The 100% pineapple fruit juice (Happy Day juice, Rauch Hungaria, Hungary) was purchased commercially from the local market. The initial pH of the culture medium was adjusted to pH 6.7 using 4N NaOH before fermentation.

The Trypticase–Phytone–Yeast medium (TPY) contained (per liter) 10 g trypticase (BBL), 5 g phytone (BBL), 5 g glucose, 2.5 g yeast extract (Difco), 1 ml Tween 80, 0.5 g L-cysteine HCl, 2 g K₂HPO₄, 0.5 g MgCl2*6H₂O, 0.25 g ZnSO4*7H₂O 0.25 g, 0.15 g CaCl₂, and 0.03 g FeCl₃. TPY agar is the TPY medium supplemented by agar–agar in a concentration of 15 g/l. The pH value of the medium was around pH 6.0.

Beeren's agar medium contained (per liter) 44 g Columbia agar (Oxoid CM331), 5 g glucose, 0.5 g L-cysteine HCl, 5 g agar–agar, and propionic acid 5 mL. The propionic acid was added to the medium after sterilization, and the final pH was adjusted to pH 5.0 with 1N NaOH. MRS agar was prepared based on a recipe given by De Man et al. (17).

The plate counting method (18) was applied to determine the colony-forming units during both the fermentation and storage processes. The Beeren's agar and the TPY agar as well as anaerobic conditions (in Anaerobe Jar + GasPak System or in a Bugbox anaerobic chamber) were used for incubation of bifidobacteria, while the MRS agar (Scharlau, Spain) was applied for lactobacilli. Generally, the colonies of bifidobacteria and lactic acid bacteria were counted after 72–96 h and 48–72 h, respectively.

Fruit juice (50/100 ml flask) was inoculated with Bifidobacterium or Lactobacillus strains and kept under anaerobic conditions at 37°C using Anaerobe Jar + GasPak System (OXOID) or a Bugbox anaerobic chamber. Samples were taken at regular time intervals and the colony-forming unit (cfu) of bifidobacteria and lactobacilli were counted. In addition, the pH was measured. Also, the carbohydrate as well as lactic and acetic acid contents and antioxidant capacity of the samples were determined. A fructo-oligosaccharide, Raftiline, was added as a prebiotic to the pineapple juice at a concentration of 1% at 0, 16, and 24 h of fermentation.

After fermentation (24 h), all juices were placed in the refrigerator (4°C) for storage for 2 months. In the first month, samples were taken weekly, and in the second month twice weekly. Different analyses (viability, antioxidant capacity, survival in the stimulated gastro-intestinal (GI) conditions, fermentation capacity, etc.) were performed.

The carbohydrate and organic acid contents of fruit juice were analyzed by the HPLC method described in a previous study (18) with one minor modification. In this study, the BioRad Aminex87H analytical column (7.7 mm × 300 mm, BioRad, USA) was applied.

The total antioxidant capacity of the fermented pineapple juice was obtained through the ferric reducing antioxidant power (FRAP) assay described by Benzie and Strain (19). The FRAP assay measures the change in absorbance at 593 nm due to the formation of a blue-colored ferrous-tripyridyltriazine complex from a colorless oxidized ferric form by the action of electron-donating antioxidants.

Total phenolic content (TPC) was determined according to methods published by Bao et al. (20) and Jia et al. (21). The total phenolic content was expressed as mg of gallic acid equivalents (GAE) per ml of the extract.

Generally, experimental data were obtained from triplicate runs of all fermentations and duplicate analysis of samples, then were processed statistically using one-way analysis of variance (ANOVA) as well as unpaired and paired Student's t-tests. The Statistica v9.0 software package (StatSoft, USA) and R studio package (r-project.org) were used for data processing and only p < 0.05 was accepted as the statistical significance level. The mean and standard deviation (SD) of data are presented in this study.

Generally, the pH of commercially available pineapple juices ranges from pH 3.5 to pH 4.0; thus before fermentation the pH was neutralized (pH ≈6.7) by 4 N NaOH solution. The juices were inoculated with different probiotic bacteria (Lactobacillus acidophilus, Lactobacillus plantarum, and Bifidobacterium lactis) at an initial cell density of about 10⁶-10⁷ cfu/ml. Cell counts, productivity, pH, and the contents of some sugars after 24 h of fermentation are summarized in Tables 1, 2. The juices were rich in fermentable sugars such as glucose (4.4%, w/v), fructose (2.4%, w/v), and disaccharides (7.4%, w/v). Additionally, small amounts of acetic acid and citric acid were detected. Both genera Lactobacillus and Bifidobacterium were reported to have high requirements of free amino acids, peptides, vitamins, and fermentable carbohydrates for growth (22) due to a lack of proteolytic activity (16). All investigated Lactobacillus and Bifidobacterium strains were able to grow well in the pineapple juice without supplement of any nutrients, meaning this matrix in itself was a suitable medium for propagation of probiotic bacteria. Costa et al. (15) also reported that the L. casei NRRL B442 strain grew very well in native pineapple juice. In the case of lactic acid bacteria (LAB), the cell numbers passed, while in the case of bifidobacteria it reached a level of 10⁹ cfu/ml after 24 of fermentation (Table 1). Interestingly, in our previous study, these probiotic bacteria exhibited better propagation properties (10¹⁰ cfu/ml after 24 h) in apricot juice, which contains about 6 g/100 ml fructose and 3.4 g/100 ml glucose (18). Our results from the bifidobacteria were in agreement with data reported by Havas et al. (23) when they checked the performance of some new isolates of Bifidobacterium in soymilk medium. Pereira et al. (24) produced probiotic cashew apple juice by fermentation with Lactobacillus casei, and they reported a cell density in fermented juice of about 3 × 10⁸ cfu/ml. However, the Lactobacillus casei NRRL B-442 grew very well in cantaloupe juice and reached about 10¹⁰ cfu/ml (25). In a comparison of two lactic acid bacteria strains, the cell count of the Lactobacillus plantarum 299V strain was significantly higher than that of the L. acidophilus La5 strain. This can be explained by their origins. While L. plantarum was isolated from a plant source, L. acidophilus originated from animal source, thus L. plantarum should grow better in a plant matrix than L. acidophilus.

Volumetric productivities of cell yields of L. acidophilus La5, L. plantarum 299V, and B. lactis Bb-12 were 1.38 × 10⁸, 3.49^*10⁸, and 4.02^*10⁷ cfu/ml^*h, respectively. These results were lower than data reported by Bujna et al. (18) when they produced beverage of apricot juice by mono and mixed cultural fermentation with bifidobacteria and lactic acid bacteria (LAB). However, our results agree with data reported by Kun et al. (26) as well as Fonteles et al. (25).

During the fermentation process, some short chain fatty acids (SCFA) were produced in decreasing pH values. In the case of bifidobacteria, the pH dropped from pH 6.7 to pH 5.0 at 16 h and to pH 4.8 at 24 h, while in the case of LAB, the pH dropped to about pH 3.8 at 24 h of fermentation. Lactic acid bacteria produced about 150 mM lactic acid and 30 mM acetic acid in fermented pineapple juice. These values were higher than those reported by Bujna et al. (18) when they did the fermentation of apricot juice by similar LAB strains. Our results are also higher than those published by Gardner et al. (27) as well as by Di Cagno (28). The ratios of lactic acids to acetic acids in L. plantarum 299V and L. acidophilus La5 were 5.37 and 9.91, respectively, and were not affected significantly by supplementation with prebiotic carbohydrate (Table 3). Production of organic acids by Lactobacillus may depend on the composition of the medium applied (29). Zalán et al. (30) confirmed that some strains of Lactobacillus can change their fermentative profile from homofermentative to mix-acid fermentation depending on the composition of media. In our cases, both nitrogen and carbon sources in pineapple juice were sufficient for LAB, and thus very good molar ratios of lactic acid to acetic acid were obtained. The presence of lactic acid generally results in good taste for fermented beverages.

In the case of Bifidobacterium lactis Bb-12, a higher concentration of acetic acid was detected. The concentrations of lactic acid and acetic acid were 6 and 23 mM in natural juices, 15 and 21 mM in the case of supplementation with prebiotic at 16 h, and 7 and 23 mM in the case of supplementation with prebiotic at the initiation of fermentation, respectively. The highest amount of lactic acid (15 mM) was produced when the prebiotic was added at 16 h fermentation. It is well-known that Bifidobacteria ferment glucose via the so-called “bifidus” pathway, in that the phosphoketolase plays a key enzyme role. Theoretically, from 1 molecule of glucose, 1 molecule of lactic acid and 1.5 molecule acetic acid will be liberated (31). Our results (with the exception of one case) did not fit this theory, but we think it depends on the actual fermentable sugar spectrum and on sugar concentration as well as on the amount of nitrogen in the fermentation medium. Unfortunately, this mechanism is still unclear, and more studies should be carried out. It is well-known that a high concentration of acetic acid (metabolic activity of bifidobacteria) may cause the odor and vinegar-like taste of products that is unacceptable.

In the case of the B. lactis Bb-12 strain, the concentration of glucose, fructose, and disaccharides decreased after 24 h of fermentation, inferring intensive hydrolysis of disaccharides and release of monosaccharides. The concentrations of glucose and fructose at 24 h when prebiotic was added at the start of the fermentation were 2.14 g/100 ml and 0.33 g/100 ml, respectively. The concentration of fructose was zero in the case of the prebiotic added at the end of the fermentation. Some papers reported that many bifidobacteria strains are able to utilize disaccharides especially sucrose natively (18, 26). In the case of L. acidophilus La5, the concentration of fructose decreased drastically after 24 h of fermentation, while the concentration of glucose and disaccharides slightly decreased. Generally, L. acidophilus prefers glucose to fructose, and it is a homofermentative organism that utilizes glucose through the Embden Meyerhof Parnas (EMP) pathway (32). Our results showed that in a complex medium like pineapple juice, with the presence of several fermentable sugars and other compounds, the order of utilization of sugars may be changed. In this case, the preference order should be fructose > glucose ≥ sucrose. Bujna et al. (18) reported an order of disaccharide > glucose > fructose with apricot juice fermented by L. acidophilus La5 and L. casei 01 strains. In the case of Bifidobacterium lactis Bb12, the utilization rate of fructose was higher than those of glucose and disaccharides. However, this result is in disagreement with data reported by Kun et al. (26) and, in general, bifidobacteria prefers glucose to fructose and other sugars.

The concentrations of TPC and FRAP in pineapple were determined to be 0.4 mg/ml gallic acid and 2.66 mM FeSO₄/ml equivalents, respectively. Our results are in agreement with data reported by Lu et al. (33) when they studied the physico-chemical properties, antioxidant activity, and mineral contents of 26 pineapple genotypes grown in China. The fermentation of pineapple juice with L. acidophilus La 5, L. plantarum 299V, and Bifidobacterium lactis Bb12 strains caused slight changes (increase) in the concentrations of TPC and FRAP. These values (at 24 h fermentation) were about 0.45 mg/ml gallic acid and 3.0 mM FeSO₄/ml equivalents, respectively. Bujna et al. (18) recorded a slight increase in antioxidant capacity during fermentation of apricot juice with B. lactis Bb12 and L. acidophilus La5. Martin and Matar (34) also registered an increase in antioxidant activity of blueberry juice during fermentation with a novel bacterium from the fruit microflora Serratia raccinii. Probiotics can produce various metabolites with antioxidant activity such as glutathione, butyrate, folate, etc. (35, 36), but the generally accepted opinion is that the antioxidative properties of probiotic bacteria are specific features of individual strains (18).

After fermentation, the samples were centrifuged, and the collected cell mass was submitted to investigate the effects of stress factors on the survival of probiotic bacteria. All investigated probiotic bacteria exhibited very good viability against both stresses: 0.3% pepsin and 0.6% bile salts. Only a 0.01 log cfu/ml loss was detected after treatment with 0.3% pepsin for 135 min, while these probiotic bacteria kept all viable cells after further treatment with 0.6% bile salts for 2.5 h. Some authors also reported the resistance of fresh probiotic bacteria against gastric stresses (37, 38).

The storage of the pineapple beverage was carried out with and without prebiotics. In all cases, the microbial population did not change significantly in the first month. The cell counts of lactobacilli were approximately in the range of 10⁹ and 10¹⁰ cfu/ml, while the cell numbers of bifidobacteria were in the range of 10⁸ and 10⁹ cfu/ml (Figure 1). During this period, minimal changes in the pH values of beverages were detected (data not shown). This is explained by the high buffering capacity of fermented pineapple juice. Additionally, our results also show that supplementation of prebiotic fructo-oligosaccharides did not have any effects on the viability of probiotic bacteria. Oligofructose was also reported to have no influence on the physicochemical characteristics (pH, titratable acidity, color, and turbidity), acceptability, purchase intent, or storage stability of the clarified apple juice products (39, 40), thus our results completely fit with this opinion. Champagne and Gardner (37) evaluated the viability of nine strains of the Lactobacillus genus in a commercial fruit drink stored at 4°C for up to 80 days. The pH of the drink was pH 4.2, which enabled good stability of many cultures during storage. They found that the viability of the L. plantarum was excellent, meaning almost viable cells were retained after 80 days storage, whereas the viability of L. acidophilus was the worst. More than 80% of viable cells were lost after 80 d storage in this medium. L. plantarum is originated from plant source, thus this species held a better level of viability in the fruit juice. Daneshi et al. (14) found that L. acidophilus La5, L. plantarum and B. lactis Bb-12 strains lost a minimal number of cell counts (<10%) during storage at 4°C for 20 days in non-fermented milk and a carrot juice mix drink. Other studies (41–43), however, reported the decrease (about 3–4 log cfu/ml) in viability of probiotic strains incorporated into different juices such as orange juice, grape juice, passion fruit juice, pineapple juice, cranberry juice, and natural cornelian cherry juice during storage. There is no doubt that the viability of LAB varied from strain to strain, but it also strongly depends on the nature and quality of the carrier-matrix. Organic acids and flavor compounds—especially in non-fermented matrices—have negative effects on the survival of probiotic LAB (14). In our case, the fermentation definitely contributed to the viability of probiotic bacteria.

In the first month of storage at 4°C in refrigerated conditions, the total phenolic content and antioxidant capacity decreased to 0.4 mg/ml gallic acid and 2.7 mM/ml FeSO₄ equivalents, and they remained stable for next month. Patthamakanokporn et al. (44) also reported the decrease in antioxidant activity and total phenolic compounds during storage of selected fruits. The decrease in antioxidant activity as well as phenolic and anthocyanin contents may be due to the slight activity of probiotic bacteria in refrigerated temperatures as well as the presence of dissolved oxygen in samples, which resulted in oxidation of phenolic compounds (41). It has been reported that in the absence of light and oxygen, the number of phenolic compounds did not change considerably during the refrigerated storage period (44).

The survival of probiotic cells in fermented pineapple juice against stress factors was modeled and investigated. After storage of fermented pineapple juice for 1 month, the probiotic bacteria lost about 0.11 log cfu/ml viability after treatment with 0.3% pepsin for 135 min, and a further 0.1 log cfu/ml after treatment with 0.6% bile salts. These values were higher by a factor of 10 than data from the fresh fermented pineapple juice. Champagne and Gardner (37) found that the viability of fresh probiotic culture was not affected by the presence of 0.3% bile salts or of pancreatic enzymes. However, the cultures that were stored for 35 days at 4°C in the fruit drink had on average 1.2 log more viability losses than the fresh cultures when exposed to 2 h incubation at pH 2.0 to simulate gastric stress. Stored probiotic cultures in fruit juices seem to be less resistant to acidic conditions than fresh cells; thus, this should be a new research topic in the development of probiotic pineapple juice drinks.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.