Innate immune responses

Macrophages play a pivotal role in Leishmania infection. Effective clearance of all forms of leishmaniasis depends on efficient elimination of parasites by activated macrophages. Macrophages are professional phagocytes and Leishmania utilizes their phagocytic function as a strategy for internalization and subsequent replication within the macrophage phagolysosomes (Reiner and Locksley, Reference Reiner and Locksley1995). Thus, macrophages act as both host cells for leishmanial replication and effector cells that kill the parasites. Internalization of Leishmania by macrophages triggers the production of reactive oxygen species (Basu and Ray, Reference Basu and Ray2005) and leads to the generation of nitric oxide (NO) (Liew et al. Reference Liew, Li and Millott1990) and N-hydroxy-l-arginine (LOHA) (Iniesta et al. Reference Iniesta, Gomez-Nieto and Corraliza2001), which can mediate parasite killing. There are different requirements for effective killing of the different Leishmania species that cause CL. While NO and LOHA are sufficient for elimination of Leishmania major (Wei et al. Reference Wei, Charles, Smith, Ure, Feng, Huang, Xu, Muller, Moncada and Liew1995), a successful anti-Leishmania amazonensis response also requires superoxide production (Mukbel et al. Reference Mukbel, Patten, Gibson, Ghosh, Petersen and Jones2007). Additionally, infection of macrophages leads to the production of pro-inflammatory cytokines that are implicated in parasite killing. Interleukin (IL)-12 is necessary for the leishmanicidal activity of macrophages, as it leads to up-regulation of interferon gamma (IFN-γ) by T cells and NK cells, generation of Th1-type responses and T cell-dependent and -independent macrophage activation leading to an increase of inducible nitric oxide synthase (iNOS) and NO production, and subsequent parasite elimination (Trinchieri, Reference Trinchieri1998). A subversive activity of Leishmania parasites in this process is the inhibition of IL-12 production, which down-regulates the immune response to infection (Ricardo-Carter et al. Reference Ricardo-Carter, Favila, Polando, Cotton, Bogard Horner, Condon, Ballhorn, Whitcomb, Yadav, Geister, Schorey and McDowell2013).

Production of pro-inflammatory cytokines by macrophages results in the recruitment of pro-inflammatory cells such as neutrophils, mast cells and eosinophils to the site of infection.

Neutrophils are among the first cells recruited to the site of infection and are thought to participate in the containment of Leishmania parasites within an hour of infection (Belkaid et al. Reference Belkaid, Mendez, Lira, Kadambi, Milon and Sacks2000). Neutrophils are a major component of innate immunity that protects against microbial infections (Segal, Reference Segal2005). Published data on the involvement of neutrophils in Leishmania infection are contradictory, indicating either their role in resistance to leishmaniasis or disease-exacerbating activities (Peters and Sacks, Reference Peters and Sacks2009). However, it has been shown that in the context of infection initiated by the bite of an infected sandfly, neutrophils are recruited to the site of infection and phagocytose parasites, a process that is vital for disease progression (Peters et al. Reference Peters, Egen, Secundino, Debrabant, Kimblin, Kamhawi, Lawyer, Fay, Germain and Sacks2008). These findings suggest that neutrophils, apoptotic neutrophils in particular, are more likely to play a role in promoting disease progress, rather than resistance. However, other studies have found that neutrophils contribute to parasite killing through the release of neutrophil extracellular traps (NETs) (Guimaraes-Costa et al. Reference Guimaraes-Costa, Nascimento, Froment, Soares, Morgado, Conceicao-Silva and Saraiva2009). Subsequent production of mast cell-derived mediators, immunoglobulin G-mediated mechanisms and cytokines/chemokines released by macrophages and neutrophils result in the recruitment of DCs, an important component linking the innate with the adaptive immune response against Leishmania (Reiner and Locksley, Reference Reiner and Locksley1995).

The main function of DCs is the recognition and processing of foreign antigens, and subsequent presentation of antigen to T cells (Banchereau and Steinman, Reference Banchereau and Steinman1998), and as such they are considered to be gatekeepers in the defence against invading pathogens. Upon encountering antigen in the presence of pro-inflammatory cytokines, immature DCs residing in the skin undergo maturation and migrate to draining lymph nodes (Randolph, Reference Randolph2001). Skin DCs, Langerhans cells and dermal DCs are very efficient at presenting antigen to naïve T cells (Von Stebut, Reference Von Stebut2007). In the case of Leishmania infection, dermal DCs appear to present antigen directly to T cells (Ritter et al. Reference Ritter, Meissner, Scheidig and Korner2004). Small numbers of parasites are taken up directly by dermal DCs shortly after infection (Ng et al. Reference Ng, Hsu, Mandell, Roediger, Hoeller, Mrass, Iparraguirre, Cavanagh, Triccas, Beverley, Scott and Weninger2008), but the majority of the DCs become infected through contact with parasitized neutrophils (Peters and Sacks, Reference Peters and Sacks2009). Several weeks post-infection the number of DCs (CD11c⁺ cells) in the lesion increases through recruitment (Belkaid et al. Reference Belkaid, Mendez, Lira, Kadambi, Milon and Sacks2000) and infected DCs are able to prime naïve CD4⁺ and CD8⁺ T cells (Woelbing et al. Reference Woelbing, Kostka, Moelle, Belkaid, Sunderkoetter, Verbeek, Waisman, Nigg, Knop, Udey and von Stebut2006). Activated DCs migrate to draining lymph nodes where apart from T cells, they also activate resting NK cells and trigger IFN-γ production (Bajenoff et al. Reference Bajenoff, Breart, Huang, Qi, Cazareth, Braud, Germain and Glaichenhaus2006). However, Leishmania parasites have evolved complex mechanisms to avoid DC functions, leading to the down-regulation of DC activation during infection. Amastigote infection of DCs results in reduced phosphorylation and degradation of vital molecules in Janus kinase/Signal Transducer and Activator of Transcription (JAK/STAT), nuclear factor (NF)-κB and Interferon Regulatory Factor (IRF) pathways (Soong, Reference Soong2008), which leads to inadequate DC activation, T cell priming, impaired NK cell activation and suppression of IL-12 and IFN-γ production. These might be Leishmania-related general phenomena; however, there are species- and stage-specific differences in modulation of DC functions. While infection with L. major or L. donovani promastigotes led to production of IL-12 by murine DCs (Gorak et al. Reference Gorak, Engwerda and Kaye1998; von Stebut et al. Reference von Stebut, Belkaid, Jakob, Sacks and Udey1998), infection with Leishmania mexicana amastigotes did not lead to DC activation or IL-12 and other pro-inflammatory cytokines production (Bennett et al. Reference Bennett, Misslitz, Colledge, Aebischer and Blackburn2001). Similarly, infection with L. amazonensis amastigotes leads to down-regulation of signalling events and impaired DC function (Xin et al. Reference Xin, Li and Soong2008). In humans L. amazonensis has been shown to use Langerhans cells to skew CD4⁺ T cell function towards regulatory T (T reg) cells and to suppress protective responses (Silveira et al. Reference Silveira, Lainson, Gomes, Laurenti and Corbett2008).

Amastigote uptake by DCs at the site of infection results in the up-regulation of IL-12 (Gorak et al. Reference Gorak, Engwerda and Kaye1998), which is essential for parasite elimination within DCs (von Stebut et al. Reference von Stebut, Belkaid, Jakob, Sacks and Udey1998) and for the effector functions of macrophages (Belkaid et al. Reference Belkaid, Butcher and Sacks1998). Uptake of amastigotes by DCs also leads to surface up-regulation of MHC I, MHC II and co-stimulatory molecules. The ability of DCs to present antigens through the MHC I and II pathways leads to stimulation of Leishmania-specific CD4⁺ and CD8⁺ T cell responses (Flohe et al. Reference Flohe, Lang and Moll1997; Belkaid et al. Reference Belkaid, Butcher and Sacks1998; von Stebut et al. Reference von Stebut, Belkaid, Jakob, Sacks and Udey1998). Although other cell sub-sets, including macrophages and B cells, are able to present leishmanial antigens, antigen presentation by DCs is essential for acquired resistance to Leishmania.

Adaptive immune response

T cells play an essential role in the generation of effector and memory responses to intracellular pathogens. In general terms, protective immunity is associated with a cell-mediated immune response, whereas non-protective responses have a strong humoral component in the absence of cell-mediated immunity. The protection against CL is intimately linked to development of Th1-type immunity and IFN-γ production. Experimental studies established a clear-cut dichotomy between Th1-mediated protection and Th2-mediated susceptibility. In resistant C57BL/6 mice, resolution of the disease is mediated as a consequence of IFN-γ release by Th1 cells and up-regulation of NO in macrophages that harbour parasites (Bogdan et al. Reference Bogdan, Rollinghoff and Diefenbach2000). Conversely, persistence of lesions in susceptible BALB/c mice is due to CD4⁺ T cell differentiation to Th2-type effector cells and the production of IL-4, which in turn promotes antibody responses and suppresses macrophage activation, resulting in parasite survival and replication (Scott, Reference Scott1989). The Th1 response is linked to IFN-γ production; however, it is functionally heterogeneous. It has been shown that a high frequency of CD4⁺ T cells producing only IFN-γ is not sufficient for resistance to infection. The quality and magnitude of the response are crucial factors influencing a protective outcome and are controlled by the type of antigen-presenting cells (APCs), amount of antigen and the duration for which it is presented to the immune system as well as cytokine milieu (Iezzi et al. Reference Iezzi, Karjalainen and Lanzavecchia1998; O'Garra, Reference O'Garra1998; Steinman and Hemmi, Reference Steinman and Hemmi2006). The Th1 response mounted by CD4⁺ T cells which are single-positive, i.e. producing only IFN-γ or tumour necrosis factor (TNF), have limited aptitude to develop into memory cells compared with IL-2 producing cells. Hence, their capacity to provide long-term durable protection is rather limited. On the other hand, IFN-γ and TNF are known to synergize in order to more efficiently kill parasites (Bogdan et al. Reference Bogdan, Moll, Solbach and Rollinghoff1990); therefore, multifunctional CD4⁺ T cells that simultaneously produce multiple cytokines are more likely to be involved in resistance to infection. Indeed, the frequency of multifunctional CD4⁺ T cells (IFN-γ ⁺ TNF⁺ IL-2⁺) correlates with the degree of protection following vaccination (Darrah et al. Reference Darrah, Patel, De Luca, Lindsay, Davey, Flynn, Hoff, Andersen, Reed, Morris, Roederer and Seder2007). These data indicate that functional heterogeneity of the Th1 response to Leishmania plays a significant role in resistance to infection.

The Th1/Th2 dichotomy has been questioned in recent times since there is accumulating evidence that early IL-4 responses might not be required to promote susceptibility and there is considerable complexity in the mechanisms responsible for acquired immunity (Sacks and Noben-Trauth, Reference Sacks and Noben-Trauth2002). Resistant C57BL/6 mice produce IL-4 early at the onset of the infection. This increase in IL-4 did not impact the mounting of an unimpaired Th1 response and disease resolution (Scott et al. Reference Scott, Eaton, Gause, di Zhou and Hondowicz1996). In several cases, resistance to infection in BALB/c mice following immunization has not been linked to a strong Th1 response (Uzonna et al. Reference Uzonna, Spath, Beverley and Scott2004b; Kedzierski et al. Reference Kedzierski, Curtis, Doherty, Handman and Kedzierska2008), and in some cases high pre-challenge IFN-γ levels did not correlate with protection (Stober et al. Reference Stober, Lange, Roberts, Alcami and Blackwell2005). A recent report implicated keratinocytes and epidermal cytokine expression as decisive factors in the generation of Th1 immunity (Ehrchen et al. Reference Ehrchen, Roebrock, Foell, Nippe, von Stebut, Weiss, Munck, Viemann, Varga, Muller-Tidow, Schuberth, Roth and Sunderkotter2010). The critical events that influence Th1/Th2 differentiation were thought to occur in the lymph nodes early during infection; however, it was also acknowledged that the skin, as a primary site of infection, could influence the immune response (Sunderkotter et al. Reference Sunderkotter, Kunz, Steinbrink, Meinardus-Hager, Goebeler, Bildau and Sorg1993). During the first few hours of infection, Leishmania induces several cytokines in keratinocytes and the gene expression profile of cells differs in susceptible and resistant mice. In particular, production of IL-4 by epidermal cells can explain the somewhat controversial role this cytokine plays in induction of Th1/Th2 responses. While IL-4 is associated with a Th2 response and susceptibility to leishmaniasis (Himmelrich et al. Reference Himmelrich, Launois, Maillard, Biedermann, Tacchini-Cottier, Locksley, Rocken and Louis2000), it is also able to induce the production of IL-12 by DCs, but only when present early during the infection (Biedermann et al. Reference Biedermann, Zimmermann, Himmelrich, Gumy, Egeter, Sakrauski, Seegmuller, Voigt, Launois, Levine, Wagner, Heeg, Louis and Rocken2001). Therefore, an early transient IL-4 production by keratinocytes is essential for induction of Th1 response against L. major, by acting in a paracrine fashion on DCs, which then produce IL-12 upon migration to the lymph node (Ehrchen et al. Reference Ehrchen, Roebrock, Foell, Nippe, von Stebut, Weiss, Munck, Viemann, Varga, Muller-Tidow, Schuberth, Roth and Sunderkotter2010). It has also been shown that IL-6, a major inflammatory cytokine, plays an important role in Leishmania protection. High levels of IL-6 of keratinocyte origin have been detected in resistant strains, and mice with IL-6 deficiency in the non-haematopoietic compartment display a Th2 skewing and non-healing phenotype (Ehrchen et al. Reference Ehrchen, Roebrock, Foell, Nippe, von Stebut, Weiss, Munck, Viemann, Varga, Muller-Tidow, Schuberth, Roth and Sunderkotter2010).

Susceptibility and resistance to infection are also influenced by T reg cells (CD4⁺CD25⁺), which reside in the skin where they suppress harmful immune responses to infectious agents, counteract inflammatory responses and limit tissue damage (Belkaid and Rouse, Reference Belkaid and Rouse2005). During L. major infection, T reg cells accumulate in the dermis where they suppress the ability of the effector T cells to eliminate parasites. This process has been linked to the production of IL-10 (Belkaid et al. Reference Belkaid, Piccirillo, Mendez, Shevach and Sacks2002a), a cytokine that is also implicated in the maintenance of parasite persistence (Belkaid et al. Reference Belkaid, Hoffmann, Mendez, Kamhawi, Udey, Wynn and Sacks2001). High levels of IL-10 produced by antigen-driven T reg cells lead to lack of vaccine efficacy despite the presence of strong Th1 responses (Stober et al. Reference Stober, Lange, Roberts, Alcami and Blackwell2005). In humans, T reg cells have been found in lesions of CL patients (Campanelli et al. Reference Campanelli, Roselino, Cavassani, Pereira, Mortara, Brodskyn, Goncalves, Belkaid, Barral-Netto, Barral and Silva2006) and have been implicated in immunopathogenesis of the cutaneous infection (Bourreau et al. Reference Bourreau, Ronet, Darcissac, Lise, Sainte Marie, Clity, Tacchini-Cottier, Couppie and Launois2009). It has been demonstrated that CD4⁺CD25⁺Foxp3⁺ T reg cells are involved in a rapid loss of resistance to infection in immune animals following inoculation with a killed parasite vaccine (Okwor et al. Reference Okwor, Liu, Beverley and Uzonna2009). These data clearly point to the important regulatory role that T reg cells play in resistance and susceptibility to CL.

Cytotoxic activity and cytokine production are two major effector functions of CD8⁺ T cells that contribute to the disease outcome in Leishmania infections. The majority of data do not indicate a protective role for CD8⁺ T cells in controlling primary infection (Huber et al. Reference Huber, Timms, Mak, Rollinghoff and Lohoff1998). However, they clearly play a role in resistance to infection by inducing Th1 responses via cytokine production (IFN-γ) or in recall responses to secondary infection (Muller et al. Reference Muller, Kropf, Etges and Louis1993). IFN-γ-producing CD8⁺ T cells are fundamental for the development of a Th1 response and thus contribute to healing in C57BL/6 mice (Belkaid et al. Reference Belkaid, Von Stebut, Mendez, Lira, Caler, Bertholet, Udey and Sacks2002b; Uzonna et al. Reference Uzonna, Joyce and Scott2004a). Besides cytokine production, CD8⁺ T cells are thought to participate in controlling the infection through cytotoxic mechanisms, such as granzyme and perforin production and Fas/FasL pathways (Ruiz and Becker, Reference Ruiz and Becker2007). In human CL, recruitment of CD8⁺ T cells producing granzyme A to the site of infection is associated with tissue damage, despite this being a consequence of anti-parasitic action (Faria et al. Reference Faria, Souza, Duraes, Carvalho, Gollob, Machado and Dutra2009). Also in Leishmania braziliensis-caused CL, CD8⁺ T cells were shown to play a harmful role contributing to disease immunopathology via their cytotoxic activity leading to tissue destruction (da Silva Santos et al. Reference da Silva Santos, Boaventura, Ribeiro Cardoso, Tavares, Lordelo, Noronha, Costa, Borges, de Oliveira, Van Weyenbergh, Barral, Barral-Netto and Brodskyn2013). It is still not known what is the exact route of CD8⁺ T cell activation in leishmaniasis, since the parasites reside in a parasitophorous vacuole inside the host macrophages and it is not clear how these cells present antigen through MHC I (Bertholet et al. Reference Bertholet, Goldszmid, Morrot, Debrabant, Afrin, Collazo-Custodio, Houde, Desjardins, Sher and Sacks2006). The most likely mechanism is cross-presentation, which has been well documented for macrophages and DCs (Rodriguez et al. Reference Rodriguez, Regnault, Kleijmeer, Ricciardi-Castagnoli and Amigorena1999; Houde et al. Reference Houde, Bertholet, Gagnon, Brunet, Goyette, Laplante, Princiotta, Thibault, Sacks and Desjardins2003), a process suggested to occur during Leishmania infection (Bertholet et al. Reference Bertholet, Goldszmid, Morrot, Debrabant, Afrin, Collazo-Custodio, Houde, Desjardins, Sher and Sacks2006) and one which the parasite is also able to block in order to evade immunity (Matheoud et al. Reference Matheoud, Moradin, Bellemare-Pelletier, Shio, Hong, Olivier, Gagnon, Desjardins and Descoteaux2013).

Development of humoral immune responses is often linked to susceptibility to Leishmania infection, and in general antibodies are not considered to be a major factor in resistance to disease. B cell depletion using anti-IgM antibodies enhanced resistance to Leishmania in BALB/c mice (Sacks et al. Reference Sacks, Scott, Asofsky and Sher1984). Administration of IL-7, a B cell stimulant, to BALB/c mice increased B cell numbers and enhanced disease severity (Hoerauf et al. Reference Hoerauf, Solbach, Rollinghoff and Gessner1995). Furthermore, B cell-deficient (μMT) mice lacking B cells through the targeted disruption of the immunoglobulin M locus are more resistant to infection with Leishmania than their wild-type counterparts (Smelt et al. Reference Smelt, Cotterell, Engwerda and Kaye2000). In addition, adoptive transfer of B cells and serum into BALB/c μMT mice has shown that antigen presentation of specific B cells rather than immunoglobulin effector functions are involved in the susceptible phenotype of BALB/c (Ronet et al. Reference Ronet, Voigt, Himmelrich, Doucey, Hauyon-La Torre, Revaz-Breton, Tacchini-Cottier, Bron, Louis and Launois2008). Conversely, μMT mice on both C57BL/6 and BALB/c genetic backgrounds showed no difference in disease phenotype and CD4⁺ T cell polarity when compared with wild-type hosts (Brown and Reiner, Reference Brown and Reiner1999). Recent studies demonstrated that B cells are required for susceptibility and the development of Th2 cell responses in BALB/c mice during L. major infection (Ronet et al. Reference Ronet, Voigt, Himmelrich, Doucey, Hauyon-La Torre, Revaz-Breton, Tacchini-Cottier, Bron, Louis and Launois2008). The ability of B cells to skew the immune response towards a Th2 phenotype was linked to their capacity to present antigen to T cells. In addition, it has been shown that IL-10 produced by B cells may play a role in susceptibility to cutaneous infection by inhibiting IL-12 production by DCs in vitro (Ronet et al. Reference Ronet, Hauyon-La Torre, Revaz-Breton, Mastelic, Tacchini-Cottier, Louis and Launois2010). Thus, the importance of B cell-mediated responses in Leishmania infection is controversial, but evidence points toward the involvement of B cells in disease susceptibility, at least in the mouse model of disease.

Cytokines

As described above, a wide range of cytokines and chemokines are involved in the immune response to Leishmania including, but not limited to, IL-4, IL-10, IL-12, IL-13, TNF and IFN-γ. The profile and timing of cytokine production correlates with the clinical outcome of Leishmania infection. A variety of immune cells express cytokines, mostly CD4⁺ T cells (Th1 and Th2), but also CD8⁺ T cells, CD4⁻CD8⁻ double-negative T cells (Gollob et al. Reference Gollob, Antonelli, Faria, Keesen and Dutra2008), NK cells, DCs and macrophages (Liese et al. Reference Liese, Schleicher and Bogdan2008), mast cells (Maurer et al. Reference Maurer, Lopez Kostka, Siebenhaar, Moelle, Metz, Knop and von Stebut2006) and regulatory B cells (Ronet et al. Reference Ronet, Hauyon-La Torre, Revaz-Breton, Mastelic, Tacchini-Cottier, Louis and Launois2010).

The exemplary Th2 cytokine in leishmaniasis is IL-4, which drives the Th2 response and promotes susceptibility through inhibition of macrophage activation and abrogation of IL-12 expression. The role of IL-4 in susceptibility to Leishmania has been illustrated in studies using transgenic and knockout mice. C57BL/6 IL-4 transgenic mice are more susceptible to infection than wild-type mice. Targeted disruption of the IL-4 gene or depletion of IL-4 in susceptible BALB/c mice renders them more resistant to L. major infection (Kopf et al. Reference Kopf, Brombacher, Kohler, Kienzle, Widmann, Lefrang, Humborg, Ledermann and Solbach1996). Additionally, disruption of the IL-4 receptor on CD4⁺ T cells promotes resistance in BALB/c mice (Radwanska et al. Reference Radwanska, Cutler, Hoving, Magez, Holscher, Bohms, Arendse, Kirsch, Hunig, Alexander, Kaye and Brombacher2007). However, the role of IL-4 as a major factor contributing to the disease susceptibility is controversial. One study indicated that BALB/c IL-4-deficient mice remained susceptible to disease in the absence of this cytokine (Noben-Trauth et al. Reference Noben-Trauth, Kropf and Muller1996), whereas other studies showed that the same mice were resistant to Leishmania infection (Kopf et al. Reference Kopf, Brombacher, Kohler, Kienzle, Widmann, Lefrang, Humborg, Ledermann and Solbach1996; Alexander et al. Reference Alexander, Brombacher, McGachy, McKenzie, Walker and Carter2002). These data question whether cytokines other than IL-4 might affect Th1 development during the infection. More recently, IL-4 has been identified as a negative regulator of chemokine production involved in Th1-type cell recruitment to the site of infection (Lazarski et al. Reference Lazarski, Ford, Katzman, Rosenberg and Fowell2013). Short-term blocking of IL-4 led to changes in Th1-associated chemokine gene expression and correlated with increased accumulation of IFN-γ producers. Therefore, elevated expression of IL-4 is thought to limit anti-microbial Th1 response in the inflamed tissue and prolong pathogen survival.

IL-13 shares a number of characteristics with IL-4 and both share a common signalling pathway through IL-4 receptor alpha (Brombacher, Reference Brombacher2000). IL-13 has been demonstrated to have disease-promoting properties and to act independently of IL-4 (Matthews et al. Reference Matthews, Emson, McKenzie, Jolin, Blackwell and McKenzie2000; Alexander et al. Reference Alexander, Brombacher, McGachy, McKenzie, Walker and Carter2002), indicating that IL-13 and IL-4 effects might be additive. High levels of IL-13 might prevent the onset of Th1 response by inhibiting IL-12 production by macrophages and skew immune responses towards deleterious Th2-type profiles. In L. mexicana-induced disease, studies with IL-13 knockout mice implicated this cytokine in preventing disease resolution by inhibiting IL-12R expression (Alexander et al. Reference Alexander, Brombacher, McGachy, McKenzie, Walker and Carter2002).

IL-10 is a major immunosuppressive cytokine in leishmaniasis, and as already discussed, is essential for parasite persistence (Belkaid et al. Reference Belkaid, Hoffmann, Mendez, Kamhawi, Udey, Wynn and Sacks2001) and can exacerbate infection (Murphy et al. Reference Murphy, Wille, Villegas, Hunter and Farrell2001; Belkaid et al. Reference Belkaid, Piccirillo, Mendez, Shevach and Sacks2002a). It is a potent suppressor of macrophage activation, inhibits DC maturation (O'Garra and Vieira, Reference O'Garra and Vieira2007) and is produced by a plethora of cells of the immune system (Moore et al. Reference Moore, de Waal Malefyt, Coffman and O'Garra2001). The ability of vaccinated mice to down-regulate IL-10 secretion has been linked to protection following inoculation with SIR-2-deficient Leishmania infantum parasites (Silvestre et al. Reference Silvestre, Cordeiro-Da-Silva, Santarem, Vergnes, Sereno and Ouaissi2007) and a phosphomannomutase (PMM) knockout line of L. major (Kedzierski et al. Reference Kedzierski, Curtis, Doherty, Handman and Kedzierska2008). IL-10 knockout mice are highly resistant to L. major, whereas IL-10 transgenic mice on the resistant background become susceptible (Groux et al. Reference Groux, Cottrez, Rouleau, Mauze, Antonenko, Hurst, McNeil, Bigler, Roncarolo and Coffman1999; Kane and Mosser, Reference Kane and Mosser2001). IL-10's crucial role in suppression of the immune response has been demonstrated in L. mexicana and L. amazonensis infections, although effective resolution of infection with these New World species requires neutralization of both IL-4 and IL-10 (Padigel et al. Reference Padigel, Alexander and Farrell2003). It has also been shown that IL-10 differentially influences the quality, magnitude and protective efficacy of Th1 cells depending on the vaccine platform (Darrah et al. Reference Darrah, Hegde, Patel, Lindsay, Chen, Roederer and Seder2010). Interestingly, co-expression of IL-10 and IFN-γ by Th1 CD4⁺ cells prevents pathogen eradication and contributes to chronic infection (Anderson et al. Reference Anderson, Oukka, Kuchroo and Sacks2007). IL-10 secreted by T cells has been shown to affect immune activation early in infection and a lack of T cell-specific IL-10 leads to enhanced protection following vaccination (Schwarz et al. Reference Schwarz, Remer, Nahrendorf, Masic, Siewe, Muller, Roers and Moll2013).

IL-9 also plays a role in Leishmania disease susceptibility. Produced mainly by Th2 clones (Demoulin and Renauld, Reference Demoulin and Renauld1998), induction of IL-9 can be either IL-4 dependent or independent (Kopf et al. Reference Kopf, Le Gros, Bachmann, Lamers, Bluethmann and Kohler1993; Monteyne et al. Reference Monteyne, Renauld, Van Broeck, Dunne, Brombacher and Coutelier1997). During L. major infection, IL-9 synthesis was observed from 4 weeks onward only in susceptible BALB/c, but not in resistant C57BL/6 or DBA mice (Gessner et al. Reference Gessner, Blum and Rollinghoff1993a; Nashed et al. Reference Nashed, Maekawa, Takashima, Zhang, Ishii, Dainichi, Ishikawa, Sakai, Hisaeda and Himeno2000). IL-9 neutralization in BALB/c mice resulted in a diminished Th2 response and a shift towards protective Th1 responses. This led to enhanced effector functions, including increased NO production by macrophages, implicating IL-9 as a susceptibility factor in leishmaniasis (Arendse et al. Reference Arendse, Van Snick and Brombacher2005).

Transforming growth factor-β (TGF-β) is a regulatory cytokine that controls initiation and resolution of inflammatory responses (Li et al. Reference Li, Wan, Sanjabi, Robertson and Flavell2006). Different Leishmania species have been shown to induce TGF-β production from macrophages and release the active form of TGF-β from the latent complex (Mougneau et al. Reference Mougneau, Bihl and Glaichenhaus2011). This cytokine is important for determining susceptibility to experimental leishmaniasis (Barral-Netto et al. Reference Barral-Netto, Barral, Brownell, Skeiky, Ellingsworth, Twardzik and Reed1992), and anti-TGF-β treatment promotes resolution of L. major infection in mice by augmenting NO production (Li et al. Reference Li, Hunter and Farrell1999). Overall, mouse studies have indicated that TGF-β inhibits Th1 responses and leads to increased susceptibility to leishmaniasis. This is achieved by suppression of NO production and inhibition of TNF and IFN-γ.

Recently, increased levels of IL-17 have been detected in patients with CL (Bacellar et al. Reference Bacellar, Faria, Nascimento, Cardoso, Gollob, Dutra, Scott and Carvalho2009). The most prominent role of IL-17 is the induction of pro-inflammatory responses via production of cytokines such as IL-6, TGF-β or TNF. In the absence of IL-10, L. major-infected mice display increased levels of IL-17 and neutrophil infiltration. It has been postulated that IL-17 exacerbates pathology and its production is up-regulated by IFN-γ and controlled by IL-10 (Gonzalez-Lombana et al. Reference Gonzalez-Lombana, Gimblet, Bacellar, Oliveira, Passos, Carvalho, Goldschmidt, Carvalho and Scott2013).

Cytokines with the ability to influence Th1 development, such as IL-12 or IFN-γ, play a protective role in leishmaniasis. IL-12's capacity to redirect early Th2 responses and promote resistance is well documented (Heinzel et al. Reference Heinzel, Rerko, Hatam and Locksley1993). IL-12 promotes resistance through macrophage activation and NO production, and is necessary for the priming of naïve T cells towards the Th1 pathway. Resistant mice depleted of IL-12 through the use of anti-IL-12 antibodies become more susceptible to infection, and administration of IL-12 to susceptible mice promotes resistance to infection (Heinzel et al. Reference Heinzel, Rerko, Hatam and Locksley1993). In addition, genetic disruption of the Il12 gene leads to up-regulation of deleterious IL-4 responses and the establishment of progressive disease (Mattner et al. Reference Mattner, Magram, Ferrante, Launois, Di Padova, Behin, Gately, Louis and Alber1996). It has been suggested that IL-12 is required for optimal proliferation and production of IFN-γ by Th1 cells, both of which are significantly enhanced in the presence of IL-12, or that IL-12 can promote Th1 cell survival (Scott et al. Reference Scott, Artis, Uzonna and Zaph2004). Recent data indicate that the central memory CD4⁺ T cells generated during L. major infection require IL-12 for IFN-γ production and differentiation into Th1-type, and in the absence of IL-12 these cells became IL-4 producers (Pakpour et al. Reference Pakpour, Zaph and Scott2008). The majority of IL-12 is produced by APCs such as macrophages, DCs and neutrophils (von Stebut and Udey, Reference von Stebut and Udey2004); however, L. major has the ability to selectively block IL-12 production by macrophages (Carrera et al. Reference Carrera, Gazzinelli, Badolato, Hieny, Muller, Kuhn and Sacks1996). DCs are the major source of IL-12 in leishmaniasis, which acts in combination with DC-derived IL-1α/β to influence Th1 development and promote resistance to cutaneous infection (Von Stebut et al. Reference Von Stebut, Ehrchen, Belkaid, Kostka, Molle, Knop, Sunderkotter and Udey2003).

Similarly to IL-12 deficiency, during IFN-γ deficiency the immune response will default to a Th2-type profile, leading to susceptibility to L. major (Wang et al. Reference Wang, Reiner, Zheng, Dalton and Locksley1994). NK cells are the primary source of early IFN-γ production (Scharton and Scott, Reference Scharton and Scott1993), which plays a role in rapid development of Th1 response. Nevertheless, these cells are not essential for resistance to the cutaneous infection, since efficient IL-12-dependent IFN-γ production by CD4⁺ T cells has been reported in the absence of NK cells (Satoskar et al. Reference Satoskar, Stamm, Zhang, Satoskar, Okano, Terhorst, David and Wang1999). IFN-γ is a key cytokine that triggers the antileishmanial functions of macrophages via induction of NO production, and can activate macrophages alone or in synergy with TNF or IL-7 (Nacy et al. Reference Nacy, Meierovics, Belosevic and Green1991; Gessner et al. Reference Gessner, Vieth, Will, Schroppel and Rollinghoff1993b). Resistant mice display elevated levels of IFN-γ compared with susceptible mice, while targeted disruption of the IFN-γ gene (Wang et al. Reference Wang, Reiner, Zheng, Dalton and Locksley1994) or disruption of the ligand binding chain of the IFN-γ receptor (Swihart et al. Reference Swihart, Fruth, Messmer, Hug, Behin, Huang, Del Giudice, Aguet and Louis1995) in C57BL/6 mice results in increased susceptibility to Leishmania infection. However, contradictory data on the role of IFN-γ exist as some studies show that administration of IFN-γ to BALB/c mice at the time of infection does not affect the susceptibility of BALB/c mice to leishmaniasis (Sadick et al. Reference Sadick, Heinzel, Holaday, Pu, Dawkins and Locksley1990). Additionally, non-healing lesions in C57BL/6 mice were observed despite a strong Th1 response characterized by high IFN-γ, NO expression and low IL-4 production, but in the presence of higher IL-10 and Foxp3 mRNA expression (Anderson et al. Reference Anderson, Mendez and Sacks2005).

TNF is a pro-inflammatory cytokine produced primarily by activated macrophages, and is also produced by fibroblasts, T and B cells. It mediates resistance by controlling intracellular pathogen replication as well as limiting the duration of the inflammatory response (Havell, Reference Havell1989). Synergizing with IFN-γ, TNF-α activates macrophages to exert iNOS-dependent leishmanicidal activity (Liew et al. Reference Liew, Li and Millott1990). Mice resistant to Leishmania produce high levels of TNF in the draining lymph nodes, whereas susceptible mice produce no or minimal TNF (Titus et al. Reference Titus, Sherry and Cerami1989).

IL-27 is a cytokine produced upon exposure to inflammatory stimuli and is functionally and structurally related to IL-12 (Boulay et al. Reference Boulay, O'Shea and Paul2003). It has been implicated in regulation of T cell functions and IFN-γ production and, as a consequence, in promoting Th1 responses (Chen et al. Reference Chen, Ghilardi, Wang, Baker, Xie, Gurney, Grewal and de Sauvage2000). Resistant mice lacking WSX-1 (a component of the IL-27 receptor) produce increased levels of IL-4 following L. major infection and a delayed Th1 response (Yoshida et al. Reference Yoshida, Hamano, Senaldi, Covey, Faggioni, Mu, Xia, Wakeham, Nishina, Potter, Saris and Mak2001). However, the requirement for IL-27 appears to be transient and important only early in infection since WSX-1 knockout mice are able to control lesion development and resolve infection (Artis et al. Reference Artis, Johnson, Joyce, Saris, Villarino, Hunter and Scott2004). IL-23 is a pro-inflammatory cytokine that also shows homology to IL-12 (Oppmann et al. Reference Oppmann, Lesley, Blom, Timans, Xu, Hunte, Vega, Yu, Wang, Singh, Zonin, Vaisberg, Churakova, Liu, Gorman, Wagner, Zurawski, Liu, Abrams, Moore, Rennick, de Waal-Malefyt, Hannum, Bazan and Kastelein2000). IL-23-deficient mice showed increased susceptibility to bacterial and parasitic infections (Tan et al. Reference Tan, Bealgey, Fang, Gong and Bao2009) and IL-23 is involved in regulation of IFN-γ production (Langrish et al. Reference Langrish, McKenzie, Wilson, de Waal Malefyt, Kastelein and Cua2004). In leishmaniasis, IL-27 and IL-23 might play a complementary protective role with other Th1 cytokines. Human patients with L. major infection and healing CL lesions display elevated levels of IL-27 and IL-23 compared with patients with non-healing lesions (Tolouei et al. Reference Tolouei, Ghaedi, Khamesipour, Akbari, Baghaei, Hasheminia, Narimani and Hejazi2012).

Type I IFNs (IFN-α/β) are pro-inflammatory cytokines that are involved early in L. major infection as regulators of the innate response, NO production and IFN-γ expression (Diefenbach et al. Reference Diefenbach, Schindler, Donhauser, Lorenz, Laskay, MacMicking, Rollinghoff, Gresser and Bogdan1998). Type I IFNs signal through STAT2, and during CL infection reduced STAT2 phosphorylation and increased degradation have been detected due to the activity of parasite proteases (Xin et al. Reference Xin, Li and Soong2008).

Taken together, the vast array of immune cells, chemokines (Oghumu et al. Reference Oghumu, Lezama-Davila, Isaac-Marquez and Satoskar2010) and cytokines involved in the immune response to Leishmania clearly highlights the complexity of the disease. The murine model of CL, which mimics many aspects of the human disease, has been used to dissect the role of cytokines and T helper cell responses. In human CL a clear dichotomy in T cell responses has not been reported, with patients displaying mixed Th1 and Th2 immune responses (Ajdary et al. Reference Ajdary, Alimohammadian, Eslami, Kemp and Kharazmi2000). Similarly, in human VL there is no strong association between Th1 responses and resistance to disease, with patients showing co-existing Th1- and Th2-type responses (Khalil et al. Reference Khalil, Ayed, Musa, Ibrahim, Mukhtar, Zijlstra, Elhassan, Smith, Kieny, Ghalib, Zicker, Modabber and Elhassan2005). It appears that in humans, the outcome of CL disease is influenced by the balance between the two T cell populations and is further affected by the host genetic factors, inoculum size and parasite strain.

Experimental models of VL: insights into organ-specific immunity

Clinical studies examining the immune response to VL infection are limited by the difficulty in directly accessing infected tissues in patients. Many studies have investigated the systemic response to VL infection by examining circulating peripheral blood mononuclear cells (PBMCs) and serum cytokine levels. A key feature of VL patients is a depressed cell-mediated immune response, as PBMCs taken from VL patients do not proliferate in vitro and do not produce IFN-γ when stimulated with leishmanial antigens (Sacks et al. Reference Sacks, Lal, Shrivastava, Blackwell and Neva1987). This has limited the utility of in vitro approaches; however, the recent development of whole-blood assays to detect cytokine production from infected patient samples could be applied to study immune correlates of disease status (Singh et al. Reference Singh, Gidwani, Kumar, Nylen, Jones, Boelaert, Sacks and Sundar2012). To better understand the immune response to VL, experimental murine models of infection have been developed, as rodents are competent hosts for both L. donovani and L. infantum. These models have provided much insight into the organ-specific immune responses generated in the bone marrow, liver and spleen during VL. In addition, low-dose infection models using the infective metacyclic form of the parasite have been developed, which more accurately reflect the natural route of transmission (Ahmed et al. Reference Ahmed, Colmenares, Soong, Goldsmith-Pestana, Munstermann, Molina and McMahon-Pratt2003).

Cytokine responses during VL

The majority of people infected with visceralizing Leishmania maintain an asymptomatic infection, but the mechanisms that mediate effective control of the disease are relatively unknown. A strong cytokine response is induced during VL and the production of IFN-γ appears crucial for the control of parasites and the development of resistance to infection (Squires et al. Reference Squires, Schreiber, McElrath, Rubin, Anderson and Murray1989). However, multiple cytokines and chemokines are produced in response to VL infection with elevated levels of IFN-γ, TNF, IL-6, IL-8, IL-10, IL-12, IL-15, IL-18, IL-33, IP-10 and MIG observed in the serum of VL patients (Sundar et al. Reference Sundar, Reed, Sharma, Mehrotra and Murray1997; Kurkjian et al. Reference Kurkjian, Mahmutovic, Kellar, Haque, Bern and Secor2006). Thus, active infection is not due to an absence of a pro-inflammatory response, but is associated with the presence of both Th1 and Th2 cytokines. Thus the Th1/Th2 paradigm, which is thought to be important in CL, does not have such a clear role in determining the resistance/susceptibility profiles in human infection or in experimental models of VL (Miralles et al. Reference Miralles, Stoeckle, McDermott, Finkelman and Murray1994). Whilst clinical studies using samples from the peripheral blood of patients are informative, they may not necessarily reflect the events or immune mechanisms occurring in infected visceral organs. Studies with experimental rodent models demonstrate that organ-specific immune responses play a significant role in host defence of VL with defined patterns of tissue tropism and differential responses developing in the liver and spleen (Engwerda and Kaye, Reference Engwerda and Kaye2000).

Adaptive immune system: contributions of B and T cells to VL

B cells are not considered to play a significant protective role during Leishmania infection and have been implicated in exacerbating VL clinical disease (Galvao-Castro et al. Reference Galvao-Castro, Sa Ferreira, Marzochi, Marzochi, Coutinho and Lambert1984; Bohme et al. Reference Bohme, Evans, Miles and Holborow1986). In contrast, T cells are critical for effective antileishmanial host responses. Immunocompromised mice lacking functional T cells, such as nude mice (Stern et al. Reference Stern, Oca, Rubin, Anderson and Murray1988), severe combined immunodeficiency (SCID) mice (Kaye and Bancroft, Reference Kaye and Bancroft1992) and recombinase activating gene (RAG) knockout mice (Alexander et al. Reference Alexander, Kaye and Engwerda2001) all show enhanced susceptibility to L. donovani infection, which can be overcome via reconstitution of T cell populations. Effector CD4⁺ T cells are responsible for the production of cytokines that are critical for the activation of macrophages and the initiation of effective host protective responses. Cytotoxic CD8⁺ T cells play a host protective role, and are required for effective clearance of parasites (Stern et al. Reference Stern, Oca, Rubin, Anderson and Murray1988) and the generation of memory responses (Stager et al. Reference Stager, Smith and Kaye2000). Antigen-specific CD4⁺ and CD8⁺ cells are activated during infection, in both humans (Mary et al. Reference Mary, Auriault, Faugere and Dessein1999) and mice (Joshi et al. Reference Joshi, Rodriguez, Perovic, Cockburn and Stager2009), and are required for optimal host response to infection. Administration of antigen-specific CD8⁺ T cells to L. donovani-infected mice significantly decreased parasite burdens in the liver and spleen (Polley et al. Reference Polley, Stager, Prickett, Maroof, Zubairi, Smith and Kaye2006), and the induction of CD8⁺ T cell responses is being explored as a therapeutic intervention (Maroof et al. Reference Maroof, Brown, Smith, Hodgkinson, Maxwell, Losch, Fritz, Walden, Lacey, Smith, Aebischer and Kaye2012). Interestingly, in an intradermal model of VL the clearance of parasites from the skin correlated with the infiltration and activation of both CD4⁺ and CD8⁺ T cells, analogous to the initiation of inflammatory responses and resolution observed in cutaneous infection (Ahmed et al. Reference Ahmed, Colmenares, Soong, Goldsmith-Pestana, Munstermann, Molina and McMahon-Pratt2003).

VL and the liver: the granuloma response

The hallmark clinical manifestation of VL is a gross enlargement of the abdomen due to splenomegaly and hepatomegaly, and is observed in almost all VL patients. In experimental mouse models, hepatosplenomegaly is also a feature, and is associated with parasite infection of these tissues. Infection of the liver is evident at one week following L. donovani inoculation, peaking at 3–4 weeks post-infection and then resolving with minimal damage to the tissue (Polley et al. Reference Polley, Stager, Prickett, Maroof, Zubairi, Smith and Kaye2006). This acute resolving infection of the liver is associated with initial dominant reactive oxygen intermediate (ROI) and iNOS responses (Murray and Nathan, Reference Murray and Nathan1999). Liver parasite burdens are initially higher in mice deficient in phagocyte oxidase (gp91 phox^−/−) or nitric oxide synthase 2 (NOS2^−/−); however, gp91 phox^−/− are able to resolve infection in the liver, whilst NOS2^−/− sustain a progressive infection (Murray and Cartelli, Reference Murray and Cartelli1983). This indicates that macrophages use both reactive oxygen and nitrogen intermediates in the initial effort to limit L. donovani replication in the liver, with RNIs appearing to play a more critical role in the resolution of liver parasite burdens.

Effective immune responses to VL in the liver are critically dependent on the formation of granuloma structures, which serve to coordinate and deliver cellular and soluble host defence factors to the infected tissue. The granuloma environment produces a focus for antileishmanial immune mechanisms, in terms of activating and sustaining appropriate parasite killing. During human VL, the presence of granulomas in the liver correlates with the ability to control and maintain infection at a sub-clinical level. In experimental models of VL, liver granulomas increase in number and size in the first month, leading to the clearance of parasites and the resolution of infection during the second month of infection (McElrath et al. Reference McElrath, Murray and Cohn1988). Whilst the majority of parasites are cleared from the liver, sterile cure is never achieved, though the liver is resistant to re-infection. The induction of immunosuppression can re-activate infection, which has been observed in the case of HIV patients (Lachaud et al. Reference Lachaud, Bourgeois, Plourde, Leprohon, Bastien and Ouellette2009) or patients receiving immunosuppressive therapies following organ transplant (Antinori et al. Reference Antinori, Cascio, Parravicini, Bianchi and Corbellino2008).

The core of the liver granuloma develops from tissue resident Kupffer cells that are recruited from the sinusoids during the acute phase of the inflammatory response (Beattie et al. Reference Beattie, Peltan, Maroof, Kirby, Brown, Coles, Smith and Kaye2010a). Kupffer cells are the major phagocytic population within the liver and the prime target for Leishmania infection. The generation of antileishmanial responses in the infected Kupffer cell is dependent on granuloma formation to provide the microenvironment for intracellular L. donovani killing (Murray, Reference Murray2001). Infected Kupffer cells fuse with other mononuclear phagocytic cells to form the core of the granuloma, resulting in the secretion of chemokines and the infiltration and recruitment of leucocytes. Monocytes and neutrophils migrate to the liver within the first few days of infection and form a cellular mantle around the infected Kupffer cells in the developing granuloma. Experiments using depleting monoclonal antibodies towards monocytes and neutrophils demonstrate that these cells are essential for parasite killing as the maturation of hepatic granulomas is delayed in their absence (Cervia et al. Reference Cervia, Rosen and Murray1993). The arrival of mononuclear cells leads to the recruitment of CD4⁺ and CD8⁺ T cells, which are essential for intact granuloma responses (Murray et al. Reference Murray, Stern, Welte, Rubin, Carriero and Nathan1987). B cells accumulate in granulomas over time in an antigen-independent manner and engage in long-lasting interactions with T cells (Moore et al. Reference Moore, Beattie, Dalton, Owens, Maroof, Coles and Kaye2012). Interestingly, histological analysis of liver tissue shows that the formation and maturation of granulomas is asynchronous with mature granulomas possessing complete mononuclear cell cuffing observed alongside infected Kupffer cells that have failed to initiate granuloma formation. Mechanisms regarding the differential timing of granuloma formation and the inability of some infected Kupffer cells to induce appropriate host defence responses are not well understood. Upon resolution of infection, empty or sterile granulomas are evident in the mouse model, which then undergo an involution phase, restoring normal liver tissue function (Murray, Reference Murray2001).

Chemokines and chemokine receptors have an important role in the development of protective immune responses in the liver due to their ability to attract Th1 cytokine-producing cells. Increased production of CCL3 (MIP1a), CCL2 (MCP-1) and CXCL10 (IP-10) occurs in the liver early during infection, and these factors are most likely produced by the infected Kupffer cell (Cotterell et al. Reference Cotterell, Engwerda and Kaye1999). The central role of chemokines in granuloma formation is highlighted by experiments demonstrating that administration of CCL2, CCL3 or IP-10 during experimental VL infection results in accelerated granuloma maturation in the liver and reduced parasite burdens (Dey et al. Reference Dey, Majumder, Bhattacharyya Majumdar, Bhattacharjee, Banerjee, Roy and Majumdar2007). Furthermore, mice lacking CCL3 or its receptor CCR5 show enhanced susceptibility to L. donovani infection (Sato et al. Reference Sato, Kuziel, Melby, Reddick, Kostecki, Zhao, Maeda, Ahuja and Ahuja1999). Initial chemokine production and cell recruitment to the granuloma is T cell-independent, but sustained chemokine production and granuloma maturation requires the presence of infiltrating T cells.

Both CD4⁺ and CD8⁺ T cells are critical for granuloma formation, and the increase in CD4⁺ and CD8⁺ T cell numbers in the liver during VL infection may reflect expansion of resident populations as well as recruitment from the spleen (Stanley and Engwerda, Reference Stanley and Engwerda2007). During L. donovani infection T cells undergo high rates of apoptosis (Alexander et al. Reference Alexander, Kaye and Engwerda2001), suggesting that immune responses are continually generated throughout the course of infection rather than being governed by long-lived effector T cell populations. In animals that lack T cells, such as SCID mice, the absence of sustained chemokine production results in a failure of granuloma formation and the uncontrolled growth of parasites in the liver (Engwerda et al. Reference Engwerda, Smelt and Kaye1996). Nude mice, which lack functional T cells, also fail to control parasite growth and T cell transfer can restore resistance (Stern et al. Reference Stern, Oca, Rubin, Anderson and Murray1988). CD8⁺ T cells contribute to the control of liver parasite burdens through their role in granuloma formation (McElrath et al. Reference McElrath, Murray and Cohn1988; Stern et al. Reference Stern, Oca, Rubin, Anderson and Murray1988) and are essential for control in the liver during re-challenge experiments (Murray et al. Reference Murray, Squires, Miralles, Stoeckle, Granger, Granelli-Piperno and Bogdan1992). The activity of CD8⁺ T cells may involve perforin and FasL-dependent lysis of parasitized macrophages as well as the secretion of pro-inflammatory cytokines and chemokines (Tsagozis et al. Reference Tsagozis, Karagouni and Dotsika2003). The dynamics of CD8⁺ effector T cells in the liver during L. donovani infection have been visualized using intra-vital two-photon microscopy and CD8⁺ T cells were observed to accumulate in granulomas in an antigen-specific manner (Beattie et al. Reference Beattie, Peltan, Maroof, Kirby, Brown, Coles, Smith and Kaye2010a). This study also demonstrated that infected Kupffer cells are the main APC for CD8⁺ T cells in the liver, and suggest that a sustained interaction with antigen-specific CD8⁺ T cells may instigate lysis of the infected host cell (Beattie et al. Reference Beattie, Peltan, Maroof, Kirby, Brown, Coles, Smith and Kaye2010a). However, Leishmania parasites have been shown to evade protective immune responses by inducing functional CD8⁺ T cell exhaustion, driving CD8⁺ T cell anergy and cell death during experimental (Joshi et al. Reference Joshi, Rodriguez, Perovic, Cockburn and Stager2009) and human (Gautam et al. Reference Gautam, Kumar, Singh, Singh, Rai, Sacks, Sundar and Nylen2014) VL.

The predominant host protective role of CD4⁺ T cells during VL is the production of cytokines and chemokines that support granuloma formation and parasite killing. Host defence in the liver is critically mediated by pro-inflammatory Th1-type cytokines, including IL-2 (Murray et al. Reference Murray, Stern, Welte, Rubin, Carriero and Nathan1987), IL-12 (Ghalib et al. Reference Ghalib, Whittle, Kubin, Hashim, el-Hassan, Grabstein, Trinchieri and Reed1995), IFN-γ (Squires et al. Reference Squires, Schreiber, McElrath, Rubin, Anderson and Murray1989), TNF (Tumang et al. Reference Tumang, Keogh, Moldawer, Helfgott, Teitelbaum, Hariprashad and Murray1994), lymphotoxin (LT) (Engwerda et al. Reference Engwerda, Ato, Stager, Alexander, Stanley and Kaye2004) and granulocyte/macrophage colony stimulating factor (GM-CSF) (Murray et al. Reference Murray, Cervia, Hariprashad, Taylor, Stoeckle and Hockman1995a). IL-2 is a potent T cell growth factor, which enhances granuloma tissue reactions and parasite clearance during experimental L. donovani infection largely through the induction of IFN-γ (Murray et al. Reference Murray, Miralles, Stoeckle and McDermott1993). Production of IFN-γ by T cells to generate protective responses in the liver is also dependent on IL-12. Control of parasitaemia is lost in the absence of IL-12, and is associated with reduced IFN-γ production and arrested granuloma formation (Murray, Reference Murray1997). IL-12 may also exert antileishmanial effects independently of IFN-γ, as administration of IL-12 to IFN-γ knockout mice still resulted in parasite killing (Taylor and Murray, Reference Taylor and Murray1997). IL-12 is thought to play an important role in the regulation of cellular immune responses in human VL. PBMCs from patients with active VL are unable to produce IFN-γ in response to Leishmania antigens in vitro; however, the addition of IL-12 restored in vitro IFN-γ production (Ghalib et al. Reference Ghalib, Whittle, Kubin, Hashim, el-Hassan, Grabstein, Trinchieri and Reed1995).

During experimental VL, IFN-γ plays a critical role in the early immune responses that induce tissue granuloma formation and effectively control parasite replication. The neutralization of IFN-γ during infection results in poor cellular assembly of granulomas and an increased parasite burden in the liver (Squires et al. Reference Squires, Schreiber, McElrath, Rubin, Anderson and Murray1989). Impaired granuloma formation was also observed in mice deficient in IFN-γ and was associated with an inability of infected Kupffer cells to recruit monocytes and T cells to the liver (Taylor and Murray, Reference Taylor and Murray1997). Therapeutic administration of IFN-γ can activate macrophages in vivo, but requires the presence of T cells for antileishmanial activity (Murray et al. Reference Murray, Hariprashad, Aguero, Arakawa and Yeganegi1995b). Experiments in mice showed that administration of IFN-γ increased the efficacy of antimony chemotherapy (Murray, Reference Murray1990), and IFN-γ has been used as an adjunct therapy for severe or refractory cases of clinical VL (Badaro and Johnson, Reference Badaro and Johnson1993). Whilst IFN-γ plays a crucial role in the initiation of the granulomatous response early in infection, mice deficient in IFN-γ are capable of reducing liver parasite burdens in the later stages of infection. An early IFN-γ response leads to induction of IL-12 and expression of TNF, and it appears that the late-developing IFN-γ independent antileishmanial mechanism is mediated by TNF (Taylor and Murray, Reference Taylor and Murray1997).

TNF is essential for the formation and maturation of the hepatic granuloma response (Murray et al. Reference Murray, Jungbluth, Ritter, Montelibano and Marino2000). TNF is required for survival in experimental VL models, as L. donovani infection is only fatal in the mouse model in the absence of TNF (Tumang et al. Reference Tumang, Keogh, Moldawer, Helfgott, Teitelbaum, Hariprashad and Murray1994). Mice lacking TNF succumb to overwhelming infection after 6 weeks, with accelerated parasite growth in the liver, impaired hepatic granuloma formation and an enhanced inflammatory response (Murray et al. Reference Murray, Jungbluth, Ritter, Montelibano and Marino2000). Neutralization of TNF during L. donovani infection promotes parasite persistence in the liver, indicating that TNF is required for hepatic resolution (Tumang et al. Reference Tumang, Keogh, Moldawer, Helfgott, Teitelbaum, Hariprashad and Murray1994). TNF is produced by infected Kupffer cells throughout the time course of infection (Engwerda et al. Reference Engwerda, Smelt and Kaye1996) and is essential for leucocyte recruitment. LTα, a member of the TNF superfamily of cytokines, is also required for the control of parasite growth in the liver. LTα plays a key role in granuloma formation, facilitating the trafficking of lymphocytes from the perivascular areas of the liver to the infected Kupffer cells (Engwerda et al. Reference Engwerda, Ato, Stager, Alexander, Stanley and Kaye2004). The role of LTα in the liver is distinct from that of TNF, as CD4⁺ T cells that express both TNF and LTα are needed for efficient killing of parasites within assembled granulomas. Other members of the TNF superfamily, such as CD95L, also contribute to host protective immune responses during VL (Alexander et al. Reference Alexander, Kaye and Engwerda2001).

Whilst the emphasis on liver immune defence is generally focused on the production of Th1 cytokines, the co-expression of Th2 cytokines may also contribute to host protective responses. For example, the induction of IL-4 is essential for the formation of mature granulomas and for effective parasite killing (Stager et al. Reference Stager, Alexander, Carter, Brombacher and Kaye2003). The suppressive effect of immunoregulatory cytokines may limit inflammatory tissue damage in the liver, but generally these cytokines down-regulate critical antileishmanial responses, particularly those dependent on IFN-γ. The production of TGF-β (Wilson et al. Reference Wilson, Young, Davidson, Mente and McGowan1998), IL-6 (Murray, Reference Murray2008), IL-10 (Murphy et al. Reference Murphy, Wille, Villegas, Hunter and Farrell2001), IL-27 (Rosas et al. Reference Rosas, Satoskar, Roth, Keiser, Barbi, Hunter, de Sauvage and Satoskar2006) and IL-33 (Rostan et al. Reference Rostan, Gangneux, Piquet-Pellorce, Manuel, McKenzie, Guiguen, Samson and Robert-Gangneux2013) impairs effective control of parasite growth in the liver. Mice deficient in IL-6 showed an enhanced ability to control infection with earlier, and more rapid, parasite killing associated with increased levels of circulating IFN-γ and accelerated granuloma formation (Murray, Reference Murray2008). Expression of IL-33, an IL-1 family member, is increased in the liver during human VL and patients have increased IL-33 serum levels. Lower liver parasite burdens were observed in mice in the absence of IL-33 signalling mice, which was associated with a strong induction of IFN- γ and IL-12 (Rostan et al. Reference Rostan, Gangneux, Piquet-Pellorce, Manuel, McKenzie, Guiguen, Samson and Robert-Gangneux2013).

IL-10 is a critical component of the host immune response that inhibits resistance to VL and promotes disease progression. Human VL disease is strongly associated with increased production of IL-10 in a variety of clinical settings and elevated IL-10 levels correlate with the development of pathology (Ghalib et al. Reference Ghalib, Piuvezam, Skeiky, Siddig, Hashim, el-Hassan, Russo and Reed1993). The absence of IL-10 leads to enhanced resistance to experimental VL infections in mice (Murphy et al. Reference Murphy, Wille, Villegas, Hunter and Farrell2001). IL-10 has multiple effects on the immune system and suppresses the production of key cytokines, IL-12 and IFN-γ (Murphy et al. Reference Murphy, Wille, Villegas, Hunter and Farrell2001; Murray et al. Reference Murray, Moreira, Lu, DeVecchio, Matsuhashi, Ma and Heinzel2003). Regulation of cellular immune responses by IL-10 includes the suppression of macrophage activation (Bogdan et al. Reference Bogdan, Vodovotz and Nathan1991) and impaired intracellular killing of Leishmania (Bhattacharyya et al. Reference Bhattacharyya, Ghosh, Jhonson, Bhattacharya and Majumdar2001). While there are multiple cellular sources of IL-10 during VL infection, a population of Th1-like CD4⁺ T cells that make IL-10 has been associated with disease progression (Stager et al. Reference Stager, Maroof, Zubairi, Sanos, Kopf and Kaye2006). Conventional DCs that make both IL-10 and IL-27 can induce the production of IL-10 from effector Th1-like CD4⁺ T cells and enhance immunopathology (Owens et al. Reference Owens, Beattie, Moore, Brown, Mann, Dalton, Maroof and Kaye2012).

NK cells and NKT cells participate in the early innate immune responses in the liver and contribute to the control of parasitaemia (Kirkpatrick et al. Reference Kirkpatrick, Farrell, Warner and Denner1985; Svensson et al. Reference Svensson, Zubairi, Maroof, Kazi, Taniguchi and Kaye2005). CD1d-dependent activation of NKT cells occurs during L. donovani infection and these cells also respond with a rapid production of IFN-γ. CD1d-deficient mice show an increased susceptibility to parasitism (Amprey et al. Reference Amprey, Im, Turco, Murray, Illarionov, Besra, Porcelli and Spath2004). During infection, Kupffer cells can activate invariant NKT (iNKT) cells by engagement of CD47 (Beattie et al. Reference Beattie, Svensson, Bune, Brown, Maroof, Zubairi, Smith and Kaye2010b) and iNKT cells are essential for regulating chemokines, such as CXCL10 (Svensson et al. Reference Svensson, Zubairi, Maroof, Kazi, Taniguchi and Kaye2005). There is increasing interest in the development of therapies that enhance iNKT cell function during visceral infection to direct early immunity towards the establishment of a protective host response. However, activation of iNKT cells using the glycolipid antigen α-galactosylceramide (α-GalCer) prior to infection did not show any beneficial effect during VL. Moreover, α-GalCer administered during an established VL infection exacerbated hepatic disease, and was associated with a decrease in IFN-γ-producing CD8⁺ T cells (Stanley et al. Reference Stanley, Zhou, Amante, Randall, Haque, Pellicci, Hill, Smyth, Godfrey and Engwerda2008b). Thus, it appears that iNKT cell activation by α-GalCer hinders disease resolution in the liver, and therapies aimed at modulating NKT cell function may be of limited clinical use.

VL and the spleen: suppression and susceptibility

The spleen is a major organ for the induction of immune responses to infection, and also a site for the killing of parasites during VL. However, prevalent clinical features of human VL include splenomegaly and suppression of antigen-specific immune responses (Zijlstra and el-Hassan, Reference Zijlstra and el-Hassan2001). This immunopathology is recapitulated in experimental murine models where splenomegaly is associated with the persistence of parasites and remodelling of the lymphoid tissue (Polley et al. Reference Polley, Zubairi and Kaye2005). The kinetics of experimental VL infection display a distinct organ-specific pattern: the liver displays an acute resolving infection, attributed to effective granuloma tissue responses, while VL parasites persist in the spleen resulting in a chronic, unresolved state of infection.

The spleen is a highly organized secondary lymphoid organ, consisting of a specialized marginal zone (MZ), which separates the red pulp and white pulp region. The macrophages in the MZ, the marginal metallophilic macrophages (MMMs) and the MZ macrophages (MZMs), are the main phagocytic cell populations responsible for the clearance of parasites during experimental L. donovani infection. The antileishmanial activity of these specialized splenic macrophages is dependent on interferon regulatory factor-7 (IRF-7) (Phillips et al. Reference Phillips, Svensson, Aziz, Maroof, Brown, Beattie, Signoret and Kaye2010).

Acute immune responses generated in the spleen play a key role in the control of L. donovani parasites in the liver during the early phase of infection. The spleen is an important site for DC priming, and DCs are the critical source of early IL-12 following VL infection (Gorak et al. Reference Gorak, Engwerda and Kaye1998). A transient and rapid burst in IL-12 has been observed as early as 5 h post-infection (Ato et al. Reference Ato, Maroof, Zubairi, Nakano, Kakiuchi and Kaye2006) and is a crucial event for the generation of effective antiparasitic immunity (Engwerda et al. Reference Engwerda, Murphy, Cotterell, Smelt and Kaye1998). Vascular cell adhesion molecule-1 (VCAM -1) and its ligand very late antigen-4 (VLA -4) are involved in the initiation of early IL-12 secretion from DCs. Blockade of VCAM-1 or VLA-4 suppressed the production of IL-12 by splenic DCs and reduced parasite-specific T cell responses in the spleen. This was also associated with lower levels of IFN-γ, TNF and NO production in the liver and significantly higher liver parasite burdens (Stanley et al. Reference Stanley, Dalton, Rossotti, MacDonald, Zhou, Rivera, Schroder, Maroof, Hill, Kaye and Engwerda2008a). Migratory DCs may directly phagocytose parasites; however, it is most likely that splenic DCs acquire antigen and are activated by infected macrophages in the MZ. Upon activation DCs migrate to the T cell areas in the peri-arteriolar lymphoid sheets (PALS) and IL-12-producing DCs are observed in the T cell area of the spleen during VL infection. The production of IL-12 by DCs is essential for the activation of effector T cell populations, and the total CD4⁺ T cell population in the spleen is expanded during experimental infection (Polley et al. Reference Polley, Zubairi and Kaye2005). T cells are the dominant leucocyte population in the spleen of VL patients, as compared with normal healthy control aspirates that show a predominance of B cells (Nylen et al. Reference Nylen, Maurya, Eidsmo, Manandhar, Sundar and Sacks2007).

Chemokine-dependent encounters between DCs and T cells in the spleen are crucial for effective responses to L. donovani infection. Mice deficient in CCL19 and CCL21 show impaired DC migration in the spleen and a decreased production of IL-12 during L. donovani infection. These defects in early DC activation in the spleen were associated with reduced migration of effector T cells to the liver and impaired granuloma formation (Ato et al. Reference Ato, Maroof, Zubairi, Nakano, Kakiuchi and Kaye2006). Exogenous administration of IP-10 restores T cell proliferative capacity, leading to decreased parasite burdens in the liver and spleen. IP-10 treatment during experimental VL induced strong expression of iNOS2 and mediated parasitic killing through increased NO synthesis (Gupta et al. Reference Gupta, Bhattacharjee, Bhattacharyya, Bhattacharya, Adhikari, Mukherjee, Bhattacharyya Majumdar and Majumdar2009). Together the data demonstrate the importance of chemokines in promoting early DC and CD4⁺ T cell interactions in the spleen and inducing protective immunity against L. donovani.

Neutrophils may also play a protective role in the acute response in the spleen, as the absence of neutrophils results in a decrease in IFN-γ-producing CD4⁺ and CD8⁺ T cells and an enhanced parasite burden in the spleen. This antileishmanial effect appeared to be specific to the spleen as the absence of neutrophils had only minor effects on parasite growth in the liver (McFarlane et al. Reference McFarlane, Perez, Charmoy, Allenbach, Carter, Alexander and Tacchini-Cottier2008). Neutrophils do not appear to play a significant role in the chronic stage of infection as long-term administration of neutrophil-depleting antibody does not significantly increase parasite burdens in either organ (Rousseau et al. Reference Rousseau, Demartino, Ferrua, Michiels, Anjuere, Fragaki, Le Fichoux and Kubar2001).

During experimental L. donovani infection in mice, no resolution of infection occurs in the spleen and animals maintain chronic parasite burdens in this tissue. There is evidence of profound immune dysfunction in the spleen with an impairment of antigen-specific T cell responses, increased T cell apoptosis (Alexander et al. Reference Alexander, Kaye and Engwerda2001) and the production of regulatory cytokines, such as IL-10 (Stager et al. Reference Stager, Maroof, Zubairi, Sanos, Kopf and Kaye2006) and TGF-β (Wilson et al. Reference Wilson, Recker, Rodriguez, Young, Burnell, Streit and Kline2002). NK cells are negative regulators of cell-mediated immunity in the spleen and show enhanced secretion of IL-10 in the chronic phase of infection (Maroof et al. Reference Maroof, Beattie, Zubairi, Svensson, Stager and Kaye2008). Marginal zone B cells in the spleen have been shown to suppress antigen-specific CD8 and CD4T cell responses during the early stages of VL (Bankoti et al. Reference Bankoti, Gupta, Levchenko and Stager2012). B cells also suppress NK cells and inhibit the generation of effector memory CD8T cells after L. donovani infection (Bankoti et al. Reference Bankoti, Gupta, Levchenko and Stager2012).

Whilst the production of TNF is crucial for the induction and maintenance of host protective responses in the liver, TNF is a key mediator of pathology in the chronically infected spleen. During the later stages of VL, high numbers of TNF-producing cells are present in the spleen and TNF production is observed in both the red and white pulp regions (Engwerda et al. Reference Engwerda, Murphy, Cotterell, Smelt and Kaye1998). TNF is the principal cytokine responsible for the breakdown of splenic architecture following experimental L. donovani infection, contributing to remodelling of the MZ (Engwerda et al. Reference Engwerda, Ato, Cotterell, Mynott, Tschannerl, Gorak-Stolinska and Kaye2002) and the loss of stromal cells from the PALS (Ato et al. Reference Ato, Stager, Engwerda and Kaye2002). Infection-induced re-modelling of the MZ is associated with a dramatic and rapid loss of MZMs, whilst MMMs undergo repositioning within the sinus. In mice lacking TNF or mice treated with TNF neutralizing monoclonal antibodies, MZMs were preserved, indicating that the loss of MZMs is a TNF-dependent process (Engwerda et al. Reference Engwerda, Ato, Cotterell, Mynott, Tschannerl, Gorak-Stolinska and Kaye2002). Evidence for the role of TNF in disease pathogenesis in human VL arises from studies of TNF polymorphisms. A study in northern Brazil examined polymorphisms in the TNFA promoter (TNF1 and TNF2 alleles) in neighbourhoods with ongoing transmission. The presence of the TNF2 allele was more frequent in individuals with progressive disease, whilst the TNF1 allele was associated with asymptomatic infection. The presence of the TNF2 susceptibility allele was associated with higher levels of serum TNF as compared with the TNF1 allele, suggesting that increased TNF is involved in the progression of human VL (Karplus et al. Reference Karplus, Jeronimo, Chang, Helms, Burns, Murray, Mitchell, Pugh, Braz, Bezerra and Wilson2002).

The activation of B cells is a key clinical indicator of VL infection, with patients displaying polyclonal hypergammaglobulinaemia (Ghose et al. Reference Ghose, Haldar, Pal, Mishra and Mishra1980), polyclonal B cell activation and increased circulating immune complexes. The role of immunoglobulins during VL is controversial, as large amounts of immunoglobulins to both parasite-specific and non-specific antigens are produced during infection, including auto-antibodies (Galvao-Castro et al. Reference Galvao-Castro, Sa Ferreira, Marzochi, Marzochi, Coutinho and Lambert1984). These immunoglobulins are not thought to be protective as elevated levels of total antibody correlate with disease pathology (Anam et al. Reference Anam, Afrin, Banerjee, Pramanik, Guha, Goswami, Saha and Ali1999) and have been implicated in the development of anaemia (Pontes De Carvalho et al. Reference Pontes De Carvalho, Badaro, Carvalho, Lannes-Vieira, Vinhaes, Orge, Marsochi and Galvao-Castro1986) and autoimmunity (Galvao-Castro et al. Reference Galvao-Castro, Sa Ferreira, Marzochi, Marzochi, Coutinho and Lambert1984). Experimental models of VL using B cell-deficient mice have demonstrated that B cells are not required for the control of parasite burdens. Additionally the re-constitution of mice with immunoglobulin leads to disease exacerbation through complement activation and signalling (Deak et al. Reference Deak, Jayakumar, Cho, Goldsmith-Pestana, Dondji, Lambris and McMahon-Pratt2010). However, B cells may have some regulatory role to play in suppressing immunopathology, as the absence of B cells leads to sustained neutrophil-mediated pathology of the liver (Smelt et al. Reference Smelt, Cotterell, Engwerda and Kaye2000).

Follicular DCs (FDCs), a resident stromal cell population, play a key role in the organization of lymphoid follicles in the spleen and facilitate the germinal centre (GC) reaction. FDCs are involved in B cell activation, proliferation and maturation through presentation of antigen and production of regulatory signals such as chemokines. During the chronic stage of L. donovani infection, the FDC network is destroyed and there is a concomitant loss of GC (Smelt et al. Reference Smelt, Engwerda, McCrossen and Kaye1997). The complete absence of FDCs is associated with the infiltration of heavily parasitized macrophages into the splenic white pulp regions. It has been hypothesized that the B cell function may become dysregulated in the absence of FDCs and thus the loss of FDCs may contribute to the hypergammaglobulinaemia observed during VL.

Impaired DC migration plays a major role in the pathogenesis of VL and alterations to stromal cell populations directly contribute to immunosuppression during the chronic stage of L. donovani infection. Splenic DCs increase in number during the chronic phase of infection but fail to migrate from the MZ to the PALS. This impaired migration is due to a disruption in the fibroblastic reticular cell (FRC) network that guides T cell and DC migration in the T cell zone of the spleen. The changes to the splenic FRC network are due to a TNF-dependent loss of podoplanin (gp38)⁺ stromal cells (Ato et al. Reference Ato, Stager, Engwerda and Kaye2002). Down-regulation of CCR7 from the DC cell surface also impairs DC migration in the spleen during VL. TNF is also implicated in this process, as enhanced levels of TNF increase IL-10 production, and IL-10 directly induces the loss of CCR7 expression on the DC surface (Ato et al. Reference Ato, Stager, Engwerda and Kaye2002). A potential therapeutic role for DCs has been proposed, as adoptive transfer experiments show that administration of in vitro activated DCs can reduce parasite burdens in the spleen. The efficacy of DC therapy relies on both IL-12 and IL-6, with IL-6 thought to suppress the expansion of IL-10-producing T cells (Stager et al. Reference Stager, Maroof, Zubairi, Sanos, Kopf and Kaye2006). However, recent studies demonstrate that some populations of DCs may contribute to splenic pathology, as targeted deletion of DCs during the established phase of infection improved disease resolution (Owens et al. Reference Owens, Beattie, Moore, Brown, Mann, Dalton, Maroof and Kaye2012).

Interventions that preserve splenic structure during VL have been shown to improve the host response to chemotherapy by enhancing parasite killing. Treatment of experimental VL with receptor tyrosine kinase inhibitors reduced splenomegaly, prevented vascular remodelling and restored the integrity of the microarchitecture of the spleen. Importantly, the maintenance of splenic architecture during infection improved the host response to drug treatment, with a 10-fold reduction in the amount of antimony required to clear infection (Dalton et al. Reference Dalton, Maroof, Owens, Narang, Johnson, Brown, Rosenquist, Beattie, Coles and Kaye2010).