Interaction of the innate immune system with GPI residues of the VSG molecule

African trypanosomes have evolved a structured monomolecular surface coat that covers the entire plasma membrane of the parasite, including the flagellum (Vickerman and Luckins, Reference Vickerman and Luckins1969; Cross, Reference Cross1975). This surface coat consists of a highly ordered and densely packed array of ~10⁷ identical GPI-anchored VSG homodimers that both determine the antigenic phenotype of the parasite and prevent Ab and innate immune elements such as complement from binding to subsurface invariant epitopes of the coat or to membrane determinants. Despite extensive primary sequence variation among VSGs that comprise different surface coats (see papers on antigenic variation by Schwede and Carrington and Black et al. in this special issue), secondary and tertiary structural features of these molecules (and the amino acids residues that determine such features) are highly conserved within the ordered coat structure (Metcalf et al. Reference Metcalf, Blum, Freymann, Turner and Wiley1987; Freymann et al. Reference Freymann, Down, Carrington, Roditi, Turner and Wiley1990; Carrington et al. Reference Carrington, Walters and Webb1991; Reinitz et al. Reference Reinitz, Aizenstein and Mansfield1992; Blum et al. Reference Blum, Down, Gurnett, Carrington, Turner and Wiley1993; Chattopadhyay et al. Reference Chattopadhyay, Jones, Nietlispach, Nielsen, Voorheis, Mott and Carrington2005).

During infection, GPI anchors are cleaved by activation of an endogenous GPI-phospholipase C that appears to be localized to an external linear array along the parasite flagellum (Mensa Wilmot et al. Reference Mensa Wilmot, Hereld and Englund1990; Field et al. Reference Field, Menon and Cross1991; Carrington et al. Reference Carrington, Carnall, Crow, Gaud, Redpath, Wasunna and Webb1998; Hanrahan et al. Reference Hanrahan, Webb, O'Byrne, Brabazon, Treumann, Sunter, Carrington and Voorheis2009). When GPI-anchored VSG flows across the membrane into this site, the activated enzyme cleaves the anchor and releases VSG from the membrane with residual glycosylinositolphosphate (GIP) residues attached to the shed VSG (GIP-sVSG) (see Fig. 1) (Webb et al. Reference Webb, Carnall and Carrington1994,Reference Webb, Carnall, Vanhamme, Rolin, Van Den Abbeele, Welburn, Pays and Carrington1997; Hanrahan et al. Reference Hanrahan, Webb, O'Byrne, Brabazon, Treumann, Sunter, Carrington and Voorheis2009). The release of GIP-sVSG seems to occur during episodes of high parasite burden in the presence or absence of functional adaptive immunity, and is readily detectable in host tissues (see Fig. 2A). It is the GIP residues that primarily activate cells of the innate immune system during infection and, as discussed in more detail below, there appear to be two critical points of modulation of host immunity: regulation of the MP and DC secreted cytokine profiles and alteration of DC and MP antigen-presenting cell functions (Tachado et al. Reference Tachado, Gerold, Schwarz, Novakovic, McConville and Schofield1997; Magez et al. Reference Magez, Stijlemans, Radwanska, Pays, Ferguson and De Baetselier1998, Reference Magez, Stijlemans, Baral and De Baetselier2002; Paulnock and Coller, Reference Coller and Paulnock2001; Mansfield and Paulnock, Reference Mansfield and Paulnock2005; Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b).

Fig. 1. Membrane-associated GPI-PLC cleaves the GPI anchor to release soluble VSG with GIP substituents linked to the C-terminal residue of the molecule adapted from Magez et al. (Reference Magez, Stijlemans, Baral and De Baetselier2002) and Paulnock and Coller (Reference Coller and Paulnock2001).

Fig. 2. (A) Detection of GIP-sVSG in the sera of wt and immunodeficient mice infected with Trypanosoma brucei rhodesiense LouTat 1. Western Blots with Ab specific for the cross-reacting determinant (CRD) of the GIP anchor residue in wt mice or against VSG in RAG1 KO mice (Paulnock and Coller, Reference Coller and Paulnock2001; Coller et al. Reference Coller, Mansfield and Paulnock2003). (B) Binding of fluorochrome labeled GIP-sVSG to MP membranes with subsequent downstream signaling events and activation of pro-inflammatory genes (adapted from Leppert et al. (Reference Leppert, Mansfield and Paulnock2007) and Lopez et al. (Reference Lopez, Demick, Mansfield and Paulnock2008)). Activation of MPs and DCs is dependent upon SR-A mediated, clathrin-dependent internalization of GIP-sVSG and delivery to late endolysosomal compartments. Aadapted from Leppert et al. (Reference Leppert, Mansfield and Paulnock2007) and Mansfield and Paulnock (2010).

GIP-sVSG-induced activation events

Studies in our laboratory and elsewhere have demonstrated that during early stages of infection MPs and DCs display multiple hallmarks of a ‘classical’ or pro-inflammatory, activation profile; this includes production of trypanocidal molecules (ROI, RNI, and TNF-α), pro-inflammatory cytokines (including IL-6, IL-12 and IFN-γ), and changes in functional capacity (including increased expression of MHC class II and co-stimulatory molecules for APC activity) (Magez et al. Reference Magez, Stijlemans, Radwanska, Pays, Ferguson and De Baetselier1998, Reference Magez, Stijlemans, Baral and De Baetselier2002; Paulnock and Coller, Reference Coller and Paulnock2001; Coller et al. Reference Coller, Mansfield and Paulnock2003; Harris et al. Reference Harris, Cooney, Mansfield and Paulnock2006; Barkhuizen et al. Reference Barkhuizen, Magez, Atkinson and Brombacher2007; Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008; Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b).

The initial event triggering this activation profile is binding of GIP-sVSG to Type A scavenger receptors (SR-A) on the MP and DC membrane. Binding subsequently triggers internalization and targeting of the receptor-GIP-sVSG complex to an endolysosomal compartment. This event initiates a cascade of subcellular signaling leading to activation of NFkB and MAPK pathways and expression of a subset of pro-inflammatory genes (Fig. 2B) (Coller and Paulnock, Reference Coller and Paulnock2001; Paulnock and Coller, Reference Coller and Paulnock2001; Coller et al. Reference Coller, Mansfield and Paulnock2003; Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008). These signaling events are modulated by additional factors, including (1) Traf6-, MyD88- and TLR-dependent down-regulation of GIP-sVSG-induced NFkB responses (Leppert et al. Reference Leppert, Mansfield and Paulnock2007); (2) augmentation of the response in the presence of IFN-γ (Coller et al. Reference Coller, Mansfield and Paulnock2003; Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008); and (3) modulation by CpG DNA released from senescent or damaged trypanosomes (Harris et al. Reference Harris, Cooney, Mansfield and Paulnock2006, Reference Harris, Mansfield and Paulnock2007).

An additional key component of MP and DC activation during infection is interaction of IFN-γ with the IFN-R. This receptor-ligand association generates the formation and phosphorylation of STAT1 homodimers and activation of downstream pro-inflammatory genes (Paulnock, Reference Paulnock1992, Reference Paulnock1994, Reference Paulnock2000; Paulnock et al. Reference Paulnock, Demick and Coller2000). Studies focused on understanding the nature of the interaction between the host and parasite factors in terms of innate cell activation have shown that the relative timing of MP and DC exposure to GPI substituents and to IFN-γ is critical for the ability of cells to become activated for APC functions and, for MPs, to kill the parasites. If there is coincident exposure to significant levels of both IFN-γ and GPI, augmentation of cellular activation and transcription of pro-inflammatory genes occurs (Magez et al. Reference Magez, Stijlemans, Radwanska, Pays, Ferguson and De Baetselier1998; Coller et al. Reference Coller, Mansfield and Paulnock2003; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008; Freeman, Mansfield and Paulnock, unpublished observations); this event may predominate in early infection with IFN-γ derived from Th1 cells and, potentially, NK cells (Demick, Suresh and Mansfield, unpublished observations) and may help to control the early explosive growth of trypanosomes within host tissues so the host survives the first wave of infection. However, if cells are exposed to GPI substituents prior to, or in the absence of, sufficient levels of IFN-γ, STAT1 is not phosphorylated in response to IFN-γ and the production of trypanocidal factors as well as APC functions are inhibited (Coller et al. Reference Coller, Mansfield and Paulnock2003; Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b; Freeman, Mansfield and Paulnock, unpublished observations). These events likely occur as the infection progresses, when increasing numbers of parasites release GIP-sVSG prior to sufficient activation of new T cell responses for production of IFN-γ. Finally, there is nascent evidence that not only does activation of the innate immune system trigger a potent IFN-α response but that this response ultimately negatively impacts on the host ability to produce IFN-γ during infection (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008). The sections below explore some of the primary data and their extrapolations that relate to the information presented above.

Negative regulation of GPI-mediated signaling

As noted above, GIP-sVSG is delivered to an intracellular endolysosomal compartment of MPs and DCs by SR-A-mediated uptake. SR-As are expressed primarily but not exclusively on cells of the monocyte/MP lineage, some DCs, and on other cells; they are trimeric integral membrane glycoproteins that demonstrate broad ligand-binding properties including altered self proteins and many different types of microbial PAMPs. While scavenger receptors were once thought to serve simply as non-opsonic phagocytic receptors for many ligands, it is now clear that there are direct signaling events generated by ligand engagement of these receptors, including SR-A [as shown in our earlier parasite-unrelated studies; (Coller and Paulnock, Reference Coller and Paulnock2001) and by others (Todt et al. Reference Todt, Hu and Curtis2008; DeWitte-Orr et al. Reference DeWitte-Orr, Collins, Bauer, Bowdish and Mossman2010)]; tyrosine phosphorylation and activation of MAPK and NFkB pathways appear to be hallmarks of SR-A-mediated signaling. Further, SR-A and other scavenger receptors have now been shown to interact functionally and synergistically with other pattern recognition receptors (PRRs) of the innate immune system, including TLRs, to generate fully functional innate immune responses to different microbial pathogens (Bowdish and Gordon, et al. 2009; Bowdish et al. 2009; DeWitte-Orr et al. Reference DeWitte-Orr, Collins, Bauer, Bowdish and Mossman2010; Stewart et al. Reference Stewart, Stuart, Wilkinson, van Gils, Deng, Halle, Rayner, Boyer, Zhong, Frazier, Lacy-Hulbert, Khoury, Golenbock and Moore2010; Yew et al. Reference Yew, Carsten and Harrison2010).

PAMP recognition by combinatorial activation of multiple PRRs appears to be an evolutionarily conserved but poorly understood and complex phenomenon (Amit et al. Reference Amit, Garber, Chevrier, Leite, Donner, Eisenhaure, Guttman, Grenier, Li, Zuk, Schubert, Birditt, Shay, Goren, Zhang, Smith, Deering, McDonald, Cabili, Bernstein, Rinn, Meissner, Root, Hacohen and Regev2009). To date it is clear that TLR-independent signaling is triggered by GIP-sVSG binding to SR-A, and that NFkB and MAPK signaling pathways are activated after internalization (Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Freeman, Mansfield and Paulnock, unpublished observations; Mansfield and Paulnock, 2010). This GIP-sVSG–SR-A ligand/receptor interaction activates subsequent TLR-dependent signaling events that then down-regulate the initial response (Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Mansfield and Paulnock, unpublished observations). While scavenger receptor interactions in other studies have been shown to promote positive TLR-mediated signaling effects by catalyzing activation, dimerization or multimerization of TLRs, our work is the first to show that SR-A-mediated signaling triggers a negative regulatory TLR signaling cascade that modifies the initial SR-A signaling event (Leppert et al. Reference Leppert, Mansfield and Paulnock2007). We know that TLRs are not required for SR-A induced signaling because the absence, inhibition or ablation of TLR signaling results in enhanced NFkB/MAPK signaling by GIP-sVSG [see examples in Fig. 3; (Leppert et al. Reference Leppert, Mansfield and Paulnock2007)]. However, the biological relevance of the SR-A-mediated activation response alone is not entirely clear and must be elucidated; currently we presume it to be a positive and protective effect early in infection but a counter-protective effect as infection progresses (Coller et al. Reference Coller, Mansfield and Paulnock2003; Mansfield and Paulnock, Reference Mansfield and Paulnock2005).

Fig. 3. GIP-sVSG-induced NFkB signaling is enhanced in the absence of TLR signaling (A): irreversible Traf6 inhibitor; (Leppert et al. Reference Leppert, Mansfield and Paulnock2007) and TLR9 inhibitory ODN (B): TLR9 iODN 2088; (Mansfield and Paulnock, 2010).

Our recent preliminary studies show clearly that TLR9 is the intracellular receptor to which GIP-sVSG is delivered and that directs the negative regulatory effect on GIP-sVSG induced signaling. Using TLR-deficient cells and known TLR inhibitors, we have ruled out the engagement of TLR2, TLR4 and other factors in this negative regulation, and have shown that TLR9 is the relevant receptor engaged by GIP (see Fig. 3B) (Mansfield et al. unpublished observations). It is not known which substituent of GIP serves as the TLR9 ligand. An early concern was that contamination of sVSG preparations with trypanosome CpG DNA was the active factor since CpG DNA released during infection contributes to MP activation (Harris et al. Reference Harris, Mansfield and Paulnock2007), but HPLC purification of GIP-sVSG, the use of DNase treatments, and the knowledge that trypanosome CpG DNA is stimulatory only in significantly high concentrations (Harris et al. Reference Harris, Cooney, Mansfield and Paulnock2006; Leppert et al. Reference Leppert, Mansfield and Paulnock2007), revealed this was not the case. It is our current hypothesis that the phosphate residue of GIP, exposed as part of the GPI anchor residuum after GPI-PLC cleavage of GPI-mfVSG, is the most likely candidate for TLR9 interaction, based on recent studies elaborating such interactions in other systems (Gangloff and Gay Reference Gangloff and Gay2008).

Down-modulation of IFN-γ-mediated MP activation

As mentioned above, the interplay and timing of parasite and host factors are critical for control of the parasite burden in host tissues and for regulation of host immune responses that provide such protection. Studies in our laboratory have shown that GIP-sVSG substituents are capable of blocking production of trypanocidal factors and parasite killing by interfering with IFN-γ-dependent activation of MPs and also by inhibiting APC function; these inhibitory events may be central to down-stream parasite evasion of host adaptive immunity.

Exposure of MPs to GIP-sVSG induces the rapid production of pro-inflammatory and protective factors including TNF-α, RNI and ROI and this response can be enhanced by initial or prior treatment of the cells with IFN-γ (Magez et al. Reference Magez, Stijlemans, Radwanska, Pays, Ferguson and De Baetselier1998; Coller et al. 2001; Paulnock et al. 2001; Coller et al. Reference Coller, Mansfield and Paulnock2003; Leppert et al. Reference Leppert, Mansfield and Paulnock2007; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008). However, the level and timing of exposure of MP to GIP-sVSG vs IFN-γ ultimately determines the MP response at the level of induced gene expression. Treatment of MPs with GIP-sVSG prior to IFN-γ exposure results in a marked reduction of IFN-induced responses, including transcription of inducible nitric oxide synthase and production of NO (Fig. 4) (Coller et al. Reference Coller, Mansfield and Paulnock2003); this effect is mediated through a dose-related inhibition of STAT1 phosphorylation, which is required for signaling through the IFN-γ receptor (Fig. 4A). Thus, in this manner, regulation of the ability of MPs to respond to IFN-γ during infection is a central mechanism by which the parasites overcome host resistance.

Fig. 4. (A) Inhibition of IFN-γ mediated STAT1 phosphorylation in MPs by prior exposure to GIP-sVSG. (B) Suppression of iNOS transcription and NO production in MPs exposed to GIP-sVSG prior to activating doses of IFN-γ. Adapted from Coller et al. (Reference Coller, Mansfield and Paulnock2003).

These results suggest that cyclical exposure of innate immune cells to varying levels of GIP-sVSG as the infection progresses can result in changes in the MP response to IFN-γ during the course of infection. Given that infection with Trypanosoma brucei spp. can result in a chronic course of infection in relatively resistant animals and humans, this may be one mechanism not only for permitting parasites to escape from destruction in the tissues, but also for controlling host tissue pathology. However, a key question is when these changes in MP responses, and the accompanying changes in functional activity, occur and at what level they continue to fluctuate over the course of infection. Studies in a Trypanosoma congolense experimental model system as well as in human infections with Trypanosoma brucei spp. have suggested that during the later stages of infection there is a permanent shift in the MP activation state from a ‘classical’ to an ‘alternative’ activation phenotype, with the latter state associated with production of anti-inflammatory cytokines including IL-10 (Uzonna et al. Reference Uzonna, Kaushik, Gordon and Tabel1999; Kaushik et al. Reference Kaushik, Uzonna, Zhang, Gordon and Tabel2000; MacLean et al. Reference MacLean, Odiit and Sternberg2001; Sternberg et al. Reference Sternberg, Rodgers, Bradley, MacLean, Murray and Kennedy2005). A similar trend is observed in a T. brucei brucei model system, but only when GPI-PLC^−/− parasites, which do not shed GIP-sVSG from their VSG coat, are used (Webb et al. Reference Webb, Carnall, Vanhamme, Rolin, Van Den Abbeele, Welburn, Pays and Carrington1997). Our work and the work of others (Stijlemans et al. Reference Stijlemans, Baral, Guilliams, Brys, Korf, Drennan, Van Den Abbeele, De Baetselier and Magez2007) suggests that it is unlikely that a full ‘alternatively activated’ MP response is induced in wild type Trypanosoma brucei rhodesiense infection, as a Th2 cell cytokine response, associated with alternatively activated MPs and capable of IL-4 production, has not been detected during chronic infection, even in IL-12 knockout mice that lack the ability to induce early Th1 polarization (Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b). Thus, IL-10 may clearly have a non-Th2 origin in trypanosomiasis. However, MPs from chronically infected animals clearly are impaired in their ability to produce pro-inflammatory cytokines when they subsequently encounter stimulatory molecules such as trypanosome GIP-sVSG, CpG DNA or LPS, while at the same time IL-10 production is enhanced in both infected animals and humans (Uzonna et al. Reference Uzonna, Kaushik, Gordon and Tabel1999; Kaushik et al. Reference Kaushik, Uzonna, Zhang, Gordon and Tabel2000; Namangala et al. Reference Namangala, Brys, Magez, De Baetselier and Beschin2000; MacLean et al. Reference MacLean, Odiit and Sternberg2001; Sternberg et al. Reference Sternberg, Rodgers, Bradley, MacLean, Murray and Kennedy2005; Harris et al. Reference Harris, Cooney, Mansfield and Paulnock2006; Barkhuizen et al. Reference Barkhuizen, Magez, Ryffel and Brombacher2008; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008; Mansfield et al. unpublished observations); this impairment is seen in experimental mice by the time the second wave of parasitaemia appears, around two weeks post-infection.

This response phenotype is similar to that observed during induced tolerance to TLR ligands and certainly provides a significant advantage to the parasite in terms of survival (Medvedev et al. Reference Medvedev, Kopydlowski and Vogel2000). Overall, the response capability of innate immune cells clearly is diminished early in infection and may be the partial basis for subsequent alterations in APC function described below. Thus it is clear that MP plasticity during infection results in a wide range of activation responses at various times in trypanosomiasis. The ability of GIP-sVSG exposure to dramatically reduce the ability of MPs to respond to subsequent activation by IFN-γ may be the most important of these changes and has obvious implications for host resistance, since IFN-γ activation of MPs for the production of trypanocidal or trypanostatic factors in the extravascular environment is key for host protection.

Negative regulation of IFN-γ production

An additional outcome of GPI-sVSG induced cellular activation that impacts on IFN-γ-mediated host resistance is the induction of Type I interferon (IFN-αβ) production during infection. We have recently shown by microarray analysis of naïve MPs exposed to GIP-sVSG in vitro that a prominent component of the cytokine response induced by GIP-sVSG encompassed genes associated with the Type I interferon (IFN-αβ) response including IRF-7 and IFN-α (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008), suggesting that GIP-sVSG recognition by MPs results in modulation in the IRF family of transcription factors in a manner similar to the modulation of STAT1 signaling observed in earlier experiments. The subsequent production of IFN-αβ (De Gee et al. Reference De Gee, Sonnenfeld and Mansfield1985; Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008) may set in motion the IRF-7- and IFN-dependent modulation of host resistance during late stage infection that inhibits IFN-γ production by Th1 cells (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008). The pattern observed in MPs treated with GIP-sVSG in vitro was similar to that which was observed in MPs taken directly from trypanosome infected mice within 72 h post-infection, and suggests a role for IFN-αβ in GIP-sVSG-mediated MP activation events. Results of in vivo studies with trypanosome infection in IFNAR1^−/− mice, which lack the receptor for Type I interferons, and UBP43^−/− mice, which are hyper-responsive to Type I interferons, suggest that IFN-αβ may play a pivotal role in host resistance. Results obtained using UBP43^−/− mice are particularly informative, as these mice are more susceptible to African trypanosome infection than their wild type counterparts, initially displaying MP activation profiles that promote control of parasite burden during the early phases of infection but, as the disease progresses, promoted down-regulation of host IFN-γ production and host resistance (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008). These results correlate well with studies in a variety of other non-viral microbial systems that suggest a role for IFN-αβ in modulating the course of infection, although the majority of the bacterial, fungal, and protozoan pathogen systems involve intracellular microbes.

Alterations in APC function during infection

For nearly a century the concept of trypanosome immunology encompassed a central paradigm in which Ab was the primary mechanism that protected the host against different parasite variant antigenic types (VATs) during infection. The old paradigm first began to unravel with publication of studies on resistant and susceptible bone marrow chimera mice in which Ab was not functionally linked to overall host resistance (De Gee and Mansfield, Reference De Gee and Mansfield1984). Subsequently classical genetic studies employing crosses between resistant and susceptible animals, as well as results from recombinant inbred strains, were used to demonstrate that Ab-mediated control of parasitaemia by itself was not genetically linked to the host resistance phenotype (De Gee et al. Reference De Gee, Levine and Mansfield1988). Thus, although VSG-specific Abs provide an important mechanism for controlling trypanosome numbers in the blood (and perhaps also provide an unerringly specific means to select for new VATs in the process!), this event alone is insufficient to provide significant host resistance.

These seminal results laid the foundation for subsequent studies in which resistant and susceptible animals were examined for T cell-mediated immune responses to trypanosome antigens. Direct evidence for VSG-specific T cell responses came from our experimental studies with Trypanosoma brucei rhodesiense infections of mice (Schleifer et al. Reference Schleifer, Filutowicz, Schopf and Mansfield1993a,b). VSG-activated T cells were shown to be CD4⁺ lymphocytes that expressed the CD3⁺ TCR membrane complex, were APC-dependent and MHC-II restricted (Schleifer et al. Reference Schleifer, Filutowicz, Schopf and Mansfield1993a; Hertz et al. Reference Hertz, Filutowicz and Mansfield1998; Schopf et al. Reference Schopf, Filutowicz, Bi and Mansfield1998; Dagenais et al. Reference Dagenais, Demick, Bangs, Forest, Paulnock and Mansfield2009a). The cytokine profiles, both transcriptional and protein secretory, were those associated with highly polarized Th1 cell responses characterized by IL-2 and IFN-γ production without detectable IL-4 or IL-5 responses. Thus, VSG preferentially stimulates a polarized Th 1 cell cytokine response during trypanosome infection.

VSG-specific Th1 cells are the primary source of IFN-γ produced during infection. IFN-γ is linked to host resistance against Brucei group trypanosomes through macrophage activation events, as noted above, that lead to the release of factors such as TNF-α, ROI and RNI that are cytotoxic for trypanosomes (De Gee et al. Reference De Gee, Sonnenfeld and Mansfield1985; Magez et al. Reference Magez, Lucas, Darji, Songa, Hamers and De Baetselier1993, Reference Magez, Geuskens, Beschin, del Favero, Verschueren, Lucas, Pays and de Baetselier1997, 1999, 2006; Lucas et al. Reference Lucas, Magez, De Leys, Fransen, Scheerlinck, Rampelberg, Sablon and De Baetselier1994; Hertz et al. Reference Hertz, Filutowicz and Mansfield1998,1999; Drennan et al. Reference Drennan, Stijlemans, Van den Abbeele, Quesniaux, Barkhuizen, Brombacher, De Baetselier, Ryffel and Magez2005; Dagenais et al. Reference Dagenais, Demick, Bangs, Forest, Paulnock and Mansfield2009a). Infection of mice lacking the IFN-γ gene was shown to result in susceptibility similar to that of scid mice, which have no adaptive immune system, despite the fact that such IFN-γ KO animals made Ab responses that controlled trypanosomes in the bloodstream like wt mice (Hertz et al. Reference Hertz, Filutowicz and Mansfield1998; Schopf et al. Reference Schopf, Filutowicz, Bi and Mansfield1998). The current presumption is that IFN-γ activation of macrophages with concomitant production of microbicidal factors is responsible for control of the parasite burden within extravascular tissues where Ab is relatively ineffective in killing trypanosomes. Thus, the current paradigm is that VSG-specific Th1 cells and IFN-γ regulate the major component of host resistance to African trypanosomes.

Given that APC function is critical to the development, maintenance and memory responses of VSG-specific Th cells, and that GIP-sVSG has the capacity to alter both innate and adaptive immune responses, it is only natural that antigen-presenting cells have come under observation for functional abnormalities. To date, several studies have demonstrated that APC functions are significantly altered during experimental trypanosomiasis (Paulnock et al. Reference Paulnock, Smith, Mansfield and Alan1989; Namangala et al. Reference Namangala, Brys, Magez, De Baetselier and Beschin2000; Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b). Recently a detailed analysis of APCs involved in the activation of naïve Th cells and induction of the polarized Th1 response in trypanosomiasis showed that DCs are the primary cell population capable of activating T cells to both parasite and non-parasite antigens (Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b). However, a key finding was that, during the process of APC activation of VSG-specific T cells, APCs rapidly became unable to process or present additional exogenous antigens to T cells (see Fig. 5A) (Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b). Subsequent analyses revealed that the uptake and internalization of exogenous antigens by APCs was not significantly altered; rather, it appeared that existing peptide (presumably VSG)-MHC II complexes were not being recycled from the membrane and that formation of new intracellular MHC II-peptide complexes was unstable (Freeman et al. 2010). Of central importance to the control of parasites was the inability of APCs to present antigenic peptides from newly encountered VSG molecules expressed by variants arising during the infection (Dagenais et al. Reference Dagenais, Freeman, Demick, Paulnock and Mansfield2009b; Freeman, Mansfield and Paulnock, unpublished observations). In these studies it was of interest that all the relevant co-stimulatory molecules were highly up-regulated on the APCs, and that during the period of APC dysfunction these cells still expressed T cell activating MHC II-VSG peptide complexes in which the peptides were derived from the initial infecting variant type. While the mechanistic basis for this remains to be determined, the GPI substituents of VSG seem to be intimately involved in the process (Mansfield and Paulnock, Reference Mansfield and Paulnock2005; Mansfield et al. unpublished observations). Thus, along with other host innate immune cell modulations induced by trypanosomes and their molecules, APCs are also affected by infection and contribute to the overall down-modulation of protective Th1 cell responses and IFN-γ-dependent activation of MPs that lead to trypanocidal activity.

Fig. 5. (A) APCs from trypanosome infected mice were pulsed with HEL or HEL peptides and exposed to HEL-specific T cell hybridomas to assess processing and presentation of HEL antigen. (B) DCs from mice infected with one trypanosome variant, and which express the relevant MHC II-VSG peptides, are unable to process and present exogenous VSGs from other variants. Adapted from Dagenais et al. (2009a,b).

Trypanosome virulence suppresses host resistance

The African trypanosomes undergo cellular differentiation and biological variation throughout their life cycle in which the organisms change surface coat glycoproteins, alter biochemical pathways, remodel cellular morphology and exhibit distinct host specificities. These changes are regulated by the expression of discrete genes and proteins (Vickerman and Luckins, Reference Vickerman and Luckins1969; Rifkin, Reference Rifkin1978; Duszenko et al. Reference Duszenko, Ferguson, Lamont, Rifkin and Cross1985; Donelson, Reference Donelson1987; Hajduk et al. Reference Hajduk, Moore, Vasudevacharya, Siqueira, Torri, Tytler and Esko1989, Reference Hajduk, Hager and Esko1992, Reference Hajduk, Smith and Hager1995; Cross, Reference Cross1990; Van-der-Ploeg et al. Reference Van-der-Ploeg, Gottesdiener and Lee1992; Borst and Rudenko, Reference Borst and Rudenko1994; De Greef et al. 1994; Steverding et al. Reference Steverding, Stierhof, Chaudhri, Ligtenberg, Schell, Beck Sickinger and Overath1994; Smith et al. Reference Smith, Esko and Hajduk1995; Hager and Hajduk, Reference Hager and Hajduk1997; Bitter et al. Reference Bitter, Gerrits, Kieft and Borst1998; Borst et al. Reference Borst, Bitter, Blundell, Chaves, Cross, Gerrits, van Leeuwen, McCulloch, Taylor and Rudenko1998; Xong et al. Reference Xong, Vanhamme, Chamekh, Chimfwembe, Abbeele, Pays, Van Meirvenne, Hamers, De Baetselier and Pays1998; Milner and Hajduk,Reference Milner and Hajduk1999; Raper et al. Reference Raper, Fung, Ghiso, Nussenzweig and Tomlinson1999, Reference Raper, Portela, Lugli, Frevert and Tomlinson2001; Molina Portela et al. Reference Molina Portela, Raper and Tomlinson2000; Navarro and Gull, Reference Navarro and Gull2001; Gerrits et al. Reference Gerrits, Mussmann, Bitter, Kieft and Borst2002; Mussmann et al. Reference Mussmann, Janssen, Calafat, Engstler, Ansorge, Clayton and Borst2003; Landeira et al. Reference Landeira, Bart, Van Tyne and Navarro2009; Mansfield et al. 2010). Many molecular elements of trypanosome antigenic variation, cellular differentiation, metabolism, morphology and host specificity may be defined as primary virulence factors for all African trypanosomes since parasites in which the relevant genes have been deleted, mutated or knocked down are unable to infect or survive in specific mammalian or insect hosts; for instance see Leal et al. (Reference Leal, Acosta-Serrano, Morita, Englund, Bohme and Cross2001), Hoek et al. (Reference Hoek, Zanders and Cross2002) and Wang et al. (Reference Wang, Bohme and Cross2003). A peripatetic focus of our laboratory, however, has been on relative virulence of trypanosomes for their mammalian hosts. It is well known that different isolates, species, subspecies and laboratory lines of trypanosomes exhibit remarkable variation in pathogenicity and virulence for genetically defined host species (McNeillage and Herbert, Reference McNeillage and Herbert1968; Mulligan, Reference Mulligan1970; Clayton, Reference Clayton1978; Barry et al. Reference Barry, Le Ray and Herbert1979; Levine and Mansfield, Reference Levine and Mansfield1981, Reference Levine and Mansfield1984; Inverso and Mansfield, Reference Inverso and Mansfield1983). The underlying question has been whether such biological differences are immutable characteristics associated with genetically distinct populations of trypanosomes, or whether intraclonal biological variation occurs in trypanosomes that impacts on the course of infection within an individual host animal.

We addressed this question in earlier studies in which inbred mice individually were infected with a single Trypanosoma brucei rhodesiense LouTat 1 trypanosome (Inverso and Mansfield, Reference Inverso and Mansfield1983; Inverso et al. Reference Inverso, De Gee and Mansfield1988; Mansfield and Paulnock, 2010). Multiple VATs were isolated from the blood at intermediate and later times during the ensuing infection; these organisms were subcloned and characterized as to VSG phenotype and relative pathogenicity. Different VATs, which represented antigenically distinct daughter cell populations clonally derived from a single LouTat 1 ‘parental’ cell in a single animal, were then used to infect resistant mouse strains. The surprising result was that each daughter cell population exhibited a different virulence phenotype compared to the parental clone. For example, infection with the parental LouTat 1 caused death in approximately 62 days post-infection, while infection with LouTat 1.3, LouTat 1.4 and LouTat 1.5 caused death in approximately 44, 30, and 28 days, respectively; trypanosomes taken from terminal peaks of parasitaemia were highly virulent, causing death in less than 10 days (Inverso and Mansfield, Reference Inverso and Mansfield1983; Inverso et al. Reference Inverso, De Gee and Mansfield1988; Johnson et al. 2010).

These results clearly demonstrated that VATs arising during trypanosome infection expressed virulence phenotypes dramatically different from the infecting VAT. In essence, daughter cells arising within a trypanosome population expressed the capacity to transcend host genetic resistance characteristics and render a relatively resistant animal into a more susceptible one. Such virulence expression appears to be a progressive phenotype in chronic infections since low virulence trypanosomes always give rise to more virulent trypanosomes. Thus, trypanosomes appear to have evolved a molecular ‘Virulence Rheostat’ that is inherent to every trypanosome cell and that, at least in experimental model systems, ultimately kills the infected host (Inverso and Mansfield, Reference Inverso and Mansfield1983; Inverso et al. Reference Inverso, De Gee and Mansfield1988, Reference Inverso, Uphoff, Johnson, Paulnock and Mansfield2010; Mansfield Reference Mansfield2006; Johnson et al. 2010). This mechanism may have evolved in order to provide trypanosomes the ability to ultimately escape destruction regardless of the disparate genetic resistance traits displayed by animal or human hosts upon which infected tsetse flies feed (Mansfield, Reference Mansfield2006; Arnold et al. Reference Arnold, Jackson, Waterfield and Mansfield2007).

We have developed a tractable model system for the molecular exploration of trypanosome virulence in which we compared low virulence parental cells to intermediate and high virulence daughter cells within the T. b. rhodesiense LouTat 1 serodeme. Several important features have emerged: First, virulence is not linked to the VSG coat, the mechanism of VSG gene expression or the specific chromosome telomeric VSG gene expression site (Inverso et al. Reference Inverso, De Gee and Mansfield1988; Inverso et al. Reference Inverso, Uphoff, Johnson, Paulnock and Mansfield2010). Second, high virulence expression results in the abrogation of measureable host innate and adaptive resistance (Mansfield and Paulnock, Reference Mansfield and Paulnock2005; Mansfield, Reference Mansfield2006; Johnson et al. 2010; Mansfield et al. 2010); for example, while a spectrum of relative resistance is observed when genetically distinct mouse strains are infected with low virulence organisms, such as LouTat 1 (Levine and Mansfield, Reference Levine and Mansfield1981, Reference Levine and Mansfield1984; De Gee et al. Reference De Gee, Levine and Mansfield1988), infection of these animals with high virulence trypanosomes derived from the low virulence parental cells such as LouTat 1A (Inverso and Mansfield, Reference Inverso and Mansfield1983; Inverso et al. Reference Inverso, De Gee and Mansfield1988, Reference Inverso, Uphoff, Johnson, Paulnock and Mansfield2010; Johnson et al. 2010) eliminates this spectrum and results in rapid death. Similarly, infection of immunodeficient scid mice with low virulence trypanosomes reveals a basal level of innate resistance, but this resistance is also ablated when such animals are infected with more virulent trypanosomes. Third, measureable and repeatable differences in the pattern of intraclonal parasite gene and protein expression occur as trypanosomes become more virulent with time of infection (Inverso et al. Reference Inverso, Uphoff, Johnson, Paulnock and Mansfield2010; Johnson et al. 2010), and some genes and proteins are predicted to impact on host immunity. This pattern is, however, not reversible in infection of the mammalian host but must, somehow, be reset to a low level of virulence in nature, perhaps in the intermediate host and vector, the tsetse fly. Fourth, and relevant to the present discussion regarding the innate immune system, trypanosomes displaying an elevated virulence phenotype do not activate the same pattern of genes in the infected host as do low virulence organisms (see Table 1). Microarray analyses of immune system genes activated early in infection reveal that while virulent organisms activate some of the same genes as less virulent organisms (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008), it is a small subset of the total array of genes normally activated [Table 1; (Lopez et al. Reference Lopez, Demick, Mansfield and Paulnock2008; Johnson et al. 2010; Mansfield and Paulnock, 2010)]. Thus, virulent trypanosomes do not trigger the same danger signals and activation events as those triggered by less virulent organisms; this is not related to GPI residues expressed by high vs low virulence trypanosomes, as these appear to be identical in composition and biological activity in vitro (Mansfield and Paulnock, 2010). In summary, in addition to the disparate and multiple mechanisms that trypanosomes have evolved to alter host immunity, the trypanosome has another trick up its sleeve with the intraclonal expression of virulence which renders all animals highly susceptible to infection.

Table 1. Microarray analyses of the immune transcriptome in mice infected with Trypanosoma brucei rhodesiense LouTat 1 (low virulence) vs LouTat 1A (high virulence). LouTat 1 induces the expression of all genes shown with the exception of CXCL19 (*), while LouTat 1A induces only a subset of the genes (underlined). Both organisms express the same VSG gene and surface coat, and exhibit identical GPI anchor residues (Inverso, De Gee et al. Reference Inverso, De Gee and Mansfield1988; Inverso, Uphoff et al. Reference Inverso, Uphoff, Johnson, Paulnock and Mansfield2010).