Identification and initial characterization of the murine gammaherpesvirus 68 gene M3, encoding an abundantly secreted protein.

Several viruses, including members of the gammaherpesvirus family, encode proteins that are secreted into the extracellular environment. We have identified an abundant 44-kDa secreted protein that is present in the supernatant of fibroblasts infected with murine gammaherpesvirus 68 (gammaHV68; also referred to as MHV-68) but not in that of uninfected fibroblasts. Sequence analysis of the amino terminus and of internal peptides revealed that this protein is encoded by the gammaHV68 M3 open reading frame (ORF). The amino-terminal sequence of the secreted protein starts at residue 25 of the M3 ORF, consistent with the first 24 residues functioning as a signal peptide. Northern blot analysis revealed a single abundant approximately 1.4-kb early-late lytic transcript encoded by the M3 ORF. Analysis of a partial cDNA clone and subsequent analyses of products of rapid amplification of cDNA ends coupled with S1 nuclease protection assays demonstrate that the M3 protein is encoded by an unspliced, polyadenylated mRNA initiating at bp 7294 and terminating at bp 6007 of the gammaHV68 genome. The 3' end of the M3 transcript maps 9 bp downstream of a consensus polyadenylation signal. Thus, the predicted M3 ORF is a functional gene that encodes an abundant secreted protein which is a candidate for interacting with host cellular receptors or cytokines.

Size of IgG-opsonized particles determines macrophage response during internalization.

It is generally assumed that particles > 1 micron elicit a phagocytic response. To determine whether this is the case, we examined the uptake and transport of IgG-opsonized polystyrene beads of defined size, ranging from 0.2 to 3 microns, by mouse bone marrow-derived macrophages. The kinetics of opsonized bead internalization were comparable for each of the different beads examined. We used rhodamine phalloidin to examine particle-induced assembly of F-actin phagocytic cups by fluorescence microscopy. Phagocytic cup formation was size dependent in a nonlinear fashion. Less than 30% of 0.2- to 0.75-micron particles and greater than 80% of 2- and 3-micron particles were associated with F-actin. Cells treated with 0.25 micron cytochalasin D showed decreased phagocytic cup formation and a linear decrease in bead uptake as a function of particle surface area. In contrast, potassium depletion, which preferentially inhibits clathrin-mediated endocytosis, was more effective at inhibiting the uptake of smaller beads. Thus, with increasing particle size, IgG-opsonized particle uptake became less clathrin dependent and more actin dependent. The kinetics of ligand delivery to lysosomes was measured using an immunoprecipitation assay based on the intermixing of internalized anti-dinitrophenol (DNP) IgG with DNP-derivitized beta-glucuronidase (DNP-beta-glu) incorporated into lysosomes. Soluble mannosylated anti-DNP IgG was delivered to lysosomes after an 8-min lag period. The kinetics of anti-DNP IgG-opsonized beads showed a size-dependent response, where beads sized 0.2, 0.5, and 0.75 micron showed a lag period prior to delivery to lysosomes. In contrast, beads 1.0 micron or larger showed no lag in delivery to lysosomes. Since beads that had no lag in delivery to lysosomes also showed high levels of phagocytic cup formation, this suggests that phagocytic cups may be important in the rapid delivery of internalized particles to lysosomes.

Conversion of an extracellular cytolysin into a phagosome-specific lysin which supports the growth of an intracellular pathogen.

Listeria monocytogenes is a facultative intracellular pathogen which secretes a pore-forming cytolysin, listeriolysin O (LLO), necessary for intracellular growth. Clostridium perfringens is an extracellular pathogen which secretes a related cytolysin, perfringolysin O (PFO). When PFO is secreted by intracellular L. monocytogenes, it is toxic to the infected host cell. PFO-mediated toxicity renders the infected host cell permeable to gentamicin and leads to the death of the intracellular bacteria. In this study, we selected for L. monocytogenes mutants in which PFO supported the intracellular growth of L. monocytogenes. Six independent mutants were isolated, each containing a single amino acid change within the PFO protein. Three classes of PFO mutations were identified, all capable of mediating lysis of the vacuole but without a toxic effect upon the infected host cell. The first class had a severe defect in haemolytic activity. The second class had a change in the pH optimum of PFO. The third class had nearly wild-type levels of haemolytic activity, but had a decrease in protein half-life in the host-cell cytosol. Acquisition of single amino acid changes in PFO were sufficient to convert an extracellular cytolysin into a vacuole-specific lysin which mediated growth of L. monocytogenes in cultured cells.

Homology requirements for double-strand break-mediated recombination in a phage lambda-td intron model system.

Many group I introns encode endonucleases that promote intron homing by initiating a double-strand break-mediated homologous recombination event. A td intron-phage lambda model system was developed to analyze exon homology effects on intron homing and determine the role of the lambda 5'-3' exonuclease complex (Red alpha beta) in the repair event. Efficient intron homing depended on exon lengths in the 35- to 50-bp range, although homing levels remained significantly elevated above nonbreak-mediated recombination with as little as 10 bp of flanking homology. Although precise intron insertion was demonstrated with extremely limiting exon homology, the complete absence of one exon produced illegitimate events on the side of heterology. Interestingly, intron inheritance was unaffected by the presence of extensive heterology at the double-strand break in wild-type lambda, provided that sufficient homology between donor and recipient was present distal to the heterologous sequences. However, these events involving heterologous ends were absolutely dependent on an intact Red exonuclease system. Together these results indicate that heterologous sequences can participate in double-strand break-mediated repair and imply that intron transposition to heteroallelic sites might occur at break sites within regions of limited or no homology.