In order to obtain the CDs of duCIITA, one set of specific polymerase chain reaction (PCR) primers was designed to identify the duCIITA sequences based on the predicated genes of chicken CIITA in the Genbank (Table 1). Total RNA (1 μg) was extracted from the spleen of healthy Cherry Valley ducks with the TRIzol reagent (Takara, Dalian, China) and the cDNA was obtained with the HiScript R II One-Step RT-PCR kit (Vazyme, Nanjing, China). The PCR conditions were pre-denatured 1 cycle of 94°C for 5 min; 35 cycles of 94°C for 30 s, 57°C for 30 s and 72°C for 4 min; and the final extension at 72°C for 10 min. Eventually, the full-length cDNA of duCIITA was sequenced by the Shanghai Invitrogen Biotechnology Co., Ltd.

The homology analysis of the duCIITA cDNA sequence was performed using the BLAST program of NCBI. The amino acid sequence of each species’ CIITAs was found on NCBI and the protein number is shown in Table 2. Meg Align software was used to analyze the similarity and evolution of the CIITAs’ amino acid sequence. The structure of the amino acid sequences of duCIITA was predicted by the SMART online tool¹. Multiple amino acid sequence alignments were performed using ClustalW2² and edited with the online tool Boxshade³. The phylogenetic analysis was generated using the MEGA5.0 software and the tree was constructed by the neighbor-joining method with the bootstrapping over 1000 replicates.

The bacterial strain of APEC O1:K1 used in this study was stored by the Environmental Microbiology Laboratory in Shandong Agricultural University and was isolated from the clinical infected ducks that suffered from the colibacillosis. The bacterial strain was cultivated in Luria-Bertani (LB) agar at 37°C for 18 h. A single colony was inoculated in 5 mL of LB broth and cultivated at 37°C for 18 h with agitation.

Duck embryo fibroblast cells were derived from the 12-day-old Cherry Valley duck embryo and cultured in Dulbecco’s modified Eagle medium (DMEM) (Gibco, Grand Island, NY, United States), which included 10% fetal bovine serum (Transgen, Beijing, China), and was incubated at 37°C in an atmosphere with 5% (v/v) CO₂.

One-day-old healthy Cherry Valley ducks were purchased from a farm in Tai’an, and housed in isolation for the experiment until 4 weeks old. The ducks were tested for several common diseases, such as Newcastle disease, avian influenza, duck Tembusu virus, and the duck plague virus, and were declared to be healthy. In addition, all the ducks used in this study were confirmed to be negative for the APEC O1K1.

All ducks used in this experiment were handled according to the guidelines of the Ethics Committee on Animal Experiments of Shandong Agricultural University and the appropriate biosecurity guidelines, the number of the approval protocol being “SDAUA-2015-004.” At 4 weeks old, 3 healthy ducks were randomly selected and their tissues, including the heart, liver, spleen, lung, kidney, brain, duodenum, jejunum, ileum, cecum, rectum, cerebellum, thymus, bursa of Fabricius, muscular stomach, glandular stomach, trachea, esophagus, skin, and muscle were collected for the duCIITA tissue distribution analyses by quantitative real-time PCR (qRT-PCR).

To confirm whether the duCIITA was involved in the innate immune response caused by the APEC O1K1, 40 four-week-old Cherry Valley ducks were randomly divided into two groups. The experimental group was infected neck subcutaneously with 0.3 mL APEC O1K1 (10⁹ CFU/mL) per duck, and the control group was inoculated with the same volume of 0.65% normal saline (Dozois et al., 2003). At 1, 2, and 3 days post-infection (dpi), three ducks with the significant clinical symptoms were killed, and the same numbers of ducks were also killed in the control group. A part of samples from the liver, spleen, and brain were treated with 10-fold serial dilution, then the diluent were plated onto eosin methylene blue nutrient agar and incubated at 37°C for 12 h to calculate intracellular bacterial CFU; a second part were fixed with 4% formalin solution at room temperature for histopathological examination; and another part of the samples were immediately preserved in liquid nitrogen for duCIITA and inflammatory cytokines detection by qRT-PCR. The clinical symptoms of the remaining ducks in the infected group were observed to 14 dpi before they were euthanized.

Three Si-CIITAs (pSi-CIITA-1, pSi-CIITA-2 and pSi-CIITA-3) sequences targeting the 2095–2116, 2217–2238, and 1371–1392 bp positions of duCIITA and negative control Si-RNA (pSi-NC) were synthesized by GenePharma (GenePharma, Shanghai, China). 1 μg pSi-CIITA or pSi-NC was transfected with TransIL-LT1 Transfection Reagent to DEFs (Mirus Bio, CA, United States). After 24 h post-transfected (hpt), cell samples were directly collected for RNA extraction. PSi-NC acted as the control and the silencing efficiency of the Si-CIITAs was analyzed by qRT-PCR. The Si-RNA sequences are seen in Table 3.

To investigate the antibacterial ability of duCIITA, DEFs were incubated in 6-well plates at 37°C. Before pSi-CIITA and pSi-NC transfection, the cells were growing with 80% confluent. After 24 hpt, DEFs were infected with 1 × 10⁶ CFU/mL APEC O1K1 for 3 h, and then washed three times with PBS. The cells were then cultured with DMEM containing gentamicin (100 μg/mL) for 3 h to kill the extracellular APEC O1K1. The cell samples were collected with 500 μL 1% (v/v) Triton PBS and treated with 10-fold serial dilution. Then the processing of diluent was as above.

Total RNA of the above tissues and DEFs was extracted and reversed with the above method. QRT-PCR primers of duCIITA were designed based on the sequences of duCIITA obtained in this study with the primer3 software⁴ and selected by the dissociation curves. The qRT-PCR primers sequences of inflammatory cytokines IL-1β, IL-6, IL-8, and IL-10 are seen in Table 1. The duCIITA and ILs expression were normalized by the endogenous gene of β-actin (Table 1). QRT-PCR was performed using the ChamQ^TM SYBR^® qPCR Master Mix (Vazyme, Nanjing, China) and carried out with the 7500 Fast Real-Time PCR System (Applied Bio-systems, Foster City, CA, United States). The PCR consisted of 20 μL volume, with conditions as follows: one cycle of pre-denatured at 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 10 s and extension at 60°C for 34 s. Then, a dissociation curve was analyzed. Each sample was analyzed in triplicate.

The relative expression levels of the duCIITA were detected using the duck β-actin gene as the endogenous reference gene and were calculated using the 2^-ΔΔC_T method (Livak and Schmittgen, 2001). All data were represented as means ± standard deviations, and statistical analyses were performed by the Graph Pad Prism 5 software (Graph Pad Software Inc., San Diego, CA, United States). The differences were evaluated using Student’s t-test. A value of P < 0.05 was considered as significant and P < 0.01 was considered to be highly significant.

The CDs of duCIITA consists of a 3648 bp and encodes 1215 amino acids. The duCIITA cDNA sequence showed a highly similar to the Gallus gallus CIITA. And the duCIITA contained characteristic domains of NLRs: a central NACHT domain and an LRR domain at its C-terminus (Figure 1). Multiple alignment analysis showed that duCIITA had seven MHC II transactivator domains in 564–582, 594–615, 618–638, 724–750, 790–802, 809–826, and 902–917 residues (Figure 1). The phylogenetic tree indicated that there were four major branches: mammals, reptiles, birds, and fishes (Figure 2A). The duCIITA and Falco cherrug CIITA were in the same small branch, which has the closest relationship. However, the genetic relationship between duCIITA has the most distant relationship with fishes (Figure 2A). DuCIITA had the highest identity to the Anas platyrhynchos (94.9%), 78.6% identity with G. gallus and Meleagris gallopavo (Figure 2B). But the duCIITA amino acid sequence showed a relatively lower identity with mammals CIITAs, only 29.5–40.6% (Figure 2B). These results showed that the duCIITA amino acid sequence was closely related to the CIITAs of other birds.

The ducks infected with APEC O1K1 exhibited the typical clinical symptoms, including listlessness, anorexia (Figure 3A), and diarrhea (Figure 3B). At necropsy, the heart and liver surface had a layer of yellowish–white fibrinous exudate attachment (Figures 3C,D), splenomegaly and hemorrhage (Figure 3E), and peritoneal fibrosis adhesions (Figure 3F). One duck died at 4 dpi in the infected group, and the condition of the remaining ducks was improved. Finally, they were euthanized at 14 dpi. No clinical symptoms or necropsy lesions were observed in the control group throughout the experiment. In addition, the E. coli rapid replication in the liver, spleen, and brain at 1 dpi, and reached its peak at 2 dpi with 10⁸ CFU/g. However, the bacteria content of all the tested tissues dramatically declined at 3 dpi in our previous study (Li et al., 2016).

As shown in Figure 4A, there was an obvious fibrinous exudate in the epicardial surface and a large amount of necrotic heterophilic granulocyte and histiocytic cells were revealed in the caseous exudate. In the liver, the infected ducks displayed a severe diffuse vacuolated degeneration, a serious hepatocellular necrosis, and lymphocyte infiltration (Figure 4B). In the spleen, the tissue were loose, lymphocytic necrosis, and heterophilic granulocyte infiltration was discovered (Figure 4C). No obvious microscopic lesions were observed in the brains of the infected ducks, except for the perivascular inflammatory infiltrates (Figure 4D). In addition, no obvious microscopic lesions were discovered in the corresponding tissues from the control group (Figures 4E–H).

To investigate the expression profile of duCIITA mRNA in normal tissue, 20 kinds of tissues were separated from 3 four-week-old healthy ducks. The muscular stomach was chosen as the standard tissue, and the duCIITA mRNA was expressed in all the tested tissues (Figure 5). It was strongly expressed in the lung and cerebellum, and weakly expressed in the rectum and esophagus (Figure 5). The wide expression of duCIITA indicates that the duCIITA might be extensively involved in the host immune response of healthy Cherry Valley ducks.

To confirm whether duCIITA was involved in the host antibacterial responses, we investigated the mRNA expression of duCIITA in the liver, spleen, and brain during APEC O1K1 infection. In the liver, the mRNA expression of duCIITA was highly significantly increased in a short time, and increased to 83.4- and 277.9-fold at 1 and 2 dpi, respectively (P < 0.01). Then, it reached a peak at 3 dpi (445.5-fold, P < 0.01) (Figure 6A). In the spleen, the expression level of duCIITA was significantly down-regulated at 1 and 3 dpi (P < 0.01) and increased a little at 2 dpi (1.6-fold, P > 0.05) (Figure 6B). In the brain, the expression of duCIITA was significantly up-regulated at 1 and 2 dpi (5.9- and 4.2-fold, respectively, P < 0.05), and down-regulated at 3 dpi (P < 0.05) (Figure 6C). These results suggest that duCIITA may be involved in the host innate immunity to APEC O1K1 infection.

To identify the inflammatory cytokines could involve in the antibacterial innate immune response of the APEC O1K1 infected ducks, the mRNA expressions of IL-1β, IL-6, IL-8, and IL-10 in the liver, spleen, and brain were measured by the qRT-PCR method. As shown in Figure 7A, the proinflammatory cytokines IL-1β mRNAs expression levels were all up-regulated in the liver, spleen, and brain during the three tested days. Among them, the highest up-regulated level was in the spleen with 151.1-fold at 2 dpi (P < 0.01) (Figure 7A), and the expression of IL-1β also peaked with 50.4-fold at 2 dpi in the liver (P < 0.01) (Figure 7A). In the brain, the mRNA expression of IL-1β had a sustained up-regulation at 1–3 dpi (Figure 7A). The proinflammatory cytokine IL-6 mRNA expression level also showed up-regulated at most of the time in the tested tissues. The greatest increase of IL-6 was in the spleen at 2 dpi (191.1-fold, P < 0.01) (Figure 7B); in the liver, similar to the spleen, the mRNA expression trend of IL-6 reached its peak at 1 dpi with 8.9-fold (P < 0.05) (Figure 7B), and then the up-regulated times gradually decreased (Figure 7B); however, the mRNA expression of IL-6 in the brain always remained at a background level at 1–3 dpi (Figure 7B). Unlike IL-6, the mRNA expression of proinflammatory cytokine IL-8 had a stable and significant up-regulation in the brain (Figure 7C), and in the spleen, was found to be significantly up-regulated at 1 dpi with 19.6-fold (P < 0.01) (Figure 7C). As one of the major suppressible inflammatory cytokines, the IL-10 mRNA expression also had a different degree of increase in the liver, spleen, and brain; among them, the highest up-regulation was expressed in the liver with 12.1-fold at 1 dpi (P < 0.05) (Figure 7D).

To further explore the antibacterial function of duCIITA, DEFs were transfected with pSi-CIITA or pSi-NC for 24 h. The cells were then collected for silencing efficiency analysis. As shown in Figure 8A, both pSi-CIITA-2 and pSi-CIITA-3 were able to decrease the mRNA expression of duCIITA in DEFs, but the pSi-CIITA-3 had stronger interference ability (Figure 8A), so we chose the pSi-CIITA-3 for further study. DEFs were infected with APEC O1K1 after transfection, and then processed as above. The number of E. coli in DEFs transfected with pSi-CIITA was significantly higher than with the pSi-NC group (P < 0.05) (Figure 8B). In addition, after 24 hpt, cells were infected with 1 × 10⁶ CFU/mL E. coli and collected for inflammatory cytokine detection. As shown in Figure 8C, the expression of inflammatory cytokines IL-6, IL-8, and IL-10 in pSi-CIITA group was significantly lower than that in the pSi-NC group; however, there was no significant change in the IL-1β mRNA expression. These results suggest that the knockdown of duCIITA could weaken the antibacterial activity and reduce the inflammatory cytokine production of DEFs.