African swine fever is an economically important haemorrhagic fever of domestic pigs which is caused by a large DNA virus (ASFV). Virus isolates vary in virulence, the most virulent isolates causing 100% mortality but isolates of reduced virulence have emerged as the virus has circulated in domestic pigs. Pigs that recover from infection may remain persistently infected for long periods. The natural hosts for ASFV are warthogs, bushpigs and soft ticks of the Ornithodoros species (O. moubata in Africa and O. erraticus in southern Europe). The virus persistently infects these wildlife hosts showing no signs of disease. The ability of the virus to persist in its hosts shows that it has mechanisms to evade host defence systems. The main target cells for virus replication in vivo are macrophages. ASFV is a cytoplasmic, icosahedral, virus containing a double-stranded DNA genome of 170 to 193 kbp and is the only member of a new virus family, the Asfarviridae.
Evasion of host defences
The ASFV genome encodes 160 to 175 proteins including enzymes required for replication and transcription of the virus genome and virus structural proteins. The virus also encodes proteins, which are not essential for virus replication in cells, but play roles in virus host interactions that are important for virus survival and transmission in its hosts. Amongst these are genes which help the virus to evade host defence systems [ 1, 2, Chapman et al., 2008].
The virus replicates in macrophages and by modulating their function can profoundly affect the host response to infection. Host defence evasion proteins encoded by ASFV include two homologues of host proteins that inhibit apoptosis (Bcl 2 and IAP homologues) [ 3, 4, 5]. These inhibit apoptosis of infected cells hence enabling production of progeny virions.
The ASFV A238L protein inhibits key pathways involved in activating transcription of immunomodulatory genes. A238L protein inhibits signalling pathways in infected cells which are involved in the transcriptional activation of host immune response genes. A238L binds to and inhibits host calcineurin (CaN) phosphatase thus inhibiting pathways dependent on CaN. A238L also inhibits transcriptional activation mediated by interaction of a number of transcription factors with the p300 transcriptional co-activator and thus acts to inhibit transcriptional activation of a wide range of host genes. [ 6, 7, 8, 9]. We are currently investigating the mechanism by which the A238L protein functions and its effects on cellular functions (Abrams et al., 2008, Silk et al., 2007).
Another ASFV immunomodulatory protein, called CD2v, is a transmembrane protein that resembles the host CD2 protein in its extracellular domain. This protein causes binding of infected cells and virus particles to red blood cells, thus facilitating virus dissemination in infected pigs and hiding virus and infected cells from the immune system. Expression of CD2v is also required for the ability of ASFV to inhibit proliferation of lymphocytes, which are not themselves infected with virus [ 10, 11, 12]. We have shown that the cytoplasmic tail of CD2v binds to an adaptor protein, SH3P7, which is involved in regulating protein transport and signalling pathways in cells (Kay-Jackson et al ,2004). Modulation of these pathways in infected macrophages may alter the pattern of cell surface and secreted proteins providing a possible mechanism for the immunomodulatory activity of CD2v. We are currently defining the consequences of this interaction on host cell function. Using virus mutants we have shown that expression of the CD2v protein is required for efficient virus replication in the soft tick vector. The protein appears to act by enhancing uptake of virus across the tick gut, and this is mediated by binding of virus particles to red blood cells (Rowlands et al., 2009).
The DP71L protein resembles the Herpes simplex virus neurovirulence factor ICP34.5 and the host DNA damage response protein GADD34 over a C-terminal domain. These proteins act as regulatory subunits of protein phosphatase 1 (PP1) by binding to PP1 and targeting the dephosphorylation of substrate proteins. All target PP1 to reverse the phosphorylation of translation initiation factor eIF-2 which is induced by stress and virus infection and leads to inhibition of general translation. The DP71L protein is also localised in the nucleus and we are investigating the role of this protein in the nucleus (Goatley et al., 2004).
Genome comparisons of high and low virulence isolates
We have determined the complete coding sequences of several high and low virulence ASFV isolates to identify genome changes that may be related to virulence (Chapman et al., 2008). This analysis showed that the low virulence isolate OURT88/3 has a deletion near the left end of 6 multigene 360 and 2 multigene 530 genes . These genes have been implicated in virulence and cell tropism by deletion from the genome of a virulent isolate and their deletion from the OURT88/3 genome is a likely explanation for the reduced virulence of this isolate. In addition the OURT88/3 has interruptions in genes encoding two immunomodulatory proteins the CD2v gene and C-type lectin gene. The OURT88/3 isolate provides a model for understanding the mechanisms and antigens involved in conferring protective immunity. A current project is investigating the function of the MGF 360 and 530 genes by studying their localisation and interaction with cellular proteins.
The sequence of the highly virulent isolate introduced to Georgia in 2007 has been determined to provide information for vaccine development (Rowlands et al., 2007, unpublished results).
Host factors involved in asfv responses and replication
Defining the response of macrophages to ASFV infection is critical to understanding mechanisms of virus pathogenesis and immune evasion. We have used porcine microarrays and quantitative reverse transcriptase PCR to study macrophage transcriptional responses following infection with low and high virulence ASFV isolates. Most of the host genes whose expression is differentially regulated showed increased expression levels at 4 hours post-infection but a return to base levels at later times post-infection. These included genes expected to be activated as part of the host response to infection and included those encoding proinflammatory cytokines, chemokines, cell surface proteins, components of signalling pathways and stress response proteins. These results are consistent with an initial activation of host defence response genes followed by a shut-off in their expression as virus-encoded proteins which inhibit host transcription factor activation are expressed (Zhang et al., 2006). The functional consequences of these changes in transcription are under investigation including analysis of the chemokine response to infection in vitro and in vivo.
Currently no vaccine is available against ASFV. Improved understanding of ASFV genes involved in virulence and immune evasion have made the development of a rationally attenuated ASFV vaccine feasible. A strategy has been developed to delete these and other genes from the genome of the Benin 97/1 isolate. The Vaccinology Group have developed assays which correlate with induction of protective immunity and these can be used to identify the best candidate vaccine strains. An alternative approach carried out in collaboration with Arizona State University and the Vaccinology Group involves screening of the ASFV proteome for antigens which induce a protective immune response.