INTRODUCTION
Protozoan parasites of the Leishmania genus are causative agents of a broad spectrum of diseases generally termed as leishmaniases. Disease manifestations range from cutaneous leishmaniasis caused by Leishmania major, Leishmania aethiopica, Leishmania tropica, Leishmania mexicana and Leishmania amazonensis, to muco-cutaneous leishmaniasis caused by Leishmania braziliensis and Leishmania guyanensis and to potentially deadly visceral leishmaniasis or Kala-azar caused by the Leishmania donovani complex (L. donovani and Leishmania infantum) (Murray et al. Reference Murray, Berman, Davies and Saravia2005). Leishmaniasis is associated with 2·4 million disability-adjusted life years and 70 000 deaths per year (Desjeux, Reference Desjeux2004). Extracellular flagellated promastigotes of Leishmania live and multiply in the mid-gut of the female Phlebotomine sandfly. Promastigotes enter the skin of the vertebrate host by the bite of the sandfly, are phagocytosed by immune cells, and transform into non-motile ovoid-shaped amastigotes that are able to propagate within the harsh environment of the macrophage phagolysosome.
Not all members of Leishmania genus are parasites of mammals. Certain species of Leishmania including Leishmania tarentolae are lizard parasites. Whether these parasites should be included within the genus Leishmania has been a matter of debate. However, molecular phylogenetic studies have suggested that lizard Leishmania, alternatively named Sauroleishmania, have evolved from Old World Leishmania after separation of the New World Vianna subgenus (Croan et al. Reference Croan, Morrison and Ellis1997). L. tarentolae is a lizard parasite strain, which has never been found associated with any leishmaniasis in humans and is thus considered as non-pathogenic to humans. L. tarentolae has been used as a model organism for studying unique features of Kinetoplastids such as RNA editing (Landweber and Gilbert, Reference Landweber and Gilbert1994; Maslov et al. Reference Maslov, Avila, Lake and Simpson1994). However, as Noyes et al. (Reference Noyes, Chance, Croan and Ellis1998) have also indicated previously, comparative studies of lizard Leishmania and virulent Leishmania can be of critical importance towards understanding the pathogenicity of Leishmania and its relationships with the host. Leishmania spp. form complex host-parasite interactions and are able to alter the host immune response for their benefit. In order to explain the absence of pathogenicity in L. tarentolae, we tested whether known virulence factors of pathogenic Leishmania species were present and or expressed in L. tarentolae.
A number of Leishmania's virulence factors such as GP63 (Joshi et al. Reference Joshi, Kelly, Kamhawi, Sacks and McMaster2002; McGwire et al. Reference McGwire, O'Connell, Chang and Engman2002; Yao et al. Reference Yao, Donelson and Wilson2003; Campbell et al. Reference Campbell, Kurath and Fleischmann1992) and cysteine proteases (CPs) (Alexander et al. Reference Alexander, Coombs and Mottram1998; Rosenthal, Reference Rosenthal1999; Mundodi et al. Reference Mundodi, Somanna, Farrell and Gedamu2002; Mottram et al. Reference Mottram, Coombs and Alexander2004) have been characterized and their roles have been extensively studied; some others such as LPG3 (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002; Ouakad et al. Reference Ouakad, Bahi-Jaber, Chenik, Dellagi and Louzir2007b), Amastins (Salotra et al. Reference Salotra, Duncan, Singh, Subba Raju, Sreenivas and Nakhasi2006) and A2 (Zhang and Matlashewski, Reference Zhang and Matlashewski1997, Reference Zhang and Matlashewski2001; Zhang et al. Reference Zhang, Mendez, Ghosh, Myler, Ivens, Clos, Sacks and Matlashewski2003) have also been identified. GP63 (also called Major Surface Protease) is one of the most extensively studied virulence factors of Leishmania. GP63 is a surface endopeptidase with a wide range of substrate and pH specificity, well suited for its dual expression in promastigote and amastigote stages. Targeted gene deletion studies have shown a crucial role for GP63 in the pathogenicity of Leishmania; gp63 nul mutants of L. major showed reduced pathogenicity and were weakened in forming lesions in BALB/C mice in comparison to wild types (Joshi et al. Reference Joshi, Kelly, Kamhawi, Sacks and McMaster2002). As another group of virulence factors of Leishmania, the cysteine proteases have been studied mostly in L. mexicana due to their crucial role in its pathogenicity. Most studies on cysteine proteases in L. mexicana have been done on the most abundant Cysteine Protease families A, B and C. Similar to gp63 −/−L. major, Δcpb L. mexicana have impaired infectivity against BALB/c mice (Mottram et al. Reference Mottram, Brooks and Coombs1998; Alexander et al. Reference Alexander, Coombs and Mottram1998). Furthermore, Williams et al. (Reference Williams, Tetley, Mottram and Coombs2006) have shown that CPB and CPA are essential for parasite autophagy during promastigote-amastigote transformation. The dominant surface molecule of promastigotes is lipophosphoglycan (LPG). This molecule plays important roles in the early stages of the infection, wherein the parasite is still in the promastigote form, namely interaction with the complement system, phagocytosis and protection of promastigotes against lysis in the phagolysosome (Spath et al. Reference Spath, Garraway, Turco and Beverley2003). LPG3, a molecular chaperone, is part of the metabolic pathway of synthesis of LPG as well as phosphoglycan residues that are added to extracellular proteins, such as GP63, GP46, sAP and GPI-anchored proteins (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002). The expression level of surface proteins such as GP63 and GP46 is greatly reduced in lpg3 nul mutants (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002). Based on these findings it was proposed that the role of LPG3 in virulence should be exerted through its role in the synthesis of LPG, GPI-anchored proteins (such as GP63) and other phosphoglycan-bearing (PG) molecules (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002). Amastins belong to a large gene family of surface proteins that are developmentally regulated in the amastigote stage of Leishmania (Wu et al. Reference Wu, El-Fakhry, Sereno, Tamar and Papadopoulou2000). Although the biological function of amastins remains still unknown, it is hypothesized that amastins may play a role in proton or ion traffic across the membrane to adjust the cytoplasmic pH under the harsh conditions of the phagolysosome (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). Like amastins, A2 is an amastigote-specific protein, originally discovered in L. donovani. The exact function of A2 protein in Leishmania is unknown. However, gene silencing of A2 in L. donovani dramatically decreased its viability in mammalian macrophages and deteriorated its pathogenicity (Zhang and Matlashewski, Reference Zhang and Matlashewski1997). Interestingly, over-expression of A2 in L. major resulted in visceralization of the infection (Zhang and Matlashewski, Reference Zhang and Matlashewski1997; Garin et al. Reference Garin, Meneceur, Pratlong, Dedet, Derouin and Lorenzo2005).
In this study, we tested whether well-characterized virulence factors in the pathogenic Leishmania species (e.g. L. major, L. infantum and L. braziliensis) were present and or expressed in the non-pathogenic L. tarentolae species in order to gain a better insight into the factors responsible for the lack of pathogenicity in L. tarentolae. We found that all the well-known virulence factors that we tested here, with the only exception of A2, were present and expressed in L. tarentolae at ratios similar to those of the pathogenic species.
MATERIALS AND METHODS
Preparation of stationary phase promastigotes
L. major (MRHO/IR75/ER), L. infantum (MCAN/98/LLM-877) and L. tarentolae (ATCC 30267) promastigotes were grown at 26°C in complete M199 medium containing 10% heat-inactivated FCS, 40 mm HEPES, 0·1 mm adenosine, 0·5 μg/ml haemin and 50 μg/ml gentamicin, and were allowed to multiply until they reached a density of 2–3×107 following 5 days of growth at 26°C. The stationary phase promastigotes were then utilized for macrophage cells (MQ) infection and RNA extraction.
Probe preparation
The ORF of cysteine protease type I (GenBank Accession no. U43706) was digested using BamHI-HindIII from vector pGEM-cpb (Rafati et al. Reference Rafati, Salmanian, Hashemi, Schaff, Belli and Fasel2001). The ORF of GP63 from L. major (GenBank Accession no. AF039721) was digested using XbaI from a vector kindly provided by Dr Mahboudi (Pasteur Institute of Iran). LPG3 ORF (GeneDB Accession no. LmjF29.0760) was digested using EcoRI-SalI from pGEM-lpg3 vector prepared previously in our lab (unpublished). The A2 ORF (Accession no. S68693) was digested with BamHI from a vector kindly provided by Dr Greg Matlashewski (McGill University, Canada). Partial sequences from amastins Lmj24.1270, Lmj28.1390, Lmj30.0870 were digested by EcoRI from vectors described previously (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). Following digestion and extraction of fragments from agarose gel, probes were prepared by random primer extension using Klenow enzyme (Roche). In certain sections of library screening, amastin probes were prepared by a non-radioactive Digoxigenin (DIG) labelling (Roche) in a similar process.
Total DNA preparation and hybridizations
Total DNA of L. tarentolae (strain ATCC 30267), L. major (strain MRHO/IR75/ER) and L. infantum (MCAN/98/LLM-877) was prepared from log-phase parasites using the standard proteinase K (Roche) and phenol extraction method (Sambrook and Russell, Reference Sambrook and Russell2001) and digested using PstI (Roche). Southern blotting was done according to standard methods (Sambrook and Russell, Reference Sambrook and Russell2001). Resulting Nylon membranes (Roche) were hybridized to 32P-labelled probes in 20 ml of Church buffer (0·5 m phosphate-buffered saline (PBS) containing 1% BSA, 1 mm EDTA and 7% SDS) at 65°C overnight. Non-stringent and stringent washes were done in 25 ml of NaPi buffer for 15 min at 25°C and 65°C, respectively. When necessary, non-stringent washing at 42°C was also performed for confirmation. Radiofilms (Kodak) were exposed overnight or for 2 days for confirmation.
Library screening
One μg of the previously described cosmid library of L. tarentolae (Genest et al. Reference Genest, Haimeur, Légaré, Sereno, Roy, Messier, Papadopoulou and Ouellette2008) was transformed to E. coli DH5α by chemical transformation. Colony lifting was done according to standard methods (Sambrook and Russell, Reference Sambrook and Russell2001) and the resulting nylon membranes were hybridized to the probe for L. major's amastin Lmj24.1270 under non-stringent conditions as stated in the previous section. The transformation and screening was performed 6 times in order to achieve high genome coverage. Cosmids that were hybridized to the amastin probe were digested using PstI and re-blotted against the probe and the hybridizing fragment was subcloned in pGEM-2 vector (Promega) after dephosphorylation (Shrimp Alkaline Phosphatase, Roche). The cloned fragment was sequenced using universal T7 and SP6 primers.
Sequence analysis
Orthologues of the sequences acquired from L. tarentolae were found in the genomes of L. major, L. infantum and L. braziliensis using GeneDB BLAST search tool (for each gene, the hit with the lowest E-value was chosen). Pairwise sequence alignments were performed on translations of the acquired sequences at EBI website. Multiple sequence alignment was performed using web-based CLUSTALW at EBI bioinformatics tools. Transmembrane helices of L. tarentolae's amastin were predicted using TMHMM server version 2 (Krogh et al. Reference Krogh, Larsson, von Heijne and Sonnhammer2001).
Amastigote preparation
Peritoneal exudate cells from naïve BALB/c mice were used as a source of resident macrophages (MQ). Briefly, the peritoneal cavity was washed with 5 ml of cold RPMI 1640 (Sigma) medium, and then the medium containing peritoneal macrophages was recovered. Cells were cultured in RPMI medium 1640 supplemented with 5% FCS, 2 mm glutamine, 50 mm 2-mercaptoethanol, 10 mm HEPES and 40 μg/ml gentamicin at a density of 106 cells/well, and incubated at 37°C in 5% CO2 for 24 h. In order to infect MQ, the stationary phase Leishmania promastigotes were added at the ratio of 10:1 (parasites versus macrophages). The cultures were washed after 4 h to exclude free parasites and then incubated for 5 days. Infected attached MQs were collected using cold PBS 1X (Breton et al. Reference Breton, Tremblay, Ouellette and Papadopoulou2005).
RNA extraction
Stationary phase promastigotes and amastigotes of L. tarentolae, L. major and L. infantum were utilized for RNA purification using RNeasy Mini kit (Qiagen). Briefly, after washing the promastigotes twice with PBS, total RNA was extracted according to the supplier kit. For total amastigote RNA isolation, the infected murine peritoneal macrophages were detached from the plate with cold PBS and lysed according to QIAGEN RNAeasy kit. To prevent co-purifying genomic DNA with RNA, on column DNase treatment was done using RNase free DNase set (Qiagen). The quality of the RNA was assessed by spectrophotometric analysis using NanoDrop ND-1000. Integrity of the RNA was checked on 1% formaldehyde agarose gels. To ensure that the obtained amastigote RNA is free of host (macrophage) RNA, we looked for the presence of murine ribosomal RNA on the gel.
cDNA synthesis
One μg of total RNA from each sample was reverse-transcribed to cDNA using the Omniscript RT kit (Qiagen) and random hexamer primers as described by the supplier. After preparation of the mixture, samples were incubated at 37°C for 1 h, heat-inactivated at 70°C for 6 min and subsequently stored in −20°C.
Real-time PCR
cDNAs from all samples were utilized as a template for the real-time polymerase chain reaction. Quantification of the mRNA corresponding to LPG3, Amastin and CPB was done in the 7500 real-time PCR system (Applied Biosystems) using QuantiFast SYBR Green master mix from Qiagen. Amplification of the Amastin fragment (185 bp) from L. major (Accession no. LmjF28.1400) and L. infantum (Accession no. LinJ28.1460) mRNA was carried out using the same primers (AmastinF, 5′-GCGGTGGAGACGTGCTG-3′ and AmastinR, 5′-GTGTAGGTAGAGTTGTCGCAGTTGT-3′). Amplification of a 100 bp region from the L. tarentolae Amastin transcript (GenBank Accession no. EU041954.1) was accomplished using a primer pair specific to this parasite (Amastin-tF, 5′-GCTTACCTGATGATGGCGGATA-3′ and Amastin-tR, 5′-CCAGTTGCGATGTGCAGATC-3′). The primers for LPG3 (LPG3F, 5′-TCTTCATCACGGACGAGTTCC-3′ and LPG3R, 5′-TTGCGCACAAGTTTCTTCTT-3′) were designed from a conserved region in a multiple sequence alignment of L. major, L. infantum, L. donovani and L. braziliensis LPG3 genes (Accession nos LmjF29.0760, LinJ29.0920, AF369892 and LbrM29-V2.0780, respectively) and were successfully used in amplifying the LPG3 transcript of L. tarentolae. The primers CPB-F, 5′-AACTTCGAGCGCAACCT-3′ and CPB-R, 5′-GGTCAAAGAACTTCGTGATC-3′ also served for amplifying an 85 bp fragment of L. infantum (LinJ08.0980) and of L. major (LmjF08.1080). Likewise, utilizing the primers CPB-LtF: 5′-AGCGGCGGGCTGATGT-3′ and CPB-LtR: 5′-GAGCACTCGGGCGCATAG-3′, a 115 bp region of L. tarentolae CPB (Accesion. no. EU860161) was successfully obtained. The primers that served to amplify a 100 bp rRNA region (LmjF32.3420 and LinJ32.3980 for L. major and L. infantum, respectively) were rRNA-F: 5′-CGGTTTTTCCCTGATGTGGTAA-3′ and rRNA-R: 5′-GGAAGCGTGGTCGACGTAGA-3′. Finally, GAPDH from all species was amplified using the 5′-TCGCTCTCGTGGACATGAGCA-3′ (forward) and 5′- CTTGAGCGACACGCCGTCCA-3′ (reverse). PCR reactions were performed at 95°C for 5 min followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Melting curve analysis demonstrated specific single PCR products for all reactions according to the amplicon melting temperatures. Furthermore, 2 housekeeper genes, GAPDH and rRNA45 were employed as internal controls for normalization. All the primers were designed utilizing Primer Express software v.2 (Applied Biosystems). Relative quantity of all genes was determined through ΔΔCT Method (SDS software, Applied Biosystems). RT negative controls were also performed to check for the presence of contaminated DNA.
SDS-PAGE and immunoblotting
Stationary phase promastigotes were collected from in vitro cultures by centrifugation (10 min, 1500 g, 4°C), washed in PBS and lysed in SDS-PAGE sample buffer containing 100 mm Tris-HCl, 4% SDS, 0·2% bromophenol blue, 20% glycerol and 200 mm β-mercaptoethanol. The lysate was separated on a 12·5% SDS-PAGE polyacrylamide gel and then transferred to a nitrocellulose membrane (Bio-Rad system). In addition, rLPG3 (Rafati et al. unpublished data), rCPB (Rafati et al. Reference Rafati, Salmanian, Hashemi, Schaff, Belli and Fasel2001) and synthetic amastin signature (Rafati et al. Reference Rafati, Hassani, Taslimi, Movassagh, Rochette and Papadopoulou2006) were used as a control. After protein transfer, the membrane was blocked overnight at 4°C in TBS (10 mm Tris-HCl, pH 7·4, 150 mm NaCl) containing 2·5% BSA for 2 h, then incubated overnight with either a rabbit anti-CPB polyclonal antibody (1:500), a mouse anti-LPG3 polyclonal antibody (1:900) or a mouse anti-amastin polyclonal antibody (1:600) in TBS. The membranes were washed 3 times with TBS, 0·1% Tween 20 and incubated for 1.5 h with either peroxidase-conjugated anti-rabbit (1:4000 Sigma) or peroxidase-conjugated goat anti-mouse IgG (1:6000, Sigma) as secondary antibodies. Unbound secondary antibodies were removed by washing as described above. Diaminobenzidine tetrahydrochloride (DAB, Sigma) was used as the substrate to detect the desired bands of the proteins.
RESULTS
Presence of known virulence factors in the non-pathogenic Leishmania tarentolae species
First, we looked whether the known virulence factors of pathogenic Leishmania species CPB, GP63, LPG3, A2 and Amastin were present in lizard non-pathogenic L. tarentolae. Total DNA from the pathogenic parasites L. major and L. infantum and from the non-pathogenic species L. tarentolae was digested, transferred to nylon membranes and hybridized with gene-specific probes corresponding to the coding regions of the above virulence factors. The results presented in Fig. 1 demonstrate that the genes for the cysteine peptidase B family, GP63 family, amastin family and LPG3 are present in L. tarentolae. Interestingly, the band patterns appear to be similar between the three species especially for CPB and GP63. The observed multiple banding in lanes corresponding to CPB and GP63 could be due to the presence of multiple copies of these genes in the genomes of these parasites. A similar pattern in L. tarentolae suggests that these factors are also present in multiple copies. Detection of the amastin genes in L. tarentolae required low-stringency conditions for hybridization. Screening of a L. tarentolae library was performed to seek homologues of the amastin LmjF24.1270, which is highly conserved between L. major and L. infantum. Partial screening led to the cloning of a 900 bp fragment from the genome of L. tarentolae. Sequencing and BLAST revealed that this fragment corresponds to the C-terminal region of the L. major's Lmj28.1400 orthologue (sequence submitted to GenBank, Accession no. EU041954). Additional analyses revealed that the retrieved sequence does not cover the whole ORF of the gene but a region close to its C-terminal. Interestingly, among all the virulence factors examined, the amastigote-specific A2 gene appears to be absent in L. tarentolae. Low-stringency washing and prolonged exposure was done to confirm this result (data not shown).
Southern blot analyses of various virulence factors in the human-pathogenic Leishmania major (LM) and L. infantum (LI) and the non-pathogenic L. tarentolae (LT) species. With the exception of the amastigote-specific gene A2, the rest of the virulence genes tested are present in the genome of L. tarentolae and show similar digestion patterns, which suggests a similar copy number and genomic organization.
Genomic structure and organization of known virulence factors in the non-pathogenic Leishmania tarentolae species
In this study, a partial sequence of an amastin homologue and a cysteine protease B gene and the full sequence of the LPG3 ORF from L. tarentolae were obtained (sequences submitted to NCBI with Accession nos EU041954, EU860161 and unpublished data, respectively). A BLAST search with these sequences throughout the genomes of L. major, L. infantum and L. braziliensis and subsequent pairwise alignment on protein sequences determined the degree of conservation between these genomes (Fig. 2). Heat maps shown in Fig. 2 provide a visual representation of the degree of homology of amastin, CPB and LPG3 genes between different Leishmania species. The black colour represents 65·4% similarity (Fig. 2B) (lowest similarity observed in all alignments between L. tarentolae amastin and L. braziliensis amastin) and the white colour represents 100% similarity (Fig. 2B). Figure 2A shows that LPG3 is highly conserved among all the species (lowest 83·5% between L. braziliensis and L. infantum) in comparison to amastin and CPB (Fig. 2C). Evaluation of protein homology was performed using the ClustalW software. Sequence alignment of amastin orthologues from L. major (LmjF28.1400), L. infantum (LinJ28.1460), L. braziliensis (LbrM28_V2_1550) and L. tarentolae (EU041954) revealed that amastin has the least level of conservation among the above-mentioned parasites (40·07%) with 73·95% sequence similarity only in the most conserved regions (e.g. transmembrane helices) (Fig. S1A, see online version). Amastin proteins are part of a large family of surface proteins in Leishmania that share a common structure in L. major and L. infantum. This structure consists of 4 transmembrane helices and a highly conserved 11-residue sequence between transmembrane helices 1 and 2 that is named the amastin signature (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). These conserved regions are also present in L. tarentolae (Fig. S1A, see online version). In addition, prediction of transmembrane helices using the web server TMHMM resulted in the same transmembrane topology to the other members of the amastin gene family with the exception of the last transmembrane region, which is not covered in the retrieved sequence (Fig. S1B, see online version).
Heat maps illustrating similarity of (A) LPG3, (B) amastin and (C) CPB among different Leishmania species. Each square represents one pairwise alignment; brighter colours represent higher similarity (white for 100%). Black represents 65·4% similarity, which is the lowest value observed in all alignments (L. tarentolae amastin vs L. braziliensis amastin). Heat maps were generated using Mayday (Dietzsch et al. Reference Dietzsch, Gehlenborg and Nieselt2006).
Similarly, alignment of the partial sequence of L. tarentolae CPB (EU860161) shared with L. major CPB (LmjF08.1030), L. infantum CPB (LinJ08_V3.0960) and L. braziliensis (LbrM08_V2.0820) resulted in 59·53% identity and 77·52% similarity (Fig. S1C, see online version). The maximum level of sequence similarity was obtained for LPG3 in which 86·25% similarity between the L. tarentolae, L. major (LmjF29.0760), L. infantum (LinJ29.0920) and L. braziliensis (LbrM29_V2_0780) orthologues (Fig. S1D, see online version) was estimated.
Expression patterns of known virulence factors in the non-pathogenic L. tarentolae species
As previously reported (Breton et al. Reference Breton, Tremblay, Ouellette and Papadopoulou2005), the L. tarentolae strain used in this study can efficiently infect human or murine macrophage cell lines as well as intraperitoneal macrophages freshly isolated from BALB/c mice, human PBMCs and immature monocyte-derived dendritic cells, and it can last for 5 days inside macrophages although with a gradual decrease in infection levels. Therefore, a relative quantification method (ΔΔ CT method) was applied to evaluate the relative expression ratio of the Amastin, CPB and LPG3 genes in both promastigote and amastigote forms of the human-pathogenic L. major and L. infantum species and of the lizard-non-pathogenic L. tarentolae. The rRNA45 and GAPDH genes were utilized as an internal control for normalizing the expression of the candidate genes. rRNA45 was reported to be expressed stably between promastigote and amastigote forms of L. major (Akopyants et al. Reference Akopyants, Matlib, Bukanova, Smeds, Brownstein, Stormo and Beverley2004; Ouakad et al. Reference Ouakad, Bahi-Jaber, Chenik, Dellagi and Louzir2007a). Results on LPG3 expression analysis showed that this gene is expressed in both life stages of the above-studied parasites. However, there were some differences in LPG3 expression levels between amastigotes/stationary-phase promastigotes of L. major (1·665) and L. infantum (1·862) versus L. tarentolae (0·926, Fig. 3). For the CPB gene, a similar pattern of expression was observed in the 3 species. Indeed, in all 3 species, CPB expression is slightly increased in the amastigote stage compared to stationary phase promastigotes (1·234, 1·351 and 1·17 for amastigotes/stationary promatigotes of L. major, L. infantum and L. tarentolae, respectively). While amastin of L. infantum shows higher expression in amastigotes than in stationary phase promastigotes (1·360), in L. major it appears that this amastin homologue is more expressed in promastigotes (ratio of amastigotes/stationary phase promastigotes of 0·57, Fig. 3). Similar observations regarding amastin expression were previously reported by Rochette et al. (Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). Interestingly, it is shown here that stationary phase promastigotes of L. tarentolae express this amastin gene homologue in higher levels than amastigotes (ratio of stationary phase promastigotes/amastigotes of 1·77).
Expression ratios of Amastin, LPG3 and CPB genes in Leishmania tarentolae, L. major and L. infantum amastigotes and stationary phase promastigotes. Based on these results, LPG3 shows a slight increase in expression in amastigotes of L. major (amastigotes/promastigotes 1·65) and L. infantum (amastigotes/promastigotes, 1·86) compared to stationary phase promastigotes; whereas, in L. tarentolae expression of LPG3 seems to be constitutive (amastigotes/promastigotes, 0·93). CPB, similarly to all studied species shows a small increase in its expression in amastigote forms (1·23, 1·35 and 1·17 for amastigotes/promatigotes of L. major, L. infantum and L. tarentolae, respectively). Lastly, amastin (dark bars) is down regulated in L. major (amastigotes/promastigotes, 0·57) and L. tarentolae amastigotes (amastigotes/promastigotes, 0·56), while L. infantum shows a slight increase in amastigotes (amastigotes/promastigotes, 1·36). Standard deviation form CT is shown as error bars.
We confirmed the expression of CPB, Amastin, as well as LPG3 mRNAs in both developmental stages of L. tarentolae by real time RT-PCR. As shown in Table 1, the obtained CT values of the studied genes (Amastin, CPB and LPG3) in L. tarentolae demonstrated that the level of expression is close to that in L. major and L. infantum. The minimum range of 19·89 corresponds to the LPG3 of L. major amastigotes and the maximum of 26·31 is relevant to CPB of L. infantum amastigotes (Table 1).
C T values from two developmental stages of the three Leishmania species are shown
(According to these results, expression levels of the studied genes in L. tarentolae are comparable to those of the other pathogenic species.)
Furthermore, we compared whether the proteins encoded by Amastin, CPB and LPG3 genes were differentially expressed between L. major, L. infantum and L. tarentolae species. As shown in Fig. 4(A), the profiles of all Leishmania lysates were similar according to Coommasie blue staining. Western blot analysis using specific polyclonal antibodies for each recombinant protein including rLPG3 (Fig. 4B1), rCPB (Fig. 4B2) and the Amastin signature in the form of synthetic peptides (Fig. 4B3), demonstrated that each of these virulence factors was expressed at the protein level in L. tarentolae. Although the reactivity using anti-CPB (Fig. 4B2) showed minor differences, its presence in the non-pathogenic L. tarentolae is indisputable. As elicited from Fig. 4, all 3 proteins were expressed constitutively in promastigotes of both the pathogenic (L. major, L. infantum) and non-pathogenic (L. tarentolae) Leishmania species.
SDS-PAGE and Western blot analysis of LPG3, CPB and amastin proteins in Leishmania tarentolae, L. infantum and L. major. (A) Lysates of L. tarentolae (lane 2), L. infantum (Lane 3), L. major (lane 4), rLPG3 (Lane 5), rCPB (Lane 6) and amastin signature (Lane 7) were subjected to electrophoresis on 12·5% SDS-PAGE and stained with Coomasie blue. (B) Same samples were transferred to a nitrocellulose membrane and incubated with different polyclonal antibodies including anti-LPG3 (B1), anti-CPB (B2) and anti-amastin (B3) as described in the Materials and Methods section.
DISCUSSION
Current research has shown that the presence or absence of certain virulence factors of Leishmania can have a dramatic effect on the fate of Leishmania's infection. In addition, clinical manifestations of leishmaniasis can vary depending on the species and even on strains (Salotra et al. Reference Salotra, Duncan, Singh, Subba Raju, Sreenivas and Nakhasi2006). Several investigations have been conducted aiming at identifying and characterizing these genes and their potential role in the pathogenesis of this disease (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005; Alexander et al. Reference Alexander, Coombs and Mottram1998; Mottram et al. Reference Mottram, Coombs and Alexander2004; Zhang and Matlashewski, Reference Zhang and Matlashewski1997, Reference Zhang and Matlashewski2004; Requena et al. Reference Requena, Alonso and Soto2000). Therefore, determining the role of these factors during infection provides a precious knowledge that leads to a better understanding of host-parasite interactions as well as to the identification of targets for treatment and vaccination. Recently, comparative transcriptomics together with proteomic analyses of different virulent Leishmania spp. have been brought into focus (McNicoll et al. Reference McNicoll, Drummelsmith, Muller, Madore, Boilard, Ouellette and Papadopoulou2006; Paape et al. Reference Paape, Lippuner, Schmid, Ackermann, Barrios-Llerena, Zimny-Arndt, Brinkmann, Arndt, Pleissner, Jungblut and Aebischer2008; Rosenzweig et al. Reference Rosenzweig, Smith, Opperdoes, Stern, Olafson and Zilberstein2008; Leifso et al. Reference Leifso, Cohen-Freue, Dogra, Murray and McMaster2007; Rochette et al. Reference Rochette, Raymond, Ubeda, Smith, Messier, Boisvert, Rigault, Corbeil, Ouellette and Papadopoulou2008).
Although Leishmania pathogenicity is most likely the result of a high complexity with more than one mechanism involved, it is possible to identify virulence factors by genomic comparisons between pathogenic and non-pathogenic Leishmania species. It has been reported that L. tarentolae is likely to be pathogenic for lizards but the mechanism(s) is unknown (Dollahon et al. Reference Dollahon and Janovy1973; Orlando et al. Reference Orlando, Rubio, Sturm, Campbell and Floeter-Winter2002). However, L. tarentolae is unable to cause disease in mice, even in immunocompromized mice (Breton et al Reference Breton, Tremblay, Ouellette and Papadopoulou2005), and has never been associated with a Leishmania disease in humans. Here, we investigated whether known virulence factors of Leishmania were present and also expressed in the lizard L. tarentolae non-pathogenic species to humans. Among them, we tested the cysteine proteases, GP63, LPG3, A2 and amastins and found that, with the exception of A2, all the other proteins were expressed in L. tarentolae. Using antisense RNA, it was shown that the L. donovani A2 gene is essential for survival of amastigotes both in vitro cultured macrophages as well as in mice (Zhang and Matlashewski, Reference Zhang and Matlashewski1997). Furthermore, using a recombinant A2 protein as vaccine against L. amazonensis infection in susceptible mice, a robust protection has been obtained (Coelho et al. Reference Coelho, Tavares, Carvalho, Chaves, Teixeira, Rodrigues, Charest, Matlashewski, Gazzinelli and Fernandes2003). Because A2 protein is truncated in L. major (Ghedin et al. Reference Ghedin, Charest and Matlashewski1998; Zhang et al. Reference Zhang, Mendez, Ghosh, Myler, Ivens, Clos, Sacks and Matlashewski2003), it was suggested that A2 might contribute to the tropism of L. donovani to visceralization (Zhang et al. Reference Zhang, Mendez, Ghosh, Myler, Ivens, Clos, Sacks and Matlashewski2003). The absence of A2 in L. tarentolae is intriguing and it may be associated to some extent with the loss of pathogenicity of lizard Leishmania in mammals. This needs to be experimentally validated, however.
Similar to L. major and L. infantum, L. tarentolae's GP63 is a multi-copy gene. The role of GP63 in pathogenesis has been widely studied (Bordier, Reference Bordier1987; Medina-Acosta et al. Reference Medina-Acosta, Beverley and Russell1993; McMaster et al. Reference McMaster, Morrison, MacDonald and Joshi1994; Yao et al. Reference Yao, Donelson and Wilson2003) although its function in disease development and also pathogenesis has not been clearly defined. Chen et al. (Reference Chen, Kolli, Yadava, Lu, Gilman-Sachs, Peterson and Chang2000) have shown that down regulation of GP63 expression in L. braziliensis using antisense RNA strategy led to impaired internalization of macrophage cells in vitro. On the contrary, a similar study on the effect of L. major gp63 −/− in disease progression resulted in impaired macrophage infection, although delayed infectivity of challenged BALB/c mice was observed (Joshi et al. Reference Joshi, Kelly, Kamhawi, Sacks and McMaster2002). On the basis of this finding, it was proposed that GP63 might contribute to additional pathogenic pathways such as migration through the extracellular matrix. Interestingly, it has also been shown (Gregory et al. Reference Gregory, Godbout, Contreras, Forget and Olivier2008) that L. tarentolae is able to perform GP63-dependent cleavages in host proteins during infection.
In this study we also showed that LPG3, CPBs and amastins were expressed in L. tarentolae. LPG3 is an endoplasmic reticulum chaperone and homologue of human GRP94 (Larreta et al. Reference Larreta, Soto, Alonso and Requena2000), which is believed to be crucial for proper functioning of LPG synthesis pathway (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002). In L. major, it has been shown that mutants lacking this gene are avirulent (Spath et al. Reference Spath, Epstein, Leader, Singer, Avila, Turco and Beverley2000), while in L. mexicana, it has been demonstrated that LPG is not essential for macrophage infection as well as lesion development in mice (Ilg, Reference Ilg2000). It is worth mentioning that in contrast to the pathogenic Leishmania, LPG appears to be absent from the lizard parasite L. tarentolae (Previato et al. Reference Previato, Jones, Wait, Routier, Saraiva and Mendonça-Previato1997). In this study, we have shown that LPG3 is preferentially expressed in amastigotes of L. major and L. infantum, which is consistent with similar findings from L. donovani (Descoteaux et al. Reference Descoteaux, Avila, Zhang, Turco and Beverley2002). Nonetheless, in L. tarentolae, we found that LPG3 gene is constitutively expressed. Species-specific differences even between pathogenic Leishmania species have recently been reported (Rochette et al. Reference Rochette, Raymond, Ubeda, Smith, Messier, Boisvert, Rigault, Corbeil, Ouellette and Papadopoulou2008). These differences point to a differential gene regulation possibly by sequences in the 3′UTR, which generally are less conserved between species than the protein-coding sequences (Peacock et al. Reference Peacock, Seeger, Harris, Murphy, Ruiz, Quail, Peters, Adlem, Tivey, Aslett, Kerhornou, Ivens, Fraser, Rajandream, Carver, Norbertczak, Chillingworth, Hance, Jagels, Moule, Ormond, Rutter, Squares, Whitehead, Rabbinowitsch, Arrowsmith, White, Thurston, Bringaud, Baldauf, Faulconbridge, Jeffares, Depledge, Oyola, Hilley, Brito, Tosi, Barrell, Cruz, Mottram, Smith and Berriman2007). Multiple sequence alignment of the LPG3 protein between the 4 studied Leishmania species showed a high level of conservation for the LPG3 orthologues (86·25%).
Among Leishmania cysteine proteinases (CPA, CPB and CPC), the role of CPB in pathogenesis has been extensively highlighted, as L. mexicana deficient in the cpb array genes showed reduced infectivity in experimentally infected BALB/c mice (Frame et al. Reference Frame, Mottram and Coombs2000; Alexander et al. Reference Alexander, Coombs and Mottram1998). It has been reported that recombinant hybrid proteins encompassing CPA and CPB could elicit protection against L. major in BALB/c mice (Zadeh-Vakili et al. Reference Zadeh-Vakili, Taheri, Taslimi, Doustdari, Salmanian and Rafati2004). Furthermore, the vaccine trial in dogs using cocktails of CPA and CPB gave satisfactory results (Rafati et al. Reference Rafati, Nakhaee, Taheri, Taslimi, Darabi, Eravani, Sanos, Kaye, Taghikhani, Jamshidi and Rad2005), although further study is required. Alongside these findings, intensive efforts have been conducted towards identifying the role of CPB in virulence. Results on L. mexicana revealed that CPB promotes survival and growth of intracellular parasites living in phagolysosomes through inhibition of the NF-kB pathway and degradation of MHC class II molecules thereby preventing IL-12 production and eliciting IL-4 expression (Mottram et al. Reference Mottram, Coombs and Alexander2004). However, infection of L. major-resistant mouse strains with L. mexicana resulted in lesion development due to the ability of L. mexicana to cleave a component of NF-kB, a capacity that seems to be absent in L. major (Cameron et al. Reference Cameron, McGachy, Anderson, Paul, Coombs, Mottram, Alexander and Plevin2004; Buxbaum et al. Reference Buxbaum, Denise, Coombs, Alexander, Mottram and Scott2003). Moreover, these studies demonstrated that infection of macrophages with wild type or cpb −/−L. mexicana elicited similar levels of IL-12 production, suggesting that other antigen-presenting cells (APCs) may be involved in vivo, instead of macrophages or even modulate other Th1-inducing cytokines (such as IL-18) (Fukao et al. Reference Fukao, Matsuda and Koyasu2000). Given the above-discussed functions of CPB, it is surprising that the non-pathogenic L. tarentolae does possess and express CPB in its genome. Besides, Southern blot analysis suggested that CPB is a multi-copy gene in L. tarentolae, similar to the pathogenic species. Our analysis showed that CPB is expressed slightly more in the amastigote stage in all 3 parasitic species. Such a high level of similarity in CPB expression, even at the protein level, between pathogenic and non-pathogenic species is a noticeable finding that needs additional studying, especially at the level of protein function. Characterization of this protein in L. tarentolae will be interesting due to the differences in protein sequence and possibly structure in comparison to the L. major and L. infantum CPB. Heat map analysis and alignments of the obtained partial sequence of CPB from L. tarentolae with its equivalent in L. major, L. infantum and L. braziliensis have predicted 71–77% similarity. Meanwhile, the maximum sequence similarity of 88·4% was estimated between L. major and L. infantum. Based on this fact, one could conclude that CPB should have a high degree of heterogeneity over different Leishmania species, which is compatible with previous findings on its copy number (Mottram et al. Reference Mottram, Coombs and Alexander2004) and variation in the C-terminal extension (Hide et al. Reference Hide, Bras-Goncalves and Bañuls2007; Nakhaee et al. Reference Nakhaee, Taheri, Taghikhani, Mohebali, Salmanian, Fasel and Rafati2004; Rafati et al. Reference Rafati, Nakhaee, Taheri, Ghashghaii, Salmanian, Jimenez, Mohebali, Masina and Fasel2003).
Finally, we analysed the presence and organization of the amastigote-specific amastin gene family in L. tarentolae. Amastins are the largest family of developmentally regulated surface proteins in Leishmania (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). Although their role has not yet been characterized, it is suggested that these proteins may be involved in the transportation of ions, metals and nutrients by the internalized parasites within the phagolysosome (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). As suggested by Southern blot hybridization, L. tarentolae also has multiple copies of this gene family. However, the ongoing sequence of the L. tarentolae genome indicated a significantly smaller number of amastins in this non-pathogenic species to humans compared to the pathogenic Leishmania (Papadopoulou et al., unpublished observations). Heat map analysis and sequence alignment of the L. tarentolae amastin homologue revealed that this putative amastin protein does possess the 11 amino acid amastin signature sequence, which is characteristic of all amastin proteins. The results of real-time PCR showed that amastin is differentially expressed between amastigotes and stationary-phase promastigotes. Meanwhile, L. major and L. tarentolae express this gene preferentially in stationary phase promastigotes, whereas in L. infantum this gene is preferentially expressed in amastigotes as previously reported (Rochette et al. Reference Rochette, McNicoll, Girard, Breton, Leblanc, Bergeron and Papadopoulou2005). Western blot analysis confirmed amastin protein expression in L. tarentolae promastigotes, as well. It is worth mentioning that like LPG3 and CPB, expression levels of amastin in L. tarentolae are comparable to those of the pathogenic species. The L. tarentolae amastin homologue shows 65–72% sequence homology to the closest orthologues in L. major, L. infantum and L. braziliensis as compared to 89·8% sequence identity between L. major and L. infantum. The lower degree of similarity of the L. tarentolae amastin could affect interactions with other proteins or substrates and hence modulate amastin's function. Amastin's putative role in pathogenesis has been recently suggested by DNA microarray data, indicating that amastin genes were differentially expressed between different L. donovani strains isolated from PKDL and VL patients (Salotra et al. Reference Salotra, Duncan, Singh, Subba Raju, Sreenivas and Nakhasi2006). However, the exact role of amastin in the disease progression has not yet been identified.
In summary, to our knowledge, this work provides the first comparative genomic analysis and expression profiles of important virulence genes, such as CPB, LPG3 and Amastin in stationary phase promastigotes and amastigotes between pathogenic Leishmania spp. and L. tarentolae, the lizard species non-pathogenic to humans. Our data confirmed that these virulence genes were expressed and the proteins produced in L. tarentolae. At this level, we cannot exclude that differences in protein sequence or species-specific differential regulation, mainly at the post-translational level of these virulence factors, might alter the interaction of these proteins with other cellular partners, which could result in changes in the parasite's virulence. It is also possible that differences in the thermophysiology between lizard and mammalian Leishmania species could account for the loss of virulence of L. tarentolae in mammals. It is not known what genes are involved in this process but it has been reported that some dehydrogenases and aldolases showed differential expression between lizard- and human-infecting Leishmania species (Janovy et al. Reference Janovy1972, Ghosh et al. Reference Ghosh and Honigberg1976). This interesting aspect requires further investigation, which is outside of the scope of this study. It is also likely that other, yet uncharacterized virulence factors, absent or differentially regulated in the lizard L. tarentolae could account for the loss of pathogenicity in this species. Whole-genome comparative analyses between pathogenic and non-pathogenic Leishmania species will shed more light in that direction.
The financial support by the Iran Research Council of the Republic Presidentship is gratefully acknowledged. In addition, this project was supported by Pasteur Institute of Iran grant number 282. We wish to thank Mr Alizadeh for technical assistance.
