Comparative analysis of two antifungal Lactobacillus plantarum isolates and their application as bioprotectants in refrigerated foods

Authors


Correspondence

Douwe van Sinderen, Department of Microbiology, University College Cork, Cork, Ireland. E-mail: d.vansinderen@ucc.ie

Abstract

Aims

To compare the technological robustness of two antifungal Lactobacillus plantarum isolates and to assess their ability to inhibit growth of the spoilage yeast Rhodotorula mucilaginosa in two different refrigerated foods.

Methods and Results

The effects of freeze-drying, thermal treatments and varying salt concentrations on the viability of two antifungal lactic acid bacteria (LAB) were examined. Antifungal compound(s) contained in the supernatant of both isolates were compared to commercially available food preservatives. Both isolates were used as dairy starter adjuncts in yoghurt and inoculants in orange juice to determine the antiyeast activity towards R. mucilaginosa. Yeast growth was retarded by the tested isolates in both food settings with one of the isolates, Lact. plantarum 16, being the most potent inhibitor.

Conclusions

Both lactobacilli exhibited considerable robustness to withstand processing treatments commonly encountered in a food industrial setting. The isolates were shown to possess potent antifungal activity in both in vivo and in vitro food models.

Significance and Impact of the Study

The studied antifungal lactobacilli may represent safer and consumer-friendly alternatives to the use of chemical preservatives. This is the first report of antifungal Lact. plantarum exerting protective potential in yoghurt and orange juice.

Introduction

Fungal contamination of foods prevails as a principal problem for food manufacturers. Food preservation methods such as the use of chemicals or irradiation may impart undesirable properties to foods as well as raising concerns among consumers (Brewer and Rojas 2008). Furthermore, the production of mycotoxins and the development of chemical-resistant fungi are problematic for the food industry (Brul and Coote 1999). Consequently, the need for alternative strategies and consumer-friendly preservation methods has become a focal point for the food sector.

The application of lactic acid bacteria (LAB) that produce antifungal activity is a promising substitute to the use of chemical preservatives because of their GRAS (Generally Regarded As Safe) status and widespread exploitation in various food fermentations (Rouse and van Sinderen 2008b). The number of reports concerning antifungal LAB, the inhibitory compounds they produce and their application as biopreservatives has expanded in recent years (Mauch et al. 2010; Adebayo and Aderiye 2011). An array of antifungal compounds from LAB has been described including proteinaceous compounds, cyclic dipeptides, hydroxyl fatty acids, phenyllactic acid and bacteriocin-like substances (Okkers et al. 1999; Strom et al. 2002; Sjorgen et al. 2003; Falguni et al. 2010; Prema et al. 2010). More recently, novel compounds such as dodecalacetone have been associated with fungistatic activities (Yang et al. 2011). The application of LAB as antifungal agents in foods has also been well documented in recent years. Antifungal metabolite-producing isolates of LAB have been successfully applied in the preservation of a variety of foods such as cheese, cucumbers, apples, corn and soybeans (El-Shafei et al. 2008; Sathe et al., 2007; Rouse et al. 2008a; Khanfari et al. 2007; Yang and Chang 2010). The preservation of breads using antifungal LAB has also been demonstrated (Dal Bello et al. 2007; Gerez et al. 2009; Rizzello et al. 2011). Ryan et al. (2011) recently reported that sourdough bread fermented with the antifungal strain Lactobacillus amylovorus DSM19280 possessed an extended shelf life of 2 weeks compared to control breads fermented with a non-antifungal strain or those treated with calcium propionate. In addition, sourdough fermented with Lact. plantarum 1A7 did not exhibit fungal contamination until 7 weeks' storage (Coda et al. 2011). The aerobic quality of silage can also be improved by the addition of Lact. buchneri strains, which have been shown to reduce yeast counts (Driehuis et al. 1999; Reich and Kung 2010).

Food spoilage attributable to yeasts represents a significant problem for both the beverage and dairy industries. Their ability to survive at low temperatures and pH levels allows them to cause frequent spoilage of foods such as yoghurts, cheeses and juices (Cangella et al. 1998). Spoilage yeasts have been associated with a variety of foods including meats and fermented vegetables (Savard et al. 2002; Nielsen et al. 2008). Growth of spoilage yeasts in foodstuffs results in deterioration of the optical, physical and organoleptic properties of foods (Loureiro and Querol 1999), aside from serious human health implications in immunocompromised patients (de Llanos et al. 2006).

The aim of this study was to examine two antifungal metabolite-producing Lact. plantarum isolates with respect to their technological properties/robustness and efficacy as food preservatives, as well as determining the antiyeast capabilities against Rhodotorula mucilaginosa in refrigerated foods.

Materials and methods

Cultures and growth conditions

Lactobacillus plantarum 16 (which was deposited at the National Collection of Industrial, Food and Marine Bacteria culture collection as NCIMB41875) and 62 (deposited as NCIMB41876) are antifungal isolates from steep water and sauerkraut, respectively (S. Crowley, unpublished data), and were cultivated in MRS broth for 24–48 h at 30°C under anaerobic conditions. Rhodotorula mucilaginosa was obtained from the UCC culture collection and was cultivated in Sabouraud-4%-dextrose broth (Merck Ltd, Darmstadt, Germany) at 30°C with agitation for 16 h. Penicillium expansum was grown on Sabouraud-4%-dextrose agar (Merck Ltd) at 30°C aerobically until sporulation occurred. Streptococcus thermophilus ST00011 and Lactobacillus delbrueckii ssp. bulgaricus CH2 were cultivated in 10% reconstituted skimmed milk (RSM) at 42 and 37°C, respectively (Table 1).

Table 1. Bacterial and fungal cultures used during this study
CultureSource
Lactobacillus plantarum 16Steep water
Lactobacillus plantarum 62Sauerkraut
Streptococcus thermophilus ST00011UCC Culture Collection
Lactobacillus bulgaricus CH2UCC Culture Collection
Penicillium expansum DSMZ
Rhodotorula mucilaginosa UCC Culture Collection

Viability tests using various stress conditions

Freeze-drying and subsequent storage of lyophilized cells at −80°C

The effect of freeze-drying on the viability of Lact. plantarum strains 16 and 62 was assessed over a 4-week period. Each isolate was grown for 48 h in MRS broth and viable cell counts performed to determine cell numbers before freeze-drying. Serial dilutions were performed in quarter-strength Ringers solution with appropriate dilutions plated onto MRS agar. 10-ml samples of the 48-h cultures were taken, and cells were collected by centrifugation at 4400 g for 15 min, washed twice in quarter-strength Ringers and finally resuspended in 10% reconstituted skimmed milk (RSM) to a final volume of 10 ml. The cultures were freeze-dried overnight in a freeze drier (Labconco Freezone 6; Labconco, Kansas City, MO), and lyophilized cells were consequently stored at −80°C. Cells were then resuspended in quarter-strength Ringers, and viable cell counts were determined following freeze treatment at regular intervals over a 4-week period on MRS agar at 30°C.

Salt tolerance

The salt tolerance of Lact. plantarum strains 16 and 62 was assessed by the inoculation of each strain into MRS broth (Oxoid Ltd, Basingstoke, UK) supplemented with a particular (0, 2·5, 5·0, 7·5 or 10%) sodium chloride concentration. Viable cell counts were determined at regular intervals over a 24-h period on MRS agar at 30°C.

Low temperature

The ability of Lact. plantarum strains 16 and 62 to survive low temperatures was examined adapted from the study by Sheehan et al. (2006). Briefly, each isolate was grown overnight in MRS broth, and 1·5 ml of culture was harvested by centrifugation at 4000 g for 10 min. Cells were washed twice in quarter-strength Ringers and finally resuspended to the original volume in MRS broth, 10% RSM or sodium chloride solution (0·9%). Samples were stored at −20°C until further use. Viable cell counts were performed on MRS agar before freezing and following five freeze-thaw cycles.

Comparative analysis of antifungal activity

The cell-free supernatant (CFS) of both isolates was compared to commercially available preservatives according to the method of Yang and Chang (2010). Lactobacillus plantarum strain 16 or 62 was grown anaerobically in 200 ml MRS broth for 48 h at 30°C anaerobically (1% inoculum of a culture containing c. 107 CFU ml−1). Cells were removed by centrifugation at 4400 g for 15 min. The supernatant was freeze-dried overnight and was then concentrated 20 times relative to the original volume using 20 mmol l−1 sodium acetate (pH 4·0) in line with Yang and Chang (2010). The following commercial preservatives were used: sodium benzoate (0·1, 0·5 and 1·0%), potassium sorbate (0·1, 0·5 and 1·0%), acetic acid (1·0, 3·0 and 5·0%), benzoic acid (0·1 and 0·5%), calcium lactate (1·0, 2·0 and 3·0%) and calcium propionate (1·0, 2·0 and 3·0%). The concentrated supernatant was compared to the commercial preservatives using the paper disc assay. Briefly, sterile paper discs (Sigma-Aldrich, Schnelldorf, Germany) were soaked in concentrated cell-free supernatant (cCFS) or test preservative and placed onto Sabauroud-4%-dextrose agar that had been seeded with either R. mucilaginosa or P. expansum (c. 10CFU ml−1). The plates were incubated at 30°C for 48 h aerobically and subsequently examined for zones of clearance surrounding the preservative-containing discs. Twenty millimolar sodium acetate (pH 4·0) was also tested for inhibitory activity as a control.

Orange juice trial

Lactobacillus plantarum 16 and 62 were grown in MRS broth for 48 h under anaerobic conditions. One millilitre of each of these cultures was harvested at 4400 g for 10 min. Collected cells were then washed twice in quarter-strength Ringers and resuspended in the same volume of orange juice (commercial product), which was then inoculated into 40 ml of orange juice to give a final concentration of c. 107–108 cells ml−1 of orange juice. Each trial was set up in triplicate and stored at 4 or 25°C for 4 weeks. Viable cell counts were determined at regular intervals over the 4-week period to determine both LAB and yeast levels in the orange juice. LAB were recovered on MRS agar at 30°C, while R. mucilaginosa was enumerated on Sabouraud-4%-dextrose agar at 30°C. 40 ml orange juice inoculated with R. mucilaginosa was used as a control for the trial.

The effect of thermal treatments commonly employed in the food industry on the viability of Lact. plantarum strains 16 and 62 was investigated in orange juice according to the study by Sheehan et al. (2007). Briefly, 1-ml volumes of orange juice containing c. 10CFU ml−1 of Lact. plantarum 16 or 62 were exposed to a heat treatment at either 90°C for 1 min or 72°C for 30 s in a water bath. Viable cell counts were performed before and after each heat treatment on MRS agar.

Growth in 10% Reconstituted Skimmed Milk (RSM)

Both lactobacilli were assessed for their ability to maintain cell viability in 10% RSM and to determine their potential as starter culture adjuncts in fermented milk products. Lactobacillus plantarum 16 or 62 were grown in 10% RSM for 48 h at 30°C. Ten per cent RSM was inoculated with 1% of each milk-grown culture to give a final concentration of c. 105 CFU ml−1. Both cell numbers and pH levels were monitored at intervals over a 6-day period at 30°C. Each sample was assessed in triplicate.

Yoghurt production

A small-scale production of yoghurt was used to assess the role of Lactobacillus plantarum strains 16 and 62 as antifungal adjuncts. Forty millilitre volumes of 10% RSM were pasteurized, following which Streptococcus thermophilus ST00011 and Lactobacillus delbrueckii subsp. bulgaricus CH2 were inoculated at 1%, while Lact. plantarum 16 or 62 were inoculated to give a final concentration of c. 10CFU ml−1. Fermentation was allowed at 37°C for 4–5 h until a pH of 4·6 was reached, after which the resulting yoghurts were stored at 4°C. After 16 h storage at 4°C, each of the yoghurts was inoculated with R. mucilaginosa to give a final viable count ranging between 101 and 10CFU ml−1. Control yoghurts were also prepared without Lact. plantarum 16 and 62. LAB were enumerated on MRS agar and yeast cells recovered on Sabouraud-4%-dextrose agar supplemented with 10 μg ml−1 chloramphenicol to prevent bacterial growth.

Effect of cCFS from Lact. plantarum 16 on R. mucilaginosa

The effect of cCFS on growth of R. mucilaginosa was examined as adapted from Guo et al. (2011). A 5-ml suspension of R. mucilaginosa cells (c. 10CFU ml−1) was harvested at 5500 rpm to obtain pelleted cells that were consequently resuspended in an equal volume of Sabouraud dextrose broth. The SDB suspension contained 20× cCFS, prepared as described above, to produce a final concentration of 5 or 10% (w/v), and was incubated at 30°C aerobically with gentle agitation. Samples were taken at regular intervals over a 96-h period and examined at either 400 or 1000× magnification under the microscope. Acidified control samples were also set up using the appropriate volume of concentrated MRS broth adjusted to pH 3·6 using lactic acid (pH 3·6 = pH of cCFS from Lact. plantarum 16).

Effect of 10% cCFS from Lact. plantarum 16 and 62 on the growth of R. mucilaginosa in orange juice and yoghurt

The influence of cCFS from Lact. plantarum 16 and 62 on the growth of R. mucilaginosa in orange juice and yoghurt was assessed. Orange juice and yoghurt samples were prepared as outlined above (except in the case of the yoghurts where no antifungal adjuncts were used). The yoghurt and orange juice samples were supplemented with cCFS from both isolates to a final amount of 10% (w/v). The orange juice and yoghurts were subsequently spiked with R. mucilaginosa (3–5 × 101 CFU ml−1) and stored at 4°C. Orange juice/yoghurt containing no cCFS served as controls. Yeast levels were monitored at regular intervals over a 30-day period stored at 4°C.

Statistical analysis

The unpaired student t-test was conducted to determine significant differences (P ≤ 0·05) between R. mucilaginosa levels in control and test samples for yoghurt and orange juice trials. The null hypothesis is being that there was no difference between yeast levels.

Results

Effect of freeze-drying and subsequent storage of lyophilized cells at −80°C on viability of Lact. plantarum 16 and 62

Lactobacillus plantarum strains 16 and 62 were assessed for their ability to survive various treatments such as freeze-drying and low-temperature exposure. The ability of industrially relevant strains to survive various treatments that are frequently applied during large-scale food production is a desirable attribute. Firstly, the viability of both Lact. plantarum isolates following exposure to freeze-drying was investigated. After a period of 4 weeks storage as lyophilized cells at −80°C, the viability of Lact. plantarum 16 decreased from 1·8 × 109 CFU ml−1 to 1·43 × 10CFU ml−1, exhibiting a 79·4% survival rate. Lactobacillus plantarum 62 were found to display a 76·0% survival after 4 weeks' storage (Fig. 1), thus showing that both strains possess excellent survival properties under such storage conditions.

Figure 1.

Percentage survival after freeze-drying of (a) Lactobacillus plantarum 16 and (b) Lact. plantarum 62. Error bars represent standard deviation of three replicates. One day post-treatment (□), 1 week post-treatment (image), 2 weeks post-treatment (■), 3 weeks post-treatment (image), 4 weeks post-treatment (image).

Salt tolerance

Growth of both strains was comparable in MRS supplemented with 0–7·5% sodium chloride reaching viable counts of 109 CFU ml−1 after 24 h of growth, while 10% sodium chloride was found to exert a bacteriostatic effect upon Lact. plantarum 16 and Lact. plantarum 62 (both maintaining viability levels in the order of 107 CFU ml−1) (data not shown). The ability to survive such an osmotic challenge thus makes these strains amenable to application in dried or salted foods.

Low-temperature viability

The viability of each isolate after five freeze-thaw cycles was evaluated in three types of media: 10% RSM, 0·9% sodium chloride and MRS broth. Both strains showed similar patterns of survival in 10% RSM, and this appeared to be the optimal cryoprotectant for both strains. Lactobacillus plantarum 16 cell numbers decreased from 1·5 × 109 CFU ml−1 to 1·3 × 109 CFU ml−1 in 10% RSM, showing a survival rate of 87% compared to a 48% survival rate for Lact. plantarum 62 in 10% RSM. 0·9% sodium chloride solution proved to have the least protective properties with only 15·4% of Lact. plantarum 16 cells surviving; though, under similar conditions, Lact. plantarum 62 showed a higher survival rate of 33·4%. Upon storage in MRS broth, viable cell numbers decreased from 1·5 × 109 CFU ml−1 to 9·0 × 108 CFU ml−1 for Lact. plantarum 16 (60% survival), while final Lact. plantarum 62 viable counts were found to be 2·0 × 109 CFU ml−1 from an initial number of 4·6 × 109 CFU ml−1 (43% survival) (data not shown).

Commercial preservatives

Lactobacillus plantarum 16 and 62, steep water and sauerkraut isolates, respectively, were previously shown to possess broad spectrum antifungal activities towards eight food spoilage fungi including Rhizopus stolonifer and P. expansum (S. Crowley, unpublished data). We wanted to investigate the technological robustness and potential of both lactobacilli as biopreservatives. For this purpose, we first compared the antifungal activity contained within cCFS from both strains with that exhibited by commercially available chemical preservatives. P. expansum and R. mucilaginosa were selected as target fungi. cCFS obtained from each of the two Lactobacillus strains was found to induce fungicidal effects upon the target organisms and was as effective as an antifungal agent as 5% acetic acid (equivalent to 833 mmol l−1), 0·5% potassium sorbate or 1% sodium benzoate (Fig. 2). Additional preservatives such as calcium propionate, benzoic acid and calcium lactate were also examined. Under the conditions tested, the latter preservatives (at concentrations of up to 3%) were found to be less effective as antifungal agents compared to cCFS from the two Lact. plantarum strains (data not shown). The control, 20 mmol l−1 sodium acetate, showed no inhibitory activity towards the target fungi.

Figure 2.

Comparison of antifungal compound(s) produced from Lactobacillus plantarum 16 (a) and 62 (b) to 0·1% (c), 0·5% (d) and 1·0% (e) sodium benzoate against Rhodotorula mucilaginosa.

Orange juice trials

To demonstrate the ability of the antifungal isolates to increase the shelf life of orange juice, we performed three different trials. Varying inoculation levels and storage temperatures were examined during the course of the trials. Firstly, the effects of two different storage temperatures were examined. Lactobacillus plantarum 16 exerted the best protective properties against R. mucilaginosa when inoculated into orange juice stored at 25°C. At room temperature, yeast levels in the control juice reached 1·9 × 107 CFU ml−1. The orange juice seeded with Lact. plantarum 16 resulted in a fungistatic effect with a five-log reduction in yeast cell numbers when the antifungal isolate was introduced at a final concentration of 5·7 × 10CFU ml−1 (P < 0·001). A two-log reduction in R. mucilaginosa levels was observed in the orange juice containing the antifungal Lact. plantarum 62 seeded at a level of 6·5 × 10CFU ml−1 (P < 0·001) (Fig. 3a). During the experimental period, R. mucilaginosa numbers were below the limit of detection (101 CFU ml−1) from days 8 to 25 and 8 to 14 for Lact. plantarum 16 and 62, respectively, with a slight increase in numbers after this period (Fig. 3a). After 30 days' storage at room temperature, there was a noticeable difference in the appearance of the control juice as compared to the juice containing either of the protective cultures. The control sample displayed visible spoilage with discolouration and malodorous properties compared to the juices supplemented with Lact. plantarum 16 or 62 which retained the original colour and odour throughout the trial.

Figure 3.

(a) Effect of Lactobacillus plantarum 16 and 62 on the growth of Rhodotorula mucilaginosa in orange juice at 25°C. (b) Effect of Lact. plantarum 16 and 62 on the growth of R. mucilaginosa in orange juice at 4°C (high-yeast inoculum). (c) Effect of Lact. plantarum 16 and 62 on the growth of R. mucilaginosa in orange juice at 4°C (low-yeast inoculum). Error bars represent standard deviation of three replicates. Lactobacillus plantarum 16 (CFU ml−1) (image), Lact. plantarum 62 (CFU ml−1) (image), R. mucilaginosa control (image), R. mucilaginosa levels in juice containing Lact. plantarum 62 (image), R. mucilaginosa levels in juice containing Lact. plantarum 16 (image). P < 0·01 (**), P < 0·001 (***).

The effect of varying inoculation levels of 101 and 102 CFU ml−1 on the protective capabilities of both isolates was also examined in orange juice stored at 4°C. Rhodotorula mucilaginosa levels reached 1·7 × 107 CFU ml−1 in the control juice not containing any protective cultures. Juice challenged with yeast levels of 5 × 102 CFU ml−1 at 4°C Lact. plantarum 16 reduced the contaminant levels by one log (P < 0·001). However, only a partial decrease in yeast cell numbers was observed in the juice inoculated with Lact. plantarum 62 compared to the control (P < 0·01) (Fig. 3b). The effect of a lower level of Rmucilaginosa (5 × 101 CFU ml−1) was also investigated in a separate trial. In the juice containing Lact. plantarum 16 (1 × 10CFU ml−1), the numbers of viable yeast cells reached a level that was three log less than that found in juice without the antifungal LAB (P < 0·001). The juice containing Lact. plantarum 62 elicited a 10-fold reduction in the viable count of R. mucilaginosa as compared to the control (P < 0·001) (Fig. 3c). Optimal inhibition was observed when the juice was stored at room temperature with LAB levels of 1 × 109 CFU ml−1 and a contamination level of 5–6 × 101 CFU ml−1.

Effects of heat treatment on the survival of LAB in orange juice

Pasteurization conditions of 90°C for 1 min were found to completely inactivate the isolates, and therefore, a milder heat treatment was investigated to assess the survival capabilities of the two antifungal strains. Following treatment at 72°C for 30 s, Lact. plantarum 16 cell numbers decreased from 1·83 × 109 CFU ml−1 to 4·06 × 104 CFU ml−1, while Lact. plantarum 62 cell numbers decreased from 2·3 × 109 CFU ml−1 to 8·0 × 104 CFU ml−1, in line with expectations of a treatment resembling pasteurization (data not shown).

Yoghurt production

The ability of the two antifungal isolates to grow in milk was investigated. Both strains were inoculated into 10% RSM to a final viable count of 2–3 × 105 CFU ml−1. A two-log increase in viable cell numbers was observed for both strains over a 7-day period at 30°C. Lactobacillus plantarum 16 reached cell numbers of 4·9 × 107 CFU ml−1, while Lact. plantarum 62 reached 8·7 × 107 CFU ml−1. Final pH levels of 4·6 and 5·5 were observed for Lact. plantarum 16 and 62, respectively, which represents a significant drop from the initial pH value of the medium (pH = 6·33). The antiyeast activity of both strains was investigated in a yoghurt model employing R. mucilaginosa as the spoilage organism. Both Lact. plantarum strains maintained good viability in the yoghurt during 4 weeks' storage at 4°C with levels of c. 108 CFU ml−1 maintained throughout the trial as well as exerting a fungistatic effect on the target organism. On day 1 of the trial, a yeast viable count of 1·3 × 102 CFU ml−1 was determined, while following 30 days of storage, yeast levels of 2·1 × 102 CFU ml−1 were seen, showing little increase in viable cell numbers throughout the 4-week storage period (P < 0·001) (Fig. 4a). Lactobacillus plantarum 62 also showed a partial decrease in yeast cell numbers, with a 10-fold reduction compared to the control yoghurt (P < 0·01) (Fig. 4b). The control yoghurt (fermented without antifungal metabolite-producing LAB) contained 1·1 × 106 CFU ml−1 of yeast cells after 30 days storage at 4°C.

Figure 4.

(a) Effect of Lactobacillus plantarum 16 on the growth of Rhodotorula mucilaginosa in yoghurt at 4°C. Lactobacillus plantarum 16 (image), R. mucilaginosa in control (image), R. mucilaginosa levels in yoghurt containing Lact. plantarum 16 (image). P < 0·001 (***). (b) Effect of Lact. plantarum 62 on the growth of R. mucilaginosa in yoghurt at 4°C. Lactobacillus planatrum 62 (image), R. mucilaginosa in control (image), R. mucilaginosa levels in yoghurt containing Lact. plantarum 62 (image). P < 0·01 (**). Error bars represent standard deviation of three replicates.

Effect of cCFS from Lact. plantarum 16 on the growth of R. mucilaginosa

The impact of cCFS from the most potent inhibitor (Lact. plantarum 16) on the growth of R. musilaginosa was investigated by monitoring morphological changes as visualized by microscopic examination. Addition of cCFS (10% final concentration) from Lact. plantarum 16 exerted a fungicidal effect by inhibiting budding and growth of yeast cells of R. mucilaginosa compared to the control sample. No viable cells were recovered after a period of 96 h, and yeast cells were found to be reduced by 50% in size (2·5 μm) compared to the MRS-treated control (5·0 μm) (Fig. 5a,b) with a final cell concentration of 5·6 × 107 CFU ml−1. Addition of CFS at lower amounts (5% final concentration) resulted in a reduction of cell size and evident cell damage compared to the control sample (data not shown). However, there was nearly a two-log reduction in yeast cell numbers with a final concentration of 1·2 × 10CFU ml−1 and 5·6 × 10CFU ml−1 in the control and treated sample, respectively. Both control and test sample contained c. 1·1 mol l−1 lactate as determined by HPLC.

Figure 5.

(a) Addition of 10% MRS broth control to Rhodotorula mucilaginosa yeast cells. (b) Addition of 10% cCFS from Lactobacillus plantarum 16 to R. mucilaginosa yeast cells.

Effect of cCFS from Lact. plantarum 16 and 62 on the growth of R. mucilaginosa in orange juice and yoghurt

cCFS from Lact. plantarum 16 exerted a fungicidal effect on R. mucilaginosa in both orange juice and yoghurt trials. By days 4 and 7, no yeast cells were recovered in the yoghurt and orange juice samples containing cCFS (10% final concentration) from Lact. plantarum 16, respectively. The cCFS provided protection for the 30-day trial period with yeast cell levels reaching final numbers of 1·6 × 10CFU ml−1 in the control samples (Fig. 6). In conjunction with Lact. plantarum 16, cCFS produced by Lact. plantarum 62 induced a fungicidal effect on R. mucilaginosa cells in the yoghurt sample. No viable yeast cells were detected in the test yoghurts after 4 weeks stored at 4°C compared to the control that reached final cell levels of 1·0 × 106 CFU ml−1 after the test period. In orange juice samples, cCFS from Lact. plantarum 62 exerted a fungistatic effect with yeast cell numbers remaining in the range of 101 CFU ml−1 over 4 weeks. The addition of active CFS from either isolate represents an additional possibility of increasing the shelf life of both orange juice and yoghurt.

Figure 6.

(a) Effect of 16 and 62 cCFS on the growth of Rhodotorula mucilaginosa in orange juice. Control juice (image), juice containing cCFS from Lactobacillus plantarum 16 (image), juice containing cCFS from Lact. plantarum 62 (image). P < 0·001 (***). (b) Effect of 16 and 62 cCFS on the growth of R. mucilaginosa in yoghurt. Control yoghurt (image), yoghurt containing cCFS from Lact. plantarum 16 (image), yoghurt containing cCFS from Lact. plantarum 62 (image). P < 0·001 (***).

Discussion

The rise in consumer preference for more naturally preserved foods has ignited significant interest in the area of food biopreservation using antifungal LAB. In the current study, the assessment of two broad spectrum antifungal isolates to be used as natural food biopreservatives was examined.

Each isolate displayed remarkable tolerance to various treatments commonly encountered in food processing including high levels of sodium chloride. This feature exonerates their suitability for application in certain foods as they can maintain sufficient viability. Furthermore, they were shown to survive freeze-drying, which is of particular importance in the commercial application of such strains. These results are consistent with the findings of Schoug et al. (2006), who showed that the antifungal strain Lact. coryniformis Si3 had a 60% survival rate after optimization of freeze-drying conditions. cCFS from Lact. plantarum 16 or 62 was found to be more effective as an antifungal agent (against certain fungi) than both sodium benzoate and potassium sorbate at their FDA maximum permitted level. This interesting and practical attribute is in line with findings reported by Yang and Chang (2010). The emergence of yeasts resistant to the aforementioned preservatives has been reported, highlighting the need for novel antifungal approaches (Mihyar et al. 1997).

A variety of yeast species have been identified from orange fruits and juices, including Candida tropicalis and R. mucilaginosa (Las Heras-Vazquez et al. 2003). Rhodotorula mucilaginosa has previously been implicated as a spoiler of dairy products and orange juice, and both Lact. plantarum 16 and 62 have previously displayed inhibitory activities towards this selected yeast (S. Crowley, unpublished data). To the best of our knowledge, this is the first report describing the use of an antifungal producing LAB as a bioprotectant against yeast spoilage in orange juice. Many juices can be stored at room temperature in supermarkets, and so the shelf life of juice stored under these conditions, as well as at refrigeration temperatures, was investigated. Both isolates showed high levels of acid tolerance in orange juice (pH 3·3) over a 4-week period maintaining levels of c. 108 CFU ml−1. Similar findings have been observed by Sheehan et al. (2007) where commercial strains of LAB, including L. casei, were examined for their growth in fruit juices. These strains showed viability of between 107 and 108 CFU ml−1 after a 4-week period in orange juice. Champagne and Gardner (2008) also documented comparable results of lactobacilli stored in fruit juice. Storage at room temperature proved to be the optimal condition for inhibitory effects of both strains, which was also visually apparent. However, antifungal activities were also observed upon storage at 4°C albeit to a lesser extent as compared to the results obtained at room temperature. Inhibitory activities of both Lactobacillus strains were noted to be dependent on initial yeast numbers. In agreement with Sheehan et al. (2007), both Lact. plantarum strains were inactivated after treatment at 90°C for 1 min in orange juice. After treatment at 72°C for 30 s, a four-log reduction in Lact. plantarum 16 and 62 cell numbers was observed. However, previous experiments (S. Crowley, unpublished data) have determined that the antifungal compound(s) produced by Lact. plantarum 16 and 62 are heat stable at temperatures of 121°C for 15 min leading to potential application of the metabolites prior to pasteurization or addition of the isolates post heat treatment. Both isolates showed successful growth and reduction of pH in 10% RSM and therefore were investigated for their potential as preservatives in yoghurt. Contaminant levels of 101–102 CFU ml−1 R. mucilaginosa were chosen for such trials to produce a yoghurt that is representative of everyday storage and being prone to low levels of fungal contamination. The two Lact. plantarum strains provided a fungistatic effect on the spoilage yeast in the in vivo model, with Lact. plantarum 16 providing the most effective protection, which is in agreement with the findings of the orange juice trial.

LAB have previously been implicated as antifungal adjuncts in conjunction with propionibacteria in yoghurts. The antifungal strains Lact. paracasei subsp. paracasei SM20, SM29 and SM63 were shown to inhibit the growth of a number of spoilage yeasts in yoghurt in combination with a mixed culture of Propionibacterium jensenii (Schwenninger and Meile 2004). To the best of our knowledge, this is the first report of antifungal Lact. plantarum strains being used as antifungal adjuncts for increased shelf life of yoghurt. A recent report investigated the inhibitory effects of potassium sorbate as an antifungal against a variety of moulds and yeasts in yoghurt. The preservative was found to reduce fungal counts by four logs when applied at a concentration of 0·2% (Al-Ashmawy and Ibrahim 2009). This suggests that Lact. plantarum 16 under the conditions tested is as effective as 0·2% potassium sorbate as an antifungal adjunct in yoghurt.

Both isolates have been tested in combination for a synergistic antifungal effect; however, it was found that no synergism exists between the isolates. Sensory analysis of both yoghurt and juice samples protected with Lact. plantarum 16 and 62 is required to assess whether the protective LAB would alter the organoleptic properties or quality of the final products. Additionally, safety evaluation of the application of both isolates in a food matrix would need to be determined. Recently, an in vivo mouse acute oral toxicity safety assay deemed Lact. plantarum DW3, an antifungal starter used in plant fermented beverages, most likely to be safe for human consumption (Kantachote et al. 2010). Previous experiments have determined that the compound(s) in question are nonproteolytic and inactivated after neutralization (S. Crowley, unpublished data). An investigation of the antifungal compounds produced by Lact. plantarum 16 is currently underway and will provide a deeper insight into the diversity of compounds capable of inhibiting fungal spoilage of foods; however, the use of the aforementioned Lact. plantarum isolates as alternative bioprotectants to chemical preservatives is promising.

Acknowledgements

S. Crowley is the recipient of a Lauritzson Foundation scholarship. D. van Sinderen is a recipient of a Science Foundation Ireland (SFI) Principal Investigator award (Ref. no. 08/IN.1/B1909).

Ancillary