Molecular characteristics of fowl adenovirus strains detected in broiler chickens on diets without immunostimulant supplements

Four Ross 308 broiler farms in the centre of the Silesia voivodeship of Poland were enrolled. The study indicated the same environmental conditions, normal diet feed supplier, water sources, supplying hatchery and husbandry conditions on each studied farm. The broilers on farm AO received antibiotics in drinking water and a normal diet. The chickens on farm P&AO were supplemented with probiotics, prebiotics, vitamins, microelements and antibiotics in their drinking water and ate a normal diet. Farm P chickens’ drinking water was dosed with probiotics, prebiotics, vitamins and microelements, and their diet was the normal one; and on farm C chickens were fed the normal diet and given drinking water to which neither antibiotics nor supplements were added. This farm was the control. The broilers were examined on the 1^st, 21^st and 42^nd days of life. The scheme of the supplementation with probiotic supplements and antibiotics was the same as that described by Tomczyk et al. (32).

This study was approved by and conducted in accordance with the guidelines of the Local Ethics Committee. All samples were collected under permission in accordance with the institutional guidelines of this committee.

The reference strain represented the type species FAdV-1/A and was provided by Charles River (Shrewsbury, MA, USA). It was used as a positive control in PCR studies.

Cultures of CEF were prepared from 9–11 day-old specific pathogen–free chicken embryos (Lohmann, Ankum, Germany) according to a standard protocol for research conducted in the Department of Poultry Diseases at the Polish National Veterinary Research Institute. Modified Eagle’s medium, 10% foetal bovine serum and 1% antibiotic mix (Antibiotic-Antimycotic; Gibco, Thermo-Fisher, Paisley, U.K.) were used. A monolayer of CEF culture was obtained after 24 h incubation at 37°C in an atmosphere of 5% CO₂.

Samples from internal organs of n = 60 birds per flock were taken from each of the four broiler farms. Liver, spleen, gizzard and caecal tonsil samples were collected aseptically during necropsy examinations. Twenty chickens were examined per flock on day 1, day 21 and day 42 of life. Homogenates prepared from internal organs as a 1:1 dilution in modified Eagle’s medium with an addition of 1% antibiotic mixture (Antibiotic-Antimycotic, Gibco) were passed through a 0.45 μm syringe filter (Minisart; Sartorius, Göttingen, Germany). Samples were preserved at −20°C for the next step of the studies.

Homogenates obtained from internal organs of experimental birds were used for CEF inoculation. Prepared CEF cultures were incubated at 37°C for 7 d in a 5% CO₂ atmosphere. The cytopathic effect characteristic for FAdV infection was observed and images of it were captured in daily microscopic examinations. The third and final passage of the examined strains was used for the analysis in the next step of the studies. DNA was isolated from the CEF cultures infected with the examined strains using a QIAamp Mini Kit (Qiagen, Hilden, Germany) as described in the manufacturer’s instructions. The adenovirus FAdV-1 reference strain was the positive control. The negative control was DNA isolated from uninfected CEF cultures which had been stored at −20°C.

The amplification of the HVR1–4 hypervariable regions of the L1 loop of the hexon gene of FAdV was carried out in a PCR as described by Niczyporuk et al. (17). The reaction was performed in duplicate with a 25 μL volume of final reaction mix.

The PCR amplicons were purified using NucleoSpin Extract II (Macherey-Nagel, Hoerdt, France) and Sanger sequenced by Genomed (Warsaw, Poland). Molecular analysis of the examined strain sequences was performed by the alignment of the nucleotide sequences of the amplified fragments originating from the hexon gene of two adenovirus field strains. The aligned sequences were compared with the twelve fowl adenovirus reference sequences obtained from the GenBank database sing the Basic Local Alignment Search Tool (BLASTn) from the National Centre for Biotechnology Information (NCBI) and MEGA version 11 software (31). Phylogenetic analyses were carried out and evolutionary associations of the nucleotides (nt) and amino acids (aa) sequences after the translation process were inferred and then the neighbour-joining method (23, 30) with 1,000 bootstrap replications was used. The nt and aa sequence similarities were calculated using the MegAlign module of the Clustal W algorithm included in MEGA 11.

The evolutionary distances in the phylogenetic tree were calculated by the p-distance method. The molecular studies included two adenovirus sequences from the field strains and twelve adenovirus reference sequences. All ambiguous, absent or incorrect nt and aa positions in the constructed alignments were removed from each sequence pair before analyses. The phylogeny was analysed in MEGA 11.

A pair-wise distance method was applied to compute the phylogenetic distance between a given group of different adenovirus type species. Analysis of distance estimation was determined with the maximum-composite-likelihood method with substitutions including the transition/transversion rate. The composite likelihood was the sum of related log likelihoods.

Twelve reference sequences were selected from the NCBI GenBank database for use in the analysis. These sequences’ accession numbers were AF339915 (in the case of the 2/D sequence), FJ360747 (the 11/D sequence), AF339921 (6/E), AF339918 (8a/E), AF339922 (7/E), AF339924 (10/C), AF339919 (5/B), KT862811 (8b/E), AF339916 (3/D), AF339923 (9/D), AF339917 (4/C) and MK050972 (1/A).

This test was applied to the two field adenovirus nucleotide sequences and twelve nucleotide sequences of adenovirus reference strains.

Fowl adenovirus strains were isolated from tissue specimens of 20 three-week-old Ross 308 broiler chickens from farm C that received neither probiotics, prebiotics, vitamins, microelements nor antibiotics. The virus strains were detected in the liver and intestines. No strains were detected in any tissue samples from farm P, where chickens received probiotics and prebiotics, farm P&AO where birds took probiotics, prebiotics and antibiotics, or farm AO where only antibiotics were administered (Table 1).

The chickens from farm AO showed clinical signs connected with liver and intestine damage with mortality during fattening. The most typical clinical signs of infection caused by adenoviruses were seen on farm C. The foremost sign was depression and apathy, and additionally ruffled feathers, a drooping head and a crouched position were evident in infected chickens. In some cases nervous signs of infection and digestive tract lesions were observed. In necropsies, anaemic combs and wattles and reduced weight gain were also noted.

The necropsy showed petechiae and ecchymosis in the skeletal muscles, and hepatomegaly with a pale brownish-to yellowish colour and fragile tissue consistency were observed. Splenomegaly was also indicated.

Two FAdV type species – FAdV1/A and FAdV-5/B – were detected in the clinical samples of examined birds. The hexon gene L1 loop HVR1–4 regions of the examined strains were submitted to GenBank with the accession numbers MW353018 of type 1/A and MW353019 of 5/B.

The pair wise distance of the examined nt sequences was indicated as 1.285. Pair-wise distances are presented in Table 2.

Estimated pair-wise distances and the related substitution parameters are presented in Table 3.

The codon usage in the HVR1–4 region of the sequences of the hexon gene of the two field and twelve reference sequences were studied. Cytosine was the most frequent nucleotide amongst adenovirus type species, comprising between 24.1% and 32.3% of nucleotides. Cytosine appeared most often in the first position of the codon in every sequence, and the percentage range was 15.6%–36.9%. It appeared in the second position of the codon between 30.5% and 34.3% of instances, whereas in reference strain sequences it did so between 29.7% and 33.4%. Guanine was more frequent in the third position of the codon in each reference strain sequence (36.1%–48.0%) than in field sequences (25.7%–27.6%). Results for adenovirus sequences of field type/species comparison to other reference strain sequences are presented in Table 4 as codon locations on the 1^st+2^nd+3^rd+ non-coding regions.

The relative synonymous codon usage values indicated the measure of gene expression level (16, 17, 21). The investigation of the synonymous codon usage in the examined region revealed differences in this region depending on the strain and adenovirus type species. The relative synonymous codon usage (RSCU) was estimated as the ratio of the indicated frequency of examined codons to the frequency expected in the synonymous codons for the same amino acids. This parameter is given following the codon frequency in Table 5. The average codon is equal to 1,303.

The estimated value of the shape parameter for the gamma distribution was 1.3314. All substitution patterns in examined regions and rates of substitution were calculated with the Tamura–Nei model (+G) (29). The model’s results are presented in Table 6. Each entry is the probability of substitution (r) from one base (row) to another (column). Rates of different substitutions which are transitional were indicated in bold. The substitutions are indicated in italics. Relative values named as instantaneous (r) should be considered as simplicity sum of (r) values that were made equal to 100. A gamma distribution was used to model phylogenetic rate differences among sites (5 categories (+G)). The mean evolutionary rates were 0.16, 0.44, 0.77, 1.23 and 2.49 substitutions per site. The nucleotide frequencies were 20.89% for A, 24.74% for T/U, 26.03% for C and 28.34% for G. For calculating maximum likelihood (ML) values, a tree topology was constructed. The maximum log likelihood for the analysis was estimated as −4,809.387. A total of 500 positions were included in the final round are presented in Table 6. The table facilitated the maximum-likelihood estimation of the parameters and the variance-covariance matrix in examined fowl adenovirus strain sequences presented different type species.

Models with the lowest Bayesian information criterion scores were used to describe the substitution pattern (10). The Akaike information criterion value corrected, maximum-likelihood value, and the number of parameters (including branch lengths) are presented in Table 6. Non-uniformity of calculated evolutionary rates among sites may be modelled by using a gamma distribution (+G) with five rate categories and certain fraction of sites is evolutionary analysis invariable (+I). Applicable estimates of gamma shape parameter and the estimated fraction of invariant sites were indicated and are presented. Assumed or designated values of transcription/transversion bias (R) are shown. In the analysis, the frequencies (f) and rates of base substitutions (r) for each examined nucleotide pair were calculated. Relative instantaneous (r) values could be considered the evaluation. For straightforwardness, the sum of (r) values was made equal to 1 for each examined model. For estimating ML values, a tree topology was calculated. The analysis was based on 14 nucleotide sequences.

The results of Tajima’s D neutrality test, the purpose of which was to identify sequences which do not fit the neutral theory model at equilibrium between mutation and genetic drift, are shown in Table 7.

The evolutionary comparisons on adenovirus sequences were created by using the neighbour-joining method by Saitou and Nei (23). The phylogenetic tree was created with the evaluation of branch length at 6.69858462 and was calculated by bootstrapping with 1,000 replicates. The phylogenetic tree was created with adequate branch lengths in the same units as those of the evolutionary distance used to estimate the phylogenetic tree. The evolutionary distances were calculated using the ML method (30) and are in the points of the number of substitutions per site. Sequence analysis HVR1–4 region of the L1 loop of the hexon gene was performed, and the obtained sequences MW353018-FAdV-1/A and MW353019-FAdV-5/B were found to be closely related. The first of these was closely related to the sequence of fowl adenovirus 1/A, MK050972, having 98% identity with it, and the second was closely related to the type species FAdV-5/B AF339919 with 99.8% identity (Fig. 1).

Fig 1.