1. Introduction
Bagaza virus (
Orthoflavivirus bagazaense, BAGV) is a flavivirus belonging to family
Flaviviridae, genus
Orthoflavivirus, close related to Israel turkey meningoencephalomyelitis virus (ITMV) and grouping in the Ntaya serocomplex. As other flaviviruses, it is mainly a mosquito-borne pathogen that affects birds (which are amplifying hosts), especially those belonging to Phasianidae family, such as turkeys, pheasants and partridges [
1].
Bagaza virus was first isolated from
Culex mosquitoes in Bagaza district, Central African Republic, in 1966 [
2]. Since then, it has been detected in several African countries, India, Middle East and, more recently, in Europe [
3,
4,
5,
6,
7,
8,
9,
10,
11]. Despite several countries have reported the detection of BAGV in mosquitos, the first isolation of this virus in a vertebrate host was in Spain in 2010 [
9], associated with a high mortality outbreak in the south of the country that affected partridges and pheasants. More recently, in 2016-2017, BAGV has been noticed in Himalayan monal pheasants (
Lophophorus impejanus) in South Africa [
6], and in 2021 in a corn bunting and several red-legged partridges in Portugal [
5]. On the other hand, the Israel turkey meningoencephalomyelitis virus (ITMV), detected in turkeys in Israel [
12], is so close genetically to BAGV that both have been proposed as belonging to the same virus species [
13].
Regarding pathogenesis, depending on the infected species, BAGV causes: apathy, weakness, unresponsiveness, impaired vision, severe hemolytic process and significant weight loss, among other disease signs [
1,
14,
15,
16]. Remarkably, BAGV can be transmitted rather efficiently by direct contact among red-legged partridges, at least under experimental conditions [
1]. Mortalities of 30% in red-legged partridges [
1] and 40% in grey partridges [
14] have been observed upon BAGV infection, which implies an enormous impact both at socio-economic levels and in the abundance of the natural populations, as well as on the ecosystems of the Iberian Peninsula [
17,
18]. Although, based on serological detection in encephalitic patients from India during the acute phase of the infection, BAGV has been proposed as a zoonotic pathogen [
8], it was found unable to infect mice in experimental conditions [
1], which considerably stands against this claim.
The genome of BAGV consists of a linear, single stranded, positive sense RNA molecule of 10,900-11,000 nucleotides in length. This molecule encodes for a single polyprotein of 3427 amino acids, that is further processed into 3 structural proteins: capsid (C), pre-membrane (prM), and envelope (E), and seven non-structural proteins (NS): NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5 [
18].
In Spain, BAGV was identified for the first time in September 2010, due to an unusual outbreak associated with high mortality rates in red-partridges (
Alectoris rufa) and common pheasants (
Phasianus colchicus) from Cádiz province (the southernmost province of the country) [
9]. Circulation of the virus was confirmed serologically in the following seasons (2011-2012) [
19]. After that, BAGV was not detected again in Spain until 2019 [
10] and again in 2021, in mosquito [
20] and in red-legged partridges. Moreover, BAGV has been reported for the first time in Portugal in September 2021, in a corn bunting (
Emberiza calandra) and several red-legged partridges [
5].
Bagaza virus is a neglected arbovirus of whom little is known apart from the information provided above. This study aimed at: 1) elucidating the origin of the recently re-emerged BAGV strains detected in Spain after nearly a decade of epidemiological silence, 2) establishing relationships with other circulating strains from different countries, 3) raising knowledge about the epidemiological situation and dispersal behavior of this pathogen. For this purpose, we have undertaken the molecular characterization and phylogenetic analysis of four Spanish BAGV isolates, obtained from outbreaks occurred in red-legged partridges in 2019 and 2021, in Cádiz, southern Spain.
4. Discussion
Bagaza virus was first identified in Europe, concretely in southern Spain, in 2010, during an unusual high mortality event affecting red-legged partridges and common pheasants [
9]. The next outbreak in Europe occurred 9 years after, in 2019, in red-legged partridges in the same region of Spain. In 2021, another outbreak occurred in the same Spanish province, affecting the same avian species, whereas in Portugal, BAGV was firstly detected in the same 2021 season, particularly in a corn bunting and several red-legged partridges [
5] in an area that is close to the affected Spanish territory. Although this virus has been circulating in several African countries since at least the last three decades, it is striking why BAGV has been detected neither molecularly nor serologically in Spain during almost a decade since its first introduction, despite active and passive surveillance carried out in the territory. In this regard, the present study aimed to shed light into the epidemiological mechanisms by which this virus spreads between two continents by analysing and characterising the full genome sequences of four BAGV isolates from Spain. The phylogenetic analyses performed identify two genetic clusters, or genotypes, of BAGV, here named BAGV-Genotype 1 and BAGV- Genotype 2, integrating strains that have been circulating in different continents (Africa, Asia and Europe), in mosquito and avian populations.
On the one hand, BAGV- Genotype 1 comprises more ancient and longstanding variants, including the earliest isolates of ITMV from Israel (1958-1959), which were the responsible of an epizootic affecting turkeys in that country [
23]. After that, the virus continued circulating in that territory, affecting turkey flocks, and becoming an emerging problem for the poultry industry in the country due to the high economic losses it caused. To cope with it, a live-attenuated vaccine was developed [
24]. This genotype was also detected in mosquitoes during an outbreak of human encephalitis in 1996 in India, which, together with serological data, suggested a zoonotic potential for this virus [
8], an aspect that still remains unclear. Besides, Genotype 1 representatives had been detected in
Culex mosquito pools from West African countries, Senegal (1989-2014) and Ivory Coast (1988) [
7], confirming a continued presence of this cluster in those regions. More recently, in 2019, this genotype emerged for the first time in Europe, concretely in Spain, in the same territory where the other genotype of BAGV (Genotype 2) was initially detected nine years before, in 2010, affecting red-legged partridges.
On the other hand, BAGV-Genotype 2 seems to have evolved more recently. Members of this cluster were first identified in 2010 in Israel and Spain [
9,
13], affecting phasianids (turkeys, pheasants and partridges). Later on, another strain from this cluster was reported in mosquitoes (
Culex quinquefasciatus) in Zambia in 2013. Genotype 2 strains were later identified in two south-western African countries, particularly in
Cx. univittatus mosquitos from Namibia (2018) [
11] and in monal pheasants from South Africa (2016-2017) [
6]. Another member of this genotype was also noticed in
Cx. perexiguus from United Arab Emirates (2018) [
3] (not included in the phylogenetic analyses due to the short length of this sequence). More recently, in 2021, strains of this genotype arose simultaneously in Spain, in red-legged partridges and mosquitoes [
20], and in Portugal, affecting wild birds [
5]. Lastly, in 2023, a Genotype 2 strain was identified again in South Africa in
Aedes dentatus mosquitoes.
These analyses confirm that there are at least two main genotypes of BAGV actively circulating since decades in birds and mosquito populations in Africa and Asia, and more recently in Europe. More in detail, Portuguese authors propose four groups within the BAGV/ITMV monophyletic clusters, namely G1, G2, G3, and G4, separated by low intra genetic distances [
18]. Following this proposal, BAGV-Genotype 1 would be divided into 2 groups: G1 gathering Israeli isolates from 1959 to 1995, while G2 would comprise Senegalese ones, as well as representatives from Central African Republic, India, Ivory Coast, and the Spanish isolates from 2019. On the other hand, BAGV-Genotype 2 would be divided also into 2 groups: G3 would gather Israeli isolates from 2010, while G4 would include isolates from the Iberian Peninsula, i.e. Spain, 2010 and 2021, and Portugal 2021, as well as African isolates from Namibia, Zambia, and South Africa.
These results reflect a remarkable dispersal capacity of BAGV through several territories and environments, as well as its ability to establish endemic cycles in different countries. Furthermore, they also support previous observations indicating that BAGV and ITMV likely belong to the same viral species [
13,
18]. Despite that, the International Commitee for Viral Taxonomy (ICTV) still classify them into two different viral species within the
Orthoflavivirus genus (
https://ictv.global/report/chapter/flaviviridae/flaviviridae/orthoflavivirus, ICTV 2023 release), a concept that is becoming obsolete as new sequence data become known. In addition, the study of the polyprotein supports the phylogenetic results, displaying each BAGV-Genotype a specific amino acidic signature in seven positions.
Focusing on the Spanish isolates, the emergence of BAGV in Spain in 2019 after almost a decade of silence was due to a new independent introduction of a BAGV-Genotype 1 variant from an African territory, probably related to Senegal. This connection between isolates from Senegal and Spain has also been observed in other flaviviruses such as West Nile virus [
25,
26], although this observation could be biased to the available sequences origin. Surprisingly, two years later, in 2021, a BAGV-Genotype 2 variant re-appeared in the territory, after eleven years undetected. This re-emergence of Genotype 2 was likely caused by a new introduction/s of BAGV in the Iberian Peninsula, as judged by the highly homologous amino acid pattern that the Spanish and Portuguese 2021 isolates share in the polyprotein, suggesting a common-close origin. Furthermore, this common pattern clearly differs from the original isolates from Spain (2010) as well as from other BAGV-Genotype 2 representatives. Moreover, Portuguese and Spanish 2021 isolates are not 100% homologous, neither at nucleotide nor at amino acid level, in fact, they differ in eight amino acid positions. Consequently, the almost simultaneous detection of BAGV-Genotype 2 in neighbouring Portuguese and Spanish territories might be due either to two separate introductions of the same viral strain, during the 2021 season, or by a single introduction that has been able to evolve and move near the border between these countries.
Hence, these findings allowed to reconstruct the main events related to the appearance of BAGV in the Iberian Peninsula. Firstly, they suggest that there have been, at least, three independent introductions of BAGV in Spain. The first one occurred in 2010, caused by a BAGV-Genotype 2. Then, after a silent period, a new cluster (Genotype 1) emerged in the same territory in 2019; finally, Genotype 2 re-emerged in 2021, affecting Spanish and Portuguese territories. Whether these variants are able to settle down and establish an enzootic cycle is still unknown, as more data are needed to confirm this hypothesis.
Regarding the impact of BAGV in animal health, this is an important pathogen not only for poultry farming (turkeys) but also for wild birds, and particularly for some species of game birds that are raised in farms for hunting purposes, such as pheasants and partridges, sustaining an economically important activity in some areas. For this reason, the presence of BAGV in red-legged partridges implies a great concern in nations such as France, Italy, Portugal and Spain, as this is a species of high economic-relevance in those countries, in fact, it is the only autochthone partridge species in Portugal [
18]. Taking into account the results of this study and considering that phasianids are highly susceptible to BAGV infection, it is necessary to reinforce surveillance activities in the south of the Iberian Peninsula in order to provide an early warning and to apply the appropriate control measures to reduce the number of affected animals and the subsequent economic losses. Of note, BAGV has been detected in the Iberian Peninsula only when outbreaks occur in these avian species, and only one mosquito pool was found positive in 2021 [
20], despite the intense flavivirus vector surveillance carried out in Southern Spain. This fact contrasts with the situation in Africa, where most BAGV data come from mosquito detections.
In summary, this study confirms the re-introduction of two BAGV variants in Spain, after a long period of epidemiological silence. As the red-legged partridge is highly susceptible to BAGV infection and disease, it would constitute a suitable target species for BAGV surveillance.