A. M. Field

B19 parvovirus has been shown to persist in some immunocompromised patients, and treatment with specific antibodies can lead to decreased quantities of circulating virus and hematologic improvement. A defective immune response to B19 parvovirus in these patients was shown by comparison of results using a capture RIA and immunoblotting. In normal individuals, examination of paired sera showed that the dominant humoral immune response during early convalescence was to the virus major capsid protein (58 kD) and during late convalescence to the minor capsid species (83 kD). In patients with persistent parvovirus infection, variable titers against intact particles were detected by RIA, but the sera from these patients had minimal or no IgG to capsid proteins determined by Western analysis. Competition experiments suggested that this discrepancy was not explicable on the basis of immune complex formation alone and that these patients may have a qualitative abnormality in antibody binding to virus. In neutralization experiments, in which erythroid colony formation in vitro was used as an assay of parvovirus activity, sera from patients with poor reactivity on immunoblotting were also inadequate in inhibiting viral infectivity. A cellular response to purified B19 parvovirus could not be demonstrated using proliferation assays and PBMC from individuals with serologic evidence of exposure to virus. These results suggest that production of neutralizing antibody to capsid protein plays a major role in limiting parvovirus infection in man.

B19 parvovirus is pathogenic in humans, causing fifth disease, transient aplastic crisis, some cases of hydrops fetalis, and acquired pure red cell aplasia. Efforts to develop serologic assays and vaccine development have been hampered by the virus's extreme tropism for human bone marrow and the absence of a convenient culture system. We constructed recombinants containing either the major (VP2) or minor (VP1) structural proteins of B19 in a baculovirus-based plasmid, from which the polyhedrin gene had been deleted; these recombinant plasmids were used to generate recombinant infectious baculovirus. Subsequent infection of insect cells in vitro resulted in high-level expression of either B19VP1 or VP2. Parvovirus capsids were obtained by self-assembly in cell cultures coinfected with either VP1- and VP2-containing baculoviruses or, surprisingly, VP2-containing baculoviruses alone. Empty B19 capsids composed of VP1 and VP2 could replace serum virus as a source of antigen in a conventional immunoassay for detection of either IgG or IgM antiparvovirus antibodies in human serum. Immunization of rabbits with capsids composed of VP1 and VP2 resulted in production of antisera that recognized serum parvovirus on immunoblot and neutralized parvovirus infectivity for human erythroid progenitor cells. Baculovirus-derived parvovirus antigen can substitute for scarce viral antigen in immunoassays and should be suitable as a human vaccine.

Urine samples from 1,235 pregnant women were examined by light microscopy for cytologic evidence of virus infection. Smears of urine sediment from 40 women (3.2%) were observed to contain inclusion-bearing cells; polyomavirus infection was confirmed by virologic methods in 24 (60%). A polyomavirus was isolated from 12 women. Five isolates were identified as JC virus and one as BK virus. Another isolate designated AS virus appeared to be unique. Serologic studies on the 40 women were consistent with a high frequency of reactivation of JC virus, and virus excretion was related to gestation. The evidence suggests that selective excretion of JC virus may occur in pregnancy. Among 390 pregnant women without inclusion-bearing cells in their urine, 78 (20%) had a high or rising titer of serum antibody to JC or BK virus or both, a result suggesting virus reactivation, but virus excretion was not detected. In contrast to other reports, no evidence was found for transmission of BK virus to the fetus.

Phages coding for production of Vero cytotoxins VT1 or VT2 in strains of Escherichia coli serotype O157.H7 or O157.H- were morphologically indistinguishable. Their genome size and restriction enzyme digests of the phage DNA were similar. These phages were clearly different in these respects from a VT1-encoding phage isolated from a strain of E. coli O26.H11 (H19). However the VT1 region cloned from the phage originating in the E. coli O157.H7 strain was identical to the VT1 region previously cloned from the phage carried by H19. Sequences encoding VT2 that were cloned from the phage in E. coli O157.H- have been mapped and the VT2 region identified by transposon insertion. The cloned regions coding for VT1 or VT2 production had no similarities in the presence of restriction enzyme sites over a distance of about 2 kb, and two VT1-specific probes spanning a region of about 1.4 kb did not hybridize under stringent conditions with cloned VT2 DNA. A 2 kb HincII fragment contained the VT2 genes but hybridized to VT1-encoding phages and recombinant plasmids via flanking phage DNA. A 0.85 kb AvaI-PstI fragment was a specific probe for VT2 sequences and did not hybridize under stringent conditions to phages or plasmid recombinants encoding VT1.