Host Genes that Influence Pathogenesis

Fv1

The Fv1 gene was identified using the susceptibility of inbred mice to FV-induced erythroleukemia as a screen (Odaka and Yamamoto 1965; Lilly 1967). The Fv1 locus has two principal codominant alleles called Fv1ⁿ and Fv1^b. Animals or cells carrying the Fv1ⁿ allele are permissive for a subset of ecotropic murine retroviruses which are called N-tropic; those carrying the Fv1^b allele are permissive for B-tropic viruses. Heterozygous Fv1ⁿ/b animals restrict replication of both types of viruses. Some murine viruses such as Mo-MLV are classified as NB-tropic because they replicate in both Fv1ⁿ and Fv1^b cells. Fv1 affects an early step in the viral life cycle related to integration or events just preceding integration (Jolicoeur and Baltimore 1976; Sveda and Soeiro 1976; Jolicoeur and Rassart 1980; Yang et al. 1980; see Chapter 5. The Fv1 gene product appears to be related to an endogenous gag sequence (Best et al. 1996). It interacts with determinants on the gag-encoded CA protein (Rommelaere et al. 1979; Boone et al. 1983; DesGroseillers and Jolicoeur 1983; Ou et al. 1983) and probably affects the function of the preintegration complex. Fv1 restriction can be overcome by infecting cells with high titers of virus (Hartley et al. 1970; Pincus et al. 1971; Boone et al. 1989), but the restriction is sufficient to have drastic effects on pathogenesis, particularly by viruses that must replicate extensively to induce disease (Mayer et al. 1978). Restriction mechanisms similar to those mediated by the Fv1 locus have not been identified in other animals, suggesting that effects on the integration of particular viral strains usually do not influence disease induction by nonmurine retroviruses.

Fv4 and Rmcf

The Fv4 or Akvr1 locus (Suzuki 1975; Gardner et al. 1980; Odaka et al. 1981) was identified based on the ability of the Fv4^r allele to restrict disease induction by blocking infection with the ecotropic MLVs. Additional analyses revealed that Fv4 (Akvr1) corresponds to an endogenous ecotropic provirus (Ikeda and Odaka 1984; Ikeda et al. 1985; Dandekar et al. 1987). The glycoprotein encoded by this provirus blocks the ecotropic viral receptor, thereby blocking infection and disease induction. Inheritance of this provirus is largely responsible for controlling spread and disease induction by ecotropic viruses in the Lake Casitas wild mice (Gardner et al. 1980, 1991), highlighting the part such genes can play in controlling naturally occurring infections.

A second gene exerting an effect similar to that of Fv4 is Rmcf (Hartley et al. 1983). Animals carrying the dominant Rmcf^R allele do not generate MCF viruses because the expression of a polytropic env gene-encoded glycoprotein blocks the viral receptor and interferes with replication of polytropic viruses (Ruscetti et al. 1981; Bassin et al. 1982). The Rmcf locus probably corresponds to an endogenous provirus (Buller et al. 1987; Frankel et al. 1989; see Chapter 8; however, it may also restrict infection by regulating the expression of one or more endogenous proviruses. Effects of the Rmcf locus are more apparent in cells of the erythroid and myeloid lineages than in the lymphoid lineage (Buller et al. 1988, 1989), and Rmcf does not affect induction of lymphoid tumors by Mo-MLV (Brightman et al. 1991). This difference suggests that the Rmcf product is more prominently expressed in erythroid and myeloid cells as compared to lymphocytes. This locus also affect CNS disease induced by a variant Fr-MLV that has a polytropic host range (Buller et al. 1990).

Any animal that expresses endogenous viruses could control infection in a manner analogous to that observed for Fv4 and Rmcf. However, the chicken is the only other animal known to use this mechanism to modulate viral infection. Some of the chicken ev loci, corresponding to proviruses belonging to the E envelope subgroup (see Chapter 8, prevent infection with other subgroup-E viruses (Robinson et al. 1981). This restriction does not influence disease in chickens because subgroup-E viruses are usually nonpathogenic (Motta et al. 1975). Two other ev loci, ev2 and ev3, protect birds from infection with subgroup-A ALV (Crittenden et al. 1982, 1984), but the mechanism by which this control is mediated remains unknown.

Receptor Expression and Other Mechanisms

An additional way in which viral replication can be controlled is at the level of viral receptor expression. In the chicken, a gene called tva encodes the receptor for subgroup-A viruses, and two other loci, tvb and tvc, control the expression of the receptors for subgroups B through D (see Chapter 3. Clearly, inheritance of these genes influences the ability of ALVs belonging to the different subgroups to infect chickens and induce disease. Theoretically, differential expression of viral receptors could control infection in other hosts. AKR and NZB mice carry genes that restrict replication of Cas-Br-E MLV in the CNS but not in lymphoid organs or the vasculature of the CNS (Hoffman and Morse 1985; McAtee and Portis 1985). This phenomenon might be partly controlled by changes in receptor expression or differential expression of endogenous viruses that could block the viral receptor. However, no evidence supporting either of these mechanisms has been reported.

Fv2

One of the well-characterized genes in this group is Fv2 (Lilly 1970; Odaka 1970). This locus affects the susceptibility of mice to erythroleukemia by influencing the frequency and cell cycling of late BFU-E and CFU-E, two classes of erythropoietin-sensitive erythroid precursors (Suzuki and Axelrad 1980; Behringer and Dewey 1985). The product of the locus is unknown. Genetic mapping has excluded the erythropoietin receptor (Hoatlin et al. 1990). However, several lines of evidence suggest that the product may be a part of this receptor complex. The effects of Fv2 are evident only when animals are injected with viruses that affect erythropoietin-sensitive cells, including a retrovirus expressing Epo itself (Hoatlin et al. 1990; Longmore and Lodish 1991). In addition, a retroviral vector expressing a constitutively active form of the erythropoietin receptor causes disease only in Fv2^s mice (Hoatlin et al. 1990). Recent work has identified mutant strains of FV-P that overcome Fv2-mediated resistance. The gp55-like protein encoded by these mutants lacks a portion of the ecotropic-specific region and interacts with the erythropoietin receptor in a fashion that bypasses events involving the Fv2-encoded molecule (Majumdar et al. 1992; Kozak et al. 1993; Hoatlin et al. 1995). Thus, the Fv2 product may be complexed with the Epo receptor in such a way that it blocks the ability of gp55 to interact with the Epo receptor in Fv2^r mice.

W and Sl

The W and Sl loci affect erythroleukemia induction by altering the availability of FV target cells (Bennett et al. 1968; Steeves et al. 1968; MacDonald et al. 1980). The W locus corresponds to the c-kit proto-oncogene (Chabot et al. 1988; Tan et al. 1990). This gene encodes a receptor PTK that binds to stem cell factor (SCF) or kit ligand, a molecule required for differentiation of early hematopoietic cells, particularly those of the myeloid lineage. Stem cell factor is encoded by the Sl locus (Anderson et al. 1990; Copeland et al. 1990; Huang et al. 1990). Thus, mutations affecting either the receptor or its ligand disrupt normal hematopoiesis and reduce the frequency of FV target cells. These mutations do not affect induction of lymphoid tumors because the lymphoid arm of hematopoiesis is normal in animals carrying these mutations.

Genes Affecting Lymphoid Development

Mutations affecting lymphoid cell development also influence disease induction. nu/nu mice do not develop thymic lymphomas because they lack a normal thymus and the normal complement of T cells (for review, see Reth 1995). These animals are also resistant to MAIDS (Mosier et al. 1987) and MMTV-induced mammary tumors (Tsubura et al. 1988), diseases in which B cell– T cell interactions have a critical role (see sections on Oncogenesis and Retrovirus-induced Immunodeficiencies). Interestingly, nu/nu mice are also resistant to neurological disease induced by ts1 Mo-MLV (Prasad et al. 1989) but not Cas-Br-E MLV (Oldstone et al. 1977), a difference that may reflect the part played by the thymus in replication of ts1 Mo-MLV but not Cas-Br-E MLV.

xid/xid mice carry a mutation in a nonreceptor PTK called BTK and display defects in antigen-mediated B-cell responses and antigen presentation (for review, see Tsukada et al. 1994; Khan et al. 1995). These animals are resistant to MAIDS (Hitoshi et al. 1993), but the ways in which defects in BTK signaling affect MAIDS development remain a matter of speculation. SCID mice lack mature B and T cells because of a defect in the catalytic subunit of the DNA protein kinase involved in antigen receptor gene rearrangement (Blunt et al. 1995; Kirchgessner et al. 1995). These animals fail to develop both MAIDS and MMTV-induced mammary tumors for reasons identical to those described for nu/nu mice. SCID mice do develop thymic lymphomas because the microenvironment present in the thymus is normal and the animals contain normal numbers of T-cell precursors (Custer et al. 1985). The possibility that endogenous viruses are involved in this process has not been investigated.

MHC Genes

Several genes that map to the MHC region affect susceptibility to a variety of retroviral infections. This region, called H-2 in the mouse and HLA in humans, encodes the class I proteins that present antigen to CTLs and the class II products that present antigen to T helper cells. The rFv1 locus (Cheseboro and Wehrly 1978), which influences the susceptibility of adult mice to FV-induced disease, corresponds to an allele of the H-2D structural gene which encodes a class I product. rFv1 influences the effectiveness of the CTL response mounted against FV and alters the outcome of disease in adult mice (Miyazawa et al. 1992a,c). An unidentified gene distinct from rFv1 affects development of MAIDS by a similar mechanism (Makino et al. 1990). The rFv2 locus, another gene modulating FV-induced disease, corresponds to the Q/TL region of the MHC locus, but the mechanism by which this region impacts FV-induced disease is not known (Miyazawa et al. 1992b). A third effect of the MHC locus has been mapped to the Aβ class II molecule. This product influences the T helper cell response to the env glycoprotein encoded by Fr-MLV (Miyazawa et al. 1992a). The H-2E locus also modulates development and recovery from FV in complex ways that involve effects on the T-cell repertoire and on helper T cells (Perry et al. 1994). Less is known about MHC-linked genes that affect other MLVs, but resistance to radiation leukemia virus (RadLV), Gross-MLV, and the neurotropic MLVs has been mapped to the MHC region (Lonai and Haran-Ghera 1977; Meruelo et al. 1977; Hoffman et al. 1984; Hoffman and Morse 1985).

MHC genes influence the development of retrovirus-induced diseases in chickens and probably have important roles in other hosts. Regression of RSV-induced sarcomas is influenced by R-Rs1, a dominant gene mapping to the chicken MHC region (Collins et al. 1977; Schierman et al. 1977; Gebriel et al. 1979). A number of other MHC haplotypes have been associated with tumor regression, induction of erythroblastosis, and ALV shedding (Bacon et al. 1981a,b, 1985; Collins et al. 1985, 1986; Bacon 1987; Schierman and Collins 1987; Malin et al. 1993). In humans, there is some suggestion that certain HLA haplotypes protect against TSP/HAM (Usuku et al. 1988), perhaps by facilitating a strong immune response against HTLV-1. However, the low incidence of this disease and the strong geographic bias complicate a thorough analysis of this issue.

Non-MHC-linked Genes

A number of partially characterized genes that influence the immune response also have important roles in determining the consequences of retroviral infections. The rFv3 gene locus affects susceptibility of adult mice to FV-induced disease. Animals carrying the rFv3^r allele make a particularly vigorous antivirus antibody response (Cheseboro and Wehrly 1979; Morrison et al. 1986) that seems to mediate this resistance. The Fv3 locus has been mapped to chromosome 15 (Hasenkrug et al. 1995), but the gene has not been identified and the mechanism by which the locus influences antibody production remains unknown. Another, as yet unmapped, gene appears to act in a similar fashion to modulate the latent period required for MLV-induced neurologic disease (Buller et al. 1990). However, the mechanism by which this gene influences antibody production is not yet known. The Av1 and Av2 loci control the ability of Ab-MLV to induce disease in certain strains of adult mice (Risser et al. 1978, 1985). These genes are distinct from those that control Mo-MLV-induced thymoma development and do not affect replication of the helper virus in Ab-MLV stocks, but little else is known about them. Another gene, Fhe, influences the types of leukemia that arise in particular strains of adult mice injected with FV (Silver and Fredrickson 1983). Studies addressing the mechanism by which this effect is mediated have not been reported.