G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS Molecular Immunology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Immunology of bovine respiratory syncytial virus in calves夽 Efrain Guzman, Geraldine Taylor ∗ The Pirbright Institute, Ash Road, Pirbright, Woking, Surrey GU24 0NF, UK
a r t i c l e
i n f o
Article history: Received 6 October 2014 Received in revised form 28 November 2014 Accepted 7 December 2014 Available online xxx Keywords: Respiratory syncytial virus Calves Respiratory disease Vaccines Immunity Pathogenesis
a b s t r a c t Bovine respiratory syncytial virus (BRSV) is an important cause of respiratory disease in young calves. The virus is genetically and antigenically closely related to human (H)RSV, which is a major cause of respiratory disease in young infants. As a natural pathogen of calves, BRSV infection recapitulates the pathogenesis of respiratory disease in man more faithfully than semi-permissive, animal models of HRSV infection. With the increasing availability of immunological reagents, the calf can be used to dissect the pathogenesis of and mechanisms of immunity to RSV infection, to analyse the ways in which the virus proteins interact with components of the innate response, and to evaluate RSV vaccine strategies. Passively transferred, neutralising bovine monoclonal antibodies, which recognise the same epitopes in the HRSV and BRSV fusion (F) protein, can protect calves against BRSV infection, and depletion of different T cells subsets in calves has highlighted the importance of CD8+ T cells in viral clearance. Calves can be used to model maternal-antibody mediated suppression of RSV vaccine efficacy, and to increase understanding of the mechanisms responsible for RSV vaccine-enhanced respiratory disease. © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Bovine respiratory syncytial virus (BRSV) is a major cause of respiratory disease in young calves, and is responsible for significant economic losses to the farming industry throughout the world (Valarcher and Taylor, 2007). BRSV is genetically and antigenically closely related to human (H)RSV, which is the single most important cause of lower respiratory tract disease in young infants. BRSV and HRSV are enveloped, non-segmented, negativestranded RNA viruses belonging to the genus Pneumovirus, within the family Paramyxoviridae. Of the 11 proteins encoded by the viral genome, three transmembrane glycoproteins, the large attachment glycoprotein (G), the fusion protein (F) and the small hydrophobic protein (SH), are located in the viral envelope. The matrix protein (M) is present on the inner face of the envelope, and the nucleoprotein (N), phosphoprotein (P), RNA polymerase (L) and M2-1 constitute the nucleocapsid. The M2-2 protein is expressed at low
Abbreviations: BRSV, bovine respiratory syncytial virus; HRSV, human RSV; BAL, bronchoalveolar lavage; rVV, recombinant vaccinia virus; rBHV-1, recombinant bovine herpesvirus -1; ALC, afferent lymph dendritic cells; APC, antigen-presenting cells; moDC, monocyte-derived dendritic cells; mAbs, monoclonal antibodies; MDA, maternally-derived antibodies; FI-HRSV, formalin-inactivated human respiratory syncytial virus. 夽 This article belongs to Special Issue on Non-rodent Animal Models. ∗ Corresponding author. Tel.: +44 0 1482231402. E-mail address: [email protected] (G. Taylor).
levels in infected cells; however, it is not clear if it is incorporated into virions. The two non-structural proteins, NS1 and NS2, are present in high levels in infected cells. Although BRSV and HRSV have a highly restricted host range, the pathogenesis and epidemiology of infection by these viruses have many features in common. In temperate climates, both viruses cause annual winter outbreaks of respiratory disease, with a peak incidence of severe disease in naïve infants and calves 1–6 months of age. The majority of calves and children become infected by 1–2 years of age (Stott and Taylor, 1985). BRSV and HRSV replicate primarily in ciliated airway epithelia cells and type II pneumocytes (Johnson et al., 2007; Viuff et al., 1996; Welliver et al., 2008), and induce a wide range of proinflammatory cytokines and chemokines (Bermejo-Martin et al., 2008; Rosenberg and Domachowske, 2012; Valarcher and Taylor, 2007), which direct the expression of cellular adhesion molecules and recruit neutrophils and lymphocytes to the lung, resulting in bronchiolitis and interstitial pneumonia. While no single animal model system can fully reflect the disease in another species, as a natural host of BRSV infection, the calf provides significant advantages over semi-permissive animal models for the study of the immunology and pathogenesis of RSV infection. Speciesspecific differences in host molecules involved in virus replication and differences in the way the virus interacts with components of the innate immune response may be responsible for the reduced efficiency of virus replication seen in non-native hosts and may influence the outcome of RSV infection. BRSV infection of calves therefore provides an ideal model to study the pathogenesis of and
http://dx.doi.org/10.1016/j.molimm.2014.12.004 0161-5890/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS
2
E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
mechanisms of immunity to RSV. The large size of calves allows for frequent collection of blood and mucosal secretions from individual animals to analyse kinetic changes in the immune response, and to determine the role of cells present in low frequency in the development of immune responses. In addition, cells involved in mucosal immune responses can be analysed by taking serial biopsies from mucosal tissues by endoscopy or by surgical cannulation of lymphatic vessels that drain the oro-nasopharynx and lungs.
2. Innate immune responses to BRSV BRSV infection induces a well-defined pathology which begins with a fever, cough, and often a mucoid nasal discharge. The fever, which can increase to as high as 40 ◦ C, is accompanied by depression, increased respiratory rate and anorexia. Auscultation of the lungs in the most severe cases will show the presence of wheeze. Overproduction of mucus is particularly detrimental as the airways can quickly become obstructed. Whilst an appropriate innate immune response can be beneficial, uncontrolled responses lead to disease. RSV initially encounters the respiratory epithelia and elements of the innate immune response are activated resulting in the induction of pro-inflammatory cytokines and chemokines. Increased levels of expression of IL-12, IFN␥, TNF␣, IL-6, IL-18, IL-8, RANTES, MCP-1, MIP-1␣, IFN␣ and IFN mRNA have been detected in pneumonic lesions from BRSV-infected calves (Sacco et al., 2012; Valarcher and Taylor, 2007). Similarly, TNF-␣ and IFN␥ have been detected in bronchoalveolar lavage (BAL) from BRSV-infected calves with clinical signs of respiratory disease (Antonis et al., 2010). In agreement with data observed in human epithelial cells, microarray analyses of gene expression in BRSV-infected bovine turbinate epithelial cells have also shown up-regulation of proinflammatory cytokines (Das et al., 2005; Gershwin, 2012).
2012). However, a clear role for BSRV F in stimulating NK-B has not been defined (Lizundia et al., 2008). TLR7 has also been proposed to mediate responses against RSV. Activation of TLR7 in HRSV-infected mouse macrophages results in increased expression of interleukin (IL)-12 and IL-23, two important T cell differentiation factors (Lukacs et al., 2010). Although ssRNA can activate bovine TLR7 in vitro, activation of TLR7 by BRSV has not yet been investigated (Buza et al., 2008). In HRSV-infected infants, cytokines and chemokines such as TNF␣, IL-1, IL-6, IL-8 MIP-1␣ and RANTES released during RSV infection by both airway epithelial cells and lung macrophage populations (Becker et al., 1991; El-Sahly et al., 2000) have been shown to correlate with disease severity. Bovine dendritic cells (DC) and monocytes infected with BRSV showed increased levels of RANTES, MIP1␣–2␣–3␣, MCP-2 mRNA (Werling et al., 2002b) and increased production of TNF␣ and IL-1 (Taylor et al., 2014), suggesting that the mechanisms involved in cytokine production by RSV infection are conserved in humans and cattle. IL-1, which is produced by BRSV- and HRSV-infected macrophages and epithelial cells (Bermejo-Martin et al., 2008; Fach et al., 2010; Taylor et al., 2014; Werling et al., 2002a), plays an important role in inflammation by orchestrating the proinflammatory response. IL-1 and IL-18, which enhance IL-12 and IFN␥ production and regulate innate and acquired immune responses, are produced as cytosolic precursors which require proteolytic cleavage induced by the inflammasome for activation and secretion. HRSV activates the nucleotide-binding oligomerization domain like receptor 3 (NLR3)/ASC inflammasome. In mouse bone marrow macrophages, this is initiated by TLR-2/Myd88/NFB signalling, and reactive oxygen species (ROS) and K+ ion efflux (Segovia et al., 2012). However, studies in human lung epithelial cells indicated that activation of the NLRP3/ASC inflammasome by HRSV was initiated by TLR-4 (Triantafilou et al., 2013). The mechanisms involved in induction of IL-1 in BRSV-infected cells have not yet been analysed.
2.1. Recognition of RSV by Pattern Recognition Receptors Studies in small animal models of RSV infection have indicated that RSV activates a number of Toll-like receptors (TLRs) on hematopoietic cells. However, since TLRs are species-specific, data obtained in mice does not necessarily relate to other species, including man and cattle. TLR2 is involved in the recognition of a wide array of microbial molecules from bacteria, mycoplasma and yeast. HRSV interacts with mouse TLR2 and TLR6 promoting neutrophil migration and dendritic cell (DC) activation within the mouse lung (Murawski et al., 2009). However, the extracellular domain of TLR2, which is responsible for ligand recognition, is significantly different in mice, man and cattle (Willcocks et al., 2013), suggesting that the effects observed in laboratory animals may not apply to man or cattle. TLR3 recognises intracellular dsRNA and although the production of dsRNA during RSV replication has not been demonstrated directly, the TLR3 pathway is activated in HRSV-infected mouse cells, resulting in chemokine upregulation (Rudd et al., 2005) and increased sensitivity to other TLR3 ligands (Groskreutz et al., 2006). Although the effect of TLR3 activation in bovine cells has not been extensively studied, BRSV induces the expression of MCP-1, MIP-1␣ and IL-10 mRNA in ␥/␦ T cells apparently in response to TLR3 activation (McGill et al., 2013). TLR4 and its co-receptor CD14 have been shown to interact with HRSV in both human and mouse immune cells. TLR4-null mice infected with HRSV showed delayed clearance of HRSV compared to wild type mice (Kurt-Jones et al., 2000) and it was later shown that the RSV F protein activates the TLR4/CD14 complex resulting in NF-B-mediated inflammation in both mouse and human cells (Awomoyi et al., 2007; Haeberle et al., 2002; Marr and Turvey,
2.2. Type I interferons Type I IFNs are specialised cytokines that are released by host cells in response to pathogens. Various pattern recognition receptors (PRR) such as RIG-I receptors, TLR, NOD-like receptors, detect RSV infection and induce the production of type I IFN which in turn has antiviral effects on neighbouring uninfected cells. However, RSV is more resistant to the anti-viral effects of type I IFNs than other paramyxoviruses (Atreya and Kulkarni, 1999; Schlender et al., 2000) and NS1 and NS2 of both human and bovine RSV have been shown to suppress induction of type I IFN induction in vitro (Bossert et al., 2003; Schlender et al., 2000). HRSV and BRSV with single or double deletions of the NS genes (NS1, NS2 and NS1NS2) have a reduced ability to replicate in cell culture and in animals (Bossert et al., 2003; Jin et al., 2000; Valarcher et al., 2003). NS1 and NS2 antagonize both the cellular antiviral response as well as the induction of IFN. The NS proteins appear to function independently and in combination to target a number of different factors involved in the IFN induction and response pathways (Barik, 2013). Although both HRSV and BRSV NS proteins inhibit activation of IRF3 (Bossert et al., 2003; Spann et al., 2005), the mechanisms by which these viruses inhibit IFN may differ. Thus, HRSV lacking NS1 (HRSVNS1) is more effective at inhibiting induction of IFN than HRSVNS2 (Spann et al., 2004), which is the converse of the relative contributions of the individual BRSV NS proteins to inhibition of IFN induction (Valarcher et al., 2003).
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
2.3. NK cells Natural killer (NK) cells are key components of the innate immune system, with cytotoxic and cytokine producing abilities. Bovine NK cells have been characterised as NKp46+ /CD3− lymphocytes. Like primate NK cells, cattle NK cells have an expanded family of KIR genes and a single Ly49, and recent studies have indicated a role for NK cells in mycobacterial and protozoan infections (Boysen and Storset, 2009). Studies in mice infected with HRSV have indicated an important role for NK cells in controlling RSV replication and in the development of lung pathology (Li et al., 2012). However, the role of NK cells in BRSV infection in calves has not been studied.
3. Immunomodulation by RSV Several RSV-encoded proteins, such as NS1, NS2, G and SH, modulate aspects of the innate and adaptive immune response and are not essential for virus replication in vitro. For example, the NS proteins suppress the induction of type I IFN and counteract the anti-viral effects of type I IFN (Lo et al., 2005; Schlender et al., 2000; Valarcher et al., 2003). In addition to their anti-viral effects, type I IFNs also play a major role in orchestrating the adaptive immune response to virus infection, either indirectly via upregulation of cytokines and chemokines or through direct action on NK cells, DCs and lymphocytes (Durbin et al., 2013). BRSV lacking either the NS1 or NS2 gene appear to be similarly attenuated in calves (Valarcher et al., 2003). However, the greater immunogenicity of recombinant BRSV lacking NS2 (rBRSVNS2) compared with that of rBRSVNS1, appeared to be associated with the greater ability of rBRSVNS2 to induce type I IFN. Similarly, the magnitude of the HRSV-specific pulmonary CTL response in mice infected with HRSV lacking the NS2 gene was significantly greater than that of mice infected with virus lacking the NS1 gene or even wild-type HRSV (Kotelkin et al., 2006). Although there is only 30% amino acid identity between the BRSV and HRSV G proteins, they share similar features in that the G protein of both viruses is highly glycosylated and is produced as a membrane-anchored and as a secreted protein. The secreted form of the HRSV G protein can act as a decoy and has been shown to inhibit antibody-mediated neutralisation of HRSV (Bukreyev et al., 2008). The G protein ectodomain of both HRSV and BRSV has a central conserved cysteine noose region flanked by two more variable domains which are heavily glycosylated with N-linked and O-linked sugars. The extensive O-linked glycosylation is characteristic of cellular mucins, and may act to reduce the immunogenicity of the G protein. The cysteine-rich region contains a CXC3 motif, which has been reported to mimic the CXC3 chemokine, fractalkine, and suppresses the influx of immune cells into the lungs of RSVinfected mice (Harcourt et al., 2006; Tripp et al., 2001). In addition, the HRSV G protein can interact with DC-SIGN on human DCs and suppress their activation (Johnson et al., 2012), and can inhibit TLR-mediated production of IL-6 and IL-1 by human monocytes and mouse macrophages (Polack et al., 2005). As seen following infection of human monocytes with HRSV lacking the G protein (Polack et al., 2005), infection of bovine monocytes with BRSV lacking the G protein induced the secretion of higher levels of IL-1 than cells infected with wild-type virus (Taylor et al., unpublished observations). The immunomodulatory properties of the G protein have been highlighted by the observation that expression of HRSV G protein by recombinant vaccinia virus (rVV) increased the virulence of the rVV in the lungs of mice (Taylor et al., 1991). Similarly, expression of the BRSV G protein by recombinant bovine herpes virus-1 (rBHV-1) increased the virulence of the rBHV-1 in the lower, but not the upper, respiratory tract of calves (Taylor et al., 1998). The ability of the G protein to inhibit components of the
3
innate and adaptive immune response would be expected to limit the induction of RSV-specific immunity following RSV infection. The SH protein also appears to play a role in inhibition of the innate immune response. Thus, BRSV lacking the SH protein induces higher levels of apoptosis, and IL-1 and TNF␣ secretion than wildtype virus (Taylor et al., 2014). Similarly, the HRSV SH protein has been shown to inhibit apoptosis and TNF␣ signalling (Fuentes et al., 2007; Lin et al., 2003). Since apoptosis enhances antigen presentation (Li et al., 2013), and IL-1 plays an important role in orchestrating the adaptive immune response (Ichinohe et al., 2009; Schmitz et al., 2005), inhibition of apoptosis and IL-1 secretion by the SH protein may also act to limit induction of RSV-specific immunity. The greater level of sequence variation in the NS1, G and SH proteins between HRSV and BRSV, compared with the other viral proteins (Valarcher and Taylor, 2007), may be related to the adaptation of the viruses to their respective hosts. This suggestion is supported by the observation that the NS proteins display a differential ability to counteract type I IFN in cells from different hosts (Bossert and Conzelmann, 2002). Thus, replication of a chimeric BRSV containing HRSV NS genes was severely attenuated in bovine IFN competent cells, whereas it replicated like BRSV in IFN-incompetent Vero cells or in IFN-competent human HEp-2 cells. The reduced ability of RSV to counteract anti-viral responses in non-native hosts is illustrated by the observation that replication of HRSV in mouse embryo cells could be overcome by adding anti-mouse IFN serum to the medium (Hanada et al., 1986). Similarly, replication of HRSV in the lungs of mice, which is typically restricted to type I alveolar pneumocytes, can occur in bronchiolar epithelium in mice lacking both MAVS and MyD88, which play an essential role in induction of type I IFN and other proinflammatory cytokines (Bhoj et al., 2008). These observations highlight the importance of studying RSV infections in a natural host.
4. Interaction of RSV with dendritic cells Antigen uptake, processing and presentation by professional antigen presenting cells (APC) are required for the initiation, maintenance and recall of immune responses. Monocytes, macrophages and DCs are the most commonly studied APC in all species. DC are the most potent APC and various models are used to investigate the interaction of pathogens with DC. Large numbers of monocytes, obtained from peripheral blood or spleens of small animals, can be cultured, in vitro, in the presence of cytokines such as granulocytemacrophage colony stimulating factor (GM-CSF) and interleukin-4 (IL-4) to produce monocyte-derived DC (moDC). These cells are phenotypically similar in many ways to tissue-resident DC, but functionally appear to behave differently. Surgical cannulation of lymphatic vessels can provide cells, including DC, that migrate from peripheral tissues to the local lymph nodes. Surgical cannulation of lymphatic vessels was originally described in rabbits (Ehrich, 1942) and was later adapted for sheep (Hall and Morris, 1962) and cattle (Cronkite et al., 1964). The thoracic duct has been the most commonly cannulated vessel, particularly in small animals, since this is relatively easy to identify and cannulate (Ionac, 2003). Cells obtained from the lymph have the benefit of being terminally differentiated and thus are not subjected to culture or maturation in vitro. Afferent lymph dendritic cells (ALDC) that drain the skin have been used to investigate the interaction of bovine DC with pathogens and vaccines (Cubillos-Zapata et al., 2011; Guzman et al., 2012; Hope et al., 2012), and vector-borne viruses (Hemati et al., 2009). Because of the relatively small size of the lymphatic vessels that drain the skin, this technique can only be performed in large animals such as sheep, cattle and pigs (Bertho et al., 2011; Bujdoso et al., 1989; Hope et al., 2006). Bovine ALDC can be subdivided into two main
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS
4
E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
subpopulations, those expressing high levels of CD172a (SIRP␣) and those expressing low levels or no CD172a. Although both of these ALDC subsets can take up BRSV, the CD172ahigh population is more efficient in presenting BRSV antigen to resting, memory CD4+ T cells than the CD172alow/neg population (Howard et al., 1997). For the study of immune responses in the respiratory tract and the interaction of respiratory viruses with immune cells that drain the upper and lower respiratory tract, surgical cannulation of the lymphatic vessels that drain the oro-nasal area (Schwartz-Cornil et al., 2006) and the lungs (Yen et al., 2009) provide access to invaluable biological material, including numerous APC. These models can be used to assess immune responses to infections originating in the respiratory mucosae or for optimizing intra-lung delivery of drugs or vaccines (Scheerlinck et al., 2006; Yen et al., 2006), and have been used to analyse the immune response of calves to BRSV (see Section 7.2). Although bovine moDC do not support BRSV replication (Werling et al., 2002a), a high proportion of moDC appear to undergo apoptosis following exposure to BRSV and the remaining viable cells express higher levels of IL-2, IL-12 and IL-15 mRNA than cells cultured with heat-inactivated BRSV (Werling et al., 2002a,b). Ovine neonatal pulmonary DC appear to support virus replication, at least in vitro (Fach et al., 2007), and replication of HRSV has been observed in various human DC (Johnson et al., 2011, 2006). Human moDC can be productively infected with HRSV, resulting in their maturation as shown by up-regulation of CD80, CD83, CD86, and HLA class II molecules. However, HRSV infection of moDC results in impaired CD4+ T cell activation characterised by a reduction in cell proliferation and ablation of cytokine production in activated T cells (de Graaff et al., 2005). The ability to isolate different populations of DC from cattle provides the opportunity to study the interaction of BRSV with DC in more detail.
5. Neonatal models of RSV infection Since RSV is a disease of young infants, a number of animal models have been developed to study pathogenesis, vaccine efficacy and immune responses to RSV in neonates. Studies on the pathogenesis of RSV infection in BALB/c mice have shown that there is less pulmonary inflammation and disease in neonatal (
ARTICLE IN PRESS Molecular Immunology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Immunology of bovine respiratory syncytial virus in calves夽 Efrain Guzman, Geraldine Taylor ∗ The Pirbright Institute, Ash Road, Pirbright, Woking, Surrey GU24 0NF, UK
a r t i c l e
i n f o
Article history: Received 6 October 2014 Received in revised form 28 November 2014 Accepted 7 December 2014 Available online xxx Keywords: Respiratory syncytial virus Calves Respiratory disease Vaccines Immunity Pathogenesis
a b s t r a c t Bovine respiratory syncytial virus (BRSV) is an important cause of respiratory disease in young calves. The virus is genetically and antigenically closely related to human (H)RSV, which is a major cause of respiratory disease in young infants. As a natural pathogen of calves, BRSV infection recapitulates the pathogenesis of respiratory disease in man more faithfully than semi-permissive, animal models of HRSV infection. With the increasing availability of immunological reagents, the calf can be used to dissect the pathogenesis of and mechanisms of immunity to RSV infection, to analyse the ways in which the virus proteins interact with components of the innate response, and to evaluate RSV vaccine strategies. Passively transferred, neutralising bovine monoclonal antibodies, which recognise the same epitopes in the HRSV and BRSV fusion (F) protein, can protect calves against BRSV infection, and depletion of different T cells subsets in calves has highlighted the importance of CD8+ T cells in viral clearance. Calves can be used to model maternal-antibody mediated suppression of RSV vaccine efficacy, and to increase understanding of the mechanisms responsible for RSV vaccine-enhanced respiratory disease. © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Bovine respiratory syncytial virus (BRSV) is a major cause of respiratory disease in young calves, and is responsible for significant economic losses to the farming industry throughout the world (Valarcher and Taylor, 2007). BRSV is genetically and antigenically closely related to human (H)RSV, which is the single most important cause of lower respiratory tract disease in young infants. BRSV and HRSV are enveloped, non-segmented, negativestranded RNA viruses belonging to the genus Pneumovirus, within the family Paramyxoviridae. Of the 11 proteins encoded by the viral genome, three transmembrane glycoproteins, the large attachment glycoprotein (G), the fusion protein (F) and the small hydrophobic protein (SH), are located in the viral envelope. The matrix protein (M) is present on the inner face of the envelope, and the nucleoprotein (N), phosphoprotein (P), RNA polymerase (L) and M2-1 constitute the nucleocapsid. The M2-2 protein is expressed at low
Abbreviations: BRSV, bovine respiratory syncytial virus; HRSV, human RSV; BAL, bronchoalveolar lavage; rVV, recombinant vaccinia virus; rBHV-1, recombinant bovine herpesvirus -1; ALC, afferent lymph dendritic cells; APC, antigen-presenting cells; moDC, monocyte-derived dendritic cells; mAbs, monoclonal antibodies; MDA, maternally-derived antibodies; FI-HRSV, formalin-inactivated human respiratory syncytial virus. 夽 This article belongs to Special Issue on Non-rodent Animal Models. ∗ Corresponding author. Tel.: +44 0 1482231402. E-mail address: [email protected] (G. Taylor).
levels in infected cells; however, it is not clear if it is incorporated into virions. The two non-structural proteins, NS1 and NS2, are present in high levels in infected cells. Although BRSV and HRSV have a highly restricted host range, the pathogenesis and epidemiology of infection by these viruses have many features in common. In temperate climates, both viruses cause annual winter outbreaks of respiratory disease, with a peak incidence of severe disease in naïve infants and calves 1–6 months of age. The majority of calves and children become infected by 1–2 years of age (Stott and Taylor, 1985). BRSV and HRSV replicate primarily in ciliated airway epithelia cells and type II pneumocytes (Johnson et al., 2007; Viuff et al., 1996; Welliver et al., 2008), and induce a wide range of proinflammatory cytokines and chemokines (Bermejo-Martin et al., 2008; Rosenberg and Domachowske, 2012; Valarcher and Taylor, 2007), which direct the expression of cellular adhesion molecules and recruit neutrophils and lymphocytes to the lung, resulting in bronchiolitis and interstitial pneumonia. While no single animal model system can fully reflect the disease in another species, as a natural host of BRSV infection, the calf provides significant advantages over semi-permissive animal models for the study of the immunology and pathogenesis of RSV infection. Speciesspecific differences in host molecules involved in virus replication and differences in the way the virus interacts with components of the innate immune response may be responsible for the reduced efficiency of virus replication seen in non-native hosts and may influence the outcome of RSV infection. BRSV infection of calves therefore provides an ideal model to study the pathogenesis of and
http://dx.doi.org/10.1016/j.molimm.2014.12.004 0161-5890/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS
2
E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
mechanisms of immunity to RSV. The large size of calves allows for frequent collection of blood and mucosal secretions from individual animals to analyse kinetic changes in the immune response, and to determine the role of cells present in low frequency in the development of immune responses. In addition, cells involved in mucosal immune responses can be analysed by taking serial biopsies from mucosal tissues by endoscopy or by surgical cannulation of lymphatic vessels that drain the oro-nasopharynx and lungs.
2. Innate immune responses to BRSV BRSV infection induces a well-defined pathology which begins with a fever, cough, and often a mucoid nasal discharge. The fever, which can increase to as high as 40 ◦ C, is accompanied by depression, increased respiratory rate and anorexia. Auscultation of the lungs in the most severe cases will show the presence of wheeze. Overproduction of mucus is particularly detrimental as the airways can quickly become obstructed. Whilst an appropriate innate immune response can be beneficial, uncontrolled responses lead to disease. RSV initially encounters the respiratory epithelia and elements of the innate immune response are activated resulting in the induction of pro-inflammatory cytokines and chemokines. Increased levels of expression of IL-12, IFN␥, TNF␣, IL-6, IL-18, IL-8, RANTES, MCP-1, MIP-1␣, IFN␣ and IFN mRNA have been detected in pneumonic lesions from BRSV-infected calves (Sacco et al., 2012; Valarcher and Taylor, 2007). Similarly, TNF-␣ and IFN␥ have been detected in bronchoalveolar lavage (BAL) from BRSV-infected calves with clinical signs of respiratory disease (Antonis et al., 2010). In agreement with data observed in human epithelial cells, microarray analyses of gene expression in BRSV-infected bovine turbinate epithelial cells have also shown up-regulation of proinflammatory cytokines (Das et al., 2005; Gershwin, 2012).
2012). However, a clear role for BSRV F in stimulating NK-B has not been defined (Lizundia et al., 2008). TLR7 has also been proposed to mediate responses against RSV. Activation of TLR7 in HRSV-infected mouse macrophages results in increased expression of interleukin (IL)-12 and IL-23, two important T cell differentiation factors (Lukacs et al., 2010). Although ssRNA can activate bovine TLR7 in vitro, activation of TLR7 by BRSV has not yet been investigated (Buza et al., 2008). In HRSV-infected infants, cytokines and chemokines such as TNF␣, IL-1, IL-6, IL-8 MIP-1␣ and RANTES released during RSV infection by both airway epithelial cells and lung macrophage populations (Becker et al., 1991; El-Sahly et al., 2000) have been shown to correlate with disease severity. Bovine dendritic cells (DC) and monocytes infected with BRSV showed increased levels of RANTES, MIP1␣–2␣–3␣, MCP-2 mRNA (Werling et al., 2002b) and increased production of TNF␣ and IL-1 (Taylor et al., 2014), suggesting that the mechanisms involved in cytokine production by RSV infection are conserved in humans and cattle. IL-1, which is produced by BRSV- and HRSV-infected macrophages and epithelial cells (Bermejo-Martin et al., 2008; Fach et al., 2010; Taylor et al., 2014; Werling et al., 2002a), plays an important role in inflammation by orchestrating the proinflammatory response. IL-1 and IL-18, which enhance IL-12 and IFN␥ production and regulate innate and acquired immune responses, are produced as cytosolic precursors which require proteolytic cleavage induced by the inflammasome for activation and secretion. HRSV activates the nucleotide-binding oligomerization domain like receptor 3 (NLR3)/ASC inflammasome. In mouse bone marrow macrophages, this is initiated by TLR-2/Myd88/NFB signalling, and reactive oxygen species (ROS) and K+ ion efflux (Segovia et al., 2012). However, studies in human lung epithelial cells indicated that activation of the NLRP3/ASC inflammasome by HRSV was initiated by TLR-4 (Triantafilou et al., 2013). The mechanisms involved in induction of IL-1 in BRSV-infected cells have not yet been analysed.
2.1. Recognition of RSV by Pattern Recognition Receptors Studies in small animal models of RSV infection have indicated that RSV activates a number of Toll-like receptors (TLRs) on hematopoietic cells. However, since TLRs are species-specific, data obtained in mice does not necessarily relate to other species, including man and cattle. TLR2 is involved in the recognition of a wide array of microbial molecules from bacteria, mycoplasma and yeast. HRSV interacts with mouse TLR2 and TLR6 promoting neutrophil migration and dendritic cell (DC) activation within the mouse lung (Murawski et al., 2009). However, the extracellular domain of TLR2, which is responsible for ligand recognition, is significantly different in mice, man and cattle (Willcocks et al., 2013), suggesting that the effects observed in laboratory animals may not apply to man or cattle. TLR3 recognises intracellular dsRNA and although the production of dsRNA during RSV replication has not been demonstrated directly, the TLR3 pathway is activated in HRSV-infected mouse cells, resulting in chemokine upregulation (Rudd et al., 2005) and increased sensitivity to other TLR3 ligands (Groskreutz et al., 2006). Although the effect of TLR3 activation in bovine cells has not been extensively studied, BRSV induces the expression of MCP-1, MIP-1␣ and IL-10 mRNA in ␥/␦ T cells apparently in response to TLR3 activation (McGill et al., 2013). TLR4 and its co-receptor CD14 have been shown to interact with HRSV in both human and mouse immune cells. TLR4-null mice infected with HRSV showed delayed clearance of HRSV compared to wild type mice (Kurt-Jones et al., 2000) and it was later shown that the RSV F protein activates the TLR4/CD14 complex resulting in NF-B-mediated inflammation in both mouse and human cells (Awomoyi et al., 2007; Haeberle et al., 2002; Marr and Turvey,
2.2. Type I interferons Type I IFNs are specialised cytokines that are released by host cells in response to pathogens. Various pattern recognition receptors (PRR) such as RIG-I receptors, TLR, NOD-like receptors, detect RSV infection and induce the production of type I IFN which in turn has antiviral effects on neighbouring uninfected cells. However, RSV is more resistant to the anti-viral effects of type I IFNs than other paramyxoviruses (Atreya and Kulkarni, 1999; Schlender et al., 2000) and NS1 and NS2 of both human and bovine RSV have been shown to suppress induction of type I IFN induction in vitro (Bossert et al., 2003; Schlender et al., 2000). HRSV and BRSV with single or double deletions of the NS genes (NS1, NS2 and NS1NS2) have a reduced ability to replicate in cell culture and in animals (Bossert et al., 2003; Jin et al., 2000; Valarcher et al., 2003). NS1 and NS2 antagonize both the cellular antiviral response as well as the induction of IFN. The NS proteins appear to function independently and in combination to target a number of different factors involved in the IFN induction and response pathways (Barik, 2013). Although both HRSV and BRSV NS proteins inhibit activation of IRF3 (Bossert et al., 2003; Spann et al., 2005), the mechanisms by which these viruses inhibit IFN may differ. Thus, HRSV lacking NS1 (HRSVNS1) is more effective at inhibiting induction of IFN than HRSVNS2 (Spann et al., 2004), which is the converse of the relative contributions of the individual BRSV NS proteins to inhibition of IFN induction (Valarcher et al., 2003).
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
2.3. NK cells Natural killer (NK) cells are key components of the innate immune system, with cytotoxic and cytokine producing abilities. Bovine NK cells have been characterised as NKp46+ /CD3− lymphocytes. Like primate NK cells, cattle NK cells have an expanded family of KIR genes and a single Ly49, and recent studies have indicated a role for NK cells in mycobacterial and protozoan infections (Boysen and Storset, 2009). Studies in mice infected with HRSV have indicated an important role for NK cells in controlling RSV replication and in the development of lung pathology (Li et al., 2012). However, the role of NK cells in BRSV infection in calves has not been studied.
3. Immunomodulation by RSV Several RSV-encoded proteins, such as NS1, NS2, G and SH, modulate aspects of the innate and adaptive immune response and are not essential for virus replication in vitro. For example, the NS proteins suppress the induction of type I IFN and counteract the anti-viral effects of type I IFN (Lo et al., 2005; Schlender et al., 2000; Valarcher et al., 2003). In addition to their anti-viral effects, type I IFNs also play a major role in orchestrating the adaptive immune response to virus infection, either indirectly via upregulation of cytokines and chemokines or through direct action on NK cells, DCs and lymphocytes (Durbin et al., 2013). BRSV lacking either the NS1 or NS2 gene appear to be similarly attenuated in calves (Valarcher et al., 2003). However, the greater immunogenicity of recombinant BRSV lacking NS2 (rBRSVNS2) compared with that of rBRSVNS1, appeared to be associated with the greater ability of rBRSVNS2 to induce type I IFN. Similarly, the magnitude of the HRSV-specific pulmonary CTL response in mice infected with HRSV lacking the NS2 gene was significantly greater than that of mice infected with virus lacking the NS1 gene or even wild-type HRSV (Kotelkin et al., 2006). Although there is only 30% amino acid identity between the BRSV and HRSV G proteins, they share similar features in that the G protein of both viruses is highly glycosylated and is produced as a membrane-anchored and as a secreted protein. The secreted form of the HRSV G protein can act as a decoy and has been shown to inhibit antibody-mediated neutralisation of HRSV (Bukreyev et al., 2008). The G protein ectodomain of both HRSV and BRSV has a central conserved cysteine noose region flanked by two more variable domains which are heavily glycosylated with N-linked and O-linked sugars. The extensive O-linked glycosylation is characteristic of cellular mucins, and may act to reduce the immunogenicity of the G protein. The cysteine-rich region contains a CXC3 motif, which has been reported to mimic the CXC3 chemokine, fractalkine, and suppresses the influx of immune cells into the lungs of RSVinfected mice (Harcourt et al., 2006; Tripp et al., 2001). In addition, the HRSV G protein can interact with DC-SIGN on human DCs and suppress their activation (Johnson et al., 2012), and can inhibit TLR-mediated production of IL-6 and IL-1 by human monocytes and mouse macrophages (Polack et al., 2005). As seen following infection of human monocytes with HRSV lacking the G protein (Polack et al., 2005), infection of bovine monocytes with BRSV lacking the G protein induced the secretion of higher levels of IL-1 than cells infected with wild-type virus (Taylor et al., unpublished observations). The immunomodulatory properties of the G protein have been highlighted by the observation that expression of HRSV G protein by recombinant vaccinia virus (rVV) increased the virulence of the rVV in the lungs of mice (Taylor et al., 1991). Similarly, expression of the BRSV G protein by recombinant bovine herpes virus-1 (rBHV-1) increased the virulence of the rBHV-1 in the lower, but not the upper, respiratory tract of calves (Taylor et al., 1998). The ability of the G protein to inhibit components of the
3
innate and adaptive immune response would be expected to limit the induction of RSV-specific immunity following RSV infection. The SH protein also appears to play a role in inhibition of the innate immune response. Thus, BRSV lacking the SH protein induces higher levels of apoptosis, and IL-1 and TNF␣ secretion than wildtype virus (Taylor et al., 2014). Similarly, the HRSV SH protein has been shown to inhibit apoptosis and TNF␣ signalling (Fuentes et al., 2007; Lin et al., 2003). Since apoptosis enhances antigen presentation (Li et al., 2013), and IL-1 plays an important role in orchestrating the adaptive immune response (Ichinohe et al., 2009; Schmitz et al., 2005), inhibition of apoptosis and IL-1 secretion by the SH protein may also act to limit induction of RSV-specific immunity. The greater level of sequence variation in the NS1, G and SH proteins between HRSV and BRSV, compared with the other viral proteins (Valarcher and Taylor, 2007), may be related to the adaptation of the viruses to their respective hosts. This suggestion is supported by the observation that the NS proteins display a differential ability to counteract type I IFN in cells from different hosts (Bossert and Conzelmann, 2002). Thus, replication of a chimeric BRSV containing HRSV NS genes was severely attenuated in bovine IFN competent cells, whereas it replicated like BRSV in IFN-incompetent Vero cells or in IFN-competent human HEp-2 cells. The reduced ability of RSV to counteract anti-viral responses in non-native hosts is illustrated by the observation that replication of HRSV in mouse embryo cells could be overcome by adding anti-mouse IFN serum to the medium (Hanada et al., 1986). Similarly, replication of HRSV in the lungs of mice, which is typically restricted to type I alveolar pneumocytes, can occur in bronchiolar epithelium in mice lacking both MAVS and MyD88, which play an essential role in induction of type I IFN and other proinflammatory cytokines (Bhoj et al., 2008). These observations highlight the importance of studying RSV infections in a natural host.
4. Interaction of RSV with dendritic cells Antigen uptake, processing and presentation by professional antigen presenting cells (APC) are required for the initiation, maintenance and recall of immune responses. Monocytes, macrophages and DCs are the most commonly studied APC in all species. DC are the most potent APC and various models are used to investigate the interaction of pathogens with DC. Large numbers of monocytes, obtained from peripheral blood or spleens of small animals, can be cultured, in vitro, in the presence of cytokines such as granulocytemacrophage colony stimulating factor (GM-CSF) and interleukin-4 (IL-4) to produce monocyte-derived DC (moDC). These cells are phenotypically similar in many ways to tissue-resident DC, but functionally appear to behave differently. Surgical cannulation of lymphatic vessels can provide cells, including DC, that migrate from peripheral tissues to the local lymph nodes. Surgical cannulation of lymphatic vessels was originally described in rabbits (Ehrich, 1942) and was later adapted for sheep (Hall and Morris, 1962) and cattle (Cronkite et al., 1964). The thoracic duct has been the most commonly cannulated vessel, particularly in small animals, since this is relatively easy to identify and cannulate (Ionac, 2003). Cells obtained from the lymph have the benefit of being terminally differentiated and thus are not subjected to culture or maturation in vitro. Afferent lymph dendritic cells (ALDC) that drain the skin have been used to investigate the interaction of bovine DC with pathogens and vaccines (Cubillos-Zapata et al., 2011; Guzman et al., 2012; Hope et al., 2012), and vector-borne viruses (Hemati et al., 2009). Because of the relatively small size of the lymphatic vessels that drain the skin, this technique can only be performed in large animals such as sheep, cattle and pigs (Bertho et al., 2011; Bujdoso et al., 1989; Hope et al., 2006). Bovine ALDC can be subdivided into two main
Please cite this article in press as: Guzman, E., Taylor, G., Immunology of bovine respiratory syncytial virus in calves. Mol. Immunol. (2014), http://dx.doi.org/10.1016/j.molimm.2014.12.004
G Model MIMM-4544; No. of Pages 9
ARTICLE IN PRESS
4
E. Guzman, G. Taylor / Molecular Immunology xxx (2014) xxx–xxx
subpopulations, those expressing high levels of CD172a (SIRP␣) and those expressing low levels or no CD172a. Although both of these ALDC subsets can take up BRSV, the CD172ahigh population is more efficient in presenting BRSV antigen to resting, memory CD4+ T cells than the CD172alow/neg population (Howard et al., 1997). For the study of immune responses in the respiratory tract and the interaction of respiratory viruses with immune cells that drain the upper and lower respiratory tract, surgical cannulation of the lymphatic vessels that drain the oro-nasal area (Schwartz-Cornil et al., 2006) and the lungs (Yen et al., 2009) provide access to invaluable biological material, including numerous APC. These models can be used to assess immune responses to infections originating in the respiratory mucosae or for optimizing intra-lung delivery of drugs or vaccines (Scheerlinck et al., 2006; Yen et al., 2006), and have been used to analyse the immune response of calves to BRSV (see Section 7.2). Although bovine moDC do not support BRSV replication (Werling et al., 2002a), a high proportion of moDC appear to undergo apoptosis following exposure to BRSV and the remaining viable cells express higher levels of IL-2, IL-12 and IL-15 mRNA than cells cultured with heat-inactivated BRSV (Werling et al., 2002a,b). Ovine neonatal pulmonary DC appear to support virus replication, at least in vitro (Fach et al., 2007), and replication of HRSV has been observed in various human DC (Johnson et al., 2011, 2006). Human moDC can be productively infected with HRSV, resulting in their maturation as shown by up-regulation of CD80, CD83, CD86, and HLA class II molecules. However, HRSV infection of moDC results in impaired CD4+ T cell activation characterised by a reduction in cell proliferation and ablation of cytokine production in activated T cells (de Graaff et al., 2005). The ability to isolate different populations of DC from cattle provides the opportunity to study the interaction of BRSV with DC in more detail.
5. Neonatal models of RSV infection Since RSV is a disease of young infants, a number of animal models have been developed to study pathogenesis, vaccine efficacy and immune responses to RSV in neonates. Studies on the pathogenesis of RSV infection in BALB/c mice have shown that there is less pulmonary inflammation and disease in neonatal (
