The extracellular domain of CR2, the Epstein-Barr virus (EBV)/C3d receptor of B lymphocytes, contains 15 or 16 tandemly arranged short consensus repeat elements (SCR). Recombinant CR2 proteins containing SCR 1 and 2 fused to Staphylococcus aureus protein A (PA-CR2) and to murine complement factor H SCR 20 (CR2FH) were expressed in Eschenichia coli and in insect cells, respectively. These recombinant CR2 molecules retained functional activity as indicated by their ability to bind to C3dg in an enzyme-linked immunosorbent assay and to inhibit EBV gp350/220 binding to B cells. PA-CR2 and CR2FH were as efficient in blocking EBV gp350/220 binding as the full-length CR2 extracellular domain, indicating that the first two SCR of CR2 contain the majority of the ligand binding activity of the receptor. PA-CR2 and CR2FH inhibited EBV-induced B-cell proliferation in vitro and blocked the development of EBV-induced lymphoproliferative disease in severe combined immunodeficient mice reconstituted with human lymphocytes. These studies indicate that soluble forms of truncated CR2 proteins may have potential therapeutic value in the treatment of EBV-induced lymphoproliferative disorders in humans that involve viral replication.

Early bone marrow infection of Moloney murine leukemia virus (M-MuLV)-infected mice was studied. Previous experiments indicated that early bone marrow infection is essential for the efficient development of T lymphoma. In order to identify the cellular pathway of infection in the bone marrow, infection of mice with a helper-free replication-defective M-MuLV-based retroviral vector was carried out. Such a vector will undergo only one round of infection, without spreading to other cells; thus, cells infected by the initially injected virus (directly infected cells) can be identified. For these experiments, the BAG vector that expresses bacterial β-galactosidase was employed. Neonatal NIH/Swiss mice were inoculated intraperitoneally with ca. 106 infectious units of a BAG vector pseudotyped with M-MuLV proteins, and bone marrow cells were recovered 2 to 12 days postinfection. Single-cell suspensions were tested for infection by staining with X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) or by immunofluorescence with an anti-β-galactosidase antibody. Two sizes of infected cells were evident: large multinucleated cells and small nondescript (presumptively hematopoietic) cells. Secondary stains for lineage-specific markers indicated that the large cells were osteoclasts. Some of the small cells expressed nonspecific esterase, which placed them in the myeloid lineage, but they lacked markers for hematopoietic progenitors (mac-1, gr-1, sca-1, and CD34). These results provide evidence for primary M-MuLV infection of osteoclasts or osteoclast progenitors in the bone marrow, and they suggest that known hematopoietic progenitors are not primary targets for infection. However, the subsequent spread of infection to hematopoietic progenitors was indicated, since bone marrow from mice infected in parallel with replication-competent wild-type M-MuLV showed detectable infection in small cells positive for mac-1 or CD34, as well as in osteoclasts.

The nature of Moloney murine leukemia virus (M-MuLV) infection after a subcutaneous (s.c.) inoculation was studied. We have previously shown that an enhancer variant of M-MuLV, Mo+PyF101 M-MuLV, is poorly leukemogenic when used to inoculate mice s.c., but not when inoculated intraperitoneally. This attenuation of leukemogenesis correlated with an inability of Mo+PyF101 M-MuLV to establish infection in the bone marrow of mice at early times postinfection. These results suggested that a cell type(s) is infected in the skin by wild-type but not Mo+PyF101 M-MuLV after s.c. inoculation and that this infection is important for the delivery of infection to the bone marrow, as well as for efficient leukemogenesis. To determine the nature of the cell types infected by M-MuLV and Mo+PyF101 M-MuLV in the skin after a s.c. inoculation, immunohistochemistry with an anti-M-MuLV CA antibody was performed. Cells of developing hair follicles, specifically cells of the outer root sheath (ORS), were extensively infected by M-MuLV after s.c. inoculation. The Mo+PyF101 M-MuLV variant also infected cells of the ORS but the level of infection was lower. By Western blot analysis, the level of infection in skin by Mo+PyF101 M-MuLV was approximately 4- to 10-fold less than that of wild-type M-MuLV. Similar results were seen when a mouse keratinocyte line was infected in vitro with both viruses. Cells of the ORS are a primary target of infection in vivo, since a replication defective M-MuLV-based vector expressing β-galactosidase also infected these cells after a s.c. inoculation.