Emergence of influenza: Expecting the unexpected

Introduction

Dr. Reginald G. Thomson was the founding dean of the Atlantic Veterinary College and had an impressive career (1). After studying veterinary medicine at the Ontario Veterinary College, he obtained his PhD at Cornell University and in the same year, 1965, became a diplomate of the American College of Veterinary Pathologists. Dr. Thomson was a distinguished researcher and prolific writer. He published more than 60 research papers, mainly on pathology of respiratory disease, was editor of the Canadian Journal of Comparative Medicine (now Canadian Journal of Veterinary Research), and wrote two textbooks, “General Veterinary Pathology” and “Special Veterinary Pathology.” The latter formed the basis of the current textbook “Pathologic Basis of Veterinary Disease” (2).

Furthermore, Dr. Thomson had an international vision for veterinary medicine. This is demonstrated by his links with universities in Kenya, Nigeria, and Iraq, his sabbatical leaves spent in Africa and Asia, as well as the rotations on international veterinary medicine and foreign animal diseases he helped establish at the Atlantic Veterinary College (1). Such an international vision is necessary to meet the current challenges to our society, in which the growth of the global human population on one hand, and the growth in average consumption per person on the other, have many far-reaching effects, such as climate change, water shortage, deforestation, depletion of fish stocks (3), and an increased rate of emergence and re-emergence of infectious diseases (4). The subject of this report is about just one of these emerging diseases, influenza. In this lecture, I have two objectives: one is to inform you of new and unexpected aspects of the versatile micro-organism, influenza A virus; the other is to show how veterinary medicine, and specifically veterinary pathology, can make important contributions to knowledge of diseases not only in domestic animals, but also in humans and wildlife.

Influenza A virus is a virus species belonging to the family Orthomyxoviridae (5). It is an enveloped, negative-strand RNA virus with a genome consisting of 8 segments. The segmented character of the genome allows reassortment, that is, if 2 viruses simultaneously infect the same cell, progeny viruses may be produced that have gene segments from both parent viruses. The envelope of the virus contains two surface antigens: the hemagglutinin, which is important for attachment of the virus to its target cell, and the neuraminidase, which is an enzyme that allows progeny viruses to be released from the surface of the cell that generated them. Influenza A viruses are categorized based on the subtype of their hemagglutinin (H1 to H17) and of their neuraminidase (N1 to N10).

The original reservoir for all influenza A viruses, with the exception of H17N10, originating from a bat species (6), are wild waterbirds (7). From this reservoir, influenza A viruses occasionally jump to other species, both domestic birds and mammals. Usually, this leads to individual cases of infection or short-lived epidemics, such as in harbor seals (Phoca vitulina). Rarely, the virus can maintain itself in the new host species, as is the case in domestic pigs, horses, and humans (8). In the last hundred years, such events occurred four times in the human population: H1N1 in 1918, H2N2 in 1957, H3N2 in 1968, and H1N1 in 2009 (9).

The emergence of H5N1 virus in humans, Hong Kong 1997

In 1997, highly pathogenic avian influenza (HPAI) virus of the subtype H5N1, which was causing an epidemic in poultry in Hong Kong, spread to humans, resulting in 6 deaths of 18 confirmed cases (10). This was unusual because this subtype had never been recorded in humans before, and, in the rare cases when avian influenza viruses had infected humans, they had never caused severe illness (11). Some scientists speculated that, as in poultry, this virus caused systemic infection in humans and therefore was so virulent (12). To address this question, we inoculated HPAI H5N1 virus into cynomolgus macaques (Macaca fascicularis) as a primate model for H5N1 influenza in humans, and performed both virological and pathological analyses. We found that HPAI H5N1 virus caused severe pneumonia in macaques, but did not spread to other organs. Based on those results, the hypothesis that H5N1 virus spread systemically in humans, as it did in chickens, was falsified (13,14).

The question remained how an avian influenza virus was able to directly infect a human. The dogma at that time was that this was not possible, because avian influenza viruses had a preference for cell receptors, α-2,3-linked sialic acids, that were absent from the human respiratory tract, which expressed α-2,6-linked sialic acids, the preferred cell receptor of human influenza viruses (15). It was then thought that domestic pigs, which were known to have receptors for both human and avian influenza viruses in their respiratory tract, were necessary as an intermediate “mixing vessel” for reassortant viruses with gene segments originating from both human and avian influenza viruses (16).

Until that time, virus-cell interaction had been studied mainly in the trachea, as a representative of the whole respiratory tract. We hypothesized that cell receptors differed according to the level of the respiratory tract. We tested this hypothesis by use of a sensitive method to determine the degree of influenza virus attachment to epithelial cells at different levels of the respiratory tract (17). By use of this so-called “virus histochemistry” technique, we found that HPAI H5N1 virus did not attach to human tracheal epithelium, which was consistent with the then current dogma. However, it was able to attach very well to respiratory epithelium in deeper parts of the respiratory tract: to Clara cells in the bronchioles, and to type II pneumocytes and alveolar macrophages in the pulmonary alveoli. Seasonal human influenza viruses, which are endemic in the human population, showed a nearly opposite binding pattern: abundant attachment to human tracheal epithelium, and lack of attachment to Clara cells, type II pneumocytes, and alveolar macrophages. This study proved that cells in the lower part of the human respiratory tract did have receptors for avian influenza viruses. In addition, it suggested that this pattern of attachment was responsible, in part at least, for severe pneumonia being the primary disease in human HPAI H5N1 virus infection (17). Other researchers working on the same question by use of a different technique, lectin histochemistry, arrived at the same answer (18).

We took this research one step further by examining the correlation between airborne transmissibility of influenza viruses in humans and pattern of virus attachment along the full length of the respiratory tract, including the nasal cavity. We found that all easily transmissible influenza viruses bind extensively to the ciliated epithelial cells of nasal cavity and trachea, whereas poorly transmissible influenza viruses do not. This is a strong indication that tropism for upper respiratory tract epithelium is one of the characteristics that an influenza virus needs to have in order to become easily transmissible among humans, and thus able to cause a pandemic (19).

H5N1 virus in felids, Thailand 2003

At the end of 2003 and beginning of 2004, when HPAI H5N1 virus had reached Thailand in its gradual but relentless spread across southeast Asia, mortality associated with feeding on infected chickens was reported in domestic cats, zoo leopards (Panthera pardus), and zoo tigers (Panthera tigris)(20). This was unusual, because felids were not known to become ill from influenza virus infection, and in fact experimental infections of domestic cats in the past had not shown any evidence of clinical disease (21). To study this unexpected development under controlled conditions, we inoculated HPAI H5N1 virus into domestic cats in the laboratory. By use of virological and pathological analyses, we showed that, regardless of the route of inoculation (intratracheal, by feeding on infected chicken, or by direct contact), the cats contracted a productive infection and developed a severe pneumonia (22). The results of this experiment showed that in one fell swoop, HPAI H5N1 virus had added a new family of mammals to the host range in which influenza A viruses cause disease.

Surprisingly, besides causing severe respiratory disease, nearly every extra-respiratory organ we examined had evidence of severe inflammation and necrosis, which colocalized with the expression of influenza virus antigen by immunohistochemistry (23). So, in contrast to the macaque experiment performed 4 years previously, HPAI H5N1 virus did cause systemic disease in domestic cats, the first time this had been shown in any mammal other than domestic ferrets and laboratory mice. This study placed the sporadic human case reports of H5N1 influenza not presenting with respiratory disease (24) in a different perspective. There was a well-documented case by Menno de Jong et al (25), who reported a child presenting with neurologic symptoms without any respiratory symptoms. An HPAI H5N1 virus was detected in cerebrospinal fluid from this child, who died in coma. Might HPAI H5N1 virus in humans spread outside the respiratory tract after all, and if so, how?

A partial answer to this question came from experimental HPAI H5N1 virus infections in ferrets. To our surprise, we found little evidence of pneumonia in ferrets inoculated by the intranasal route, which is a common method used in this species. Instead, there was abundant evidence of encephalitis (26). Other articles on experimental HPAI H5N1 virus infection in ferrets also stated that, despite being severely ill, ferrets had surprisingly mild pulmonary lesions, but did show neurologic signs (27–28). Together, these studies suggested that intranasal inoculation of H5N1 virus into ferrets was causing central nervous system disease rather than the respiratory disease it was supposed to be modelling. We subsequently performed a new experiment to determine how HPAI H5N1 virus was reaching the brain. This experiment showed that HPAI H5N1 virus, while not binding to respiratory epithelium in the nasal cavity, had a strong tropism for the olfactory epithelium, and from there travelled along the olfactory nerve, through the cribriform plate, and to the olfactory bulb, from which it spread to other parts of the brain per continuitatem(29). This study showed that HPAI H5N1 virus has, after the lower respiratory tract and the gastrointestinal tract, yet another route by which it can enter mammalian organisms. The olfactory route of entry might explain human cases, in which patients present with neurologic symptoms in the absence of respiratory disease (25). Another important conclusion from this study was that intranasal inoculation is not suitable in a ferret model for pneumonia from HPAI H5N1 virus infection, but that intratracheal inoculation should be used instead (26).

Potential role of wild birds in spread of H5N1 virus from Asia to Europe, 2005

While HPAI H5N1 virus had slowly been expanding across southeast Asia between 2002 and 2004, in 2005 it suddenly expanded rapidly to western China, and from there to central and western Europe, as well as the Middle East and Africa. Wild birds were clearly involved in this rapid expansion, but it was not clear whether they were innocent bystanders or whether they had an active part to play in spreading the virus over a long distance (30). Until then, wild birds had never been involved in the epidemiology of HPAI outbreaks (31).

To help resolve this question, we performed a series of experimental inoculations on wild duck species considered to be suspects as vectors of HPAI H5N1 virus: Eurasian wigeon (Anas penelope), gadwall (Anas strepera), common teal (Anas crecca), mallard (Anas platyrhynchos), tufted duck (Aythya fuligula), and common pochard (Aythya ferina). To test the hypothesis that these birds could excrete virus without showing evidence of disease, we inoculated them intratracheally and intra-esophageally with a low dose of virus, and followed them virologically and pathologically (32). There were three unexpected results: first, there was a neat divide in pathogenicity between diving ducks (Aythya spp.) and dabbling ducks (Anas spp.): diving ducks became ill and some died; dabbling ducks showed no clinical signs whatsoever. This emphasizes the saying by Gary Wobeser in his book on waterfowl diseases (33): “All ducks were not created equal.” Second, all species shed more virus from the throat than from the cloaca. By histological examination, we tried to find out the area from which they were shedding. Most likely, virus in the throat was coming from replication in the air sacs, which was the most frequently positive tissue with the highest titers. Virus in the cloaca was likely coming from liver and/or pancreas: no birds of any species had virus replication in intestinal epithelium. That was the third unexpected result, because intestinal epithelium is the main site of low pathogenic avian influenza (LPAI) virus replication (34), and the change from low to high pathogenic avian influenza virus is supposed to expand tissue tropism, not to switch it.

This work on HPAI H5N1 virus in wild ducks revealed gaps in our knowledge on the dynamics of LPAI virus infection and associated lesions in wild waterbirds. We were interested, therefore, to determine the pattern of infection and disease in naturally infected birds. The main difficulty in such a study is to select the birds that are naturally infected with LPAI virus. Low pathogenic avian influenza virus infection shows no known clinical signs, and even at the best of times, only a small percentage of any group of birds is infected. To get around this, we caught 50 mallards, swabbed them, and kept them for 24 h while samples were transported to the lab and the polymerase chain reaction (PCR) was run. Based on the PCR results, we selected 19 positive birds and 4 negative controls, necropsied them, and collected samples for virological and pathological analyses. The outcome of this study was that virus replicated mainly in the bursa of Fabricius, less in the intestinal epithelium, and nowhere else (35). Unexpectedly, none of the virus-positive tissues had any evidence of inflammation or necrosis despite detailed histopathological examination. Later, we performed a similar study on black-headed gulls (Chroicocephalus ridibundus), which gave similar results (36). These studies suggest that LPAI virus infection in wild birds is non-pathogenic. From an evolutionary point of view, it is interesting to know whether LPAI virus has any effects at all on the infected bird. For example, if it caused diarrhea, this might increase virus excretion and thus provide a selective advantage to the virus (37).

Conclusions

In summary, there have been substantial advances in our knowledge of influenza in recent years. Avian influenza viruses have been shown not only to be able to be transmitted directly from birds to humans, but also to cause severe and in some cases fatal disease. One unusual avian influenza virus, HPAI H5N1 virus, has been shown to be capable of entering the mammalian body via the digestive tract and the olfactory tract, in addition to the known route of entry via the respiratory tract. This virus also has been shown to be capable of causing severe disease outside the respiratory tract in multiple mammalian species, particularly felids and humans. Finally, we have increased our understanding of the effect of LPAI virus infection in the wild bird reservoir, and realized that the switch from low to high pathogenic avian influenza virus is more complex than merely an expansion of tissue tropism.

In addition, the studies I have described show that veterinary medicine in general, and veterinary pathology in particular, can form a strong research partnership with human medicine and wildlife ecology to increase our knowledge of emerging infectious diseases. I hope that this example may act as an encouragement to young veterinary graduates interested in pursuing a career in this multidisciplinary field of science. To quote the editor of The Lancet (38): “One thing is clear: given that all new infectious diseases of human beings to emerge in the past 20 years have had an animal source, veterinary science and animal husbandry are as important for disease control as clinical medicine.”