Multigene Families in Trypanosoma cruzi and Their Role in Infectivity

Abstract

The Trypanosoma cruzi genome contains the most widely expanded content (∼12,000 genes) of the trypanosomatids sequenced to date. This expansion is reflected in the high number of repetitive sequences and particularly in the large quantity of genes that make up its multigene families. Recently it was discovered that the contents of these families vary between phylogenetically unrelated strains. We review the basic characteristics of trans-sialidases and mucins as part of the mechanisms of immune evasion of T. cruzi and as ligands and factors involved in the cross talk between the host cell and the parasite. We also show recently published data describing two new multigene families, DGF-1 and MASP, that form an important part of the scenario representing the complex biology of T. cruzi.

INTRODUCTION

The Trypanosoma cruzi genome project forms part of a major initiative known as TriTryp, under the aegis of which the genomes of Leishmania major, Trypanosoma. brucei, and T. cruzi have been sequenced and compared (41). Although these three kinetoplastids diverged between 500 and 200 million years ago, their genomes show a long scale of genetic conservation, or synteny (a high degree of synteny suggests phylogenetic proximity), and T. brucei and L. major have 68 and 75% of their genomes, respectively, in common with that of T. cruzi (29, 35). Interest in sequencing the T. cruzi genome arose at a series of meetings held in Brazil, Argentina, and the United Kingdom in 1993, 1994, and 1997; the CL-Brener strain was chosen for sequencing, and the draft genome was finally published in 2005 (28). On the basis of a haplotype analysis of this strain, it was calculated that the genome is composed of 12,000 genes (a higher dosage than that found in the other kinetoplastids sequenced to date) and it has been possible to assign functions to 50.8% of them (28). Approximately 50% of the genome is composed of repetitive sequences, mainly of large multigene families and retrotransposons (28). Various biological, biochemical, and molecular studies have shown that T. cruzi is a very heterogeneous taxon, and comparisons of field and laboratory isolates have revealed numerous genetic and protein polymorphisms (13, 44, 53). These intraspecific variations give rise to evident linkage disequilibrium, which indicates a clonal population structure of this parasite (13, 66). This, together with evidence that genetic exchange occurs in the parasite, is indicative of the complex structure of these populations (65). The current classification characterizing T. cruzi isolates phylogenetically proposes six discrete typing units (DTUs), I to VI (75), with DTU I being subdivided into at least five genotypes on the basis of spliced-leader intergenic region polymorphism (22, 30).

T. cruzi genome sequencing has shown that the synteny between kinetoplastid genomes is inconsistent at certain points, suggesting the existence of species-specific genes and sequences present in large nonsyntenic islands (>600 kb) of genes encoding surface proteins such as trans-sialidases (TSs), mucins, mucin-associated surface proteins (MASPs), dispersed gene family 1 (DGF-1), and gp63 peptidases, all with retrotransposons and retrotransposon hot-spot protein gene inserts (29).

Recent comparisons of the genetic content of phylogenetically spaced strains made by comparative genomic hybridization and using the CL-Brener strain as the template have demonstrated a high variation in copy numbers among the various strains of T. cruzi, which affect mainly the highly populated sites of repeat regions, including the T. cruzi multigene families mentioned above (47). In addition, the genome of a new strain of T. cruzi, Silvio X10/1 (DTU I), which belongs to a phylogenetic group somewhat distant from the originally sequenced CL-Brener strain (DTU VI), has been sequenced. This sequencing revealed a contraction in the gene content of the Silvio X10/1 multigene families, which theoretically increases the plasticity of the T. cruzi genome (4, 33). On the basis of evidence provided by currently available markers (microsatellites, spliced-leader intergenic regions, and mitochondrial gene polymorphisms) to determine phylogenetic separation between the different strains of T. cruzi, the association between the DTUs and the specific T. cruzi biodemes proves not to be a strong one, suggesting that the individual biological traits of strains of this parasite may be determined by variations in the content of multigene families (Fig. 1). This high variability could confer upon the parasite the ability to invade a significant number of cell types, as well as to have various tissue tropisms, causing the different types of heart conditions and megasyndromes associated with Chagas' disease.

Fig 1.

Schematic representation of the diversity of gene dosage in multigene families in strains belonging to different DTUs. Shown are the different factors involved in the genetic variability observed; in the center is a schematic representation of the variations in the gene contents of multigene families in various strains of T. cruzi. The colored horizontal bars represent the gene numbers of the multigene families in each lineage. *, Silvio X10/1 (DTU I) genome sequence available; **, CL-Brener (DTU VI) genome sequence available.

MULTIGENE FAMILIES AND INFECTIVITY

TSs.

The TS superfamily is composed of glycosylphosphatidylinositol (GPI) proteins anchored to the surface of kinetoplastid species such as T. cruzi, T. brucei, and T. rangeli. The family contains 1,430 genes, half of which are apparently functional and represent one of the largest phenomena of gene expansion described in the kinetoplastids (28). Many of these genes are found near the telomeric regions, which implies that part of the gene family expansion is due to their chromosomal location and to the immune system pressure to which the TSs are exposed (40). A comparative analysis of the TS sequences deposited in the genome of T. cruzi has led to the differentiation of a total of eight different groups forming individual clusters of proteins (34). The expression of these genes is due in part to the regulation of mRNA stability and to efficient translational control of TS family transcripts by 3′ untranslated regions (UTRs). These UTRs could cause expression efficiency to differ between the various stages, with greater expression efficiency in the trypomastigote blood forms resulting from intracellular replication, although it has been recently shown that expression levels are not homogeneous among genes belonging to different groups of TSs (34, 38).

One of the best-characterized functions of this family is the TS enzyme, which was first described in 1980 (56). Subsequent studies with wheat germ lectins and monoclonal antibodies to the N terminus of the TS enzyme (where the conserved Ser-X-Asp-X-Gly-X-Thr-Trp motifs present in bacterial neuraminidases are located) successfully demonstrated that T. cruzi is unable to synthesize sialic acid and uses the TS to transfer this molecule to itself from host cell glycoconjugates, which are the β-galactosidase (β-Gal) mucin residues from the parasite acceptors of that molecule (12, 52, 58, 60, 62, 63). The TS enzyme also exerts neuraminidase activity, which occurs only when suitable Gal acceptors are present and represents ∼5% of the total activity of the TS enzyme (55). The difference between the active and inactive enzymes resides in a mutation in the Tyr342 residue needed to prompt the catalytic reaction. The motifs that catalyze the protein enzyme activity are located in the N-terminal region (16). The sialic acid bond triggers a conformational switch in the enzyme that facilitates binding by reorienting the Tyr119 residue, thus activating the enzyme by repositioning the Tyr342/Leu36 residue (14). This enzyme activity may be the result of TS evolution from an ancestral trypanosome hydrolase, conferring protection from the attack of anti-alpha-galactosyl antibodies through the negatively charged coat of the surface of the parasite produced by the sialylation of mucins of the trypomastigote form (14, 54).

In addition to the transfer of sialic acid residues to the parasite, another proven function of these glycoproteins is that they bind to host cell membrane receptors, thus stabilizing the contact between the host cell and the parasite. Its ability to bind to host cell receptors resides in the conserved motif VTVxNVxLYNR, known as the FLY domain, present in 371 members of the gene family. Interestingly, by using phage display, it has recently demonstrated that this motif harbors patterns for binding to different organs, remarkably resembling the tissue tropism described in human disease and animal models (28, 45, 68). In these experiments, the authors confirmed that FLY binds strongly to heart, esophagus, and bladder endothelium, particularly to cytokeratins such as CK8 (expressed by muscle and epithelial cells), CK20 (epithelial cells), and vimentin (endothelial cells), and possibly to other kinds of intermediate filaments that form an important part of the host cell cytoskeleton (68).

Laminins are another type of receptor to which such proteins can bind by anchoring to an epitope recognized by the monoclonal antibody Mo H1A10, an epitope present in the gp85/TS subgroup of the TS family (36), and in fact, members of the gp85/TS subgroup such as gp82, gp90, and gp35/50 have been related to the parasite's ability to invade its host cell. gp82 is able to promote invasion by a bidirectional signaling cascade, leading to the mobilization of intracellular Ca²⁺ in both parasite and host cells. Furthermore, the binding of gp82 to target cells induces the disruption of the actin microfilaments that is required for the parasite's entry into the cell (31, 59). The importance of TS proteins in the T. cruzi invasion process has been widely demonstrated by using specific antibodies against TS epitopes, sialic acid acceptors, and the catalytic domain of TS; interaction between the parasites and cells in the presence of antibodies against these epitopes reduces the infection of host cells (32, 46, 61, 72).

Another well-characterized function of TSs is to act as a parasite-derived neurotrophic factor, by the union with Trk receptors of mammalian cells, acting as an antiapoptotic factor in these cells (18, 23). The TrK signal helps to mediate differentiation, cell survival, and the regeneration of both the central and peripheral nervous systems (37). The mechanism resides in a domain of 21 amino acids that binds to the nerve growth factor TrKA receptor and produces its autophosphorylation and thereby activates phosphatidylinositol 3-kinase/Akt antiapoptotic signaling (17, 19, 20). In this way, the interaction between PNDF and TrKC receptors, which are receptors for the entry of T. cruzi into neural, glial, and epithelial cells, was recently demonstrated (73, 74). This function is independent of sialic acid, which was demonstrated through the use of sialic acid mutant CHO cells and by the ability of the parasite to trigger the Akt response in the host cell cytosol, where there is no sialic acid available to the parasite, upregulating prosurvival mechanisms such that cells become resistant to oxidative stress (21, 74). Therefore, TSs are part of the mechanism of survival in host cells, which facilitates the persistence of the parasite in infected tissues.

Another key function of a group of proteins of the TS superfamily named complement regulatory proteins (CRPs) is immune evasion, which is brought about by restricting the activation of the complement pathway and therefore the lysis of the parasite. This CRP binds to C3b and C4b via noncovalent interactions, suggesting that interaction with these molecules may be important for C3 decay acceleration activity (50). The members of this subgroup are highly antigenic and can be coexpressed, since it has been demonstrated that the serum of patients with Chagas' disease can recognize more than one isoform isolated from an individual cell line (11). The exact mechanism that enables the protein to bind to its target is still unknown, but it has been suggested that the C-terminal domain serves as an extension of the functional domain beyond the cell membrane, favoring recognition and interaction with C3b. The N-linked glycosylation of the CRP functional domain appears to be important for this recognition to take place (11, 51).

Although TSs are linked to the membrane by GPI anchors, they can be released into the bloodstream through the action of a phosphatidylinositol phospholipase C that acts on these GPI anchors (12). The release of this enzyme alters the surface sialidase profile, triggering multiple biological activities. The TSs that have been shed then transfer sialic acid residues from the platelet surface, causing thrombocytopenia during the acute phase of the disease. A single intravenous injection into mice of 5 μg of the enzyme lowered platelet counts by 50% (69). The enzymatic activity of TSs could also lead to apoptosis of immune cells in the spleen, thymus, and peripheral lymph nodes (43, 49, 67).

Due to the parasite's need to obtain sialic acid from host glycoconjugates, together with a lack of any biosynthetic alternative to obtain a monosaccharide and the various biological activities of the enzyme during the pathogenesis of Chagas' disease, this enzyme is a good candidate for designing specific drugs against T. cruzi (5, 48).

Mucins (TcMUCs).

During its different stages, the external layer of T. cruzi is coated with a series of glycoproteins, many of which belong to the mucin (TcMUC) family. Mucins are the most commonly expressed components on the surface of T. cruzi (∼2 × 10⁶ copies per parasite) and are the third most widely expanded gene family in the genome, comprising about 863 genes (201 of which are pseudogenes) (1, 28). These proteins have two main functions, to protect the parasite from both the vector's and the host's defensive mechanisms and to ensure the anchorage point and invasion of specific cells and tissues (12). Mucins contain O-glycosylated sugars which account for up to 60% of the total glycoprotein weight and attach themselves to the external lipid surface of the plasma membrane via GPI anchors (12). All TcMUC genes encode proteins containing a signal peptide, binding sites for GPI anchors and a hypervariable (HV) central region. This region contains a core polypeptide comprising 50 to 200 amino acids that are rich in serine/threonine (Ser/Thr) residues and have binding sites for N-acetylglucosamine (GlcNAc) residues. Many of the genes belonging to this family are pseudogenes and, as do the functional genes, account for the highest number of mutations in the central HV domain (15). The central region may be grouped into repeat units in which the consensus structures are T(6-8)KP(1-2) (family 1), KNT7ST3S(S/K)AP (family 2), and DQT17-20NAPAKDT5-7NAPK (family 3; also called L and S subfamilies) (12). Alternately, some gene products are organized into repeat sequences in the central region but are rich in Thr, Ser, and Pro (family 2) residues (26). These residues in TcMUC group II are targets for the O-glycosylated sugars and consequently potential sialic acid acceptor sites, which may explain why the amastigote and trypomastigote stage mucins (mammal host stages) show higher glycosylation and are larger than those expressed by the epimastigote forms in the insect gut (12).

One of the main characteristics of this family is the protective coating formed around the surface of the parasite, making it impervious to protease and glycosidase attack and allowing it to survive by resisting enzyme digestion during parasite development and growth in the insect vector. During this epimastigote stage, the potential sialic acid residue binding sites of mucins are not biologically significant because they are eliminated from the surface of the parasite inside the insect's gut (12). In contrast, the sialylation of mucins during the developmental stages of T. cruzi in mammals actively protects the parasite against the host's humoral immune response and may give the parasite a negative charge, as explained above (12). Although the O-oligosaccharide composition of mucins has not yet been completely clarified, it is known that during the mammal-dwelling stage, mucins expose epitope terminals of Gal(α1,3)Gal, which are among the main targets of the humoral immune system during infection by the parasite (2). Nevertheless, the negative charge conferred by sialic acid residues acquired by TSs neutralizes any complement-independent lysis induced by anti-α-galactosyl antibodies, thus allowing the parasite to avoid lysis (1). Many TcMUC II genes are linked to TS genes, which may explain the coordinated expression of functionally related TS and TcMUC on the surface of blood form trypomastigotes (15). Mucins are divided into two major groups of genes known as TcMUC I and TcMUC II, although other variants have been discovered, such as the T. cruzi small mucin-like gene (TcSMUG) family (12, 27). TcSMUGs exhibit much lower internal variability than do the other two families, which may be explained by their expression during the biological stages in the insect vector, where they are free from immune system pressure (13). Recent studies have shown that there is high variability in the expression and processing of TcSMUGs between parasite strains (71). Nevertheless, the HV region of these proteins has a mosaic of variants that are expressed simultaneously in the same parasite and are recognized by the serum of the infected host, which can lead to anergy in the protective response of CD4⁺ cells (57). Mucins are also able to induce the synthesis of the proinflammatory cytokines interleukin-12 and tumor necrosis factor, as well as nitric oxide production by macrophages. The motifs for the GPI anchors of these proteins are responsible for the immune stimulation of the proinflammatory response by binding specifically to Toll-like receptor 2 (TLR2) on the macrophage surfaces, although other TLRs may also be involved (3). Another trait conferred by its HV region is to provide binding ligands for the parasite for a variety of cell types. Treatment with monoclonal antibodies against mucins inhibits the entry of T. cruzi (61), although the true role of these mucins in its invasion may be to cooperate with other surface molecules of the trypomastigote such as the TSs mentioned above, because these TSs are more highly expressed during the trypomastigote and amastigote stages (1).

DGF-1.

This is the fifth largest gene family in the genome of T. cruzi and contains 565 genes and 136 pseudogenes (28). Many members of this family are located in subtelomeric chromosome regions where the variability in their sequences may be favored (39, 40). According to its genealogy, the family can be divided into at least three groups, with phenomena of gene duplication, recombination, and hybridization that would stimulate the generation of diversity in terms of gene conversion (39). Bioinformatic studies of various T. cruzi gene families and multiple point mutation models over 1,000 generations have shown that this family has undergone a series of point mutations and shows the highest rate of variability in its sequences (9). Gene conversion events between members of this family suggest that its pseudogenes may be responsible for stimulating sequence diversity and maximizing amino acid variability at the nucleotide level (9).

Because epidermal growth factors and arginine-glycine-aspartic acid (RGD) tripeptide domains are present in DGF-1 genes, it is believed that they may act similarly to integrins (39). These proteins have been identified by proteomic analysis in the cell culture-derived trypomastigote stage, appearing as N-glycosylated proteins (7). The use of antibodies against the DGF-1.2 protein has revealed its location in the cytoplasm of all of the forms of the parasite, with greater abundance in amastigotes than in trypomastigotes and epimastigotes (42). The location of the protein varies during differentiation from the trypomastigote to the amastigote stage, where it finally becomes clustered and bound to the inner side of the plasma membrane. At least 22 DGF-1 members have been identified in the trypomastigote form and a total of 39 have been identified in the contractile vacuole of the epimastigote stage (42). This high number, along with their location, points to the importance of the role that this family might play in the biology of the parasite (70).

MASPs.

MASPs are encoded by the second largest gene family in the T. cruzi genome. This family of genes and proteins was first identified during the analysis and sequencing of the T. cruzi genome and received its name from the position of its members among large TS and mucin gene groups, particularly those members downstream from the TcMUC II mucins, to which they are similar in structure though not in sequence (10, 28). Members of this family are characterized by two highly conserved N- and C-terminal regions, as well as a signal peptide and a GPI anchor site (Fig. 2). Their central region varies considerably in length, as well as in sequence, and there are few common motifs found in that region (10).

Fig 2.

Cell pathway of MASP synthesis. Shown are the basic composition of the mRNA and the structures of the immature and mature forms of the protein (top), with the conserved and HV regions indicated. At the top are the mature protein regions at the membrane of T. cruzi without the conserved signal peptide (SP) and C-terminal regions (processed at the Golgi apparatus and endoplasmic reticulum). On the left are two immunofluorescence images of the MASPs (obtained with anti-SP antibodies) and the surface location of HV MASP regions (obtained with antibodies against the anticatalytic region of the MASP52 protein). N, nucleus; K, kinetoplast; SL, spliced leader.

The MASP-encoding gene family in the T. cruzi genome comprises 1,377 genes and 433 pseudogenes distributed in arrays of up to 600 kb, although the total number may be higher due to the fragmentation of the CL-Brener genome current assembly (6). A unique trait of this family is the formation of chimeras that contain conserved MASP motifs at either the N or the C terminus with N- or C-terminal domains of mucins or C terminals of TSs or else lack the N- or C-terminal MASP domain (28). The distribution of the genes for these proteins among those for other families of surface proteins may help maintain genetic diversity, preventing the sequence homogenization of MASP genes (10).

Although not it is always the case, many MASP-encoding genes are (like other surface protein-encoding gene families) located at internal chromosome sites in blocks in which synteny is not conserved compared to other trypanosomatids such as L. major or T. brucei. The variability expressed by MASPs is not due to association with telomeres, as in the case of the variable surface glycoprotein of T. brucei, because other mechanisms generate variability in these proteins in T. cruzi. Nevertheless, the sequence variability found in MASPs may be important for providing the parasite with a source of antigen and thus aiding immune evasion (10).

The UTRs of this gene family are highly conserved, which suggests that such conservation constitutes sites, together with motifs, within the same HV region (although there is no location with a minimum sequence identity in the latter region) for homologous recombination and thus for replication (10). It has been suggested that one of the main mechanisms behind the formation of the different members of the MASP family is retrotransposition by mobile elements of the TcTREZO type. This mobile element is specific to this gene family, having its insertion sites at the conserved 5′ and 3′ ends of the MASP-encoding genes and pseudogenes (64).

MASPs have sites for N- and O-glycosylation which undergo extensive posttranslational modifications, thus perhaps explaining why proteomic analyses have been unable to detect many members of this family even though protein expression may be much higher than that obtained on the basis of the number of genes in the family (8). Fourteen members of this family have been reported and in some cases are expressed simultaneously in trypomastigotes, as well as in amastigotes and epimastigotes (7). They have been found in the membrane of trypomastigotes, although they can in some cases be shed into the culture medium (10). Our group has investigated the involvement of the MASP family in T. cruzi infectivity and found that IgG antibodies specific to the catalytic region of the MASP52 protein significantly reduce the parasite's capacity to infect host cells. This protein was also overexpressed in the infectious parasitic stages and located in the plasma membrane, demonstrating that the MASP family plays an important part in the mechanism of T. cruzi invasion of the host cell (24).

Although it is presumed that this family is expressed to a greater degree in the infective stages of the parasite (10, 24, 25), we have recently demonstrated clear differences at the level of gene expression among strains belonging to different DTUs (19), in agreement with previous work that reveal the great diversity in the dosage of the MASP-encoding gene family among different lineages (Fig. 1) (4, 27, 39). Using the conserved N-terminal sequence of the signal peptide of this family, we found considerable changes in expression during the parasite's intracellular cycle, particularly when it left the parasitophorous vacuole 24 h after infection (19). The data gathered from the overexpression of MASPs in intracellular parasites beginning at 24 h postinfection, prior to the division of the amastigotes, suggest that some of the proteins in this extensive family play a major biological role in the survival and multiplication of intracellular amastigotes.

Information provided recently by our group and that of Bartholomeu (10) supports the hypothesis that the expression of this type of protein is very often clonal, which is to say that within a population of parasites, the repertory and expression of MASPs vary with regard to both quality and time (9, 19). The causes of this variability remain unknown, although the absence of any detectable RNA polymerase II promoters and the polycistronic organization of transcripts of T. cruzi render it possible that the regulation of the expression of the MASP family might be tied to posttranscriptional mechanisms, thus opening the way to future research in this field.

CONCLUDING REMARKS

The great expansion of surface protein families in T. cruzi may be due to the greater complexity of its life cycle than that of other kinetoplastids. The need to establish contact with a large number of surface receptors on the different cells the parasite invades, together with the various extra- and intracellular environments involved in the many biochemical and physiological adaptations required of T. cruzi during its biological cycle, means that the organism can adapt to various stress situations, thus increasing the number of surface molecules able to interact and “connect” with the external medium surrounding this protozoan. This field of research is currently undergoing significant advances due to the emergence of high-throughput technologies, which will be useful in throwing more light on the development and evolution of multigene families within the complex genomic context of T. cruzi.