Nutritional, Microbial, and Allergenic Changes during the Fermentation of Cashew ‘Cheese’ Product Using a Quinoa-Based Rejuvelac Starter Culture

Fermentation has been applied to a multitude of food types for preservation and product enhancing characteristics. Interest in the microbiome and healthy foods makes it important to understand the microbial processes involved in fermentation. This is particularly the case for products such as fermented cashew (Anacardium occidentale). We hereby describe the characterisation of cashew samples throughout an entire fermentation production process, starting at the quinoa starter inoculum (rejuvelac). The viable bacterial count was 108 -109 colony forming units/g. The nutritional composition changed marginally with regards to fats, carbohydrates, vitamins, and minerals. The rejuvelac starter culture was predominated by Pediococcus and Weissella genera. The ‘brie’ and ‘blue’ cashew products became dominated by Lactococcus, Pediococcus, and Weissella genera as the fermentation progressed. Cashew allergenicity was found to significantly decrease with fermentation of all the end-product types. For consumers concerned about allergic reactions to cashew nuts, these results suggested that a safer option is for products to be made by fermentation.


Introduction
Food fermentation began as a method of preservation, but subsequently developed into a mechanism of altering taste, texture, and nutritional content. The process relies on the metabolic activity of microorganisms to yield by-products that enhance food preservation [1,2]. Fermentation can be undertaken by microorganisms that are either inoculated, already present, or mixed between the two types. [3]. The benefits of convenience and product consistency associated with inoculated fermentation is often matched with a decrease of flavour richness, leading to more bland products [4]. An industrial reality is that the process still mostly depends on spontaneous fermentations to create the enhanced flavour profile that is desired. However, this technique can result in inconsistent products due to a lack of control of the viability of the microorganisms present in the raw ingredients.
Fermentation makes food last longer; one method is through acid production that decreases pH and creates an environment unsuitable for harmful bacteria to survive in. The opportunity exists to  and 'blue' cashew cheese-like product.

Product Sampling
Two product batches were followed for sample collection, termed 'brie' and 'blue'. Three samples of a thousand grams were taken from pre-fermentation, post-fermentation, and final product at the production facility (Nuts for Cheese, London, ON, Canada) and were collected from each batch. Eight-hundred grams of each sample were sent for nutritional analysis, five grams were used for pH testing, five grams were used for microbial culturing, five grams were used for allergenicity testing, and one gram was used for microbiota composition analysis. For allergenicity testing, five different varieties ('cheddar', 'blue', 'brie', herb, and red) were followed and fifty gram samples were collected at 0 h, 24 h, 48 h, and 72 h post inoculation. For the red batch, only the 72 h timepoint was collected, because it has the same cashew processing as the herb batch up until 72 h. Five grams were used for allergenicity tests. All samples were stored at 4 °C until needed.

Product Sampling
Two product batches were followed for sample collection, termed 'brie' and 'blue'. Three samples of a thousand grams were taken from pre-fermentation, post-fermentation, and final product at the production facility (Nuts for Cheese, London, ON, Canada) and were collected from each batch. Eight-hundred grams of each sample were sent for nutritional analysis, five grams were used for pH testing, five grams were used for microbial culturing, five grams were used for allergenicity testing, and one gram was used for microbiota composition analysis. For allergenicity testing, five different varieties ('cheddar', 'blue', 'brie', herb, and red) were followed and fifty gram samples were collected at 0 h, 24 h, 48 h, and 72 h post inoculation. For the red batch, only the 72 h timepoint was collected, because it has the same cashew processing as the herb batch up until 72 h. Five grams were used for allergenicity tests. All samples were stored at 4 • C until needed.

Nutritional Analysis
Samples of the 'brie' product were sent to an accredited food testing laboratory (SGS Canada Inc., Burnaby, BC, Canada) for nutritional analysis by standard methods. The standards and respective protocols used include AFVAN-SLMF-0010 and AOAC 996.06; AFVAN-SLMF-0014 and AOAC 990.03; AFVAN-SLMF-0013 and AOAC 974.29 and AOAC 976.26; AFVAN-SLMF-0018 and AOAC 982.14; AFVAN-SLMF-0022 and AOAC 2011.14; AFVAN-SLMF-0009 and journal of AOAC vol 86, No. 2, 2003 pg 367-373; AFVAN-SLMF-0012 and AOAC 926.08; AFVAN-SLMF-0022 (ICP-MS) and AOAC 926.08 (AOAC International, Rockville, MD, USA). Analyses were conducted for nutritional label content (moisture, fat, protein, carbohydrate, cholesterol, sugar, dietary fibre, iron, calcium, sodium, potassium, ash content, and energy), as well as vitamins B1, B2, B6, B12, and minerals (copper, magnesium, zinc, selenium, manganese, nickel, and phosphorous). All analyses were performed for pre-fermentation, post-fermentation, and final product samples except for the mineral scan, which was performed on the final product sample, only because metals are not degraded or produced by bacteria and yeast [19]. In addition, the pre-fermentation, post-fermentation, and final product sample of 'brie' and 'blue' batches were tested for pH. Five grams of each sample were mixed with 50 mL of water and then the pH of the mixture was read with a pH probe at 25 • C (Beckman Coulter, UT). The pH of each mixture was recorded four times and the average was calculated. The pH of the original samples accounting for dilution was then calculated.

Culturing of Microorganisms
One gram of each sample was manually homogenized with 10 mL of phosphate-buffered saline (PBS) in a sampling bag (Whirl-Pak, Nasco, ON). The homogenized solutions were then serially diluted 10-fold to 10 −6 in PBS. All dilutions were then each spread plated onto de Man, Rogosa, and Sharpe (MRS; BD Difco, MD), Rogosa (BD Difco, MD), Sabouraud dextrose agar (SDA; BD Difco, MD), and glucose yeast calcium carbonate agar plates (GYC) for the detection of acetic acid bacterial producers, yeast, and fungi [20]. Plates were aerobically incubated at 37 • C for 3 days.

PCR and 16S rRNA Gene Analysis of Bacterial Isolates and 18S rRNA Gene Analysis of Yeast Isolates
Colonies that appeared morphologically different on agar plates were isolated and their DNA was extracted using the InstaGene matrix (Bio-Rad, CA) for PCR. Suspected bacterial DNA was subjected to PCR using the bacterial ribosomal 16S primers pA (AGAGTTTGATCCTGGCTCAG) and pH (GCGACAAACCACCTACGAG). Suspected yeast DNA was subjected to PCR using the yeast ribosomal 18S primers F-566 (CAGCAGCCGCGGTAATTCC) and R-1200 (CCCGTGTTGAGTCAAATTAAGC). Both bacterial and yeast PCR were carried out using Q5 High-Fidelity DNA Polymerase (New England BioLabs, MA). The bacterial PCR conditions used were an initial denaturation at 95 • C for 4 minutes, followed by 35 cycles of 95 • C for 30 seconds, 55 • C for 30 seconds, and 68 • C for 1 minute. The yeast PCR conditions were an initial denaturation at 94 • C for 4 minutes followed by 40 cycles of 94 • C for 1 minute, 54 • C for 1 minute, and 72 • C for 1 minute. The PCR products were then run on a 1% agarose (w/v) electrophoresis gel for 80 minutes at 80V. The gel was stained with an ethidium bromide wash (0.5 µg/mL) and visualized on a MultiImage Light Cabinet (Alpha Innotech Corporation, CA). The PCR amplicons were then purified with the Invitrogen Purelink PCR Purification Kit (ThermoFisher Scientific, CA). The purified PCR amplicons were analysed with a DeNovix spectrophotometer (model DS-11 (M), DeNovix, DE) for purity and then sent to the London Regional Genomics Centre (Robarts Research Institute, London, ON, Canada) for Sanger sequencing. DNA sequences were compared to sequence databases using the Basic Local Alignment Search Tool (BLAST) for species identification.

Microbiota Extraction, Sequencing, and Analysis
DNA was extracted using the DNeasy Powersoil Kit (Qiagen, MD). The Beckman BioMek 3000 Laboratory Automation Workstation was set to aliquot 10 µL (2.3 pmol/µL) of 32 primers (16 left and 16 right) with unique barcodes into a 96 well plate. The V4 region of the 16S ribosomal RNA gene was amplified using the primers (5'-3') ACACTCTTTCCCTACACGACGCTCTTCC GATCTNNNNxxxxxxxxGTGCCAGCMGCCGCGGTAA and (5 -3 ) CGGTCTCGGCATTCCTGCTG AACCGCTCTTCCGATCTNNNNxxxxxxxxGGACTACHVGGGTWTCTAAT (where xxxxxxxx is a sample-specific nucleotide barcode, the preceding sequence is a portion of the Illumina adapter sequence for library construction). Then, 2 µL of template DNA and 20 µL of Promega GoTaq Colourless Master Mix (Promega, WI) were added to the well plates. The PCR conditions used were 2 minutes at 95 • C, then 25 cycles of 95 • C for 1 minute, 50 • C for 1 minute, and 72 • C for 1 minute. PCR amplicons were sent to the London Regional Genomics Centre (Robarts Research Institue, London, ON, Canada), where samples were DNA quantified (Quant-it, Life Technologies, ON) and cleaned with QIAquick (Qiagen, MD). DNA was sequenced with the MiSeq Illumina®platform processed with dada2 (version 1.8) and custom R (version 3.4.1) scripts (Dr. Greg Gloor, github.com/ggloor/miseq_bin).

Cashew Allergenicity Testing of Products at Different Stages of Production
For allergen content testing, the BioFront Technologies Cashew ELISA Kit (BioFront Technologies, FL), designed for the sensitive detection of cashew allergens in food products, was adapted for the assay. Preliminary studies using this ELISA assay indicated that samples required the significant dilution of samples in the order of 10 −5 . Other steps of the ELISA were followed as per the manufacturers' instructions and read using a spectrophotometer microplate reader (BioTek Eon Microplate, VT). A standard curve was created, and sample allergen quantities were calculated against these values.

Nutritional Analysis
The starting cashew mixtures were both pH 3.0, marginally decreasing to 2.8 in the 'brie' final product and 2.9 in the 'blue' final product (Table 1). In order to document the possible nutritional changes during nut fermentation, different 'brie' timepoint samples (pre-fermentation, post-fermentation, final product) were nutritionally analysed for comparison. The fats, vitamin B12, and moisture decreased when comparing the final product to pre-fermentation samples (Table 1). Conversely, the carbohydrates, energy, sodium, calcium, iron, potassium, vitamin B1, and vitamin B6 increased from pre-fermentation samples. The trans-fat, omega 3 fat, cholesterol, sugars, fibre, vitamin C, and vitamin B12 were not detected at high enough levels to yield a discrete value throughout all three samples. Copper, magnesium, zinc, selenium, manganese, nickel, and phosphorous were only analysed from the final product for quantification and they were present at low levels.

Viable Microbial Counts
'Brie' samples showed approximately a one log increase of CFU when grown on Rogosa (lactobacilli), SDA (yeast), and GYC (lactic acid bacteria) ( Figure 3A). 'Brie' samples grown on Nutrients 2020, 12, 648 6 of 12 MRS (semi-selective for lactobacilli-and related bacteria) showed approximately a two log increase ( Figure 3A). The highest count of 10 9 CFU/mL was on MRS media at the product step. The other three media showed increasing counts from 10 7 to 10 8 CFU/mL following fermentation. Conversely, the samples from 'blue' product showed no log decreases in counts on SDA and MRS, and remained at 10 8 CFU/mL. The 'blue' samples did show a log increase on Rogosa and a two log decrease on GYC ( Figure 3B). pH of 'blue' 3.0 2.9 2.9 -% unit represents % of sample analyzed; pH values represent the mean of 4 pH readings, each taken at 25 °C.

Viable Microbial Counts
'Brie' samples showed approximately a one log increase of CFU when grown on Rogosa (lactobacilli), SDA (yeast), and GYC (lactic acid bacteria) ( Figure 3A). 'Brie' samples grown on MRS (semi-selective for lactobacilli-and related bacteria) showed approximately a two log increase ( Figure  3A). The highest count of 10 9 CFU/mL was on MRS media at the product step. The other three media showed increasing counts from 10 7 to 10 8 CFU/mL following fermentation. Conversely, the samples from 'blue' product showed no log decreases in counts on SDA and MRS, and remained at 10 8 CFU/mL. The 'blue' samples did show a log increase on Rogosa and a two log decrease on GYC ( Figure 3B).

Identification of Bacterial Isolates
Bacteria were isolated and identified as Weissella cibaria, Leuconostoc citreum, and Lactococcus lactis ( Table 2). The W. cibaria isolates were identified in both rejuvelac and samples from the 'blue' product. W. cibaria was found to grow on both MRS and GYC, while L. citreum grew on both SDA and GYC.

Microbiota Characterisation of Fermentation Stages and Products
The microbiota composition of samples taken from the rejuvelac, miso paste, pre-fermentation, post-fermentation, and final product were examined. Over 17 genera were detected. Comparison of the relative abundance of bacterial composition obtained by 16S rRNA gene sequencing methodologies illustrated that the fermentation of the product alters the microbiota composition, to a few genera dominating the final product ( Figure 4). The starting fermented cultures of rejuvelac and miso also influenced the microbial composition of the final product. In the samples from 'brie', Pediococcus became the most prominent genus in the final product, whereas in the other 'blue' samples, the composition was dominated mostly by Weissella, Lactococcus, and Pediococcus.

Cashew Allergenicity.
In all five batches that were tested, there was a decrease in cashew allergen content when comparing pre-fermentation (0 h) to the finished product stage (72 h) ( Figure 5). Most of the reduced allergy activity occurred within the first 48 hours of fermentation. The 'brie' samples started with an allergen concentration of 266,800 ppm and decreased by 35% at 72 h. As for 'blue' samples, the allergen concentration started at 165,200 ppm and decreased by 54% after 72 h. Following the same 0 h to 72 h comparison, the 'cheddar' started at 159,850 ppm and decreased by 66%, herb started at 248,750 ppm and decreased by 49%, and red started at 248,750 ppm and decreased by 29%.

Cashew Allergenicity
In all five batches that were tested, there was a decrease in cashew allergen content when comparing pre-fermentation (0 h) to the finished product stage (72 h) ( Figure 5). Most of the reduced allergy activity occurred within the first 48 hours of fermentation. The 'brie' samples started with an allergen concentration of 266,800 ppm and decreased by 35% at 72 h. As for 'blue' samples, the allergen concentration started at 165,200 ppm and decreased by 54% after 72 h. Following the same 0 h to 72 h comparison, the 'cheddar' started at 159,850 ppm and decreased by 66%, herb started at 248,750 ppm and decreased by 49%, and red started at 248,750 ppm and decreased by 29%.

Discussion
In this study, a variety of bacterial genera were identified in the production process of fermented cashews. Weissella were dominant in rejuvelac and 'blue' samples, Lactococcus in 'blue' samples, and Leuconostoc in 'brie' samples. As the pH of the fermented cashew mixes ranged from 2.8-3.0, the bacteria were clearly able to tolerate acidic environments and grow on SDA media. Conversely, yeast were not cultured on SDA media perhaps due to low level presence in the rejuvelac to begin with. Certain bacteriocins produced by lactic acid bacteria also have antifungal properties [21].
In the cashew 'brie' product, Pediococcus, Weissella and Lactobacillus accounted for nearly 60% of the relative bacterial composition of the final product. These organisms originated from the rejuvelac and they have also previously been isolated from quinoa [22]. Comparing the pre-fermentation, postfermentation, and final product samples of cashew 'brie', these three genera became predominant. Pediococcus is commonly found in fermented food, as it plays a role in souring fermented food through the metabolism of carbohydrates into lactic acid [23]. Weissella strains are able to metabolize carbohydrates for energy and produce lactic acid and acetic acid through this sugar metabolism [24]. Lactobacillus is another genus that is common in fermentation, as it is also capable of producing lactic acid [25]. These three genera each contribute to the maintenance of acidity of the cashew product.
As for the 'blue' product, Weisella, Pediococcus, and Lactococcus cumulatively contributed over 80% of the relative bacterial composition in the final product. Lactococcus was present because it is also capable of lactic acid fermentation, thereby contributing to the acidity of the final product, similar to Weissella and Pediococcus [1].
The rejuvelac used for both 'brie' and 'blue' products had Enterobacteriaceae present at low levels, which may be attributed to contaminated quinoa because Enterobacteriaceae have been found to be

Discussion
In this study, a variety of bacterial genera were identified in the production process of fermented cashews. Weissella were dominant in rejuvelac and 'blue' samples, Lactococcus in 'blue' samples, and Leuconostoc in 'brie' samples. As the pH of the fermented cashew mixes ranged from 2.8-3.0, the bacteria were clearly able to tolerate acidic environments and grow on SDA media. Conversely, yeast were not cultured on SDA media perhaps due to low level presence in the rejuvelac to begin with. Certain bacteriocins produced by lactic acid bacteria also have antifungal properties [21].
In the cashew 'brie' product, Pediococcus, Weissella and Lactobacillus accounted for nearly 60% of the relative bacterial composition of the final product. These organisms originated from the rejuvelac and they have also previously been isolated from quinoa [22]. Comparing the pre-fermentation, post-fermentation, and final product samples of cashew 'brie', these three genera became predominant. Pediococcus is commonly found in fermented food, as it plays a role in souring fermented food through the metabolism of carbohydrates into lactic acid [23]. Weissella strains are able to metabolize carbohydrates for energy and produce lactic acid and acetic acid through this sugar metabolism [24]. Lactobacillus is another genus that is common in fermentation, as it is also capable of producing lactic acid [25]. These three genera each contribute to the maintenance of acidity of the cashew product.
As for the 'blue' product, Weisella, Pediococcus, and Lactococcus cumulatively contributed over 80% of the relative bacterial composition in the final product. Lactococcus was present because it is also capable of lactic acid fermentation, thereby contributing to the acidity of the final product, similar to Weissella and Pediococcus [1]. The rejuvelac used for both 'brie' and 'blue' products had Enterobacteriaceae present at low levels, which may be attributed to contaminated quinoa because Enterobacteriaceae have been found to be present in cereal grains and are also common contaminants of food in general [26][27][28]. Although Enterobacteriaceae were initially present in the rejuvelac, by the end of cashew fermentation it was no longer detectable. Reduction of Enterobacteriaceae may be attributed to the activity of bacteriocins produced by lactic acid bacteria, which are the main bacteria present in both 'brie' and 'blue' products. Bacteriocins produced by lactic acid bacteria have demonstrated bactericidal and bacteriostatic properties against common food-contaminating Enterobacteriaceae genera, such as Salmonella and Listeria [29,30]. An alternate explanation for the reduction of Enterobacteriaceae would be that the cashew mix is more acidic than the starting rejuvelac culture, and some Enterobacteriaceae growth is inhibited by acidity [31].
Previous studies have demonstrated that many fermented foods are a reliable source for lactic acid bacteria, with counts of 10 5 -10 7 CFU/mL or g [32]. These include the genera found in this study. The lactic acid preservative effects lower the pH levels past the point of optimal growth for other microorganisms [33]. In both the 'brie' and 'blue' batches, the pH essentially did not change, further enhancing food safety during cashew fermentation [34].
The bacterial counts stayed within 10 7 -10 9 CFU/mL or 10 8 -10 10 CFU/g depending upon the stage of production. The increase in counts after fermentation can be explained by the proliferation of bacteria in the nutrient rich cashew mix. These counts are higher than the levels generally found in fermented foods which usually contain 10 5 -10 7 CFU/mL or g of lactic acid bacteria [32]. The exception is fermented dairy products with cell counts up to 10 9 CFU/mL [32]. The high presence of bacteria might imply that the bacteria are metabolizing molecules into by-products that increase the product's nutrient content. However, this was not noted in this study. Where the basic need for food is no longer caloric intake, the prevention of nutritional diseases and maintenance of health has become a critical part of North American dietary culture [35]. As a result of this, being able to design food beyond just taste, but also to include necessary nutrients and biologically active molecules could be a leverage point for food manufacturers.
Nutritional analysis revealed an increase then decrease of nutrient levels, which may be attributed to the production and then usage by the microorganisms present. For example, a decrease in fats may be attributed to lipolysis by bacteria. Lipolysis is a natural occurrence in dairy fermentation and it involves the breakdown of triglycerides into fatty acids, which can aid flavour development [36]. Since total fat is the largest component of cashew nuts by weight, contributing approximately 48.3%, there are high levels available for lipolysis in the cashew product [37].
Metal analyses were only performed on the final product sample because they can neither be created or destroyed by microorganisms [19]. The levels for copper, magnesium, zinc, selenium, manganese, nickel, and phosphorous were quite low with regards to nutritional requirements. The highest was 2059 ppm of phosphorous or approximately 0.006177 mg per serving. The recommended dietary allowance for adults in Canada for phosphorous is 700 mg per day [38].
A potentially important finding was that cashew fermentation yielded a decrease in cashew allergen in all varieties of product that were tested. Most of the decrease appeared in the first 48 hours. From 48 h to 72 h, a slight increase was observed; however, this may be due to the evaporation of moisture from the product as they are dried from 44.0% to 43.4%, which proportionally increases the cashew content. The decrease in cashew allergen is likely from proteolytic cleavage by a bioactive molecule produced during fermentation, as this is the mechanism shown to decrease peanut allergenicity [39,40]. Finding methods to overcome nut allergies is a field of increasing interest as the prevalence of nut allergies is a major contributor to anaphylaxis in children [18]. Immunotherapy is a major avenue that is pursued when attempting to mitigate nut allergies, but it has its drawbacks with limited results [16]. An alternative to immunotherapy is decreasing nut allergens in food so that an allergic reaction is never reached. Methods to remove cashew allergens such as with p-aminobenzamidine and magnetic 6% cross-linked agarose beads have yielded success in decreasing some cashew allergens, but not all [14]. With peanuts, biological methods to decrease allergenicity have been investigated and enzymatic cleavage of peanut allergens, such as seed storage proteins, Ara h1, Ara h 2, and Ara h 3, have been shown to decrease along with its IgE binding reactivity (17)(18)(19). Cashews have similar proteins to peanuts as their major allergens and therefore may be subject to alike enzymatic cleavage to decrease allergenicity. The findings suggest that bacteria degraded one or more of the three major cashew allergen proteins, Ana o 1, Ana o 2, and Ana o 3 [41][42][43][44], which are associated with higher rates of anaphylaxis compared to peanuts [45]. Currently, there is no enzymatic treatment of cashew allergens. The microbiota sequencing identified lactic acid bacteria as predominant throughout the cashew product samples, and it appears that their proteolytic activity was involved in the reduction in allergy activity [46]. Although the cashew allergen load decreased during fermentation, it did not decrease past the threshold of eliciting cashew allergies of approximately 60 ppm per cashew 'cheese' serving size [47]. Despite this, decrease in cashew allergen via fermentation is a method that could be useful in decreasing allergens.
In summary, the study cashew products retained their nutritional composition after fermentation, and the starter rejuvelac culture was important in generating the end-product. The reduction in cashew allergens with fermentation provided an additional benefit for consumers.