Alphaherpesviruses and Chemokines: Pas de Deux Not Yet Brought to Perfection

The coexistence of viruses and their hosts implies constant and mutual evolutionary pressure. In addition to the fundamental systems necessary for viruses to replicate and spread, viruses have developed accessory systems to escape killing by the host's immune system. Herpesviruses have been coevolving with their hosts over millions of years and are exquisitely well adapted to their respective partners. Biological criteria have long been used to subdivide the family Herpesviridae into three subfamilies, namely Alpha-, Beta-, and Gammaherpesvirinae. Members of the Alphaherpesvirinae have a narrow in vivo host range, a short replication cycle, and the capacity to establish lifelong, latent infections, primarily but not exclusively in neurons of sensory ganglia (50). Their linear, double-stranded DNA genomes vary between 124 and 177 kbp in length and generally consist of regions of unique sequences flanked by direct or inverted repeat sequences. The subfamily includes human pathogens as well as a number of animal viruses of considerable agricultural and economic importance (Table 1). The human pathogens herpes simplex virus type 1 (HSV-1), HSV-2, and varicella-zoster virus (VZV) are the causative agents of cold sores, genital ulcerous disease, and chickenpox/shingles, respectively. Some of the animal herpesviruses can cause diseases with potentially devastating economic consequences. Infection with equine herpesvirus type 1 (EHV-1) results in respiratory disease, abortion, and neurological disorders; bovine herpesvirus type 1 (BHV-1) leads to respiratory infections and abortions in cattle; pseudorabies virus (PRV) (suid herpesvirus 1) infection (Aujeszky's disease) is characterized by respiratory and neurological disorders, abortion, and infertility in swine; and Marek's disease virus (MDV), an oncogenic alphaherpesvirus, causes massive immunosuppression and invariably lethal T-cell lymphomas in unvaccinated chickens.

TABLE 1.

Genera and members of the subfamily Alphaherpesvirinae

Genus	Virus	Abbreviation	Host	Gene/protein with virokine/viroceptor/vCKBP function	Reference
Simplexvirus	Human herpesvirus type 1	HSV-1	Human
	Human herpesvirus type 2	HSV-2	Human	gG peptides	9, 10, 11
	Bovine herpesvirus type 2 (bovine mammilitis)	BHV-2	Cattle	(immunomodulator?)
	Ateline herpesvirus type 1	AtHV-1	Spider monkey
	Cercopithecine herpesvirus type 1	CeHV-1	Macaque
	Cercopithecine herpesvirus type 2	CeHV-2	Macaque
	Cercopithecine herpesvirus type 16	CeHV-16	Baboon
	Saimiriine herpesvirus type 1	SaHV-1	Marmoset
	Macropodid herpesvirus type 1	MaHV-1	Wallaby
	Macropodid herpesvirus type 2	MaHV-2	Wallaby
Varicellovirus	Human herpesvirus type 3 (varicella-zoster virus)	VZV	Human
	Cercopithecine herpesvirus type 9	CeHV-9	Macaque
	Equid herpesvirus type 1 (equine abortion herpesvirus)	EHV-1	Horse	gG (viroreceptor/vCKBP)	14, 57, 58, 59
	Equid herpesvirus type 3 (equine coital exanthema virus)	EHV-3	Horse	gG (vCKBP)	14
	Equid herpesvirus type 4 (equine rhinopneumonitis virus)	EHV-4	Horse
	Equid herpesvirus type 6	EHV-6	Donkey
	Equid herpesvirus type 8	EHV-8	Donkey
	Suid herpesvirus type 1 (pseudorabies virus)	PRV	Pig
	Bovine herpesvirus type 1 (infectious bovine rhinotracheitis)	BHV-1	Cattle	gG (vCKBP)	14
	Bovine herpesvirus type 5 (bovine encephalitis herpesvirus)	BHV-5	Cattle	gG (vCKBP)	14
	Bubaline herpesvirus type 1	BuHV-1	Water buffalo
	Ovine herpesvirus type 1 (sheep pulmonary adenomatosis-associated herpesvirus)	OvHV-1	Sheep
	Caprine herpesvirus type 1	CapHV-1	Goat	gG (vCKBP)	14
	Cervid herpesvirus type 1	CerHV-1	Reindeer	gG (vCKBP)	14
	Rangiferine herpesvirus type 1	RanHV-1	Reindeer	gG (vCKBP)	14
	Phocid herpesvirus type 1	PhoHV-1	Seal	gG (unknown)
	Felid herpesvirus type 1	FeHV-1	Cat	gG (viroreceptor/vCKBP)	19, 20
	Canid herpesvirus type 1	CaHV-1	Dog	gG (unknown)
Mardivirus	Gallid herpesvirus type 2 (Marek's disease herpesvirus)	MDV	Chicken	vIL-8 (virokine)	18, 22, 37, 48
	Gallid herpesvirus type 3 (Marek's disease herpesvirus 2)	GaHV-3	Chicken
	Meleagrid herpesvirus type 1	HVT	Turkey
Iltovirus	Gallid herpesvirus type 1 (infectious laryngotracheitis virus)	ILTV	Turkey, chicken	gG (vCKBP?)	24, 25
	Psittacid herpesvirus type 1 (Pacheco disease virus)	PsHV-1	Parrot

Infection with herpesviruses, as is the case with most viruses, normally stimulates the production of cytokines and chemokines, and some of the components of the immune system for viral subversion are ligands and receptors of the cytokine and chemokine network (4, 56). These secreted proteins mediate and regulate fundamental processes such as immune responses, inflammation, and hematopoiesis and play a crucial role in leukocyte migration during both innate and adaptive immune responses. Certain cytokines, such as interferons and tumor-necrosis factor, result in intracellular signals that can lead to an antiviral state and/or apoptosis of the cell and thereby limit viral replication (13). Several cytokines aid in enhanced immune recognition and modulate immune responses that protect against viral infection, and they can even mediate the killing of infected cells by natural killer (NK) cells or cytotoxic T lymphocytes (29).

Chemokines are chemoattractant molecules that regulate the traffic and effector functions of leukocytes and are key regulators of inflammation and immune surveillance (5). Functionally they can be divided into two major groups: housekeeping chemokines, which are expressed constitutively, and proinflammatory chemokines, which are typically inducible. The physiological activities of chemokines are mediated by selective recognition and activation of chemokine receptors belonging to the seven-membrane domain, G protein-coupled receptor superfamily (GPCRs) (52). In addition, chemokines bind to glycosaminoglycans (GAGs) through distinct binding sites. Chemokine binding to GAGs on cells, particularly endothelial cells, results in chemotactic chemokine gradients that allow the correct presentation of chemokines to leukocytes and therefore enable target cells to cross the endothelial barrier and migrate into tissues (15, 16, 44) (Fig. 1).

FIG. 1.

Chemokine functions. Chemokines are produced at sites of infection and form chemotactic gradients by interacting with glycosaminoglycans at the surfaces of endothelial cells. Leukocytes expressing the appropriate chemokine receptors (seven-transmembrane cell surface G protein-coupled receptors) respond to the chemokines and migrate to sites of infected tissue. The presentation of chemokines to leukocytes by chemotactic gradients is required for correct presentation of chemokines in vivo and leukocyte migration through the vascular endothelium into infected or damaged tissue.

Given the central role of cytokines and chemokines in antiviral defense, it is not surprising that herpesviruses have evolved strategies to subdue pivotal elements of this network to their service. For the Beta- and Gammherpesvirinae, several virus-encoded proteins with cytokine/chemokine modulatory properties have been identified based on their sequence similarities with host cytokines and chemokines (2, 3, 36, 39, 53). In many cases, viral cytokine/chemokine modulators are derived from host genes and were originally pirated during ancestral virus infections. Consequently, they have evolved as virus constituents that allow their carriers, the viruses themselves, to modify or evade the antiviral defense. Interestingly, of the Alphaherpesvirinae, only MDV has been shown to express a viral chemokine modulator, called viral interleukin 8 (vIL-8), with homology to a chicken gene (37, 48). Several studies investigating evolutionary relationships within the Herpesviridae have shown that the alphaherpesviruses are the most recently evolved, and it has been proposed that within this subfamily, MDV is the original alpha class antecedent species, which was later transferred from birds into mammals (32, 42). These evolutionary considerations raise the possibility that mammalian alphaherpesviruses may be too “early” in their coevolutionary relationships with their hosts to have hijacked genes encoding chemokines. On the other hand, and more likely, it has been noted that molecular mimicry by viral proteins does in fact resemble the interspecies diversity of the host immune pathways themselves (45). Many alphaherpesviruses cause infections that are initiated through the respiratory or genital route and are restricted to immunologically privileged sites, such as the central and peripheral nervous system, where host immune responses are more repressed (50). This implies that alphaherpesviruses might have fashioned virus-encoded proteins, which account for immunomodulatory functions that are different from those of other subfamilies, and adapted them to their very specific and unique needs. Indeed, alphaherpesviruses are well known for the expression of the glycoprotein E-glycoprotein I complex, an Fc receptor-like molecule targeting the constant region of immunoglobulins, and the expression of glycoprotein C, which binds complement factor C3b. These viroreceptors were shown to allow viruses to avoid recognition and destruction by the complement system in vitro and in vivo. However, complement immune evasion strategies used by alphaherpesviruses have previously been reviewed extensively and are therefore beyond the scope of this review (28, 38), where we will focus on more recent findings on alphaherpesviral interactions with other immunomodulatory functions. Here we give an updated overview of the recent developments with chemokine interference by the Alphaherpesvirinae, more specifically the alphaherpesvirus-encoded vIL-8 and glycoprotein G (gG).

In spite of the absence of alphaherpesviral mimicry of cytokines and chemokines, with the notable exception of the virokine vIL-8 encoded by MDV, there are recent data indicating that alphaherpesviruses are in fact capable of effectively modulating the chemokine network to their benefit. Several members of the Alphaherpesvirinae subfamily express gG, a viral protein shown to interfere with a broad range of chemokines that appears to intercept chemokine networking at different levels (14, 63). It is these viral factors that have garnered attention lately, and we will provide a description of their properties and putative functions.

vIL-8

MDV, or gallid herpesvirus type 2 (GaHV-2), is the only alphaherpesvirus shown to encode and express a virokine, vIL-8 (37, 48). Most likely, vIL-8 was pirated from the chicken genome after the divergence of the members of the Mardivirus genus, since nononcogenic close relatives of MDV, gallid herpesvirus type 3 (GaHV-3) and meleagrid herpesvirus type 1 (herpesvirus of turkeys), do not harbor an IL-8-like gene. Two copies of vIL-8 in each of the long repeat regions are present in the MDV genome. vIL-8, which is encoded by three exons (I, II, and III), shares significant homology with cellular CXC chemokines like IL-8, also designated CXCL8, and GRO-α. Exon I of vIL-8 is rich in hydrophobic residues and serves as a signal peptide, while exons II and III contain the CXC motif and a three-amino-acid motif (DKR) that determines specificity (48). Chicken IL-8, originally designated chicken chemotactic and angiogenic factor, is the product of the 9E3/CEF4 gene and shares a high level of amino acid similarity with human IL-8 (40). In contrast to human IL-8, which is chemotactic for neutrophils, however, chicken IL-8 predominantly targets cells of the monocyte/macrophage lineage (40). Similar to chicken IL-8, vIL-8 encoded by MDV also functions as a chemoattractant for chicken peripheral blood mononuclear cells when expressed and tested in chemotaxis assays in vitro (48).

The chemoattractant specificity of vIL-8 is an excellent example of a cellular gene that is pirated and tailored to the needs of the virus by strong and regulated expression at early times after virus uncoating. Upon entry into the chicken and passage to lymphoid organs by hijacking of antigen-presenting cells, MDV requires B and activated T cells for efficient replication. It is in the former where the virus lytically replicates and the latter where MDV establishes latency and induces transformation. It is unknown exactly what function vIL-8 serves during MDV pathogenesis (Fig. 2). It has been suggested that secretion of vIL-8 by infected cells might help recruit lymphocytes to initially infected cells that function as “virus ferries” and carry MDV from the periphery to primary lymphatic organs. The recruitment of lymphocytes helps increase the efficiency of early virus replication, since MDV spreads from cell to cell only, which requires quite intimate contacts between infected and new target cells. Alternatively, vIL-8 may act as a mimicry molecule, helping to evade the immune system by antagonizing host IL-8 responses. Still a third possibility is that vIL-8 expression augments viral replication by binding to a receptor on infected cells and activating a transcriptional/translational cascade inducing MDV promoters. Based on experiments using MDV vIL-8 deletion mutants, the first function is favored. Deletion of both copies of vIL-8 in the very virulent RB-1B MDV strain showed that, while in vitro replication in tissue-cultured cells was unaltered, in vivo replication was severely impaired (18, 30, 48). Likewise, Cui et al. (22) showed that the numbers of infected cells in lymphoid organs (bursa of fabricius, thymus, and spleen) were significantly lower in viruses lacking vIL-8. Consistent with the behavior of deletion mutants, recombinant vIL-8 strongly binds predominantly to B but also to T lymphocytes, as demonstrated with a baculovirus-expressed vIL-8 tagged with human Fc (J. P. Kamil and N. Osterrieder, unpublished observation). Thus, it appears that MDV maintains and utilizes vIL-8 for its replication. According to current knowledge, other alphaherpesviruses have not subverted a cellular chemokine for their purposes, although some of the mammalian species, such as VZV and EHV-1, also exhibit strong lymphotropism and would seem to have a vested interest in such a mechanism for manipulating the chemokine environment and attracting putative targets or excluding unwanted visitors.

FIG. 2.

Potential functions of MDV vIL-8. There are three proposed functions of vIL-8. (A) The first and most likely possible function of vIL-8 is the attraction of B and T lymphocytes to infected cells by secretion of vIL-8 and migration of uninfected cells by a chemoattractant gradient. (B) Another possibility is the secretion of vIL-8 from infected cells, which antagonizes chicken IL-8 binding to the IL-8 receptor, thus blocking the function of the chicken chemokine. (C) A third possibility is vIL-8 binding to chemokine receptors and the subsequent activation of transcriptional and translational cascades that leads to enhanced viral replication and/or migration of infected cells.

gG

gG homologues in several alphaherpesviruses have been described and are expressed as nonessential membrane-anchored proteins with type I topology (8). gG is unusual compared to other herpesvirus glycoproteins, since it also gets secreted into the media of infected cells. gG can therefore exist in three isoforms: a full-length membrane-bound form, a smaller membrane-bound form, and a secreted form (26). The latter two isoforms appear to be the results of a proteolytic cleavage event of the full-length membrane-bound form (26). Alphaherpesviral gG can interfere at different distinct stages of chemokine action and therefore constitutes yet another immunoevasion tool used by alphaherpesviruses (Fig. 3). The full-length, membrane-anchored gGs of feline herpesvirus type 1 (FeHV-1) and equine herpesvirus type 1 (EHV-1) can function as viroreceptors and are capable of binding a broad range of chemokines (14, 19). Cleaved gG of several alphaherpesviruses has been described as functioning as a viral chemokine binding protein (vCKBP) and has recently been classified as the prototype of a new subfamily, vCKBP-4 (63). By using cross-linking assays with supernatants from infected cells and recombinant chemokines, it was shown that gG of EHV-3, BHV-1, BHV-5, Rangiferine herpesvirus type 1, Caprine herpesvirus type 1, and Cervid herpesvirus type 1 (Table 1) also bind a plethora of chemokines, with each virus, however, having its own signature of specificities (16). In addition, it has been shown for EHV-1, BHV-1, and FeHV-1 that gG-chemokine interactions prevent the binding of chemokines to GPCRs, thereby neutralizing chemokine activity (14, 19). Moreover, gG can inhibit chemokine activity by blocking the interaction of chemokines with heparin, although gG does not appear to bind heparin directly but rather indirectly through the cross talk of chemokines with GAGs (14). By preventing chemokine-GAG interactions, gG specifically disrupts preestablished chemokine gradients, and in combination with the prevention of chemokine-receptor binding, efficiently controls the local microenvironment of infected tissues. We will now discuss what is known about the general roles of the gGs of the different alphaherpesviruses and how they interfere with the chemokine network.

FIG. 3.

Potential functions of gG. Alphaherpesviral gG functions as a secreted vCKBP, which prevents chemokines from interacting with both chemokine receptors and glycosaminoglycans. (A) As a result, chemokine gradients are neutralized and chemotaxis of leukocytes into virus-infected tissues is inhibited. gG expressed on infected cells might also function as a viroreceptor, thereby sequestering chemokines from the extracellular milieu around infected tissues (B) or promoting infected-cell proliferation or migration (C).

HSV-1 AND HSV-2 gGs

No chemokine binding of HSV-1 and HSV-2 gGs has been reported to date based on the observation that supernatants from HSV-1- or HSV-2-infected cells are unable to cross-link chemokines of murine or human origin (14). For HSV-1, this might simply be related to the fact that its gG is not secreted into the media of infected cells (51). An HSV-1 gG deletion mutant was evaluated in vivo and displayed only marginal attenuation in the mouse ear model, suggesting that the role of gG during HSV-1 pathogenesis might be limited (7). In contrast to HSV-1 gG, the HSV-2 gG homologue is secreted into the media as a 34-kDa moiety representing the ectodomain of the protein (6, 54). Although no specific function has been ascribed to HSV-2 gG as a whole, peptides derived from gG have been shown to possess proinflammatory properties. These gG-derived peptides are not only chemoattractants for monocytes and neutrophils but also have profound downregulatory effects on NK cells (9-11). Still, it remains unclear whether the native HSV-2 gG protein has the same proinflammatory properties as gG-derived peptides and whether (regulated) proteolytic degradation of HSV-2 gG would release peptides with such activities. In addition, the gGs of both simplex viruses have been described to display additional functions which are unrelated to chemokine binding or any other immunomodulatory function. HSV-1 gG appears to be required for infection of polarized epithelial cells through apical surfaces (57). More recently, it was suggested that HSV-2 gG is directly involved in HSV-2 attachment to cells, since gG present in the viral envelope was shown to interact with sulfated polysaccharides, including cell surface GAGs (41).

BHV-1 AND BHV-5 gG

BHV-1 and BHV-5 gGs are nonstructural proteins that are present on the plasma membranes of infected cells and are secreted as 65-kDa polypeptides. In addition, secreted gG can be found as protein species ranging from 90 to 240 kDa when linked to GAGs (27, 33). BHV-1 gG is nonessential for viral growth but essential for cell-to-cell spread in bovine kidney cells (47, 58). Moreover, BHV-1 gG has been proposed to be important for maintenance of intact cell-to-cell junctions (46). The binding of BHV-1 and BHV-5 gG to chemokines was demonstrated using cross-linking assays with both supernatants of infected cells and baculovirus-expressed gG (14). In addition, recombinant BHV-1 and BHV-5 gG inhibited migration of human neutrophils induced by CXCL1 or of alpha interferon-treated human lymphoma cells mediated by CCL-3 (14). In vivo studies using BHV-1 mutants devoid of gG showed significant attenuation and increased immunogenicity in cattle (31). However, since no rescuant virus was used in this particular study and the expression of adjacent genes was not investigated, it is difficult to determine conclusively whether BHV-1 gG plays an important role in pathogenicity, let alone which function, if any, can be attributed to a gG-chemokine interaction.

PRV gG

PRV secretes a nonstructural viral glycoprotein of approximately 99 kDa, which was formerly referred to as gX but more recently renamed gG for its similarity with the gG homologues of other alphaherpesviruses (12). Since PRV gG is not required for efficient growth in vitro and in vivo, gG mutants have been suggested as useful marker vaccines to distinguish between vaccinated and infected pigs, mostly in combination with attenuating mutations in other glycoprotein genes (55). Most gG deletion mutants did not exhibit altered virulence in pigs (34, 43), but one gG mutant based on the PRV Bartha strain did show impaired cell-to-cell spread in vitro and reduced virulence in vivo. This effect, however, was later explained by reduced expression of the upstream US3 gene, which encodes a serine/threonine protein kinase (23). Therefore, in the models employed in the PRV system, gG was shown not to play a major role in PRV pathogenesis, and to our knowledge, experiments to determine the potential role of PRV gG as a vCKBP have not yet been done.

FeHV-1 gG

Recently, gG encoded by FeHV-1, an alphaherpesvirus of cats, has been evaluated for its possible chemokine binding properties. It was first shown that FeHV-1 secretes gG into the culture medium and that secreted gG not only displays high-affinity binding to a broad range of chemokines but is also capable of blocking chemokine activity by preventing chemokine interaction with GPCRs (19). In addition, it has been demonstrated that the membrane-bound form of gG, expressed on the surfaces of infected cells, also binds to a number of chemokines with high affinity (19). It is possible that membrane-bound gG acts as a bona fide viroreceptor, providing a decoy that prevents the interaction of chemokines with cellular receptors and inhibits the biological activity of chemokines. In addition, FeHV-1 gG is a structural protein and is present on the surfaces of virus particles (20). This observation begs the speculation that membrane-bound gG, besides functioning as a viroreceptor, might also play a role in virus attachment to cells which present chemokines bound to GAGs. The FeHV-1 gG homologue may be a determinant for cell and tissue tropism in vivo and/or may aid in virus entry. Although it has been shown that FeHV-1 gG can act as a vCKBP when present on the virion surface, preincubation of virions with chemokines, including CXCL1, CCL3, or XCL1, did not alter the infectivity of FeHV-1, and these data would therefore not support a role for gG in cell and tissue tropism in the chosen in vitro system (20). However, a cell-type-specific interaction between FeHV-1 gG and GAG-bound chemokines on target cells is easily conceivable. Experiments would need to be repeated with feline lymphocytes or other target cells under different conditions.

ILTV gG

ILTV (Iltovirus) causes acute respiratory disease in poultry, and its gG has been identified as a secreted, glycosylated protein of 32 kDa (35). Although no experiments have been performed to evaluate the role of ILTV gG as a vCKBP, some interesting observations have been made while using a gG deletion mutant in the natural host, the chicken. It was shown that gG-deficient ILTV was significantly attenuated in chickens with respect to clinical signs, weight loss, and mortality. The wild-type phenotype was completely restored upon reinsertion of gG, and expression of the adjacent genes was not altered by the genetic manipulations (25). In addition, it was observed that the degree of inflammatory cell infiltration in the tracheas of chickens was increased in the absence of gG, strongly suggesting that ILTV gG may have an immunomodulatory role and act as a vCKBP in vivo (25). In a follow-up study, the same gG deletion mutant was shown to protect specific pathogen-free chickens against clinical signs subsequent to challenge with virulent ILTV, demonstrating the mutant's potential to serve as a new modified live vaccine candidate (23) against this poultry disease affecting the upper and lower airways.

EHV-1 AND EHV-4 gG

Both EHV-1 and EHV-4 are economically important pathogens of horses, and each encodes gG as membrane-associated and secreted forms, the latter representing moieties of approximately 55 to 60 kDa (17, 26, 59). The full-length, membrane-anchored form of EHV-1 gG has vCKBP properties, since recombinant gG expressed on the surfaces of insect cells was capable of binding human CXCL1 and CXCL8 (14). Secreted EHV-1 gG has also been shown to bind a broad range of chemokines with high affinity and in a species-independent manner (14). The potential role of EHV-1 gG in chemotaxis and cell trafficking has since then been extensively studied, both in vitro and in vivo. In line with what has been described for other alphaherpesviruses, gG of EHV-1 was found to be dispensable for virus replication in cultured cells (62). In the murine BALB/c model of EHV-1 infection, no significant differences in virulence between a gG deletion mutant and its revertant virus were detected when high doses of infectious virus were used. A clear phenotype was observed, however, when the gG deletion mutant was applied to mice at lower doses of infection. Intriguingly, at these lower doses of infection (1 × 10³ to 1 × 10⁴ infectious units/animal), the gG deletion mutant induced more-severe clinical signs and a more pronounced inflammatory response in the lungs of infected mice than wild-type or revertant viruses (62).

The vCKBP activity of gG was also studied in more detail using chemotaxis assays in vitro. First, it was demonstrated that baculovirus-expressed full-length EHV-1 gG was capable of inhibiting CXCL8-induced chemotaxis of human neutrophils (14). In a following study, this observation was extended to equine cells and equine chemokines, and it was shown that secreted EHV-1 gG (both from the supernatant and baculovirus-expressed) was capable of interfering with the chemotaxis of equine neutrophils induced by equine CXCL8 (59). In contrast, gG was unable to interfere with the CCL2-induced chemotaxis of equine monocytes (59). Other studies demonstrated a functional interference of EHV-1 gG with the chemotaxis of murine neutrophils and macrophages induced by the CXCL-8 relative KC and the proinflammatory chemokine CCL3, respectively (59, 60). Moreover, gG was shown to have a significant effect on the migration of immune cells into murine airways in vivo (59, 60). Interestingly, a reinfection experiment in which mice were inoculated with a gG deletion mutant and subsequently challenged with wild-type virus revealed that the presence of gG-specific antibodies not only had a protective effect but was also able to control the vCKBP activity of gG (60). This observation was supported by in vitro data showing that the presence of gG-specific antibodies could restore chemokine-induced chemotaxis (60). This finding seems to suggest that gG-specific antibodies can control gG's vCKBP function and might be important in preventing EHV-1 from evading the immune system. These findings also put into question the use of gG deletion mutants as modified live virus marker vaccines for protection against EHV-1 infections in particular and possibly alphaherpesvirus infections in general.

Whereas EHV-1 gG clearly has vCKBP activities both in vitro and in vivo, no such role was found for its EHV-4 counterpart (14, 59). EHV-4 is a close relative of EHV-1, and their gG amino acid sequences share 72% homology, although approximately 100 amino acids of the ectodomains are highly divergent and harbor type-specific epitopes (21). In general, the structural features of gG important for binding to chemokines remain undetermined to date, but preliminary data with baculovirus-expressed EHV-1/EHV-4 gG chimeric proteins indicate that the binding epitope for chemokine binding is located in the extracellular and hypervariable region of EHV-1 gG (G. R. Van de Walle and N. Osterrieder, unpublished observation). The observation that EHV-1 gG is a vCKBP but that gG of the closely related EHV-4 does not show chemokine binding properties is very interesting, especially when one takes into account the different pathogenetic patterns of these two equine herpesviruses. Infection with EHV-1 can lead to multiorgan clinical signs, whereas EHV-4 infection is predominantly associated with highly localized and mild upper respiratory disease (49, 61). This leads us to hypothesize that the ability of gG to interfere with the chemokine network might contribute to the dissemination and virulence of EHV-1. In turn, the inability of EHV-4 gG to stop or modulate the host's first line of defense may help restrict EHV-4 to the upper airways. However, we cannot formally exclude that EHV-4 gG possesses (restricted) chemokine binding properties, since not all EHV-4 gG-chemokine interactions have been fully explored to date.

CONCLUDING REMARKS

In this review, we have discussed recent developments in the area of immunomodulatory proteins encoded by alphaherpesviruses, specifically those targeting chemokine signaling. To date, MDV expressing vIL-8, the viral counterpart of cellular IL-8, appears to be the only alphaherpesvirus modulating the chemokine network by molecular mimicry of a host protein. This implies that the more recent mammalian alphaherpesviruses use other strategies to manipulate the action of chemokines, as seems to be the case with gG, a vCKBP that not only interferes with a broad range of chemokines but can also intercept chemokine networking on other levels. Still, one cannot exclude the possibility that the mammalian alphaherpesviruses might actually encode viral proteins with similarities to host cytokines or chemokines which are not yet identified, as new host molecules involved in immunity are discovered on a regular basis. This growing knowledge about host genes and the ever more comprehensive annotation of host genomes sequenced in their entirety urges the virologist to constantly follow new developments and discoveries in genomics and immunology, as findings there might give them new insights into genes with possible immune evasion properties encoded by viruses. Every new discovery will not only aid in a better understanding of the viruses' “antiimmune” systems but will also aid in unraveling the complexity of the host immune systems with which viruses have established close relationships.