Results and discussion

Rotavirus pathogenesis is complex and involves several mechanisms

The outcome of an infection with rotavirus is clearly dependent on both host and viral factors, and these factors can affect one or more of the several stages of pathogenesis (Table 1). Host factors have been dissected by analysis of the outcome of infection in animals inoculated with well-defined viral strains. Both natural and experimental rotavirus infections are characterized by viral replication in enterocytes in the small intestine, with subsequent cell lysis and attendant villus blunting, depressed levels of mucosal disaccharidases, watery diarrhoea and dehydration. Many studies have demonstrated malabsorption in

TABLE 1 Stages in enteric virus pathogenesis

Entry into the host Primary replication

Local or disseminated host cellular responses (signalling) Spread through host Host immune response Cell injury

Stability and survival of virus in the gastrointestinal tract Adsorption and penetration into enterocytes (receptors) Uncoating, transcription, translation replication, assembly, release Effect of viral proteins on cell function (NSP4)

enterocytes and correlated this dysfunction with destruction of enterocytes and histopathological changes in the intestine, which are generally seen 24-36 h after infection (Graham et al 1984, Davidson et al 1977). However, malabsorption cannot be the entire basis of rotavirus pathogenesis because it fails to explain the early watery diarrhoea that occurs prior to the detection of villus blunting and other histological changes in the intestine (Collins et al 1989, Theil et al 1978, McAdaragh et al 1980, Mebus 1976, Ward et al 1996). In addition, some animals exhibit diarrhoea in the absence of clear histopathological changes (neonatal mice infected with homologous rotavirus strains; Burns et al 1995), and other adult rotavirus-infected animals (rabbits) show typical histological changes in the intestine but do not get diarrhoea (Ciarlet et al 1998a). Finally, oral administration of epidermal growth factor to rotavirus-infected piglets can restore the intestinal mucosa and enzyme activities, but such treatments do not hasten the resolution of diarrhoea (Zijlstra et al 1994). Thus, a clear association of mucosal damage and diarrhoea is lacking. One explanation for this is the fact that the infection is patchy, so one might observe mucosal changes in one part of the intestine but these would be insufficient to cause diarrhoeal disease. In other cases, animals with mucosal damage may release fluid into the intestinal lumen but compensatory physiological mechanisms (colonic reabsorption of fluid) may decrease fluid loss so diarrhoea is not observed. In 1988, Osborne proposed that enteroctye interactions induced a localized response that triggered the production of endogenous, neuroactive, hormonal substances of pathophysiological importance (Osborne et al 1988). Vascular damage due to villus ischaemia was suggested to be involved, possibly being mediated by release of a vasoactive agent from infected epithelial cells. Recent studies are beginning to provide a molecular basis for such new mechanisms of pathogenesis.

The rotavirus particle is composed of three concentric protein layers surrounding the eleven segments of double-stranded RNA which encode the six structural viral proteins VP1—VP4, VP6, and VP7, and six non-structural proteins, NSP1—NSP6. Each genome segment, with the exception of gene 11, which encodes two viral proteins (NSP5 and NSP6), codes for a single viral protein. The innermost core layer is formed by VP2 and encloses the genomic RNA and enzymatic complexes found at the fivefold vertices, which contain VP1 (the RNA-dependent RNA polymerase) and VP3 (a guanylyltransferase and methylase). The intermediate layer is made up of the most abundant rotavirus protein, VP6. The outer layer consists of the glycoprotein VP7 and the haemagglutinin and cell attachment protein VP4.

Viral factors involved in virulence have been dissected by several approaches. Analyses of reassortants that contain a single gene from one 'virulent' parental virus and other genes from another 'avirulent' parental virus have implicated specific viral genes in virulence. These putative virulence genes code for both structural (VP3, VP4, VP7) and non-structural (NSP1, NSP2, NSP4) proteins (reviewed in Burke & Desselberger 1996). The role of some of these proteins (e.g. NSP1) in virulence may vary depending on the host species; for example, NSP1 appears to be a virulence factor for mice but not for rabbits or piglets (Broome et al 1993, Ciarlet et al 1998b, Bridger et al 1998). The genes implicated in virulence that code for the surface proteins of the virus are likely to be involved in virus stability, virus attachment and penetration into cells. The precise roles for the inner capsid protein VP3, which functions as the capping enzyme for viral RNA, and the two non-structural proteins NSP1 and NSP2 in virulence remain unclear. However, these proteins may affect replication efficiency and two of them (VP3 and NSP1) are recognized targets for cytotoxic T lymphocytes. NSP4 is a virulence gene because it functions as an enterotoxin (see below).

Rotaviruses enter polarised intestinal epithelial cells by distinct mechanisms depending on whether or not virus binding to cells is sensitive or resistant to treatment with neuraminidase

A critical initiating event in pathogenesis involves the entry of virus into cells. This process is complex for rotavirus and likely involves two viral proteins and multiple proteins on cells. Recent analyses of the interactions of viruses with polarized intestinal cells, grown on permeable filters, have detected differences in the early steps of viral infection based on whether or not the virus requires interaction with sialic acid or other surface moieties that are sensitive to treatment with neuraminidase (Fig. 1; Ciarlet et al 2000a). Most rotaviruses bind cells in a neuraminidase-independent manner (Ciarlet & Estes 1999). Rotavirus strains (all human strains tested to date, and bovine WC3, porcine Gottfried) that infect cells

FIG. 1. Experimental design to study rotavirus replication in polarized epithelial cells. Human intestinal Caco-2 cells were grown on permeable 0.4 ^m transwell membrane filters until they became polarized. The polarized cells were infected with virus by adding rotavirus to the apical or basolateral surfaces and the infectious process was monitored by immunofluorescence and by measuring the transepithelial cell resistance (TER).

FIG. 1. Experimental design to study rotavirus replication in polarized epithelial cells. Human intestinal Caco-2 cells were grown on permeable 0.4 ^m transwell membrane filters until they became polarized. The polarized cells were infected with virus by adding rotavirus to the apical or basolateral surfaces and the infectious process was monitored by immunofluorescence and by measuring the transepithelial cell resistance (TER).

0 0

Post a comment