Magang goose (Anser domesticus) were purchased from a farm in Guangdong Province, China, and 12-week-old goose were used in this experiment. The goose was maintained under natural light and temperature conditions for at least 2 weeks and fed commercial diets according to the recommended nutritional standard for the breed. Food and water were sterilized before being fed to the goose. Latex beads of different diameters (0.5 μm, 0.8 μm, 1.0 μm, and 2.0 μm) were purchased from Sigma-Aldrich. S. aureus, Aeromonas hydrophila, and E. coli were cultured according to routine laboratory protocols. Entire experiments were performed following recommendations of the Laboratory Animals (Ministry of Science and Technology of China 2006) and approved by the Animal Ethics Committee of Zhongkai University of Agriculture and Engineering.

The goose erythrocytes were isolated and purified according to a previously described procedure with minor modification (20). Briefly, blood samples were collected from a wing vein using heparinized syringes and mixed with 0.9% buffered saline (0.9 g NaCl in 100 ml H₂O) containing heparin sodium (0.1 mg/ml). After washing with 0.9% buffered saline by centrifugation, the cell suspension was layered onto a 34%/51% Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden) density gradient and centrifuged at 500 × g for 30 min at 4°C. After the plasma and white blood cells were removed (21), the RBC pellets were collected, washed three times, and finally resuspended in 0.9% buffered saline. The 5% trypan blue solution was used to make sure the cell viability >95%. The isolated erythrocytes were cultured in L15 medium containing 10% calf serum (Gibco, Carlsbad, USA), 100 U/mL penicillin (Gibco, Carlsbad, USA), 100 µg/mL streptomycin (Gibco, Carlsbad, USA), and 50 µg/mL gentamicin (Gibco, Carlsbad, USA).

An aliquot of purified erythrocytes (1 × 10⁶cells/ml) was incubated with beads of different sizes (0.1 μm, 0.5 μm, 1.0 μm, and 2.0 μm, Sigma, USA) in a 1:10 ratio for 2 h at 37°C. Later, the suspensions were washed three times with sterile 0.9% buffered saline to remove non-adherent beads and observed under a fluorescence microscope (Leica DMI8, Germany). The adhesion rate was analyzed using a FACScan flow cytometer (BD Biosciences, USA). Next, another group of purified erythrocytes incubated with 0.8-μm beads (Sigma, USA) at the ratio of 1:10. The mixture was incubated for 1 h, 2 h, and 4 h at 37°C. After washing, bright-field photographs were obtained under a fluorescence microscope.

The purified goose red blood cells (1 × 10⁶ cells/ml) were incubated with latex beads (0. 8-μm diameter) at a ratio of 1:10. After incubation at 37°C for 1 h, 2 h, and 4 h, the beads were washed three times with 0.9% sterile buffered saline and fixed in 2.5% glutaraldehyde at 4°C for at least 2 h. Finally, scanning electron microscopy (SEM) (Hitachi SU8010, Japan) and transmission electron microscopy (TEM) images (Hitachi H-7000FA, Japan) were used to observe the phagocytic activity of goose RBC. Scanning electron microscope magnification (× 8.00 k and × 20.0 k) and transmission electron microscopy magnification (× 4.0 k).

Fluorescein isothiocyanate (FITC) was stained and labeled with S. aureus, A. hydrophila and E. coli at a ratio of 1:100, respectively, and incubated in a constant temperature incubator at 37°C for 30 min sheltered from light. After that, the labeled bacterial solution was resuspended to a concentration of 10⁷ CFU/ml by using 0.9% sterile buffered saline. Purified erythrocytes (1 × 10⁶ cells/ml) were incubated with fluorescein isothiocyanate (FITC)-labeled S. aureus, A. hydrophila, and E. coli (1 × 10⁷ CFU/ml) at a ratio of 1:10 in an incubator at 37°C for 1 h, 2 h, and 4 h. Later, the suspensions were washed three times with 0.9% sterile buffered saline to remove unattached bacteria, and the phagocytosis rate was analyzed by a FACScan flow cytometer (BD Biosciences, USA).

Experiments were performed according to the instructions of the kit, and levels of ROS and iNOS in erythrocytes were measured with the fluorescent probe 2’-7’-dichlorodihydrofluorescein-diacetate (DCFH-DA, Sigma, USA), which penetrates cells and hydrolyzes to non-fluorescent dichlorodihydrofluorescein (DCFH). NO production was measured by the fluorescence probe DAF-FM (Sigma, USA). Briefly, purified erythrocytes (1 × 10⁵ cells/ml) were incubated with S. aureus (10⁶ CFU/ml), A. hydrophila (10⁶ CFU/ml), and E. coli (10⁶ CFU/ml) at 1 h, 2 h, and 4 h, and suspensions were collected at these time points. The mixtures were incubated with 1 μmol/ml DCFH-DA (Sigma, USA) or DAF-FM (Sigma, USA) for 25 min at 37°C in a sheltered room, and then, the cells were washed three times with 0.9% sterile buffered saline and resuspended in 500 μL 0.9% sterile buffered saline for analysis with a FACScan flow cytometer (BD Biosciences, USA).

The purified erythrocytes (1 × 10⁶ cells/ml) were incubated with S. aureus (10⁷ CFU/ml), A. hydrophila (10⁷ CFU/ml), and E. coli (10⁷ CFU/ml) in an incubator at 37°C for 1 h, 2 h, 4 h, and 8 h. Afterwards, total erythrocytes RNA were obtained with RNAiso Plus according to the manufacturer’s protocol. Approximately 1 µg of total RNA was used to generate cDNA. qRT-PCR was performed using HIScript^®Q Select RT SuperMix (Vazyme, Nanjing, China) according to the manufacturer’s instructions, and then stored at -20°C. Primers were designed by Primer Premier 5.0, and listed in Table 1 .

The qRT-PCR was performed in a qTOWER3 real-time PCR thermocycler (Analytik Jena AG, Germany), following the protocol for the AceQ qPCR SYBR Real-time PCR Premixture (Vazyme, Nanjing, China). In this study, two subunits of NADPH enzymes, NOX1 and NOX5, were selected for expression analysis, and β-actin was used as an internal reference gene. The qRT-PCR procedures were 5 min at 95°C, followed by 40 cycles consisting of 15 sec at 95°C, 30 sec at 60°C, and 20 sec at 72°C, and lastly, for 5 min at 4°C. There were three replicates for each treatment, and all technical experiments were triplicated (22). The 2^-ΔΔCT method was used to calculate the mRNA levels (23).

We further investigated the antimicrobial activity of erythrocytes, which was performed by the previously described method, with minor modifications (12). Briefly, each 1 ml of erythrocytes (5 × 10⁶ cells/ml) was incubated with S. aureus and E. coli at 37°C for 1 h, 2 h, 4 h, and 8 h, after which the infected cells were serially diluted 10-fold with sterile saline, spread on nutrient agar plates, and incubated overnight at 37°C. The antibacterial activity of the erythrocytes was determined by observing how the number of bacterial colonies on the agar plates differed from the control group.

The activities of superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GSH-Px) were measured independently in the purified goose erythrocytes (5 × 10⁶ cells/ml) after treated with S. aureus, A. hydrophila, and E. coli for 1, 2, and 4 h. Bacterial concentration adjusted to (5×10⁷ CFU/ml) at 10-fold dilution. SOD, CAT, and GSH-Px assay kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China) were used to determine the response of antioxidant system in goose erythrocytes SOD, CAT, and GSH-Px according to the manufacturer’s instructions.

The data were statistically analyzed using SPSS software (version 23.0). All data are expressed as the mean ± SEM. Statistical significance was measured by one-way analysis of variance (ANOVA) and t-test, using graphs drawn with GraphPad Prism 7 software. The differences were statistically significant at P < 0.05 (*) or P < 0.01 (**).

The fluorescence microscopy was used to detect the ability of the goose erythrocytes to uptake and attach different size of beads (0.1 μm, 0.5 μm, and 1.0 μm). As shown in Figure 1A , numerous fluorescence of different size of beads were recorded over the surface or inside of the individual erythrocytes. Followed, flow cytometric analysis showed that goose erythrocytes were able to uptake different diameters of fluorescent latex beads with different efficiencies ( Figures 1B, C ). Specifically, the erythrocyte phagocytosis rates of 0.1 μm, 0.5 μm, 1.0 μm, and 2.0 μm microbeads were 94.1%, 30.4%, 10.9%, and 0.4%, respectively. After incubated with different sizes of fluorescent beads for 2 h, it was observed that 0.1 μm beads were more easily phagocytosed, followed by 0.5 μm beads. For 2.0 μm beads, the goose erythrocytes phagocytosis rate was much lower compared to the smaller beads (0.1 μm, 0.5 μm, and 1.0 μm) ( Figures 1B, C ).

To further test the phagocytic ability of goose erythrocytes, 0.8 μm latex beads were incubated with erythrocytes at 37°C for 1 h, 2 h, and 4 h. Under the optical microscopy, the 0.8 μm latex beads were found to be surrounded the goose erythrocytes ( Figure 2A ). Furthermore, SEM observation data showed that beads had tightly bound to erythrocytes and formed distinct convex vesicle structures over the surface of the RBC ( Figure 2B ). The SEM analysis was performed to characterize the phagocytosis of goose erythrocytes, and the results showed that some 0.8 μm latex spheres attached to the surface of individual erythrocytes and then formed invaginations, which were partially phagocytosed ( Figure 2C ). To further confirm the phagocytic activity of erythrocytes, which were observed under TEM, and it was clear that latex beads were also visible within the erythrocytes ( Figure 2D ).

To investigate the phagocytosis and adhesion ability of goose erythrocytes towards bacteria, we incubated goose erythrocytes with S. aureus, A. hydrophila, and E. coli for 1 h, 2 h, and 4 h. By flow cytometry analysis, the phagocytosis and adhesion activity of E. coli by erythrocytes was increased from 1 h to 4 h ( Figures 3B, D ), but decreased at 4 h ( Figure 3F ), and the phagocytosis and adhesion activity for S. aureus by goose erythrocytes increased with time ( Figures 3A, C, E ). For A. hydrophila, the phagocytosis and adhesion rate at 2 h and 4 h were significantly higher than that of in 1 h ( Figures 3C, E ).

To further investigate the antimicrobial mechanism of goose erythrocytes, we measured the ROS and iNOS levels of erythrocytes stimulated by three kinds of bacteria for 0.5, 1, 2, 4 and 6 h. The flow cytometry analysis data revealed that the stimulation of S. aureus significantly increased the ROS production compared to control groups, and reached the highest level at 1 h ( Figures 4A, B ). In A. hydrophila stimulation groups, the production of ROS has a highest level at 4 h, and the production of ROS in A. hydrophila stimulation groups were significantly higher than that in control groups ( Figures 4A, C ). Similar to S. aureus stimulation groups, the stimulation of E. coli significantly increased the production of ROS at all tested time points, and at 1 h showed a highest level ( Figures 4A, D ). In determination of iNOS assay, the results data showed that the stimulation of S. aureus significantly increased the iNOS production at all tested time points, and at 0.5 h showed the highest iNOS levels ( Figures 5A, B ). In A. hydrophila and E. coli stimulation groups, the data showed that the highest iNOS level both at 6 h, and the production of iNOS at all time points were significantly higher than that of in control groups ( Figures 5A, C, D ).

NOX family NADPH oxidases are proteins that transfer electrons across biological membranes. Usually, the electron acceptor is oxygen and the product of the electron transfer reaction is superoxide. Therefore, the biological function of NOX enzymes is to generate ROS. The homologs of the catalytic subunit of NADPH oxidase, namely NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1, and DUOX2 which are all involved in the regulation of ROS in macrophages. To further verify whether NADPH in erythrocytes is involved in the regulation of ROS, we measured the transcript levels of two subunits of NADPH enzymes, NOX1 and NOX5 through qRT-PCR. The results showed that after stimulation by S. aureus, A. hydrophila, and E. coli, NOX1 transcript levels were significantly upregulated at 1 h and 4 h with a significant difference (P < 0.01), and NOX5 transcript levels were significantly upregulated at 2 h and 4 h with a significant difference (P < 0.01) ( Figure 6 ).

To demonstrate the antimicrobial ability of goose erythrocytes, we performed antibacterial experiments with erythrocytes acting against S. aureus and E. coli. The results showed that the erythrocytes exhibited strong activity against both bacteria, with erythrocytes showing stronger antibacterial activity against S. aureus at 2 h and 4 h compared to the control ( Figure 7A ), and against E. coli at 2 h and 4 h ( Figure 7B ).

To investigate the influence of bacterial incubation to the antioxidant system in goose erythrocytes, the antioxidant indicators including CAT, GSH-Px, SOD, and other antioxidant enzyme activities were measured, whose basic function is to scavenge superoxide radicals such as superoxide anion (O^2-) and hydrogen peroxide (H₂O₂). The results showed that the incubation of three bacteria resulted in an increase in the CAT activity with time, especially a significant increase at 2 h and 4 h with significant differences (P < 0.01) ( Figure 8A ). All three bacteria stimulated GSH-Px activity, which increased at 1 h, 2 h, and 4 h, with a significant increase by S. aureus at 4 h with significant differences (P < 0.01) ( Figure 8B ). A significant increase in SOD activity was also observed at 2 h and 4 h with significant differences (P < 0.01) ( Figure 8C ).

To study the behavioral characteristics of goose erythrocytes during bacterial infection, TLR4, IL-1β, IL-6, and IL-8 were measured by qRT-PCR after S. aureus, A. hydrophila, and E. coli stimulation. We found that TLR4 gene expression was significantly upregulated in goose erythrocytes after 2 to 4 h of stimulation by the three bacteria ( Figure 9A ), but there was no significant difference (P > 0.05) in S. aureus stimulation at 2 h ( Figure 9B ). The expression of the IL-1β gene was significantly upregulated after A. hydrophila stimulation for 1 h, while the expression of the IL-1β gene was significantly upregulated after S. aureus and A. hydrophila stimulation for 4 h, with significant differences (P < 0.01). However, the expression of the IL-1β gene was upregulated but was not significantly different (P < 0.01) after E. coli stimulation from 1 to 4 h ( Figure 9B ). In addition, IL-6 gene expression levels were significantly upregulated during E. coli stimulation from 2 to 4 h with significant difference (P < 0.01), while for S. aureus stimulation at 4 h, IL-6 gene expression levels were significantly upregulated with significant difference (P < 0.01). In contrast, IL-6 gene expression levels were upregulated during A. hydrophila stimulation for 2 to 4 h, but were not significantly different (P < 0.01) ( Figure 9C ). IL-8 gene expression levels were significantly upregulated during E. coli stimulation for 1 h, while after S. aureus and A. hydrophila stimulation for 2 to 4 h, IL-8 gene expression levels were significantly upregulated with a significant difference (P < 0.01) ( Figure 9D ).