Co-infection of Chlamydia psittaci with H9N2, ORT and Aspergillus fumigatus contributes to severe pneumonia and high mortality in SPF chickens

Since 2007, most areas of China have seen outbreaks of poultry airsacculitis, which causes hugely economic losses to the poultry industry. However, there are no effective measures to combat the problem. In this study, 105 rations were collected to isolate Aspergillus spp. from the diseased farms. In subsequent experiments, SPF chickens were inoculated with Ornithobacterium rhinotracheale (ORT), Chlamydia psittaci (C. psittaci) and Aspergillus fumigatus (A. fumigatus), and mortality rate, body weight gain and lesion score were evaluated. Of these ration samples, 63 (60.0%) were A. fumigates, 21 (20.0%) were Aspergillus niger (A. niger) and 11 (10.5%) were Aspergillus candidus (A. candidus). Furthermore, SPF birds infected with C. psittaci, ORT, H9N2 virus and A. fumigatus conidia exhibited a mortality rate of 40%, while simultaneous co-infection with C. psittaci, ORT and A. fumigatus resulted in a mortality rate of 20%. The avian airsacculitis was manifested in the C. psittaci + ORT/A. fumigatus, C. psittaci + H9N2 + ORT/A. fumigatus and C. psittaci + H9N2/A. fumigatus groups while others had transient respiratory diseases without mortality. Our survey indicates that feed-borne A. fumigatus is prevalent in poultry rations. The combination of C. psittaci, ORT, H9N2 and A. fumigatus conidia contributes to the replication of avian airsacculitis by aggravating the severe damage to the air sacs and lungs of chickens.

In addition to C. psittaci infection, H9N2 AIV has been associated with respiratory distress in birds 13 . It has also demonstrated that the co-infection of C. psittaci and H9N2 AIV induces 35% mortality rate with severe pneumonia in SPF chickens 14 . H9N2 AIV has been circulating world-wide in multiple avian species and humans 15 . In a pilot study, ORT, H9N2 and Streptococcus zooepidemicus have been isolated from commercial birds 3,16 . In the broiler industry, it is very difficult to select a vaccine strain that can work as a potent vaccine in the face of widely circulating mutants of an avian influenza virus 17 . Failure of Mycoplasma Galliscepticum (MG), IBV immunization strategy, outbreaks of avian airsacculitis are still a huge threat to chicken health, although the inactivated H9N2 vaccine, attenuated MG vaccine and live vaccine against IBV have been implemented 3,14 .
In a recent case, A. fumigatus has been isolated and identified from broiler grower ration that is above the nutrient standards during an outbreak of avian airsacculitis in Northern China 18,19 . In addition, an aflatoxin-induced immunosuppression has been reported in concurrent infection with H9N2 avian influenza virus 20 . However, co-infection with four pathogens and their potential role in airsacculitis have not been recorded. Our current experiment in a specific pathogen free (SPF) chicken model constitutes an exploration into the pathogenesis of avian airsacculitis and disease control strategies in poultry industry.

Materials and Methods
Determination of Aspergillus spp. in poultry feeds. Broiler feeds were collected from poultry farms suffered with severe diarrhea during 2015 and 2016, including starter (34), grower (36) and finisher (35) feed formulations. The samples were sent to the laboratory at China Agricultural University, China and examined upon arrival. Five-hundred grams of each sample were collected from different locations within the same diet and then mixed thoroughly for the inoculation of Sabouraud dextrose agar (SDA) medium, which contains chloramphenicol (0.5%) (Oxoid, Beijing, China) 21 . Fumonisin B1, B2 and B3 were subsequently determined using UPLC/ ESI-MS/MS as previously described 22,23 . Identification of Aspergillus fumigatus by conventional PCR. The amplification was performed using specific primers 24 . All oligonucleotides used in this study were designed based on the relevant sequences of the internal transcribed spacer 1 (ITS1) ribosomal DNA. The primers were expected to amplify a fragment 521 bp in length, including the whole ITS1 region. Moreover, three oligonucleotide primers based on specific regions of ITS1 were designed for identifying A. fumigatus, A. flavus and A. niger. The A. fumigatus specific primers were made up of primers designated ASPU (5′-ACTACCGATTGAATGGCTCG-3′) and Af3r (5′-CATACTTTCAGAACAGCGTTCA-3′). The A. niger set was composed of ASPU and Ni1r (5′-ACGCTTTCAGACAGTGTTCG-3′), and the A. flavus set consisted of ASPU and Fl2r (5′-TTCACTAGATCAGACAGAGT-3′). However, the A. candidus specific primers were designed according to the sequence of the 10-deacetylbaccatin III 10-O-acetyltransferase (dbat) gene (NCBI accession number EU883596.3). The designed primers were 5′-GAAGCCGGAAGACCCTTTATAC-3′ and 5′-CCCACCCAAAGTCTACTTCATC-3′. The PCR was carried out in a final volume of 25 μl containing 12.5 μl 2 × Taq PCR MasterMix (Aidlab Biotechnologies Co., Ltd,Beijing,China), 2 μl DNA template, 0.2 μl of each primer, and ultra-pure water was added up to 25 μl. PCR conditions were as follows: one pre-denaturation cycle at 94 °C for 5 min, 25 cycles of denaturation at 94 °C for 1 min, annealing at 60 °C for 15 s, elongation at 72 °C for 15 s and one post-elongation cycle at 72 °C for 10 min. PCR products were visualized in 1.20% agarose gel stained with ethidium bromide under UV transillumination.
Animals and ethics statement. Eighty 10-day-old specific pathogen free (SPF) birds were procured from Beijing Merial Vital Bridge Laboratory Animal Co., Ltd and kept in negative-pressure isolators at China Agricultural University (Beijing, China). All animals were maintained, inoculated and euthanized according to the protocol approved by an ethical board at China Agricultural University on Institutional Animal Care and Use Committee (IACUC). This protocol follows humane protocols that minimize pain in the animals. Briefly, the   birds were observed 2 or 3 times per day for signs, such as severe depression (ruffled feathers and reluctance to move, not moving when prodded and/or severe respiratory distress) or extreme injuries unrelated to treatment. Animals displaying these symptoms were removed and euthanized. No animals were died of influenza or unrelated injuries for the entire duration of the study. All animals were euthanized at the end of study in a CO 2 chamber using 100% CO 2 at a flow rate of 10-30% of the Chamber Volume per Minute (CV/min). The animals were observed for the absence of breathing and lack of the heartbeat. The CO 2 flow was maintained for at least 1 min after respiratory arrest. After confirmation of death, an additional secondary physical euthanasia (i.e. cervical dislocation) was performed before tissue collection and carcass disposal.
Experimental infection of SPF chickens with C. psittaci, H9N2, ORT and A. fumigatus conidia. Specific pathogen free (SPF) White Leghorn chickens were randomly divided into 8 groups (10 birds per group). At the age of three weeks, group 1 animals were received simultaneous intratracheal inoculation (it) of 1 × 10 8.5 IFU C. psittaci, 1 × 10 8.0 CFU ORT and intranasal administration (in) of 1 × 10 8.0 EID 50 H9N2, and then inoculated with 1 × 10 8.0 CFU of the fungal spores per os (po). A second group of birds were inoculated with C. psittaci (it), H9N2 (in) and fungal conidia (po). Group 3 birds were given C. psittaci (it), ORT (in), and fungal spores (po) while group 4 birds were administered C. psittaci (it) and A. fumigatus conidia (po). Similarly, group 5 birds were inoculated with H9N2 (in) and then fed fungal spores (po), group 6 birds were administered ORT (in) and fungal conidia (po) and group 7 birds were only received A. fumigatus conidia (po). SPF chickens from group 8 were orally given physiological saline and served as the control group. All birds were observed daily, and the immune organ indices were determined at 5 and 10 days post-inoculation.

Lesion scores of air-sacs and lungs.
Air-sac and lung lesion scoring (Table 1 and Table 2) were determined as previously recorded 25,26 .
Flow cytometric analysis of T-lymphocyte subsets. The splenic lymphocytes were collected for the measurement of the CD4+/CD8+ ratio as previously described 27 . Briefly, 1 × 10 6 splenic lymphocytes were incubated with anti-chicken CD3-SPRD, CD4-FITC and CD8-RPE (Southern Biotech, USA) at 4 °C for 30 min. Subsequently, lymphocytes were washed 3 times with phosphate buffer saline containing 1% fetal bovine serum, then resuspended and analyzed by FacsCalibur and CellQuest software (Becton Dickinson, USA). Viable lymphocytes were calculated based on light scatter (forward and side scatter) characteristics, and 10,000 events were analyzed for positive staining with SPRD, FITC and RPE antibodies.
Statistical analysis. Data were analyzed using IBM SPSS Statistics 22 statistical software, and results were expressed as means ± SD. One-way ANOVA, and Duncan multiple comparison and two-tailed independent Student t tests were performed for comparison between groups. Multiple group analysis included the multiple comparison correction (Bonferroni) was conducted. Differences between groups with P values of <0.05 or <0.01 were considered to be statistically significant.   (Fig. 1). Obviously, A. fumigatus isolates accounted for significant statistical differences as compared to other Aspergillus species (P < 0.01). Interestingly, higher A. fumigatus conidia were determined both in the grower and finisher diets in comparison with those of the starter diets (P < 0.05).

Isolation and identification of Aspergillus fumigatus. Mycological examination of the 105 samples
indicated that the fungus grew rapidly at 25 °C, and the texture of colonies varied from wooly to cottony, to somewhat granular with smoky gray color. However, the color of very mature colonies turned to slate gray. The surface color was smoky gray at the beginning and became slate gray by aging ( Fig. 2A). Under microscopy, conidial heads were in the form of compact columns in an undisturbed culture. Conidiophores had smooth-walled, greenish, up to 300 µm long, and terminated in a dome (Fig. 2B). Based on general morphology, cultural and molecular characteristics, fungal spores isolated from grower diets were identified as A. fumigatus and amounted to 1.1 × 10 4 CFU/g, while the spores from the starter and finisher diets were 4.0 × 10 6 CFU/g and 1.0 × 10 5 CFU/g, respectively. The putative Aspergillus-specific primers were employed to identify Aspergillus spp. As we speculated, the product of 521 bp in length, including the whole ITS1 region was amplified by PCR from all tested samples. Furthermore, species-specific primer-sets for A. fumigatus, A. flavus, A. niger and A.candidus were tested. A. fumigatus was successfully identified by PCR reactions (Fig. 2C). Finally, The product of ITS1 region and the specificity of putatively species-specific primer-sets for A. fumigatus were successful amplified.
after PCR identification, the A. fumigatus strains were successfully isolated from the feed samples by single colony purification out of Aspergillus spp. isolates.
Effect of co-infection on relative weight gain. Within five days post infection (p.i.) with C. psittaci, H9N2, ORT and A. fumigatus, chickens exhibited a decreased appetite and poor activity compared to the healthy control group. The relative body weight gains in those animals infected with several pathogens were 66.4%, 72.2%, and 83.9% in the C. psittaci + H9N2 + ORT/A. fumigatus, C. psittaci + H9N2/A. fumigatus and C. psittaci + ORT/A. fumigatus groups, respectively. However, the relative weight gains were 82.7%, 85.4%, 96.4% and 88.6% in the C. psittaci/A. fumigatus, H9N2/A. fumigatus and ORT/A. fumigatus groups, correspondingly. Interestingly, birds infected with the C. psittaci + H9N2 + ORT/A. fumigatus or C. psittaci + H9N2/A. fumigatus exhibited a significant lower relative body weight gain as compared to those of the ORT/A. fumigatus or healthy control group (P < 0.01) (Fig. 3). C.psittaci, H9N2, ORT and A. fumigatus. In addition to the poor activity in the birds infected with C. psittaci, ORT, H9N2 and the fungal spores, 8 out of 10 chickens (80.0%) suffered from severe breathing difficulties. Subsequently, 4 birds died and the mortality rate  . Survival rates in the chicken inoculated with different pathogenic combinations Birds displayed 60.0% survival rates in the C. psittaci + H9N2 + ORT/A. fumigatus group as compared to 70% live birds in the C. psittaci + H9N2/A. fumigatus group and 80.0% live birds in the C. psittaci + ORT/A. fumigatus group, respectively. Statistical differences were found in the three and four infectious agent groups compared to the one or two infectious agent groups as well as the control group ( ** P < 0.01).

Morbidity and mortality post-infection with
was approximately 40%. Meanwhile, of the chickens inoculated with C. psittaci, H9N2 and A. fumigatus conidia, 5 birds (50%) suffered from typical respiratory distress on day 6 p.i., and the mortality rate was nearly 30.0%. Moreover, 2 chickens out of 10 (20.0%) were died due to the combined infection with C. psittaci, ORT and A. fumigatus conidia. In general, the birds infected with two pathogens as well as A. fumigatus exhibited typical pneumonia, decreased activity, and recovered thereafter. Intriguingly, transient breathing difficulty, decreased activity and appetite with no deaths were recorded in the C. psittaci/A. fumigatus, H9N2/A. fumigatus or A. fumigatus group during the observation period (Fig. 4).

Effect of co-infection on lesions of air-sacs and lungs.
At necropsy, severe hemorrhagic discharges in the tracheal and bronchial obstruction were observed in the C. psittaci + H9N2 + ORT/A. fumigatus, C. psittaci + H9N2/A. fumigatus or C. psittaci + ORT/A. fumigatus group in comparison with the C. psittaci/A. fumigatus, ORT/A. fumigatus or H9N2/A. fumigatus group. Moreover, fibrinous airsacculitis, pericarditis, peritonitis and haemorrhagic pneumonia were evident in the chickens co-infected with 3 or 4 microbial isolates. As for lesion score of air sacs, chickens inoculated with C. psittaci + H9N2 + ORT/A. fumigatus exhibited higher lesion score compared to C. psittaci/A. fumigatus or ORT/A. fumigatus on day 10 p.i. (P < 0.01). However, the H9N2/A. fumigatus and A. fumigatus groups displayed the lowest air-sac lesion scores compared to other groups (Fig. 5).
With respect to lung lesions, severe haemorrage and necrosis were noted in the C. psittaci + H9N2 + ORT/A. fumigatus, C. psittaci + H9N2/A. fumigatus or C. psittaci + ORT/A. fumigatus group when compared to those in the ORT/A. fumigatus or C. psittaci/A. fumigatus group (P < 0.01). However, lower lung lesions were observed in the A. fumigatus and H9N2/A. fumigatus groups (Fig. 5).

Effect of the co-infection on immune organ index.
As for the immune organ indices of the chickens with the above 3 pathogen inoculations, both spleen and bursa indices were substantially elevated compared to those with ORT/A. fumigatus or A. fumigatus alone.
Lately, the spleen, thymus as well as bursal indices were significantly reduced in the C. psittaci + H9N2 + ORT/A. fumigatus, C. psittaci + H9N2/A. fumigatus or C. psittaci/A. fumigatus group as compared to those of the H9N2/A. fumigatus or ORT/A. fumigatus group (P < 0.01). On day 10, the bursal index of C. psittaci/A. fumigatus group was lower than the H9N2/A. fumigatus or control group (P < 0.05). However, there were no significant differences in the immune organ index between the C. psittaci + H9N2 + ORT/A. fumigatus and C. psittaci + H9N2/A. fumigatus groups (Fig. 6).

Effect of co-infection on subsets of spleen lymphocytes.
At day 5 post-inoculation, a significantly higher CD4+/CD8+ ratio was detected in the C. psittaci + H9N2 + ORT/A. fumigatus or C. psittaci + H9N2/A. fumigatus group when compared to that of the ORT/A. fumigatus or H9N2/A. fumigatus group (P < 0.01). Additionally, the ratio of CD4+/CD8+ was significantly reduced in the C. psittaci + H9N2/A. fumigatus and C. psittaci/A. fumigatus groups compared to that of the H9N2/A. fumigatus or A. fumigatus group (Fig. 7).

Discussion
Our survey of chickens in the diseased farms indicated that a highly contamination of Aspergillus spp. was prevalent in the chicken rations. Of the rations sampled, 60.0%, 20% and 10.5% were positive for A. fumigatus, A. niger and A. candidus, respectively while A. fumigatus spores were highly reported both in the grower and finisher rations, suggesting a severe fungus-elicited high risk to the poultry health as well as a potential food safety problem for human beings. Interestingly, the dominant A. fumigatus might pose an immune suppression and a secondary microbial infection. Both a significant decrease in relative weight gain and a severe airsacculitis with 40% mortality rate were replicated in the birds inoculated with C. psittaci + H9N2 + ORT/A. fumigatus. Similarly, post-inoculation with 2 or 3 pathogens in SPF chickens could lead to the same symptoms with a milder degree. The respiratory signs and mortality rates were characterized as those seen in the clinical outbreak of seasonal avian pneumonia in broiler poultry under cold temperature. No deaths occurred in the birds infected with C. psittaci/A. fumigatus, H9N2/A. fumigatus and A. fumigatus conidia alone, and these chickens only suffered from transient respiratory distresses. In a previous record, no clinical signs and mortality rates were noted in SPF chickens whereas more than 50% of mortality rate was occurred in broilers by aerosol inoculation with A. fumigatus 28 . The results of the present study might suggest that co-infection with C. psittaci, H9N2, ORT and A. fumigatus are responsible for the current severe airsacculitis with high mortality rates in commercial broilers. It is the first time that clinical airsacculitis has been well replicated using the above clinical four pathogenic isolates in a SPF chicken model. A post-mortem examination implied that C. psittaci and ORT infection targeted both air sacs and lungs while H9N2 infection induced upper airway lesions at the junctions of the primary and secondary bronchi and lungs in comparison with the lung damage caused by an A. fumigatus infection. Based on the dominant lesions in the co-infection, C. psittaci infection might worsen avian airsacculitis by inducing immune suppression characterized by the decreased immune organ indices and ultimately lowered CD4+/CD8+ ratios 14 . After replacement of the grower ration with A. fumigatus-contaminated ration, the immune suppression was aggravated by the C. pisttaci, H9N2 and ORT infection, contributing to the 40% mortality rate. Relatedly, 10%-30% mortality rate was reported in broilers with C. psittaci infection 29 . Moreover, the birds developed upper respiratory disease and dyspnoea, and C. psittaci infection preceded the ORT infection 9 . During acute C. psittaci infection, aMPV infection might play a role in exacerbating the respiratory disease. However, no clear interaction could be established after aMPV infection in latently C. psittaci-infected SPF turkeys 30 . In our pilot survey, an average of 68.4% C. psittaci-specific positive antibodies were detected in the affected chickens, suggesting that highly infection with C. psittaci might contribute to an avian respiratory disease. In recent study, birds inoculated with C. psittaci/ H9N2 induced 37.5% mortality with typical breath difficulty, and immune suppression was identified in virulent C. psittaci-infected SPF chickens 14 .
In comparison with C. psittaci infection, A. fumigatus contamination is often trivial in the co-infection, particularly at the level of 1.1 × 10 4 CFU/g in broiler rations, and more than 2.3 × 10 4 CFU/g in turkey commercial feed 31 . A. fumigatus is the predominant mold pathogen of the immunosuppressed bird, characterized by reducing NDV antibody, phagocytic activity and reactive oxygen species (ROS) production 19 . Once the immune system is compromised, the conidia of A. fumigatus can germinate into hyphae and establish a focus of infection within the lung. In addition to A. fumigatus contamination, aflatoxins (AF) produced by Aspergillus flavus and Aspergillus parasiticus are common contaminants of feed raw materials 32 and previous studies have also reported the presence of aflatoxins in poultry tissues in Pakistan 33,34 . In our survey, A. flavus and A. parasiticus were isolated much less frequently than A. fumigatus from the chicken ration samples. Therefore, a study of the association between mycotoxin and feed-borne A. fumigatus is urgently needed.
Clinically, outbreaks of avian airsacculitis occur frequently in poultry flocks even after immunization with many vaccines, including inactivated H9N2 vaccine, attenuated MG, live IBV and NDV vaccine. This dilemma suggests the involvement of other pathogens during respiratory diseases. Our study implies that multiple infectious agents may contribute to respiratory diseases. We propose that C. psittaci triggers primary immune suppression and aggravates the compromised birds by the consistent feed-borne A. fumigatus, contributing to the secondary infection with H9N2 and ORT and severe respiratory distress with high mortality rates.
Our study is the first to report typical airsacculitis with a 40% mortality rate in SPF chickens infected with C. psittaci, ORT, AIV H9N2 and A. fumigatus. During co-infection, both C. psittaci and A. fumigatus induce suppression by the impairment of immune organs and Th1/Th2 imbalance, while the combination of H9N2 with ORT might play an exacerbating role in the respiratory disease by causing lung damage. Therefore, it is urgent to eradicate immune suppression caused by C. psittaci infection and feed-borne A. fumigatus. Furthermore, the combination of C. psittaci, ORT, H9N2 and A. fumigatus infection should be considered when developing prevention and treatment strategies for avian airsacculitis.