Numerous microencapsulation strategies have been examined for the ability to protect probiotic bacteria from environmental stresses (11, 13), but there remains considerable scope for the development of microencapsulants that fulfill the many demands of a successful probiotic encapsulant. These include protection against oxygen, heat, and other environmental stresses during drying, formulation, and storage; protection against low pH and proteases during gastric transit; efficient release of the bacteria within the gastrointestine; production of small capsules with suitable sensory properties; and the use of materials that are inexpensive, stable, and of food grade.

In the current investigation, a Bifidobacterium infantis strain was microencapsulated within a novel, film-forming protein-carbohydrate-lipid emulsion. Nondigestible prebiotic carbohydrates, which specifically stimulate the growth and/or activity of beneficial populations of bacteria within the gut (3, 12), formed part of the encapsulant formulation to produce a synbiotic (9) combination. The capacity of the encapsulant to protect the viability of the bacteria was evaluated during nonrefrigerated storage and during simulated gastrointestinal transit.

Bifidobacterium infantis Bb-02 was obtained from C. Hansen A/S (Hoersholm, Denmark) as a freeze-dried powder containing 3.2 × 10¹⁰ CFU/g. The freeze-dried bacteria were microencapsulated within a film-forming protein-carbohydrate-oil emulsion according to the rationale and methodologies described by Crittenden et al. (4). Briefly, oil-in-water emulsions were prepared, containing canola vegetable oil (Crisco; Goodman Fielder, Australia), caseinate (Alanate 180; New Zealand Milk Products, New Zealand), and prebiotic fructo-oligosaccharides (FOS) (Raftilose P95; ORAFTI, Belgium) plus either dried glucose syrup (DGS) (Maltostar 30; Manildra, Australia) or microfluidized resistant starch (RS) (Hylon VII; National Starches). The emulsions were heated to 98°C for 30 min to promote Maillard reaction products, which improve film-forming and oxygen-scavenging properties (17). The mixture was then cooled to 10°C before the addition of the probiotic bacteria at 8% (wt/wt) prior to spray drying. Both encapsulated and nonencapsulated B. infantis cells were spray dried using a Drytec (Tonbridge, United Kingdom) laboratory-scale spray dryer with an inlet temperature of 160°C and an outlet temperature of 65°C. The final formulation (wt/wt) of the dried powders was as follows: 8% probiotic, 32% oil, 20% caseinate, 20% FOS, and either 20% DGS or 20% RS.

The water activities of powdered samples were measured after 2 months of storage in sealed aluminum foil pouches. Duplicate samples were measured using a Decagon CX-2 water activity meter (Decagon Devices, WA). The particle size distribution of the microcapsules was measured using a Malvern Mastersizer 2000 instrument (Malvern Instruments Ltd., United Kingdom).

The survival of encapsulated and nonencapsulated B. infantis cells was assessed during nonrefrigerated storage over a 5-week period. The bacterial samples were stored in open containers exposed to an ambient atmosphere maintained at 25°C and 50% relative humidity. The survival of the bacteria was assessed after 2 and 5 weeks in four repeat experiments.

The ability of the encapsulant to protect B. infantis during gastrointestinal transit was assessed using a two-stage in vitro model simulating conditions in the human stomach and small intestine (22). Details of the model are described by Crittenden et al. (4), but briefly, 1.0-g samples of encapsulated or nonencapsulated bacteria were mixed with 10 ml of simulated gastric fluid (SGF) at pH 1.2 and incubated for 2 h at 37°C. Following incubation in SGF, the pH of the samples was adjusted to 6.8, and the bacteria were diluted 10-fold in simulated intestinal fluid (SIF) at pH 6.8 and incubated for a further 3 h at 37°C. Viable counts of the bacteria were compared before and after passage through both stages of the in vitro model. Survival was assessed in four repeat experiments.

Viable bacteria were enumerated in duplicate samples cultured anaerobically on reinforced clostridial agar for 48 h at 37°C. In cases where bacteria had to first be released from the microcapsules, the capsules were dissolved by incubation in SIF for 1 h at 37°C. The detection limit of the analysis was 3 × 10² CFU/g of encapsulated material. The statistical significance of differences between treatments was assessed using one-way analysis of variance and the Tukey test.

The spray-dried microcapsules produced were small (15 to 20 μm in diameter), with low water activities (0.2 to 0.3). Scanning electron micrographs of unwashed microencapsulated bacterial powders showed that no free, nonencapsulated bacteria were present, indicating that the encapsulation efficiency was high (Fig. 1).

FIG. 1. Scanning electron micrograph (magnification, ×1,000) of spray-dried encapsulated Bifidobacterium infantis.

Microencapsulation significantly protected the viability of bacteria when they were stored in an open container at 25°C and 50% relative humidity (Fig. 2). Similarly, microencapsulation significantly protected the bacteria during simulated gastrointestinal transit (Fig. 3). The addition of the encapsulant materials to nonencapsulated bacteria increased survival in the gastrointestinal model but afforded substantially less protection than the formulated microcapsules. Microscopic examination of the microcapsules showed that the bacteria remained entrapped within the capsule material in SGF and were released when transferred to SIF (Fig. 4).

FIG. 2. Survival of Bifidobacterium infantis Bb-02 during storage for 5 weeks at 25°C and 50% relative humidity. •, nonencapsulated bacteria; , bacteria encapsulated in caseinate-FOS-oil-DGS emulsion; , bacteria encapsulated (more ...)

FIG. 3. Effect of microencapsulation on survival of Bifidobacterium infantis in an in vitro model simulating gastrointestinal transit. The encapsulant used was the caseinate-FOS-oil-DGS emulsion. The data represent the means of four experiments. Error bars represent (more ...)

FIG. 4. Light micrographs (magnification, ×1,000) of Gram-stained microcapsules (caseinate-FOS-oil-DGS emulsion) containing Bifidobacterium infantis after 2 h in SGF at pH 1.2 (A) and following subsequent transfer to SIF at pH 6.8 for 3 h (B).

The small diameters of the microcapsules produced in the current study (<20 μm) were comparable to those of microcapsules produced to protect probiotics using cellulose acetate phthalate (8) or starch (15). These microcapsules are much smaller than those produced using gelling materials such as alginate (7, 13, 19) or gellan and xanthan gums (20) (typically >100 μm) and provide advantages with regard to minimizing adverse impacts on texture and mouth feel when incorporated into foods (1).

Numerous microencapsulation strategies have been evaluated for the ability to protect probiotic viability during storage and intestinal transit (11, 13). Unfortunately, it is difficult to directly compare the degrees of effectiveness of different approaches due to substantial variations in the durabilities of the probiotic strains tested and differences in the methods used to challenge the probiotics with environmental conditions. A strain of B. infantis was selected for the current study since it is a species that is autochthonous to the human gastrointestinal tract and is relatively sensitive to environmental conditions such as low pH, temperature, and oxygen in comparison to many commercial probiotics (2, 16). Encapsulation with alginate, carrageenan, and modified starch has generally failed to substantially improve the survival of acid-sensitive bifidobacteria in models simulating the acidic conditions in the stomach (5, 7, 10, 14, 15, 18, 19). However, the encapsulant tested in the current study was highly effective in protecting B. infantis during simulated gastrointestinal transit. Additionally, the capsules remained intact in SGF and dissolved rapidly in SIF, suggesting that the bacteria would be released in the small intestine. The fact that encapsulated B. infantis was also protected during nonrefrigerated storage with exposure to oxygen and humidity is particularly encouraging, as it demonstrates the potential of this technology to extend the application of environmentally sensitive probiotics to nonrefrigerated, long-shelf-life food and pharmaceutical products.

We gratefully acknowledge the advice provided by Martin Playne of Melbourne Biotechnology, Australia, and technical support from Sieh Ng and Jenny Rusli of Food Science Australia. We thank Christine Coombs of CSIRO Textile and Fiber Technology, Australia, for producing the scanning electron micrograph of the encapsulated bacteria.