Lentivirus-induced Equine Anemia

The first retrovirus associated with disease was equine infectious anemia virus (EIAV), isolated as a filterable agent in 1904 (Vallee and Carre 1904). This virus induces a chronic recurrent anemia in horses and remains an important veterinary problem today. EIAV infections share some similarities with those involving other lentiviruses. An acute, self-limiting disease usually develops soon after EIAV infection and is followed by a chronic phase in which macrophages are the major reservoir of virus (McGuire et al. 1971; Rice et al. 1989; Sellon et al. 1992). In addition, infected horses contain a population of viruses with particular variants being prominent at different points postinfection (Fig. 19) (Kono et al. 1971; Montelaro et al. 1984; Salinovich et al. 1986; Payne et al. 1987; Kim and Casey 1992). Despite these parallels, EIAV-induced disease follows a unique course and involves pathogenic mechanisms that are distinct from those involved in other known lentivirus-induced diseases (for review, see Montelaro et al. 1993).

Figure 19

Course of EIAV-induced anemia in horses. The graph illustrates the successive waves of fever, emergence of viral variants (V1, V2, V3, V4), and the humoral immune response to the variants in an EIAV-infected horse. (Light color) Humoral (more...)

The acute phase of EIAV infection occurs within 1 month of infection and is marked by a short period of fever and viremia (McGuire et al. 1971; for review, see Montelaro et al. 1993). Very few animals succumb to this episode and most progress to the chronic stage of the infection. The chronic phase of EIAV disease lasts between 1 and 12 months and is marked by recurrent, increasingly severe, episodes of a debilitating, febrile, hemolytic anemia that correlate with the presence of high titers of circulating cell-free EIAV (Fig. 19) (McGuire et al. 1971; Issel and Coggins 1979; Orrego et al. 1982). Interactions between erythrocytes and EIAV particles, probably via the Env protein (Sentsui and Kono 1988), induce erythrocyte lysis in several ways. Complement coats EIAV-bound erythrocytes and is activated either directly or by anti-EIAV antibodies, causing lysis (McGuire et al. 1969a,b; Sentsui and Kono 1987a; Perryman et al. 1988). The coated cells can also be engulfed by macrophages (Sentsui and Kono 1987b). In addition to these effects, the virus appears to suppress the differentiation of erythroid precursors (Perryman et al. 1988; Swardson et al. 1992).

Most horses survive the recurrent bouts of anemia and become asymptomatic virus carriers 6–12 months after infection. This state usually lasts for the remainder of the animal's life, a very rare consequence of other lentiviral infections. Detecting virus directly in asymptomatic animals, including those in this final carrier phase, is difficult (McGuire et al. 1971; Rice et al. 1989; Sellon et al. 1992), but low levels of replication can be found using PCR-based approaches (Kim and Casey 1994). Consistent with these data, whole blood from inapparent carriers can transfer the infection efficiently (Coggins et al. 1972; Issel et al. 1982). The mechanisms that control the transition to the asymptomatic state are not understood, but an intact immune response is probably required (Kono et al. 1976; Perryman et al. 1988).

Very little is known about the viral determinants of EIAV pathogenesis. An infectious clone of the virus was difficult to obtain (Rushlow et al. 1986; Yaniv et al. 1986; Kawakami et al. 1987; Cunningham et al. 1993) and the one that is available does not induce disease in horses (Whetter et al. 1990). However, the unique mechanisms of pathogenesis and the intriguing ability of most infected horses to contain the infection makes this a particularly interesting disease for further study. Elucidating the mechanisms by which this is accomplished could provide clues to strategies that might control other lentiviral infections.

Lentiviral Infections of Sheep and Goats

Caprine arthritis encephalitis virus (CAEV) and VMV are two unrelated lentiviruses (Sonigo et al. 1985; Braun et al. 1987; Salterelli et al. 1990) that use similar pathogenic mechanisms to induce arthritis, pneumonia, mastitis, and CNS disease in goats and sheep, respectively (see Tables 5 and 6) (for review, see McGuire et al. 1990; DeMartini et al. 1993; Narayan et al. 1993). CAEV is most commonly associated with arthritis in adult goats (Cork et al. 1974; Crawford et al. 1980). This disorder shares many features with rheumatoid arthritis, and thus CAEV-infected goats provide a model system in which to study this important human disease. VMV usually causes pneumonia in sheep (DeMartini et al. 1993). These diseases arise 3–5 years after naturally occurring infections and follow a prolonged course in which a vigorous inflammatory response causes tissue damage. This response involves macrophages and CD8⁺ and CD4⁺ T cells (Cordier et al. 1992; Wilkerson et al. 1995a,b), and the site(s) of inflammation determines the type of disease that results. The neurological diseases associated with these infections occur because of a similar inflammatory response within the CNS.

These viruses replicate and persist in macrophages (Klevjer-Anderson and McGuire 1982; Narayan et al. 1982, 1983; Anderson et al. 1983; Gendelman et al. 1985, 1986; Cheevers et al. 1988; Zink et al. 1990). The macrophages found in inflammatory lesions are infected, and a subset of the cells express viral products (Staskus et al. 1991b; Brodie et al. 1992); whether these products initiate and sustain the inflammatory response is not known. Arthritic, CAEV-infected animals have very high titers of antibodies that react with the SU protein compared to healthy infected goats, and large numbers of B cells are found in the lesions (Johnson et al. 1983; Gogolewski et al. 1985; Knowles et al. 1990; Wilkerson et al. 1995a). In addition, although healthy infected goats contain CD4⁺ and CD8⁺ T cells that respond well to CAEV-encoded proteins, the response of cells from arthritic animals is suppressed (Perry et al. 1995). Whether similar events occur in the inflammatory lesions observed in VMV-infected sheep has not been fully investigated.

Investigators originally assumed that levels of viral expression were extremely low for long periods of time in animals infected with CAEV and VMV (for review, see Haase 1986). However, the high frequency with which variants are generated during these infections (Clements et al. 1982; Ellis et al. 1987; Stanley et al. 1987; Cheevers et al. 1991) suggests that these viruses, like other lentiviruses, undergo multiple cycles of replication (see Chapter 11. However, very low levels of cell-free virus are produced because replication occurs preferentially in tissue macrophages, not circulating monocytes (Narayan et al. 1983; Peluso et al. 1985; Gendelman et al. 1986; Gabudza et al. 1989). These differences in replication correlate with developmentally regulated expression of particular transcription factors during myeloid differentiation (Gabuzda et al. 1989; Shih et al. 1992).

Information concerning the viral determinants that mediate VMV- and CAEV-induced disease is scant. Multiple strains of both viruses exist and unknown characteristics influence the type of tissue involved in the disease (Querat et al. 1984; Lairmore et al. 1987, 1988; Cheevers et al. 1988; Roy et al. 1991; Staskus et al. 1991a). Infectious molecular clones of CAEV and VMV are available (Sonigo et al. 1985; Braun et al. 1987; Salterelli et al. 1990), but testing the pathogenic potential of large numbers of isolates is problematic because of the long latent period and the lack of an appropriate small animal model. In addition, as with other lentiviruses (Meyerhans et al. 1989), selective pressures applied during growth of these viruses in culture have probably altered their properties. Another complication inherent in these experiments is that additional variants will almost certainly be generated in the infected animals. This variation reflects selective pressures of the immune response and selection for viruses that have an inherent replicative advantage (for review, see Burns and Desrosiers 1994). It may be difficult to distinguish determinants that appear to be linked to pathogenicity because they allow efficient replication from those that stimulate the pathogenic inflammatory response.

Wasting Induced by Avian Retroviruses

Most avian retroviruses are usually thought of as oncogenic viruses (see above Oncogenesis). However, the major economic and agricultural problem associated with infection by several strains of ALVs and many strains of REV is an ill-defined wasting syndrome characterized by poor growth, anemia, and immunosuppression resulting from atrophy of the bursa and thymus (Mussman and Twiehaus 1971; Witter 1984; Barth and Humphries 1988). The disease occurs most commonly in young birds that have been infected with large doses of virus. The immunosuppressive effects of the virus appear to relate directly to thymic and bursal atrophy and may reflect the ability of these viruses to lyse infected cells (Keshet and Temin 1979; Weller and Temin 1981). However, analyses of one set of chimeric viruses indicate that both poorly and highly pathogenic strains infect the same cell types and replicate to similar titers in infected birds (Filardo et al. 1994). In addition, large differences in the amounts of viral protein expressed in the infected bursas were not observed, suggesting that differences in viral replication may not control wasting. Consistent with this interpretation, differences in pathogenesis could not be linked to differences in LTR sequences characteristic of the two isolates or to any particular viral gene (Filardo et al. 1994). These data suggest that wasting is induced by interactions involving several different viral gene products.

Anemia Induced by FeLV-C

Severe, FeLV-C-induced aplastic anemia affects more pet cats than any other retrovirus-induced disease (Hoover et al. 1974; Mackey et al. 1975; for review, see Hardy 1980). FeLV-C strains arise in naturally infected cats and are recombinants between endogenous FeLVs and the exogenous FeLV-A strain (see above Oncogenesis, Tumor Induction by Simple C-type Retroviruses That Lack Oncogenes). Sequences in the amino-terminal portion of the SU protein found in the prototype FeLV-C/Sarma strain and other FeLV-C isolates that induce anemia are linked to disease induction (Reidel et al. 1986, 1988; Dornsife et al. 1989; Rigby et al. 1992). These sequences include those that determine host range, but there is no evidence suggesting that anemogenic strains and closely related nonpathogenic strains infect different types of hematopoietic cells (Brojatsch et al. 1992; Dean et al. 1992; Rigby et al. 1992). Although env sequences have an important role in anemia induction, additional, unmapped viral determinants are also important for rapid induction of anemia (Rigby et al. 1992). Anemic cats contain reduced numbers of BFU-E and CFU-E, and FeLV infection suppresses the generation of these precursors in vitro (Onions et al. 1982; Testa et al. 1983; Wellman et al. 1984; Rojko et al. 1986; Abkowitz et al. 1987a,b). Anemia-inducing strains replicate particularly well in macrophages, and the infected cells secrete high levels of tumor necrosis factor α, which may suppress hematopoiesis in the infected animals (Khan et al. 1993). However, whether these mechanisms have key roles in disease induction has not been addressed fully.

Osteopetrosis

Several isolates of ALV and MAV induce osteopetrosis, a disorder in which the osteoblasts of the long bones grow abnormally (for review, see Smith 1982; Payne 1992). As a consequence, affected chickens are stunted, walk with an abnormal gait, and have thick shank bones (Fig. 20). Outgrowth of bone tissue does not reflect a malignant process and is polyclonal in nature (Robinson and Miles 1985). The infected cells accumulate high levels of unintegrated DNA and large amounts of viral proteins (Robinson and Miles 1985; Aurigemma et al. 1989; Foster et al. 1994). Accumulation of large amounts of viral DNA has been linked to failed mechanisms of superinfection resistance and cytolysis in other instances (Weller and Temin 1981; Donahue et al. 1991; Kristal et al. 1993; Reinhart et al. 1993; see above Retrovirus-induced Immuno-deficiencies, FAIDS). However, osteoblasts infected with pathogenic ALV establish superinfection resistance in vitro, suggesting that other mechanisms contribute to this aspect of osteopetrosis (Foster and Robinson 1994). Analyses of chimeric viruses suggest that induction of osteopetrosis by ALV is influenced by sequences at the 5′ end of the gag gene (Robinson et al. 1986, 1992b). In contrast, sequences within the env gene appear to play a part in osteopetrosis induced by MAV isolates (Joliot et al. 1993). Proposing a model that accounts for all of these features requires further experimentation and a better understanding of the ways in which infection interferes with bone development.

Figure 20

Chickens afflicted with osteopetrosis. (Top) A normal chicken (left) and a chicken suffering from osteopetrosis (right). (Courtesy of H. Robinson.) (Bottom) Skeleton of a chicken with ALV-induced osteopetrosis (left) and a (more...)