Molecular Characterization of Isolated Chicken Infectious Anemia Viruses in Central China

Background: Chicken infectious anemia (CIA) is an immunosuppressive disease induced by the chicken infectious anemia virus (CIAV) causing heavy economic losses once outbreak. This study conducted a systematic analysis of the epidemiology and pathology of CIA in Henan province, central China. Methods: A total of 437 clinical tissue samples and 120 poultry disease-related live attenuated vaccines were collected during 2017-2019; of which 45 were positive for CIAV nucleic acid, with a positive rate of 8.08%. Results: Our results showed that genome sequences similarity among a total of 12 CIAV isolates was high, and ranged from 97.1% to 99.3%, and their similarity to the vaccine strains Cux-1 and Del-Ros ranged from 97.8% to 98.6%. However, There were non-synonymous amino acid mutations in the locus of the major capsid proteins VP1, VP2 and VP3 among all isolates. The subsequent sequence analysis indicated that the isolates of HN-4 and HN-8 showed genetic recombination and follow up animal experiments revealed that HN-4 might be a high pathogenic strain. Conclusions: Our results reveal that both eld infection and vaccine contamination promote epidemiology of CIAV in China. Some dominant epidemic viruses have undergone recombination and evolution. This study provides important information for CIAV prevention and control in poultry industry development. ORF: PBS:

neglected in clinical settings [10]. However, the latent infection caused chicken immunosuppression, which enhanced the invasion of other opportunist pathogens, including bacteria and viruses, which lead to disorders of chicken growth and development [11]. Signi cantly, CIAV can transmit both horizontally and vertically. Horizontal transmission occurs in the ock usually from the feces, danders or feathers, mainly through oral infection. More importantly however, vertical infection is caused from breeding eggs, which can easily result in an outbreak of CIA in the offspring. Furthermore, previous studies have highlighted the fact that vertical transmission still exists, even if there is an immune response in infected hens [12].
CIAV infection has spread globally causing severe economic losses. For example, 32 CIAV sequences were identi ed in 2010 in South Korea [13]. In India, 351 serum samples (86.88%) were positive in 404 serum samples collected from chicken ocks in 11 poultry farms [14], and in northern Vietnam, there were 74 (62.2%) positive tests in 119 samples collected from 64 farms [15]. In China, CIAV has been detected in the poultry farms of most provinces, and has been successfully isolated both from live vaccines and speci c pathogen free (SPF) chickens [16], which may explain the reason for the CIAV contamination found in the eld. Interestingly, co-infection of four immunosuppressive pathogens including CIAV has been detected in Shandong Province [17,18]. Recently, 1,187 clinical samples from major poultry farms nationwide from 2017 to 2019, have been tested for co-infection with six immunosuppressive pathogens including CIAV [19]. These surveys have con rmed the widespread nature of CIAV in China, which has seriously affected the development of the poultry industry. However, there is a paucity of information relating to CIAV infection in the central region of China. Therefore, an epidemiological investigation of CIA was carried out on samples collected from chicken farms and live attenuated vaccines in the markets of Henan Province, and the genomes of CIAV epidemic strains in this region were systematically analyzed. Furthermore, the pathogenesis of one recombinant strain has been investigated further.

Sample collection
The tissue samples were collected from chicken ocks produced throughout Henan province, China, by the Poultry Disease Research Institute of Henan Agricultural University. In brief, tissues from suspected chickens containing CIAV were cut into small pieces and place into a sterile tube. Samples were then homogenized three times in phosphate buffer saline (PBS) using small steel balls in an automated tissue grinder. Next, the samples were freeze/thawed three times to encourage the release of the virus. The viral supernatant was isolated by centrifugation at 8000 rpm and then transferred into a new sterile tube. A total of 120 non-CIAV live attenuated vaccines were collected from 24 manufacturers and ve different batches for CIAV detection.

Virus ampli cation
Marek's disease lymphoblastoid cell (MSB1) (A kind gift from Professor Jun Ji, Nanyang Normal University, Henan province) were cultured in 1640 culture medium with 10% fetal bovine serum (Solarbio, Beijing, China). When the cell con uency reached 5×10 5 cells/cm 3 , they were centrifuged at 1000 rpm for 5 min and washed twice with sterile PBS to remove residual serum. Then the CIAV containing supernatant was added for virus adsorption for 1 h. After this step, MSB1 cells were cultured in 5 mL of 1640 medium containing 2% fetal bovine serum and When visible cytopathic changes could be seen, the infected cells were passaged blind ve times on MSB1 cells, for viral ampli cation.

NucleicAcid extraction and CIAV detection
Genomic DNA was extracted from the chicken samples using a commercial kit (TIANamp Genomic DNA Kit, Tiangen, Beijing, China) and used according to the manufacturer's instructions. Three pairs of primers for ampli cation of the CIAV genome were designed and synthesized by Primer Premier 5.0 (Sangon Biotech, Shanghai, China). The rst primer pair were as follows: forward primer 5'-GCATTCCGAGTGGTTACTATTCC-3', reverse primer 5'-CGTCTTGCCATCTTACAGTCTTAT-3'; and these primers were expected to produce a predicted amplicon size of 842 bp. The second primer pair were as follows: forward primer 5'-CGAGTACAGGGTAAGCGAGCTAAA-3', reverse primer 5'-TGCTATTCATGCAGCGGACTT-3'; and these primers produced a predicted amplicon size of 990 bp. The nal set of primers were as follows: forward primer 5'-GAAAATGAGACCCGACGAGCAACAG-3', reverse primer 5'-GATTCGTCCATCTTGACTTTCTGTG-3', and these primers produced a predicted amplicon size of 736 bp. The polymerase chain reaction (PCR) reaction was performed in a total volume of 50 μL, containing 18 μL of 2 x Taq Master Mix, 1 μL (10 μm/L) of each primer, 25 μL distilled water, and 3 μL template DNA.
The PCR cycling conditions for ampli cation were as follows: 95°C pre-denaturation for 5 min; 95°C denaturation for 30 s, 60°C annealing for 30 s, and 72°C extension for 30 s. Ampli cation was performed over 35 cycles with a nal extension at 72°C for 10 min. The PCR products were then resolved using 1% agarose gel electrophoresis, included both positive and negative controls.

Sequencing analysis
The PCR products were puri ed with a PCR gel recovery kit and cloned into PMD18-T vectors and sent to Sangon Biotech for sequencing, and then aligned using the BLSAT program on the National Center for Biotechnology Information (NCBI) website (www.ncbi.nlm.nih.gov) to reference strains (Table 1). Next, the MEGA 6 and RDP4 softwares were used to analyze the sequences and recombination of the isolates, which incorporated RDP, GENECONV, BootScan, MaxChi, Chimera, SiScan, Phyl-Pro, LARD And 3Seq. SPF chickens were wing-banded upon hatching, and housed in isolators, and randomly sorted into experimental groups. An additional group was used as control. To determine the pathogenic properties of the NH-4 isolate, experimental groups from one-day-old SPF chickens were inoculated with 0.1 ml of HN-4 in the muscle tissue, whereas the chickens in the control group were inoculated with an identical volume of physiological saline. All chickens that had died during the experiment (21 days), or were euthanized at the end of the experiment, were necropsied and examined for CIAV-related pathology.
To evaluate the effect of HN-4 on body weight and lymphoid organs, three chickens in each group were euthanatized at 7, 14 and 21 days post-infection, and their body weight and lymphoid organ weights (thymus and bursa) were measured. The ratio of the lymphoid organs was expressed as the weight of lymphoid organ divided by the body weight for each chicken, multiplied by one hundred.
For hematoxylin-eosin (H&E) staining, tissues were formalin xed and para n embedded and 6-8 µmthick sections were prepared. H&E staining was carried out following regular procedure and sections were examined under a microscope.

Data analysis
The ratio of the lymphoid organs and survival rates were analyzed using Graphpad Prism version 8.0.1 software (GraphPad Software, Inc. La Jolla, CA). Statistical t-test analysis was used for each data point, which represented a triplicate average and a value of P<0.05 was considered statistically signi cant.

Complete Gene Sequence Analysis of isolated CIAVs
A total of 41 samples were detected to be CIAV positive from 437 tissue samples collected from June 2017 to December 2019 in Henan province. The overall positive rate was 9.4% (95% CI: 6.8%-12.5%). Four CIAV positive vaccines in 120 non-CIAV live attenuated vaccines were detected, and the overall positive rate was 3.33% (95% CI: 0.9%-8.3%) ( Table 2). A total of 12 CIAV strains were isolated and identi ed by PCR from chicken samples and as shown in Fig. 1A, the amplicon sizes were 842 bp, 990 bp and 736 bp, respectively, and this was consistent with prediction.
Four strains of CIAV virus (HN-1, HN-8, HN-9, HN-12) had a similarity between 98.1% and 99.7%, of which, HN-1 and HN-9 had the highest similarity. These four isolates are more closely related to the American strain L14767.1, with a similarity of over 98%. Compared to the vaccine strains that have been widely used worldwide, the 12 isolates had high sequence similarity with the Del-Ros vaccine strain, and ranged from 97.8% to 98.6%, and among which, the HN-11 isolate had the highest similarity with the vaccine strain. The similarity between the isolated strains and the Cux-1 vaccine strain worldwide was less than 98% (Fig. S1).
A phylogenetic tree, represented by two large branches, was constructed based on the complete genome sequence of 44 reference strains and 12 CIAV isolates from Henan. All of the Henan isolates were on the same large branch, which was closely related to the CIAV strains derived from Asia. After subdivision, the Henan CIAV isolate became located to three different small branches. The four strains HN-3, HN-5, HN-7, and HN-10 however, were on the same branch as the CIAV strains isolated from Shandong and Liaoning. The HN-11 and the Guangdong isolated strain GD-1-12 were on the same small branch. Both the HN-11 and the Guangdong strain, GD-1-12 evolved from the classical attenuated strain C369, isolated from Japan, and the attenuated strain, CIAVV89-69 isolated from Korea. However, the four strains HN-3, HN-5, HN-7, and HN-10 and the CIAV strains isolated from Shandong and Liaoning were closely related to the classical attenuated strain C369 and CIAVV89-69. The three strains HN-2, HN-4 and HN-6 appear on the same branch as the HN1405 strain isolated from Shandong. HN-1, HN-8, HN-9, and HN-12 are on the same branch as the HN1504 strain. All of the CIAV strains isolated in Henan are far removed from the internationally prevalent vaccine strain Cux-1, and are not on the small branch with the Del-Ros vaccine strain (Fig. 1B).
Amino acid sequence analysis of the VP1 and VP2 proteins Amino acid sequence analysis based on the VP1 sequence (450 aa) showed that, the overall variation in VP1 protein from all of the isolates was 0-3.4%. The similarity in the VP1 amino acid sequence for the HN-1 and HN-12 strains reached 100%, and the sequence similarity with the American strain L14767.1 was very high (Fig. S2). When comparing the VP1 protein amino acid evolutionary tree ( Fig. 2A), and the nucleotide sequence evolutionary tree (Fig. 2B), we found that the VP1 nucleotide tree was roughly the same as the whole genome nucleotide evolutionary tree. When we compared the amino acid trees, we found that the three strains, HN-3, HN-4, and HN-7 were on the same branch as GD-1-12, while the HN-3 and HN-7 isolates were on the same branch as the HN-5, and HN-10 isolates. Finally, the HN-4 isolate was on the same branch as the HN-2, and HN-6 isolates. Therefore, it can be inferred that there may exist site mutations, or sequence recombinations in these strains at the amino acid level. HN-1, HN-8, HN-9, and HN-12 still represent a large branch, which is not different from the phylogenetic tree at the nucleotide level.
VP2 is the most conserved CIAV protein and there are no signi cant differences at the amino acid level between the Henan isolates and the reference strain, and the co-e ciency of variation was less than 2.3% (Fig. S3). The VP2 nucleotide phylogenetic tree (Fig. 2C) and amino acid phylogenetic tree (Fig. 2D) displayed no obvious phylogenetic pattern. Among them, HN-1, HN-6, HN-7 and HN-8 had an obvious developmental trend within the VP2 amino acid evolutionary tree, which may be related to mutations at their amino acid sites.
From the results seen using similarity analyses, the VP3 protein from the Henan isolates and reference strain, are more conservative at the amino acid level with a difference co-e cient of 0-1.7% (Fig. S4). The nucleotide phylogenetic tree (Fig. 2E) and the amino acid phylogenetic tree (Fig. 2F) of VP3 had no obvious phylogenetic pattern. The HN-1 and HN-2 strains become extended in the amino acid phylogenetic tree, which may be due to the variation in some amino acid sites in these two isolates.
Overall, there were no signi cant differences in apoptotic proteins in CIAV from the Henan isolates and the other isolates.

Analysis of major amino acid sites of VP1, VP2 and VP3 proteins
Previous studies have con rmed that the hypervariable region in the VP1 sequence is at amino acid position 13, in which amino acids 139 and 144 have an effect on the replication rate and infection e ciency of the virus in infected cells. When the 139th and 144th amino acids of the strain were glutamine (Q), the proliferation rate was signi cantly slower. In this study, the amino acid sites for the VP1 protein in 12 isolates, main vaccine strains, classical attenuated strains, domestic virulent strains, and previously isolated strains from Henan were all analyzed. Among these, the 12 isolates from Henan, and HN-1, HN-8, HN-9, and HN-12) strains all carried Q139 and Q144, and the other eight CIAV isolates (HN-2, HN-3, HN-4, HN-5, HN-6, HN-7, HN-10, and HN-11) carried the amino acid lysine (K) at position 139 and glutamate (E) at position 144. All the viruses isolated from Henan contained Q, at position 394, suggesting that the isolates from Henan had high pathogenicity. The amino acid sites for the four isolates HN-1, HN-8, HN-9, and HN-12 were the same as those of the HN1504 strain on the VP1 protein, which further provides evidence that the four strains (HN-1, HN-8, HN However, whether this mutation in uences its pathogenicity, needs further investigation (Table 3). The amino acid sites of the VP2 and VP3 proteins were relatively conserved with only a few mutated amino acid positions. In VP2, ve alternative amino acid mutations were observed: glycine (G) 31 to E mutation at position 31 and alanine (A) 53 to V mutation at position 53 in the HN-1 isolate, a serine (S) 13 to arginine (R) mutation in HN-6 isolate, a lysine (K) 102 to E mutation in HN-7 isolate, and a G 24 to E mutation at position 24 in the HN-8 isolate. In VP3, a P 18 to S mutation was observed in HN-1 isolates, and an A 54 to G mutation was observed in HN-2 isolates (Table 4).

Recombination sequence analysis
Recombination analysis between all isolates and reference strains by RDP 4 software showed that recombination events occurred between different strains. Among them, HN-4 and HN-8 strains may be two potential recombinant strains. The isolate HN-4 may be a potential recombinant strain between the Korean isolate CIAVV89-69 and the Henan isolate HN-2. Bootscan analysis of the sequence of the HN-4 strain and its major and minor parental strains was conducted. The Korean strain CIAVV89-69 was found to be the major parental strain, while the Henan strain HN-2 represented the minor parental strain, and its recombination breakpoint was mapped to position 787 (initial breakpoint) and 1707 (termination breakpoint) (Fig. 3A).
Isolate HN-8 may represent a potential recombinant strain between the Brazilian isolate RS-BR-15 and the Henan isolate HN-6. Here, we used Bootscan analysis of the sequence of the HN-4 strain and its major and minor parental strains. We found that the Henan isolate HN-6 was the major parental strain, while the Brazilian isolate RS-BR-15 was the minor parental strain, and its recombination breakpoint was mapped to position 36 (initial breakpoint) and 2155 (termination breakpoint) (Fig. 3B).

Pathogenesis of HN-4
To examine the biological characteristics of the CIAV eld isolates, we conducted analyses of their pathogenicity by infecting one-day-old SPF chickens with the HN-4 strain. As expected, there were no pathogenic symptoms in the control group. In the experimental group, however, no death was seen from day 1 to day 13, but chickens showed depression, were somnolent, displayed abnormal development and had a signi cantly lower body weight (P<0.001) (Fig. 4A). Chickens in the experiment group died continuously 14 days post infection, with a total of 10 in 20 deaths seen up until the end of the experiment (21 days) (Fig. 4B). At necropsy, no obvious pathological changes were seen in the liver and kidney, but signi cant atrophy was seen in the spleen, bursa and thymus, when compared to the control group (Fig. 4C). The tibia of chickens infected with HN-4 was thin and fragile and had yellowish colored bone marrow (Fig. 4D).
To further evaluate the pathogenicity of the HN-4 strain, histological lesions were examined on day 16 in the bursa, thymus, liver, spleen, and kidney. Our ndings revealed that there was a signi cantly reduced number of lymphocytes in the Bursa, Thymus, spleen, and kidney. Whereas liver cells displayed metatropy when compared to the control group (Fig. 4E). These results revealed that HN-4 may be a highly pathogenic strain.

Discussion
CIAV infection of its natural host, the chicken, results in a low mortality rate, however, CIAV infection does cause important symptoms such as a poor immune response and even immune failure in its host, resulting in the invasion of other secondary pathogens, which can then induce a huge economic loss to the poultry industry [20]. Therefore, it is vital to perform epidemiological and pathological investigations to help control the spread of CIA. We performed a PCR survey of CIAV infection, which was carried out on 437 samples collected from September 2017 to December 2019, with a positive rate of 9.4% (95% CI: 6.8%-12.5%), con rming the prevalence of CIA in the eld, in the Henan province region of China.
In addition, the safety of a potential vaccine is closely related to the health of chickens [21]. Here, we investigated the extent of CIAV contamination in the live attenuated vaccines sold at market in Henan Province. A molecular epidemiological survey was conducted on vaccine samples from 24 manufacturers, with a total of 120 attenuated vaccines. The results showed that the overall positive rate was 3.3% (95%CI: 0.9%-8.3%), indicating that the necessity for biological products quality control. In the present study, CIAV was mainly tested in the Newcastle disease vaccine, bivalent live attenuated vaccine of Newcastle disease, Avian infectious bronchitis, and Fowlpox live vaccines. Possible CIAV contamination in other important avian diseases, also needs to be further investigated. Unfortunately, we did not isolate or identify live CIAV from these vaccines.
We did however, isolate and identify 12 CIAV strains from clinical chicken samples, and then analyzed their genome sequences. The nucleotide variations among all the isolates ranged from 0.75% to 3.9%, indicating little difference between these strains and reference strains. The high identity seen in the Henan isolates suggested that these strains are highly conserved. Phylogenetic results revealed that the 12 CIAV strains were in one large branch, most of which were Asian isolates. These isolates were further divided into three branches, where HN-3, HN-5, HN-7, HN-10, and HN-11 were all on the same branch as the Japanese classical attenuated strain C369 and the Korean attenuated strain CIAV 89-69 [22]. Among them, HN-11 was on the same small branch as Guangdong isolate GD-1-12 [23]. The three strains HN-2, HN-4 and HN-6 were on the same branch as the former Henan HN1405 strain, whereas the HN-1, HN-8, HN-9 and HN-12 strains were on the same branch as the former Henan HN1504 strain, indicating that these strains might be the dominant strains in Henan province. However, there were no similar branches seen for the international vaccine strains Cux-1 and Del-Ros [24]. These results indicate that the Henan isolates may be epidemic strains, after a natural recombination and evolutionary event in Asia.
Of the three intact proteins found in the isolates, VP1 protein had the most amino acid site variations. Previous studies have con rmed that higher virulence strains are seen when position 394 in VP1 protein is Q, which represents the main genetic determinant of CIAV virulence [25]. In this study, all isolates had Q at this position, indicating that the viral strains have strong virulence. HN-4, HN-6, and HN-8 have A 290 to P mutations at position 290, which has never been previously reported. However, VP2 and VP3 were relatively conserved, which may not affect their basic functions. These results suggested that the antigenic variation seen in VP1 may be important for virulence.
Recombination is the main source of mutation and the driving force for virus evolution [24]. We tested the 12 isolated CIAV strains and found that there were genetic recombination events in the HN-4 and HN-8 strains. Recombination occurred both in the coding region and the non-coding region, suggesting a higher level of adaptability and epidemic transmission. Importantly, HN-4 infection induced 50% of death and signi cant pathological symptoms after inoculation of one-day-old chickens in a 21 day long experiment.
These results suggest that HN-4 might be a highly virulent epidemic strain induced by recombination.

Conclusion
In conclusion, our study has focused on the isolation, identi cation and analysis of different CIAV strains prevalent in Henan province, and found that there were speci c amino acid site variations and gene recombinations. These recombination events may be responsible for the evolution of virulence in the epidemic strains, whose biological characteristics need further investigation.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. 2021412SupplementaryMaterials.docx