Partially Hydrolysed Whey Has Superior Allergy Preventive Capacity Compared to Intact Whey Regardless of Amoxicillin Administration in Brown Norway Rats

Background It remains largely unknown how physicochemical properties of hydrolysed infant formulas influence their allergy preventive capacity, and results from clinical and animal studies comparing the preventive capacity of hydrolysed infant formula with conventional infant formula are inconclusive. Thus, the use of hydrolysed infant formula for allergy prevention in atopy-prone infants is highly debated. Furthermore, knowledge on how gut microbiota influences allergy prevention remains scarce. Objective To gain knowledge on (1) how physicochemical properties of hydrolysed whey products influence the allergy preventive capacity, (2) whether host microbiota disturbance influences allergy prevention, and (3) to what extent hydrolysed whey products influence gut microbiota composition. Methods The preventive capacity of four different ad libitum administered whey products was investigated in Brown Norway rats with either a conventional or an amoxicillin-disturbed gut microbiota. The preventive capacity of products was evaluated as the capacity to reduce whey-specific sensitisation and allergic reactions to intact whey after intraperitoneal post-immunisations with intact whey. Additionally, the direct effect of the whey products on the growth of gut bacteria derived from healthy human infant donors was evaluated by in vitro incubation. Results Two partially hydrolysed whey products with different physicochemical characteristics were found to be superior in preventing whey-specific sensitisation compared to intact and extensively hydrolysed whey products. Daily oral amoxicillin administration, initiated one week prior to intervention with whey products, disturbed the gut microbiota but did not impair the prevention of whey-specific sensitisation. The in vitro incubation of infant faecal samples with whey products indicated that partially hydrolysed whey products might confer a selective advantage to enterococci. Conclusions Our results support the use of partially hydrolysed whey products for prevention of cow’s milk allergy in atopy-predisposed infants regardless of their microbiota status. However, possible direct effects of partially hydrolysed whey products on gut microbiota composition warrants further investigation.


2 DNA extraction and amplicon sequencing of the 16S rRNA gene
DNA was extracted from faeces or small intestine content by DNeasy PowerLyzer PowerSoil Kit (Qiagen, Hilden, Germany) according to the manufacture's protocol. Mechanical lysis of bacteria was conducted at 30 cycles/s twice for 5 min using bead beater MM300 (Retsch, VWR, Haan, Germany).
The V3-region of the 16S rRNA gene was amplified using a universal forward primer (PBU 5'-Aadapter-TCAG-barcode-CCTACGGGAGGCAGCAG-3') with a unique 10-12 bp barcode for each sample (IonXpress barcode as suggested by the supplier, Life Technologies, Carlsbad, CA, US) and a universal reverse primer (PBR 5'-trP1-adapter-ATTACCGCGGCTGCTGG-3') and Phusion High-Fidelity DNA polymerase (Thermo Fisher Scientific, Waltham, MA, US). PCR products were purified by HighPrep™ PCR Clean-up System (Magbio, Gaithersburg, MD, US) according to the manufacture's protocol. DNA concentrations were determined with Qubit HS assay (Life Technologies). Finally, a library was constructed by mixing an equal amount of PCR products from each sample. Sequencing of all samples was performed on a 318-chip for Ion Torrent sequencing using the Ion OneTouch™ 200 Template Kit v2 DL (Life Technologies).

Preparation of defined culture mix
Frozen stocks of Bifidobacterium longum ssp. infantis (NCIMB 702205), Lactobacillus rhamnosus (ATCC 53103) and Enterococcus faecalis (DSM 20478) were thawed and plated on Bifidus Selective Medium (BSM) agar plates for 2 days (B. longum), de Man, Rogosa and Sharpe (MRS) plates for 2 days (L. rhamnosus) or blood agar plates for 1 day (E. faecalis). Single colonies of these were inoculated in Gifu Anaerobic Medium (GAM) broth and incubated anaerobically overnight. Finally, the three cultures were mixed to obtain an equal optical density of each and added glycerol in saline to a final concentration around 15% (v/v) and frozen at -80°C in aliquots.

Real-time PCR conditions
The 16S rRNA-targeting primers used in this study are listed in Table S2. Total reaction volume of 11 µL containing 5.5 μL LightCycler® 480 SYBR Green I Master (Roche), 2.2 pmol of each of the primers (TAG Copenhagen, Denmark), 2 ng template DNA, and nuclease-free water purified for PCR (Qiagen). The reaction conditions were: Pre-incubation at 95°C for 5 min followed by 45 cycles of 95°C for 10 s, 60°C for 15 s and 72°C for 45 s. Lastly, a melting curve was generated (95°C for 5 s, 68°C for 1 min and increasing the temperature to 98°C with a rate of 0.11°C/s with continuous fluorescence detection). The qPCR was run in 384-well format on a LightCycler® 480 II (Roche Applied Science) and analysed using the LightCycler® 480 software.

Real-time PCR data handling
For each incubation replicate, the mean threshold cycle (Ct) value between qPCR triplicates was used to calculate the relative abundance of the genera Bifidobacterium, Lactobacillus and Enterococcus relative to the total bacteria (ntarget/ntotal) using 2 ΔCt as described elsewhere (4). ΔCt is the Ct value of the bacterial target normalised to the Ct value of the total bacterial population in the same incubation sample. Furthermore, the ratio between the relative abundance of a bacterial target in incubation with different whey products relative to iW (ntreated/niw) were calculated using 2 ΔΔCt , where ΔΔCt is the ΔCt value of a given sample normalised to the median ΔCt of three iW samples.   . Relative abundance of the most abundant bacterial genera in the small intestine of individual rats. Only those genera with a relative abundance of more than 0.05 in at least one rat are shown. The remaining genera are grouped into "Other".