A strategy for cloning and mutagenesis of an infectious herpesvirus genome is described. The mouse cytomegalovirus genome was cloned and maintained as a 230 kb bacterial artificial chromosome (BAC) in E. coli. Transfection of the BAC plasmid into eukaryotic cells led to a productive virus infection. The feasibility to introduce targeted mutations into the BAC cloned virus genome was shown by mutation of the immediate-early 1 gene and generation of a mutant virus. Thus, the complete construction of a mutant herpesvirus genome can now be carried out in a controlled manner prior to the reconstitution of infectious progeny. The described approach should be generally applicable to the mutagenesis of genomes of other large DNA viruses.Cytomegalovirus (CMV) is an important human pathogen with a high prevalence in the human population that causes severe and even fatal disease in immunologically immature or immunocompromised patients (1). Because human and mouse CMV (MCMV) show a series of similarities in biology and pathogenesis (2) infection of the mouse with MCMV has become an extensively used in vivo model to study the pathogenesis of CMV infection. The 235-kb genomes of both human and mouse CMV are the largest genomes of mammalian DNA viruses. Sequence analysis of the human and mouse CMV genomes revealed a similar genetic organization and a coding capacity for presumably more than 220 polypeptides (3-5). However, information on the function of the majority of CMV gene products is still rather limited. This is in sharp contrast to the alphaherpesviruses, where the study of a wealth of viral mutants contributed significantly to the understanding of viral gene functions (reviewed in ref. 6). There is a lack of CMV mutants because due to the large genome size and slow replication kinetics construction of CMV recombinants turned out to be difficult.The technique of insertional mutagenesis has been developed for disruption and deletion of CMV genes (7,8). Because the frequency of homologous recombination in eukaryotic cells is low the technique is quite ineffective. In addition adventitious deletions and the formation of illegitimate recombinant viruses have frequently been observed (refs. 7 and 9; I.C., unpublished data). Although selection procedures have improved the original technique (9-11) generation of CMV mutants remains a laborious, time-consuming, and often unsuccessful task. Recently, the technique for construction of recombinant herpesviruses from cloned overlapping fragments (12) has been extended to CMV (13). This is a major improvement in that the technique generates only recombinant virus and obviates selection against nonrecombinant wild type (wt) virus. Still, the resultant mutant is the product of several recombination events in eukaryotic cells that are difficult to control. Correct reconstitution of the viral genome can only be verified after growth and isolation of the mutant virus.Here we describe an approach for production of CMV mutants. Construction of the mutant genome is completely inde...

Conserved molecular patterns derived from pathogenic microorganisms prime antigen-presenting dendritic cells (DC) to induce adaptive T cell responses. In contrast, virus-infected or tumor cells that express low levels of major histocompatibility complex (MHC) class I activate natural killer (NK) cells for direct killing. It is unknown whether NK cell recognition of MHC class I(low) targets can also induce adaptive T cell responses. Here, we show that MHC class I(low) targets initiate a cascade of immune responses, starting with the immediate activation of NK cells. The activated NK cells then prime DC to produce IL-12 and to induce highly protective CD8 T cell memory responses. Therefore, sensing of MHC class I(low) targets by NK cells can link innate and adaptive immunity to induce protective T cell responses and may alarm the immune system during early infection with noncytopathic viruses.

The principle by which mouse cytomegalovirus blocks antigen presentation in the MHC class I pathway was investigated. The responsible gene m152, encoding a type I transmembrane glycoprotein of 40 kDa, is a member of a gene family located in the right-hand terminal region of the 230 kb virus genome. Expression of m152 in murine and human cells arrested the export of mouse class I complexes from the ER-Golgi intermediate compartment/cis-Golgi compartment and inhibited lysis by cytotoxic T cells. The plasma membrane transport of human MHC class I molecules was not affected. The deletion of the cytoplasmic tail of gp40 did not lift its effect on class I molecule export, indicating that this protein differs in its functions from known immunosubversive viral gene products and represents a novel principle by which a herpesvirus shuts off MHC class I function.

No abstract