Материал: Bovine Viral Diarrhea Virus Diagnosis, Management, and Control

cation include NS2/3, NS4B, NS5A, and NS5B. NS2/3 possesses RNA helicase and nucleotide triphosphatase activities essential for RNA replication (Warrener and Collett, 1995; Grassmannn et al., 1999; Gu et al., 2000). NS5B contains the GDD amino acid motif characteristic of RNA-dependent RNA polymerases and has been demonstrated to possess RNA polymerase activity and be essential for RNA replication (Zhong et al., 1998; Lai et al., 1999). NS4B is an integral endoplasmic reticulum protein that, besides being part of the multiprotein complex composed of NS3, NS4B, and NS5A, can also play a role in suppression of the cytopathic phenotype by mutation of the highly conserved tyrosine residue at position 2441 (Qu et al., 2001). This apparent tripartite protein complex may be associated with the endoplasmic reticulum membrane via the transmembrane domain of NS4B (Qu et al., 2001). The function of NS5A remains elusive and is discussed later with its known interactions with cellular proteins.

Currently, little is known concerning the identity and the role of host factors in the replication of the BVDV genomic RNA. Highly conserved RNA secondary structures at both the 5’ and 3’ ends of the BVDV genomic RNA indicate functional roles (Becher et al., 2000; Yu et al., 2000). As illustrated by the function of the IRES, these secondary structures probably bind specific cellular factors, as well as viral proteins, that participate in RNA replication (Deng and Brock, 1993; Becher et al., 2000). The requirement for the secondary structure contained in the 5’ nontranslated region of the BVDV genomic RNA in viral replication has been analyzed (Yu et al., 2000). Stem-loop Ia, found outside the IRES sequences at the extreme 5’ end of the genomic RNA (Rijnbrand et al., 1997; Chon et al., 1998), appears to regulate the switching of the genomic RNA of BVDV as template for either translation or RNA replication (Behrens et al., 1998; Yu et al., 2000; Li and McNally, 2001). This was supported by findings that regulation of RNA replication and translation was outside the IRES region and was probably due to interaction with regulatory proteins or viral protein(s) associated with specific host protein translation complexes (Li and McNally, 2001).

Several reports have demonstrated interactions of cellular proteins with 3’ nontranslated region secondary structures (both plus and minus strands) of flaviviruses. Several of these proteins have been identified (Blackwell and Brinton, 1995; Blackwell and Brinton, 1997; Ito and Lai, 1997; Li et al., 2002). Two separate reports may provide evidence

for one possible function of NS5A. The first study demonstrated that the cellular protein elongation factor-1 (EF-1 ) recognized and bound to highly conserved secondary structure found on the plusstrand of the 3’ nontranslated region of West Nile virus (Blackwell and Brinton, 1997). The second demonstrated that NS5A protein of BVDV interacts specifically with EF-1 (Johnson et al., 2001). Taken together, these data indicate that binding of EF-1 to both the secondary structure of the 3’ nontranslated region and to NS5A may act to bring the genomic RNA template and the NS5A containing endoplasmic reticulum membrane-bound replication complex into the correct position or orientation for RNA replication to take place. EF1plays numerous roles in RNA sorting and regulation of expression of mRNA in eukaryotic cells and may play a role in RNA replication of BVDV.

The RNA-binding proteins TIA-1 and TIAR have been identified as cellular proteins that interact with the 3’ nontranslated region of West Nile virus (Li et al., 2002). In contrast to EF-1 , TIAR and TIA-1 bind to the 3’ nontranslated region of the minus strand of the West Nile virus genomic RNA. West Nile virus grew to significantly lower titers in murine TIAR-/- knockout cells; however, growth was not eliminated. TIA-1-/- knockout cells showed no decrease in final virus titer but these levels were reached 6 hours later than those grown in normal control cells, indicating distinct requirements for the two proteins. The identification of proteins that differentially bind to the plus and minus strands of the genomic RNA may indicate a mechanism to distinguish between the two strands during replication. These data demonstrate that host proteins that normally interact with cellular RNA are essential to the flavivirus RNA replication process. It is expected to be similar in BVDV RNA replication.

The function of some BVDV proteins may be regulated through phosphorylation by cellular protein kinases. The NS5A protein of BVDV was shown to be phosphorylated (Reed et al., 1998). The ability of NS5A to interact with both cellular and viral proteins and perhaps regulate their function may be dependent on its phosphorylation state (Kapoor et al., 1995; Reed et al., 1998). The phosphorylation of the BVDV NS5A protein indicates that it is associated with serine/threonine protein kinases (Reed et al., 1998). Kinase inhibitor studies indicated that the associated kinases belong to the CMGC family of protein kinases (Reed et al., 1997). This interaction has been postulated to regulate viral replication and possibly host gene expression, whether through trans-

port of NS5A into the nucleus or through interactions with cellular signaling pathways. An example of the latter includes the interaction with the host double-stranded RNA-activated protein kinase (PKR) by the HCV NS5A protein. This interaction inhibits the host interferon antiviral response, including inhibition of the phosphorylation of translation initiation factor eIF-2 and subsequent decrease in protein translation (Gale et al., 1997).

CELL DEATH

Apoptosis, or programmed cell death, is the genetically and biochemically defined mechanism of cellular suicide that may be induced for a number of physiological reasons. From a virological standpoint, this is one of the first lines of defense against viral infection. Its primary purpose is to kill the virus-infected cell before the virus replicates and spreads to neighboring cells. Apoptosis is also a means to kill the cell without inducing an inflammatory response that may damage surrounding tissue (Steller, 1995). Apoptosis is characterized by many cellular changes, including cellular swelling, loss of plasma membrane integrity, loss of mitochondrial membrane potential, release of compartmentalized molecules into the cytoplasm, proteolytic cleavage of proteins (both activation and inactivation of function), and degradation of nuclear DNA. Apoptosis is induced in cells by two major pathways, the extrinsic and intrinsic pathways. The extrinsic pathway is induced when an extracellular signal is received and transduced into the cytoplasm. This results in the activation of caspase 8, initiating the caspase-mediated destruction of the cell. The intrinsic pathway is activated by an intracellular stimulus that results in loss of mitochondrial membrane potential ( m) and release of cytochrome c. Cytochrome c, free in the cytoplasm, is bound by APAF-1 and, along with ATP, activates caspase 9, thus beginning the caspasemediated cellular destruction.

Many viruses encode proteins that interact with cellular defense mechanisms, resulting in the inhibition of apoptosis until viral replication steps have been conducted, allowing production of maximal levels of progeny virus. At this point, no inhibitor of apoptosis has been specifically identified in the BVDV genome. The hallmark of the cp BVDV strains is the induction of cell death in cultured epithelial cells soon after infection, generally within 24–48 hours. The inability of the cp BVDV strains to infect susceptible cells without inducing cell death is considered loss of function, because the more prevalent noncytopathic BVDV (ncp BVDV)

strains possess this ability. The loss of the ability to prevent cell death by cp BVDV strains appears to be related to the genetic changes that occur within the NS2/3 protein coding sequences that give rise to the formation of NS3 (Hoff and Donis, 1997; Lambot et al., 1998), although other mechanisms cannot be ruled out (Bruschke et al., 1997; Vassilev and Donis, 2000; Qu et al., 2001).

Late in the infection process, cells infected with cp BVDV strains show many of the classic signs of apoptosis. These include rounding of cells, cleavage of nuclear DNA to oligonucleosomal fragments, and cleavage and inactivation of poly (ADP-ribose) polymerase, a nuclear enzyme important in DNA repair (Zhang et al., 1996; Hoff and Donis, 1997). In an investigation of the mechanism of induction of apoptosis in cp BVDV-infected cells, Grummer et al. (2002a) demonstrated that cp BVDV-infected cells induced the intrinsic apoptotic pathway. This was shown by translocation of cytochrome c into the cytoplasm, increased expression of APAF-1, and increased caspase 9 activity that is indicative of APAF-1 activation. Inhibition of loss of mitochondrial m and subsequent loss of cytochrome c into the cytoplasm delayed onset of apoptosis. Additionally, treatment of cells with apoptosis inhibitors delayed induction of apoptosis (Grummer et al., 2002b). Disruption of the m disrupts normal cellular oxidation/reduction. Schweizer and Peterhans (1999) provided evidence that cp BVDV-infected cells show an increase in reactive oxygen species, indicative of oxidative stress that preceded caspase activation. Antioxidants that protected the cell from oxidative stress, prevented apoptosis. Interestingly, this had no effect on virus replication and virus titers, indicating that the loss of m and induced oxidative stress plays no role in cp BVDV replication (Schwiezer and Peterhans, 1999; Grummer et al., 2002a).

The envelope glycoprotein Erns has been reported to possess RNase activity (Schneider et al., 1993; Hulst et al., 1994), the function of which is unknown. The effect of Erns on cells infected with CSFV was investigated by Bruschke et al. (1997). This was done because of the immunomodulatory effects of RNases (Tamburrini et al., 1990; D’Alessio, 1993) and the known immunosuppression associated with pestiviral infections. Treatment of lymphocytes from porcine, bovine, ovine, and human sources with Erns completely inhibited concanavalin A activation in vitro. In addition, protein synthesis was inhibited in the Erns-treated cells without disruption of the plasma membrane. Examina-

tion of cells following treatment revealed that apoptosis was induced in the treated cells. These results suggest that Erns may, at least in part, be responsible for the leukopenia observed in pestivirus infected animals. Earlier work had shown that Erns was secreted in the medium of cultured cells (Rümenapf et al., 1993). If this is true in vivo, circulating 0.0 may be able to effect cell death in lymphocytes without direct viral infection and replication. The mode of entry into the susceptible cell remains a mystery. Langedijk (2002) demonstrated that short peptides corresponding to the C-terminus of Erns were transported across the plasma membrane and targeted to the nucleolus. The region involved, as short as 13 amino acid residues in length, appeared to be associated with conserved basic residues. These C- terminal sequences, when attached to unrelated proteins, would still function to mediate transport into the cell. The translocation was rapid, occurring in 1 minute or less, and was not dependent on energy or cell surface receptors. Delivery of the peptide to the nucleolus may indicate that Erns has a role in gene expression or protein synthesis modulation.

BVDV, like all members of the Flaviviridae, utilize the endoplasmic reticulum in their replication process with the envelope proteins being inserted into the membrane of the endoplasmic reticulum for further modification. When the endoplasmic reticulum experiences loss of Ca2+ or the accumulation of misfolded or unassembled proteins, the endoplasmic reticulum stress response is triggered. This acts to slow translation of proteins and is mediated through phosphorylation of eIF-2 by the ER kinase PERK (Ma et al., 2002). This is maintained until the problem of the accumulated proteins is resolved (Kaufman, 1999). PERK phosphorylation of eIF-2 also leads to transcriptional induction of endoplasmic reticulum chaperone proteins as well as transcriptional repression of bcl-2 (anti-apoptotic protein). If the PERK-mediated protein translation interruption is of sufficient duration, apoptosis may be induced by decreased levels of bcl-2 (Friedman, 1996). Jordan et al. (2002b) demonstrated that MDBK cells infected with cp BVDV induce the ER stress response. This response was characterized by activation of PERK, hyperphosphorylation of eIF- 2 , and decreased transcription of bcl-2. In addition, increased caspase 12 activity and decreased glutathione levels were noted. The latter would be a consequence of decreased bcl-2 protein levels and could potentially contribute to increased susceptibility to reactive oxygen species (Schweizer and Peterhans, 1999; McCullough et al., 2000). Overex-

pression of bcl-2 in cells infected with dengue virus and Japanese encephalitis virus was demonstrated to change the cellular response to the infection from apoptotic to chronic (Su et al., 2001; Su et al., 2002).

INHIBITION OF SIGNALING

CYTOKINES

The innate immune response represents the first line of defense against an invading virus and occurs at the cellular level. Interferon production and induction of apoptosis represent two mechanisms a cell may employ to attempt to limit the infection. Synthesis and release of interferon act not only at the infected cell but also to notify other cells of the danger posed by the virus. Many viruses encode proteins that target components of the innate immune system to limit the response, giving the virus time to replicate to maximal levels. BVDV appears to be no exception; however, the mechanisms utilized to bring this about remain unclear. Adler et al. (1994), working in vitro with bovine bone mar- row–derived macrophages (BBMM), demonstrated that only ncp BVDV strains primed BBMM for enhanced nitric oxide production in the presence of Salmonella. Infection of cells by cp BVDV resulted in cytopathic effect (apoptosis), demonstrating that the two biotypes acted differently in affecting specific functions of the host. In a later study, Adler et al. (1996), again working with BBMM, showed that infection with both biotypes resulted in decreased production of tumor necrosis factor (TNF ) upon stimulation with Salmonella, while other specific macrophage functions were not affected. This suggested that decline in the ability to produce TNF may contribute to immunosuppression often observed in BVDV infections. In addition, the infected BBMM also produced a substance that primed uninfected cells for decreased nitric oxide production and for apoptosis following treatment with lipopolysaccharide (Adler et al., 1997; Jungi et al., 1999). This substance was believed to be interferon, based on its physicochemical properties (Adler et al., 1997; Perler et al., 2000).

INTRACELLULAR SIGNALING INHIBITION

The production of interferons by virus-infected cells is an important function of the innate immune response. The mechanism of inhibition of interferon synthesis by ncp BVDV in infected cells remains unclear. ncp BVDV strains, but not cp BVDV strains, possess a function that inhibits interferon production in response to infection or treatment with

dsRNA, but not the response to exogenous interferons (Schweizer and Peterhans, 2001). dsRNA is a molecule produced during the infection process by most viruses and has been shown conclusively to be a specific trigger of apoptosis and interferon production (Jacobs and Langland, 1996; Kibler et al., 1997). This inhibition was observed only in cells that had been infected with the ncp BVDV strain for at lease 12 hours, indicating that inhibition was virus-induced. ncp BVDV inhibited induction of apoptosis by dsRNA but not by staurosporine or actinomycin D. In addition, this inhibition was through an unknown intracellular mechanism and not by inhibition of uptake of the dsRNA. This suggested that some component encoded by ncp BVDV interfered specifically with intracellular signaling. By using different cell lines and different strains of ncp BVDV, Schweizer and Peterhans (2001) demonstrated that interference with apoptosis and interferon synthesis was a general and important function of ncp strains.

Baigent et al. (2002) extended the above studies to further characterize the lack of intracellular response in cells infected with ncp BVDV. As described by Schweitzer and Peterhans (2001), prior infection with ncp BVDV did not block the response to exogenous interferons but could block interferon and MxA gene transcription in response to dsRNA or infection by the heterologous virus Semliki Forest virus (SFV). Prior infection with ncp BVDV did not block induction of apoptosis by cp BVDV or SFV. The nature of the interferon response requires transcriptional up-regulation or posttranslational modification of specific transcription factors, including NF- B, ATF2, and c-jun. Examination of these proteins revealed that NF- B was not activated, and ncp BVDV infection did not block activation of NF- B by SFV or by tumor necrosis factor-. The stress-activated kinases JNK1 and JNK2 were not activated nor were the transcription factors ATF2 and c-Jun phosphorylated. These events were not inhibited following superinfection by SFV following ncp BVDV-infection. Interferon regulatory factor-3 (IRF-3), a transcriptional activator responsible for the increased transcription of interferon genes, was translocated to the nucleus in ncp BVDV-infected cells but was shown to lack DNA binding activity in nuclear extracts (Baigent, et al., 2002). In addition, IRF-3 DNA binding activity was present in SFV-infected cells but could be blocked by prior infection with ncp BVDV.

HCV was recently shown to also inhibit IRF-3 transcriptional activation of interferon genes (Foy et

al., 2003) but by an apparently mechanistically different means. The HCV NS3/4A serine protease prevented the phosphorylation of IRF-3, blocking its translocation to the nucleus. This is in contrast to the IRF-3 in ncp BVDV-infected cells where IRF-3 is translocated to the nuclease but does not bind DNA. The two viruses appear to utilize two different mechanisms to achieve the same goal.

A recent report by Ruggli, et al. (2003) suggested that the pestivirus protein Npro may play a role in interference with the induction of the interferon response. Porcine SK-6 cells infected with wild-type CSFV possessed the ability to resist induction of apoptosis when treated with dsRNA, and macrophages were inhibited from producing an /ß interferon response. When infected with CSFV mutants lacking the Npro coding sequences, SK-6 cells were not protected from dsRNA-induced apoptosis and macrophages produced a type I interferon response in the absence of additional stimuli. In addition, reduced replication was noted for the Npro-deficient mutant. These data imply that Npro plays a role in the inhibition of the interferon response, the mechanism of which still remains to be determined.

CELLULAR REMODELING

Recent advances in functional genomics have made it possible to examine changes in gene expression in cells under a variety of conditions that allow an understanding of the basic mechanism(s) that bring about distinct cellular changes. These changes, indicated by increased or decreased transcription of specific genes, allow the dissection of the host response based on the function of the genes with altered expression levels. One functional genomics technology, serial analysis of gene expression (SAGE), was applied to examining gene expression changes that take place in BVDV-infected cells in response to the viral infection (Neill and Ridpath, 2003a, b). This analysis revealed that a number of gene expression changes occur following BVDV infection and that some of the changes act to increase efficiency of functions that are beneficial to viral replication, maturation, and release. Although expression level of genes involved in energy production and metabolism were relatively unchanged, genes encoding proteins involved in protein translation were altered in a pro-virus manner. Protein translation plays such an integral role in the replication of a virus that alterations in the translational machinery that enhances viral protein synthesis are advantageous. These changes included ribosomal proteins, elongation factors, and tRNA synthetases. In addition,

transcription of genes involved in transport of nascent polypeptides into the lumen of the endoplasmic reticulum was increased, indicating possible increased capacity for ER translocation and processing. One protein in particular, TRAM, functions in regulating the Sec61p pore complex through which the nascent polypeptide passes into the lumen of the endoplasmic reticulum (Hegde et al., 1998a). TRAM also regulates exposure of the nascent protein to the cytoplasm in response to pause transfer signals (Hegde et al., 1998b). This may be important for BVDV replication in allowing the correct processing of the structural proteins. Gene expression changes were also noted with unknown effect or benefit to the replicating virus. Several tubulin isotypes showed sharp declines in transcript numbers, indicating possible disruption of microtubule function. This may suggest a possible mechanism for the disruption of platelet production by BVDV-infected megakaryocytes. Proper microtubule function is essential for pro-platelet release, and disruption of the microtubule network would have a detrimental effect on this process.

VIRUS RELEASE

Little is currently known concerning release of infectious BVDV virions from the infected cell. This is one of the more confusing aspects of flavivirus/ pestivirus biology because of apparent differences in the mechanisms of viral egress used by different viruses. Egress by budding from the cell surface was reported for West Nile (Sarafend) virus and was demonstrated by both transmission and scanning electron microscopy (Ng et al., 1994). Examination of release of virus in polarized epithelial cells by Chu and Ng (2002) showed that West Nile (Sarafend) was released from the apical surface, and Kunjin virus was released from both apical and basolateral surfaces. In addition, microtubules were shown to play a role in the sorting and transport of viral proteins to the cell surface by West Nile (Sarafend) virus, and disruption of the microtubule network had no effect on Kunjin virus maturation and egress (MacKenzie and Westaway, 2001; Chu and Ng, 2002). Transport of Kunjin virus to the cell surface from the golgi apparatus appears to be by movement of virus-containing vesicles and fusion with the plasma membrane.

The release of mature pestivirus particles appears at this point to more closely resemble that of Kunjin virus. A study examining location of viral proteins in cells infected with CSFV (Weiland et al., 1999) showed that the viral proteins were associated only

with released viral particles. Lack of detection of viral proteins in the plasma membrane of BVDVinfected cells was confirmed by Grummer et al. (2001) using indirect fluorescence microscopy, confocal microscopy, or FACS analysis. Using subcellular fractionation techniques, the envelope glycoproteins Erns and E2 were found associated exclusively with intracellular membranes. In addition, with the possible disruption of microtubule networks in BVDV-infected cells (Neill and Ridpath, 2003), the microtubule-associated sorting of viral proteins required for surface budding would be affected. This provides further evidence for the use of an egress mechanism more closely resembling that of Kunjin virus.

BVDV REPLICATION CYCLE OVERVIEW

The mechanisms utilized by BVDV to enter a susceptible cell and replicate itself are becoming clearer. It is now possible to put together the picture of what takes place from attachment of the infecting virus particle to release of infectious progeny. As detailed in Figure 11.4, the infection process begins by attachment of the BVDV particle to the plasma membrane, probably first by attachment of Erns to a docking surface glycosaminoglycan(s) followed by binding of the LDLR E2 (Figure 11.4A). The internalization of the virion is probably mediated by the LDLR receptor through endocytosis (Figure 11.4B,C), and release of the genomic RNA occurs following acidification of the endosomal vesicle (Figure 11.4D,E). The genomic RNA, now in the cytoplasm of the cell first acts as mRNA for translation of the polyprotein at least initially taking place on the endoplasmic reticulum (Figure 11.4F). At some point, a switch occurs that causes the genomic RNA to be used as template for RNA replication rather than for protein translation (Figure 11.4G). It is presumed that some of the newly synthesized daughter RNAs are then used for protein translation while others participate in RNA replication. The genomic RNAs also interact with the capsid protein followed by recognition of the cytoplasmic domains of the envelope proteins and budding into the lumen of the endoplasmic reticulum (Figure 11.4H,I,J). The passage of the immature particles through the endoplasmic reticulum and the golgi body result in the maturation of the particles by processing and glycosylation of the envelope proteins (Figure 11.4K). The virus-laden vesicles (Figure 11.4L) move through the cytoplasm to the cell surface (Figure 11.4M) where the vesicle fuses with the plasma