Influenza is a highly contagious, usually self-limited, febrile respiratory tract infection caused by infection with influenza A or B virus and characterized by the abrupt onset of fever, myalgias, and malaise. Winter outbreaks are also typical. Recurrent outbreaks of influenza have probably occurred since antiquity, but the virus was first isolated in 1933 (Smith et al., 1933). Influenza A viruses infect a wide variety of animals, including humans, horses, pigs, ferrets, and birds. Annually, 720,000 people die from influenza and pneumonia in the US alone (Centers for Disease Control and Prevention, 1997).
Influenza A viruses are classified into subtypes based on two surface antigens, hemag-glutinin (H) and neuraminidase (N). Three of the 15 known subtypes of hemagglutinin (H1, H2, and H3) and two of the 8 known subtypes of neuraminidase (N1 and N2) are recognized as being capable of causing outbreaks of disease in man. Influenza A virus has the unique property of being able to undergo extensive antigenic variation in the viral hemagglutinin and neuraminidase antigens, so that annual outbreaks of influenza occur. Antigenic drift refers to the relatively minor changes in the influenza hemagglutinin and neuraminidase that occur each year resulting from mutations in the RNA segment coding for these proteins. Antigenic shift refers to a major change in influenza A subtype (e.g., from H2N2 to H1N1). This has occurred every 10 to 40 years and results from genetic reassortment (i.e., introduction of a new RNA segment encoding influenza hemagglu-tinin and/or neuraminidase). The current thinking is that contemporary human viruses have arisen by genetic reassortment between previous human and nonhuman influenza viruses. This reassortment is believed to occur when swine (or another host animal) are simultaneously infected with both an avian and a mammalian influenza virus. When this reassortment results in a virus with new surface proteins, it spreads rapidly because most of the population has no protective serum antibody (Webster et al., 1992).
There are currently only two options available for the control of influenza: vaccination with a killed-virus vaccine and therapy with antiviral drugs. Further studies on the immu-nopathogenesis of influenza may therefore not only further the understanding of basic immunology, but could also contribute to the control of a significant human pathogen.
There are several mammalian models of influenza: ferrets, monkeys, pigs, horses, and mice. Each model has particular advantages and disadvantages.
Ferrets can be naturally infected with human influenza isolates, develop a fever, and exhibit infection primarily of the nasal turbinates, with lesser amounts of infection of the trachea, bronchi, and lungs (Smith and Sweet, 1988; Sweet et al., 1981). Serial sampling of the same animal for virus and antibody is possible and the course of the disease is quite similar to that in humans;
however, ferrets are fairly expensive (~$100 each with $3 to $4/day board), reagents for assessment of cellular and mucosal immunology are not readily available, and they should not be used if assessment of pneumonia is a desired outcome.
Squirrel monkeys can also be infected with human isolates and develop a mild clinical disease. Although they are more closely related to humans than the other species, this advantage is negated by their expense (~$1000 each, $4 to $5/day board) and lack of commercially available reagents.
Swine influenza naturally occurs in outbreaks from late autumn to early spring. After a 1 to 3 day incubation period, there is a sudden onset of coughing, dyspnea, fever, anorexia, inactivity, prostration, and piling (huddling together). There is weight loss, but a generally low mortality rate (<1%) and the pigs begin to recover after 5 to 7 days (Easterday and Hinshaw, 1992). The pathology of the viral pneumonia is essentially identical to human cases (Urman et al., 1958). It is also a disease of major economic importance to the pig industry. Pigs can be readily infected by human isolates. The disadvantages of using pigs as an influenza model include their cost ($1/pound purchase price, $5 to $10/day board) and a limited number of Biosafety Level 2 facilities that can handle these animals.
Equine influenza is another naturally occurring disease of large animals. Unlike strains that infect humans, there is relatively little antigenic shift in equine influenza isolates; only two subtypes of equine influenza virus have been isolated from horses—H3N8 and H7N7—and only the former has been isolated in outbreaks in the past two decades. As with pigs and humans, there is a several day incubation period, followed by fever, cough, anorexia, and nasal discharge (Wilson, 1993). This model is also limited by its cost (~$500/pony, $1000+/horse, $10 to $15/day board), availability of reagents, and the number of Biosafety Level 2 facilities for large animals. Technicians trained in the care and handling of these animals, along with an equine medicine veterinarian, are also needed to assist with clinical problems that arise.
Mice are clearly the animal model of choice to study influenza virus infections. Following an intranasal inoculation of influenza virus, mice develop a progressive upper and lower respiratory tract disease with histopathology virtually identical to that seen in human disease. Further, mice are small, easily handled, rela tively short-lived, and are fairly inexpensive (~$10 each, $0.25/day board). They have a well characterized immunological response with a wide variety of inbred strains available and an increasing selection of genetically altered mice. Though most of the protocols in this unit were derived using BALB/c (H-2d) mice, many different strains have been used. Further, influenza is the only major human infection in which both immunogenicity and protection against infection can be dissected for mucosal IgA antibody, serum IgG antibody, and cellular immunity. Much that is known about the details of influenza pathogenesis and host defense were first established in mice and later confirmed in man (Bender and Small, 1992; Small, 1990). The mouse model has established the following.
Serum IgG antibody prevents lower respiratory tract infection. Passive transfer of serum antibody protects mice from a lethal challenge and prevents viral replication in the lungs (Loosli et al., 1953). This is primarily due to anti-hemagglutinin activity (Virelizier, 1975), although anti-neuraminidase (Schulman et al., 1968) and anti-matrix (Treanor et al., 1990) antibodies may also contribute to this protection.
Secretory igA antibody prevents upper respiratory tract infection. Intravenously administered polymeric IgA (pIgA) is actively transported into nasal wash fluid and protects the nose from challenge with the homologous virus (Renegar and Small, 1991b). Administration of anti-IgA antibody (anti-antibody) via nose drops to immune animals ablates this immunity, thus proving that IgA is the host defense mechanism for the nose (Renegar and Small, 1991a). Most secretory IgA is also anti-hemag-glutinin (Tamura et al., 1990; Tamura et al., 1991). Another important characteristic of antiinfluenza mucosal IgA is that it is more broadly cross-reactive than IgG against different influenza serotypes (Tamura et al., 1991; Tamura et al., 1992; Waldman et al., 1970).
Recovery from influenza infection is primarily, but not exclusively, mediated by CD8+ T-cells. As with most viral infections, recovery from influenza infection is a complex phenomenon (Doherty et al., 1997; Gerhard et al., 1997). Nude (athymic) mice do not recover from an influenza infection unless their immunity is enhanced with the adoptive transfer of anti-influenza CTL (Wells et al., 1984). Transfer of cloned anti-influenza CTL can also prevent death from a lethal virus challenge (Lukacher et al., 1984). If a high titer of antiinfluenza antibodies are injected into influ-
Animal Models For Infectious Diseases
Was this article helpful?