The role of virus-receptor interaction in the development of FeLV-related erythroid hypoplasia (pure red cell aplasia)

Feline leukaemia virus is one of the most common causes of viral disease in the domestic cat. Virus infection causes a number of diseases, including lymphoma, immunodeficiency, reproductive problems and a form of anaemia known as pure red cell aplasia (PRCA). There are three known subgroups of the virus (FeLV-A, FeLV-B and FeLV-C), as determined previously by interference assays and each subgroup displays a distinct in vitro cell tropism. FeLV-A is present in all natural isolates, whereas FeLV-B and FeLV-C are more seldom isolated (FeLV-B is present in approximately 40% of all natural isolates, while FeLV-C is present in only 1% of all natural isolates). The subgroups are determined by the amino acid sequence of the envelope glycoprotein with each subgroup of virus utilising a distinct cellular receptor. It is the specificity of the FeLV-C receptor interaction that is thought to underlie the development of PRCA with the selective destruction of erythroid progenitors by this subgroup of virus. Subgroup C envs were cloned from a panel of primary isolates (the subgroup of these isolates had been established previously by superinfection interference). The envs were then characterised in order to determine the amino acids responsible for the characteristic in vitro tropism of subgroup C viruses. These analyses revealed that the subgroup C-component was more abundant in some primary isolates than in others. Furthermore, in some of the primary isolates more than one C-component was identified, suggesting that each isolate may consist of multiple viral variants. In agreement with previous studies, the env sequences were found to be relatively conserved with the majority of changes located within discrete variable regions, and in particular, the first variable region, the "VRA" region. However, a small number of characteristic mutations (unique to the isolate) were distributed throughout the entire envelope glycoprotein. Although these mutations may simple reflect drift in the Env sequence due to the low fidelity of the viral reverse transcriptase, an alternative explanation is that they compensate for mutations within the VRA and ensure that the protein adopts a viable conformation enabling receptor recognition. The env clones that displayed an in vitro tropism similar to the prototypic strain FeLV-C/Sarma shared two common motifs in the VRA region; the loss of a six amino acid stretch (61TNVKHG66) and a valine to tryptophan substitution on position 63 (V63W) (tryptophan residues have been implicated in the formation of a hydrophobic pocket that is thought to be critical for interaction between the virus and its receptor). In contrast with previous findings, the ability to infect porcine cells was found not to be a unique property of subgroup B viruses, both subgroup A and subgroup C Env clones were isolated that were capable of mediating infection of the porcine cell line ST Iowa (for example, an A-component of primary isolate FA27 and a C-component of primary isolate FY981). The presence of an aspartic acid to asparagine mutation on position 51 (D51N) in the A-component of the isolate appeared to correlate with the expanded cell tropism. However, the C- components of primary isolate FA27, which also displayed the D51N mutation, were incapable of infecting ST Iowa cells. This may be due to additional mutations in the env resulting in an altered conformational structure of the C-component Env compared with the A-component that was capable of infecting this cell line. These findings suggest that a single mutation may not be sufficient to alter the in vitro tropism of a clone; additional mutations may ensure the correct folding of the protein and thus the accessibility of the receptor binding site. Although previous studies suggested that the amino acid sequence of the envelope glycoprotein of FeLV-A is highly conserved between isolates, it was found that minor differences in sequence were distributed throughout the entire env. These variations in the envelope glycoprotein may account for the differences in in vitro tropism displayed by the novel FeLV-A isolates. To understand fully the extent of expression of the feline FeLV-C receptor (feFLVCR) and Its possible role In the onset of PRCA, a panel of both haematopoietic and non-haematopoietic tissues was screened for feFLVCR mRNA. In contrast with the human FeLV-C receptor (thought to be expressed preferentially in haematopoietic tissues), the feFLVCR is expressed in both haematopoietic and non-haematopoietic tissues and there was no single tissue where the receptor was consistently expressed to higher or lower levels. Next, the extent of receptor expression in cells that are involved in the erythroid maturation was examined. Bone marrow samples were collected and subsequently enriched for or depleted of cells that are involved In erythroid maturation using a monoclonal antibody directed against cells of this lineage. Although only a small number of bone marrow samples were tested, the feFLVCR appeared to be expressed at high levels in the bone marrow populations that were enriched for cells of the erythroid lineage. These findings support the hypothesis that the normal function of this gene is disrupted upon infection and subsequently leads to the onset of PRCA