The H5N6 virus A/duck/Guangdong/16873/2016 (DK16873) used in this study was isolated from swabs of ducks at a live bird market in Guangdong, China in 2016. The H5N6 virus was isolated in 9-day-old specific-pathogen-free (SPF) embryonated hen eggs, using methods described in the literature (23). The 50% tissue culture infective dose (TCID₅₀) of the H5N6 virus was determined by infection of DEFs; 50% egg infective doses (EID₅₀) was determined by infection of embryonated hen eggs, and values were calculated using the Reed-Muench method (24). DEFs were derived from 9-day-old duck embryos and cultured in Dulbecco's modified Eagle medium (GIBCO, USA) complemented with 10% fetal bovine serum (GIBCO, USA). Four-week-old healthy Peking ducks were purchased from farms in Guangdong and housed in the isolators of the ABSL-3 facilities. The Peking ducks were confirmed to be serologically negative for avian influenza with a hemagglutination inhibition (HI) test.

According to the predicted sequence of duck-TRIM32 (duTRIM32) from the National Center for Biotechnology Information (Accession number: XP_005029948.1), primers were designed and used to amplify full-length coding sequence (CDS) of duTRIM32 cDNA from total spleen RNA of Peking ducks (Table 1). Amino acid sequences of different species TRIM32 were aligned using Clustal X2.1 software and edited with BOXSHADE (https://embnet.vital-it.ch/software/BOX_form.html). The functional domains of duTRIM32 were predicted using the SMART program. MEGA software (vesion 4.0) was also used to conduct phylogenetic analysis of TRIM32 amino acid sequences of 10 species.

To analyze the distribution of duTRIM32 in healthy ducks, three 4-week-old healthy ducks were euthanized humanely to collect tissues that included livers, pancreases, hearts, brains, kidneys, lungs, spleens, intestines, thymuses, bursa of Fabricius, and stomachs. The lung, spleen and brains are the major target tissues for H5 subtype AIVs infection, and our previous study have demonstrated that the host immune responses in these organs may influence the outcome of the infection (25). Therefore, we have quantified the expression of duTRIM32 in the lungs, spleens, and brains of the infected ducks to evaluate if duTRIM32 involved in the host immune response to H5N6 infection. To investigate the expression of duTRIM32 in H5N6 infected ducks, six 4-week-old ducks were inoculated intranasally with 10⁶ EID₅₀ of DK16873 H5N6 HPAIV in a volume of 200 uL. At 12 h and 3 days post-inoculation, each of the three inoculated birds were euthanized humanely to collect lungs, brains and spleens. Three control ducks were treated with PBS in the same way and were also euthanized humanely to collect lungs, brains and spleens. Total RNA of the collected tissues were extracted using an Eastep®Super Total RNA Extraction Kit (Promega, USA) according to the manufacturer's instructions. Total RNA (1 μg) was reverse-transcribed into cDNA using M-MLV Reverse Transcriptase (Promega, USA) according to the manufacturer's protocol. Quantitative real-time PCR (qRT-PCR) was conducted using a GoTaq® qPCR Master Mix (Promega, USA) and was run on a Bio-Rad CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad, USA) with the following settings: 5 min at 95°C, then 15 s at 95°C and 34 s at 62°C for 40 cycles.

The obtained sequence of full length duTRIM32 was subcloned into the pCAGGS plasmid with specific primers (Table 1). Other plasmids and luciferase reporter plasmids used in this study (duSTING-HA, pGL3-IFN-β- Luc, pRL-TK, and pCAGGS-vector) were kept in our labatory and were constructed using previously described methods (26). For the luciferase reporter assay, expression plasimids (duTRIM32-V5 and duSTING-HA), reporter plasmids, pCAGGS-vector and pRL-TK were transfected into DEFs using the FuGENE HD transfection reagent (Promega, USA). The pRL-TK plasmid (Promega, Madison, WI, USA), encoding the Renilla luciferase protein, was used as an internal control plasmid to normalize transfection efficiency. In addition, we also used pAc-GFP to observe the transfection efficiency. After 24 h, DEFs were collected and tested for luciferase activities using the dual-luciferase Assay System (Promega, USA) according to the manufacturer's instructions.

DEFs were seeded on 6-well plates and cultured until the cells were 70% confluent. DuTRIM32-V5 or pCAGGS-vector was then transfected into the corresponding well (n = 3) using the FuGENE HD transfection reagent (Promega, USA) and cultured at 37°C with 5% CO₂. After 24 h, DEFs were collected and the mRNA of IFN-β, IRF7, and Mx were tested using qRT-PCR described above.

Coimmunoprecipitation (Co-IP) assays were conducted to analyze the interaction between duTRIM32 and duSTING. For Co-IP experiments, 293T cells were seeded on 100 mm dishes (10⁷ cells/dish) overnight. Expression plasmids or empty plasmids were used to transfect 293T cells. After 24 h, cells were lysed using cell lysis buffer (P0013, Beyotime, China) containing 1 mM PMSF. Lysates were centrifuged at 12,000 g, for 15 min at 4°C and precipitated with anti-HA mAb, in conjunction with protein G Agarose beads (ThermoFisher Scientific, USA) for at least 4 h at 4°C. The beads were washed with cold PBS 5 times (10 min/wash), and subsequently eluted with SDS loading buffer by boiling for 10 min. Proteins isolated from the beads and the cell lysates were separated by SDS-PAGE. In addition, the separated proteins were electroblotted onto polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA, USA). The membranes was blocked with 5% BSA in phosphate buffer solution and was then incubated for 1 h with mouse anti-HA and anti-V5 antibodies (Sigma, German) at a dilution of 1:2,000 and 1:3,000, respectively. The goat anti-mouse-IRDye 800 (LI-COR Biosciences, Lincoln, NE USA) was used as the secondary antibody at a 1:10,000. dilution. To determine whether ubiquitination of duSTING can be detected after the interaction, the 293T cells were co-transfected with duSTING-Myc, Ub-HA, and duTRIM32-V5. The duSTING-myc and Ub-HA transfected group was used as a control group. Cell lysates were immunoprecipitated with anti-Myc mAb, in conjunction with protein G Agarose beads (ThermoFisher Scientific, USA) and the immunoprecipitates were analyzed by Western blot with mouse anti-HA antibodies (Sigma, German) at a dilution of 1:2,000. The secondary antibody was indicated before.

Antiviral experiments were conducted using previously described methods (26). DEFs were seeded on 12-well plates and cultured until cells were 70% confluent. Expression plasmids (duTRIM32-V5) or empty vectors was transfected into DEFs using FuGENE HD transfection reagent. After 24 h, DEFs were infected with DK16873 H5N6 virus at a dose of 5 TCID₅₀. Cell supernatants were collected at 12, 24, 36, and 48 HPI. DEFs were used to determine TCID₅₀ titers.

All experiments with H5N6 HPAIV were conducted in Animal Biosafety Level 3 (ABSL-3) and Biosafety Level 3 (BSL-3) facilities. The protocol (SCAUABSL2017-011) was approved by the Experimental Animal Administration and Ethics Committee of the South China Agricultural University.

GraphPad Prism 7.0 software (GraphPad Software Inc, USA) was used to perform statistical analysis. The Student's t-test was used to analyze the statistical significance of the differences. A p-value < 0.05 was considered to be statistically significant (^*p < 0.05; ^**p < 0.01).

Based on the predicted sequence of duTRIM32 on the National Center for Biotechnology Information (NCBI), primers were designed and used to amplify full-length duTRIM32 cDNA from total duck spleen RNA. The full-length duTRIM32 contains 2,049 bp, encoding 682 amino acids (Figure 1). Sequence alignment results revealed that duTRIM32 shares 98.3% similarity in amino acid sequences with goose TRIM32, 95.5% with chicken TRIM32, 84.9% with human TRIM32, 84.3 % with mouse TRIM32 and 61.5% with fish TRIM32, respectively. Phylogenetic analysis revealed that duTRIM32 is clustered into the bird clade (Figure 2). The SMART program and the well-understood structure of human TRIM32 were used to predict the functional domains of duTRIM32. Results indicate that duTRIM32 contains a conserved RING-finger domain, a BBOX domain and a coiled-coil domain, which are similar to those of mammalian TRIM32 (Figure 2).

To analyze the expression profiles of duTRIM32 in healthy ducks, we performed qRT-PCR to determine the expression of duTRIM32 in different tissues. qRT-PCR assays were conducted as described in the Materials and Methods. Results indicate the highest expression of duTRIM32 is in the liver, followed by the pancreas, heart, brain, kidney, lung, spleen, intestine, thymus, bursa of Fabricius, and stomach (Figure 3). Therefore, our results indicated that duTRIM32 is ubiquitously expressed in all the tested tissues of healthy ducks.

To investigate if duTRIM32 is involved in the host immune response to viral infection, we quantified the mRNA levels of duTRIM32 in the lungs, spleens, and brains of H5N6 HPAIV infected ducks using techniques described in the Materials and Methods. Results from these studies indicated that the mRNA of duTRIM32 is slightly changed at 12 HPI in the lungs, spleens and brains. Notably, duTRIM32 was found to be upregulated in these tissues of the H5N6 HPAIV infected ducks at 3 days post-infection (DPI), especially in the lungs and spleens, with a fold increases of 7.23 (P < 0.05) and 5.29 (P < 0.05), respectively (Figure 4). Therefore, our results suggest that duTRIM32 might be involved in the host innate immune response to H5N6 HPAIV infection.

To determine whether duTRIM32 functions in the type I IFN pathway, we performed a luciferase reporter assay to test the ability of duTRIM32 to activate the IFN-β promoter. The luciferase reporter assay was completed as described in the Materials and Methods. The pCAGGS is an empty plasmid and used as a negative control in this study. The results show that transfection of pCAGGS could not activate the IFN-β promoter (Figure 5). These results indicate that overexpression of duTRIM32 can significantly activate the IFN-β promoter (4.34-fold, P < 0.05) (Figure 5). In addition, we also quantified the mRNA of IFN-β, IRF7 and Mx in the duTRIM32-transfected DEFs. Results from these studies indicate that duTRIM32 significantly upregulates the expression of IFN-β (4.58-fold, P < 0.05), IRF7 (3.67-fold, P < 0.05) and Mx (5.26-fold, P < 0.05) (Figure 6). Therefore, these results reveal that duTRIM32 can activate the type I IFN pathway and upregulates the expression of the related antiviral genes.

Mammalian TRIM32 was found to interact with STING and to contribute to the expression of IFN-β mediated by duSTING, thereby positively regulating the type I IFN pathway (11). We performed a dual luciferase reporter assay to determine whether duTRIM32 contributes to duSTING-mediated expression of IFN-β. Results from these studies showed that both duTRIM32 and duSTING can activate the IFN-β promoter, with fold increases of 4.34 (P < 0.05) and 28.63 (P < 0.05), respectively. Co-transfection of duTRIM32 and duSTING exhibited stronger activation of the IFN-β promoter than the transfection of duSTING, with a fold increase of 43.03 (P < 0.05). But the increased level between duTRIM32 and duSTING co-transfection group and duSTING transfection group is <2-folds (Figure 7).

In addition, we co-transfected duTRIM32 and duSTING expression plasmids into 293T cells and the interaction between duTRIM32 and duSTING was analyzed by Co-IP assays. Results showed that duTRIM32 directly interacted with duSTING. Both duTRIM32 and duSTING were correctly expressed in 293T cells, with a molecular weight of 76 kDa and 43 kDa, respectively (Figure 8, Supplementary Material). These results confirmed that duTRIM32 can directly interact with duSTING. Furthermore, we found that duTRIM32 was demonstrated to ubiquitinate duSTING in 293T cells (Figure 9).

Therefore, our results revealed that duTRIM32 can interact with duSTING and contributes to duSTING-mediated expression of IFN-β.

To investigate the antiviral effects of duTRIM32, we transfected duTRIM32 or an empty vector in DEFs. DEFs were then infected with DK16873 H5N6 HPAIV at 24 h post-transfection. Samples were collected at 12, 24, 36, and 48 h post-infection. The viral titers of the collected samples were measured by the TCID₅₀ method in DEFs. Results indicated that the viral titers of H5N6 HPAIV in the duTRIM32-transfected group were lower than that of the empty vector transfected group. Notably, overexpression of duTRIM32 could significantly inhibit the replication of H5N6 HPAIV at 24 HPI (P < 0.05) and 36 HPI (P < 0.05) (Figure 10). Therefore, our results revealed that duTRIM32 can inhibit the H5N6 HPAIV infection.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.