Reassortment of newly emergent clade 2.3.4.4b A(H5N1) highly pathogenic avian influenza A viruses in Bangladesh

ABSTRACT

Avian influenza active surveillance was conducted in Bangladesh from January 2022 to November 2023 in live-poultry markets (LPMs) and Tanguar Haor wetlands. The predominant viruses circulating in LPMs were low pathogenic avian influenza (LPAI) A(H9N2) and clade 2.3.2.1a highly pathogenic avian influenza (HPAI) A(H5N1) viruses. Non-H9N2 LPAIs were found at Tanguar Haor and at a lower prevalence in LPMs. Starting from June 2023, we detected novel genotypes of clade 2.3.4.4b A(H5N1) viruses from ducks in LPMs. The HA, NA, and M genes of these viruses are related to those of 2020 European clade 2.3.4.4b H5N1 viruses such as A/Eurasian Wigeon/Netherlands/1/2020 (Netherlands/1). However, analyses of the other five gene segments’ sequences identified three distinct genotypes (BD-G2, BD-G3, and BD-G4). BD-G2 viruses were closely related to the clade 2.3.4.4b H5N1 viruses that have been detected in Japan and nearby regions since November 2022. BD-G3 viruses were reassortants, with gene segments from other Eurasian LPAI viruses. BD-G4 viruses were similar to BD-G2 viruses, but their NS gene was accrued from contemporary Bangladeshi clade 2.3.2.1a A(H5N1) viruses. The ability of any of the clade 2.3.4.4b viruses to displace the long-entrenched 2.3.2.1a A(H5N1) viruses in Bangladesh is unknown.

Introduction

In recent years, clade 2.3.4.4b A(H5N1) highly pathogenic avian influenza A (HPAI) viruses have been wreaking havoc on poultry and wild birds and posing a threat to human health worldwide. Since emerging, the clade 2.3.4.4b A(H5N1) viruses (represented by A/Eurasian Wigeon/Netherlands/1/2020 (Netherlands/1), have spread to many countries in Europe, Africa, Asia, and the Americas [1–4] and have undergone multiple reassortment events with low pathogenic avian influenza A (LPAI) viruses [5,6]. In January 2021, an A(H5N1) virus related to the 2020–2021 European viruses was reported in West Africa [7] and subsequently in Southern African countries [8]. Since then, it has been persistently circulating in this geographic area and in West Africa, where it reassorted with A(H9N2) viruses of the zoonotic G1 lineage [9]. Since October 2021, different genotypes of clade 2.3.4.4b A(H5N1) viruses, some of them previously identified in Europe, have been detected in South and East Asia, including in China [5]. In December 2021, North America announced the detection of 2.3.4.4b A(H5N1) viruses that were related to those circulating in Northern Europe during the 2020–2021 epidemic season [1,10]. This introduction was followed by further reassortment events with North American LPAI viruses. Tian et al., described >35 genotypes of clade 2.3.4.4b A(H5N1) viruses worldwide, 10 of which were generated in North America [6].

Bangladesh is at the overlap of two major flyways for migratory birds: The Central Asian and East Asian-Australian flyways [11–14]. The northeastern part of Bangladesh consists of marshy seasonal wetlands, termed haors, where floodplains and tributaries receive surface runoff to form seasonal lakes. Haors provide abundant aquatic vegetation for migratory waterfowl to overwinter from across Europe and Central Asia [15,16]. Commercially raised ducks in these haors commonly scavenge for food during the day, thereby making frequent contact with migratory waterfowl. Hence, resident poultry are at high risk of acquiring avian influenza A infections and often contribute to the dispersal of the vast viral gene pool [17]. Since 2008, we have conducted active avian influenza A surveillance in live-poultry markets (LPMs) in or near Dhaka, the capital city of Bangladesh. We began wild migratory bird surveillance in Tanguar Haor, a wetland ecosystem, in 2015.

Although LPAI viruses in Bangladeshi LPMs are highly dominated by viruses of the H9N2 subtype, we often find other subtypes [18,19]. The A(H9N2) viruses in LPMs are highly homogeneous and belong to the G1 lineage [20]. HPAI A(H5) viruses have been in circulation in Bangladesh since 2008, with the predominant clades found being 2.2.2. and 2.3.2.1a. Since 2011, clade 2.3.2.1a A(H5N1) viruses have been circulating in Bangladeshi LPMs and have undergone reassortment with co-circulating LPAIs, including A(H9N2) viruses, to generate multiple genotypes of virus. However, all these genotypes, except the H5N1-R1 genotype of clade 2.3.2.1a, subsequently disappeared [21]. The H5N1-R1 genotype viruses were first detected in June 2015 and have been circulating in Bangladeshi LPMs as the dominant HPAI A(H5N1) virus since February 2018 [21]. In 2016, HPAI clade 2.3.4.4b A(H5N6) viruses were first identified in domestic waterfowl from peri-urban Bangladeshi LPMs. The HPAI-positive waterfowl were reared locally in the Netrokona and Sunamganj districts, located in the northern part of Bangladesh along the central Asian flyway [22].

From the beginning of our surveillance in 2015 through 2019, the AIVs isolated in the Tanguar Haor wetland area have been of Eurasian LPAI origin. In January 2020, we isolated clade 2.3.4.4.h A(H5N6) viruses from migratory birds and domestic duck in Tanguar Haor wetlands. Three weeks after initial detection in Tanguar Haor, this same virus was also isolated from a mallard duck with no clinical sign in an LPM in Dhaka [23]. Although the 2.3.4.4 h virus was detected in an LPM, it did not replace the clade 2.3.2.1a viruses in circulation, and thereafter, there have been no more reports of their detection in Bangladesh. In December 2021, A(H5N1) viruses of clade 2.3.4.4b were isolated from ducks on free-range farms in the Tanguar Haor wetland region [24]. Genetic analyses revealed that the 2021 viruses differed from those isolated in 2016 [22] and were genetically related to A/Eurasian Wigeon/Netherlands/1/2020 (Netherlands/1), like clade 2.3.4.4b A(H5N1) viruses, except for their PB2, which was closely related to contemporary Eurasian LPAI viruses circulating in the Novosibirsk region of Russia [24].

In this study, we describe our continued longitudinal avian influenza surveillance in Bangladesh from January 2022 through November 2023. Genetic analyses revealed that along with A(H9N2) viruses, other LPAI viruses, and clade 2.3.2.1a A(H5N1) viruses, multiple genotypes of the clade 2.3.4.4b A(H5N1) viruses were present in Bangladesh. Furthermore, we found evidence for reassortment between clade 2.3.2.1a and 2.3.4.4b A(H5N1) viruses.

Materials and methods

Sample collection

Oropharyngeal, cloacal, fecal, and water samples were collected from poultry and wild birds in two distinct locations: Tanguar Haor wetlands, located in Northeastern Bangladesh, and LPMs in or near Dhaka [24]. In Tanguar Haor, oropharyngeal and cloacal swabs were collected from domestic ducks on free-range duck farms; fresh fecal droppings were collected from mixed flocks of migratory waterfowl. In Dhaka, oropharyngeal and cloacal swabs were collected from chicken, duck, and quail in retail and wholesale markets, along with water samples from poultry cages. Samples were placed in isolation media (PBS supplemented with antibiotics and 50% glycerol) [25], stored in liquid nitrogen, and then shipped at 4°C to St. Jude Children’s Research Hospital, where samples were then processed, as previously described [18,21].

Sample screening and virus isolation

Viral RNA was extracted from swabs collected by using the MagMax™-96 AI/ND Viral RNA Isolation Kit (Applied Biosystems) on the KingFisher™ Flex Purification System (Thermo Scientific). All extracted samples were screened for avian influenza A via real-time reverse transcriptase PCR (rRT-PCR) with universal matrix (M) gene-specific primers [26]. IAV-positive samples from LPMs were further screened by rRT-PCR with H5-specific primers [27]. All H5-positive swabs from LPMs were inoculated into 10-day-old embryonated chicken eggs and incubated at 35°C for 48 h. For M-gene positive but H5-negative samples, all duck swabs and selected (∼15%, sorted by species, market and lower CT value) chicken and quail swabs were inoculated into 10-day-old embryonated chicken eggs and incubated at 35 °C for 72 h. For swabs originating from Tanguar Haor, samples were screened only for M-gene via rRT-PCR. Eggs were then chilled at 4 °C overnight before allantoic fluid was harvested to test for influenza virus by using the hemagglutination assay (HA). Hemagglutination assays were performed using 0.5% chicken erythrocytes according to the World Health Organization (WHO) protocol [25].

Genome sequencing and phylogenetic analysis

Viral RNA was extracted from HA-positive allantoic fluid (isolates) by using the RNeasy Mini Kit (Qiagen), and cDNA was synthesized by using the SuperScript™ IV First-Strand Synthesis System (Invitrogen). Multi-segment RT–PCR was performed per protocol by using Phusion® High-Fidelity DNA Polymerase (New England Biolabs) and previously described primers that target the conserved 5′ and 3′ ends of each gene segment, regardless of IAV subtype [28]. Library preparations for Next-Generation Sequencing on the Illumina Miseq platform were made by using 1 ng of cDNA with Illumina's Nextera XT DNA Sample Preparation kit according to the manufacturer's protocol. A paired-end sequencing approach (2 × 150) on the Miseq (Illumina) platform was used for amplicons; reads were subsequently demultiplexed, quality-trimmed, and filtered. Both de novo and consensus sequences were generated by using CLC Genomics Workbench version 21.0.1 (GIAGEN). All contigs were then aligned and analyzed by using BioEdit Sequence Alignment Editor version 7.25 [29].

For phylogenetic analysis, sequences other than those in this study were retrieved from the National Center for Biotechnology Information (NCBI) Influenza Virus Sequence Database [30] and the EpiFlu database of the Global Initiative on Sharing All Influenza Data (GISAID) [31]. The sequences were then aligned and trimmed to equal lengths. Phylogenetic relationships were inferred by using the Maximum Likelihood method and Tamura-Nei model [32]. The initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pair-wise distances estimated by using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with the greatest log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The reliability of phylogenetic inference at each branch node was estimated by performing the bootstrap method with 1000 replications. Evolutionary analyses were conducted in MEGA 7 [33].

Hemagglutination inhibition (HI) assays

HI assays were conducted as previously described [25]. Representative A(H5N1) and A(H9N2) viruses were selected for use in antigenic evaluation and characterization based on their HA genetic similarity. The viruses chosen to represent a range of genetic clusters of Bangladeshi viruses were compared antigenically to the current WHO-recommended A(H5) and A(H9N2) candidate vaccine viruses (CVVs) and other reference viruses [34].

Results

Avian influenza A surveillance in Tanguar Haor, Bangladesh

Oropharyngeal and cloacal samples were collected from free-range ducks, and fecal samples were collected from wild birds. During January, February, and March of 2022 and February and March of 2023, 500 (250 oropharyngeal and 250 cloacal) swabs from 250 free-range ducks (from eight different farms, total 2500 samples) and 500 fecal samples from migratory waterfowl (from 9 different locations, total 2500 samples) were collected from the Tanguar Haor area each month. Of these, 637 swab samples (25.5%) from ducks and 108 fecal samples (4.3%) from migratory waterfowl were IAV-positive (M-gene PCR positive). Most IAV-positive swab samples were obtained in the months of January 2022 (46.6%) and February 2023 (63.6%). Of these, 16 A(H7N7) and 18 A(H1N7) influenza viruses were isolated from ducks. In January 2022, migratory waterfowl fecal samples exhibited a relatively higher rate of IAV-positivity (19.6%), and eight A(H10N4) and eight A(H10N5) influenza A viruses were isolated. In all other months, IAV-positivity rates were low for both swab and fecal samples, and no viruses were isolated. Most swab samples were collected from asymptomatic ducks and all viruses isolated came from birds with no clinical sign. During this study period no HPAI viruses were isolated from Tanguar Haor.

Avian influenza A surveillance in Bangladeshi LPMs

Each collection month, we collected 160 samples (oropharyngeal, cloacal, and water-trough samples) from poultry and poultry cages (90 chicken, 45 ducks, and 25 quail) at 4 LPMs in or near Dhaka, Bangladesh. The study design was to collect samples from Bangladeshi LPMs every alternate month. Due to high COVID-19 activity and associated movement restrictions, however, this was not always possible, resulting in more irregular samplings (Figure 1). During the surveillance period, on average, 54% (292/540) of duck, 56% (601/1080) of chicken, and 17% (52/300) of quail samples were IAV-positive as determined by real-time PCR. A(H9N2) viruses were isolated every month sampled; HPAI A(H5N1) were isolated every month sampled except for June 2022 and May 2023 (Figure 1).

Figure 1.

Monthly virus isolation from Bangladeshi LPMs. Each month, 160 virologic samples were collected from poultry and poultry cages at 4 LPMs in or near Dhaka, Bangladesh. All M-positive duck samples and H5-positive samples from any species were inoculated in eggs. However, only about 15% of H5-negative but M-positive chicken and quail samples were inoculated in eggs.

A total of 91, 39, and three IAVs were isolated from duck, chicken, and quail, respectively. HPAI A(H5N1) and A(H9N2) viruses were the most commonly isolated and were predominantly isolated from ducks and chicken, respectively. All three viruses isolated from quail were A(H9N2). In addition to A(H9N2) viruses, one A(H4N6), three A(H6N2), two A(H10N7), and one A(H12N6) viruses were isolated from LPMs, all from ducks. All birds were asymptomatic at the time of sampling.

Virus sequencing

To understand the genotype of each virus, the 183 IAVs isolated during this study period were sequenced, seven of which were mixed (3-H5N1-2.3.2.1a/2.3.4.4b, 3-H5N1-2.3.2.1a/H9N2, and 1-H5N1-2.3.4.4b/H10N7) and were removed from further analysis. Furthermore, sequences from 8 samples were not of high enough quality to submit to GenBank, but subtype information could be ascertained. Therefore, complete sequences of 1290 gene segments from 168 isolates [H5N1-2.3.2.1a (n = 35), H5N1-2.3.4.4b (n = 41), H9N2 (n = 36), H1N7 (n = 18), H4N6 (n = 1), H6N2 (n = 3), H7N7 (n = 16), H10N4 (n = 7), H10N5 (n = 8), H10N7 (n = 2), and H12N5 (n = 1)] were submitted to GenBank, including full genomes of 128 isolates (Supplementary Table 1).

HA and NA phylogeny

All viruses isolated from Tanguar Haor during this study period were LPAIs. As discussed, the subtypes of viruses isolated in January 2022 included A(H7N7), A(H10N4), and A(H10N5). In February 2023, we isolated A(H1N7), a subtype isolated worldwide in low frequency; these viruses have been described in detail elsewhere [35]. H7 HA and N4, N5, and N7 NAs are all Eurasian lineages and are closely related to LPAI viruses isolated from Bangladesh, China, Japan, Korea, and Russia (Supplementary Figure S1 and S2). Both HA and NA of the A(H7N7) viruses isolated from Tanguar Haor area in January 2022 (Supplementary Figures S1C and S2F) were closely related to those of the A(H7N7) viruses isolated from the same area in December 2021 [24]. Surprisingly, the HA of the A(H10N4) and A(H10N5) viruses were of North American LPAI lineages and were closely related to that of A(H10Nx) viruses isolated from Korea in 2022 and from Korea, China, and Bangladesh in 2020 (Supplementary Figure S1E).

The predominant LPAIs circulating in LPMs were A(H9N2) viruses that were isolated mostly from chicken. The HA and NA genes of the A(H9N2) viruses clustered phylogenetically with those of A(H9N2) viruses isolated from Bangladeshi LPMs since 2020 (Supplementary Figures S1D and S2B). The sole A(H4N6) virus was isolated in August 2022 from an LPM in Dhaka. Unfortunately, we could not determine the complete HA sequence of this virus; however, the partial gene sequence was most similar to the HA from A/duck/Bangladesh/50268/2021(H4N6), which was isolated in October 2021 from a Bangladeshi LPM. Phylogenetic analysis indicated that the NA gene of this virus was also closely related to that of A/duck/Bangladesh/50268/2021(H4N6) (Supplementary Figure S2E). HA and NA genes of all other LPAI viruses isolated from Bangladeshi LPMs are of Eurasian lineages and are closely related to those of viruses isolated from Bangladesh, China, Japan, Korea, and Russia. (Supplementary Figures S1 and S2).

Out of total 76 Bangladeshi A(H5N1) viruses isolated and sequenced during this study period, the HA (Figure 2) and NA (Supplementary Figure S2A) genes of 35 viruses phylogenetically clustered with those of clade 2.3.2.1a Bangladeshi A(H5N1) viruses that have been isolated since 2020. More importantly, the HA and NA genes of the other 41 A(H5N1) viruses isolated from June through November 2023, clustered with those of clade 2.3.4.4b A(H5N1) viruses isolated worldwide since 2020 and were most closely related to those of clade 2.3.4.4b A(H5N1) viruses isolated from Japan, Vietnam, China, Russia, and Tanguar Haor (Figure 2 and Supplementary Figure S1A).

Figure 2.

Phylogenetic relationship of the HA genes of the HPAI H5N1 viruses isolated from Bangladeshi LPMs. The viruses identified in this study are shown in color [red, clade 2.3.2.1a A(H5N1); pink, clade 2.3.4.4b A(H5N1) viruses]. BD, Bangladesh; TH, Tanguar Haor; $\Leftarrow$, Reference antigens used for clade 2.3.4.4b viruses HI assays. The tree is rooted to A/goose/Guangdong/1/96(H5N1). Bootstrap values ≥70 are shown on branches.

Phylogeny of internal genes

The phylogenetic trees of all six internal genes of the viruses isolated during this study were analyzed. The PB2 genes from the A(H9N2) viruses all clustered together with A(H9N2) viruses isolated from Bangladeshi LPMs since 2019 (Figure 3, shown in green). The PB2 genes of all other LPAI viruses clustered with various groups of Eurasian LPAI viruses (Figure 3, shown in blue). The PB2 genes of the clade 2.3.2.1a A(H5N1) viruses all clustered together with PB2 genes of A(H5N1) viruses isolated from Bangladesh since 2019 (Figure 3, shown in red). The PB2 genes of clade 2.3.4.4b A(H5N1) viruses segregated into two distinct phylogenetic groups; neither of them contained the clade 2.3.4.4b A(H5N1) viruses that we identified in Tanguar Haor in December 2021 (Figure 3, BD-G1) [24]. One group clustered with PB2 genes of clade 2.3.4.4b A(H5N1) viruses that have been isolated from Japan and nearby countries (Figure 3, BD-G2 & 4 shown in pink) since November 2022. The other group (BD-G3), comprising two isolates, clustered with a separate group of PB2 genes of clade 2.3.4.4b A(H5N1) and other Eurasian LPAI viruses that were also isolated from Japan and nearby countries.

Figure 3.

Phylogenetic relationship of the PB2 genes of the viruses isolated in Bangladesh. The viruses identified in this study are shown in color [red, clade 2.3.2.1a A(H5N1); pink, clade 2.3.4.4b A(H5N1); green, A(H9N2); and blue, non-H9N2 LPAI viruses]. BD, Bangladesh; TH, Tanguar Haor. The tree is rooted to A/equine/Prague/1/1956(H7N7). Bootstrap values ≥70 are shown on branches.

Phylogenetic analysis of NS genes also exhibited clustering of A(H9N2) and clade 2.3.2.1a A(H5N1) viruses with respective viruses isolated previously from Bangladeshi LPMs. The NS genes of non-H9N2 LPAIs clustered in different groups of Eurasian LPAI viruses. Similar to the positioning of the PB2 gene, the NS gene of BD-G2 A(H5N1) viruses also clustered with clade 2.3.4.4b A(H5N1) viruses that have been isolated from Japan and nearby countries (Figure 4, BD-G2 shown in pink). The NS gene of BD-G3 A(H5N1) viruses clustered separately from that of BD-G2 viruses and was instead closely related to the NS genes of Eurasian LPAI viruses isolated from Korea and Bangladesh from 2020 through 2023 (Figure 4, BD-G3 shown in pink). Of note, out of the 41 isolates belonging to clade 2.3.4.4b A(H5N1), the NS genes of 14 viruses clustered with NS genes of Bangladeshi clade 2.3.2.1a A(H5N1) viruses (Figure 4, BD-G4 shown in pink), indicating reassortment of clade 2.3.4.4b and 2.3.2.1a viruses within Bangladesh (Supplementary Figure S3).

Figure 4.

Phylogenetic relationship of the NS genes of the viruses isolated in Bangladesh. The viruses identified in this study are shown in color [red, clade 2.3.2.1a A(H5N1); pink, clade 2.3.4.4b A(H5N1); green, A(H9N2); and blue, non-H9N2 LPAI viruses]. BD, Bangladesh; TH, Tanguar Haor. The tree is rooted to A/equine/Prague/1/1956(H7N7). Bootstrap values ≥70 are shown on branches.

Antigenic analysis in HI assays

Using post-infection ferret antiserum raised against World Health Organization reference viruses and viruses isolated previously from Bangladesh, HI assays were performed for selected viruses isolated. Clade 2.3.2.1a A(H5N1) and A(H9N2) viruses isolated during this study period remained antigenically homogeneous (within subtype) and closely related to viruses that have been isolated since 2019 in Bangladeshi LPMs (data not shown). Post-infection ferret anti-sera raised against the clade 2.3.4.4b candidate vaccine virus A/Astrakhan/3212/2020(H5N8) reacted with two- to four-fold lower HI titer with Bangladeshi 2.3.4.4b viruses than it did with homologous antigen, indicating some antigenic differences (Supplementary Table S2). Post-infection ferret anti-sera raised against the representative Bangladeshi clade 2.3.4.4b A(H5N1) virus previously isolated from Tanguar Haor A/duck/Bangladesh/51601/2021(H5N1) reacted with similar HI titers to Bangladeshi 2.3.4.4b viruses isolated from LPMs as it did to the homologous antigen (Supplementary Table S2).

Discussion

HPAI A(H5N1) viruses have been endemic in Bangladesh for many years. The continued circulation of these viruses with other AIVs in poultry has driven reassortment and subsequent generation of novel genotypes of virus. Furthermore, ducks raised in free-range duck farms in wetland areas have considerable contact with wild migratory birds and then with other poultry when moved to LPMs. Recently we described the detection of clade 2.3.4.4b A(H5N1) viruses to ducks raised in free-range duck farms in Tanguar Haor wetland areas of Bangladesh [24]. Here we describe the introduction of clade 2.3.4.4b A(H5N1) viruses to Bangladeshi LPMs and their subsequent reassortment with circulating clade 2.3.2.1a A(H5N1) viruses. Although reassortment is a common trait of recent 2.3.4.4b A(H5N1) viruses circulating worldwide, most reassortment events have occurred LPAI viruses in wild bird ecosystems. The impact of the 2.3.4.4b/2.3.2.1a reassortant A(H5N1) on bird populations and human health is undetermined.

The A(H9N2) and clade 2.3.2.1a A(H5N1) viruses isolated during this study period were highly homogeneous and similar to viruses detected previously [24]. Phylogenetic analysis of LPAI viruses from LPMs and wetland regions revealed that some were closely related to viruses previously circulating in Bangladesh, whereas others [e.g. A(H1N7)] may be new introductions to Bangladesh. Furthermore, different clustering patterns of individual internal genes indicate a continuous process of reassortment among LPAI viruses.

The clade 2.3.4.4b A(H5N1) viruses isolated from Bangladeshi LPMs from June through November 2023 differ genetically from the clade 2.3.4.4b A(H5N1) viruses isolated from Tanguar Haor in December 2021 (Figure 3, BD-G1) [24] and could be divided into three genotypes (Table 1). The numerically dominant A(H5N1) genotype in the LPMs, BD-G2, appeared to be a new introduction to Bangladesh and was most similar to viruses detected in Japan. The BD-G3 genotype A(H5N1) viruses were reassortants that maintained PB1, PA, and NP genes most similar to those of A/duck/Bangladesh/51601/2021(H5N1)-like viruses but with PB2 and NS genes from other Eurasian LPAI viruses. In September 2023, we identified clade 2.3.4.4b A(H5N1) BD-G2 viruses that had acquired the NS gene from Bangladeshi clade 2.3.2.1a H5N1 viruses, forming the BD-G4 genotype. The NS gene of these viruses (BD-G4) is closely related to that of contemporary Bangladeshi clade 2.3.2.1a A(H5N1) viruses; therefore, we assert that this reassortment happened recently in Bangladesh. Finally, using concatenated sequences of all eight gene segments (PB2-PB1-PA-HA-NP-NA-M1-NS1), phylogenetic analysis of whole genome of clade 2.3.4.4b-H5N1 viruses shows segregation of viruses in four different groups indicating four different genotypes (BD-G1, -G2, -G3, and -G4, Supplementary Figure S4).

Table 1.

Genotypes of Bangladeshi clade 2.3.4.4b H5N1 viruses.

Genotype	First detected	Isolated from	Gene segments similar to
			HA	NA	PB2	PB1	PA	NP	MP	NS
BD-G1	December, 2021	Tanguar Haor	Nds/1	Nds/1	Amur/31b	Nds/1	Nds/1	Nds/1	Nds/1	Nds/1
BD-G2	June, 2023	LPMs	Nds/1	Nds/1	Buryatia/89i	KNU2021-48	Ko/KNU-34	Aichi/231111	Nds/1	China/10-14
BD-G3	August, 2023	LPM	BD-G1	BD-G1	Ko/KNU17	BD-G1	BD-G1	BD-G1	BD-G1	Ko/KNU-14
BD-G4	September, 2023	LPM	BD-G2	BD-G2	BD-G2	BD-G2	BD-G2	BD-G2	BD-G2	BD-3.2.2.1a

Nds/1, A/Eurasian Wigeon/Netherlands/1/2020(H5N1, clade 2.3.4.4b)-like; BD/51601, A/duck/Bangladesh/51601/2021(H5N1, clade 2.3.4.4b)-like; Amur/31b, A/Common Teal/Amur region/31b/2019(H3N6)-like; Burtia/89i, A/Buryatia/89i/2019(H1N1)-like; KNU2021-48, A/Mallard/South Korea/KNU2021-48/2021(H7N7); Aichi/231111, A/duck/Aichi/231111/2021(H2N5)-like; China/10-14, A/duck/China/10-14/2022(H6N6)-like; Ko/KNU17, A/Spot-billed duck/Korea/KNU17/2022(H6N5)-like; Ko/KNU-34, A/spot billed duck/Korea/KNU-34/2022(H4N6); Ko/KNU-14, A/bean goose/Korea/KNU-14/2022(H6N1)-like; BD-3.2.2.1a, Contemporary Bangladeshi clade 2.3.2.1a H5N1.

After analyzing sequences of clade 2.3.4.4b A(H5N1) viruses isolated worldwide, from their first detection in October 2020 through March 2022, Cui et al. reported a total of 16 genotypes [5]. Tian et al. subsequently described 19 additional genotypes (total = 35), with 10 distinct genotypes isolated in North America by October 2022 [6]. Sixteen more genotypes of clade 2.3.4.4b A(H5N1) viruses were identified in South Korea between October 2022 and March 2023 [36]. Together, these studies highlight the propensity that the clade 2.3.4.4b viruses have shown for reassortment. The 2.3.4.4b A(H5N1) viruses have also displaced other clades of A(H5) viruses in many regions, a trend that also appears to be occurring in Bangladesh. Within three months of our first identification of clade 2.3.4.4b A(H5N1) viruses in an LPM, viruses of this clade were detected at higher proportions than the endemically established 2.3.2.1a viruses were. Whether the 2.3.4.4b viruses eliminate the 2.3.2.1a viruses from Bangladesh remains to be seen, but it appears possible based on the experiences of other countries. Of note, Bangladesh and neighboring countries are the only region where the 2.3.2.1a A(H5N1) viruses circulate.

The production of antigenically representative CVVs is an ongoing pandemic preparedness activity of the WHO. In general, we detected a slight reduction in reactivity of the Bangladeshi 2.3.4.4b A(H5N1) viruses to ferret antiserum raised against A/Astrakhan/3212/2020 (H5N8), a WHO CVV (Supplementary Table 2). Most of the Bangladeshi 2.3.4.4b HAs have an N158D substitution (Supplementary Table 3) in an antigenic region. Furthermore, the N158D substitution disrupts an N-glycosylation site that increases binding to human-type receptors and is present in airborne-transmissible A(H5N1) viruses [37–40]. The Bangladeshi 2.3.4.4b A(H5N1) viruses lack other well-characterized mammalian-adaptive substitutions, such as PB2 627 K or 701N [41], but do contain 66S in PB1-F2, a mutation that increases viral pathogenicity in mice [42]. The clade 2.3.4.4b A(H5N1) viruses were isolated from asymptomatic ducks (mallard or khaki Campbell), and no concurrent die-offs had been reported. Although mortalities have been reported in wild mallards infected with HPAI, including in the 2021–2022 North American outbreak [43], most mallards infected with HPAI A(H5N1) in laboratory settings show few or no clinical signs, despite shedding large quantities of virus [44–47].

Overall, the data collected in this study show the diverse and dynamic nature of avian influenza viruses in Bangladesh. Specifically, we show (i) the introduction of clade 2.3.4.4b A(H5N1) viruses to Bangladeshi LPMs, (ii) reassortment of clade 2.3.4.4b A(H5N1) viruses previously identified in Tanguar Haor with Eurasian LPAI viruses, and (iii) reassortment of clade 2.3.4.4b and clade 2.3.2.1a A(H5N1) viruses in Bangladesh.

Funding Statement

This project has been funded in whole or in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services (contract number 75N93021C00016) and ALSAC. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental Data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/22221751.2024.2432351.