Cereal used

Three different lots of the same brand of commercially manufactured dry infant rice, barley, mixed grain, and oatmeal cereals were purchased from local retailers in Griffin, Ga. The three lots of each type of cereal were used in three replicate trials. Infant rice, barley, mixed grain, and oatmeal cereals contained rice flour, barley flour, wheat, rice, and oat flour, and oat flour, respectively. In addition, all cereals contained added soy oil-lecithin, tri- and dicalcium phosphate, tocopherols (vitamin E), electrolytic iron, zinc sulfate, niacinamide, riboflavin

(vitamin B2), pyridoxide hydrochloride (vitamin B6), thiamin (vitamin B1), folic acid, and cyanocobalamin (vitamin B12).

2.3. Preparation of inoculum for cereals

Cells of the ten strains of E. sakazakii in aqueous 15% glycerol solution were stored at -31°C. Each suspension was thawed and streaked with a wire loop on violet red bile glucose agar (VRBG, pH 7.4; Oxoid Ltd., Basingstoke, Hampshire, England) and incubated at 37°C for 24 h. Cells from single colonies of each strain were streaked on tryptic soy agar (TSA, pH 7.3;

Difco; Dickinson and Co., Sparks, Md.), incubated at 37ºC for 24 h, and stored at 4°C until used to prepare inocula.

To prepare inocula for cereals, each strain of E. sakazakii was grown in 9 ml of brain heart infusion broth (BHI, pH 7.4; Becton, Dickinson and Co.) at 37°C, with two successive loop transfers (ca. 10 µl) at 24-h intervals. A final transfer of 0.1 ml was made from the second 24-h culture to 400 ml of BHI broth and held in an incubator shaker (New Brunswick Scientific, Edison, N.J.) set at 62 rpm for 24 h at 37°C. The ten 400-ml BHI broth cultures were centrifuged for 15 min at 2,700 x g in a Marathon 12KBR Benchtop Refrigerated Centrifuge (Fisher Scientific, Pittsburgh, Pa.) and the supernate was decanted. Cells were resuspended in 100 ml of sterile deionized water, centrifuged a second time, and resuspended in 10 ml of sterile deionized water. To determine populations, suspensions were serially diluted in sterile 0.1%

peptone water, surface plated on TSA supplemented with 0.1% sodium pyruvate (TSAP) (Sigma-Aldrich, Inc., St. Louis, Mo.) and VRBG supplemented with 0.1% sodium pyruvate (VRBGP), and incubated at 37ºC for 24 h. Sodium pyruvate was added to aid in the recovery of

injured cells (Baird-Parker and Davenport, 1965). Colonies were counted and populations of each of the ten strains in the suspensions were calculated.

2.4. Preparation rice cereal inoculum

Ten 100-g portions of a commercially manufactured infant rice cereal (aw 0.35 - 0.41) were distributed in even layers (ca. 2.5 cm deep) in ten sterile stainless steel bowls (40 cm

diameter). Each of the ten suspensions of E. sakazakii cells was separately spray-inoculated onto the surface of separate 100-g portions. Cell suspension was applied using a chromatography reagent sprayer (Model 422530-0050, Kontes Glass Company, Vineland, N.J.). The sprayer was held 45 cm above the surface of the infant rice cereal and suspension was sprayed at ca. 2 psi using nitrogen gas as a carrier. Infant rice cereal was thoroughly stirred using a sterile spoon at 15-sec intervals during application of spray inoculum over a 150-sec period during which a total of ca. 1.0 ml of suspension per 100 g of infant rice cereal was applied. The inoculated infant rice cereal was deposited in a Ziploc^® (Johnson and Son, Inc., Racine, Wisc.) plastic bags, sealed, and vigorously mixed for 2 min.

Each 100-g portion of cereal inoculated with a single strain of E. sakazakii was analyzed for population of the pathogen within 1 h after inoculation, then stored for 7 days at 21ºC in sterile polystyrene trays (22 cm long x 15 cm wide x 5 cm high; Rubbermaid, Wooster, Ohio) placed on wire platforms above the surface of ca. 6 liters of saturated potassium acetate in a hermetically sealed plastic tub (81 cm long x 33 cm wide x 30 cm high). The depth of infant rice cereal inoculum in each tray was ca. 1.5 cm. Two brushless 12VDC cooling fans (6 cm long x 6 cm wide x 2.5 cm high; RadioShack, Fort Worth, Tex.) were attached on opposite internal sides

of the tub to facilitate air circulation. Upon equilibration, the atmospheric relative humidity inside the tub was ca. 23%.

2.4.1. Determination of populations of E. sakazakii in rice cereal inoculum

Within 1 h after inoculation of infant rice cereal and after storage at 21ºC for 7 days, triplicate 10-g samples of each of the ten portions of inoculated infant rice cereal were analyzed to determine populations of E. sakazakii. Each sample was reconstituted in 90 ml of sterile deionized water at 45°C, serially diluted in sterile 0.1% peptone water, surface plated (0.1 ml in duplicate) on TSAP and VRBGP agar, and incubated at 25°C for 48 h and 37°C for 24 h, respectively, before counting colonies and calculating log CFU/g. Portions of each of the ten infant rice cereal inocula were combined and thoroughly mixed for 2 min to give a 10-strain mixture containing approximately equal populations of each strain of E. sakazakii.

To determine the population of E. sakazakii in the rice cereal inoculum (aw 0.23) containing a ten-strain mixture, samples (10 g) were combined with 90 ml of sterile deionized water at 45°C, serially diluted, and surface plated on VRBGP and TSAP agar. Plates were inoculated at 37°C for 24 h before colonies were counted. This inoculated rice cereal, containing 7.3 - 7.5 log CFU/g, served as the high inoculum for rice cereal to be stored for up to 12 months.

To prepare an inoculum containing a low population of E. sakazakii, the high-population inoculum was serially diluted in infant rice cereal (aw 0.23) to give a population of ca. 2 log CFU/g. High-population inoculum or low-population inoculum (2 g) were added to 200 g of uninoculated low-aw infant rice cereal to give desired populations and stored for up to 9 months before analyzing for the number of E. sakazakii that survived. Rice, barley, mixed grain, and

oatmeal cereals (200 g) at high aw were combined with 2 g of high-population inoculum and stored for up to 24 weeks before analysis.

2.5. Preparation of cereals for inoculation and storage

2.5.1. Cereal at low aw

In the first experiment, survival of E. sakazakii in infant rice cereal at target aw values of 0.23, 0.43, and 0.68 was studied. Three different lots of the same brand of commercially manufactured infant rice cereal were used in three replicate trials. Infant rice cereal from each lot was divided into six of 200-g quantities and placed in polystyrene trays (22 cm long x 15 cm wide x 5 cm high) at a depth of ca. 2.5 cm. Two of the six trays of each lot were placed on wire platforms inside each of three sterile plastic tubs (81 cm long x 33 cm wide x 30 cm high)

containing ca. 6 liters of saturated solutions of potassium acetate, potassium carbonate, or lithium acetate, which created equilibrium atmospheric relative humidities of ca. 23, 43, and 68%,

respectively, for the purpose of adjusting the infant rice cereal to aw values of ca. 0.23, 0.43, and 0.68. Two brushless 12VDC cooling fans were attached to opposite internal sides of each tub to facilitate air movement. Cereals were stored, with intermittent mixing, in tubs for 14 days at 21°C, then inoculated with E. sakazakii at high or low population, as described below.

2.5.2. Cereals at high a_w

A second series of experiments was done using three different lots of the same brand of commercially manufactured infant rice, barley, mixed grain, and oatmeal cereals. This study was focused on determining the ability of E. sakazakii to survive at target aw values of ca. 0.68, 0.79, and 0.88. Each lot of the four types of infant cereal was divided into three 200-g quantities

and placed in polystyrene trays (22 cm long x 15 cm wide x 5 cm high) at a depth of ca. 2.5 cm.

Three of the nine trays of each lot and each type of infant cereal were placed on wire platforms inside each of three sterile plastic tubs (81 cm long x 33 cm wide x 30 cm high) containing ca. 6 liters of saturated solutions of lithium acetate, ammonium sulfate, or sodium benzoate, which created equilibrium atmospheric relative humidities of ca. 68, 79, and 88%, respectively, for the purpose of adjusting the infant rice, barley, mixed grain, and oatmeal cereals to aw values of ca.

0.68, 0.79, and 0.88. Two brushless 12 VDC cooling fans were installed inside each tub to facilitate air movement and equilibrium of aw with relative humidity as described above. After storage, with intermittent mixing, in the tubs for 14 days at 21°C, cereals at each aw were separately inoculated with high-population inoculum as described below.

2.6. Inoculation of cereals for storage

Infant rice cereal (200 g) initially at aw 0.30, 0.45 - 0.46, and 0.68 - 0.69 was inoculated with 2 g of the 10-strain mixture of high- or low-population inoculum, placed in a Ziploc plastic bag (17.7 cm x 20.3 cm), sealed, and vigorously mixed for 2 min to give populations of 5.03 ± 0.09 log and 0.31 log CFU/g, respectively. Ten grams of inoculated cereal at each aw and containing high or low numbers of E. sakazakii were deposited in 30-ml glass vials and sealed with polyseal liner phenolic caps (Fisher Scientific). Vials were double bagged in Ziploc plastic bags (17.7 cm x 20.3 cm), hermetically sealed in plastic tubs, and stored at 4, 21, or 30°C for up to 12 months before analyzing cereal for the presence (by enrichment) and populations of E.

sakazakii.

In the second series of experiments, infant rice, barley, mixed grain, and oatmeal cereals (200 g) adjusted to ca. aw values of 0.63 - 0.66, 0.76, and 0.82 - 0.83 were inoculated with 2 g of

the 10-strain mixture of high-population inoculum, sealed in a Ziploc plastic bag (17.7 cm x 20.3 cm), and vigorously mixed for 2 min to give a population of 5.39 ± 0.36 log CFU/g. Sixty grams of each type of inoculated cereal at each aw were deposited in 250-ml glass bottles fitted with polypropylene closures with resin/silicone septa (Innovative Chemical Co., Amherst, N.Y.).

Cereals were stored at 4, 21, or 30°C for up to 24 weeks before analyzing to determine the presence (by enrichment) and populations of E. sakazakii.

2.7. Storage of inocula

Rice cereal inocula (initial aw, 0.23) containing high (ca. 7 log CFU/g) or low (ca. 2 log CFU/g) numbers of E. sakazakii were each divided into three equal portions, placed in plastic bags, sealed in glass jars, and stored at 4, 21, and 30°C. Inocula were analyzed for initial populations of E. sakazakii on day 0 (within 1 h after inoculation) and after storage for up to 9 months.

2.8. Microbiological analysis of inoculated cereal

Each inoculated infant cereal (10 g) initially containing high or low numbers of E.

sakazakii was combined with 90 ml of sterile deionized water at 45°C and vigorously shaken for 1 min. Undiluted mixtures of cereal and water (0.25 ml, in quadruplicate and 0.1 ml, in

duplicate) and suspensions serially diluted in 0.1% sterile peptone (0.1 ml, in duplicate)

containing high-inoculum (7.3 - 7.5 log CFU/g) were surface plated on TSAP and VRBGP agar.

The VRBGP plates were incubated at 37°C for 24 h and TSAP plates were incubated at 25ºC for 48 h before positive-presumptive E. sakazakii colonies were counted. Mixtures of cereal initially containing high- or low-population inoculum (10 g) and water (90 ml) were incubated at 37ºC

for 24 h. Colonies formed on TSAP were examined for yellow pigmentation. If presumptive E.

sakazakii colonies were not detected on TSAP or VRBGP agar, 10 ml of pre-enrichment mixture was combined with 90 ml of Enterobacteriaceae enrichment broth (Becton, Dickinson and Co.) supplemented with 0.1% sodium pyruvate (EEP) and incubated at 37°C for 24 h. These

enrichments were streaked onto TSAP and VRBGP agar and incubated at 25°C for 48 h and 37°C for 24 h, respectively. Presumptive colonies of E. sakazakii formed on TSAP agar were subjected to confirmation assays using the API 20E identification system (bioMérieux,

Hazelwood, Mo.) and the Microbact 12A/B identification kit (Oxoid) according to manufacturers’ instructions.

2.9. Measurement of aw

Measurement of aw of infant cereals was done immediately after opening boxes in which they were commercially packaged, after equilibration in tubs with various atmospheric relative humidities for 14 days at 21°C, and after inoculation and storage for up to 12 months using an aw

meter (Aqua Lab model CX2, Decagon Devices, Inc., Pullman, Wash.).

2.10. Statistical analysis

All experiments were performed in triplicate. Data were analyzed using the general linear model on SAS software (Statistical Analysis Systems Institute, Cary, N.C.). The least significant difference test was used to determine significant differences (p ≤ 0.05) in populations of E. sakazakii detected in infant cereals as affected by aw, composition, storage temperature, and storage time.