Correspondence: Rosa Milagros Corrales, Département Sociétés et Santé, UR016 Caractérisation et contrôle des populations de vecteurs, Institut de Recherche pour le Développement, 911 Av. Agropolis, 34394 Montpellier, France. Tel./fax: +33 4 67 41 61 47; e-mail: email@example.com
Parasitic protozoa of the genus Leishmania are the causative agents of leishmaniasis. Survival and transmission of these parasites in their different hosts require membrane-bound or extracellular factors to interact with and modify their host environments. Over the last decade, several approaches have been applied to study all the extracellular proteins exported by an organism at a particular time or stage in its life cycle and under defined conditions, collectively termed the secretome or the exoproteome. In this review, we focus on emerging data shedding light on the secretion mechanisms involved in the production of the Leishmania exoproteome. We also describe other methodologies currently available that could be used to analyse the Leishmania exoproteome. Understanding the complexity of the Leishmania exoproteome is a key component to elucidating the mechanisms used by these parasites for exporting proteins to the extracellular space during its life cycle. Given the importance of extracellular factors, a detailed knowledge of the Leishmania exoproteome may provide novel targets for rational drug design and/or a source of antigens for vaccine development.
Trypanosomatid parasitic protozoa of the genus Leishmania are the causative agents of leishmaniasis and are transmitted by female phlebotomine sand flies. Leishmania infections can range from mild self-healing skin lesions to fatal visceral infections depending on the Leishmania species involved (McMahon-Pratt & Alexander, 2004). Currently, it is estimated that 12 million people worldwide are affected and 2 million new cases are believed to occur each year in large areas of the tropics, subtropics, and the Mediterranean basin (Stuart et al., 2008). Control of leishmaniasis has been hampered by the absence of a vaccine, limited efficacy of frontline drugs, and increased transmission as a result of coinfections with HIV (Croft et al., 2006). Therefore, a detailed understanding of all aspects of Leishmania biology is desirable to help formulate new antiparasitic strategies.
Leishmania are highly divergent protozoa with several unusual biological features, including the editing of mitochondrial transcripts and post-transcriptional regulation of gene expression (Simpson et al., 2006). To complete their life cycle, Leishmania spp. need to adapt and develop in an insect vector and a vertebrate host. In the sandfly, Leishmania replicate as motile flagellated extracellular cells known as promastigotes, which reside primarily in the insect alimentary tract. Two main forms can be distinguished: (1) multiplicative, but not mammalian-infective, procyclic promastigotes that are present in the insect midgut, and (2) nondividing, but mammalian-infective, metacyclic promastigotes found in the thoracic midgut and proboscis of the sandfly (Bates, 2007). When inoculated into a mammalian host through a sandfly bite, the metacyclic promastigotes are phagocytosed by macrophages and differentiate into intracellular aflagellate amastigotes. This amastigote form resides within a vacuole with lysosomal features that is termed the parasitophorous vacuole (Kima, 2007). Amastigotes proliferate by binary cell division and can spread to other macrophages, as well as some other phagocytic and nonprofessional phagocytic cells (e.g. dendritic cells and fibroblasts) reviewed in Kima (2007).
The different morphological forms of Leishmania represent an adaptation to the changing environmental conditions encountered by the parasites within their two hosts. These parasites require various effectors that are membrane bound or released extracellularly to ensure their survival and transmission in the different hosts. Promastigote surface coat constituents have been the focus of numerous studies, especially the lypophosphoglycans and the promastigote surface protease named glycoprotein 63 (gp63), both of which are involved in the parasite's virulence in the insect and the mammalian host (Lodge & Descoteaux, 2005; Kulkarni et al., 2006; Santos et al., 2006). Extracellular components produced by Leishmania parasites have been mainly studied for their antigenic properties. Indeed, ‘exogenous antigens’ have proved to be highly immunogenic, eliciting strong immunity and protection against infection in mice and dogs (Tonui et al., 2004; Lemesre et al., 2007). Furthermore, an immunogenic component of Leishmania major culture supernatant, the thiol-specific antioxidant (Webb et al., 1998), is one of the components of the recombinant vaccine Leish-111f for leishmaniasis in human clinical trials (Coler et al., 2007). Thus, extracellular factors represent a source of antigens for vaccine development as demonstrated in Mycobacterium tuberculosis (Horwitz et al., 2005), where, similar to leishmaniasis, cell-mediated immunity plays a pivotal role in the control of the infection (Reece & Kaufmann, 2008).
Despite the potential importance of Leishmania extracellular components in these host–parasite interactions, including the regulation of host immune responses, only a few extracellular proteins have been fully characterized. Moreover, little is known about the mechanism(s) predominantly used by Leishmania to export proteins to the extracellular environment during its life cycle.
Over the last decade, several approaches have been used to study proteins exported by an organism at a given time or under certain conditions. Collectively, these extracellular proteins are termed the secretome or the exoproteome (Hathout, 2007). To fully understand the functionality of the Leishmania exoproteome, it is important to characterize the individual proteins, their expression, and their role(s) in the different morphological forms. Indeed, the composition of the parasite exoproteome may be variable, depending on the parasite stage under consideration. It could also differ according to the relative contribution of different pathways operating in protein secretion. As these parasites are highly divergent eukaryotes, understanding these processes, some of which may represent adaptations to a parasitic lifestyle, may provide new insights into the unusual biology of trypanosomatids. Thus, Leishmania exoproteome studies may represent a promising means to identify novel targets for rational drug design and/or a source of antigens for vaccine development. This review will focus on the mechanisms involved in producing the Leishmania exoproteome, as well as the methodologies currently available and those that could be used for its study. The importance of extracellular components in host–parasite interactions is also discussed.
Origins of the secretome
The term ‘secretome’ was first introduced in a study dealing with a genome-based global survey of proteins secreted by Bacillus subtilis (Tjalsma et al., 2000). The authors defined the secretome as a subset of the proteome that included secreted proteins and the components of the cellular machinery used for protein secretion. Using computational methods, the authors predicted all the proteins B. subtilis exports by searching for signal peptides and cell retention signals in the protein sequences. Further characterization of the B. subtilis secretome using a proteomic approach showed that the original prediction correctly identified about 50% of the proteins (Antelmann et al., 2001). These studies, together with the availability of the complete genome sequences of several organisms, opened the door for the identification and analysis of the secretome both in prokaryotes and in eukaryotes.
Different approaches and definitions have been used to identify the secretome in different organisms. In eukaryotes, the term secretome has been used to describe different subsets of the proteome, including (1) all the proteins processed through the secretory pathway (Klee, 2008), (2) the proteins processed through the secretory pathway that lack transmembrane domains and/or a glycosylphosphatidylinositol anchor signal (Grimmond et al., 2003; Lee et al., 2003), or (3) the subset of proteins identified in the extracellular proteome (Zwickl et al., 2005; Chevallet et al., 2007; Paper et al., 2007). From a proteomic perspective, the mammalian secretome was defined as the quantitative map for the distribution of all proteins and lipids in the classical secretory pathway (Simpson et al., 2007). These studies reveal that the term ‘secretome’ has been used (or misused) in a variety of ways.
There is a conspicuous lack of studies involving the secretome of protozoan parasites, despite the importance of its role in parasite virulence and modulation of the host immune response.
So far, the most specific definition of secretome refers to the malaria agent Plasmodium falciparum. In this parasite, the secretome describes all the proteins exported into the host erythrocyte mediated by an endoplasmic reticulum (ER)-type signal sequence and a downstream host targeting motif or plasmodium export element (Hiller et al., 2004; van Ooij et al., 2008).
For trypanosomatids, the term ‘secretome’ was introduced recently in a proteomic approach to identify a large number of extracellular proteins in a culture medium conditioned by Leishmania donovani (Silverman et al., 2008). Likewise, a classical proteomic strategy was used to define the secretome as the naturally ‘excreted/secreted’ factors of Trypanosoma congolense and Trypanosoma evansi (Holzmuller et al., 2008; Grebaut et al., 2009). In the last two decades, several studies involving trypanosomatids have attempted to identify and characterize ‘excreted/secreted factors’ owing to their potential for vaccine development and/or drug targets. It is well known that Trypanosoma cruzi and Leishmania spp. not only secrete proteins into the extracellular space but also release several factors in vitro (Jazin et al., 1991; Yokoyama-Yasunaka et al., 1994; Jaffe & Dwyer, 2003). For example, members of the proteophosphoglycan family, structurally related to the cell surface lipophosphoglycans, and members of the metalloprotease gp63, are released in vitro into the extracellular medium of Leishmania cultures (Foth et al., 2002; Jaffe & Dwyer, 2003). However, the mechanism(s) by which these and other factors are released by Leishmania spp. are largely unknown. Thus, the term ‘excreted/secreted factors’ was used to include all the molecules found outside the cell, including bona fide secreted proteins. For the purpose of this review, we will describe the Leishmania secretome based on the definition of ‘secretion’, meaning the transport of a protein from the inside to the outside of a cell through any of the secretion mechanisms known to date for eukaryotes. Therefore, the secretome will refer to proteins that are actively secreted from the cell using a classical or a nonclassical mechanism or to secretions mediated by exosomes. The term ‘exoproteome’ will be defined as the set of Leishmania proteins present in the extracellular space, and thus includes nonsecreted proteins (e.g. extracellular proteins originating from cell lysis) and those secreted actively (the secretome). Given the complexity of the composition of the extracellular Leishmania proteome, we suggest that the use of ‘exoproteome’ is a more suitable term to avoid the misnomer of secretion and secretome.
Importance of components of the Leishmania exoproteome in the host–parasite interaction
During the Leishmania life cycle, parasite survival within the different environments provided by their hosts requires several strategies to block host defence mechanisms. In the mammalian host, amastigote survival within infected cells depends on the outcome of the parasite's interaction with the host cell (reviewed in Kima, 2007). Within infected macrophages, amastigotes reside in mature phagolysosomes, resisting acidic pH conditions and attack by lysosomal enzymes. Secreted proteases belonging to the family of cystein- or metalloproteases are generally thought to be involved in the manipulation of host immune responses in both the invertebrate and the vertebrate hosts (Mottram et al., 2004; Olivier et al., 2005; Santarem et al., 2007). Likewise, Leishmania amastigotes have complex nutritional requirements that must be scavenged from the host cell (reviewed in Naderer & McConville, 2008). Leishmania, like other trypanosomatids, are purine auxotrophs (Ortiz et al., 2007). To satisfy its essential purine requirements, Leishmania secretes a nuclease that may function externally of the parasite to hydrolyse and access host-derived nucleic acids (Joshi & Dwyer, 2007). Furthermore, there is increasing evidence that some amastigote proteins can be exported to the host cytoplasm and directly modulate host signalling pathways (Mottram et al., 2004; Naderer & McConville, 2008). Nevertheless, the mechanisms underlying this process are poorly defined.
In the insect vector, Leishmania express several virulence factors that may facilitate transmission to and infection of the mammalian host. However, the identity of these factors is still limited. Materials secreted by the parasite play pivotal roles by modifying the environment in their sandfly hosts so as to enhance their transmission success (Rogers et al., 2004). In mature Leishmania infections, the stomodeal valve is forced open and becomes blocked with parasites embedded in promastigote secretory gel (PSG) (Rogers et al., 2004). The permanent opening of the stomodeal valve is essential for colonization of the foregut and transmission by regurgitation (Rogers et al., 2004). The PSG is a viscous mixture of phosphoglycans secreted by the parasites (Ilg et al., 1996). The major component of the PSG is a secreted glycoprotein called filamentous proteophosphoglycan (Rogers & Bates, 2007). Proteophosphoglycans secreted by Leishmania parasites also include the promastigote-secreted acid phosphatase and a nonfilamentous proteophosphoglycan from amastigotes (aPPG) and promastigotes (pPPG) (reviewed in Ilg, 2000). Despite extensive studies on aspects on their biochemistry, structure, and genetics, the precise function of these molecules remains elusive. However, it has been shown that aPPG induces vacuole formation in mammalian macrophages (Peters et al., 1997a) and is an activator of complement via the mannose-binding lectin pathway (Peters et al., 1997b). Recently, a DNA vaccine encoding the N-terminal domain of L. donovani proteophosphoglycan generated an immunoprotective response against visceral leishmaniasis in rodents (Samant et al., 2009).
Secreted chitinase enzyme is involved in the parasite–vector interaction. Recently, chitinase was shown to act as a multifunctional virulence factor benefiting Leishmania mexicana throughout its entire life cycle. Joshi et al. (2005) have shown that secreted chitinase is a virulence factor enhancing parasite intramacrophage survival and cutaneous pathology in mice. Furthermore, using a molecular approach, it was shown that this enzyme not only enhances infection of the mammalian host but also parasite transmission by modifying the environment in the sandfly vector (Rogers et al., 2008). Specifically, the overexpression of chitinase enables transgenic parasites to colonize the anterior midgut of the sandfly more quickly, damage the stomodeal valve, and affect its blood feeding (Rogers et al., 2008).
A myriad of functions have been described for the gp63 from Leishmania spp. including complement resistance and attachment to the host cell (reviewed in Santos et al., 2006). Despite gp63 being one of the best-characterized surface molecules of Leishmania, its precise role in the life cycle of the parasite during colonization of the insect host remains speculative. Biochemical analysis of the extracellular gp63 of Leishmania has revealed two forms of the protein: one is a glycosylphosphatidylinositol-anchored form that can be released from the cell surface and another that is apparently directly secreted (Ellis et al., 2002; McGwire et al., 2002). Within the sandfly vector, the release of extracellular gp63 from promastigotes could enhance the hydrolysis of protein substrates and play a nutritional and/or a protective role (Kulkarni et al., 2006; Santos et al., 2006). Moreover, gp63 may function differently for distinct Leishmania species in their interactions with different invertebrate vector species (Joshi et al., 2002; Hajmova et al., 2004). Taken together, these studies highlight the importance of studying the interaction between the parasite, its vector, and the mammalian host to fully understand the transmission of Leishmania parasites. In this scenario, secreted proteins have been shown to play multifunctional roles in the transmission of the parasites and in the infection of mammalian hosts.
Secretion mechanisms in eukaryotes and their contribution to the Leishmania exoproteome
The classical secretory pathway
In eukaryotes, soluble secretory proteins typically contain N-terminal signal peptides that direct them to the translocation apparatus of the ER. The primary sequence of the signal peptide is not conserved among proteins, but possesses a conserved three-domain structure: a basic ‘N domain’, a hydrophobic ‘H domain’, and a polar ‘C domain’.
The translocation is mediated by the recognition of the N-terminal signal peptide by a signal recognition particle, which carries the nascent peptide chain to the Sec61 translocon complex within the ER membrane, across which transport takes place (Schatz & Dobberstein, 1996). Following translocation to the ER, the signal peptide is usually cleaved from the mature protein by a signal peptidase. This pathway of protein export from eukaryotic cells is known as the classical or the ER/Golgi-dependent secretory pathway (Rothman & Orci, 1992). The secretory pathway is primarily responsible for the distribution of the newly synthesized products among the endomembrane compartments, as well as delivery to the exterior of the cells. Proteins processed in this pathway are commonly referred to as ‘classically secreted proteins’ and are by default transported through the Golgi apparatus and exported by secretory vesicles. Some proteins have specific retention signals that retain them to the ER or the Golgi, or divert them to lysosomes (Nilsson & Warren, 1994). Extracellular proteins are released into the extracellular space by fusion of Golgi-derived secretory vesicles with the plasma membrane.
Several studies involving trypanosomatids have demonstrated that the basic features of their secretory pathways are very similar to other eukaryotes, despite the fact that these organisms represent one of the most divergent eukaryotic lineages (Tobin & Wirth, 1993; Clayton et al., 1995; McConville et al., 2002). However, the polarized delivery of secretory material to the flagellar pocket of trypanosomatid parasites is an unusual characteristic of the secretory pathway (Field et al., 2007).
Although several proteins in Leishmania have been identified as components of the secretion machinery, little is known about the number of proteins exported to the extracellular space through the classical pathway. The bulk of knowledge on Leishmania N-terminal signal peptide-containing proteins is centred on surface proteins, such as the surface metalloprotease gp63, and to a lesser extent on some members of the proteophosphoglycan (Gopfert et al., 1999; Foth et al., 2002; Kulkarni et al., 2006). These proteins can be released constitutively and may thus be constituents of the exoproteome.
A recent screen of an L. major cDNA library with antiserum raised against a promastigote culture supernatant led to the detection of eight proteins that bear a potential signal peptide, of which five corresponded to unknown proteins (Chenik et al., 2006). Using a genomic-based approach, the secretion of three proteins was experimentally confirmed among 13 hypothetical conserved proteins in trypanosomatids bearing an N-terminal signal peptide (Corrales RM, Mathieu-Daudé F, Garcia D, Brenière F & Sereno D, unpublished data). These studies support the notion that the classical secretory pathway is operational in Leishmania parasites for the export of proteins to the extracellular space. Nevertheless, little is known about the contribution of the classical pathway to the Leishmania exoproteome. An exhaustive analysis is required to identify all the proteins involved in the classical pathway and to characterize their function, not only in the different parasite stages but also across different Leishmania species.
Unconventional secretory pathway
Although prevalent, the classical ER–Golgi-dependent pathway is not the only means of protein export in eukaryotes. An increasing number of secreted proteins devoid of a signal peptide have been reported to be exported from eukaryote cells without the help of the ER–Golgi apparatus (reviewed in Nickel & Seedorf, 2008). The main features of nonclassical secretion are as follows: (1) the lack of conventional signal peptides, (2) the exclusion of these proteins from classical secretory organelles, such as the ER and the Golgi, and (3) secretion that is resistant to brefeldin A, a specific inhibitor of classical secretion (Nickel & Seedorf, 2008). Proteins belonging to diverse functional groups were demonstrated to be released independent of ER–Golgi. Unconventional secretory proteins include cytokines, growth factors, or other molecules with important signalling roles in physiological processes, such as inflammation, angiogenesis, cell differentiation, or proliferation (Prudovsky et al., 2008). Diverse mechanisms have been proposed to explain unconventional secretory processes, including lysosomal secretion, plasma membrane shedding, release in exosomes, as well as secretion through transporters residing in the plasma membrane (reviewed in Nickel & Seedorf, 2008).
To date, in trypanosomatids, the only protein characterized using an unconventional secretory pathway is the hydrophilic acylated surface protein B (HASPB) of L. major (Denny et al., 2000). This protein is associated with the outer leaflet of the plasma membrane in the infectious metacyclic promastigote and amastigote stages (Flinn et al., 1994). The protein is synthesized on free ribosomes in the cytoplasm and becomes both myristoylated and palmitoylated at its N-terminal SH4 domains, which is the molecular basis of how HASPB is anchored in the inner leaflet of the plasma membrane (Denny et al., 2000). The SH4 domain of Leishmania HASPB becomes phosphorylated and this modification regulates subcellular localization (Tournaviti et al., 2009). Nevertheless, it is unclear how HASPB is translocated across the membrane. HASPB has been shown to be secreted from mammalian cells, demonstrating a conserved pathway among lower and higher eukaryotes (Stegmayer et al., 2005). However, the sequence encoding the HASPB protein is not conserved among all Leishmania species and is absent from the Leishmania braziliensis genome. Heterologous expression of SH4 domains of Leishmania HASPB induced the reorganization of the plasma membrane to produce highly dynamic nonapoptotic membrane blebs (Tournaviti et al., 2007). Membrane blebs are formed by different mechanisms and have different functions, including cell movement, cytokinesis, cell spreading, and apoptosis (reviewed in Charras, 2008). Interestingly, membrane blebbing is involved in the mechanism of nonclassical secretion of cytokine interleukin-1β (Qu et al., 2007).
Although the mechanisms and molecular components of eukaryotic unconventional secretion are largely unknown, the roles of regulatory components are beginning to emerge (Nickel & Rabouille, 2009). Recently, caspase 1 and the Golgi reassembly stacking protein (GRASP) were shown to play a role in the regulation of unconventional protein secretion. Caspase 1 was shown to function as a general regulator of stress-induced unconventional secretion for several cytokines (Keller et al., 2008). Caspases play a central role in programmed cell death (PCD), but are also involved in cell proliferation and differentiation (Chowdhury et al., 2008). Although the vast majority of the proteins involved in mammalian PCD are not present in Leishmania spp. or related protozoa (Deponte, 2008), these organisms possess caspase orthologues, the metacaspases, which are involved in PCD (Lee et al., 2007; Deponte, 2008). However, metacaspases are likely to play diverse roles in protozoa, for example, the single metacaspase gene of L. major is essential for cell cycle progression (Ambit et al., 2008). Metacaspases of Trypanosoma brucei may be involved in vesicle trafficking, but apparently possess PCD-independent functions (Helms et al., 2006). Further analyses of trypanosomatid metacaspases are necessary to elucidate their roles and to provide insights into their involvement in regulatory networks, which may include nonconventional secretion.
The GRASP protein is essential for an unconventional secretion pathway during Dictyostelium discoideum cellular development (Kinseth et al., 2007) and is also required for the unconventional secretion of the integral plasma membrane α integrin in Drosophila melanogaster during epithelium remodelling (Schotman et al., 2008). Given that the GRASP family is widely conserved in eukaryotes, these unexpected findings raised several questions. Is GRASP involved in unconventional protein secretion in all eukaryotes or does its function vary depending on the cell type? Can the classical secretion pathway regulate Golgi-independent secretion? In order to resolve this important issue, the role of GRASP proteins in other cell types and organisms should be addressed. In this regard, Leishmania may represent an interesting unicellular model to further reveal the function of GRASP proteins in lower eukaryotes. Recently, a quantitative analysis of the L. donovani secretome indicated that it predominantly uses nonclassical secretion pathways to directly export protein (Silverman et al., 2008). It is tempting to suggest that some Leishmania proteins involved in the classical pathway, such as GRASP, may regulate the nonclassical secretion of proteins as demonstrated in D. discoideum. Additional studies might reveal conserved functions for GRASP proteins in lower eukaryotes, including the regulation of nonconventional secretion.
Vesicle-mediated protein export
Another pathway for protein export in eukaryotes is via the release of exosomes into the extracellular environment. Exosomes are small microvesicles, released from cells by the fusion of either multivesicular bodies or secretory lysosomes with the plasma membrane (Simpson et al., 2008). Unlike the constitutive trans-Golgi secretory pathway, exosome secretion arises directly from the endocytic pathway. Originally described as a removal mechanism of cell surface molecules in reticulocytes, several laboratories showed their importance in a range of biological processes, including development, immunology, and cancer. In addition to a common set of membrane and cytosolic molecules, exosomes harbour unique subsets of proteins linked to cell type-associated functions (reviewed in Simpson et al., 2008). In the last few years, proteomic technologies have made it possible to analyse molecular components of exosomes from a variety of cell types and body fluids. While the functional roles of exosomes are only recently becoming clear, future research is likely to indicate their importance as mediators in biological processes. In addition to exosomes, cells can also release other forms of membrane vesicles into the extracellular space: apoptotic blebs, microparticles, and microvesicles (Simpson et al., 2008). These vesicles represent a heterogeneous population of vesicles that bud directly from the plasma membrane. However, given that exosomes, apoptotic blebs, microvesicles, and microparticles exhibit the same membrane topology, it is still unclear whether these vesicles represent discrete entities with specialized functions or merely represent a size continuum of the same entity (Simpson et al., 2008). Recently, the study of the L. donovani exoproteome supported the notion that stationary-phase promastigotes export materials through an exosome-like mechanism (Silverman et al., 2008). Using a stable isotope labelling with amino acids in cells (SILAC) approach, the authors detected several proteins in the exoproteome that were previously reported as being components of exosomes isolated from B-lymphocytes and adipocytes (Silverman et al., 2008). In addition, the authors detected similar contents in exosomes from apoptotic dendritic cells that had previously been defined as apoptotic blebs (Thery et al., 2001). Given that stationary-phase promastigotes contain up to 43% apoptotic cells that are involved in the establishment and maintenance of an infection, Silverman et al. (2008) proposed that proteins released by Leishmania promastigotes through apoptotic blebs were those important for the parasite's virulence. Further studies of Leishmania exosomes may shed light on novel functions for these vesicles, such as the modulation of immune response as has been demonstrated in dendritic cells (Chaput et al., 2006). Finally, exosome studies may explain the presence of proteins in the extracellular space that lack an N-terminal signal peptide, such as histones, ribosomal proteins, and elongation factors (Chenik et al., 2006; Silverman et al., 2008).
Approaches for the study of the exoproteome and its application in Leishmania
Emerging technologies in proteomic research and genome sequencing have considerably accelerated studies of the exoproteome in eukaryotes. Generally, these methods can be categorized into four groups: (1) direct proteomic-based approaches, (2) genome-based computational predictions, (3) genetically based approaches, and (4) immune-based approaches. In the next section we describe the different approaches, their limitations, and applications in the study of the Leishmania exoproteome.
Direct proteomic-based approaches
Recently developed methods in proteomics have considerably facilitated the broad-scale analysis of proteins in biological samples, including whole tissues and culture media (reviewed in Lane, 2005). The massive progress in mass spectrometry (MS), bioinformatics, and analytical techniques has considerably facilitated analyses of the exoproteome, especially the cancer secretome (Latterich et al., 2008). Currently, three major proteomic technologies are used for secretome analyses: gel-based methods, gel-free MS-based methods, and surface-enhanced laser desorption/ionization time-of-flight mass spectrometry. Two-dimensional gel electrophoresis coupling MS is the most classic and well-established proteomic approach. This allows identification of hundreds to thousands of proteins from a single gel. Although this method remains the most efficient for separation of complex protein mixtures, the detection of secreted proteins in cell supernatants represents a challenge for current proteomics techniques. This is for several reasons. Firstly, culture media are rich in salts and other compounds, for example serum supplement, that interfere with the selective precipitation of proteins. Secondly, the nonsecreted proteins released into the culture medium upon lysis of a few dead cells contaminate the so-called secretome. To overcome these technical limitations, a strategy involving the elimination of serum constituents through washing and subsequent incubation of cells in a serum-free medium can be used. This has proved useful in studies on the secretome of different cancer cell lines (Xue et al., 2008) and for the recovery of secreted proteins from Leishmania promastigote culture supernatants after 4–6 h of incubation in serum-free media (Merlen et al., 1999; Chenik et al., 2006; Silverman et al., 2008). Nevertheless, the balance between the incubation time required for the cells to secrete detectable amounts of protein and cell survival in the serum-free medium, to avoid proteins from cell lysis, is a key factor.
Another challenge is the detection of low abundance secreted proteins. Given the low concentration of secreted proteins in culture media, supernatant concentration is necessary before subsequent proteomic analyses. Although incubation in serum-free conditioned media combined with the concentration of the supernatant by ultrafiltration can significantly reduce contamination by nonsecreted proteins, a minimal presence of contaminant proteins may easily mask the proteins of interest, especially when they are in low quantities. Consequently, the discrimination of genuinely secreted proteins from nonsecreted proteins remains the main challenge in proteomic-based technologies.
SILAC culture has emerged as an alternative methodology for accurate identification of genuinely secreted proteins in quantitative proteomics. SILAC involves culturing cells in media containing either the normal amino acid or its isotopically labelled analogue until essentially all proteins of the cell are labelled. The two populations or samples to be compared are then mixed in equal ratios and analysed by nanoflow liquid chromatography-tandem mass spectrometry (reviewed in Gevaert et al., 2008). SILAC-based strategies not only help to profile secreted proteins in mammalian cell cultures but also facilitate the distinction between secreted proteins and contaminant proteins from the culture media. The SILAC methodology was successfully used to analyse the secretome from cultures of stationary-phase promastigotes of L. donovani, by comparing the amount of any given protein secreted into the conditioned medium with the amount of the same protein that remained cell associated. A total of 358 proteins were identified using this approach, and based on quantitative analyses, the authors concluded that 151 were actively secreted by the parasites (Silverman et al., 2008). This methodology may represent a useful tool to characterize the secretome at different life-cycle stages of Leishmania and to study the related medically important parasites T. brucei and T. cruzi. However, the basis for quantitative analysis using SILAC prevents identification of secreted proteins that display very low intracellular concentrations, independent of their concentration in the conditioned medium. This drawback resulted in the nonidentification of some Leishmania proteins previously described as secreted when analysing the L. donovani secretome (Silverman et al., 2008).
Genome-based computational predictions
Computational analyses rely on the prediction of signal peptides, viewed as a hallmark of classically secreted proteins in eukaryotes. The N-terminal signal sequence or signal peptide holds a conserved set of secondary characteristics that are identifiable by computational algorithms. Over the last decade, the development of new signal peptide prediction programs has been prolific and has allowed the identification of potentially secreted proteins from eukaryotes (reviewed in Klee & Sosa, 2007). Nevertheless, these programs use a variety of algorithms and architectures, leading to different predictions. To overcome this limitation, several prediction tools, such as transmembrane domain, glycosylphosphatidylinositol-anchoring signal, and mitochondrial targeting signal predictions, can be combined to improve the in silico prediction of classically secreted proteins. A computational strategy was used to identify the mouse secretome by looking for transcripts encoding proteins with signal sequences, but without transmembrane domains (Grimmond et al., 2003). Lee et al. (2003) used a set of prediction algorithms to define the potential Candida albicans secretome through computational identification of soluble proteins that possessed N-terminal signal sequences and lacked transmembrane domains, glycosylphosphatidylinositol anchor sites, and mitochondrial targeting sequences, from the 6165 ORFs of the yeast genome. A similar multistep computational analysis was used to simultaneously predict the secretomes of zebrafish and humans, and to identify orthologues among the potentially secreted proteins (Klee, 2008). Although genome-based methods can provide a comprehensive list of potentially secreted proteins, the data obtained may contain many false-positive and false-negative candidates. In addition, these approaches make it difficult to discriminate between extracellular proteins and other proteins processed by the ER, but not released into the extracellular space, for example ER and Golgi-resident proteins and lysosomal proteins. Another limitation for computational prediction is the lack of programs to predict proteins secreted through nonclassical pathways. To date, only one program is available for the prediction of mammalian proteins targeted by the nonclassical secretory pathway (Bendtsen et al., 2004). Because of the lack of a conserved targeting signal for unconventional secretion, the development of algorithms for the prediction of nonclassical secreted proteins remains a challenge.
Although sequencing of the genome of different Leishmania species is now completed, only two genome-based studies have performed in silico prediction of potentially secreted proteins in these pathogens. Silverman et al. (2008) analysed the secretome of the L. major genome by searching for protein sequences bearing a signal peptide, lacking a transmembrane domain, and glycosylphosphatidylinositol anchor signal. Using these parameters, the authors found 217 potentially secreted proteins, of which 141 were annotated as hypothetical proteins. Remarkably, only 14 of the predicted secreted proteins were identified experimentally using a proteomics-based approach. In our group, we used a computational approach and confirmed that only 25% of the trypanosomatid proteins predicted to be secreted were detected in the extracellular environment (Corrales RM, Mathieu-Daudé F, Garcia D, Brenière F & Sereno D, unpublished data). Thus, even if computational predictions can provide a list of potentially secreted proteins, experimental validation is needed to confirm whether they are actually secreted.
Genetically based approaches: forward and reverse genetics
In eukaryotes, several forward genetic strategies based on the presence of N-terminal signal peptides in protein sequences have been used to identify proteins involved in the classical secretory pathway. A widely used method to identify secreted and transmembrane proteins is the yeast signal sequence trap (YSST). This method relies on the ability of putative signal sequences to rescue secretion of the yeast invertase enzyme lacking its endogenous signal peptide (Klein et al., 1996; Jacobs et al., 1997). Several YSST approaches have been developed and applied to lower and higher eukaryotes to identify secreted and membrane-associated proteins (Nene & Bishop, 2001; Smyth et al., 2003; Yamane et al., 2005). Nevertheless, the main drawback of YSST is the presence of false-negative proteins bearing a predictive signal sequence, which may be explained by heterologous gene expression that does not reflect the natural environment. However, this technology has led to the discovery of genes encoding secretory proteins that were not predicted to bear signal sequences in Hydra polyps (Bottger et al., 2006). Thus, genes discovered by YSST methods offer a starting point for a more detailed characterization. A morpholino-based gene ‘knockdown’ strategy was used in zebrafish to assess the role of members of the secretome in vertebrate development (Pickart et al., 2006). This study provides a framework utilizing zebrafish for the systematic assignment of biological function in a vertebrate genome. In protozoans, a remarkable reverse genetic approach was recently used to analyse the secretome in P. falciparum (van Ooij et al., 2008). Using the piggyBac transposition system, the authors validated the export of 70% of the predicted secreted proteins (van Ooij et al., 2008). In the piggyBac transposition system, integration and expression of specific DNA sequences is induced by the lepidopteran transposon and a transposase-expressing helper plasmid (Balu et al., 2005). Given that efficient transposon-mediated random mutagenesis has been reported in Leishmania (Beverley et al., 2002), this methodology might represent an interesting tool to analyse the exoproteome of amastigotes within macrophages.
An alternative methodology to identify components of the Leishmania exoproteome is based on the highly immunogenic properties displayed by Leishmania culture supernatants (Chenik et al., 2006). The authors attempted to identify extracellular proteins from L. major by screening a cDNA expression library with sera raised against culture medium supernatants. Using this strategy, 33 proteins were identified, out of which eight had a signal peptide, suggesting their involvement in the classical pathway. Remarkably, about 40% of the identified genes encode unknown proteins, pointing out the interest in characterizing the components of the Leishmania exoproteome. Unfortunately, antibody-based methodologies are biased towards the identification of the most abundant or highly immunogenic proteins, and thus typically identify only a small subset of proteins in the secretome.
Challenges and perspectives
The genome sequences of Leishmania species have provided a new starting point to understand the biology of these medically important parasites. Understanding the relative contribution of parasite genes to the wide spectrum of disease caused by Leishmania species is one of the most important questions to be answered. Given the relatively minor differences among Leishmania species in terms of gene content and RNA gene expression (Lynn & McMaster, 2008), studies of the proteome from different Leishmania species are crucial. In this regard and because of the relevance of parasite extracellular components to its interaction with its environment, the in-depth study of the exoproteome represents an open door for further exploration: (1) the function of the large number of hypothetical proteins awaiting detailed analysis; (2) the dynamics of the exoproteome throughout the Leishmania life cycle; and (3) the main secretion mechanisms used by these parasites to export proteins into the extracellular space, and their significance in different developmental forms. Compared with that of bacteria and higher eukaryotes, our knowledge of the Leishmania exoproteome is at a rudimentary stage. However, the recent study supporting nonclassical secretion pathways as the main mechanism used by L. donovani promastigotes to export proteins may represent the tip of an iceberg ready to emerge. Whether this observation is confirmed for the amastigote form and other Leishmania species may shift our research towards understanding the mechanisms and regulation of unconventional secretion. Roles for GRASP and caspase1 proteins in the regulation of Golgi-independent secretion have now been revealed. Although caspases are missing in protists, GRASP proteins are highly conserved in eukaryotes. Analysing their roles in Leishmania parasites may further shed light on the regulation of nonclassical mechanisms and on the possible cross talk between classical and nonclassical pathways (summarized in Fig. 1). This information will lead to a deeper understanding of how secretion mechanisms are used by trypanosomatids to ensure parasite survival within the different environments encountered throughout its life cycle. Together, detailed knowledge of the Leishmania exoproteome may lead to the development of new antiparasite strategies.
Financial support for this study was provided by the European Union through a grant to RMC by the ‘Programme Alßan’ European Union Programme of High Level Scholarships for Latin America (No. E05D057391AR) and by IRD DSF. We thank P. Agnew for help revising the English in the manuscript.