Trypanosoma cruzi NST candidates
We initially searched for putative nucleotide sugar transporters in the T. cruzi genome by performing Blastp searches in GeneDB 16] using characterized NSTs of different organisms as queries. We have identified a
family of eleven putative NSTs (Table 1) showing considerable similarity (e-value??1e-10) to known NSTs. Consistent with
their putative assigned function, the genes code for multi-transmembrane (TM) proteins
displaying between 7 and 10 TM domains (Table 1).
Table 1. Putative NSTs identified in the T. cruzi genome (Clone CL Brener)
Due to the hybrid nature of the diploid CL Brener strain 17], both haplotypes of nine of the genes could be identified based on similarity and
synteny (alleles are indicated in parentheses in Table 1). The alleles in each pair are at least 90 % identical to each other at the amino
acid level and, with two exceptions, code for proteins of identical or similar size
(up to three amino acids in difference). The genes TcCLB.509127.90 and TcCLB.511817.280,
however, have an extended amino terminus when compared to their respective alleles.
Comparison of the predicted protein products with the corresponding homologues in
T. cruzi Sylvio ×10/1, a non-hybrid strain whose draft genome sequence was made available
more recently 18], and also with the putative homologues in T. brucei and Leishmania spp., suggests that for both genes the amino terminal extension results from annotation
errors (data not shown).
T. cruzi UDP-GlcNAc transporter
It is not possible to determine the substrate specificity of a given NST by analysis
of its amino acid sequence 8]. To identify UDP-GlcNAc transporters, we used a yeast complementation approach in
which a Kluyveromyces lactis mutant deficient in the Golgi transport of UDP-GlcNAc was used 19]. This mutant has been used for the identification of UDP-GlcNAc transporters from
canine cells 20], Caenorhabditis elegans21] and T. brucei13]. Complementation assays are based on the less intense binding of the GS-II lectin
from Griffonia simplicifolia, which specifically recognizes terminal GlcNAc residues, to the mutant cell surface.
Although sequence similarity is not particularly useful for substrate specificity
prediction, we used the UDP-GlcNAc transporters from T. brucei13], a closely related parasite, to restrict our candidate genes. Based on phylogenetic
analysis between the putative T. cruzi transporters and known NSTs from different organisms (Fig. 1), the genes TcCLB.511277.400, TcCLB.504085.60, TcCLB.504057.120 and TcCLB.511517.150
code for the closest transporters to TbNST1, 2, 3 and 4, respectively. These genes,
indicated by arrows in Fig. 1, were then used for the complementation assays in K. lactis.
Fig. 1. Phylogenetic analysis of NSTs from different organisms and the putative T. cruzi NSTs. Experimental characterized substrates are indicated in parentheses. T. cruzi NSTs used for complementation assays in K. lactis are indicated by arrows. Sequences were aligned using ClustalX2.1 and the tree was
constructed with Mega5.2. Statistical method: Neighbor-Joining. Model: p-distance.
Number of bootstraps replications: 1000. AtUDP-GalT1, A. thaliana UDP-GalT1 (21281057); CeGDP-Fuc, C. elegans GDP-Fuc (20138279); CeC03H5.2, C. elegans C03H5.2 (24636210); CnGMT1, C. neoformans GMT1 (68137387); DmGFR, D. melanogaster GFR (7299013); HsSLC35C1, Homo sapiens SLC35C1 (13940506); HsSLC35A2-hUGT2, H. sapiens SLC35A2-hUGT2 (1669566); HsSLC35A1, H. sapiens SLC35A1 (5453621); KlUDP-GlcNAc, K. lactis UDP-GlcNAc (1373152); LmPG2, L. major LPG2 (157876110); LmLPG5A, L. major LPG5A (45649089); LmLPG5B, L. major LPG5B (45649091); ScVRG4, S. cerevisiae VRG4 (6321213); TbNST1, T. brucei NST1 (Tb927.10.13900); TbNST2, T. brucei NST2 (Tb927.11.16560); TbNST3, T. brucei NST3 (Tb927.4.1640); TbNST4, T. brucei NST4 (Tb927.6.3960). GI numbers are indicated in parentheses
Only the transporter encoded by the TcCLB.511517.150 gene, TcNST1, could rescue the
wild-type phenotype (Fig. 2a). Restoration of GSII lectin binding to levels similar to those of wild-type cells
was observed in approximately 50 % of cells transfected with TcNST1. Similar results
have been observed for the heterologous expression of NSTs in yeast, and it is likely
related to the expression of nonfunctional transporters [ 22]. The mean fluorescence intensity values are presented in Fig. 2b. For cells expressing TcNST1, the fluorescence intensity corresponded to approximately
60 % of wild-type levels, a finding that is in agreement with half of the cells not
restoring GS-II lectin binding. As a positive control, we used the K. lactis UDP-GlcNAc transporter (Kl UGT), whose activity is deficient in the yeast mutant
strain. The expression of the other genes in the K. lactis mutant cells did not increase binding of the lectin (Additional file 1). Expression of these genes in K. lactis was analyzed by western blotting using a monoclonal antibody against the histidine
tail inserted at the carboxyl-terminal of the transporters. However, our results regarding
level of expression and apparent molecular weight were not conclusive (data not shown).
Therefore, the incapacity of TcCLB60, TcCLB120 and TcCLB400 to complement the K. lactis mutant could be due to lack of proper expression in the yeast cells.
Fig. 2. Identification of a UDP-GlcNAc transporter (TcNST1) of T. cruzi by in vivo complementation assays. K. lactis mutant (Kl3) cells were transfected with TcNST1, the K. lactis UDP-GlcNAc transporter (Kl UGT, positive control) or empty vector (pE4, negative
control). Wild-type K. lactis cells were included in the analysis for comparison. Cells were grown as described
in the Methods section. After labeling with GS-II lectin (Alexa Fluor 488 conjugate),
yeast cells were separated by flow cytometry in a FACS Canto II (Becton Dickinson).
a Representative histograms showing lectin binding transfected cells. b Mean fluorescence intensities of a representative experiment in triplicate. Error
bars represent the standard deviations. * p??0.01
The TcCLB.511517.150 gene and its allele TcCLB.511737.70 are both composed of 942
nucleotides and code for a 313-amino acid polypeptide. There is only one amino acid
difference between them, in which an asparagine residue is replaced by a serine (N74S)
in TcCLB.511517.150. Sequencing of the gene from T. cruzi clone Dm28c, which was used in this work, revealed a protein identical to its homologue
in T. cruzi Sylvio ×10/1, encoded by the gene TCSYLVIO_003299. This protein differ by three amino
acids to that encoded by the gene TcCLB.511517.150 in T. cruzi CL Brener (K7N, L168V, and L234F in TcCLB.511517.150).
Many NSTs are multisubstrate transporters, including the recently characterized T. brucei transporters. TbNST4 is the closest T. brucei transporter to TcNST1, and it can transport UDP-GlcNAc, UDP-GalNAc and GDP-Man. Next,
we decided to test whether TcNST1 would also be able to transport GDP-Man because
mannose residues are important for T. cruzi glycoconjugates (GalNAc residues are not present in either T. cruzi or T. brucei glycoconjugates). The capacity to transport GDP-Man was assessed by complementation
of a Saccharomyces cerevisiae mutant partially deficient in GDP-Man transport (NDY5 strain) 23] using a fusion protein TcNST1-V5. As shown in Fig. 3a, TcNST1 could not restore growth of the yeast cells in the presence of the anionic
dye Congo red. To rule out the possibility that TcNST1 was not expressed in the yeast
cells, western blot analyses were carried out using an anti-V5 monoclonal antibody.
TcNST1 is visualized in total protein extracts of cells transfected with the vector
encoding TcNST1-V5 but not in cells transfected with the empty vector (Fig. 3b).
Fig. 3. TcNST1 cannot complement a S. cerevisiae mutant strain deficient in GDP-Man transport. aS. cerevisiae strain NDY5 was transfected with TcNST1-V5, the S. cerevisiae GDP-Man Vrg4p transporter (positive control) or empty vector (pYEST-DEST 52, negative
control). Wild-type S. cerevisiae (JPY25 6c) cells were included in the analysis for comparison. Transfected cells
were grown to exponential phase in liquid cultures of SCM-URA, adjusted to an OD of
approximately one (lanes 1 and 2) and submitted to ten-fold serial dilution (lanes
3–5). Cells were then inoculated on solid agar plates containing YP plus 2 % galactose
and 1.6 ?g/ml Congo red. Growth was assessed after 4 days at 30 °C. b Western blot analysis of strain NDY5 transfected with empty vector (lane 1) or TcNST1-V5
(lane 2). TcNST1 expression was detected with a mouse anti-V5 antibody. Expression
of histone 3 was used as an internal control
TcNST1 subcellular localization and gene expression
That TcNST1 could restore UDP-GlcNAc transport in the K. lactis mutant suggested the Golgi apparatus as its probable subcellular localization. To
investigate this hypothesis in T. cruzi, we used a chimeric protein in which green fluorescent protein (GFP) was fused to
the N-terminus of TcNST1. After transfection and selection of neomycin-resistant cells,
T. cruzi epimastigotes were analyzed by fluorescence microscope. Transfected epimastigotes
with the GFP-TcNST1 fusion protein displayed a dot signal at the anterior region of
the parasite while fluorescence emitted by transfectants expressing only GFP-Flag
(GFP fused at the C-terminus with Flag) was distributed through the cytoplasm. Wild-type
cells had little intrinsic fluorescence (Additional file 2). Next we used a monoclonal antibody against GFP to detect expression of GFP-TcNST1
in total extracts of epimastigotes. As shown in Additional file 3, the transporter fused to GFP is present in transfected epimastigotes but not in
wild-type cells. The Golgi localization of TcNST1 was confirmed by co-localization
with TcHIP, a putative T. cruzi zDHHC palmitoyl transferase recently shown to be localized to the Golgi apparatus
24]. A clear punctual signal in close contact with the kinetoplast, as would be expected
for a Golgi resident protein (Fig. 4b, d), was shown to co-localize with TcHIP (Fig. 4c, d).
Fig. 4. TcNST1 is localized to the Golgi apparatus. T. cruzi epimastigotes were transfected with a fusion construct in which TcNST1 was tagged
at the amino terminus with GFP (GFP-TcNST1). Tagged TcNST1 was localized by fluorescence
microscopy after staining with anti-GFP (4b, green) and anti-TcHIP, a Golgi marker (4c, red). Kinetoplasts (large arrow) and nuclei (arrowhead) were stained with DAPI (4d, blue). Secondary antibody controls showed no fluorescence signal (data not shown)
TcNST1 expression was analyzed by semi-quantitative RT-PCR along the life cycle of
T. cruzi and during the process of metacyclogenesis in vitro, in which epimastigotes in the
late exponential phase are differentiated into metacyclics under nutritional stress
in a chemically defined differentiation medium 25]. In this medium, called TAU3AAG, epimastigotes adhere to a substrate and are released
as metacyclics after approximately 100 h. According to the results shown in Fig. 5, TcNST1 mRNA is present in all four major life forms of T. cruzi and in intermediate forms during differentiation from epimastigotes to metacyclics.
While post-transcriptional control of gene expression is particularly important in
trypanosomatids, the presence of mRNA suggests that the transporter may be expressed
in all life forms of the parasite.
Fig. 5. TcNST1 is expressed at the mRNA level in all life forms of T. cruzi and during metacyclogenesis in vitro. RT-PCR analysis of TcNST1 mRNA expression in epimastigotes (Epi), cell-culture-derived trypomastigotes (Trypo),
amastigotes (Ama), metacyclics (Meta) and intermediary forms during metacyclogenesis:
epimastigotes in the late exponential phase (LEP), epimastigotes under nutritional
stress (Stress) and adhered epimastigotes (Adh). Expression of the 60S ribosomal protein
L9 was used as an internal control (TcCLB.504181.10). Aliquots of all samples were
pooled together and subjected to reverse transcription without reverse transcriptase
(? RT)