All animal protocols were pre-approved by the Animal Protection Committee of Northwest A&F University, and the use of animals and the experimental procedure was consulted to the Guide for the Care and Use of Laboratory Animals of NIH.

Twelve newborn litters of New Zealand White were selected based on doe’s parity and health. Two healthy male kits with similar body weights were selected from each of these 12 litters at 5-day-old and then randomly distributed into the control group and treatment group. The body weight difference between the two kits in the same pair is <2.0 g at 5 days of age. All the selected kits were artificially fed with milk from the beginning to the end of the trial. All the kits were fed five times (8:00 am, 12:00 am, 4:00 pm, 8:00 pm, and 12:00 pm) every day with a commercial milk powder (without probiotic in the powder) for pet infant. The milk powder was dissolved in warm (37°C to 38°C) boiled water (milk powder:water = 1 g:4 mL) before feeding. The volume of the dissolved milk fed to each kit each time was 5 mL (5 d), 8 mL (6 d), 10 mL (7 to 8 d), 15 mL (9 to 11 d), 20 mL (12 to 13 d), and 25 mL (14 to 15 d). Kits in the treatment group had been continuously supplemented with Lactobacillus casei RABX1 (accession number: KT944253) from 5 days of age to 13 days of age. To prevent the interference from the supplemented Lactobacilli, which would affect the investigated percentage of intestinal Lactobacilli in total bacteria, kits in the treatment group fed milk without Lactobacillus casei RABX1 from 14 days of age. Before being added to the milk, the isolated bacterium was cultured in sterile Mann-Rogosa-Sharp broth at 37°C for 12 h and harvested by centrifugation at 5,000 g for 6 min and then washed with sterile phosphate buffered saline. The number of the harvested bacteria was estimated by a spectrophotometer. Bacteria were then resuspended in the milk to reach a concentration of 5–6×10⁸ colony-forming units/mL and orally administered to kits in the treatment group. All the kits were caged in boxes with constant temperature of 30°C (moisture 50% to 70%). The kits of the same group were caged in the same box with natural lighting. The padding towels were changed twice every day and sterilized in boiled water after being washed, and the boxes were cleaned three time every day (before feeding). All the kits were weighed at 5 days and 15 days of age before morning feeding, and their health condition was recorded every day, the unhealthy kits were separated from its group as soon as their abnormality were found.

Healthy pairs of kits were slaughtered eight hours after the morning feeding at 15 days of age. Before sampling, the length of the vermiform appendix was quickly measured from its end to the junction between the cecum and vermiform appendix. Then, the content in small intestine was collected and mixed together to determine the relative proportion of intestinal Lactobacilli or Escherichia-Shigella in total bacteria (Note: the content in kit’s duodenum, jejunum and ileum wasn’t collected separately because it was too less). Two parts each were harvested from the duodenum, jejunum and ileum. One part was stored at −80°C for determination of gene expression, and another part was fixed in 4% paraformaldehyde for investigation of the intestinal morphology and degranulation of paneth cell.

Total bacterial genomic DNA in the collected intestinal content was extracted by the modified phenol-chloroform-isoamylalcohol extraction method [18]. The concentration of the extracted DNA solution was subsequently determined and then diluted to 15 ng/μL. The relative proportion of Lactobacilli or Escherichia-Shigella in total bacteria was determined by relative quantification real-time polymerase chain reaction (PCR). The reaction mixture (20 μL) consisted of 10 μL of SYBR Premix Ex Taq (Takara Biotechnology [Dalian] Co., Ltd, Dalian, Liaoning, China), 0.4 μM of each primer and 30 ng of the extracted bacterial genomic DNA. The average cycle threshold (Ct) value for Lactobacilli, Escherichia-Shigella, or total bacteria in each sample was determined in triplicate. Then, the average Ct value of Lactobacilli (or Escherichia-Shigella) after being normalized to that of total bacteria was used for calculating the relative proportion of Lactobacilli (or Escherichia-Shigella) in total bacteria by using the 2^−ΔΔCt method as previously described [18]. Real-time PCR was performed on the IQ5 Cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The primers used for detecting 16S rDNA of Escherichia-Shigella, Lactobacilli and total bacteria are listed in Table 1.

The fixed intestine sections were rinsed with water and then dehydrated in a graded series of ethanol. After being cleared in benzene twice, sections were saturated with paraffin first and then embedded in paraffin. The embedded sections were cut into thin slices (5 μm thickness, 10 slices of each sample), which were stained with hematoxylin/eosin and prepared for observation by light microscopy. A total of 10 intact, well-oriented crypt-villus units were selected in triplicate for each intestinal cross-section. Villus height was measured from the tip of the villus to the villus-crypt junction. Crypt depth was defined as the depth of the invagination between two adjacent villi. The intestine morphological measurements were performed by using an image processing and analysis system (Optimus software version 6.5, Media Cybergenetics, North Reading, MA, USA). Quantitation of degranulated paneth cells was performed following the method described by Rumio et al [14]. The fixed intestinal sections were embedded in paraffin and then cut into 5-μm slices. After staining with hematoxylin and eosin, the slices were mounted in Entellan (Merck, Darmstadt, Germany) and then observed with a ViCo microscope (Biomedica Mangoni S.n.C., Nikon Instruments S.p.A, Pisa, Italy) equipped with a digital Nikon DS-L1 camera. Paneth cells were quantitated only in crypts that were sectioned through their center and could present a clear lumen. Degranulated paneth cells were distinguished from non-degranulated paneth cells based on the increased granule dimension and presence of large darkly colored vacuoles in the cytoplasm [14]. The percentage of degranulated paneth cells in each crypt was calculated, and twenty crypts were analyzed for each intestinal segment.

Total RNA of each intestine sample stored at −80°C was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA). The cDNA was synthesized from 1 μg of total RNA using a reverse transcriptase kit (Takara Biotechnology [Dalian] Co., Ltd, China). The relative mRNA levels of toll-like receptor 9 (TLR9), DEFEN, LYZ, tumor necrosis factor alpha (TNF-α), interferon beta (IFN-β), and interleukin 6 (IL-6) in the intestine were quantified by real-time quantitative PCR, which was carried out on the IQ5 Cycler using the TaKaRa SYBR Premix Ex Taq kit (China). The reaction system contained 1 μL of synthesized cDNA, 1 μL of each primer (4 μM), 10 μL of SYBR Premix Ex Taq (Takara Biotechnology [Dalian] Co., Ltd, China) and 7 μL of nuclease-free water. The cycle threshold (Ct) value for the investigated gene in each sample was determined in triplicate, and the average Ct value was calculated. Finally, the average Ct values for each gene, after being normalized to that of glyceraldehyde-3-phosphate dehydrogenase gene, were used for quantification by the 2^−ΔΔCt method. Real-time PCR was carried out with an initial denaturation step of 95°C for 5 min, followed by 40 cycles of 95°C for 15 s and 60°C for 45 s. The primers designed using Primer 5.0 or referenced are shown in Table 2.

The data collected from the experiment were analyzed with a paired-samples t test using SPSS19.0 soft package (SPSS Inc, Chicago, IL, USA), and the results were presented as mean± standard deviation for each trait. The probability of significance is indicated by the following conventional standard abbreviations: p>0.05 for non-significance and p<0.05, p<0.01, and p<0.001 for significance at these levels.

The relative proportion of the investigated bacterium in total bacteria was measured by real-time PCR. As presented in Table 3, although the oral administration of Lactobacillus casei had been stopped for more than 24 h, suckling rabbits in the treatment group had an extremely higher relative proportion of Lactobacilli in total intestinal bacteria (p<0.001) and an extremely lower relative proportion of Escherichia-Shigella (p< 0.001) than rabbits in the control group.

Oral administration with Lactobacillus casei did not change the intestinal morphological indices, including villus height, crypt depth and the ratio of villus height to crypt depth (p> 0.05), but it significantly increased the length of the vermiform appendix in suckling rabbits (p = 0.04) (Table 4). Degranulated paneth cell was showed in Supplementary Figure S1. As presented in Figure 1, the percentage of degranulated paneth cells in the duodenum and jejunum of suckling rabbits orally administered with Lactobacillus casei RABX1 was increased by 206% and 177% (p<0.001), respectively. Degranulation of paneth cells was also observed in ileum sections, but the difference in the percentage of degranulated paneth cells between the two groups was not obvious (19%±6.1% in the control group vs 25%±5.7% in the treatment group).

Expression of TLR9, DEFEN, and LYZ in the duodenum (p< 0.001, p = 0.05, and p = 0.008, respectively) and jejunum (p< 0.001, p<0.001, and p = 0.001, respectively) was significantly increased by orally administered Lactobacillus casei RABX1 (Figure 2), but that in the ileum was not affected (p>0.05). Oral administration with Lactobacillus casei did not change the expression level of TNF-α, IFN-β, and IL-6 in the duodenum, jejunum and ileum of artificially suckling rabbits (Figure 3).