Avian influenza virus

Avian influenza is a RNA virus. Its RNA is single-stranded, segmented, and of negative-sense. Its genetic material is composed of eight segments of single-stranded RNA. These eight segments, which are actually eight genes, encode 10 proteins of the virus (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). The eight segments are surrounded by an envelope, which is covered by haemagglutinin (HA), and neuraminidase (NA) glycoproteins. Avian influenza viruses have 16 haemagglutinins and nine neuraminidases (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). Each virus has one H and one N antigen subtype in any combination, creating 144 combinations. Most of the possible combinations have been isolated from avian species (Alexander, Reference ALEXANDER2008).

The virus of avian influenza occurs in two forms: highly pathogenic avian influenza virus (HPAI) and the low pathogenic avian influenza virus (LPAI). The HPAI virus is highly virulent and is capable of causing mortality in birds up to 100% within a few days. The LPAI virus is mild and mortality is usually less than 5% (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008; Vegad, Reference VEGAD2008). A characteristic feature of avian influenza viruses is that they are dynamic and continuously evolving. They can change in two ways: via antigenic drift and/or antigenic shift (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008), although, while antigenic drift changes occur all the time, antigenic shift is rare (Wikipedia, 2012a).

Antigenic drift

When a DNA virus replicates, its adenine must pair with thymine; and cytosine with guanine. If the correct pairing does not occur, an exact copy of the virus cannot be produced. The virus, by using its corrective, ‘proofreading’ mechanism (enzyme exonuclease), tries to rectify the mistakes, whereby the wrong base is cut and removed, and in its place a correct base is inserted. RNA viruses make a lot of wrong base pairings during replication due to very poor corrective mechanisms (Turner et al., Reference TURNER, MCLENNAN, BATES and WHITE2005; Wikipedia, 2012a) and mistakes persist. This wrong pairing of a base is a mutation, and since it occurs only at one point (i.e. only one base is replaced in the gene), the change is called a ‘point mutation’. Mutation of one base can change only one amino acid, so antigenic drift therefore causes small, gradual changes.

Recently, Lu et al. (Reference LU, SHOU-CHUAN, JUN, CHUAN-QING and ZE-LIN2011) reported that low pathogenic H9N2 subtype avian influenza viruses in China have the tendency to evolve, resulting in the occurrence of drift variants which have higher virulence and variable antigenicity. The highly virulent mutants may cause death in chickens and as well as economic losses. It has been reported that antigenic drift occurs in the genes of haemagglutinin and neuraminidase (Wikipedia, 2012a), which makes the protein of haemagglutinin unrecognisable to pre-existing host immunity. Haemagglutinin is the major antigen that elicits antibodies which protect birds against death and clinical signs (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). Accumulation of mutations within the antibody-binding sites changes viruses. The resulting viruses therefore cannot be inhibited by antibodies made against previous strains. This makes it easier for transmission throughout a partially immune population (Wikipedia, 2012a).

Cattoli et al. (Reference CATTOLI, MILANI, TEMPERTON, ZECCHIN, BURATIN, MOLESTI, ALY, ARAFA and CAPUA2011) reported that the HPAI H5N1 virus has been endemic in poultry in Egypt since 2008, despite mass vaccination and culling protocols. The same authors stated that extensive circulation of the virus has resulted in a progressive genetic evolution and an antigenic drift. In poultry, occurrence of antigenic drift is less well documented, and the mechanisms remain to be clarified. To test whether H5N1 drift is driven by mechanisms similar to type A influenza viruses in humans, reassortment viruses have been generated by reverse genetics that harboured molecular changes identified in genetically divergent viruses circulating in vaccinated population. The results indicated that antigenic drift of H5N1 is driven by multiple mutations primarily occurring in major antigenic sites of the haemagglutinin molecule similar to what has been described for human influenza H1 and H3 subtype viruses.

Emergence of drift variants in India

While the HPAI virus has effectively been brought under control, LPAI virus has kept circulating and became endemic in India. In the intervening six years (2006-2012) of numerous point mutations are expected to have occurred and accumulated in LPAI, will have induced changes in the amino acid composition of haemagglutinins (and hence reduced recognition by immune processes). This will have increased virulence, producing drift variants as observed by Lu et al. (Reference LU, SHOU-CHUAN, JUN, CHUAN-QING and ZE-LIN2011) in China. What India faced in the summer of 2012 appeared to be the outbreak caused by drift variants of LPAI viruses, since the disease did not conform to any other prevailing condition. The disease in India was very severe in certain states, but mild in others. The variation may depend on several factors, including the amount of variant virus present in the area, the number of mutations it has undergone, and the degree of virulence acquired.

Clinical signs

The general pattern observed, particularly in broilers, was that healthy birds suddenly fell sick, made abnormal respiratory sounds, had severe breathing difficulties, and, while struggling to breathe, died within a few hours. The mortality varied from 5% to 100%, in certain cases. Mortality began suddenly and increased over a period of 10-20 days, but there was no fixed period or pattern. Deaths occurred in all groups - broilers, layers, and breeders, and young or adult birds. In layers, egg production declined.

In cases where disease was very severe and birds suddenly died (peracute cases), no lesions were seen. In acute cases, lesions in broilers were confined to trachea, bronchi and airsacs, and only occasionally seen in the lungs. Trachea samples presented severe haemorrhagic tracheitis, occasionally filled with blood. Necrosis and massive tissue destruction lead to the formation of caseous deposits mainly in the two primary bronchi, and sometimes in the trachea. These were the most consistent lesions and almost pathognomonic. No changes were seen in other organs. In layers, salpingitis and egg peritonitis were the most common findings. Lesions in the respiratory tract were less common and many times absent.

The history of acute disease exposure, characteristic clinical symptoms, heavy mortality, and almost pathognomonic lesions all pointed to infection by the drift variants of LPAI.

Differential diagnosis

Disease produced by the drift variants differed from the following:

1. 1. Highly pathogenic avian influenza (HPAI): HPAI is a severe disease. Its first sign is the sudden onset of high flock mortality which may go up to 100% within in a few days. Respiratory signs are less prominent than with LPAI viruses (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). Being highly virulent, HPAI virus affects almost all organs of the bird. As a result, on post mortem, haemorrhages and necrotic lesions are present in the liver, spleen, heart, kidney, and lungs. Haemorrhages are prominent on heart, breast muscles, and mucosa of the proventriculus and gizzard (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). With drift variants, lesions were confined to the respiratory tract. No lesions were seen in other organs. Moreover, mortality spread over many days and rarely reached 100%.

2. 2. Newcastle disease (ND): This disease has existed in India for past 85 years ever since it was discovered in 1927. The bird's immune system is therefore familiar with the antigens of its virus and in India, almost every farmer carries out ND vaccination, hence some degree of acquired immunity is always present. Moreover, the genome of ND virus is stable and does not undergo antigenic drift like the avian influenza virus. It is therefore unlikely that ND virus may have become highly virulent to inflict huge mortalities in chickens, across the country. Furthermore, ND lesions in the respiratory tract are characterised by congestion of trachea and some mucosal haemorrhages. Caseous plugs completely blocking the two primary bronchi, consistently and prominently seen with drift variants, seldom occurs in ND.

3. 3. Infectious bronchitis (IB): Infectious bronchitis is widespread in India, but mortality is generally negligible in the uncomplicated disease, except in very young chicks. Mortality in IB is mostly limited to chickens less than six weeks of age and usually is insignificant thereafter (Cavanagh and Gelb, Reference CAVANAGH and GELB2008). This was not the case with drift variants, in which mortality was very severe in all age groups. Moreover, lesions caused by IB consist of mild haemorrhagic tracheitis, and only rarely show caseous exudate. In contrast, complete obstruction of both primary bronchi with caseous plugs was a regular feature with drift variants, and was almost pathognomonic.

4. 4. Infectious laryngotracheitis (ILT): Presence of ILT in India has been suspected in certain areas, but not confirmed. Eradication of ILT virus, unlike drift variants, is possible due to antigenic stability of its genome. Since ILT virus strains are antigenically homogeneous, a single ILT virus vaccine produces cross-protective immunity for all ILT virus strains (Guy and Garcia, Reference GUY and GARCIA2008).

5. 5. Mycoplasmosis (Mycoplasma gallisepticum infection): The characteristic lesions in mycoplasmosis are airsacculitis and presence of caseous exudates in air sacs, but not in trachea or bronchi, as was the case with drift variants. Typical lesions and mortality pattern of drift variants rule out mycoplasmosis to be the cause of deaths.

6. 6. Infectious coryza: Infectious coryza is characterised by low mortality, and its main clinical sign is facial oedema. Lesions are confined to nasal passages and sinuses. Lesions of drift variants involving bronchi are conspicuously absent.

In conclusion, from the differential diagnosis presented, possibility of other respiratory diseases as cause of extensive and heavy mortality in the 2012 outbreak in India appears unlikely. It was concluded that the underlying cause appears to be the drift variants of LPAI viruses. However, this conclusion is based entirely on the field observations and their scientific interpretation. The exact picture will not be known until investigations are carried out using more sophisticated biotechnological techniques in any future outbreak.

The cytokine storm

When a bird's immune system detects a microorganism with new proteins (antigens), it recognises the infectious agent as a new threat (Wikipedia, 2012b). It overreacts to destroy the organism, and in doing so, kills the bird from what is known as a ‘cytokine storm’ (Us, Reference US2008; Avian influenza mutation, H5N1, 2012). Cytokine storm is the fierce and lethal interplay of cytokines (Johnson, Reference JOHNSON2005).

The drift variant of LPAI is a new virus strain, because it has acquired new proteins through mutations. When bird's immune system recognises these new proteins, it overreacts to destroy the virus and protect the bird. The immune system sends cells to the site of infection, which secrete ‘cytokines’ (Johnson, Reference JOHNSON2005; Wikipedia, 2012b). Cytokines are proteins, and when released at right time in proper amounts, cytokines can help fight infections and regulate immune processes. But many cytokines are inflammatory and damaging to the body if present in too high levels, or for too long (Avian influenza mutation, H5N1, 2012). The cytokine storm is a potentially fatal immune reaction (Wikipedia, 2012b).

Cytokines released are in two forms: those that promote inflammation such as tumour necrosis factor (TNF)-alpha, interleukin (IL)-1, interleukin (IL)-6, interferon (INF)-gamma, and several others (pro-inflammatory cytokines). Their purpose is to destroy infectious agents, and to check and control inflammation such as interleukin-10 and interleukin-1 receptor antagonist, so that there is no overreaction, i.e. anti-inflammatory cytokines (Johnson, Reference JOHNSON2005; Avian influenza mutation, H5N1, 2012). As microbes evolve, they may produce a mix of toxins that interfere with the control mechanisms of the immune system. This seems to be the case with certain strains of avian influenza. The H5N1 virus is not only partially resistant to cytokines, but it reduces the production of anti-inflammatory cytokines. In essence, it enhances acceleration of the disease while impairing the control systems until the immune system goes out of control and crashes (Avian influenza mutation, 2012).

When the immune system is fighting avian influenza virus, cytokines signal immune cells such as T-lymphocytes and macrophages to travel to the site of infection (Johnson, Reference JOHNSON2005; Wikipedia, 2012b). In addition, cytokines activate these cells, stimulating them to produce more cytokines. In some instances, the reaction becomes uncontrolled as too many immune cells are activated at one place (Wikipedia, 2012b). The cytokine storm induced by H5N1 is characterised by extreme production and secretion of large numbers and excessive levels of pro-inflammatory cytokines. This leads to their increase in the blood - hypercytokinaemia (Us, Reference US2008; Avian influenza mutation, H5N1, 2012). Cytokine storm results in the release of more than 150 known inflammatory mediators (cytokines, oxygen free radicals, and coagulation factors). Increased levels of cytokines induce an overactive inflammatory response (Johnson, Reference JOHNSON2005; Us, Reference US2008; Avian influenza mutation, H5N1, 2012).

Recently, it has been uncovered that avian influenza viruses (H5N1) are very strong inducers of various cytokines from macrophages and CD8+ T-lymphocytes, which increase inflammation (Us, Reference US2008). Further, it has been revealed that mutations in the genes of haemagglutinin and neuraminidase are responsible for the cytokine storm. It has been shown that changes in the haemagglutinin and neuraminidase glycoprotein increase immune-mediated damage caused by the highly virulent H5N1 strains (Us, Reference US2008). It is suggested that the drift variants of LPAI may act likewise and kill the bird through cytokine storm.

The immediate cause of death in the Indian outbreak

The exaggerated inflammatory response was limited to the respiratory tract, mostly in trachea and bronchi. Necrosis, and massive tissue destruction, led to formation of large caseous deposits in the two primary bronchi, and occasionally, in trachea. This resulted in their complete blockage. The bird was then unable to breathe, struggled for air, and ultimately died from asphyxiation. Thus, suffocation was the immediate cause of death, apparently brought about by the cytokine storm.

Pus in the samples from birds is semi-solid and caseous because of the presence of an anti-enzyme in their serum (Vegad, Reference VEGAD2007). This enzyme does not allow proteolytic enzymes of bird's heterophils to liquefy the pus. It can be concluded that caseous deposits in bronchi, presumably caused by the drift variants, were due to intense suppurative inflammation and massive tissue destruction, presumably induced by the cytokine storm.

Prevention: limitations of vaccination

If vaccines are available against avian influenza virus, they can help impart immunity and protection. However, protection will occur only when antigens of the vaccine virus and those of the field virus strain are the same. that is, the same virus-subtype in both. This is because protection is virus subtype specific (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). Birds are susceptible to infection with avian influenza viruses belonging to any of the 16 haemagglutinin subtypes, and there is no way to predict their exposure to any particular subtype (Swayne and Halvorson, Reference SWAYNE and HALVORSON2008). Thus, if the challenge virus strain is of a different subtype than antigens in the vaccine, protection does not occur.

Li et al. (Reference LI, YU, TIAN, YU, LIU, JING, PING and CHEN2005) reported that in mainland China, low pathogenic H9N2 avian influenza viruses caused severe economic losses in poultry industry. An inactivated vaccine derived from an early isolate has been used in layers and breeders in chicken farms since 1998. However, virus infection still occurred, sometimes even in the vaccinated flocks. To find answers, their studies involved sequencing of the 27 isolated viruses; and investigations on their antigenic variation, replication and pathogenicity in chickens and mice. Results revealed that the vaccine could not completely prevent replication and shedding of viruses. They concluded this was because most of the viruses isolated were antigenically different from the vaccine strain. This was similar to the observation in the H5N2 avian influenza virus in Mexico, where the commercial vaccine was not able to prevent virus shedding when chickens were challenged with antigenically different isolates (Lee et al., Reference LEE, SENNE and SUAREZ2004). Taken together, Li et al. (Reference LI, YU, TIAN, YU, LIU, JING, PING and CHEN2005) concluded that it is therefore necessary to evaluate how well prevailing isolates match the present vaccine and update the vaccine formulation accordingly. Recently, Lu et al. (Reference LU, SHOU-CHUAN, JUN, CHUAN-QING and ZE-LIN2011) investigated characteristics of pathogenicity and HA antigenic variations in H9N2 subtype avian influenza viruses isolated in China from 1988 to 2008. On the basis of their findings, they concluded that the antigenic mutants (drift variants) may evade immunity, which poses a new challenge to the immune prevention for the H9N2 avian influenza viruses. More recently, Sun et al. (Reference SUN, PU, FAN, SUN, WANG, ZHANG, LIU and LIU2012) observed that despite the long-term vaccination programme implemented in China, H9N2 avian influenza viruses continue to persist in chicken population, even in vaccinated flocks. The results of their studies indicated that H9N2 avian influenza viruses underwent antigenic drift and evolved into distinct antigenic groups. They concluded that persistence of H9N2 avian influenza viruses in China might be due to incomplete protection, and that the vaccine should be regularly evaluated and updated to maintain optimal protection.

Recent studies from the Republic of Korea reported similar findings. Park et al. (Reference PARK, KWON, SONG, PASCUA, BAEK, LEE, JANG, LIM, MO, MOON, KIM and CHOI2011) investigated whether currently circulating H9N2 avian influenza viruses had evolved in Korean poultry since 2007. Genetic and phylogenetic analysis revealed that reassortment viruses had emerged, generating at least five novel genotypes. Serological analysis showed that the isolates were antigenically distinct from previous Korean H9N2 viruses. On the basis of these results, they concluded that H9N2 viruses from domestic poultry have undergone substantial evolution since 2007 as a result of pressure from vaccines and natural immunity. Evolution was found to be more rapid following vaccination. They therefore suggested that periodical updating of vaccine strains based on continuous surveillance data is important to provide sufficient protection against continuous avian influenza infections.

The limitations of any vaccination programme to control drift variants are two-fold. Firstly, protection will occur only when antigens of the vaccine virus and those of the challenge virus strain are the same. Secondly, since avian influenza viruses undergo rapid evolution following vaccination, its vaccine has to be regularly evaluated and updated, based on continuous surveillance data to ensure optimal protection.

Acknowledgements

I thank Dr. Sandip Ther, Dr. Rakesh Sharma and Mr. Satish Shukla for their help.