Identification of secreted ROP kinases and their role in virulence
In Europe and North America three major T. gondii lineages are predominant, types I, II and III (Howe & Sibley, 1995). The three types differ widely in a number of phenotypes in mice such as virulence, persistence, migratory capacity, attraction of different cell types and induction of cytokine expression (Saeij et al., 2005). In immunocompetent individuals, infection with T. gondii tends to be chronic and asymptomatic, as parasites differentiate to the slow growing bradyzoite form and encyst in tissues. Types I and II, however, are associated with congenital birth defects and abortion in first-time mothers who become infected during pregnancy. They are also linked with severe disease in immunocompromised individuals such as people with HIV/AIDS, those undergoing chemotherapy or recipients of organ transplants (Sibley & Ajioka, 2008). Type III parasites are significantly less virulent and infrequently cause disease, even in the immune compromised.
Although most likely coming from a common ancestor in North America, the South American parasite population structure is slightly different (Lehmann et al., 2006; Sibley & Ajioka, 2008). A larger diversity, probably due to increased recombination events, defines this population. In South America, a major problem from T. gondii comes in the form of recurrent ocular infections that appear in otherwise healthy adults (Jones et al., 2006; Khan et al., 2006). A total of 11 haplotypes exist within the T. gondii population (Sibley & Ajioka, 2008), this chapter will focus on the three main strains that are found in Europe and North America, Types I, II and III.
Microarray analysis of T. gondii infected host cells indicate changes in transcription that can be attributed to the parasite (Blader et al., 2001). This information, combined with a wealth of data gleaned from genetic crosses of the three main European T. gondii lineages (Saeij et al., 2006; Taylor et al., 2006; Behnke et al., 2011; Reese et al., 2011) and large scale proteomics analyses (Bradley et al., 2005), has led to the identification of several proteins important for the differences in virulence between parasite types (Table 2). These proteins are members of the ROPK family and they sit in what have been called VIR loci. Briefly, the VIR1 locus codes for a rhoptry pseudokinase (TgROP5) that is present as a tandem array of gene duplications (Saeij et al., 2006; El Hajj et al., 2007a, b), the VIR3 and VIR4 loci code for active, secreted rhoptry kinases (TgROP18 and TgROP16 respectively) and VIR2 and VIR5 have not yet been examined in detail (Saeij et al., 2006; Taylor et al., 2006). Interestingly, TgROP5, TgROP16 and TgROP18 all have a higher than average level of polymorphism across the three main T. gondii strains (Peixoto et al., 2010). Within the ROPK family, some genes are expressed at a higher than average level as identified by transcript abundance and there are a greater number of genes regulated in a strain-specific manner (Peixoto et al., 2010). This information is pertinent when viewed in the context of three strains with significantly different degrees of virulence and gives weight to the theory that ROPKs are important for this. Table S1 shows the conservation of the ROPKs in coccidian species N. caninum and E. tenella.
Table 2. ROPK virulence determinants in T. gondii
| ||Type I||Type II||Type III|
Immune response to intracellular pathogens
Prior to considering some of the key effector molecules shown to be critical for manipulation of the host to promote parasite survival, it is first important to understand the influence of T. gondii mediated changes that occur upon infection of host cells and the host response to intracellular pathogens. This is a very complex situation and micro array analyses of infected host cells highlight a large number of genes that are regulated differentially depending on the infecting T. gondii type and this number also varies based on the host cell type (Saeij et al., 2007; Jensen et al., 2011). However, there are some meaningful patterns and features. Many, but not all, of these gene products are identifiers of classical or alternative macrophage differentiation (Jensen et al., 2011) including members of NFκB signalling pathways and those that lead to the activation of signal transducer and activator of transcription (STAT) factors (Saeij et al., 2007). Fittingly, one function of STATs is the regulation of genes involved in the immune response to intracellular pathogens (Quinton & Mizgerd, 2011).
Invasion into a host cell by T. gondii initially causes an increase in expression of pro-inflammatory cytokines such as IL-12. It is interesting to note that IL-12 levels vary depending on the infecting parasite strain, likely because of the combination of effector molecules produced by particular strains and their downstream effects on the host (Robben et al., 2004) (Table 2). IL-12 is mainly produced from innate immune cells such as macrophages and dendritic cells, and is involved in T helper 1 (Th-1) immunity. Upon host cell infection, IL-12 stimulates activation of STAT4 and results in the differentiation of naïve T cells into Th1 cells, which along with natural killer cells, produce large amounts of IFN-γ (Murphy et al., 2000). IFN-γ in turn activates STAT1, culminating in the expression of a number of effector genes such as p47 GTPases (Leung et al., 1996; Takaoka & Yanai, 2006), which go on to aid in the destruction of the parasite (Taylor et al., 2004). Thus, the STAT4-STAT1 axis is essential for host defences against intracellular pathogens such as T. gondii. Indeed, studies observing the immune response of mice lacking Stat4 or Stat1 found they were highly susceptible to T. gondii infection (Cai et al., 2000; Gavrilescu et al., 2004; Lieberman et al., 2004). The parasite lines used in these studies were virulent types I and II, it would therefore be interesting to know if these mice also became more susceptible to the avirulent type III parasites. Intriguingly, T. gondii infection of human foreskin fibroblasts (HFFs) results in expression of interferon inducible genes (Kim et al., 2007) regardless of the fact that HFFs do not produce IFN-γ. This indicates that the parasite stimulates expression of these genes in an alternative fashion; however, the trigger for this activation remains to be determined. One could suspect that other IFNs such as the type I interferons could trigger a similar response as IFN-γ, however, there is no significant expression of these IFNs in HFFs and this has never been shown. It would be of interest to assess if the expression of interferon inducible genes is also observed upon infection of murine fibroblasts with T. gondii. Murine fibroblasts lack Irf3 and Irf7, which are essential for the induction of type I interferons (Honda et al., 2006), making this a useful system to identify if type I IFNs play a role in this context. It has also been suggested that pro-inflammatory cytokines themselves could be capable of activating IFN-responsive genes (Kim et al., 2007), although this also requires further investigation.
As discussed above, invasion of T. gondii causes a pro-inflammatory response. In a normal immune setting, an anti-inflammatory reaction will also be stimulated to temper the pro-inflammatory reaction and prevent it from causing damage to the host. STAT3 is activated in innate immune cells through the action of the anti-inflammatory cytokine IL-10. In T. gondii-infected mice, IL-10 is secreted from immune cells (Murray, 2006), predominantly conventional CD4+ T cells (Jankovic et al., 2007, 2010). Meanwhile, Th2 cytokines IL-4 and IL-13 induce activation of STAT6 and eventually compete or even down-regulate the Th1 pro-inflammatory response. Together, STATs are regulated by various cytokines and can function in opposite ways to produce a balanced anti-T. gondii immune response.
STATs are fully activated through phosphorylation of a conserved tyrosine residue. Regulation of the STAT pathways occurs via a negative feedback mechanism composed of negative regulators, such as suppressors of cytokine signalling (SOCS) (Butcher et al., 2011; Tamiya et al., 2011). SOCS1 inhibits the kinase activity of JAK1 and JAK2, both of which are tyrosine kinases upstream of STAT1 and STAT2. This leads to the abrogation of STAT1 phosphorylation (Kobayashi & Yoshimura, 2005) and therefore abolition of antiparasite effector molecule stimulation.
To counteract the response against it, T. gondii induces SOCS1, which consequently suppresses the IFN-γ/STAT1-mediated cellular antiparasite response (Zimmermann et al., 2006). Another negative regulator of STATs, SOCS3, inhibits STAT3 phosphorylation by interfering with the association between JAKs and cytokine receptors (Kubo et al., 2003). SOCS3 is required for inhibition of IL-6/STAT3-mediated anti-inflammatory programs in innate immune cells. Mice lacking SOCS3 specifically in myeloid cells show reduced IL-12 production, and so reduced pro-inflammatory response, to T. gondii infection (Whitmarsh et al., 2011). This results in a failure to control acute toxoplasmosis. Thus, SOCS indirectly block phosphorylation of STATs, inhibiting their activation. For recent reviews on the innate immune response to intracellular pathogens and STAT signalling refer to (Sehgal, 2008; Rasmussen et al., 2009; Barber, 2011; Peng et al., 2011).
TgROP5: a multi-copy gene coding for pseudokinases
TgROP5 is a pseudokinase that is delivered to the host cytosol face of the PVM (Fig. 4a) (Hanks & Hunter, 1995; El Hajj et al., 2006). Differential permeabilisation techniques and molecular modelling of TgROP5 predicted that a proposed transmembrane helix was instead buried within the protein (El Hajj et al., 2007a, b). This set a precedent for reinterpreting the structure of the other ROPs that had been thought to carry a transmembrane spanning domain (El Hajj et al., 2007a, b). It also raised questions about their actual mechanism of membrane association.
TgROP5 was identified in genetic crosses of types I-III as being highly significant for virulence and is located within the VIR1 locus (Saeij et al., 2006; Behnke et al., 2011). It is remarkable that although TgROP5 codes for a predicted pseudokinase, it is still a major virulence determinant for T. gondii. Recently, evidence has been presented to support the role of pseudokinases in regulating cellular functions. Pseudokinases have been implicated in activation of other kinases through receptor activity, complex formation and providing a molecular scaffold to support other active agents (Boudeau et al., 2006).
TgROP5 exists as a tandem cluster of nearly identical genes, rather than a single gene, and a different number of copies are found in types I-III (Behnke et al., 2011; Reese et al., 2011). Type I has ∼6 copies, type II has ∼10 and type III has ∼4 TgROP5 genes (Reese et al., 2011). There are three major isoforms (A, B and C), with another two presenting only minor SNPs between their closest relatives (Reese et al., 2011). Each isoform has all the residues of a canonical kinase except for the aspartate residue of the His-Arg-Asp (HRD) domain (Behnke et al., 2011; Reese & Boothroyd, 2011). This domain helps to stabilise the interaction with the substrate (Kornev et al., 2006). In TgROP5, the aspartate residue has been replaced with either a histidine or an arginine residue (Behnke et al., 2011), both of these residues are basic in nature. These pseudokinases also have a glycine residue in place of the arginine and so the group of pseudokinases to which TgROP5 belongs are known as His-Gly-Basic (HGB) pseudokinases (Reese & Boothroyd, 2011). The majority of differences between TgROP5 isoforms are found in the C-terminal ATP binding pocket and substrate recognition domains (Behnke et al., 2011; Reese et al., 2011). There is evidence to suggest that the pseudokinase domain is undergoing a diversifying selection, which may be related to function, whereas the N-terminal region, associated with targeting to the PVM, remains unchanged (Reese et al., 2011). It is also interesting to note that none of the SNPs present in any of the isoforms occur in a region that would potentially restore catalytic activity (Reese et al., 2011). Type I TgROP5 isoforms (ROP5AI, ROP5BI and ROP5CI) are almost identical to the type III isoforms (ROP5AIII, ROP5BIII and ROP5CIII respectively), whereas type II isoforms (ROP5AII, ROP5BII and ROP5CII) are significantly different. At least one of the type II alleles presents a frame shift, which leads to a truncated and most likely nonfunctional protein (Behnke et al., 2011; Reese et al., 2011). Interestingly, type I and type III TgROP5 isoforms are more divergent from one another than the type II alleles, and the increased number of copies within the type II cluster may act as a compensation mechanism for the loss of other functional copies (Reese et al., 2011).
Expression of a cosmid containing the TgROP5 locus in a hypovirulent parasite strain (S22) causes a >105 fold increase in virulence in comparison to the parental strain (Reese et al., 2011). The S22 strain is a progeny line from a type II/type III parasite cross that has nonvirulent TgROP5, ROP16 and ROP18 alleles, making it useful for gain of virulence studies (Saeij et al., 2006). Deletion of the entire TgROP5 locus from a type I strain, although presenting no growth phenotype, causes a drop in virulence when compared with the parental strain (Behnke et al., 2011; Reese et al., 2011). The ΔROP5 phenotype is complemented fully by expression of a cosmid containing the entire type I TgROP5 locus (Behnke et al., 2011). Interestingly, even expression of only one or two copies of ROP5III partially restores virulence (Reese et al., 2011), showing that each allele contributes to virulence but multiple copies are required for maximum virulence.
Furthermore, a type I ΔROP5 strain complemented with a mutant TgROP5 that has the canonical kinase HRD aspartate restored (ΔROP5 + ROP5AIII(R389D)) is virulent but to a much lesser extent than those complemented with a WT allele. This result is fascinating as it shows that restoration of a canonical kinase residue is actually detrimental to TgROP5 function. Moreover, the minor virulence phenotype of the ΔROP5 + ROP5AIII(R389D) strain is not due to any restoration of catalytic activity (Reese & Boothroyd, 2011). It is possible that even the mutant TgROP5 allele can still associate with another partner in a manner that retains some function, although it has only a minor effect on virulence.
High significance was assigned to a possible interaction between TgROP5 and TgROP18 (Reese et al., 2011) and the recent published work on pseudokinases (Boudeau et al., 2006) becomes very interesting in this context. The possibility of an interaction between TgROP5 and TgROP18 suggested that ROP5 could acts as a platform or scaffold for TgROP18 at the PVM (Fig. 4).
Recently, work has been published that demonstrates the role of TgROP5 in virulence. TgROP5 appears to act as a cofactor for TgROP18 rather than a platform, as there does not seem to be a stable direct interaction between the two (Niedelman et al., 2012). TgROP5 can bind directly to IRGs and it has been shown that this alone is enough to reduce IRG burden on the PVM (Fig. 4b) (Fleckenstein et al., 2012; Niedelman et al., 2012). TgROP5 appears to bind IRGs, changing the structure to an inactive conformation and thereby preventing them from functioning to destroy the PV. TgROP18 is then able to access the essential threonine, phosphorylate it and so ensure the IRG is fully inactivated (Fig. 4b) (Fleckenstein et al., 2012). It was reported that TgROP5 binds ATP in an unusual conformation (Reese & Boothroyd, 2011), which is consistent with the new data. As TgROP5 is able to inhibit IRG activity without a highly active TgROP18 (type III parasites for example), TgROP18 cannot function without a virulent TgROP5 (as in type II) (Table 2) (Niedelman et al., 2012). Additionally, in vitro analyses confirm that TgROP18 kinase activity is enhanced in the presence of TgROP5 (Behnke et al., 2012).
In summary, TgROP5 is composed of a tandem array of genes that vary in number between types I-III. It is a secreted rhoptry pseudokinase that is essential for the virulence of type I parasites. Located on the host cytosol face of the PVM it acts as a cofactor for TgROP18 and enhances activity but may also associate with other ROPKs (Fig. 4b). TgROP5 has been shown to bind ATP but cannot complete the transfer of phosphate to a substrate even when the catalytic Asp residue is restored. This indicates that the pseudokinase status of TgROP5 was probably an evolutionary choice for a functional purpose rather than a passive drift away from active kinase ability. TgROP5 confers, independently of other ROPKs, a virulence phenotype that consists of protecting the PVM from the oligomerisation of immune-related proteins sent by the host to attack the parasite in several different cell types (Behnke et al., 2012; Fleckenstein et al., 2012).
TgROP38: impact on host genes expression
TgROP38 was identified as a member of the ROPK family through database mining of the T. gondii genome (Peixoto et al., 2010). It is differentially expressed between strains with almost no expression in a virulent type I strain, but high levels of activity in avirulent type III parasites. Type I parasites alter the expression levels of around 6000 host genes during the course of infection, type III parasites affect the expression of around 650 genes. When type I parasites overexpress an additional copy of TgROP38 under the control of a tubulin promoter, the expression of ROP38 becomes similar to that of a type III strain and the number of genes differentially expressed drops from 6000 to around 400. This level of change in gene expression is more in keeping with a type III strain (Peixoto et al., 2010), suggesting that TgROP38 has an inhibitory effect on host cell transcription. Transcription factors, apoptosis-related genes and signalling molecules, which are normally upregulated in type I infection, appear to be down-regulated when TgROP38 is overexpressed.
TgROP38 not only appears to influence the expression of a large number of genes, but also is itself one of the most highly regulated genes in the genome. Furthermore, it is induced during differentiation of the parasite from one stage to another. The changes in TgROP38 levels in the type I strain expressing a type III TgROP38 do not however appear to have an impact on growth in vitro or on virulence in vivo. It will be interesting to see future reports on the characterisation of such an influential protein.