SumnlaryThe CD45 transmembrane glycoprotein has been shown to be a protein phosphotyrosine phosphatase and to be important in signal transduction in T and B lymphocytes. We have employed gene targeting to create a strain of transgenic mice that completely lacks expression of all isoforms of CD45. The spleens from CD45-null mice contain approximately twice the number of B cells and one fifth the number ofT cells found in normal controls. The increase in B cell numbers is due to the specific expansion of two B cell subpopulations that express high levels ofimmunoglobulin (IgM) staining. T cell development is significantly inhibited in CD45-null animals at two distinct stages. The efficiency of the development of CD4-CD8-thymocytes into CD4+CD8 § thymocytes is reduced by about twofold, subsequently the frequency of successful maturation of the double positive population into mature, single positive thymocytes is reduced by a further four-to fivefold. In addition, we demonstrate that CD45-null thymocytes are severely impaired in their apoptotic response to cross-linking signals via T cell receptor (TCtk) in fetal thymic organ culture. In contrast, apoptosis can be induced normally in CD45-null thymocytes by non-TC1L-mediated signals. Since both positive and negative selection require signals through the TCR complex, these findings suggest that CD45 is an important regulator of signal transduction via the TCIk complex at multiple stages of T cell development. CD45 is absolutely required for the transmission ofmitogenic signals via IgM and IgD. By contrast, CD45-null B cells proliferate as well as wild-type cells to CD40-mediated signals. The proliferation of B cells in response to CD38 cross-linking is significantly reduced but not abolished by the CD45-null mutation. We conclude that CD45 is not required at any stage during the generation of mature peripheral B cells, however its loss reveals a previously unrecognized role for CD45 in the regulation of certain subpopulations of B cells.

All gammaherpesviruses encode a virion glycoprotein positionally homologous to Epstein-Barr virus gp350. These glycoproteins are thought to be involved in cell binding, but little is known of the roles they might play in the whole viral replication cycle. We have analyzed the contribution of murine gammaherpesvirus 68 (MHV-68) gp150 to viral propagation in vitro and host colonization in vivo. MHV-68 lacking gp150 was viable and showed normal binding to fibroblasts and normal single-cycle lytic replication. Its capacity to infect glycosaminoglycan (GAG)-deficient CHO-K1 cells and NS0 and RAW264.7 cells, which express only low levels of GAGs, was paradoxically increased. However, gp150-deficient MHV-68 spread poorly through fibroblast monolayers, with reduced cell-free infectivity, consistent with a deficit in virus release. Electron microscopy showed gp150-deficient virions clustered on infected-cell plasma membranes. MHV-68-infected cells showed reduced surface GAG expression, suggesting that gp150 prevented virions from rebinding to infected cells after release by making MHV-68 infection GAG dependent. Surprisingly, gp150-deficient viruses showed only a transient lag in lytic replication in vivo and established normal levels of latency. Cell-to-cell virus spread and the proliferation of latently infected cells, for which gp150 was dispensable, therefore appeared to be the major route of virus propagation in an infected host.Herpesviruses are large, complex pathogens that use a range of different glycoproteins to spread between cells and between hosts (34). A common first step in infection is virion adsorption to cell surface glycosaminoglycans (GAGs), such as heparan sulfate, which are widespread on epithelial surfaces (5). Herpesvirus-GAG interactions are thought to concentrate virions on cell surfaces and so promote specific protein receptor binding, which is then followed by membrane fusion (32). Cell binding by purified virions and recombinant viral glycoproteins has been studied extensively in vitro, but relatively little is known about how individual herpesvirus glycoproteins contribute to infection of the natural host. Here simple fluid-filled spaces are rare, extracellular matrix is abundant, and viral spread is probably limited to defined anatomical pathways. A significant role for direct cell-to-cell spread in herpesvirus infections is implied by pseudorabies virus that lacks glycoprotein D and infects cells poorly as free virus propagating efficiently in mice (28) and by herpes simplex virus that lacks glycoprotein E (required only for cell-to-cell spread of virus) disseminates inefficiently from the rat cornea (14).Our understanding of in vivo glycoprotein function is especially limited with gammaherpesviruses such as Epstein-Barr virus (EBV) and the Kaposi's sarcoma-associated herpesvirus (KSHV), whose narrow host ranges have largely precluded studies of pathogenesis. In vitro lytic replication systems have also proved difficult to establish. Nonetheless, it is important for practical purposes, suc...

The murine gamma-herpesvirus-68 (MHV-68) K3 protein, like that of the Kaposi's sarcoma associated herpesvirus, down-regulates major histocompatibility complex (MHC) class I expression. However, how this contributes to viral replication in vivo is unclear. After intranasal MHV-68 infection, K3 was transcribed both during acute lytic infection in the lung and during latency establishment in lymphoid tissue. K3-deficient viruses were not cleared more rapidly from the lung, but the number of latently infected spleen cells was reduced and the frequency of virus-specific CD8(+) cytotoxic T lymphocytes (CTLs) was increased. CTL depletion reversed the viral latency deficit. Thus, a major function of K3 appears to be CTL evasion during viral latency expansion.

Luciferase-based imaging allows a global view of microbial pathogenesis. We applied this technique to gammaherpesvirus infection by inserting a luciferase expression cassette into the genome of murine herpesvirus-4 (MuHV-4). The recombinant virus strongly expressed luciferase in lytically infected cells without significant attenuation. We used it to compare different routes of virus inoculation. After intranasal infection of anaesthetized mice, luciferase was expressed in the nose and lungs for 7–10 days and in lymphoid tissue, most consistently the superficial cervical lymph nodes, for up to 30 days. Gastrointestinal infection was not observed. Intraperitoneal infection was very different to intranasal, with strong luciferase expression in the liver, kidneys, intestines, reproductive tract and spleen, but none in the nose or lungs. The nose has not previously been identified as a site of MuHV-4 infection. After intranasal infection of non-anaesthetized mice, it was the only site of non-lymphoid luciferase expression. Nevertheless, lymphoid colonization and persistence were still established, even at low inoculation doses. In contrast, virus delivered orally was very poorly infectious. Inoculation route therefore had a major impact on pathogenesis. Low dose intranasal infection without anaesthesia seems most likely to mimic natural transmission, and may therefore be particularly informative about normal viral gene functions.

Herpesviruses characteristically persist in immune hosts as latent genomes, but to transmit infection they must reactivate and replicate lytically. The interaction between newly formed virions and pre-existing antibody is therefore likely to be a crucial determinant of viral fitness. Murine gammaherpesvirus-68 (MHV-68) behaves as a natural pathogen of conventional, inbred mice and consequently allows such interactions to be analysed experimentally in a relatively realistic setting. Here, monoclonal antibodies (mAbs) were derived from MHV-68-infected mice and all those recognizing infected-cell surfaces were tested for their capacity to neutralize MHV-68 virions. All of the neutralizing mAbs identified were specific for the viral glycoprotein H (gH)-gL heterodimer and required both gH and gL to reproduce their cognate epitopes. Based on antibody interference, there appeared to be two major neutralization epitopes on gH-gL. Analysis of a representative mAb indicated that it blocked infection at a post-binding step -either virion endocytosis or membrane fusion.

Gamma-herpesviruses persist in lymphocytes and cause disease by driving their proliferation. Lymphocyte infection is therefore a key pathogenetic event. Murid Herpesvirus-4 (MuHV-4) is a rhadinovirus that like the related Kaposi's Sarcoma-associated Herpesvirus persists in B cells in vivo yet infects them poorly in vitro. Here we used MuHV-4 to understand how virion tropism sets the path to lymphocyte colonization. Virions that were highly infectious in vivo showed a severe post-binding block to B cell infection. Host entry was accordingly an epithelial infection and B cell infection a secondary event. Macrophage infection by cell-free virions was also poor, but improved markedly when virion binding improved or when macrophages were co-cultured with infected fibroblasts. Under the same conditions B cell infection remained poor; it improved only when virions came from macrophages. This reflected better cell penetration and correlated with antigenic changes in the virion fusion complex. Macrophages were seen to contact acutely infected epithelial cells, and cre/lox-based virus tagging showed that almost all the virus recovered from lymphoid tissue had passed through lysM+ and CD11c+ myeloid cells. Thus MuHV-4 reached B cells in 3 distinct stages: incoming virions infected epithelial cells; infection then passed to myeloid cells; glycoprotein changes then allowed B cell infection. These data identify new complexity in rhadinovirus infection and potentially also new vulnerability to intervention.

The herpesvirus glycoprotein H (gH) and gL associate to form a heterodimer that plays a central role in virus-driven membrane fusion. When archetypal alpha-or betaherpesviruses lack gL, gH misfolds and progeny virions are noninfectious. In order to define the role that gL plays in gamma-2 herpesvirus infections, we disrupted its coding sequence in murine gammaherpesvirus-68 (MHV-68). MHV-68 lacking gL folded gH into a conformation antigenically distinct from the form that normally predominates on infected cells. gL-deficient virions bound less well than the wild type to epithelial cells and fibroblasts. However, they still incorporated gH and remained infectious. The cell-to-cell spread of gL-deficient viruses was remarkably normal, as was infection, dissemination, and latency establishment in vivo. Viral membrane fusion was therefore gL independent. The major function of gL appeared to be allowing gH to participate in cell binding prior to membrane fusion. This function was most important for the entry of MHV-68 virions into fibroblasts and epithelial cells.

Dendritic cells (DCs) play a central role in initiating immune responses. Some persistent viruses infect DCs and can disrupt their functions in vitro. However, these viruses remain strongly immunogenic in vivo. Thus what role DC infection plays in the pathogenesis of persistent infections is unclear. Here we show that a persistent, B cell-tropic gamma-herpesvirus, Murid Herpesvirus-4 (MuHV-4), infects DCs early after host entry, before it establishes a substantial infection of B cells. DC-specific virus marking by cre-lox recombination revealed that a significant fraction of the virus latent in B cells had passed through a DC, and a virus attenuated for replication in DCs was impaired in B cell colonization. In vitro MuHV-4 dramatically altered the DC cytoskeleton, suggesting that it manipulates DC migration and shape in order to spread. MuHV-4 therefore uses DCs to colonize B cells.