell within well-recognized groups of LAB and, for the majority of them, were clearly related to particular species. Most of the LAB characterized belonged to the genera Lactobacillus, Lactococcus, and Leuconostoc, with L. plantarum (24%), L. lactis(22%), and L. pseudomesenteroides (20%) totaling two-thirds of the entire LAB flora. Yet a great species diversity was observed since five other species were also identified: P. acidilactici (11%), L. brevis (11%),E. faecalis (7%), W. kimchii (3%), and P. pentosaceus (2%). Furthermore, homofermentative species accounted for two-thirds (66%) of the total LAB microflora. LAB species identified in PRS are common inhabitants of a variety of forage crops and silages. This is in agreement with the results of our previous investigations (2, 4, 6), as well as those of other authors (19,22, 24, 32), which showed that the natural fermentation processes in crop silages, including corn, sorghum, alfalfa, Italian ryegrass, alfalfa, and guinea grass silages, are dominated by species ofLactobacillus, Pediococcus, Leuconostoc, and Lactococcus. Particularly, the prevalent species in PRS, L. plantarum, is reportedly the most dominant during the fermentation of crop silage (6). Other LAB, such as leuconostocs and pediococci, have often been found at low frequencies together with Weissella strains on plant material and in various silages (4, 16, 19). They are also widely used as starter cultures or control barriers for food pathogens in vegetables (38). The presence of E. faecalis suggests a fecal origin of some PRS microorganisms. The natural habitat of lactococci is milk, but L. lactis subsp. lactis has been isolated previously from plants, vegetables, and cereals (9, 27). Silage fermentation begins with aerobic respiration immediately upon harvest of the plant. During this phase, WSC are converted to CO2, heat, and water by both plant cells and aerobic microbes (4, 19). This process will continue until the oxygen is depleted. Once anaerobic conditions are established, the anaerobic fermentation begins with the growth of heterofermentative LAB. More nutrients are retained in the silage when this phase is rapidly completed. As the pH of the ensiled mass falls below 5, the growth of heterofermentative bacteria is inhibited and there is a shift towards increasing populations of homofermentative LAB (4, 19). Rapid growth and acidification by these bacteria are the basic aim of the silage fermentation so that sugars are efficiently converted to lactic acid, leading to suppression of nonbeneficial microbes (by both direct competition and lowering of the pH). This ultimately leads to improvements in preservation of plant proteins in a digestible form, in aerobic stability of the silage during the feedout phase, and in nutritional quality in relation to intake levels and livestock productivity (8). However, the rate and extent of final pH in the ensiled forage depend largely on the type and moisture of the forage being ensiled. As opposed to commonly ensiled crops, such as corn, with paddy rice, low WSC levels (1 to 4%) and a high buffering capacity are less conducive to rapid acidification by epiphytic LAB populations and hence to adequate ensilage in the absence of additives

(10, 13, 23, 40). Moreover, under farm conditions, the populations of epiphytic LAB are not always large enough or do not have a composition suitable for promoting efficient homolactic fermentation (11). Thus, efforts to obtain PRS that has higher nutritional value and good storage properties have to be pursued with the aim of developing additives that stimulate and direct the fermentation process. Many lab-scale fermentation studies have been conducted to evaluate the effect that inoculants have on the fermentation of various crops. In general, favorable improvements in the rate of pH decline and increased lactic acid levels have been noted with legumes, grasses, and cereal silages. In previous investigations (3, 4, 6), we showed that the inoculation of various forages with homofermentativeLactobacillus spp. (predominantly L. plantarum), in combination with Pediococcus, Enterococcus, orLactococcus spp., had beneficial effects, which is in agreement with other authors' studies (14, 29, 30,39). In these studies, selective inoculants have been shown to (i) hasten the fermentation process and reduce nutrient loss, (ii) improve fiber digestibility, and (iii) reduce protein degradation, all of which yield silage with higher nutritional value. However, a recent literature review found that only a few of the published trials reported positive effect from silage inoculation, with only 28% of studies showing an increase in silage intake and only 47% showing an increase in milk production (18). Given this scientific evaluation of a 1-in-4 chance of intake improvement and a 50:50 chance of improved animal performance, it is not surprising that many producers remain unconvinced that silage inoculants can offer any real economic benefit. The effectiveness of silage inoculants depends, among other things, upon the quality (growth rate and environmental adaptability) and quantity of the microorganisms used. Research trials have in fact shown that individual strains differ in their crop preference, ability to ferment various substrates, growth potential in various moistures and temperatures, and ability to enhance fiber digestibility (19, 21, 22). However, since most research on silage fermentation is of a strategic or applied nature, little information is available on the complex microbial and biochemical processes involved. As far as PRS is concerned, it is safe to assume that homofermentative LAB isolated in this study are adapted to the particular ecosystem of rice fermentation and could therefore be suitable as inoculants for silage making from paddy rice. For instance, the greater frequencies of recovery of L. plantarum and P. acidilactici strains from PRS samples than from paddy rice forage samples (data not shown) suggested that these species participated actively in the fermentation process. Trials are on the way in our laboratory in order to develop suitable crop-specific additive combinations by screening potentially useful PRS homofermentative LAB strains for desirable characteristics in laboratory media and then performing ensilage experiments. In the presence of an adequate supply of available nutrients at an early stage, the selected competitive LAB strains should be able to compete with the epiphytic PRS microflora and ensure a highly cost-effective means of improving the quality of silage.