Synthetic CpG-ODN rapidly enriches immune compartments in neonatal chicks to induce protective immunity against bacterial infections

Oligodeoxynucleotides containing CpG motifs (CpG-ODN) induce innate immunity against bacterial infections. Despite recent advances, how CpG-ODN alone protects against bacterial infections remained elusive. Here, we report for the first time, to our knowledge, that CpG-ODN orchestrates anti-microbial protective immunity by inducing a rapid enrichment of various immune compartments in chickens. In this study, eighteen-day-old embryonated eggs were injected with either 50 µg of CpG-ODN or saline (~n = 90 per group). In the first experiment, four days after CpG-ODN treatment, chicks were challenged subcutaneously with a virulent strain of Escherichia coli (E. coli) and mortality was monitored for 8 days. We found significant protection, and reduced clinical scores in CpG-ODN treated chicks. To gain insights into mechanisms of protection induced by CpG-ODN, first we investigated cytokine expression kinetics elicited by CpG-ODN. The spleen and lung were collected from embryos or chicks (n = 3–4 per group) at 10 time points post-CpG-ODN inoculation. Multiplex gene analysis (interleukin (IL)-1, IL-4, IL-6, IL-10, IL-18, interferon (IFN)-γ, IFN-α, and lipopolysaccharide induced tumor necrosis factor (LITAF), revealed a significantly higher expression of pro-inflammatory cytokines following CpG-ODN treatment compared to the saline controls. In our study, LITAF stands out in the cytokine profiles of spleen and lungs, underscoring its role in CpG-ODN-induced protection. The third experiment was designed to examine the effects of CpG-ODN on immune cell populations in spleen, lungs, and thymus. Flow cytometry analysis was conducted at 24, 48 and 72 hrs (thymus only collected at 72 hr) after CpG-ODN administration to examine the changes in CD4+ and CD8+ T-cell subsets, monocyte/macrophage cell populations and their expression of maturation markers (CD40 and CD86). Flow cytometry data indicated a significant enrichment of macrophages, CD4+ and CD8+ T-cell subsets in both spleen and lungs of CpG-ODN treated embryos and chicks. Macrophages in spleen and lungs showed an upregulation of CD40 but not CD86, whereas thymocytes revealed significantly high CD4 and CD8 expression. Overall, the present study has demonstrated that CpG-ODN provides protection in neonatal chicks against E. coli infection not only by eliciting cytokine responses and stimulating immune cells but also through enriching immunological niches in spleen and lungs.


Results
E. coli challenge. Survival following E. coli challenge was significantly higher in groups of birds that received 50 µg of CpG-ODN compared to the saline control group (P = 0.04) (Fig. 1a). The cumulative clinical scores (CCS) of chicks following E. coli challenge with either 1 × 10 4 or 1 × 10 5 cfu of E. coli showed that the chicks that received 50 µg CpG-ODN had a significantly lower CCS compared to the saline group (P = 0.03) (Fig. 1b). Samples collected from the challenged chicks revealed heavy bacterial growth more frequently in samples from chicks in the control group than in the CpG-ODN group (Fig. 1c). Moreover, the reduction in the relative risk of mortality following E. coli challenge was 46.4% compared to the saline control group. The data of groups that received 1 × 10 4 and 1 × 10 5 cfu of E. coli were combined for clarity of presentation and analysis.
Gene expression analysis. The mechanism of action of CpG-ODN in these chicks was also elucidated by measuring gene expression patterns of pro-inflammatory (IL-1β, IL-6, LITAF, IL-18), Th1 type (IFN-γ, IFN-α) and Th2 type (IL-4, IL-10) cytokines in spleen and lungs at several hours post CpG-ODN treatment. Fold changes of mRNA levels were 1 or >1 at almost all time points in birds of CpG-ODN treated experimental group compared to the saline control group. Overtime, levels of gene expression were not consistent nor followed a specific pattern. However, most significant differences in gene expression occurred in the 32 hr to 48 hr period after treatment. Normalized gene expression of pro-inflammatory cytokines (IL-18, IL-1β and IL-6 and LITAF) tended to show a relatively higher level in CpG-ODN treated group compared to the saline treated group. The expression of IL-1β and IL-6 in the lungs of chickens was many fold higher compared to spleen at 6 through 48 hr and 12 through 48 hr, respectively. The expression of IL-18 and LITAF mainly increased between 12 through 72 hr and 32 through 72 hr, respectively in both organs (Fig. 2). The gene expression levels of Th1 type cytokines (IFN-γ, IFN-α) and Th2 type cytokines (IL-4, IL-10) were higher in lungs compared to spleens in CpG-ODN treated birds (Fig. 3).

Flow cytometry analysis. Flow cytometric analysis of spleen and lung cells isolated from CpG-ODN
treated birds showed a marked influence of CpG-ODN on antigen presenting cells (APCs) (monocyte/macrophage), expression of costimulatory markers on APCs and T lymphocyte populations.
Antigen presenting cells. Since the overall cell number was low at 24 hr post treatment, when the chicken embryos were 19 days old, we could only obtain and analyse spleen cells at this time point. CpG-ODN treatment induced a notable increase in the expression of costimulatory molecules (CD40) and the number of APCs (monocyte/macrophage cells) in spleen and lungs compared to saline treated embryos. Compared to number of APCs detected in saline control birds, the percentage of APCs in CpG-ODN treated embryos in spleen at 24, 48 and 72 hrs post treatment were 5.88, 12.6 and 11.2%, respectively ( Fig. 4a-d). A considerably higher expression of CD40 costimulatory molecule was observed in APCs in CPG-ODN treated embryos compared to saline treated embryos as indicated by a shift in the MFI to the right (Fig. 4) of CD86 in CpG-ODN group compared to the saline group, no significant difference was observed at any of the time points.
A similar trend was observed in the percentage of APCs in the lungs of CpG-ODN treated embryos compared to saline treated embryos at 48 and 72 hrs post in-ovo injections. The percentage of APC in the lungs of CpG-ODN treated embryos at 48 and 72 hrs post treatment were 1.73 and 1.74% whereas the percentage of APCs in saline treated birds at 48 and 72 hrs were 0.03 and 0.17% respectively ( Fig. 5a and b). CD40 expression was significantly higher in APC in lungs of CpG-ODN treated embryos compared to saline treated embryos and as indicated by a shift in the MFI to the right. Similar to the spleen, no increase in the expression of CD86 costimulatory molecule was observed in APCs in lungs. This striking enrichment of immune cells in the CpG-ODN treated group was noticed at all three time points (24, 48 and 72 hrs post CpG-ODN inoculation) in both spleen and lung. Groups of birds that received 50 µg CpG-ODN CpG-ODN had a significantly lower CCS compared to the saline group, (P = 0.03). Bar = mean. (c) Bacterial scores of swabs taken from airsacs of birds. Higher bacterial scores indicate greater bacterial load. Increased bacterial growth was observed more frequently in lesions from birds in the saline group than in the CpG-ODN group. The substantial number of '0' or 'no growth' of bacteria in both groups is a common finding in such experiment. CD8 + and CD4 + T cells. CD8 + and CD4 + T cell populations were analysed in spleen and lung at 72 hrs post CpG-ODN injections (21 day old embryos). Both CD8 + and CD4 + cells were markedly increased in spleen and lung of CpG-ODN treated birds compared to the saline control. The percentage of CD4 and CD8 T cells in spleen of CpG-ODN treated embryos were 1.89% and 1.09%, respectively compared to 0.22% CD4 + T cells and 0.23% CD8 + T cells in saline control (Fig. 6a). Similarly, the percentage of CD4 + and CD8 + T cells in lungs of CpG treated embryos were 0.67% and 6.01%, respectively compared to 0.43% CD4 + T cells and 0.79% CD8 + T cells in saline control (Fig. 6b). Interestingly, while the percentage of CD4 + T cells was significantly higher in CpG-ODN group compared to saline in the spleen (Fig. 6c), the percentage of CD8 + T cell increase was more prominent in lungs than CD4 + T cells (Fig. 6d). Moreover, to gain insight whether increase in the number of CD4 + or CD8 + is due to increased differentiation of CD4 + or CD8 + T cells in thymus, we analysed thymus cells. We observed that the T cell population in thymus was remarkable, where 99% of the total thymus cells were CD4 + /CD8 + double positive at the time of hatch in CpG-ODN treated embryos. In contrast, only 85% the total thymus cells were CD4 + /CD8 + double positive at the time of hatch in saline treated embryos. Both CD4 + and CD8 + expression levels were significantly higher in CpG-ODN treated group than the saline control group (Fig. 7).

Discussion
There is strong evidence to suggest that CpG-ODN administration modulates innate and adaptive immune responses in chickens enhancing resistance to several bacterial pathogens including E. coli 12,24,25 , Salmonella Typhimurium 14,28 , Salmonella Enteritidis 26,27 , and Campylobacter jejuni 37 . The immunoprotective effects of CpG-ODN have also been reported in mice against Helicobacter pylori 38 , Yersinia pestis 39 , influenza virus 40 , and Listeria monocytogenes 41 . These findings suggest that CpG-ODN induces protection in several species against a wide range of pathogens. However, how CpG-ODN treatment confers protection remains poorly understood. The present study was undertaken to gain a better understanding of the mechanisms of CpG-ODN induced host responses that enhance resistance in neonatal chickens to a bacterial pathogen. In this study, we examined how CpG-ODN, delivered by in ovo route to18-days old embryonated eggs, protects neonatal chicks, 1-day after the hatch (four days after the CpG-ODN treatment), against a virulent strain of E. coli. Consistent with our previous studies 12,24,25,42 in the present study, we found that CpG-ODN treated chicks were significantly protected against E. coli infection. During a pathogenic insult, the host immune response is orchestrated by a complex network of cytokines that shape the immune process by regulating cellular communication and function. CpG ODN has been shown to elicit superior cytokine responses in chicken immune cells in vitro compared to other TLR ligands (lipopolysaccharide or Pam3CSK4) 43 . Previous studies have suggested that the protection conferred by CpG-ODN in chickens is primarily due to an enhanced expression of cytokines and chemokines 30,35 . Therefore, we designed experiments to evaluate the effects of CpG-ODN administration on the expression profiles of cytokines involved in innate and adaptive immune responses. We examined cytokine expression kinetics of eight cytokine genes (IFN-α, IL-18, IFN-γ, IL-1β, LITAF, IL-4, IL-6, and IL-10) in spleen and lungs of CpG-ODN treated and the saline control. In our study, cytokine expression was measured using QuantiGene Plex 2.0 ® assay technique, a multiplex assay that can simultaneously measure multiple cytokine genes in a single well of the assay. We observed a substantial upregulation of cytokines following CpG-ODN treatment in spleen and lungs. The mRNA levels of all cytokines tested here tended to be higher in CpG-ODN treated chickens than in the control group at most time points.
In agreement with a previous study 44 , we observed higher expression of IFNs following CpG-ODN injections, which was especially notable in the lungs of chickens. IFNs are important cytokines in shaping immune responses against pathogens. IFN-α and IFN-γ have been shown to act synergistically to develop an antiviral state and  activate macrophages and nitric oxide secretion 45 . Like Th1 cytokines (i.e., IFNs), Th2 cytokine IL-4 and regulatory cytokine IL-10 were also upregulated following CpG-ODN injections and were also more noticeable in lungs.
A previous study reported that CpG-ODN induces a Th1 type of response in chickens 34 . However, in our study, we observed both Th1 and Th2 types of response. This finding is also supported by the recent report that demonstrated both Th1 and Th2 types of cytokine expression in chickens in response to CpG-ODN treatment 30 . Based on present data and previous studies, our data suggest that CpG-ODN in chickens can be used as modulator of both Th1 and Th2 types of cytokine. Accumulating evidence suggests that pro-inflammatory cytokines play a crucial role in promoting strong immune response against pathogens. Our findings of an enhanced expression of pro-inflammatory cytokines, IL-1β, IL-6, IL-18, and LITAF in lungs and spleen following in ovo delivery of CpG-ODN suggests that CpG-ODN promotes inflammatory responses in chickens. Our finding is also supported by recent studies showing upregulation of pro-inflammatory cytokines in chickens following CpG-ODN treatment 37,43 . IL-1β is well known for its induction of IL-6 46 and in T cell proliferation [47][48][49] . We found that among all the cytokines investigated in this study, IL-1β was the most upregulated (several fold higher compared to control) cytokine following CpG-ODN administration. Protective effect of pro-inflammatory cytokines is also evident from the study that demonstrated a strong correlation between pro-inflammatory cytokine level and the resolution of Salmonella Typhimurium infection in chickens 50 . In the present study, we observed that CpG-ODN treatment led to a more pronounced effect on the cytokine profiles in lungs compared to the spleen, which could suggest that CpG-ODN has a superior effect on the mucosal immune system. Our finding is also supported by recent studies that demonstrated enhanced expression of pro-inflammatory cytokines in response to CpG-ODN in intestinal tissues 30 and reduced intestinal colonization with Campylobacter sp. 37 . These findings suggest that CpG-ODN treatment conditions the chicken immune system against pathogenic invasions at mucosal entry levels. Our data suggest that CpG-ODN enhances mucosal immunity and support the role for CpG-ODN as a potent mucosal adjuvant 30 .
Proinflammatory cytokines such as IL-1β, LITAF, IL-18, and IL-6 are rapidly released upon TLR activation 51 . IL-18 can stimulate T-cell proliferation and IFN-γ release in chickens 52 . IL-18 has also implicated in stimulating TNF-α, IL-1β, IL-8, and IL-6 53 . It is worth mentioning that IL-18 and LITAF remained upregulated in both spleen and lungs of hatched chicks for 72 hr post-CpG-ODN treatment. Importantly, among all the cytokines tested here, LITAF remained upregulated for almost all time points studied and in both spleen and lungs. TNF-α is a multifunctional cytokine that has antitumor, antiviral and antibacterial actions. TNF-α induces upregulation of CD40 and other costimulatory molecules in macrophages and dendritic cells 54 .
In the present study, there were substantial differences in the magnitude and kinetics of cytokine expression between CpG-ODN treated and the saline control. These differences can be ascribed to both TLR signaling and cellular composition as well as the number of cells in the organs investigated. Proinflammatory cytokines such as IL-1β and TNF-α orchestrate the secretion of chemokines and leukocyte cell-surface adhesion molecules, supporting the rapid recruitment of immune cells towards inflammatory area [55][56][57] . CpG-ODN was shown to enhance cytokine gene expression in various immune cells in chickens, including bursal cells 58 , macrophages 59 , splenocytes 60 , and thrombocytes 61 . Proinflammatory cytokines such as IL-1β, TNF-α, IL-18, and IL-6 are mainly produced by macrophages when stimulated by bacterial products 62 .
Therefore, we next investigated the recruitment kinetics, and activation of macrophages in the immunological niches in spleen and lungs of CpG-ODN treated and saline controls using flow cytometry. Compared to saline control group, flow cytometry data indicated a significantly enhanced enrichment of macrophages in both spleen and lungs of CpG-ODN treated embryo and chicks. It is worth mentioning that the lungs' immunological niches had very low number of macrophages, sentinel cells that provide the first line of defense against pathogens. These data can explain why chicks are vulnerable to common bacterial infections in their neonatal life 63 , as chicks' immune systems are still not fully mature at the time of hatch. Thus, our study provides new insights into CpG-ODN-mediated antibacterial mechanisms by demonstrating that in ovo delivery of CpG-ODN enriches immunological niches in the immature immune system of the chicken embryo so that pathogenic insults can be tolerated more efficiently at hatch. Maturation of antigen presenting cells (APCs) such as macrophages and dendritic cells involves upregulation of costimulatory molecules such as CD40, CD80 and CD86, and the production of cytokines 64 . Besides, CD40 signaling was shown to activate the APCs and facilitate CD8 + T-cell priming 65 to generate protective CD8 + cytotoxic T cell (CTL) immunity 66 . Therefore, we investigated expression of CD86 and CD40 on macrophages in spleen and lungs of CpG-ODN treated or saline controls. Interestingly, we found that macrophages in CpG-ODN treated chicks were expressing significantly high CD40; however, CD86 was not very different than saline control. This interesting observation can be explained by the cytokine profiles in our study. In the present study, we found significantly higher expression of LITAF in CpG-ODN group in spleen and lungs for several days after treatment. The previous study has reported that TNFα induces upregulation of CD40 onto macrophages and dendritic cells 54 . Furthermore, it was reported that TNF-α upregulated CD40 but not CD86 in APCs such as Langerhans cells 67,68 . Thus, prolonged expression of LITAF in our study and upregulation of CD40 but not CD86 on macrophages may implicate LITAF for this observation, which corroborates previous studies on Langerhans cells 67,68 . Thus, we hypothesize that enhanced LITAF in CpG-ODN group could be the reason for increased CD40 in our study. However, further studies are needed to understand better TLR21-LITAF-CD40 axis in chickens.
We also examined the T-cell compartment in spleen and lungs of CpG-ODN treated and saline controls chicks that hatched 72 hr post-CpG-ODN treatment. Flow cytometry revealed significantly high number of both CD4 + T and CD8 + T cells in lungs and spleen of CpG-ODN treated group. These data suggest that CpG-ODN promotes the enrichment of T-cell immunological niches both in lymphoid (spleen) and non-lymphoid (lungs) organs in chickens. The increase in the number of immune cells in spleen and lungs can be due to either enhanced recruitment of T cells or increase thymic out (generation of T cells in the thymus) or both. Given that in thymus T cells are generated as CD4(+) and CD8(+) double positive cells and after thymic selection T cells egress thymus as either CD4 + T cells or CD8 + T cells. Therefore, we harvested thymus from CpG-ODN treated and saline control chicks to investigate the expression of CD4 and CD8 in thymocytes as an indicator of thymopoiesis. Interestingly, we found significantly high expression of both CD4 and CD8 in the thymocytes of CpG-ODN treated group.
A previous study in mice demonstrated that CPG-ODN treatment resulted in a systemic antigen-independent expansion of both CD4 + and CD8 + T cell subsets and suggested that increase in T cell number was the consequence of reduced T cell death in CpG-ODN treated mice 69 . However, further studies are needed to investigate the role for a reduced T cell death or potentially increased thymic output in neonatal chicks following CpG-ODN treatment to explain the increased T cell numbers in our study. Previous studies using transgenic mice reported that TNF-α is constitutively expressed in the thymus 70 and promotes murine 71 and human 72 T cell development.
In our study, LITAF expression pattern stands out in the cytokine profiles of spleen and lungs. Therefore, we hypothesize that CpG-ODN induced LITAF may have more significant roles in enriching immune compartments such as spleen and lungs with various sentinel immune cells; consequently, neonatal chicks become better equipped to fight against pathogenic insults.
In conclusion, the present study has demonstrated for the first time that in ovo delivery of CpG-ODN provides protection against E. coli infection in chickens by enriching immunological niches in spleen and lungs that enhances immunocompetence of neonatal chicks to tackle pathogenic insults. Furthermore, our data highlight the importance of LITAF in CpG-ODN induced immunoprotective mechanisms and suggest that future studies on TLR21-LITAF-mediated immune enrichment axis will help in harnessing the full therapeutic potential of CpG-ODNs against diseases.

Materials and Methods
Synthetic CpG-ODN. The  Bacteria. The bacterial strain for challenge was obtained from a field isolate of E. coli from a turkey with septicemia as described in previous studies 12,24 . This E. coli was serogroup O 2 , non-hemolytic, serum-resistant, produced aerobactin, with a K1 capsule and Type 1 pili. Aliquots of bacteria were stored at −80 °C in 50% brainheart infusion broth (BHI; Difco, Detroit, MI) supplemented with 25% (w/v) glycerol (VWR Scientific, Inc., Montreal, Quebec). Bacteria used for challenge were cultured on columbia sheep blood agar plates for 18-24 hr at 37 °C. One colony was added to 100 ml of Luria broth in a 250 ml Erlenmeyer flask. The culture was grown at 37 °C for 16-18 hrs with shaking at 150 rpm. Stationary phase culture contained approximately 1 × 10 9 colony forming units (cfu) of bacteria per ml. The cultures were further diluted in sterile saline so the concentration of bacteria required for challenge (1 × 10 5 or 1 × 10 4 cfu/bird) was obtained. Viable bacterial counts were determined by plating serial dilutions of the diluted culture in duplicate on Columbia sheep blood agar plates, incubating for 18-24 hrs at 37 °C; then counting the colonies. Animal model for E. coli challenge. Hatched chicks were allocated into an animal isolation room at the ACU. An identification tag was placed on their neck in order to identify birds that received in-ovo CpG-ODN or saline. Chicks of both groups were housed in the same condition and were kept on floor with wood shavings as litter. Water and commercial broiler ration were provided ad libitum throughout the trial. The animal room was ventilated with filtered, non-recirculated air at a rate of 10-12 changes/hr. In addition, air pressure differentials and strict sanitation were maintained in this isolation facility. One day post-hatch, the birds received either 1 × 10 4 or 1 × 10 5 cfu's of stationary -phase E. coli, in a total volume of 250 μl per bird, by subcutaneous injection in the neck, resulting in E. coli septicemia. In this model, E. coli septicemia with airsacculitis, pericarditis, perihepatitis or polyserositis develops in 60-90% of birds that are not protected by treatment intervention. The two different doses of E. coli were given to groups of birds to simulate field conditions where all birds are not exposed to a consistent dose of E. coli in a barn. Birds were assessed three times daily at the critical stage (first four days post challenge) and twice thereafter up to 8 days post challenge. Birds were observed for clinical signs and each individual bird was assigned a daily clinical score: 0 = normal; 0.5 = slightly abnormal appearance, slow to move; 1 = depressed, reluctant to move; 1.5 = reluctant to move, may take a drink and peck some; 2 = unable to stand or reach food or water; and 3 = found dead. Birds that received a clinical score of 2 were humanely euthanatized by cervical dislocation. Chicks that were found dead or euthanatized were necropsied immediately. On day 8 post E. coli challenge, the remaining birds were euthanatized by cervical dislocation. Bacterial swabs were taken immediately after necropsy from the air sacs and cultured on Columbia sheep blood agar using a typical method of inoculation and streaking on four quadrants of the plate of medium. A semi-quantitative estimate of E. coli isolation was conducted on Columbia sheep blood agar by the quadrant streaking method in which bacterial colonies thinning occurs as streaking goes clockwise from quadrant 1 to quadrant 4. Growth on these plates were recorded on a scale from 0 to 4+, where 0 = no growth; 1+ = growth of bacteria on the quadrant 1 only; 2+ = growth of the bacteria on quadrants 1 and 2; 3+ = growth of bacteria on quadrants 1, 2, and 3; and 4+ = growth of bacteria on all quadrants 1-4.
Cells for flow cytometry. Cell preparation and antibody staining for flow cytometry was done as previously described with some modifications [73][74][75] . Briefly, spleen, lung and thymus tissues were collected at 24, 48 and 72 hr  Table 1. Frozen spleen and lung tissues were processed and tissue homogenates were prepared following the manufacturer's instructions with some modifications. Briefly, 5 mg of tissue was excised, manually broken down and added to 300 µl of homogenization solution and 3 µl of Proteinase K. The tissue lysate was digested at 65 °C to release the mRNA and centrifuged to precipitate the debris. The supernatant of tissue lysates were collected and stored in −80 °C if not used the same day. The oligonucleotide capture probes for mRNA and label probes were designed by the manufacturer as requested. The tissue homogenates were added to a 96-well plate (40 µl/well) that was pre-loaded with 20 mL of the capture reagent and the respective probe set. After overnight hybridization at 54 ± 1 °C, hybridizations with bDNA pre-amplifier 2.0, bDNA amplifier 2.0, biotinylated label probe and finally substrate were subsequently carried out according to the manufacturer's instructions. Luminescence was quantified using a Luminex instrument (Bio-Rad, USA). Signals, which is the mean fluorescence intensity (MFI) generated from each bead, are proportional to the amount of each mRNA captured on the surface of each generated specific probe set 76 . The expressions of these genes were normalized with the expression of hypoxanthine-guanine phosphoribosyl transferase 1 (Hprt 1) and tubulin beta 1 (Tubb1) genes which were used as housekeeping genes (HKGs). For gene expression analysis, first, the MFI data of each gene was normalized by dividing with the average of Hprt 1 and Tubb 1HKGs. Then, using normalized MFI data fold changes were calculated in CpG-ODN group by determining how many folds expression compared to saline control group at each time point.
Flow cytometry. The lung and spleen cell populations were stained for the presence of monocyte/macrophages, CD4 + and CD8 + T cells. The monocyte/macrophages were further analysed for the expression of maturation markers (CD40 and CD86). Briefly, ~5 × 10 5 cells were incubated with mouse anti-chicken monocyte/macrophage phycoerythrin (PE) antibody at 4 °C for 30 min for detecting APCs. For determining the maturation state of monocyte/macrophages, the cells from previous step were washed three times and incubated with either mouse anti-chicken CD40 or CD86 primary antibodies separately at 4 °C for 30 min. After three washings with PBS, the cells were stained with PerCP/Cy5.5 goat anti-mouse IgG secondary antibody at 4 °C for 30 min. Another set of ~5 × 10 5 cells were also incubated with mouse anti-chicken CD8 (FITC) and CD4 (PE) together at 4 °C for 30 min to determine CD4 + and CD8 + T cells. Finally, the washed cells were suspended in 300 µl buffer in flow tubes and processed for flow cytometric analysis. Flow cytometry data were acquired by Epics XL (Beckman Coulter) and FACS Caliber (BD Bioscience), and data were analyzed with FlowJo software (Tree Star).
Statistical analysis. Survival pattern, cumulative clinical scores (CCS), bacterial percentages and cell populations from flow cytometry analysis were graphed and analyzed with the use of Prism (Prism 5.0, GraphPad Software Inc., San Diego, CA) with a significance level of P < 0.05. The survival patterns were compared using the log-rank test and chi-square statistic. For calculating CCS, the clinical score for each bird was summed over the 8-day observation period and the significance of differences among groups were tested with Kruskal-Wallis nonparametric analysis of variance. Dunnett's test following ANOVA testing was used to test for significant differences of gene expression between the CpG-ODN group and the saline control group at each time point. For testing difference of APC percentages, CD4 + and CD8 + expression between groups, we used a two way ANOVA followed with Bonferroni post-test and Student-t test with Welch's correction for unequal variance, with a significant difference of P < 0.05.