SP600125 (a JNK inhibitor), SB203580 (a p38 MAPK inhibitor), U0126 (an ERK inhibitor), and BAY11-7082 (an NF-κB inhibitor) were obtained from MedChem Express (Monmouth Junction, NJ, USA). C12-iE-DAP was obtained from InvivoGen (San Diego, CA, USA).

Rabbit kidney cells (RK-13) were cultured and maintained in Dulbecco’s modified Eagle medium (Gibco, Grand Island, MI, USA) containing 10% fetal bovine serum (TransGen, Beijing, China) at 37°C in 5% (v/v) CO₂.

The bacterial pathogen EHEC was originally isolated from clinically infected rabbits suffering from acute diarrhea, and was stored at the Environmental Microbiology Laboratory at Shandong Agricultural University.

Healthy, weaned, 35-day-old New Zealand White rabbits were raised in the same environment with sufficient room, food, and ventilation for each rabbit.

Total RNA was extracted from the RK-13 cells using TransZol (TransGen) and first cDNA synthesis was carried out using HiScript^® II Q Select RT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China). To clone rNOD1, primers were designed based on the predicated gene sequence published in GenBank (Table 1). The amino acid sequence of rNOD1 was aligned with that of other species using Clustalx. The CARD, NACHT, and LRR regions of rNOD1 were analyzed using the sample modular architecture research tool (SMART). Lastly, phylogenetic analysis of rNOD1 was conducted using the neighbor-joining method of the MEGA5.1 program with 1,000 bootstrap replications to validate the branches.

Quantitative real-time PCR (qRT-PCR) was used to determine the expression of rNOD1 in various tissues. Five healthy rabbits were euthanized and the heart, liver, spleen, lungs, kidneys, trachea, thymus, esophagus, stomach, duodenum, jejunum, ileum, cecum, colon, rectum, pancreas, appendix, mesenteric lymph nodes, lymph follicles, sacculus rotundus, brain, cerebellum, brainstem, muscles, and skin were collected. Total RNA was isolated from each sample and cDNA was prepared as described above. The primers used to evaluate the expression of rNOD1 are shown in Table 1; glyceraldehyde-3-phosphate-dehydrogenase (Gapdh) was used as an internal control.

Rabbits were infected with 10⁸ CFU of E. coli bacterial suspension by intraperitoneal injections. At 1, 2, and 3 days postinfection (dpi), five rabbits of each group were euthanized and the liver, spleen, and kidney were collected for RNA extraction. All animal experiments were conducted in duplicate.

The following sequences were amplified using the primers shown in Table 1: the full length rNOD1, rNOD1-CARD, and rNOD1 without CARD (rNOD1-delCARD). All fragments were cloned into the pCDNA3.1 (+) vector using the Hieff Clone™ Multi One Step Cloning Kit (Yeasen, Shanghai, China) to make the constructs pC-rNOD1, pC-rNOD1-CARD, and pC-rNOD1-delCARD. pCDNA3.1-empty vector was used as the control. RK-13 cells were plated in six-well plates for 12 h prior to transfection. 2 µg of pC-rNOD1, pC-rNOD1-CARD, pC-rNOD1-delCARD, or pCDNA3.1-empty were transfected using TransIL-LT1 Transfection Reagent (MirusBio, Madison, WI, USA) for 24 h, after which the cells were harvested for RNA extraction.

1 µg of siRNA or negative control (NC) siRNA were transfected with TransIL-LT1 Transfection Reagent. The si-RNA sequences used were: si-rNOD1-1, sense 5′-GGAUCAUCUUUCGCUGCUUTT-3′, antisense 5′-AAGCAGCGAAAGAUGAUCCTT-3′; si-rNOD1-2, sense 5′-GCAUGUUCAGCUGCUUCAATT-3′, antisense 5′-UUGAAGCAGCUGAACAUGCTT-3′; si-rNOD1-3, sense 5′-CCAAGAGCCUGUUUGUCUUTT-3′, antisense 5′-AAGACAAACAGGCUCUUGGTT-3′; NC, sense 5′-UUCUCCGAACGUGUCACGUTT-3′, antisense 5′-ACGUGACACGUUCGGAGAATT-3′. siRNA sequences were synthesized by GenePharma (Shanghai, China).

RK-13 cells were plated in 24-well plates for 12 h prior to transfection. The luciferase reporter plasmids pGL3-NF-κB, pGL3-IFN-β, and pGL3-ISRE were purchased from Agilent (Santa Clara, CA, USA). The pRL-TK plasmid (Promega, Madison, WI, USA) acted as an internal control to normalize transfection efficiency. Cells were transfected with reporter plasmid (100 ng/well) or pRL-TK plasmid (50 ng/well) and 1 µg of pC-rNOD1, pC-rNOD1-CARD, pC-rNOD1-delCARD, or pCDNA3.1-empty or with 500 ng of si-rNOD1 or NC siRNA for 24 h using TransIL-LT1 Transfection Reagent. After 24 h of cotransfection with NF-κB, the cells were infected with 1 × 10⁷ E. coli for 2 h. The medium was then removed, and the cells were cultured in DMEM containing gentamicin (100 µg/mL) for 3 h. After lysing and harvesting cells, luciferase activities were detected with a dual-luciferase reporter assay system (Beyotime, Wuhan, China).

RK-13 cells were plated in 24-well plates for 12 h prior to transfection. 1 µg of pC-rNOD1 or pCDNA3.1-empty or 500 ng of si-rNOD1 or NC siRNA were transfected into cells for 24 h using TransIL-LT1 Transfection Reagent. The cells were then infected with 1 × 10⁷ E. coli for 2 h, the medium was removed, and cells were washed three times with PBS containing gentamicin (100 µg/mL) to eliminate extracellular E. coli. The cells were then cultured in DMEM containing gentamicin for 3 h. RK-13 cells were lysed in 500 µL PBS containing 1% (v/v) Triton X-100 for 20 min and were then plated onto nutrient agar to calculate the intracellular bacterial CFU.

Total RNA was extracted from the liver, spleen, and kidney of rabbits, as well as from RK-13 cells, using the RNeasy Plus Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Total RNA (1 µg) was reverse transcribed using TransScriptR One-step gDNA Removal and cDNA Synthesis SuperMix for qPCR (Transgen Biotech Co., Ltd., Beijing, China). The synthesized cDNA was stored at −20°C until further use. qRT-PCR was performed using TransStart^® Tip Green qPCR SuperMix (+Dye II) (Transgen Biotech). qRT-PCR primers were designed using the Primer 3 software (http://bioinfo.ut.ee/primer3-0.4.0/) based on published target sequences (Table 1). The primers Il6 and Tnf used for qRT-PCR have been previously reported (39). qRT-PCR was carried out using a 7500 Fast Real-Time PCR system (Applied Biosystems, Carlsbad, CA, USA). The PCR conditions were as follows: 1 cycle at 94°C for 30 s, 40 cycles at 94°C for 5 s, and 60°C for 34 s. Dissociation curves analysis was performed as the final step of the PCR.

RK-13 cells were seeded on sterile glass cover slips placed in 24-well plates and cotransfected with pC-rNOD1 and GFP-LC3 plasmids for 24 h. The cells were then infected with 1 × 10⁷ E. coli for 2 h, the medium was removed, and cells were washed three times with PBS containing gentamicin (100 µg/mL) to eliminate extracellular E. coli. Cells were then cultured in DMEM containing gentamicin for 3 h, washed with PBS, fixed with 4% paraformaldehyde for 8 min, and treated with 0.1% Triton X-100 for 10 min. Cells were blocked in PBS containing 5% bovine serum albumin for 30 min, incubated with primary antibodies for 1 h at 37°C, and then incubated for 1 h with secondary antibody of Cy3-labeled goat antimouse IgG. Images were captured using a Leica TCS SPE confocal microscope with a 63× (1.3 numerical aperture) oil immersion objective.

Total protein lysates were obtained by lysing the cells with ice-cold RIPA buffer supplemented with a protease inhibitor cocktail (Beyotime). Protein was quantified using a BCA protein assay kit (Tiangen, Beijing, China), samples were run on a SDS-PAGE and transferred to PVDF membranes. After blocking with 5% skim milk for 1 h at room temperature, membranes were incubated overnight at 4°C with the primary antibodies, and then incubated with appropriate secondary antibodies for 2 h at room temperature. Images of protein blots were obtained with a ChemiDoc XRS (Bio-Rad, Marnes-la-Coquette, France) using a Western ECL Substrate kit. The density of each band was normalized to that of β-actin and quantified it using Quantity One software (Bio-Rad).

The relative expression of each gene was calculated using the 2^−ΔΔCt method. The housekeeping gene Gapdh was used as an endogenous control to normalize the expression of target genes. Each treatment was conducted in triplicate and each in vitro assay was performed in triplicate. The data were analyzed using the non-parametric Mann–Whitney U test. Statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software Inc., San Diego, CA, USA). Statistical significance was set at P < 0.05.

The complete open reading frame of rNOD1 was obtained and the sequence was submitted to GenBank (MF069503). The sequence contained 2,862 bp encoding 953 amino acids. Secondary structures were predicated using SMART, the results of which indicated that rNOD1 contained three characteristic domains: an N-terminal CARD (aa 20–104), a central NACHT domain (aa 196–368), and C-terminal LRR domains (aa 727–754, 755–782, 783–810, 811–838, 839–866, 867–894, and 895–922), all shown in Figures 1A,B.

The deduced amino acid sequence of rNOD1 was 81.7, 79.2, and 80.7% identical to those of Homo sapiens, Mus musculus, and Sus scrofa, respectively (Table 2). The rNOD1 NACHT domain was closest in sequence identity to that of other species (Table 2). A phylogenetic tree was constructed with the full length NOD1 protein and three major branches were observed (Figure 1C). Among the mammalian species, Rattus norvegicus showed the closest evolutionary relationship with rNOD1.

Quantitative real-time PCR was performed to analyze the expression levels of rNOD1 mRNA in tissues of healthy rabbits. Although rNOD1 was detected in all tested tissues, higher levels of expression were detected in lung, kidney, duodenum, jejunum, cecum, liver, spleen, ileum, rectum, pancreas, appendix, colon, heart, mesenteric lymph nodes, lymph follicles, esophagus, stomach, sacculus rotundus, trachea, and thymus than in cerebellum, brainstem, muscle, skin, and brain (Figure 2A).

To evaluate the response of rNOD1 expression to bacterial infection, we infected rabbits with E. coli and examined rNOD1 mRNA transcripts in liver, spleen, and kidney at 1, 2, and 3 dpi using qRT-PCR. As shown in Figure 2B, although the expression of rNOD1 displayed no significant difference at 1 dpi, it became significantly upregulated at 2 dpi by 3.69-fold (P < 0.05) and 3 dpi by 7.68-fold (P < 0.05) in the liver, whereas the expression of rNOD1 displayed significantly upregulated at 3 dpi by 8.01-fold (P < 0.05) in the spleen (Figure 2C). Significant upregulation of rNOD1 mRNA expression was detected in kidney at 1 dpi (4.88-fold) and 2 dpi (5.98-fold) (Figure 2D).

To evaluate the role of rNOD1 in NF-κB signaling and to clarify which domain of rNOD1 mediates this role, mutant plasmids were constructed (Figure 3A). In addition, three siRNAs (si-rNOD1-1, si-rNOD1-2, and si-rNOD1-3) were designed to target different regions of rNOD1. The knockdown efficiency of si-rNOD1-1 was 54.57% (Figure 3B).

As shown in Figure 3C, overexpression of rNOD1 significantly increased NF-κB activity by 33.54-fold (P < 0.05) compared to that induced by the empty vector. rNOD1-CARD and del-CARD mutants were unable to significantly increase NF-κB activity, even after 24 h (Figure 3C). Addition of rNOD1-specific siRNA reduced NF-κB activity (0.38-fold, P < 0.05), and rNOD1 increased IFN-β and ISRE activity by 2.80- and 4.07-fold, respectively. By contrast, addition of si-rNOD1 did not affect basal IFN-β or ISRE activity (Figures 3D,E). Finally, to evaluate the effect of rNOD1 on E. coli-induced NF-κB activity, RK-13 cells were infected with E. coli after transfected with rNOD1. rNOD1 significantly increased NF-κB activity by 3.60-fold (P < 0.05) when induced by E. coli (Figure 3F).

To investigate the induction of innate immune response by rNOD1 effector domains in RK-13 cells, the expression of cytokines and defensins was analyzed via qRT-PCR. The expressions of proinflammatory cytokines Il1b, Il6, Il8, Ifn-γ, and Tnf became significantly up-regulated with the overexpression of rNOD1 effector domains (Figures 4A,C,D,F,G). A 258.71-fold (P < 0.05) increase in Il8 expression was detected after transfection with rNOD1 (Figure 4D). The anti-inflammatory cytokine Il4 became significantly upregulated by only 2.67-fold (P < 0.05; Figure 4B) with the overexpression of rNOD1, whereas Il10 became upregulated by 3.67- and 6.20-fold (P < 0.05; Figure 4E) with overexpression of rNOD1 and del-CARD domains, respectively. Overexpression of full length rNOD1 significantly increased the expression of α-defensin (Np5) and β-defensins (Defb124, Defb125, Defb128, and Defb135) compared to overexpression of the empty vector (Figures 4H–L). Induction of α- and β-defensins by rNOD1-CARD and del-CARD mutants was significantly less than that of full length rNOD1. In particular, rNOD1 overexpression significantly increased expression of Defb125 by 6.39-fold; by contrast, overexpression of rNOD1-CARD and del-CARD mutants increased expression of Defb125 by 2.50-fold (P < 0.05) and 1.98-fold (P > 0.05), respectively.

To investigate the ability of rNOD1 to regulate an antimicrobial response to E. coli, RK-13 cells were infected with E. coli after transfection with pC-rNOD1 or empty vector and si-rNOD1 or NC siRNA. As shown in Figure 5A, RK-13 cells transfected with rNOD1 contained significantly less E. coli CFUs than those transfected with the empty vector. By contrast, cells transfected with si-rNOD1 contained significantly more E. coli CFUs than those transfected with NC siRNA (Figure 5B). To further investigate which pathway was activated by rNOD1, cells were pretreated with inhibitors of NF-κB activity and three major components of the MAPK pathway. Inhibition of the NF-κB pathway, but not the MAPK pathway eliminated the ability for rNOD1 to inhibit E. coli growth (Figure 5C). As shown in Figure 5D, the number of E. coli was inhibited after stimulated with iE-DAP.

The effect of rNOD1 overexpression or knockdown on E. coli-induced cytokine and defensin production was evaluated. Overexpression of rNOD1 increased E. coli-induced proinflammatory cytokines Il1b, Il6, Il8, Ifn-γ, and Tnf (Figures 6A,C,D,F,G). Knockdown of rNOD1 significantly reduced E. coli-induced proinflammatory cytokine expression, but significantly increased the expression of anti-inflammatory cytokines Il4 by 2.32-fold (P < 0.05; Figure 6B) and Il10 by 2.96-fold (P < 0.05; Figure 6E). Similarly, overexpression of rNOD1 increased E. coli-induced Defb124, Defb 125, and Defb128 expression, whereas knockdown of rNOD1 impaired the expression of these defensins (Figures 6H–J). No significant changes were detected in the expression of Defb135 and Np5 (Figures 6K,L).

After inhibition of the NF-κB pathway, overexpression of rNOD1 decreased the expression of proinflammatory cytokines and defensins induced by E. coli. In particular, the expression of Il1b and Il8 were significantly decreased to 0.47- and 0.65-fold, respectively (P < 0.05). Although the expression of all defensins was reduced, no changes were statistically significant. By contrast, the expression of anti-inflammatory cytokines Il4 and Il10 was increased (Figure 7).

As shown in Figure 8, stimulation with iE-DAP significantly induced the expression of immune-related genes in RK-13 cells after E. coli infection. Especially, the expression of Il6, Il8, and Tnf were significantly upregulated by 81.20-fold, 940.00-fold, and 93.80-fold, respectively (P < 0.05). Similarly, expression of defensins Defb125 (29.80-fold), Defb135 (19.18-fold), and Np5 (69.69-fold) was significantly increased.

To investigate the ability for rNOD1 to induce autophagy, the colocalization of LC3 with rNOD1 was first evaluated. LC3 and rNOD1 colocalized in RK-13 cells (Figures 9A,B), and E. coli significantly induced autophagosome formation (Figure 9B), as similar phenomena shown in Figures 9C,D. Overexpression of rNOD1 significantly increased the protein levels of LC3-II in RK-13 cells infected with E. coli (Figure 9C). A slight decrease in the protein level of LC3-II was measured after knockdown of rNOD1 (Figure 9D).