Navigation überspringen

Library

The H5 highly pathogenic avian influenza: understanding the patterns of virus spreading to/from Africa

Alice Fusaro, Bianca Zecchin, Bram Vrancken, Celia Abolnik, Rose Ademun, Yao P. Akpeli, Abdou Alassane, Joseph Adongo Awuni, Emmanuel Couacy-Hymann, M’Bétiégué Coulibaly, Emilie Go-Maro, Tony Joannis, Simon Dikmu JUMBO, Germaine Minoungou, Clement Meseko, SOULEY Maman Moutari , Deo Birungi Ndumu, Augustin Twabela, Abel WADE, Lidewij Wiersma, Gianpiero Zamperin, Adelaide Milani, Philippe Lemey, Isabella Monne

Abstract

During the 2006, 2014 and 2016 intercontinental epidemic waves of the H5 highly pathogenic avian influenza virus (HPAIV) of the Gs/GD lineage, the African continent was reached by three distinct H5 genetic clades, namely 2.2, 2.3.2.1c and 2.3.4.4. Our purpose is to investigate the role of Africa in the global spread of the HPAIV H5 and to shed light on the contributions of different avian host populations to the virus introduction and dissemination.

We generated two datasets of the hemagglutinin gene for each clade, one including representative sequences from affected regions throughout the world and from different host species, and the other comprising all the African sequences generated in this study or retrieved from public databases. Spatiotemporal data analysis of the epidemic spread and of host switching patterns were performed through Bayesian phylogeographic analyses in both discrete and continuous space for each dataset using the BEAST v1.8.4 program.

We recognized multiple introductions as well as varying paths of avian influenza incursions into the African continent during the three epidemic waves and were able to demonstrate that Africa had mainly served as an ecological sink of the H5 Gs/GD HPAIV. Within Africa, West Africa acted as a crucial hotspot for virus introduction and dissemination into the continent. A joint analysis of host dynamics and continuous spatial diffusion indicated that both migratory birds and live poultry trade may have played an important role in the spread of the virus into the continent.

Overall, this study suggests that viral sources are not stable over time in the African continent, but can change at each epidemic wave. Enforcing target surveillance in the regions identified as at high risk of virus incursion, such as West Africa, is recommended.

Poster at the Epidemics7 - International Conference on Infectious Disease Dynamics in Charleston, SC, USA (3-6 December 2019)

Modulation of highly pathogenic avian influenza virus H5N8 infection in naturally low pathogenic avian influenza virus - exposed mallard ducks

S Koethe, L Ulrich, A. Globig, T Harder, FJ Conraths, M Beer

Abstract
The 2016/2017 highly pathogenic avian influenza virus (HPAIV) H5Nx clade 2.3.4.4b outbreak was the most severe ever reported in Germany with high mortality rates not only among domestic birds, but also in wild birds.
Mallard ducks have been kept as sentinels for circulating field AIV in an aviary located at the shallows of the Baltic coast in close contact to wild waterfowl and migratory birds. The mallards were regularly tested for AIV infection. Between August 2017 and October 2018, four low pathogenic (LP) AIV infections of individual mallards were confirmed - H1N3, H3N8, H4N6 and HxN9. One month prior to the HPAIV study, all mallard ducks were tested seropositive in an IAV NP-ELISA, and two out of seven had H5-specific antibodies.
To study dynamics of HPAIV H5N8 clade 2.3.4.4b infection after a prior natural field exposure to LPAIV, seven LPAIV exposed mallards were challenged with HPAIV H5N8. The surviving ducks were re-challenged with the homologous HPAIV H5N8. For both challenge infections, virus transmission to seronegative juvenile Pekin contact ducks was investigated. Seronegative adult Pekin ducks served as a control group.
Summarizing the results, during the first challenge infection all control group ducks died peracutely or had to be euthanized due to severe neurological symptoms. However, although one mallard duck died, the mallards showed only very few clinical symptoms, but three out of four contact juvenile Pekin ducks died. H5-specific antibodies were detectable in all surviving birds. Infectious virus could be isolated from several pond and drinking water samples. In contrast, after the homologous H5N8 re-challenge, neither inoculated nor contact ducks showed any clinical symptoms, nor was an AI-specific antibody titer increase or seroconversion of contact animals determined.
In conclusion, the study simulated an HPAIV H5N8 outbreak under semi-natural conditions with LPAIV-exposed waterfowl. Mallards with preceding LPAIV immunity, did not show any severe clinical or fatal disease, but sufficient virus shedding caused fatal illness in naïve contact ducks. In contrast, H5N8-convalescent animals developed a very robust immunity when re-challenged with the homologous virus shown by the absence of virus shedding and seroconversion of the contact animals.

Talk at the European Congress of Virology in Rotterdam, The Netherlands (28 April - 1 May 2019)

Impact of the polybasic cleavage site within the HA of a recent German H7N7 virus on its pathogenicity in different poultry species

Scheibner D., Veits J., Mettenleiter T.C. and Abdelwhab E.M.

Abstract
Virulence of avian influenza viruses (AIVs) is largely dependent on the amino acid sequence of the hemagglutinin cleavage site (CS). Low pathogenic AIV (LPAIV) carrying an HA with a monobasic CS which is activated by trypsin-like proteases in the respiratory and intestinal tracts cause only local infections with mild clinical signs, if any. Some H5 and H7 subtypes exhibit high pathogenicity (HP) by acquisition of a polybasic CS after circulation of LP precursors in terrestrial poultry. The polybasic CS of HPAIV is cleaved by ubiquitous furin-like proteases causing systemic infections and high mortality. In 2015, in an outbreak in poultry in Germany LP and HP H7N7 viruses could be isolated on the same farm indicating precursor-progeny relationship. Here, we investigated the pathogenicity of these LP and HP viruses in chickens, Muscovy ducks and turkeys via oculonasal and intravenous inoculation and analyzed virulence determinants by reverse genetics. Muscovy ducks showed no or mild clinical signs, while turkeys and chickens died after infection with HPAIV. Insertion of a polybasic CS into the HA of the LP virus (LP-poly) increased pathogenicity for chickens and turkey drastically, but was not sufficient for 100% mortality. LP-poly was less virulent in turkeys than in chickens exhibiting IVPI values of 2.8 and 1.9, re-spectively. Thus, virulence determinants differ between these two species. Our results contribute to understand the pathobiology and evolution of recent LP and HP H7N7 viruses in different poultry species.

Poster at the 29th Annual Meeting of the German Society for Virology in Düsseldorf, Germany (20-23 March, 2019)

Transmission dynamics of highly pathogenic avian influenza virus A (H5N8) in Italy, 2016-2017

Alice Fusaro, Bianca Zecchin, Paolo Mulatti, Gianpiero Zamperin, Alessia Schivo, Silvia Ormelli, Sabrina Marciano, Lebana Bonfanti, Alessandra Azzolini, Giovanni Cunial, Paola Massi, Anna Moreno, Maria Lucia Mandola, Stefano Marangon, Calogero Terregino, Isabella Monne

Abstract

Between December 2016 and December 2017, Italy was affected by Highly Pathogenic Avian Influenza virus (HPAIV) H5N8 outbreaks both in domestic and in wild birds. The first epidemic wave lasted for six months, from December 2016 to May 2017, with 16 outbreaks in poultry farms and 7 in migratory birds. The second epidemic wave began on the third week of July 2017 and continued until mid-December of the same year. Sixty-seven outbreaks were observed in poultry farms, while only seven cases were reported in wild birds.

To shed light on the source of the Italian H5N8 outbreaks and to investigate the inter- and intra-farm genetic diversity of the circulating viruses, we used a next generation sequencing approach to characterize the complete genome of 115 samples collected from multiple hosts both from each infected farm and from viruses identified in wild species.

Maximum-likelihood phylogenetic trees of the eight gene segments indicated four distinct introductions of AIV genotypes at the beginning of the epidemic (December 2016-February 2017). Since March 2017 one single genotype (H5N8-A/wild duck/Poland/82A/2016-like) had been identified, with the exception of a single virus detected in a turkey farm in October, which turned out to be a reassortant virus for the NP and PA genes, likely acquired from low pathogenic H9 viruses circulating in resident wild birds. Evolutionary and phylogeographic analyses performed using the BEAST v1.8.4 package suggested that genotype H5N8-A/wild duck/Poland/82A/2016-like had further evolved into two distinct clusters, namely Italy-A and Italy-B. Italy-A had probably emerged between February-April 2017 and circulated in the north-east of Italy; differently, Italy-B seems to have appeared  between March-July 2017 and mainly spread in the north-western regions.

During the second epidemic wave, epidemiological investigations identified four clusters of secondary cases, involving 36 infected farms. Viruses isolated from each of these clusters grouped together in the phylogenetic trees of all the eight AIV genes and in our median-joining (MJ) network of the eight concatenated gene segments, thus supporting the findings of the epidemiological analysis.

In addition, we demonstrated a high intra-farm genetic diversity (0-9 nucleotide substitutions/genome). This highlighted the importance of genetically characterizing viruses from multiple hosts within a single farm, so as to correctly reconstruct the evolution and the transmission dynamics of an AI epidemic.

Overall these results proved to be instrumental to help epidemiological investigations to discriminate between new introductions and lateral spreads. This type of analyses could be performed in almost a real-time fashion, providing quickly accessible information to generate and/or corroborate hypotheses on the likely epidemiologic pattern of contacts between cases, in order to adjust AI control and prevention policies.

Talk at the 2nd National Congress of the Italian Society for Virology in Rome, Italy (28-30 November 2018)

Investigating Africa’s contribution to the global spread of H5Nx highly pathogenic avian influenza viruses

Alice Fusaro, Bianca Zecchin, Bram Vrancken, Celia Abolnik, Rose Ademun, Yao P. Akpeli, Abdou Alassane, Joseph Adongo Awuni, Emmanuel Couacy-Hymann, M’Bétiégué Coulibaly, Emilie Go-Maro, Tony Joannis, Simon Dikmu JUMBO, Germaine Minoungou, Clement Meseko, SOULEY Maman Moutari , Deo Birungi Ndumu, Augustin Twabela, Abel WADE, Lidewij Wiersma, Gianpiero Zamperin, Adelaide Milani, Philippe Lemey, Isabella Monne

Abstract

In the last two decades, four intercontinental epidemic waves of the highly pathogenic avian influenza virus (HPAIV) of the A/goose/Guangdong/1/1996 lineage have been reported. The worldwide spread of this strain has caused important damages to the poultry industry, but it also represents a serious concern for public health since the virus can be occasionally transmitted to humans. The African continent has been reached by three of these four waves, which resulted in the introduction of three distinct genetic clades, namely clade 2.2 in 2006, clade 2.3.2.1c in 2014 and clade 2.3.4.4 in 2016. The virus has become endemic in poultry in some African regions, such as Egypt and West Africa, where multiple clades are currently co-circulating, creating the opportunity of genetic reassortment and of emergence of viruses with unknown biological properties. To shed light on the potential risk of virus spread from Africa to the other continents, we compared the global HPAIV H5 transmission patterns of these clades and explored the contributions of different avian host populations to virus dissemination.

For each of the three clades, we generated a dataset which included representative sequences from several affected regions throughout the world and from different host species. Based on the spatial distribution and host origin of the available sequences, we identified 9 discrete regions - West Europe, East Europe, Middle East, East Asia, North-Central Asia, South Asia, West Africa, Central-east Africa and South Africa - and 4 host traits - domestic Galliformes, domestic Anseriformes, wild Anseriformes and other wild bird species. We reconstructed the history of epidemic spreads in space and time simultaneously with the host switching patterns through Bayesian phylogeographic analyses in both discrete and continuous space for each dataset using the BEAST v1.8.4 program.

We identified multiple introductions of clade 2.2 from Europe and of clade 2.3.4.4-B from South and North-Central Asia into the African continent, while a single virus spread from South Asia to Africa seems to have been responsible of the 2.3.2.1c incursion. A joint analysis of host dynamics and continuous spatial diffusion indicates that the incursion of the H5 clades into Africa has been driven by wild Anseriformes and domestic Galliformes hosts, thus suggesting that both migratory birds and live poultry trade may have played an important role in the spread of the virus into Africa. Our results identify only few virus migrations from Africa to the Middle East, likely through poultry trade, while no virus movement from Africa to other continents has been observed.

Overall, this study suggests that Africa mainly serves as a sink of the virus. However, the routes of virus introduction into Africa by means of both wild and domestic birds may change at each epidemic wave, making it difficult to predict the source for the next incursion.

Poster at the 2nd National Congress of the Italian Society for Virology in Rome, Italy (28-30 November 2018)

Global Origins Of African Highly Pathogenic Avian Influenza H5Nx Viruses And Intracontinental Spread

Alice Fusaro, Bianca Zecchin, Bram Vrancken, Celia Abolnik, Rose Ademun, Yao P. Akpeli, Abdou Alassane, Joseph Adongo Awuni, Emmanuel Couacy-Hymann, M’Bétiégué Coulibaly, Emilie Go-Maro, Tony Joannis, Simon Dikmu JUMBO, Germaine Minoungou, Clement Meseko, SOULEY Maman Moutari , Deo Birungi Ndumu, Augustin Twabela, Abel WADE, Lidewij Wiersma, Gianpiero Zamperin, Adelaide Milani, Philippe Lemey, Isabella Monne

Abstract

In 2006, 2014 and 2016 Africa was hit by three intercontinental epidemic waves of the highly pathogenic avian influenza virus (HPAIV) belonging to three distinct genetic clades of the A/goose/Guangdong/1/1996 lineage, namely 2.2, 2.3.2.1c and 2.3.4.4, which substantially affected the local poultry industry. In this study we compared the global and intra-African HPAIV H5 transmission patterns of these clades to shed light on the spread of the virus within Africa and to explore the contributions of different avian host populations to virus introduction and dissemination.

For each of the three clades, two datasets of the hemagglutinin gene were generated: one including representative sequences from affected regions throughout the world and from different host species, and one comprising all the African sequences generated in this study or retrieved from public databases. We reconstructed the history of epidemic spread in space and time and the host switching patterns through Bayesian phylogeographic analyses in both discrete and continuous space for each dataset using the BEAST v1.8.4 program.

We identified multiple introductions of clade 2.2 from Europe and of clade 2.3.4.4-B from South and North-Central Asia into the African continent, while a single virus spread from South Asia to Africa seems to have been responsible of the 2.3.2.1c incursion. Our results identify West Africa as the most important area of virus introduction into the continent. A joint analysis of host dynamics and continuous spatial diffusion indicates that the incursion of the H5 clades into Africa is driven by wild Anseriformes and domestic Galliformes hosts, suggesting that both migratory birds and live poultry trade may have played an important role in the spread of the virus into Africa.

This study shows that viral sources are not stable over time in the African continent, but can change at each epidemic wave, making it difficult to predict the source for the next incursion. In addition, our results indicate a strategic role of West Africa in the virus spread within the continent, which may be considered as a hotspot for H5 HPAIV surveillance.

Talk at the International Meeting on Emerging Diseases and Surveillance in Vienna, Austria (9-12 November 2018)

The African continent: an endpoint in the global spread of the highly pathogenic avian influenza H5Nx virus

Alice Fusaro, Bianca Zecchin, Bram Vrancken, Celia Abolnik, Rose Ademun, Yao P. Akpeli, Abdou Alassane, Joseph Adongo Awuni, Emmanuel Couacy-Hymann, M’Bétiégué Coulibaly, Emilie Go-Maro, Tony Joannis, Simon Dikmu JUMBO, Germaine Minoungou, Clement Meseko, SOULEY Maman Moutari , Deo Birungi Ndumu, Augustin Twabela, Abel WADE, Lidewij Wiersma, Gianpiero Zamperin, Adelaide Milani, Philippe Lemey, Isabella Monne

Abstract

In the last two decades, three of the four intercontinental epidemic waves of the highly pathogenic avian influenza virus (HPAIV) of the A/goose/Guangdong/1/1996 lineage reached the African continent, resulting in the introduction of three distinct genetic clades in African poultry, namely clade 2.2 in 2006, clade 2.3.2.1c in 2014 and clade 2.3.4.4 in 2016. To compare the global and within-Africa HPAIV H5 transmission patterns of these clades and to shed light on the contributions of different avian host populations to the virus spread, we performed Bayesian phylogeographic analyses of the hemagglutinin gene of each of the three clades both on a global and local (Africa) scale.

We identified multiple introductions of each clade into the African continent, either with the same (clade 2.2) or a mixed geographic origin (clades 2.3.2.1.c and 2.3.4.4). We find that HPAIV always first arrives in West Africa, followed by spread to Central-East (2.2 and 2.3.4.4) and/or South (2.3.4.4) Africa. A joint analysis of host dynamics and continuous spatial diffusion indicates that the incursion of the H5 clades into Africa is mainly driven by wild Anseriformes hosts, suggesting that migratory birds play a major role in the spread of the virus into Africa. With one exception, we observed no backward spread from Africa to Eurasia.

These findings highlight the strategic role of West Africa, which is located at the intersection of three flyways, as a hotspot for H5 HPAIV surveillance, particularly in wild birds, for which no or very few data are currently available.

Talk at the Virus Genomics and Evolution 2018 in Cambridge, UK (18-20 June 2018)

Genesis of a highly pathogenic virus: what can we learn from deep sequencing data?

Alice Fusaro, Gianpiero Zamperin, Adelaide Milani, Alessia Schivo, Annalisa Salviato, Isabella Monne

Abstract

The mechanism driving the evolution of a low pathogenic (LP) avian influenza (AI) virus into a highly pathogenic (HP) form is currently scarcely understood. Between April and May 2016 Italy reported two H7N7 outbreaks, one in a laying farm (farm-1) and the other in a turkey farm (farm-2). Epidemiological investigations suggested that the LP form of this strain had been introduced into farm-1 through direct contact between wild birds and free-range hens and then rapidly mutated into a HP form within the premise.

To explore the intra-farm evolution of the virus pathogenicity we genetically characterized the H7N7 positive samples collected from symptomatic and asymptomatic animals at farm-1, revealing the co-circulation of both the HP and the LP progenitor virus. A deep sequencing analysis was performed on three LPAI and five HPAI (four from farm-1 and one from farm-2) viruses.

Phylogenetic analyses showed that their genomes clustered together and were separated by long branches from viruses circulating in Eurasia. The LPAI viruses fell at the base of this cluster and possessed a total of 8 nucleotide and 6 amino acid differences compared to the HPAI viruses. Deep sequencing analysis of the eight viruses indicated that the number of genetic differences as well as the complexity of the viral population of the HP viruses from farm-1 was lower compared to the LP viruses (p<0.05), suggesting that the HP viruses had recently experienced a narrow bottleneck.

These results highlight the importance of genetic surveillance within a single farm to monitor the virus evolution.

Poster at the 10th International Symposium on Avian Influenza, Brighton, UK (15 - 18 April 2018)

Strengthening outbreak investigations using a next generation approach

Alice Fusaro, Bianca Zecchin, Gianpiero Zamperin, Alessia Schivo, Angela Salomoni, Annalisa Salviato, Silvia Ormelli, Sabrina Marciano, Lebana Bonfanti, Alessandra Azzolini, Giovanni Cunial, Paolo Mulatti, Calogero Terregino, Isabella Monne

Abstract

A big data approach to assisting the H5N8 highly pathogenic avian influenza (HPAI) outbreak investigations in Italy was used.

To shed light on the source of the Italian H5N8 outbreaks reported since December 2016 and to investigate the inter- and intra-farm genetic diversity of the circulating viruses, we used an ultradeep sequencing approach to characterize the complete genome from multiple hosts collected at each infected farm as well as from viruses identified in wild species, for a total of 115 samples and about 50 gigabases of data.

Maximum-likelihood phylogenetic trees showed that, following the multiple viral introductions that had occurred at the beginning of the epidemic (December 2016-February 2017), a single genetic group seems to have been circulating in Italy. Evolutionary and phylogeographic analyses performed using the BEAST v1.8.4 package indicate that this group has further evolved into two distinct clusters, namely Italy-A and Italy-B. Italy-A probably emerged between February-April 2017 and has been circulating ever since in the north-east of Italy; differently, Italy-B arose between March-July 2017 and has mainly spread in the north-west of Italy. Median-joining network analysis of the inter- and intra-farm genetic diversity highlights the importance to genetically characterize viruses from multiple hosts sampled at each infected farm, so as to correctly reconstruct the evolution and the transmission dynamics of an AI epidemic.

These results were instrumental to discriminate between new introductions and lateral spreads and to demonstrate the potential and challenges of using high-throughput sequencing data for surveillance and control.

Poster at the 10th International Symposium on Avian Influenza, Brighton, UK (15 - 18 April 2018)

Pathogenicity and virulence determinants of recent German H7N7 viruses in different poultry species

Scheibner, D., Salaheldin, A.H., Winter, F., Gischke, M., Veits, J., Mettenleiter, T.C., and Abdelwhab, E.M.

Abstract

Virulence of avian influenza viruses (AIVs) is largely dependent on the amino acid sequence of the hemagglutinin cleavage site (CS). Low pathogenic AIV (LPAIV) carrying an HA with a monobasic CS which is activated by trypsin-like proteases in the respiratory and intestinal tracts cause only local infections with mild clinical signs, if any. Some H5 and H7 subtypes exhibit high pathogenicity (HP) phenotype after acquisition of a polybasic CS after circulation of LP precursors in terrestrial poultry. The polybasic CS of HPAIV is cleaved by ubiquitous furin-like proteases causing systemic infections and high mortality. In 2015, in an outbreak in poultry in Germany LP and HP H7N7 viruses could be isolated on the same farm indicating precursor-progeny relationship. In this study, we investigated the pathogenicity of these LP and HP viruses in chickens, Muscovy ducks and turkeys via oculonasal and intravenous inoculation. Furthermore, virulence determinants of HPAIV were identified using reverse genetics. Muscovy ducks showed no or mild clinical signs, while turkeys and chickens died after the challenge with HP AIV. Insertion of a polybasic CS in the HA of the LP virus (LP_poly) increased pathogenicity for chicken and turkey drastically, but was not sufficient to kill all animals. LP_Poly was less virulent in turkeys than chickens exhibiting IVPI values of 2.8 and 1.9, respectively. Interestingly, LP_poly containing the NS gene from the isogenic HPAIV was as virulent and transmissible as the original HPAIV in chickens. Our results are important to better understand the pathobiology and evolution of recent LP and HP H7N7 viruses in different poultry species.

Talk at the 10th International Symposium on Avian Influenza, Brighton, UK (15 - 18 April 2018)