Negative Regulators of Inflammation Response to the Dynamic Expression of Cytokines in DF-1 and MDCK Cells Infected by Avian Influenza Viruses

The H5N1 and H9N2 avian influenza viruses (AIVs) seriously endanger the poultry industry and threaten human health. Characteristic inflammatory responses caused by H5N1 and H9N2 AIVs in birds and mammals result in unique clinical manifestations. The role of anti-inflammatory regulators, PTX3, Del-1, and GDF-15, in H5N1 and H9N2-AIV-mediated inflammation in birds and mammals has not yet been verified. Here, the expression of PTX3, Del-1, and GDF-15 in DF-1 and MDCK cells infected with H5N1 and H9N2 AIVs and their effect on inflammatory cytokines were analyzed. Infection with both AIVs increased PTX3, Del-1, and GDF-15 expression in DF-1 and MDCK cells. Infection with H9N2 or H5N1 AIV in DF-1 and MDCK cells with overexpression of all three factors, either alone or in combination, inhibited the expression of tested inflammatory cytokines. Furthermore, co-expression of PTX3, Del-1, and GDF-15 enhanced the inhibition, irrespective of the cell line. The findings from this study offer insight into the pathogenic differences between H5N1 and H9N2 AIVs in varied hosts. Moreover, our findings can be used to help screen for host-specific anti-inflammatory agents.


INTRODUCTION
Avian influenza virus (AIV), a type A influenza virus, is a threat to both the poultry industry and human health [1]. Based on their pathogenicity in chickens, AIVs can be divided into two groups-highly pathogenic avian influenza viruses (HPAIVs) and low pathogenic avian influenza viruses (LPAIVs) [2]. HPAIV H5N1 and LPAIV H9N2 are the most important causes of avian influenza in China [3,4]. Previous reports demonstrate that the clinical symptoms that arise due to H9N2 and H5N1 AIVs infection are unique. Specifically, H9N2 AIV usually results in a mild host response, while H5N1 AIV causes respiratory distress syndrome and can even result in death [5,6]. These clinical differences reflect the distinctive inflammatory responses caused by different AIV subtypes.
Inflammation is one of the most important host responses against influenza virus infection, and a core indicator of virus pathogenicity [7]. Leukocyte exudation is an important hallmark of the inflammatory response, and involves leukocyte margination, adhesion, and emigration [8][9][10]. Negative regulators of inflammation are evolutionarily conserved and play an important role in inhibiting the inflammatory response [11][12][13]. PTX3, Del-1, and GDF-15 are negative regulators of inflammation and are involved in different stages of leukocyte exudation [14][15][16]. Developmental endothelial locus-1 (Del-1) contains three epidermal growth factor-like domain repeats and two C-terminal discoid I-like domains [12,17]. Del-1 can bind to LFA-1 integrin and antagonize the interaction between LFA-1 on neutrophils and ICAM-1 on endothelial cells [18]. Consequently, Del-1 effectively blocks LFA-1-dependent neutrophil adhesion to the vascular endothelium [19,20]. Pentraxin 3 (PTX3) was the first long chain pentamer to be discovered and plays an important role in humoral immunity [21][22][23]. As an inhibitor of leukocyte aggregation, PTX3 competitively interferes with P-selectin interactions. This blocks neutrophil rolling and prevents neutrophil recruitment [24][25][26][27]. Growth differentiation factor 15 (GDF-15) is a member of the transforming growth factor-β superfamily [28,29]. The regulatory effect of GDF-15 on leukocyte aggregation is mediated by activating Cdc42 GTPase and inhibiting Rap1 GTPase, thereby blocking the activation of β2 integrin, which is induced by chemokines [30,31]. Considering an inhibitory effect of these three negative regulators on the typical phase of leukocyte exudation in the inflammatory response, how they participate in the inflammatory response caused by different avian influenza virus remains to be clarified.
To date, the anti-inflammatory effects of PTX3, Del-1, and GDF-15 in the context of AIV infection have not yet been explored. Herein, how PTX3, Del-1, and GDF-15 impact inflammation in response to AIV infection was investigated. Furthermore, whether these inflammatory responses overlap between avian and mammalian cells was also examined. To accomplish this, DF-1 and MDCK cells with overexpression of PTX3, Del-1, and GDF-15, either alone or in combination, were infected by H5N1 and H9N2 AIVs, and the expression profile for inflammatory cytokines was assessed. Collectively, the results from our study provides further explanation for the pathogenic differences between AIV strains in varied hosts.

Plasmid Construction and Cell Transfection
The primer pairs used in DF-1 cells for PTX3, Del were digested by the restriction enzymes and cloned into the corresponding sites of pEGFP-N1 vector. All recombinant plasmids were verified by sequencing.
DF-1 and MDCK cells were seeded in six-well plates at 2.5 × 10 5 cells/well, and transfected with plasmids using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. At 6-hours (h) posttransfection, fresh DMEM containing 2% FBS replaced the transfection mixture, and cells were incubated for an additional 48 h.
Cell transfection with PTX3, Del-1, and GDF-15 was performed in one of the following three ways: (1) plasmids constructs were transfected individually; (2) constructs were co-transfected at a 1:1 ratio (paired combinations); or (3) all three constructs were cotransfected together at a 1:1:1 ratio. Twenty-four hours posttransfection, cells were infected with H9N2 or H5N1 AIV as described below.

Cell Infection
Cells were infected with H9N2 (10 -4.6 TCID 50 /mL) or H5N1 (10 -6.8 TCID 50 /mL) AIV at an MOI of 0.1 for 1 h at 37 °C. After incubation, the cells were washed twice by PBS and then incubated with DMEM containing 0.2% BSA at 37 °C with 5% CO 2 . For H9N2 infection, TPCK-trypsin (0.25 mg/mL) was added to the culture medium. Cells and supernatant were collected at 0, 3, 6, 12, and 24 h post-infection. Three independent assays on each sample at specified time points and entire experiment three times were performed.

RNA Extraction, cDNA Preparation, and Real-time PCR
Total RNA was extracted, treated, quality determined, stored, and reverse-transcribed to cDNA as described in our previous reports [34,35]. Relative expression levels of IL-1β, IL-6, and TNF-α were determined by real-time PCR using DNA Engine 7500

Determination of Virus Titer and the Virus Growth Curve
The virus supernatant was diluted tenfold from 10 -1 to 10 -10 by serum-free incubation solution with 8 replicates for each dilution. After incubation, the cells were washed twice by PBS and then incubated with diluted virus at 37 °C with 5% CO 2 for 1 h, and set two rows as negative control. Then, the diluted virus was discarded and 200 μL maintenance solution was added to each well and continued to incubate. The cell growth was observed and recorded every day, and the cytopathic criteria were compared with control for continuous observation for 2 to 3 days. The titer of the virus was determined by Reed-Muench method and the growth curves of H5N1 and H9N2 AIVs were plotted respectively.

Statistical Analysis
Statistical analyses were performed using SPSS software (version 20.0), and figures were obtained using GraphPad Prism (GraphPad Software, La Jolla, CA). One-way ANOVA or Student's t-test was used to determine statistical significance between samples. Statistically significant differences were obtained when p < 0.05.

Expression of Inflammatory Cytokines in DF-1 and MDCK Cells Infected by H9N2 and H5N1 AIVs
mRNA transcripts for inflammatory cytokines including IL-1β, IL-6, and TNF-α increased significantly in DF-1 and MDCK cells infected by H5N1 and H9N2 AIVs. In DF-1 cells, IL-1β, IL-6, and TNF-α increased significantly from 6 h after infection with both AIVs (p < 0.01), but higher expression levels were observed in response to H5N1 than H9N2 infection ( Fig. 1A-C). This is consistent with the expression trend of inflammatory cytokines in DF-1 cells in our previous study [5]. In MDCK cells, IL-1β expression increased significantly from 6 h with H9N2 and H5N1 infection (p < 0.01), with higher expression in response to H9N2 AIV than H5N1 (Fig. 1D). IL-6 expression increased significantly from 3 to 24 h following both H9N2 and H5N1 infection (p < 0.05). Between 3 and 12 h post-infection, H5N1 AIV elicited significantly higher IL-6 expression than H9N2 infection (Fig. 1E). TNF-α expression increased significantly from 3 to 24 h after infection with H9N2 AIV (p < 0.01), and from 3 to 12 h after infection with H5N1 (p < 0.01). Similar to other tested cytokines, H5N1 infection caused a higher change in TNF-α expression than H9N2 infection (Fig. 1F).

Expression of PTX3, Del-1, and GDF-15 in DF-1 and MDCK Cells Infected by H9N2 and H5N1 AIVs
Upon H9N2 and H5N1 AIV infection in DF-1 cells, PTX3, Del-1, and GDF-15 expression levels increased significantly. In DF-1 and MDCK cells, PTX3 expression levels peaked at 12 h after infection by H9N2 AIV (p < 0.01), while expression levels in response to H5N1 infection steadily increased until study endpoint (p < 0.05; Fig. 2A, D). In response to H9N2 AIV infection, Del-1 expression increased and remained stable from 3 to 12 h (p < 0.05), peaking at 24 h in both cell lines (p < 0.01). H5N1 infection increased significantly Del-1 expression levels, which remained relatively stable from 3 to 24 h (p < 0.01). However, the elicited change in expression was not as robust as that observed in the H9N2 infection group (Fig. 2B, E). GDF-15 analysis revealed that  Fig. 2C, F). Altogether, these findings show that H9N2 and H5N1 AIVs infection in DF-1 and MDCK cells led to increased PTX3, Del-1, and GDF-15, and that the change in expression due to H9N2 is greater than that due to H5N1.

Effects of Overexpression of PTX3, Del-1, and GDF-15 on Proliferation of H9N2 and H5N1 AIVs
The effects of overexpression of PTX3, Del-1, and GDF-15 on proliferation of H9N2 and H5N1 AIVs in DF-1 cells were measured at different time points. For H9N2 AIV, the virus titer of PTX3, Del-1, and GDF-15 overexpression group decreased significantly from 6 to 24 h compared with the control (p < 0.05; Fig. 9A). For H5N1 AIV, the virus titer was significantly reduced from 6 to 12 h in PTX3 overexpression group (p < 0.01; Fig. 9B), and from 6 to 24 h in Del-1 and GDF-15 overexpression group (p < 0.05; Fig. 9B). GDF-15 showed the strongest inhibitory effect on the replication of both H9N2 and the H5N1 avian influenza viruses (Fig. 9).

DISCUSSION
It has demonstrated that inflammatory cytokines and chemokines are critical to the pathogenesis of avian influenza [36]. In the early stage of avian influenza virus infection, the pattern recognition receptors will sense the invasion of viral RNA and initiate innate immune response resulting in production of inflammatory cytokines and chemokines [37,38]. These inflammatory factors regulate the migration and chemotaxis of a variety of leukocytes to the site of infection and leukocyte Fig. 4 Expression profile for IL-1β, IL-6, and TNF-α in DF-1 and MDCK cells with pair expression of PTX3, Del-1, and GDF-15 infected with H9N2 AIV. * p < 0.05; ** p < 0.01. Data are presented as mean ± standard deviation (SD). ◂ Fig. 5 Dynamic expression of IL-1β, IL-6, and TNF-α in DF-1 and MDCK cells overexpressing PTX3, Del-1, and GDF-15 simultaneously infected with H9N2 AIV. * p < 0.05; ** p < 0.01. Data are presented as mean ± standard deviation (SD). exudation causes severe inflammatory pathological damage [39,40].
Previous studies have shown that different clinical manifestations arise upon infection of different influenza viruses in the same host, or identical influenza viruses in different hosts [41,42]. The underlying reason for this is differences in host inflammatory responses. The inhibitory effects of negative regulatory factors, PTX3, Del-1, and GDF-15, play an important role in inflammation development [8]. PTX3, also known as the "antibody precursor," is a humoral pattern recognition receptor that recognizes foreign microorganisms and acts as an opsonin to regulate inflammatory responses [43]. PTX3 is produced by phagocytes and stored by neutrophil granulocytes. After stimulation, PTX3 can bind to antigens and be recognized by macrophages or neutrophil Fcγ receptor IIa (FcγRIIa/CD32) and complement receptor 3 (CD11b/CD18), thus enhancing its phagocytosis activity and promoting innate immunity [44]. PTX3-deficient mice showed higher levels of neutrophil accumulation in pleural inflammation and acid-induced acute lung injury; bone marrow chimerism experiments showed that PTX3, a hematopoietic source, mediated the inhibitory effect of leukocyte rolling [45]. Consistent with the above findings, PTX3 knockout mice showed increased neutrophil infiltration and more severe myocardial necrosis injury in the cardiac ischemia-reperfusion injury model [45]. In addition, PTX3 has been proven to be an effective inhibitor of influenza virus. The surface of PTX3 structure has salivated ligand, which can mimic the structure of cell receptor used by influenza virus, thus blocking the receptor binding site of HA, and playing a series of antiinfluenza virus responses, including neutralization of virus, inhibition of hemagglutination, and inhibition of viral neuraminidase [46]. Del-1 is a secreted multi-domain protein that interacts with integrins and phospholipids and modulates different phases of the host inflammatory response depending on its expression location [47]. Due to its anti-inflammatory properties, Del-1 can prevent a variety of inflammation-related diseases, such as autoimmune encephalitis, inflammatory bone loss, lung inflammation and fibrosis, and inflammation associated with islet transplantation [47]. Del-1 has been shown to be an inhibitor of IL-17 which could mediate the recruitment of neutrophils. In periodontitis models, Del-1 inhibits the recruitment of neutrophils and the associated inflammatory response to mediate the pathological response by inhibiting the expression of IL-17 [48]. Interestingly, there is a reciprocal relationship between Del-1 and IL-17 in inflammatory tissues. IL-17 can down-regulate the expression of Del-1 in endothelial cells by regulating an important transcription factor C/EBPβ [48]. GDF-15 is a stress response protein belonging to the transforming growth factor-β superfamily. Under normal physiological conditions, GDF-15 is not expressed in many tissues, but under stress conditions such as hypoxia, inflammation, or pressure load, its expression in tissues, especially the heart, is significantly increased [49]. GDF-15 improves the inflammatory response of the heart after myocardial infarction by inhibiting leukocyte recruitment, thereby reducing infarct size; in contrast, mortality increased after myocardial rupture and infarction in GDF-15 deficient mice [50]. In addition, GDF-15 can inhibit the transendothelial migration of leukocytes and protect the myocardium from ischemia-reperfusion injury [13]. In addition to its anti-inflammatory effects, GDF-15 is closely related to the occurrence and development of tumors [51]. However, the precise roles of PTX3, Del-1, and GDF-15 in AIV-mediated inflammation in different hosts have not been verified in detail.
Different influenza viruses cause different inflammatory reactions. Species-specific differences in inflammation between mammals and birds in response to influenza viruses may be related to negative regulatory mechanisms [42]. To evaluate the role of negative inflammation regulators across different species, the response was analyzed in DF-1 and MDCK cells, which are avian and mammal cells, respectively.
Both H5N1 and H9N2 AIVs infection in DF-1 and MDCK cells caused a significant increase in the expression of PTX3, Del-1, and GDF-15, implicating these factors are in regulating inflammation. A series of experiments showed that PTX3, Del-1, and GDF-15 inhibit the expression of the pro-inflammatory cytokines, IL-1β, IL-6, and TNF-α, which are notably also increased in response to infection by various AIV subtypes. Moreover, many of the damaging pathological effects in the acute inflammation phase are caused by the combined actions of these cytokines [52,53]. Since IL-1β and IL-6 are specific to H9N2 AIV infection, and IL-1β and TNF-a are specific responses to H5N1 AIV infection [5], these cytokines were selected in this study. Fig. 6 Expression of IL-1β, IL-6, and TNF-α in DF-1 and MDCK cells overexpressing PTX3, Del-1, and GDF-15 alone that are infected with H5N1 AIV alone. * p < 0.05; ** p < 0.01. Data are presented as mean ± standard deviation (SD).

◂
For DF-1 cells, when PTX3, Del-1, and GDF-15 were individually overexpressed, the inhibitory effect on the expression of H9N2-induced inflammatory factors was strong, while, when GDF-15 was overexpressed, there was no inhibitory effect on the expression of H9N2-induced inflammatory factors. The overexpression of PTX3, Del-1, and GDF-15 all induced strong inhibition of IL-1β and IL-6 in the H5N1 infected group, while the inhibition of TNF-α was not obvious in the group with Del-1 overexpression, GDF-15 overexpression, and Del-1 and GDF-15 combined overexpression. Therefore, it is speculated that PTX3 may play the most important role in inhibiting inflammation during H5N1 infection. For MDCK cells, the overexpression of PTX3, Del-1, and GDF-15 can cause the inhibition of inflammatory cytokines in the H5N1 and H9N2 infected groups starting from 6 h. Whether it is MDCK cells or DF-1 cells, co-expression of PTX3, Del-1, and GDF-15 enhanced the inhibition. These findings suggest that initiating PTX3, Del-1, or GDF-15 could effectively inhibit the inflammatory response. Therefore, activators for the above genes would be plausible candidates to suppress AIV-mediated inflammatory responses. Overexpression of PTX3, Del-1, and GDF-15 in DF-1 cells inhibited the proliferation of H9N2 and H5N1 AIVs. The influence of PTX-3, Del-1, or GDF-15 on the proliferation of influenza virus and its possible mechanism need further research.
In summary, PTX3, Del-1, and GDF-15 attenuated the expression of inflammatory cytokines in DF-1 and MDCK cells infected H9N2 or H5N1 AIV. This partly explains the species-specific differences in clinical symptoms in response to infection by different AIV subtypes. The findings presented here contribute to the foundational knowledge that explains differences in pathogenesis by different AIV subtypes in identical or unique hosts. Additionally, given the unique cytokine expression profiles observed due to H9N2 and H5N1, the results from our study have clinical implication, as PTX3, Del-1, and GDF-15 may be used to screen for host-specific anti-inflammatory agents.

AVAILABILITY OF DATA AND MATERIALS
All data and materials included in this study are available on request to the corresponding author. Fig. 9 Inhibitory effect of overexpression of PTX3, Del-1, and GDF-15 on proliferation of H9N2 and H5N1 AIVs. * p < 0.05; ** p < 0.01. Data are presented as mean ± standard deviation (SD).