Genetic Diversity of Avian Influenza Viruses Detected in Waterbirds in Northeast Italy Using Two Different Sampling Strategies

Abstract

Avian influenza viruses (AIVs), which circulate endemically in wild aquatic birds, pose a significant threat to poultry and raise concerns for their zoonotic potential. From August 2021 to April 2022, a multi-site cross-sectional study involving active AIV epidemiological monitoring was conducted in wetlands of the Emilia-Romagna region, northern Italy, adjacent to densely populated poultry areas. A total of 129 cloacal swab samples (CSs) and 407 avian faecal droppings samples (FDs) were collected, with 7 CSs (5.4%) and 4 FDs (1%) testing positive for the AIV matrix gene through rRT-PCR. A COI-barcoding protocol was applied to recognize the species of origin of AIV-positive FDs. Multiple low-pathogenic AIV subtypes were identified, and five of these were isolated, including an H5N3, an H1N1, and three H9N2 in wild ducks. Following whole-genome sequencing, phylogenetic analyses of the hereby obtained strains showed close genetic relationships with AIVs detected in countries along the Black Sea/Mediterranean migratory flyway. Notably, none of the analyzed gene segments were genetically related to HPAI H5N1 viruses of clade 2.3.4.4b isolated from Italian poultry during the concurrent 2021-2022 epidemic. Overall, the detected AIV genetic diversity emphasizes the necessity for ongoing monitoring in wild hosts using diverse sampling strategies and whole-genome sequencing.

Keywords: avian faecal droppings; avian influenza; cloacal swabs; wetlands; whole-genome sequencing; wild birds.

Figure 1

Sampling sites in the Emilia Romagna region, northern Italy, selected for investigating the occurrence of AIVs in hunted or found dead wild birds and avian faecal droppings. Outbreaks in poultry (orange area and orange diamond) and cases in wild birds during the 2021–2022 AI epidemic are reported. Created with QGIS v.3.6.0.

Figure 2

Wild bird population sampled. (A) Bird taxa/number of faecal droppings collected in wetlands; (B) species/number of hunted or found dead wild birds included in the study according to sex; (C) species/number of hunted or found dead wild birds included in the study according to age classes.

Figure 3

Phylogenetic analysis of the H1 of the A/teal/Italy/1821-10_22VIR4622-1/2021(H1N1) isolate. The assembled full H1 sequence hereby obtained (red, bold) is compared with full coding sequences available at GenBank and/or GISAID. Human isolates of the pandemic influenza A(H1N1)2009 virus lineage and swine A(H1N1) viruses were included. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was GTR+F+I+G4.

Figure 4

Phylogenetic analysis of the H5 of the A/mallard/Italy/22VIR4203-2/2021 (H5N3) LPAI isolate. The assembled full H5 sequence hereby obtained (red, bold) is compared with full coding sequences available at GenBank and/or GISAID. The collapsed clade in the upper part of the phylogenetic tree shows the HPAIVs isolated in Europe during 2020–2023. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was GTR+F+I+G4.

Figure 5

Phylogenetic analysis of the H9 of the A/teal/Italy/1856-7_22VIR4622-7/2021(H9N2), A/teal/Italy/1828-6_22VIR4622-5/2021(H9N2), and A/teal/Italy/1821-14_22VIR4622-3/2021(H9N2) isolates. The assembled full H9 sequences hereby obtained (red, bold) are compared with full coding sequences available at GenBank and/or GISAID. Reference sequences are shown in italics. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was GTR+F+G4.

Figure 6

Phylogenetic analysis of the N1 of the A/teal/Italy/1821-10_22VIR4622-1/2021(H1N1) isolate. The assembled full N1 sequence hereby obtained (red, bold) is compared with full coding sequences available at GenBank and/or GISAID. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was TIM2+F+G4.

Figure 7

Phylogenetic analysis of the N3 of the A/mallard/Italy/22VIR4203-2/2021 (H5N3) LPAI isolate. The assembled full N3 sequence hereby obtained (red, bold) is compared with full coding sequences available at GenBank and/or GISAID. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was TIM2+F+G4.

Figure 8

Phylogenetic analysis of the N2 of the A/teal/Italy/1856-7_22VIR4622-7/2021(H9N2), A/teal/Italy/1828-6_22VIR4622-5/2021(H9N2), and A/teal/Italy/1821-14_22VIR4622-3/2021(H9N2). The assembled full N2 sequences hereby obtained (red, bold) are compared with full coding sequences available at GenBank and/or GISAID. Circles at nodes indicate the ultrafast bootstrap (UFBoot) score range and are labeled in white (UFBoot 50–69), grey (UFBoot 70–89), and black (UFBoot 90–100). The substitution model for the maximum likelihood phylogenetic reconstruction was TVM+F+G4.