Technical field of the invention

The present invention relates to probiotic bacterial Lactobacillus plantarum strains isolated from quinoa, said strains being selected from Lactobacillus plantarum ChB11 (LMG P-31891), Lactobacillus plantarum ChG33 (LMG P- 31892), Lactobacillus plantarum ChR228 (LMG P-31893) and Lactobacillus plantarum ChJ239 (LMG P-31894). The invention also relates to a composition such as a food or feed composition comprising said strains and to a method for preparing a food composition such as quinoa milk by using said strains.

Background Art

To the Bolivian indigenous people Quinoa (Chenopodium quinoa Willd.) has been one of the most important crops for thousands of years. During the last two decades quinoa has been playing an increasing role in human diets worldwide. Quinoa is cultivated in the agricultural part of the vast Andean region in Bolivia, located 4000 meters above the sea level. The Andean region has harsh environmental conditions including intense UV radiation, seasonal drought and occasional frost. Consumption of seeds is the most common use of quinoa and they have a high nutritional value containing, several vitamins, minerals, oils, high content of proteins, starch, and essential free amino acids. In addition quinoa seeds are potential source of phenolic compounds with antioxidative capacity. Because of all the health’s benefits that quinoa has, this pseudo cereal is being consumed by the European population including people suffering from celiac disease.

Lactobacillus plantarum and Lactobacillus brevis have been reported as dominant species bacteria isolated from spontaneous fermented quinoa sourdough in Argentina, the quinoa grain was imported from Bolivia. However, it should be necessary to considering that on spontaneous fermentation the naturally proliferation of native bacteria is an unpredictable process, conditioned by the growth of the dominated genus present in the substrate. Induced fermentation by Lactobacillus plantarum CRL 778 was used for ferment quinoa slurries, increasing the bioavailability of the nutrients. The fermented quinoa slurry can be used in the bakery industry. The content of probiotic bacteria on spontaneous or induced fermentation depends on the pH of the medium, and acidity in the substrate. Generally the pH should be lower than 4 to guarantee that no pathogenic bacteria are present, and the acidity is reported as a function of percentage of lactic acid synthesized by the bacteria present in the substrate.

In 2015 a research article entitled “Biodiversity and technological-functional potential of Lactic acid bacteria isolated from spontaneously fermented quinoa sourdough” by L. Ruiz Rodriguez et. al. was published in the Journal of Applied Microbiology. L. Ruiz Rodriguez et. al. used back slopping for the fermentation process and isolated Lactobacillus plantarum, Lactobacillus brevis and Pediococcus pentosaceous from the quinoa grains. The methodology used by the group can however change the microbiology environment, since is not possible to ensure which species, strains and/or quantity of microorganism had been transferred. This mechanism can therefore be interpreted as induced fermentation instead of spontaneous fermentation. The study supports the presence of lactic acid bacteria (LAB) and identified isolated species at genotypical level but not at phenotypical level.

Lactic acid bacteria are considered as GRAS (generally regarded as safe) and the microorganisms are commonly used by the Food and Pharmaceutical Industry as starter cultures for fermentations and as probiotics.

The genus Lactobacillus consist of different species and subspecies, with a catalogue of 90 different Lactobacillus spp. so far sufficiently characterized. To simplify the catalogue of Lactobacillus spp. these have been divided into three different subcategories depending on their fermentation abilities. The facultative heterofermentative (Group II) subcategory has the ability to ferment hexoses into lactic acid and in addition to this, pentoses and/or gluconate can also be fermented. Strains of the species Lactobacillus plantarum belong to either one of the two remaining subcategories: Obligate homofermentative (Group I) which can only ferment hexoses into lactic acid and obligate heterofermentative (Group III) which ferments hexoses into lactic acid, acetic acid and/or ethanol and carbon dioxide.

In many ways Lactobacillus plantarum is separated from other Lactobacillus spp., e.g. in the unusual high tolerance towards environmental stress that the species displays when surviving the passage through the low pH environment of the human stomach.

Lactobacillus plantarum is thought to have a higher adaptive ability towards a variety of different conditions than other Lactobacillus spp., something that can be credited its unusually large genome. Also, high levels of intracellular manganese reducing oxygen free radicals to hydrogen peroxide and eventually oxygen and water make Lactobacillus plantarum more resistant towards oxygen poisoning than other Lactobacillus spp. Lactobacillus plantarum has furthermore the ability to ferment a wide variety of different carbohydrates and is frequently found in the human Gl-tract. The observation is unsurprising since it is found and spontaneously multiply in most lactic acid fermented foods, with one example being sourdough. Adult humans acquire immune system tolerance towards harmless, food-associated bacteria, revealing a close relationship between human immune reactions and consumption of Lactobacillus plantarum. Strains belonging to the species Lactobacillus plantarum is also generally regarded as safe for consumption and the species does not contain any pathogenic strains.

There is a need on the planet for continuous development of healthy, vegetable and/or vegan alternative foods. For instance, the dietary vegeterian and vegan population is steadily growing throughout the world and there is a growing demand to reduce the consumption of dairy and animal products. The European population including people suffering from celiac disease is also increasing and demands new alternatives. Therefore, there is need to find new healthy alternatives to this part of the growing population, which are both healthy and have a pleasant taste.

Summary of the invention

The present invention relates in one aspect to a probiotic bacterial

Lactobacillus plantarum strain isolated from quionoa grains selected from the group consisting of Lactobacillus plantarum ChB11 having accession number LMG P-31891 , Lactobacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P-31894.

In another aspect, the present invention relates to a composition or a food compositions comprising a probiotic bacterial Lactobacillus plantarum strain selected from the group consisting of Lactobacillus plantarum ChB11 having accession number LMG P-31891 , Lactobacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P-31894.

In another aspect, the present invention relates to a method for the preparation of a quinoa milk comprising the steps: a) toasting quinoa grains; b) adding water to the toasted grains; c) mixing the water and the grains with a blender and filtering the mixture; and d) inoculation the mixture with at least one probiotic strain selected from Lactobacillus plantrum ChB11 having accession number LMG P- 31891 , Lactoacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P-31894 at 30°C for 48 h in anaerobic conditions.

Description of the figures

Figure 1. Microbiota characterization of quinoa sourdough. The microorganisms were isolated from Violet Red Bile Dextrose agar (red), Tryptic Soy agar (orange), Malt Extract agar (green) and Rogosa agar (blue).

Definitions

The term “probiotic strain” as used herein means a live microorganism (herein Lactobacillus plantarum stain) that provides health benefits when consumed by an individual, usually within the gastrointestinal tract. The term “cfu” means colony forming units and is a generally used unit in microbiology to estimate the number of viable bacteria cells in a sample. The number of cfu in a fermented drink is usually above 10 ⁸ CFU.

Detailed description of the invention

As described above the present invention relates to a probiotic bacterial Lactobacillus plantarum strain selected from the group consisting of Lactobacillus plantarum ChB11 having accession number LMG P-31891 , Lactobacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P- 31894, wherein said strains have been isolated from quinoa. There is need for alternative healthy products to the growing population of vegeterians and vegans. In addition, the population with celiac disease also requires new and healthy alternatives.

Based on the analyzes of the results as have been performed herein, it is concluded that the 4 Lactobacillus plantarum strains showed to have potential to be used as probiotic based on their metabolic scope, the ability to counteract potential pathogenic microorganisms and because Lactobacillus spp. are considered as safe bacteria. Additionally, the most remarkable fact was that two strains able to ferment Xylose are Lactobacillus plantarum. Those were named as Lactobacillus plantarum ChB11 and Lactobacillus plantarum ChG33. Besides, the other two strains identified as Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228 are different from each other.

The Lactobacillus plantarum strains isolated according to the present invention have been identified at genotypical and phenotypical level. It has been demonstrated that two of the Lactobacillus plantarum strains (Lactobacillus plantarum strains ChBU and Lactobacillus plantarum ChG33) are capable to ferment xylose, a pentose carbohydrate that is otherwise exclusively metabolized by Lactobacillus pentosus. The enzymatic action over xylose is used as a strong test to distinguish Lactobacillus pentosus from the rest of the Lactobacillus spp. A composition comprising at least one bacterial strain is also provided herein, for instance a food composition comprising at least one bacterial strain is also provided. With the growing demand for vegeterian alternatives a food composition compring at least one of the Lactobacillus plantarum strains as disclosed herein may be provided. Said food composition may be a fermented food composition such as a fermented vegetable-based beverage, functional food, food additive, dietary supplement, or nutritional product. Said fermented food composition may be fermented with at least one of the strains selected from the group Lactobacillus plantarum ChB11 having accession number LMG P-31891 , Lactobacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P-31894.

Said fermented vegetable-based beverage may for instance be a fermented quinoa milk. It is understood that the Lactobacillus plantarum strains as disclosed herein may be used for fermenting any other vegetarian beverage based on any other vegetables such as amaranth, rice, oat or wheat. Since two of the isolated strains have the capability of fermenting xylose, it would particularly suitable to ferment vegetables crops, eg quinoa, containing contents of xylose which otherwise may be difficult to ferment by other strains of Lactobacillus plantarum. Thus, healthy fermented products containing nutrients released during fermentation may be provided. It has been shown in the experimental part herein that Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239 showed high inhibition against potential pathogenic bacteria ensuring that the final fermented quinoa milk was free of those microorganism after 48 hours of fermentation. Besides, the amount of L. plantarum remained stable in quantity during storage time. In comparison to the fermentation of quinoa milk with the commercially available Lactobacillus plantarum299v strain, the microorganism Enterococcus mudtii was still present in a considerable amount compromising the hygiene quality of the final quiona milk product. The Lactobacillus strains ChB11 , ChG33, ChJ239 and ChR 228 totally inhibit the growth of Enterococcus mudtii safeguarding the hygiene quality of the formulated fermented quinoa milk.

Said composition or food composition as disclosed herein may comprise said at least one strain present in the composition in an amount of from about 1 x10 ⁶ CFU/day to about 1 x 10 ¹⁴ CFU/day, for instance 1 x10 ⁸ CFU/day to about 1 x 10 ¹² CFU/day, e.g. about 1 x10 ⁹ CFU/day, 1 x10 ¹ ° CFU/day or 1 x10 ¹¹ CFU/day. The at least one strain may be present in the composition as attenuated, inactivated, alive or dead.

A feed composition is also provided herein, which composition is suitable for animals to consume. For instance, at least one of the two bacterial strains Lactobacillus plantarum strains ChB1 1 and Lactobacillus plantarum ChG33 which are capable to ferment xylose would be suitable to add to any xylose containing cereal or vegetable. The strains could ferment such xylose containing cereal or vegetable to metabolize the xylose present in the cereal or vegetables.

The present invention also relates to a method for the preparation of a quinoa milk comprising the steps: a) toasting quinoa grains, for instance in an oven or on a stove at above 145°C; b) adding water to the toasted grains in a proportion of about 1 :8; c) mixing the water and the grains with a blender to provide a mixture and filtering the mixture; and d) inoculation the mixture with at least one probiotic strain selected from Lactobacillus plantrum ChB1 1 having accession number LMG P- 31891 , Lactoacillus plantarum ChG33 having accession number LMG P-31892, Lactobacillus plantarum ChR228 having accession number LMG P-31893 and Lactobacillus plantarum ChJ239 having accession number LMG P-31894 at 30°C for 48 h in anaerobic conditions.

A benefit of the method of the preparation of the milk is that there is no need to add any sugar such as carbohydrates to the mixture. There already is carbohydrates available from the avaiable quiona grains. In addition, stabilizers are not needed to add either. In addition, the quinoa milk as prepared has the advantages of providing a good taste, i.e. the previous bitter taste has been reduced. This is because eg that saponins are removed by the washing step. The toasting step also assists in providing a good taste.

Deposited strains

All strains were deposited at BCCM/LMG (Belgian Coordinated Collections of Micro-organisms/Laboratorium voor Microbiologie, Universiteit Gent (UGent)), Gent, Belgium on July 2nd 2020. The depositor is Asa Hakansson. The accession numbers are as follows; Lactobacillus plantarum ChB11 having accession number LMG P-31891 ; Lactobacillus plantarum ChG33 having accession number LMG P-31892; Lactobacillus plantarum ChR228 having accession number LMG P-31893; Lactobacillus plantarum ChJ239 having accession number LMG P-31894.

Other objectives, features and advantages of the present invention will appear from the following detailed disclosure, from the attached claims, as well as from the figures. It is noted that the invention relates to all possible combination of features.

Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to “a/an/the [element, device, component, means, step, etc.]” are to be interpreted openly as referring to at least one instance of said element, device, component, means, step, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.

As used herein, the term “comprising” and variations of that term are not intended to exclude other additives, components, integers or steps. Experimental part

An aim of the experimental part is to report the isolation and identification of 4 autochthonous Lactobacillus plantarum strains from the Bolivian white quinoa grains. The method used were planned and applicated in order to distinguish between Lactobacillus plantarum and Lactobacillus pentosus, two closely related strains. Methods such as, polymorphic chain reaction (PCR), genetic sequencing by Eurofins Genomics in Germany, and edition of the results using Sequence Alignment Editor (BioEdit) was used to identify the strains. Randomly applied polymorphic DNA (RAPD) was applied for identification at genotypical level, PCR reactions with specific primers for identification of Lactobacillus pentosus were tested, API 50CHL was used to identify the bacteria at phenotypical level besides the tannase activity test and the identification of phenolic compounds released from the quinoa grains after fermentation. The interpretation of the results were analyzed and used to conclude the identity between Lactobacillus species.

Examples

By way of examples, and not limitation, the following examples identify embodiments of the present invention.

Materials and Methods

Quinoa sourdough preparation

Six samples were prepared mixing the quinoa grains with demineralized water (1 :2.5 v/v) using a blender (Electrolux, Great blending TruFlowTM blades, ESB5400BK) to obtain a homogenous semiliquid dough distributed into a 500 mL glass flask (Reagent bottle with Screw Cap, clear). The presence of oxygen was minimized by filling the glass flasks until maximum capacity and sealing them hermetically. The incubation was set up at 30°C for 8 days. Samples were withdrawn daily, and the pH was monitored at the same time using a Metrohm 744 pHmeter (Metrohm Ltd, Herisau, Switzerland), previously calibrated according to the manufacturer recommendations. The demineralized water and the glass bottles were autoclaved at 121 °C for 15 min and stored at 4°C. The blender was washed with tap water and rinsed thoroughly with ethanol 70%, before to be used. Microbiota characterization by plate count

The samples were collected, promptly tenfold diluted and plated by duplicate. In general, 10 grams, per bottle, were withdrawn every day and transferred into 90 mL sterile bacteriological peptone water and homogenized by vortex. Serial dilutions were made and obtained samples were plated (100 μL) on: Rogosa agar (Oxoid) incubated anaerobically (Gas Pack Anaerobic system, BBI, Becton Dickinson and company, USA) at 37°C for 72 h for Lactobacillaceae, Violet Red Bile Dextrose Agar (VRBD, Merck, Germany) was incubated aerobically at 37°C for 24 h for Enterobacteriaceae, Tryptic Soy Agar (TSA, Fluka, Missouri, USA) was incubated aerobically at 30°C for 72 h for total aerobic count, and Malt Soy Agar (Sigma Aldrich, India) was incubated at 26°C for 7 d for yeast and fungus. The number of viable cells were recorded, and from the appropriated dilution (20 ≥ 200), ten colonies were randomly picked each day. The cells were streaked onto clean plates and incubated following the procedure previously explained for each agar. Following purification, the isolates were re-cultured into 5 mL Man Rogosa and Sharpe Broth (MRS-Broth, Merck, Germany) and incubated at 37°C overnight (max. 18 h) for the multiplication of Lactobacillaceae. The viable cells from TSA and VRBD agar were re-cultured in Tryptic Soy broth (TSB, Sigma Aldrich, India) incubated at 30°C for 48 h, finally Malt Broth (Fluka, Germany) was incubated at 26°C for 5 d. The cells were washed twice with a 0.85% sterile NaCI solution (8263 xg for 5 min). The clean cells were stored in freezing media (4·28 mM-K2HPO4, 1· 31 mM-KH2PO4, 1 ·82 mMNa- Citrate, 0·87 mM-MgSO4-7H2O and 1 ·48 mM-98 % glycerol) at -80°C. (Fak et al., 2012). Type strains

Three type strains used as reference: Lactobacillus pentosus CCUG 33455T (also register as ATCC 8041 by the American Type Culture Collection), Lactobacillus paraplantarum CCUG 35983T, and Lactobacillus plantarum ATCC 14917 (also register as CCUG 30503T) were purchased from the Culture Collection University of Gothenburg, Sweden. DNA extraction

Fresh cells multiplied in broth were used to extract the DNA applying the glass bead beating method. Briefly: 200 μL was transferred into 1.5 mL tubes and centrifuged at 20.8xg for 3 min (Thermo Scientific, HeraeusPico21 , centrifuge). The pellet was cleaned with 500 μL of 0.85% NaCI (g/L) sterile solution followed by a second clean with 500 μL of sterile milli-q water. The pellet was reconstituted on 500 μL of sterile milli-q water and 8 to 10 glass beads (0.2mm diameter) were added. The cell wall was broken using an Eppendorf Mixer 5432 (Eppendorf, Hamburg, Germany) for 45 min at 4°C. The DNA was separated from the pellet by centrifugation (20.8xg, 1 min). Identification of the microorganisms by 16S rRNA gene sequencing The 16S rRNA genes were amplified using ENV1 (5'- AGA GTT TGA TII TGG CTG AG -3', Escherichia Coli, 8-27 bp) and ENV2 (5'- CGG ITA CCT TGT TAC GAC TT -3', Escherichia coli, 151 1 -1492 bp) as forward and reverse primers. A total volume of 25 μL containing 2.5 μL template DNA and 22.5 μL of PCR master mix (0.2 mM of both primers, 2.5 μL of 16S PCR reaction buffer with 1.5 mM MgCI2 [Roche Diagnostic GmbH, Mannheim, Germany], 200 mM of each deoxyribonucleotide triphosphate [dNTP, Qiagen, Germany], and 2.5 U of Taq DNA polymerase [Roche Diagnostic, Mannheim, Germany] and nuclease free water [Promega, Sweden] to completed the final volume). Amplification was achieved in a PCR Mastercycle 5333 (Eppendorf) according to the following profile: incubation at 94°C for 3 min, denaturation at 94 °C for 60 s; annealing temperature at 50 °C for 45 s and elongation at 72°C for 120 s for 1 cycle. A total of 30 cycles were performed with an extension step at 72°C for 10 min, thereafter, cooled down to 4°C. To confirm the amplification products, 1 .5 μL of PCR product mixed with 1 μL dye (6X Orange DNA Loading Dye) were gel electrophoresed on 1 .5%, (w/v) agarose gel (Sigma) in TAE buffer (50X, VWR Chemicals, USA, pH 8.3) using as molecular weight markers 100 bp (GelPilot 100bp Plus Ladder, Qiagen). The gels were stained on GelRedTM bath (30 μL [Biotium, USA] dissolved in 100 mL distilled water) for 15 min. The DNA bands were observed through UV chamber (Transilluminator UVP, USA). The PCR products were sequenced at Eurofins Genomics (Ebersberg, Germany) on an ABI 3130x1 Genetic analyser (Applied biosystems, Foster City. CA, USA) using ENV1 as sequencing primer. The sequenced genes were edited by BioEdit Sequence Alignment Editor (Michigan State University, USA) (Hall, 1999) and the resulted genes were submitted to RDP data base (http://rdp.cme.msu.edu.) in order to obtain the identity of the microorganism confirmed by the percentage of similarity. Randomly Amplified Polymorphic DNA (RAPD)

P73 (ACGCGCCCT) containing 80% of G+C was chosen as primer and was previously described by Johansson and Quednau. A total volume of 50 μL solution composed by 2 μL templated DNA and 48 μL of master mix (0.2 mM of each deoxyribonucleotide triphosphate [dNTP, Qiagen, Germany], 2.5 μL reaction buffer with 1.5 mM MgCI2 [Roche Diagnostic GmbH, Mannheim, Germany], 0.8 mM primer [Eurofins Genomics, Germany] and 2.5 U of Taq DNA polymerase [Roche Diagnostic, Mannheim, Germany] and nuclease free water [Promega, Sweden]) were centrifuged and covered with mineral oil. Four cycles at the temperature profile: 94 °C, 45 s; 30 °C, 120 s; 72 °C, 60s; followed by 26 cycles at 94 °C, 5s; 36 °C, 30s; 72 °C, 30s. The extension step was at 75 °C for 10 min and cooled down to 4 °C. The amplified products were electrophoresed on 1 .5 % (w/v) agarose gel using DNA Molecular Weight Marker VI 0.15-2.1 kbp (Roche Diagnostic GMbH Mannhem, Germany) as ladder. The gels were photographed using Panasonic DMC- LX100 camera, and the pictures were transferred to the computer by Panasonic imagen app.

PCR specific primers for Lactobacillus oentosus spp, gene amplification Two pair primers specific to distinguish Lactobacillus pentosus were used separately on a final volume of 25 μL master mix containing 2.5 μL of template DNA and 22.5μL PCR master mix (0.25 mM of each primer [Eurofins genomic], 2.5 μL of 16S PCR reaction buffer with 1 .5 mM MgCI2 [Roche Diagnostic, Mannheim, Germany], 1 ,5 μL of MgCI2 [25mM, Qiagen Germany], 200 mM of each deoxyribonucleotide triphosphate [dNTP, Qiagen, Germany], 2.5 U of Taq DNA polymerase [Roche Diagnostic, Mannheim, Germany] and nuclease free water to complete the final volume). The recA gene was amplified according to the method developed by Torriani et al., using the primers pentF (5'- CAG TGG CGC GGT TGA TAT C -3', forward) and pREV (5'- TCG GGA TTA CCA AAC ATC AC - 3', reverse) and a temperature gradient settled in a PCR Mastercycler 5333 (Eppendorf), described as follows: incubation at 94 °C for 3 min, denaturation at 94 °C, 30 s; annealing at 56 °C, 10 s; and elongation at 72 °C, 30s; final extension at 72 °C for 5 min during 30 cycles and cooled down to 4 °C (Torriani and Felis, 2001 ). The 16S rRNA region was amplified using the base pair primer 16S (5'- GCT GGA ATC ACC TCC TTT C - 3', forward) and Lpe (5'- GTA TTC AAC TTA TTA CAA CG - 3', reverse) designed by Berthier et al. and, the PCR reaction was settled in a PCR Mastercycler gradient (Eppendorf) during 30 cycles starting with an incubation at 94 °C, 5 min; denaturation at 94 °C, 1 min; annealing at 53 °C, 1 min; and elongation at 72 °C, 1 min and cooled down to 4 °C (Berthier and Ehrlich, 1998). The amplified regions were observed on 2% (w/v) using 100bp plus molecular weight. The gels were stained on GelRedTM bath for 15 min, as previously described on section 3.5., and the gels were photographed as described on section 3.6. Taxonomic identification using API 50 CHL Fermentation Assay

The capacity of Lactobacillus spp., strains to ferment different carbohydrates was evaluated using API 50CH strips and API 50CHL medium as inoculum (API System, bioMerieux Marcy-l’Etoile, France) according to the manufacturer instructions. Overnight cultures of lactobacilli isolates were grown in 10 ml MRS broth at 37oC. The multiplied cells were washed with sterile physiological saline solution (0.9 % w/v of NaCI) and centrifuged (20.8xg, 5 min), and the pellets were suspended on API 50 CHL medium and vortexed prior to transferring the mixture to the 50 wells on the API 50CH strips (the well 0 served as a control). To generate anaerobic conditions all wells were overlaid with sterile paraffin oil (Merck, KGaA, Germany) and to enhance humidity the strips were moistened with water, previously autoclaved, covered and incubated at 37°C (Boyd et al., 2005). Changes in color from violet to yellow (positive) or blue (negative) were monitored after 1 , 2 and 7 days. The results were graded as complete change to yellow by 1 or blue by 0. The well 25 (Esculin Ferric Citrate) is considered as positive if turns black. The results were analyzed using APIWEBTM (https://apiweb.biomerieux.com).

Lactobacillus spp., tannase enzymatic capacity

The capacity to degrade tannins and decarboxylate gallic acid was evaluated based on the colorimetric method developed by Osawa et al. (Osawa et al., 2000) with minor’s modifications described below. The colors of the solutions were judged visually and confirmed spectrophotometrically on the UV/VIS range. The reactions were conducted in darkness and negative controls containing all the elements, but the bacterium was added to the analysis. Methilgallate degradation

The cells were re-culture on Rogosa agar (anaerobically, 37 °C, 72 h). The fresh cells were harvested using a 10 μL loop and transferred into 1 mL nutrient solution with pH = 4.6 (33 Mm/L of NaH2PO4 [Merck, Darmstadt, F. R., Germany] and 20mM/L of methilgallate [Sigma Aldrich, USA]) and incubated at 37 °C for 24 h. Afterwards, the samples were alkalinized adding 1 mL of a saturated solution at pH 8.6 of NaHCO3 (Merck, Darmstadt, Germany) and expos oxygen for 1 h at room temperature. The color was visually evaluated before and after alkalinisation. Additionally, the maximum absorbance at 440nm using a microplate reader (SPECTROstartNano, BMG LABTECH, Germany) was measured by triplicate using 360 μL of sample per well. The color of the solution of the negative control was uncolored.

Gallic acid decarboxylation

From the freezing media, 50 μL were transferred into 5 mL of MRS Broth and incubated overnight (12-18 h, 37 °C). After cell multiplication, the tubes were vortexed and 50 μL were transferred into 10 mM/L final concentration of gallic acid (Sigma Aldrich, 3,4,5-Trihydroxybenzoic acid monohydrated) in 10 mL of MRS broth and incubated at 37°C for 72 h. Thereafter, the tubes were vortexed, and 2 mL aliquot were alkalinized with equal amount of saturated solution of NaCHO3 (pH 8.57). The samples were incubated at 37°C for 1 h aerobically. The change in color to orange or brown were judged as positive reaction otherwise the solution became dark blue or dark green, the same as the negative control.

Statistical calculations

SigmaPlot version 14.0 (SYSTAT Software, Point Richmond, USA) was used for the statistical analysis. The number of viable count and pH between samples were evaluated by Kruskal Wallis One Way Analysis of Variance (ANOVA) on Ranks or a Mann-Whitney Ranks Sum test when required. Results are presented as median and interquartile range and p-values < 0.05 were considered significant.

Results

Quinoa sourdough preparation

The appearance of the liquid sourdough did not present sedimentation but in contrary the viscosity of the mixture increased. Resistance to flow was observed compared to before fermentation. The pH decreased from 6.20 (6.19 - 6.22) to 4.34 (4.21 - 4.36) after the first 24 h of incubation and this is the first change on the pH values statistically significant (p = 0.002). Thereafter, a continuous decrease without a significant statistical difference was shown until the fourth day (pH = 4.06; 4.04 - 4.08) and fifth day (pH = 3.98; 3.93 - 4.18) were the pH dropped down slightly below 4, becoming evident a second interval of time were the change on the pH was statistically significant (p = 0.002). Consecutively, the upcoming days the pH remined statistically stable. See table 1 .

Microbiota characterization by plate count

The samples were incubated at the same time and samples were withdrawn every 24 hours and plated on Rogosa, TSA and VRBD agar immediately after being tenfold diluted. The initial pH was 6.42 and decreased to 4.10 after 192 hours (8 days). No presence of mold or yeast growing over the quinoa grain was observed. The odor of the samples can be described as bitter and the color of the medium changed from uncolored to light yellow. The autochthonous microorganisms increased in amount while the pH decreased during the time elapses. Before fermentation the number of Enterobacteriaceae and Lactobacillaceae were below the limit of detection (< 1 ). Nevertheless, the number on total aerobic count was 6.09 (6.04-6.17) log CFU/g, and for yeast and mold was 4.55 (4.39-4.60) log CFU/g. The colonies were diverse in shape, size, and color, varying between white and yellow. After 24 hours of fermentation a sharp increase of the number of viable cells were registered for all the agars compared with time zero. The number of viable cells of Enterobacteriaceae was 7.24 (7.13-7.34) log of CFU/g, total aerobic count was 9.35 (9.24 -9.40) log CFU/g, Lactobacillaceae was 6.22 (6.05-6.30) log CFU/g, and for yeast and mold it was 9.22 (9.08-9.44) log CFU/g. Through the fermentation time, the native microbiota community continued to increase and changed gradually. It became noticeable, that on Tryptic soy agar and Malt agar there was a loss of the microbial diversity. The shape, size, and color of the colonies were mostly homogeneous.

Furthermore, an equilibrium point between the number of log CFU/g on Tryptic soy agar, Rogosa and Malt agar was registered after 96 hours of fermentation with p = Malt-Rogosa = 0.987; p = Malt-TSA = 0.908; p = TSA - Rogosa = 0.908, and no statistical difference was found between the three pairs pursued for a dominance presence by the Lactobacillaceae family at the same time as a decreased number of VRBD viable cells was registered. Enterobacteriaceae became undetectable (below the limit of detection) after 120 hours of fermentation, 0.00 (0.00-1.23) log CFU/g (table 1). Table 1 . Daily progression of pH and microorganisms through the incubation time of quinoa semiliquid sourdough

Detection limit <1

Identification of the microorganisms by 16S rRNA gene sequencing

A total of 1400 viable cells were isolated and divided as 420 isolates from Tryptic Soy agar, Malt Extract agar and Rogosa and 240 viable cells isolated from Violet Red Bile Dextrose agar. The percentage of similarity to consider the identity of the microorganism was between 0.999 - 1 percent for Lactobacillaceae and between 0.997 - 1 percent for the rest of the microorganisms. The presence of the same bacterial strains was found on more than one agar and their detection depended on time and sample dilution. For example, Stenotropomonas maltophilia was isolated from Violet Red Bile Dextrose agar for a period of four consecutive days while from Malt extract agar they were detectable for two days. However, a correlation of time occurred between the first 24 and 48 hours of fermentation for Violet Red Bile Dextrose agar and Malt Extract agar. Violet Red Bile Dextrose agar plated samples were less diluted compared to Malt agar as the fermentation process proceeded. Klebsiella michiganensis and Klebsiella oxytoca were able to grow on Tryptic Soy agar, Malt extract and Violet Red Bile Dextrose agar with an overlapping time of one day. A similar correlation was observed for the Lactobacillaceae family. Those lactic acid bacteria were isolated from Rogosa, Tryptic Soy agar and Malt Extract agar at the highest dilution. The identified microorganism were Pediococcus pentosaceous, Lactobacillus plantarum/Lactobacillus pentosus, Lactobacillus brevis to mention the most frequent isolated species (Figure 1 ).

Randomly Amplified Polymorphic DNA

The closely related Lactobacillus pentosus CCUG 33455T, Lactobacillus plantarum ATCC 14917T and Lactobacillus paraplantarum CCUG 35983T were clearly discriminated using the primer P73. The band patterns are characteristic for each type strain. Those patterns can be compared to fingerprints and were used as reference to compare and differentiate the undefined bacterium Lactobacillus plantarum/Lactobacillus pentosus due to the 0.999 percentage of similarity according to the 16S rRNA gene sequencing results in our study. Therefore, according to the band pattern and size of the bands the isolated microorganisms from quinoa sourdough, compared to the type strains, belonged to the genera Lactobacillus plantarum.

Randomly Amplified Polymorphic DNA bands patterns present for Lactobacillus ChB11 , Lactobacillus ChG33, Lactobacillus ChJ239 and Lactobacillus ChR228. Electrophoresed gel showing the amplified bands patterns for the microorganisms identified as Lactobacillus plantarum/Lactobacillus pentosus spp., isolated from quinoa sourdough compared to the bands patterns of the type strains Lactobacillus pentosus CCUG 33455T (L5), Lactobacillus plantarum ATCC 14917 (Lp), and Lactobacillus paraplantarum CCUG 35983T (Lpp). The Molecular Weight Marker VI (Ld) used as size reference expressed in kpb.

PCR specific primers for Lactobacillus pentosus spp. gene amplification

The developed method and the designed primers by Torriani et.al. specific for Lactobacillus pentosus were used and evaluated on the isolated cells from quinoa sourdough. Unfortunately, the pair primers pentF-pREV also attached and amplified regions belonging to Lactobacillus plantarum and Lactobacillus paraplantarum, apart from Lactobacillus pentosus, the strain of interest. The size of the amplicons from the used type strains correlates with the sizes reported by Torriani et.al. 218 bp for Lactobacillus pentosus CCUG 33455T and 318 bp for Lactobacillus paraplantarum CCUG 35983T and with double bands at 218 bp and 118 bp for Lactobacillus plantarum ATCC 14917T. The double band generated during the electrophoresis was also observed for the cells isolated from quinoa sourdough. The results are not consistent regarding the specificity of the primers. However, for the purpose and scope of this study the results can be interpreted as the Lactobacillus spp., isolated from quinoa belongs to the specie Lactobacillus plantarum. Regarding the method and designed primers by Berthier et.al., no bands were observed on the gels except for Lactobacillus pentosus CCUG 33455T. The size of the amplicons were approximately at 220 bp which could correlate to the amplicons size reported by Berthier et.al., at 205 bp for the type strain Lactobacillus pentosus ATCC 8041 T, an homologous strain to Lactobacillus pentosus CCUG 33455T strain used in this study. Neither Lactobacillus paraplantarum CCUG 35983T, Lactobacillus plantarum ATCC 14917T or the Lactobacillus spp. isolated from quinoa sourdough showed amplicons on the gel after the electrophoresis. The results confirmed the specificity of the primers pair straightening the identity of the Lactobacillus spp., at species level from quinoa dough as Lactobacillus plantarum.

Taxonomic identification using API 50CHL

The 4 Lactobacillus plantarum spp. strains identified by analytical methods previously described were phenotypically characterized using API 50CH. The test was repeated twice to confirm the reactions. None of the bacteria were able to fermented erythritol (2), D-arabinose (3), L-xylose (7), D-adonitol (8), methyl-[3D-xylopyranoside (9), L-sorbose (14), dulcitol (16), inositol (17), inulin (33), amidon (starch, 36), glycogen (37), xylitol (38), D-lyxose (41 ), D-fucose (43), L-fucose (44), D-arabitol (45), L-arabitol (46), potassium 2-ketogluconate (48) or potassium 5-ketogluconate (49). The 4 strains have shown to be facultatively heterofermentative lactobacilli based on their ability to ferment pentoses, such as L-arabinose (4) and D-ribose (5), as well as their ability to growth in presence of oxygen. The 2 strains were able to ferment D-xylose (6) and in the case of glycerol (1 ), the status changed from (?) to (t) meaning that the reaction was considered as positive after interpretation of the results regarding glycerol. The change in solution color was not so noticeable; however, if the color of the solution change by 25% to green it should be considered as positive. For potassium gluconate (47), it changed from (?) to (t) for the two strains able to ferment xylose (6). Additionally, as the results were analyzed and compared with the reference (Lp.5*), the identity of the microorganisms was more precisely confirmed as Lactobacillus pentosus for the Lactobacillus ChB1 1 and ChG33 (see table 2). In the case of the other 2 strains Lactobacillus ChJ239 and ChR228 the fermentation of potassium gluconate (47) was interpreted as negative (t). The carbohydrate utilization patterns by the isolated lactobacilli are listed on table 2. The results were compared to the apiweb™ data base, and viable cells were grouped according to their capacity to ferment the carbohydrates. From the 4 bacteria tested, 2 belong to Lactobacillus pentosus and 2 to Lactobacillus plantarum. All the strains were able to fermented L-arabinose (4), D-ribose (5), D-galactose (10), D-glucose (1 1 ), D-fructose (12), D-mannose (13), D- mannitol (18), D- sorbitol (19), methyl-αD-mannopyranoside (20), acetylglucosamine (22), amygdaline (23), arbutin (24), esculin ferric citrate (25), salicin (26), D- cellobiose (27), D-maltose (28), D-lactose (bovine origen) (29), D- sacharose (sucrose) (31 ), D-trehalose (32), D- melezitose (34), gentibiose (39), D- turanose (40).

Table 2. Phenotypical characterization of Lactobacillus ChB1 1 , ChG33, ChJ239 and ChR228 evaluating the metabolic capacity on a variety of carbohydrates.

* Lactobacillus pentosus (Lp.5* ) and Lactobacillus plantarum 1 and 2 (Lp.1 , Lp.2*) are type strains from apiweb™ (API® 50CHL V5.2) database. Means that the microorganisms are more than 50% capable to ferment the carbohydrate in question but less than 100%. The microorganism is less than 50% able to ferment the carbohydrate in question but more than 0%. The microorganism is 50% capable to ferment the carbohydrate in question.

Lactobacillus spp, tannase enzymatic activity

The four Lactobacillus ChB11 , ChG33, ChJ239 and ChR228 strains showed to possess tannase enzymatic activity and decarboxylate gallic acid. A maximum absorbance over 2, in general, was detected on a wavelength range of 380 to 440 nm for the species that could degrade methilgallate. (Table 3).

Table 3. Tannase activity evaluation of Lactobacillus strains isolated from quinoa grains.

+/- Positive or negative result on methilga late degradation and gallic acid decarboxylation The negative control color solution for methilgallate was uncolored and dark green for gallic acid. Results

The experiments report the characterization of the microbiota of commercial white quinoa grains imported from Bolivia to Sweden for Kung Markatta. The presence or not of oxygen besides the temperature during fermentation are factors that could affect the proliferation of some autochthonous bacteria. To find suitable conditions to characterize the microbiota on quinoa grains spontaneous fermentation was controlled aerobically and anaerobically at 30°C for 192 hours.

The value of the pH was used as a reference to control the fermentation process of the quinoa grains. At optimum fermentation, the pH should go below 4. Lactic acid bacteria can resist acid environments compared to other strains. Also, it was observed that during fermentation the percentage of the acidity increased. The number of Enterobacteriaceae viable cells decreased until being undetectable and lactic acid bacteria displaced the number of total aerobic bacteria dominating, assuring an optimum fermentation.

In general 16S rRNA gene sequencing was not enough to distinguish between Lactobacillus plantarum and Lactobacillus pentosus. For that reason, it was necessary to do deeper tests to distinguish between both strains. RAPD electrophoretic band profiles for the 4 lactic acid bacteria isolated from Rogosa agar gave clear distinct patterns for all the strains in comparison to the type strains. According to the type strain Lactobacillus plantarum ATCC 14917 compared with the 4 strains isolated from quinoa grains, all had similar band patterns.

To distinguish between the species Lactobacillus plantarum and Lactobacillus pentosus the selected isolates were evaluated for their capacity to ferment carbohydrates using API 50CHL test. The main point to identify some bacteria as Lactobacillus pentosus was the ability of those to fermented D-Xylose (DXYL). However, Methyl-αD-Mannopyranoside (MDM), D-Melezitose (MLZ), D-Tagatose (TAG), D-Raffinose (RAF) and D-Melibiose (MEL) were also taken i consideration. Glycerol (GLY) was partially fermented by lactobacilli strains isolated from quinoa grains in this research work hence, glycerol was also suitable to discriminate between L. plantarum and L. pentosus. Lactobacillus plantarum ChB11 and Lactobacillus plantarum ChG33

Two different Lactobacillus plantarum strains ChB11 and Lactobacillus plantarum ChG33 were identified as dominating species isolated from white quinoa grains imported from Bolivia. It is the first time that a Lactobacillus plantarum can ferment D-Xylose, a carbohydrate that only the Lactobacillus pentosus species should be able to use as source. This enzymatic action can be used to identify Lactobacillus pentosus from the rest of the LAB group. The strains are facultative heterofermentative (Group II) and they grow under both aerobic and anaerobic conditions. They were isolated from Rogosa agar after 72 hours of incubation at 37°C. However, Lactobacillus plantarum ChB11 and Lactobacillus plantarum ChG33 were also isolated from Malt agar, incubated aerobically at 26°C for 7 days. Malt agar consists of a mixture of maltose, glycerol, and peptone. The capacity to grow on malt agar is uncommon for the species showing that Lactobacillus plantarum ChBB1 and Lactobacillus plantarum ChG33 were able to adapt to the environment, use the available nutrients and grow in presence or not of oxygen.

Lactobacillus plantarum ChB11

The bacteria Lactobacillus plantarum ChB11 was identified to phenotypical level using the API 50CHL test, showing 91 .3% of similarity. The strain is able to fermented L-arabinose, D-ribose, D-galactose, D-glucose, D-fructose, D-mannose, D-mannitol, D-sorbitol, methyl-αD-mannopyranoside, N- acetylglucosamine, amygdaline, arbutin, esculin ferric citrate, salicin, D- cellobiose, D-maltose, D-lactose (bovine origen), D-Melibiose, D- sacharose (sucrose), D-trehalose, D-melezitose, D-Raffinose, gentibiose, D-turanose, D-Tagatose and partially ferment L-Rhamnose and Potassium Gluconate. The growth of bacteria belonging to the family

Enterobacteriaceae, including potential pathogenic bacteria expressing pro- inflammatory lipopolysaccharides on their cell wall, are inhibited by Lactobacillus plantarum ChB11 due to their possibilities of decreasing pH. The bacteria decrease the pH below 4 after 48 hours of fermentation (pH = 3.45 ± 0.013). Also, Lactobacillus plantarum ChB11 express DL-lactate racemase activity which catalyzes the conversion of D-Lactate to L-lactate increasing the percentage of the last one through time. In general, DL- lactate racemase enzymes expression have been found in the Lactobacillus plantarum spp. species.

Lactobacillus plantarum ChB11 increases the concentration of polyphenols such as rutin, vanillic acid, quercetine, kampherol, and luteolin increase at the same time as the antioxidant capacity improve. The identification of the phenolic compounds was done using High Performance Liquid Chromatography-HPLC, on a mixture of acetonitrile:methanol (formic acid 1 %):water as mobile phase, 20 μL of sample was injected and the UV spectra obtained for each molecule was compared against the corresponding standard. The phytase enzymatic activity will be evaluated and probably Lactobacillus plantarum ChB11 increases the availability of minerals by degradation of phytate through the fermentation process. The production vitamins, including vitamin B and folates, essential for humans, are also expected to increase in amount during growth of the Lactobacillus plantarum ChB11. It is also believed that the strain can survive the gastrointestinal passage, which will be demonstrated through re-isolation from human samples after 2 weeks of consumption. The possibility to re-isolate Lactobacillus plantarum ChB11 from humans will then, to our knowledge, be reported for the first time.

Lactobacillus plantarum ChG33

Lactobacillus plantarum ChG33 was identified to phenotypical level using the API 50CHL test, showing 93.1% of similarity. The strain is able to fermented L-arabinose, D-ribose, D-galactose, D-glucose, D-fructose, D-mannose, D- mannitol, D-sorbitol, methyl-αD-mannopyranoside, acetylglucosamine, amygdaline, arbutin, esculin ferric citrate, salicin, D-cellobiose, D- maltose, D-lactose (bovine origen), D-Melibiose, D-sacharose (sucrose), D-trehalose, D-melezitose, D-Raffinose, gentibiose, D- turanose, D-Tagatose and partially ferment L-Rhamnose and Potassium Gluconate. Lactobacillus plantarum ChG33 cannot be solely identified by 16S rRNA gene sequencing but using phenotypical identification the species can be identified and from other strains of the same species. Lactobacillus plantarum ChG33 can be distinguished from Lactobacillus plantarum ChB11 in their ability to ferment methyl-αD-glucopyranoside. Lactobacillus plantarum ChG33 is able to decrease the pH to values below 4 after 48 hours of fermentation (pH = 3.56 ± 0.012). Also for this strain and during the fermentation process the concentration of phenol compounds as well as the antioxidant capacity will probably increase. The resistance to oxidative stress was evaluated using the tannase test, and Lactobacillus plantarum ChG33 was capable to degrade methilgallate and gallic acid to an acceptable amount.

Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228

The strains Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228 were identified as dominating species isolated from white quinoa grains from Bolivia. Lactobacillus spp. are commonly isolated from plant material. However, the phenotypical identification between strains from the same species is overestimated for Lactobacillus plantarum, indicating the necessity of identification also based on PCR-based methods. Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228 belongs to group I (obligate homofermentative) and they grow under both aerobic and anaerobic conditions. They were isolated from Rogosa agar after 72 hours of incubation anaerobically at 37 °C. Both microorganisms were also isolated from Malt agar, incubated aerobically at 26 °C for 7 days. Malt agar consist of a mixture of maltose, glycerol, and peptone. The capacity of growing on Malt agar has not previously been reported for Lactobacillus spp. This adaptability of the strains can be explained by the ecological niche from where they have been isolated. The quinoa plant grows under extreme environmental conditions, where the supply of oxygen and water is limited. The autochthonous flora is therefore highly adaptable to new environmental conditions such as different mediums, variable temperature, and times for growing. It is believed that the strains can survive the gastrointestinal passage, which will be demonstrated through re-isolation from human saliva and fecal samples after 2 weeks of consumption. It is also believed that the strains will be able to affect the population size and expression of activation-, homing, regulation, and memory markers as well as receptors for gram-positive and gram-negative bacterial cells on B-cells, T-cells, macrophages, and dendritic cells in mesenteric lymph nodes and Peyer s patches.

Lactobacillus plantarum ChJ239

Lactobacillus plantarum ChJ239 was identified at phenotypical level using API 50CHL, where 99.9% similarity confirms that the microorganism belongs to the species Lactobacillus plantarum. The strain is able to ferment L-arabinose, D-ribose, D-galactose, D-glucose, D-fructose, D- mannose, D-mannitol, D-sorbitol, methyl-αD-mannopyranoside, N- acetylglucosamine, amygdaline, arbutin, esculin ferric citrate, salicin, D- cellobiose, D-maltose, D-lactose (bovine origen), D-melibiose, D- saccharodise, D-trehalose, D-melezitose, gentiobiose, D-turanose and undefined potassium gluconate (assuming a partial positive reaction of 50%, represented by a light green color on the indicator color change). Lactobacillus plantarum ChJ239 expresses DL-lactate racemase activity which catalyzes the conversion between enantiomers. Lactobacillus plantarum ChJ239 synthesize L-lactate and D-lactate because of the fermentation and after 48 hours the pH decreases to 3.87. At this low pH value, the growth of Enterobacteriaceae is inhibited. The tannase test was used for evaluated the tolerance on oxidative stress. Lactobacillus plantarum ChJ239 was able to resist the stress showing an acceptable percentage of decarboxylation of methilgallate and reduction of gallic acid. During fermentation, the concentration of phenolic compounds such as rutin, vanillic acid, quercetine, kampherol, and caffeic acid increased; however, the amount of saponins seemed to decrease since it was not possible to find compounds belonging to this group. Nevertheless, at the same time the antioxidant capacity increased. The precision of the identification of phenolic compounds by HPLC improved after the enzymatic action of the bacteria. The production of vitamins, mainly vitamin group B, folates and phytase activity will be evaluated and the hypothesis is that Lactobacillus plantarum ChJ239 increases the bioavailability of minerals.

Lactobacillus plantarum ChR228

In general, the characteristics of Lactobacillus plantarum ChR228 are like Lactobacillus plantarum ChJ239; however, Lactobacillus plantarum ChJ239 can ferment D-melibiose and methyl-αD-Mannopyranoside while Lactobacillus plantarum ChR228 is not. Additionally, Lactobacillus plantarum ChR228 can ferment potassium gluconate.

Conclusion

The present experiments demonstrated that the microbiota of quinoa grains is mainly constituted by the autochthonous bacteria Lactobacillus pentosus, Lactococcus lactis, Enterococcus mudtii, Enterococcus hirae and Pediococcus pentosaceous. It is important to highlight that it is the first time that a species of Lactobacillus plantarum is reported as part of the dominating autochthonous bacteria on quinoa grains capable to ferment Xylose. Also because of the metabolic capabilities, the microbiota of quinoa grains increases the nutritional properties and benefits on consuming this grain.

The general procedure was based on taking samples before fermentation, after fermentation and every 7 days until to completed 28 days of storage time. The sampling times were, zero time (before fermentation), at 48 hours (fermentation time) and 7, 14 and 28 days. The samples were withdrawing at the same time in each occasion, with a variance of 30 min.

Preparation of Fermented Quinoa Milk

Quinoa milk

The fermented quinoa milk is prepared using white quinoa grains, imported from Bolivia, and lactic acid bacteria as discloseed herein Lactobacillus plantarum ChB11 (LMG P-31891), Lactobacillus plantarum ChG33 (LMG P- 31892), Lactobacillus plantarum ChR228 (LMG P-31893) and Lactobacillus plantarum ChJ239 (LMG P-31894), isolated from the quinoa grains, categorized as probiotic.

2300 grams of quinoa grains was washed with water in a proportion of 1 :3 (V:V) for 30 minutes. The mixture was agitated few times and the water was replaced after 15 minutes in contact with the quinoa grains. The procedure was repeated twice. The water was discarded and the quinoa grains were washed down under running water. The procedure was repeated 3 times observing the discarded water looks clear and without foam. The washed quinoa grains were ready to be toasted.

The quinoa grains, still humid, were dried over a pant at 225 °C on the stove (level 7, Elektro 8H Helios) for at least 10 to 15 minutes until dryness. Then the temperature was decreased to 195 °C (level 5, Elektro εH Helios) and the quinoa grains were toasted for 20 to 25 minutes. The temperature during the toasting was measured (Multi-Thermometer) on the quinoa grains between 140 to 145°C.

The toasted quinoa grains were cooled down, at room temperature for at least 30 minutes, spreading the toasted quinoa grains over a table. The table was covered with a new clean baking paper to protect the quinoa grains from any contamination coming from the surface of the table.

The quinoa milk was prepared using 100 grams of toasted quinoa grains with 800 ml of water. The water was autoclaved at 121 °C for 15 minutes and stored at 4°C, overnight, before to be used. The toasted quinoa grains were mixed using a blender (Electrolux, blender ESB5400BK, 5 speeds setting) at maximum speed, for 15 minutes. The mixture was filtrated through a juice strainer cloth (Menuett, plastic stand). Total volume after filtration 900 ml of quinoa milk.

Inoculation of Quinoa milk

The quinoa milk, 1000ml, was inoculated by the 4 Lactobacillus plantarum ChB11 (LMG P-31891 ), Lactobacillus plantarum ChG33 (LMG P-31892), Lactobacillus plantarum ChR228 (LMG P-31893) and Lactobacillus plantarum ChJ239 (LMG P-31894), obtaining 4 formulations of fermented quinoa milk. The quinoa milk was inoculated with 0.2% of concentrated bacteria and fermented at 30°C during 48 hours at anaerobic conditions. The probiotic bacteria, used to induce the fermentation, were isolated from white quinoa grain. Those strains were identified at genotypical and phenotypical level. Physicochemical properties The physicochemical properties, such as color and viscosity, was measured for the four strains.

Color

Konika Minolta (United Kingdom) colorimeter was used for measured the color of the quinoa milk at room temperature. The samples were analyzed by triplicated. The parameters luminosity or lightness (L), redness (a) and yellowness (b) were used as a reference for determine the color of the product. The results were analyzed using Spectra Magic software and, the results were classified according to the whiteness index (Wl), and color difference defined as AE.

The mathematic equations used for classified the color of the quinoa milk are:

The total color difference (ΔE), which represents the overall difference in color of the quinoa milk was calculated using equation 2. The results should be classified, according to Cserhalmi et al., as not noticeable (0 to 0.5), slightly noticeable (0.5 to 1 .5) and noticeable (>1 .5).

Viscosity

The viscosity of the quinoa milk was measured using a rheometer with a sensor system of coaxial cylinders were the shear stress was measured as a function of the shear rate between 0.001 to 50 pa, for 900 seconds at 18 °C of temperature. pH of the quinoa milk

The pH of the fermented quinoa milk was measured using a digital pH meter (744pHmeter/Metrohom-Kebolab). The pH-meter was calibrated before to be used according to the manufacturer recommendations using buffer solutions at pH 7 and 4.

Acidity

The lactic acid produced during fermentation was measurement using Enzytech D/L lactic acid kit (R-Biopharm Darmstadt, Germany) following the instructions recommended for the manufacturer. However, the volume of the samples was changed from ml to μL. Briefly, 15 ml of samples (n=6) were centrifuged at 6000 rpm for 5 minutes. Aliquots of the supernatant was extracted and diluted 100 times with Milli-Q water. A volume of 224pl for D- lactic acid and 226 μL for L-lactic acid is required. The solution content 100 μL glycylglycine buffer, 20 μL of NAD solution, 2 μL of GPT suspension, 10 μL of sample or 10pl Milli-Q water (blank) and, 90pl of Milli-Q water was mixed and the solution was stained for 5 minutes. The absorbance (A ₁ ) of the mixture was measured. After, 2 μL of the enzyme D-lactate dehydrogenase solution was added and incubated for 30 minutes and the absorbance (A ₂ ) was measured. Finally, 2μl of the enzyme L-lactate dehydrogenase solution was added and incubated for 30 minutes more and, the last absorbance (A ₃ ) was measured. The incubation of the samples was done at room temperature and the absorbance of the samples was measured at 340nm using a Spectrostar Nano Multiplate Reader (BMG Labtech, Germany) by triplicate. The concentration of D-lactic acid or L-lactic acid was calculated applying the following equation:

First calculate the absorbance difference.

ΔA _{D -lactic acid} = (A ₂ - A ₁ ) ΔA _{L -lactic acid} = (A ₃ - A ₂ ) Use the absorbance difference from previous equation, and used on the question 3:

Where:

V= final volume (ml) v = sample volume (ml)

MW = molecular weight of the substance to be assayed (g/mol) d = light path (cm) ε = extinction coefficient of NADH at 340 nm = 6.3 [lxmmol ^-1 xcm ^-1 ]

ΔA = ΔA _{D -lactic acid} or ΔA _{L -lactic acid} Nutritional properties

Ash

The ash content was determined incinerating the samples at 550°C during 16 to 20 hours in a Haraeus burning chamber. The melting pot was dried and weighted before transferring the samples into them. The samples were weighted before and after incineration. The ash content was calculated following the equations below:

% Ash (dry basis) = ( M _ASH /M _DRY )*100 (4)

% Ash (wet basis) = (M _ASH /M _WET )*100 (5)

Moisture

The moisture content is used to analyze the water content. The water was evaporated from the samples during approximately 2 hours followed by heating in a dryng oven (Termaks drying oven) at 105 °C for 24 hours. The samples were weighted before and after elimination of water. The equation used to determine the moisture content, expressed as percentage (%) is: % moisture = (M _initial — M _dried )/M _initial *100 (6) Protein.

The protein analyzer (Flashea 112 series) was used to determine the protein content on the fermented quinoa milk. 1 ml of sample, by triplicated, was used to be analyzed following the protocol for liquids where the nitrogen content is measured and the content of protein can be calculated using a conversion factor of 6.25. This conversion factor is stablished by defect for the equipment and the protein analyzer is based on the Kjeldhal method, were the total content of nitrogen from the sample is detected by the equipment.

Lipids or Fat

Approximately, 15 ml of sample were collected on a 50ml tube (sterilized Eppendorf tube) and centrifuged at 13600 rpm for 10 minutes at 4°C. The supernatant was discarded and between one or two ml of ethanol was added to the solid portion. Then, the samples were dried at 105 °C overnight. Between 2.5 or 3 grams of dry sample (S) was weighted on a porous thimble. 80 ml of petroleum benzene, was used per sample to extract the fat, and transferred to metal containers (W1 ). The equipment used was a Soxtechs 2055 and has a capacity for 6 samples.

The content of fat is calculated according to equation 7:

% crude fat =

Where

W ₁ = weight of empty flask in (g).

W ₂ = weight of the flask and the extracted fat (g).

S = weight of the sample (g)

Total Carbohydrates

The calculation of the total carbohydrates depends on the result of previous determination. The Total carbohydrates can be calculated according to equation 8:

% Carbohydrates= 100 - % of moisture - %protein - %lipids - % ashes (8) Microbiological Characterization of the Fermented Quinoa Milk

In order to evaluate the safeties and hygiene quality of the fermented quinoa milk, two agars medium were used. Violet Red Bile Dextrose (VRBD) agar is use for evaluate the presence of Enterobactericeae, the plates were incubated at 37 °C for 24 hours and, Tryptic Soy Agar (TSA) was used for determine the total bacterial count, the plates were incubated at 30 °C for 72 hours. Samples of quinoa milk were withdrawing and mixed with peptone water (1 :9, v:v). Serial dilutions were made and from the proper dilution the samples were plated on VRBD, TSA and Rogosa. The survivability of the probiotic bacteria, used for induce the fermentation, was evaluated controlled every seven days, starting after the 48 hours of fermentation. The Rogosa plates were incubated on anaerobic conditions at 37°C for 72 hours.

The viable colonies were recorded from each agar and used to calculate the number of colony forming units (CFU).

Statistical analyses

The statistical analyses was performed using SigmaPlot software. The data were analyzed comparing the values for each group in each sampling occasion, meaning before fermentation, after fermentation and during storage using ANOVE on rank basis. Whitney rank sum test was applied to compare between different groups. The statistical value was presented as mediam, 25- 75 percentages.

Preparation of Fermented Quinoa Milk

The quinoa grains were toasted with constant movement of the grains until a nutty smell can be perceived and the color of the grains change from white to golden. The quinoa milk was inoculated immediately after being filtrated and collected into 1 liter bottles. The concentration of viable bacteria found in the fermented quinoa milk is listed in table 4.

Table 4. Content of probiotic bacteria after 48 hours of fermentation.

In comparison, for Lactobacillus plantarum299v, the number of viable cells expressed as Log CFU/ml is lower compared to the strains Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239andChR228.

Physicochemical properties

Color

The measurement in color is directly correlated with the production of lactic acid on fermented products. A major production of lactic acid may or is the probability that the product became whitish. The amount of D-or L-lactic acid is listed on table 4. For the formulation of fermented plant base beverage is important because if the color of the drink is similar or close to be whitelooking like milk, helping on the acceptability during consumption. Additionally, it is easier to change the color of the drink, if the final product is whitish. The use of the Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228 isolated from the white quinoa grains improved the color of the drink with a noticeable change to white after and during fermentation. According to a classification system it is noticeable if a change of the color is grater than 1 ,5. On the case of our four strains the change of the color is noticible. The value on the ΔE was ranged between 2,01 and 4.95. The whiteness index (Wl) can be considered extremely good if the value is over 50, meaning that the color is closely correlated to white. The whiteness index (Wl) is still stable during the following days, 14 and 28, meaning that the 48 hours of fermentation are essential to ensure the final color of the product.

Table 5. Whiteness index for the 4 lactobacillus strains after 48 hours of fermentation.

Table 6. pH before fermentation (cero time) after fermentation and during storage time

• *The decreasing of the pH reach to this value after 72 hours (3 days) of fermentation.

• **These values will be completed ones the samples are collected.

Acidity. Most Lactic acid bacteria, mainly Lactobacillus species has the potential to produce D-or L-lactic acid mean while the enzymes are active in equal proportions. Additionally, some of them have the capacity to transform D- lactic acid into L-lactic acid or vice versa. It was observed that Lactobacillus plantarumChB11andLactobacillus plantarumChG33 had the capacity to transform L-lactic acid into D-lactic acid in some point between the 14 days and 28 days of fermentation. As can be observed on Table 7, the amount of D-Lactic acid decrease drastically at 28 days, were the amount of D_lactic acid decrease and the amount of L-lactic acid remained stable. This capacity was not observed for Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228. However, those last-mentioned strains synthesize D-and L-lactic acid in equal amount during fermentation.

Table 7. concentration of D-and L-lactic acid production during fermentation The quantity of lactic acid produced during fermentation and the decrease of the pH are the two mainly parameters considered and controlled during fermentation. The production of lactic acid, both L-and D-, and the change of the pH during quinoa milk fermentation with the commercial probiotic bacteria Lactobacillus plantarum299v is shown in table 7. It was observed that the production of lactic acid and the change of the pH are significantly different regardless of the Lactobacillus plantarumChB1 1 , Lactobacillus plantarumChG33, Lactobacillus plantaru mChJ239and Lactobacillus plantarumChR228 isolated from the quinoa grains. The production of each lactic acid, both L-and D, is higher in comparison with the lactic acid produced by Lactobacillus plantarum 299v.The values can be observed in table 7.

Nutritional properties

Table 8. Physicochemical properties before, after fermentation and during storgage time

Microbiological characterization of the fermented quinoa milk.

Quinoa grains contains, as a part of their autochthonous microbiota, beneficial and pathogenic bacteria. The majority of, them are inactive and undetectable. However, a change on the environmental conditions, such as temperature, can active them and encourage their multiplication. Some of the pathogenic bacteria identified during fermentation was Pantoea agglomerans, Klebsiella michiganensis, Pseudomonas spp., Staphylococcus warnery, Bacillus subtilis and Enterococcus mundttiand Enterococcusspp. The presence of those potential pathogenic bacteria can compromise the quality of the product and, most important, the human health if they are consumed. Lactobacillus plantarumChB11 , Lactobacillus plantarumChG33, Lactobacillus plantarumCh239 had shown the potential to growth and inhibit those potential pathogenic bacteria ensuring the quality and safeness of the fermented quinoa milk.

Comparison of fermented quinoa milk with the 4 strains and Lactobacillus plantarum 299V

Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239 showed to have highest inhibition against potential pathogenic bacteria ensuring that the fermented quinoa milk was free of those microorganism after 48 hours of fermentation. Besides, they remained stable in quantity during storage time. In comparison to the fermentation of quinoa milk with Lactobacillus p/antarum299v the microorganism Enterococcus mudtii was still present in a considerable amount compromising the hygiene quality of the product. The Lactobacillus strains ChB11 , ChG33, ChJ239 and ChR 228 totally inhibit the growth of Enterococcus mudtii safeguarding the hygiene quality of the formulated fermented quinoa milk.

Regarding Lactobacillus plantaru mChR228, it has shown to inhibit the growth of the potential pathogenic bacteria present at the beginning in the quinoa milk, but not against Bacillus subtilis, a microorganism that was found in a concentration below the limit of acceptance, but in comparison to Lactobacillus plantarumChB11 , Lactobacillus plantarumChG33, Lactobacillus plantarumChJ239 wherer it was completely inhibited. Discussion

The general procedure for preparing vegetable beverages, mainly known as vegetable milk, is based on cooking by boiling the cereal or grains. The cooked grains are mixed with water and then the mixture is filtered and, as last step, the none soluble particles are discarded. This procedure was applied on the preparation of vegetable milks such as oat, rice or wheat including quinoa. However, in the special case of quinoa, the described procedure above was consistency for the color, obtaining a whitish color similar to cow milk, but regarding the flavor and aroma of the product it was not well accepted. The aroma and flavor are mainly influenced by the saponins molecules, contained on the surface of the grain, bestowing bitterness to the vegetable drink.

In order to improve the characteristics of the formulated quinoa milk, the grains were washed to remove saponins. It is highly recommended to wash the grains, to decrease the saponins content and also to decrease the risk of bacterial contamination that could occur at any point. As a result, the aroma of the vegetable milk improved. The taste of the vegetable milk was improved since the bitter flavor decreased until to be undetectable.

The decreasing of the temperature between the drying procedure and the toasting of the grains was for avoiding the burning of the quinoa grains According to the FDA (Food and Drog Administration) and WHO (World Health Organization) to consider a fermented drink as functional the minimum amount of viable bacteria, expressed as CFU, cannot be below 10 ⁸ . The increase of the bacteria concentration during fermentation time (48 hours) is over the minimum required.

Lactobacillus plantarumChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239 and Lactobacillus plantarum ChR228 also improved the color of the fermented quinoa drink. The formulated fermented quinoa drink is whitish, safe and hygienic to consume. Based on those parameters, it is possible to state that Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239and Lactobacillus plantarum ChR228 improve the nutritional properties of the quinoa milk transforming the the product to functional food drink.

Health benefits consuming Lactobacillus plantarum strains isolated from white quinoa grains

The Lactobacillus strains identified as Lactobacillus plantarum ChB11 , Lactobacillus plantarum ChG33, Lactobacillus plantarum ChJ239and Lactobacillus plantarum ChR228 were isolated from quinoa grains.

20 healthy volunteers were asked to participate in a human study consuming 250 decilitres of quinoa milk fermented with Lactobacillus plantarum ChB11 during 2weeks. The volunteers also were asked to have a wash out period of two weeks, consisting in the not consumption of any product containing live bacteria or probiotic bacteria, previous the study started, to avoid false positives results. Faeces and saliva samples were collected one day before the study started and after 2 weeks consumption of the fermented beverage.The following preliminary results were observed1: ) The samples were analysed using T-RLFP, next q-PCR and generation sequencing. 2) Using q-PCR it was showed that the consumption of Lactobacillus plantarum ChB11 increased the amount of Lactobacillaceae after consumption of the fermented quinoa drink. 3)The Lactobacillus plantarum ChB11 can survive the gastrointestinal track if the increasing of the amount of Lactobacillaceae corresponds to Lactobacillus plantarum ChB11 . 4)The healthy volunteers were asked to answer a questionnaire regarding to symptoms. If during consumption of fermented quinoa drink, they could observe or felt any change on the consistency of the stool. Lactobacillus plantarum ChB11 can survive the passage of the gastrointestinal track. Therefore the strain can be categorized as probiotic.