Introduction

Avian influenza is an infectious disease of birds. All birds are thought to be susceptible to infection with the viruses causing avian influenza, but many wild bird species carry these viruses with no apparent signs of harm. Some bird species, including domestic poultry (hens, turkeys, quails etc), are especially susceptible to these viruses and develop disease. In poultry, these viruses cause two distinctly different forms of disease: one is common and mild and is named Low Pathogenic Avian Influenza (LPAI); the other by contrast is rare but highly lethal and is named Highly Pathogenic Avian Influenza (HPAI). HPAI, which was first identified in Italy in 1878, is characterized by sudden onset of severe disease, rapid spread, and a mortality rate that can approach 100% within 48 hours (WHO, 2006a). Because of its contagious and lethal nature, HPAI was called Fowl Plague, together with Newcastle Disease. Once HPAI is established in domestic poultry, it is a highly transmissible disease and wild birds are no longer essential actors for its spread. Infected birds excrete virus in high concentration in their faeces and also in nasal and ocular discharges. Once introduced into a domestic flock, the virus is passed on from flock to flock by the usual methods involving the movement of infected birds, contaminated equipment, egg flats, feed trucks, and service crews.
An unprecedented epizootic of HPAI is raging across three continents, gaining momentum as time goes by. Not only the geographical scale of this epizootic is unheard of reaching global proportion as a panzootic but also the spectrum of avian and mammalian species infected by the culprit virus and its pathogenicity, which is variable in some species from time to time, are extraordinary. To the date of writing this proposal, the Office International des Epizooties (OIE) website lists 34 (or 7) hit by the circulation of influenza H5N1 (or H5 respectively), 9 (or 0) of them belonging to the European Union (OIE, 2006). The cradle of the current global epizootics is in the Far East, where outbreaks started in 2003. Virus spread has been unstoppable since then among wild bird as well as in domestic flocks. The areas of high density of poultry in the regions of the world where it first started match with the territories of outbreaks eg East China, North and South Vietnam, Thailand and Indonesia (in particular Sumatra) (FAO, 2006). The epizootics spread towards the west part of the Eurasian continent through migrating wild birds but probably also through human activities related to poultry either from industrial settings or backyards during the 2005 summer. The arrival of A(H5N1) virus in Siberia (2006) was a sign that its progression would probably reach far beyond considering the importance of that region of the world for wild aquatic birds. Since then, the virus westward progression has first reached countries neighbouring the European Union such as Romania (07/10/2005), Turkey (12/10/2005) and Croatia (21/10/2005) then it entered UE territories (first Italy – 11/02/2006), causing its first poultry outbreak (in France) on the 23rd February 2006 (OIE, 2006) where 400 animal died from the disease and 700 were culled. During this time, HPIA due to the A(H5N1) virus caused epizootics in Nigeria (08/02/2006) and soon in the neighbouring country Niger (28/02/2006). So far, it is estimated that more than 140 million poultry (as of September 2005) were culled as part of aggressive control measures. In spite of these, outbreaks have recurred as it seems that the causative virus has become endemic in many parts of Asia. It has established an ecological niche in poultry, making it extremely difficult to control.
            Besides the animal health issue, this unique global epizootic is also associated with human health issues. To the 10th March, 176 human cases of infection by the HPIA causing the outbreaks in poultry have been reported, 97 of which are dead (WHO, 2006b). All human cases occurred in countries where there were outbreaks of HPAI in poultries. Almost all human infections can be linked to contact with infected poultry, but isolated instances of inefficient transmission from a human being to another one may have occurred in Viet Nam in 2004, and possibly in Thailand and Indonesia, in peculiar circumstances including prolonged contacts in household. Because influenza A viruses, whether from human or animal hosts, share most features and are genetically unstable either by mutations and/or reassortments, the rare zoonotic passages of these viruses from birds to human pose a threat of the emergence of a HPAI, which could become adapted to human and trigger a full blown pandemic. Until the viruses remain as they are, prevention and control of the animal disease is the best way to avert a potential pandemic.

Etiological agent: Influenza A Virus (H5N1)

The viruses causing avian influenza viruses (AIVs) are influenza viruses, which are single stranded negative RNA viruses. Their genome is segmented into 7 or 8 fragments. Influenza viruses (IVs) are split into three types or taxonomic genera Influenzavirus A, Influenzavirus B et Influenzavirus C. Only influenza A viruses infect birds. They are themselves divided into subtypes according to the nature of their two surface glycoproteins: the haemagglutinin (H) and the neuraminidase (N). There are 16 ‘molecular species’ of H and 9 of N, a subtype being defined by the combination HxNy. IVs causing HPAI belong to two categories of subtypes: H5Ny and H7Ny. The current global epizootic is caused by a HPAI virus (HPAIV) A(H5N1).
IVs are prone to mutate and they evolve by accumulating mutations within their genome in the course of time. Recent genetic analyses reveal that the long lasting circulation of the H5N1 viruses in domestic poultry in Asia has resulted in the establishment of distinct regional virus sublineages (Vietnam/Thailand/Malaysia; Guangdong; Hunan; Yunan and Indonesia) due to cumulated mutations (Chen et al., 2006). Mutations can be reflected at the aminoacid level and can be beneficial for the virus. Some viral phenotypes can be changed by single or plural mutations. In poultry, the insertion of a multibasic cleavage site between the two subunits of the hemagglutinin transforms a LPAIV into a HPAIV (Ito et al., 2001). Mutations can also alter the virus receptor binding site, which recognises sialic acid (SA) by the H at the surface of the target cells. SAs can be of various chemical compositions (NeuAc or Neu Gc) and be linked to the underlying carbohydrate moieties by different kind of linkages, namely α2,3 and α2,6 for glycoproteins or α2,8 for gangliosides. Human IVs link preferentially to α2,6 bound SA found on human cell surfaces, whereas AIVs link rather to α2,3 bound SA found on avian cell outer membranes (Rogers & Paulson, 1983). Such mutations can contribute to breech the species barrier. However, the direct transmission of AIVs to humans in 1997 in Hong Kong suggested that other mutations are also required for AIVs adapting to humans. (Claas et al., 1998, Suarez et al., 1998, Subbarao et al., 1998). In particular, residue 627 of PB2 being a determinant of cold sensitivity in RNA replication of AIVs, mutations at this position can alter the biological behaviour of a mutant AIV in a human host towards adaptation (Massin et al., 2001, Naffakh et al., 2000). This is probably true for other mammals. IVs may also vary through another kind of genetic accidents: reassortments. This drastic phenomenon happens as follows. During a co-infection of one host by two different viruses, a hybrid viral particle can be formed, at the moment of the formation by budding of the new virions including some genomic segments of one parental virus and the rest of its genome segments from the other parental virus. To date, available genetic data indicate that the currently dominating AIV A(H5N1) bears haemagglutinins and neuraminidases, which are genetically derived from those of 1997 A(H5N1) viruses. However, the rests of the genome segments, making up the genotype of the currently dominating AIV, is different from the AIV which caused outbreaks in poultry and human cases in 1997 (Goose/Guangdong or Gs/Gd genotype) (Li et al., 2004a 14098). Indeed, a study suggested that a series of reassortments occurred since then leading to the genesis of a dominating and hegemonic genotype designated as genotype Z, which was first detected in virus circulating in 2002 (Li et al., 2004b 14098). The same study suggested that domestic ducks in Southern China played a central role in the genesis of genotype Z and its perpetuation. It also indicated that wild birds were involved in the geographic distribution of these viruses in Asia. A recent study confirmed the continued dominance of H5N1 genotype Z virus in most regions in Asia (Chen et al., 2006). However, viruses isolated from some provinces of Southern China (Guangdong, Guangxi, and Hunan) are more diverse, containing genotypes Z, V,W, and G (Li et al., 2004c).
The phenotype of A(H5N1) AIVs has also been changing from a IVs infecting but not killing Anatidae species to infecting and killing them and again infecting but not killing these hosts (Hulse-Post et al., 2005). A(H5N1) AIVs seem to have widened their host spectrum as H5N1 AIVs are apparently distributed widely not only in aquatic birds but also in terrestrial avian species such as tree sparrows (Kou et al., 2005).

Perpetuation of Avian Influenza Viruses in nature

Although there are some data about AIVs in wild avian species, we are lacking of scientific knowledge about the global ecology of influenza viruses in nature. There is one basic alternative: either AIVs require environmental reservoirs such as contaminated waters or AIVs sustain an ongoing never ending circulation among wild birds from one year end to the other. Data collected by Ito and collaborators (Ito et al., 1995) suggest that influenza viruses have been maintained in waterfowl population by water-borne transmission and revealed the mechanism of year-by-year perpetuation of the viruses in the lakes where they breed. Previous studies have shown that still waters could serve as relays for AIVs as they can be so heavily contaminated that AIVs could be isolated without water concentration (Hannoun & Devaux, 1980, Hinshaw et al., 1979). Moreover, studies relating to AIVs circulation in wild migrating and non migrating birds suggested that AIVs was detectable in lakes and pond as well as in birds during autumn and not in spring, implying that the source of contamination is in the North (Hannoun & Devaux, 1980). More recent data collected in northern Europe were consistent with these previous data (ADME Osterhaus, personal communication). Altogether, these data support an essential role of contaminated lakes and ponds as environmental reservoirs of AIVs. Data generated through Tasks 3 and 4 together with data generated by this project will probably allow the scientific community to better understand the global year round cycle of AIVs in nature with the relative importance of each compartment (biological and ‘non biological’ reservoirs). Since A(H5N1) AIVs can cause significant amount of infection in terrestrial birds (Kou et al., 2005), these should be taken into account in the overall virus survival in the environment outside waters.

Survival of AIVs in its environment

                  Data from the literature on influenza virus survival are rather limited, often very old and sometimes not confirmed from one study to another or even contradictory. Generally speaking, viral survival with respect to the temperature and other physicochemical parameters was assessed by infectious titre reduction as measured on hen embryonnated eggs or in cell culture using traditional virological methods. Experimental protocols tried to mimic environmental conditions by exposing virus preparation to liquid manures and physicochemical agents. There is no virological data whatsoever in relation with the mechanism of virus survival outside its hosts.
                  The first demonstration of isolation of AIVs from cloacal swabs, birds faeces and non concentrated lake waters in summer in Canada was published in 1979 (Hinshaw et al., 1979). Much later Ito and collaborators isolated on hens embryonated eggs AIVs from Alaskan lakes, during September when most ducks had left, with viral titres as high as 1,8 à 2,8 log EID50/mL. The same AIVs subtypes were isolated from cloacal swabs and waters.
                  Stallknecht and coll. experimentally studied the persistence of viable AIVs in water using 5 strains isolated from ducks in Louisiana (Stallknecht et al., 1990b). They incubated AIVs in distilled water at the concentration of 106 TCID50/g as determined using chick embryo fibroblasts. Infectious virus was recovered after 60 days when incubated at 17°C, after 3 days at 28°C. Virus half life at 4°C was 75 days. From these data, linear regression indicated that viral preparations could remain infectious for 207 days at 17°C and 102 at 28°C, which was considerable. Webster and coll. studied virus survival in bird faeces (Webster et al., 1978) after experimental infection. Faeces were collected 3 days post infection and viral titres were determined using embryonnated eggs. Infected faeces were incubated in sterile tubes at 4°C and 22°C during 32 days, with a titration every 5 days. The results were that viable virus could be recovered after 32 days at 4°C with a drop in titre from 6,8 to 3,3 log/mL whereas it could be recovered only after 8 days at 22°C with the same drop in titres. Virus was undetectable by isolation after 13 days of incubation. When the same faeces were first diluted in Mississipi water at pH 6.8 at 4 and 22°C, virus titres were stable at 8.1 log/mL until day 7 dropping to 4.3 log/mL at day 32 and dropped to 3.6 log/mL at day 4, the virus being undetectable at day 7 respectively at 4°C and at 22°C. From these two studies, it appears that tested AIVs survived slightly better in the raw faeces than in faeces diluted in river water. However, virus survival was better in distilled water than in the faeces either raw or diluted in river water.
Because mixed pig and poultry farms are not uncommon, it is interesting to assess the potential survival of AIVs in manure liquid abundantly produced by pig farms. Fitchner and coll. presented data at a symposium in 1987 suggesting that A(H5N2) virus survival in liquid manure in normal weather conditions lasted at least 105 days. In a study comparing various enveloped viruses, swine IVs were more resistant than the bovine virus diarrhoea virus and 2 herpesviruses: the infectious bovine rhinotracheitis virus and Aujesky’s disease virus (Haas et al., 1995). In this study, swine influenza virus spiked in pig liquid manure at an initial concentration of 5,8 log TICD50/50mL remain viable during 9 weeks at 5°C, 2 weeks at 20°C, more than 24 hours at 35°C and 40°C, more than 2,5 hours at 50°C and 1 hour at 55°C. It has been suggested that AIVs survived better in Specific Pathogen Free poultry liquid manure than in normal poultry wastes (Lu et al., 2003). The impact of temperature was also tested in this study. Very few works have been published in relation to the survival of IVs in the air and surfaces. A study showed that human IVs retained their infectivity 24 to 48 hours on smooth surfaces (stainless steel, plastic) but less than 8 to 12 hours in porous surfaces such as paper handkerchiefs, newspaper paper. It was also shown that virus deposited on human hand skin did not survive more than 5 minutes (Bean et al., 1982). The latter results are difficult to interpret because the kind of experiment done is difficult on a methodological viewpoint (virus recovery being one critical step).