Kinjiro Morimoto

Reinvestigation of the role of the rabies virus glycoprotein in viral pathogenesis using a reverse genetics approach

The rabies virus glycoprotein (G) gene of the highly neuroinvasive and neurotropic strains SHBRV-18, CVS-N2c, and CVS-B2c was introduced into the non-neuroinvasive and less neurotropic SN-10 strain to provide further insight into the role of G in the pathogenesis of rabies. Phenotypic analyses of the recombinant viruses revealed, as expected, that the neurotropism of a particular rabies virus strain was a function of its G. Nevertheless, the pathogenicity of the recombinant viruses was, in every case, markedly lower than that of the wild-type viruses suggesting that while the G dictates neurotropism, other viral attributes are also important in pathogenesis. The low pathogenicity of the recombinant viruses is at least in part due to a strong increase in transcription activity. On the other hand, the production of infectious virus by the R-SHB18 recombinant virus-infected cells was significantly delayed by comparison with SHBRV-18 wild-type virus infected-cells. Replacement of the R-SHB18 G cytoplasmic domain, transmembrane domain, and stem region with its SN-10 G counterparts neither results in a significant increase in budding efficiency nor an increase in pathogenicity. These results suggest that an optimal match of the cytoplasmic domain of G with the matrix protein may not be sufficient for maximal virus budding efficiency, which is evidently a major factor of virus pathogenicity. Our studies indicate that to maintain pathogenicity, the interactions between various structural elements of rabies virus must be highly conserved and the expression of viral proteins, in particular the G protein, must be strictly controlled.

Genetic engineering of live rabies vaccines.

Rabies virus is not a single entity but consists of a wide array of variants that are each associated with different host species. These viruses differ greatly in the antigenic makeup of their G proteins, the primary determinant of pathogenicity and major inducer of protective immunity. Due to this diversity, existing rabies vaccines have largely been targeted to individual animal species. In this report, a novel approach to the development of rabies vaccines using genetically modified, reverse-engineered live attenuated rabies viruses is described. This approach entails the engineering of vaccine rabies virus containing G proteins from virulent strains and modification of the G protein to further reduce pathogenicity. Strategies employed included exchange of the arginine at position 333 for glutamine and modification of the cytoplasmic domain. The recombinant viruses obtained were non-neuroinvasive when administered via a peripheral route. The ability to confer protective immunity depended largely upon conservation of the G protein antigenic structure between the vaccine and challenge virus, as well as on the route of immunization.

The Central Nervous System Inflammatory Response to Neurotropic Virus Infection Is Peroxynitrite Dependent

We have recently demonstrated that increased blood-CNS barrier permeability and CNS inflammation in a conventional mouse model of experimental allergic encephalomyelitis are dependent upon the production of peroxynitrite (ONOO−), a product of the free radicals NO· and superoxide (O2·−). To determine whether this is a reflection of the physiological contribution of ONOO− to an immune response against a neurotropic pathogen, we have assessed the effects on adult rats acutely infected with Borna disease virus (BDV) of administration of uric acid (UA), an inhibitor of select chemical reactions associated with ONOO−. The pathogenesis of acute Borna disease in immunocompetent adult rats results from the immune response to the neurotropic BDV, rather than the direct effects of BDV infection of neurons. An important stage in the BDV-specific neuroimmune response is the invasion of inflammatory cells into the CNS. UA treatment inhibited the onset of clinical disease, and prevented the elevated blood-brain barrier permeability as well as CNS inflammation seen in control-treated BDV-infected rats. The replication and spread of BDV in the CNS were unchanged by the administration of UA, and only minimal effects on the immune response to BDV Ags were observed. These results indicate that the CNS inflammatory response to neurotropic virus infection is likely to be dependent upon the activity of ONOO− or its products on the blood-brain barrier.

Rabies Virus-Induced Activation of Mitogen-Activated Protein Kinase and NF-κB Signaling Pathways Regulates Expression of CXC and CC Chemokine Ligands in Microglia

ABSTRACT Following virus infection of the central nervous system, microglia, the ontogenetic and functional equivalents of macrophages in somatic tissues, act as sources of chemokines, thereby recruiting peripheral leukocytes into the brain parenchyma. In the present study, we have systemically examined the growth characteristics of rabies virus (RV) in microglia and the activation of cellular signaling pathways leading to chemokine expression upon RV infection. In RV-inoculated microglia, the synthesis of the viral genome and the production of virus progenies were significantly impaired, while the expression of viral proteins was observed. Transcriptional analyses of the expression profiles of chemokine genes revealed that RV infection, but not exposure to inactivated virions, strongly induces the expression of CXC chemokine ligand 10 (CXCL10) and CC chemokine ligand 5 (CCL5) in microglia. RV infection triggered the activation of signaling pathways mediated by mitogen-activated protein kinases, including p38, extracellular signal-regulated kinases 1 and 2 (ERK1/2), and c-Jun N-terminal kinase, and nuclear factor κB (NF-κB). RV-induced expression of CXCL10 and CCL5 was achieved by the activation of p38 and NF-κB pathways. In contrast, the activation of ERK1/2 was found to down-regulate CCL5 expression in RV-infected microglia, despite the fact that it was involved in partial induction of CXCL10 expression. Furthermore, NF-κB signaling upon RV infection was augmented via a p38-mediated mechanism. Taken together, these results indicate that the strong induction of CXCL10 and CCL5 expression in microglia is precisely regulated by the activation of multiple signaling pathways through the recognition of RV infection.

Overexpression of Cytochrome c by a Recombinant Rabies Virus Attenuates Pathogenicity and Enhances Antiviral Immunity

ABSTRACT The pathogenicity of individual rabies virus strains appears to correlate inversely with the extent of apoptotic cell death they induce and with the expression of rabies virus glycoprotein, a major inducer of an antiviral immune response. To determine whether the induction of apoptosis by rabies virus contributes to a decreased pathogenicity by stimulating antiviral immunity, we have analyzed these parameters in tissue cultures and in mice infected with a recombinant rabies virus construct that expresses the proapoptotic protein cytochromec. The extent of apoptosis was strongly increased in primary neuron cultures infected with the recombinant virus carrying the active cytochrome c gene [SPBN-Cytoc(+)], compared with cells infected with the recombinant virus containing the inactive cytochrome c gene [SPBN-Cytoc(−)]. Mortality in mice infected intranasally with SPBN-Cyto c(+) was substantially lower than in SPBN-Cytoc(−)-infected mice. Furthermore, virus-neutralizing antibody (VNA) titers were significantly higher in mice immunized with SPBN-Cyto c(+) at the same dose. The VNA titers induced by these recombinant viruses paralleled their protective activities against a lethal rabies virus challenge infection, with SPBN-Cytoc(+) revealing an effective dose 20 times lower than that of SPBN-Cyto c(−). The strong increase in immunogenicity, coupled with the marked reduction in pathogenicity, identifies the SPBN-Cyto c(+) construct as a candidate for a live rabies virus vaccine.

High level expression of a human rabies virus-neutralizing monoclonal antibody by a rhabdovirus-based vector

Humans exposed to rabies virus must be promptly treated by passive immunization with anti-rabies antibody and active immunization with rabies vaccine. Currently, antibody prepared from pooled human serum or from immunized horses is utilized. However, neither of these reagents are readily available, entirely safe, or consistent in their biological activity. An ideal reagent would consist of a panel of human monoclonal antibodies. Such antibodies are now available, their only drawback being the cost of production. Using recombinant technology, we constructed a rabies virus-based vector which expresses high levels (approximately 60 pg/cell) of rabies virus-neutralizing human monoclonal antibody. The vector is a modified vaccine strain of rabies virus in which the rabies virus glycoprotein has been replaced with a chimeric vesicular stomatitis virus glycoprotein, and both heavy and light chain genes encoding a human monoclonal antibody have been inserted. This recombinant virus can infect a variety of mammalian cell lines and is non-cytolytic, allowing the use of cell culture technology routinely employed to produce rabies vaccines.

Mechanisms of virus-induced neuronal damage and the clearance of viruses from the CNS.

The pathogenic mechanisms underlying virus-induced neurological disease are complex but fall into two general categories, neuronal damage or dysfunction, resulting: (1) from within, as a direct consequence of the virus infection, and (2) from without, due to the indirect action of resident and invading immune/inflammatory cells responding to viral antigens. Our studies have focused on two viruses that can cause acute, lethal neurological diseases representative of these classes: rabies virus (RV) and Borna disease virus (BDV). RV infection induces significant electrophysiological changes in the CNS and sleep alterations but is accompanied by only minor histopathological changes in the CNS (Gourmelon et al. 1986, 1991). In contrast, acute Borna disease is associated with extensive neuropathology including astrogliosis, perivascular cuffing, monocytic infiltration of the brain parenchyma, and massive neuronal loss (Right et al. 1990; Gosztonyi and Ludwig 1995; Stitz et al. 1995; Morimoto et al. 1996), which are dependent on a BDV-specific immune response. The profound difference in the pathogenesis of rabies and Borna disease is evidenced by the fact that immunosuppression either has no effect or is detrimental to the outcome of rabies infection but is therapeutic in Borna disease (Stitz et al. 1995; Morimoto et al. 1996). Despite the clear differences in neuropathology between rabies and Borna disease, there are immune strategies for each that can evidently clear these viruses from the CNS thereby preventing a lethal outcome to the infections (Dietzschold et al. 1992; Dietzschold 1993; Richt 1994). Further knowledge of the different pathogenic processes underlying rabies and Borna disease is key to the development of therapeutic strategies for diverse CNS viral diseases. In this chapter we discuss the response of CNS resident cells to infection, the link between these responses and the induction of virus-specific immunity, and the mechanisms through which virus can be cleared from the CNS.

Studies on the rabies virus RNA polymerase : 2. Possible relationships between the two forms of the non-catalytic subunit (P protein)

We investigated the relationship between the two forms of rabies virus P protein, a non‐catalytic subunit of rabies virus RNA polymerase. The two displayed different electrophoretic mobilities as 37‐ and 40‐kDa polypeptides, hence termed as p37 and p40, respectively. Double labeling experiments with [3H]leucine and [32P]orthophosphate demonstrated that p40 was much more phosphorylated than p37. Treatment of the virion proteins with alkaline phosphatase eliminated only p40, and not 37‐kDa polypeptide. The p37 was a major product of the P gene, and was accumulated in the infected cell and incorporated into the virion. On the other hand, p40 was apparently detected only in the virion, and little detected in the cells. Treatment of infected cells with okadaic acid, however, resulted in significant accumulation of p40 in the cell, suggesting that p40 was continuously produced in the cell but dephosphorylated quickly. We detected both 37‐ and 40‐kDa products in P cDNA‐transfected animal cells, while only a 37‐kDa product was produced in Escherichia coli. Incubation of 37‐kDa products from E. coli with the lysates of animal cells in vitro resulted in the production of a 40‐kDa product, which was also shown to be suppressed by the heparin. From these results, it is suggested that p40 is produced by the hyperphosphorylation of a 37‐kDa polypeptide, which depends on certain heparin‐sensitive cellular enzyme(s) and occurs even in the absence of the other viral gene products, and that p40 is reverted quickly to p37 in the infected cells, probably being dependent on some virus‐induced factor(s).