点击显示 收起
【关键词】 Molecular
Trypanosoma brucei encodes a relatively high number of genes of the equilibrative nucleoside transporter (ENT) family. We report here the cloning and in-depth characterization of one T. brucei brucei ENT member, TbNT9/AT-D. This transporter was expressed in Saccharomyces cerevisiae and displayed a uniquely high affinity for adenosine (Km = 0.068 ± 0.013 µM), as well as broader selectivity for other purine nucleosides in the low micromolar range, but was not inhibited by nucleobases or pyrimidines. This selectivity profile is consistent with the P1 transport activity observed previously in procyclic and long-slender bloodstream T. brucei, apart from the 40-fold higher affinity for adenosine than for inosine. We found that, like the previously investigated P1 activity of long/slender bloodstream trypanosomes, the 3'-hydroxy, 5'-hydroxy, N3, and N7 functional groups contribute to transporter binding. In addition, we show that the 6-position amine group of adenosine, but not the inosine 6-keto group, makes a major contribution to binding (G0 = 12 kJ/mol), explaining the different Km values of the purine nucleosides. We further found that P1 activity in procyclic and long-slender trypanosomes is pharmacologically distinct, and we identified the main gene encoding this activity in procyclic cells as NT10/AT-B. The presence of multiple P1-type nucleoside transport activities in T. brucei brucei facilitates the development of nucleoside-based treatments for African trypanosomiasis and would delay the onset of uptake-related drug resistance to such therapy. We show that both TbNT9/AT-D and NT10/AT-B transport a range of potentially therapeutic nucleoside analogs.
Purine salvage is an essential function for Trypanosoma spp., in that they are unable to synthesize the purine ring de novo (Berens et al., 1995). Consequently, the Trypanosoma brucei brucei nucleoside transporters are of both physiological and pharmacological interest, and several purine transport activities and transporter genes have been identified (de Koning et al., 2005). Most of the pharmacological interest has centered on the TbAT1/P2 aminopurine transporter because of its unusual substrate profile (high affinity for both adenine and adenosine but not for oxypurines or pyrimidines) and its involvement in the accumulation of the major trypanocidal drugs by the parasite (Carter and Fairlamb, 1993; de Koning, 2001). The detailed understanding of the molecular interactions involved in substrate-transporter binding allowed the synthesis of novel trypanocides specifically targeted to the trypanosome (Tye et al., 1998; Soulère et al., 2000; Baliani et al., 2005). One major drawback of this strategy, however, is that loss of TbAT1/P2 activity is already associated with resistance to the current front-line drugs against African trypanosomiasis, better known as sleeping sickness (Delespaux and de Koning, 2007; Lüscher et al., 2007).
In contrast, the P1 transport activity has received little attention, although it has been clear for some time that the T. brucei brucei genome contains multiple equilibrative nucleoside transporter (ENT) family genes encoding P1-type transporters (Sanchez et al., 1999, 2002, 2004), and the completion of the T. brucei brucei genome revealed several additional ENT genes. The simultaneous expression of several related purine nucleoside transporters would be of interest because it would delay the onset of resistance to purine antimetabolites accumulated through multiple transport proteins. Bloodstream trypanosomes are already known to express at least two high-affinity hypoxanthine transporters (de Koning and Jarvis, 1997) that seem to prevent the induction of resistance to the hypoxanthine analog allopurinol (Natto et al., 2005). Second, the presence of multiple nucleoside transporters in the parasite would allow the efficient uptake of a wider spectrum of purine antimetabolites, especially because nutrient transporters of kinetoplastids, including Trypanosoma spp. and Leishmania spp., display much higher affinity for their substrates and are usually energy-dependent and monodirectional (for review, see Lüscher et al., 2007). The pharmacological potential of the P1 activity was very recently demonstrated by Geiser et al. (2005), who showed that 3'-deoxyadenosine (cordycepin) is an extremely potent trypanocide and that its uptake is mediated by both the P1 and P2 activities.
In the current article, we report for the first time that multiple, pharmacologically distinct but very similar P1-type transport activities can be identified in the various T. brucei brucei life cycle stages-distinguished by sensitivity to dipyridamole and 5'-deoxyadenosine. Moreover, we report the cloning of two P1-type transporters from T. brucei brucei and characterize them while expressed in the yeast Saccharomyces cerevisiae. The cloning and partial characterization of one, NT10/AT-B, has previously been reported by Sanchez et al. (2004), and we report here a fuller characterization in yeast as well as a first demonstration of this transport activity in T. brucei brucei as the main P1 transporter in procyclic trypanosomes. We also demonstrate that this transporter is capable of mediating uptake of multiple antimetabolites.
Most notably, we cloned a new P1-type transporter, NT9/AT-D, that showed a uniquely high affinity for adenosine and 2'-deoxyadenosine when expressed in yeast. We present a model explaining this exceptional feature on the basis of the transporter's interactions with its substrate.
Materials. [2-3H]Adenosine (0.92 TBq/mmol), [5,6-3H]uracil (1.78 TBq/mmol), and [8-3H]-hypoxanthine (1.18 TBq/mmol) were purchased from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, UK). [2,8-3H]-Adenine (1.19 TBq/mmol) was bought from PerkinElmer Life and Analytical Sciences (Boston, MA). [3H]7-Deazaadenosine (11 Ci/mmol), [3H]3-deazaadenosine (12.1 Ci/mmol), and [2, 8-3H]inosine (1.23 TBq/mmol) were obtained from Moravek Biochemicals (Brea, CA). Standard chemicals, nucleosides, and nucleobases were obtained from Sigma (St. Louis, MO) unless otherwise stated. 6-Chloropurine riboside and 7-deaza-2'-deoxyadenosine were from TriLink BioTechnologies (San Diego, CA); tubercidin (7-deazaadenosine) and cordycepin (3'-deoxyadenosine) were from Fluka (Buchs, Switzerland); cytidine and adenine-9--D-arabinofuranoside (AraA) were from ICN; 1-Deazaadenosine was a generous gift from Prof. G. J. Koomen (University of Amsterdam, The Netherlands).
Organisms. T. brucei brucei of the pleomorphic strain TREU927 were used throughout this study. The S. cerevisiae strain MG887-1 (fcy2-) used in this project was auxotrophic for uracil and does not possess a functional purine transporter (Gillissen et al., 2000). JM109 competent Escherichia coli were obtained from Promega (Madison, WI).
Cloning of NT9/AT-D and NT10/AT-B. Several ENT homologues were identified from TBLASTN searches of the Wellcome Trust Sanger Institute sequence databases (http://www.sanger.ac.uk/projects/T_brucei). The cloning and characterization of one of these was previously reported as TbNBT1 (Burchmore et al., 2003). Two others were designated AT-B and AT-D and the following PCR primers from regions flanking the ORF were designed: AT-B: forward, 5'-CTTCACCATCTAGCTGAGTCC-3'; reverse, 5'-TTACCTCCTCTTAGGGACAG-3'; AT-D: forward 5'-GATGCTCGGTTTTTATTCAG-3'; reverse, 5'-ATGTCACACTGGTTATTTCAGC-3'. Genomic DNA was isolated from procyclic T. brucei brucei s927 grown in SDM79 culture medium supplemented with 10% fetal calf serum (Brun and Schönenberger, 1979), following standard protocols. The ORFs of AT-B and AT-D were amplified by PCR using Pfu proofreading DNA polymerase (Stratagene, La Jolla, CA), cloned into the vector pGEM-Teasy (Promega) and amplified in E. coli JM109. For both genes, the cloned products of two independent PCR reactions were sequenced on both strands (MWG Biotech); the two independent clones were found to be identical. The cloning and partial characterization of AT-B as TbNT10 has since also been reported by Sanchez et al. (2004).
Expression and Characterization of T. brucei brucei Nucleoside Transporters in S. cerevisiae. The complete AT-B or AT-D ORF was excised from pGEM-Teasy using NotI and subcloned into the NotI site of the yeast expression vector pDR195 (Rentsch et al., 1995), which was transformed into a uracil auxotrophic fcy2-strain of S. cerevisiae, MG887-1, and selected on uracil-free medium, as described previously (Burchmore et al., 2003). Transformed yeast was plated onto yeast nitrogen base without ammonium salts and amino acids, supplemented with 4 mM hypoxanthine or 1 mM adenosine as the sole nitrogen source. Purine transport in (transformed) yeast was performed with cells grown in complete minimal medium without uracil, to a density of 1.5 to 2 OD600 units, as described previously for S. cerevisiae (Burchmore et al., 2003) and T. brucei brucei (Wallace et al., 2002). The method used an incubation of 107 cells with an equal volume of buffer containing radiolabeled permeant and (where indicated) an inhibitor at appropriate concentrations. Incubations are stopped by the addition of excess volume ice-cold unlabeled permeant at saturating concentrations and immediate centrifugation, separating the cells from radiolabel by pelleting them under an oil layer. Nonspecific association of radiolabel (zero uptake) was determined in the presence of 1 mM unlabeled permeant at 0°C and subtracted to yield mediated uptake.
Isolation and Staining of Short-Stumpy Bloodstream Forms. Adult female mice (ICR Swiss strain) were immunocompromised with cyclophosphamide (200 mg/kg) by intraperitoneal injection. After 24 h, mice were inoculated i.p. with 1 x 106 of T. brucei strain 927 bloodstream forms. The blood was collected from the mice after 9 days. Using air-dried blood smear slides for routine Giemsa staining and NAD diaphorase assay, the levels of stumpy form trypanosomes were assessed microscopically. The remainder of the blood was centrifuged at 2500g for 15 min and trypanosomes from the buffy coat layer were purified by DE52 chromatography as described previously (Wallace et al., 2002).
NAD diaphorase assay was performed to confirm the presence of stumpy forms, which react positively, whereas long-slender forms do not (Vickerman, 1985). In brief, reaction solutions were made fresh just before use containing 4 mg of NADH disodium hydrate (Sigma-Aldrich), 0.8 ml of 0.1 M phosphate buffer, pH 7.3, and 1.2 ml of water, and thin air-dried blood smears were prepared. The slides were fixed at 4°C for exactly 5 min with 0.1 M cacodylate buffer, pH 7.2, containing 2.5% glutaraldehyde, and were then rinsed quickly in distilled water. Slides were incubated for 1 h with the reaction solutions (with and without NADH) before being mounted in glycerol and viewed under oil at 1000-fold magnification phase contrast microscopy. T. brucei brucei procyclic forms were used as positive control in this assay.
RT PCR of mRNA Isolated from Different Life Cycle Stages of T. brucei. Parasites from different life cycle stages of T. brucei [1) procyclic from a 2-day old culture of T. brucei strain 427 in SDM79/10% FCS; 2) a mid-log culture of long-slender bloodstream forms of T. brucei strain 427 in HMI-9/20% FCS (Invitrogen, Carlsbad, CA); 3) long-slender or short-stumpy bloodstream forms of T. brucei strain 927 isolated from infected mouse blood on a Whatman DE52 column] were harvested by centrifugation at 1000g for 3 min at 4°C and resuspended at 1 x 107 cells/ml. 1 ml of Tri reagent (Sigma) was added to 1 ml of cell suspension and mixed gently, incubated for 5 min at room temperature, and extracted with RNase-free chloroform (Sigma) and with isopropanol. The RNA was collected by centrifugation (13,000g for 15 min at 4°C) and washed with 500 µl of 75% ethanol. The pellet was resuspended in 30 µl of diethylpyrocarbonate-treated water and incubated at 55°C for 10 min. RNA samples were stored at -70°C for experimental use.
First-strand cDNA synthesis was performed using a SuperScript II reverse transcriptase (RT) (Invitrogen) according to the manufacturer's instructions. Control tubes were treated exactly in the same way but without adding the SuperScript II RT enzyme. Samples of the resulting cDNA and control samples were stored at -20°C until used in RT-PCR with primers to amplify AT-B, AT-D, and TbNBT1. Purity and integrity of the RNA was verified using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA).
Data Analysis. The Prism 3 software package (GraphPad, San Diego, CA) was used to calculate the kinetic parameters, given as mean and S.E, using nonlinear regression. Ki inhibition constants and Gibbs free energy of interaction G0 were calculated as described previously (de Koning and Jarvis, 1999; Wallace et al., 2002). Errors given in tables and shown as bars in graphs are standard errors. In transport experiments, slopes were calculated by linear regression and determined to be significantly different from 0 or not using an F test calculated by the Prism 3 software.
Molecular Cloning of Additional Members of the ENT Family from T. brucei brucei. TBLASTN searches, using the amino acid sequence of TbAT1, of genomic DNA sequence databases assembled at the Wellcome Trust Sanger Institute (http://www.sanger.ac.uk/Projects/T_brucei/) yielded several unknown ORFs that aligned with known ENT family transporters TbNT2 (Sanchez et al., 1999) and TbAT1 (Mäser et al., 1999). These ORFs were provisionally named AT-A through AT-E. AT-B and AT-D seem to belong to a phylogenetic subgroup with the P1-type nucleoside transporters, TbNT2 through TbNT7, and purine nucleoside transporters from Crithidia fasciculata and Leishmania donovani (de Koning et al., 2005). Hydropathy plots (using TMPRED, http://www.ch.embnet.org/software/TMPRED_form.html) suggested structures with 11 transmembrane domains and a large intracellular loop, consistent with the general ENT topology (Sankar et al., 2002). The amino acid sequences of AT-B and AT-D are displayed alongside TbNT2 and TbAT1 in Fig. 1 of the supplementary material available on-line.
Fig. 1. Expression of NT10/AT-B and NT9/AT-D in different life-cycle stages of Trypanosoma brucei brucei. A, amplification of the NT9/AT-D and NT10/AT-B fragments from procyclic forms of T. brucei brucei s427 with the expected size of 1.4 kb. B, represents the gel using long-slender bloodstream forms, showing no bands for AT-B or AT-D. C, amplification of NT10/AT-B fragment from short-stumpy bloodstream form of T. brucei 927 with the expected size of 1.4 kb. The same fragment of TbNBT1 (positive control) was amplified in all stages of the life cycle with a size of 0.7 kb. The Invitrogen 1-kb DNA ladder was used to estimate sizes. Negative control was a parallel preparation produced in the absence of reverse transcriptase.
AT-B has since been independently reported as NT10 by Sanchez et al. (2004), and the AT-D sequence has also been identified as a potential ENT family homolog by Landfear et al. (2004) and named NT9. We will from here on refer to them as NT9/AT-D and NT10/AT-B, respectively. The two transporters were amplified by PCR and cloned into the vector pGEM-T Easy (Promega) as described above under Cloning of NT9/AT-D and NT10/AT-B, and subcloned into the yeast expression vector pDR195 as described previously (Burchmore et al., 2003). Orientation and integrity of the construct was verified using an appropriate series of restriction digests, after which the constructs were transformed into the Saccharomyces cerevisiae strain MG887.1 (Burchmore et al., 2003).
Expression of NT9/AT-D and NT10/AT-B during the Trypanosome Life Cycle. To assess expression of NT9/AT-D and NT10/AT-B in the various T. brucei life cycle stages, RT-PCR was performed on mRNA isolated from cultured procyclic forms (PC, s927), cultured long-slender forms (LS, EATRO 427 monomorphic strain) and short-stumpy bloodstream forms (SS, s927 pleomorphic strain) isolated from infected mice at peak parasitemia. The nucleobase transporter gene TbNBT1 served as positive control because it is known to be expressed in PC and LS forms (Burchmore et al., 2003). PCR products of the predicted size for NT9/AT-D, NT10/AT-B, and TbNBT1 were identified from mRNA of PC trypanosomes (Fig. 1A), whereas only the TbNBT1 product was amplified from LS mRNA (Fig. 1B). The presence of a high proportion of SS in the infected blood preparation was confirmed using the NAD diaphorase assay (Vickerman, 1985). All procyclic T. brucei brucei were stained in this assay, and more than 80% of parasites in the SS preparation (see Fig. 2 of Supplementary material). Because the LS trypanosomes do not express NT9 or NT10, any product amplified with NT9 or NT10 primers must be derived from SS. Only NT10/AT-B and TbNBT1 expression was detected in this preparation (Fig. 1C).
Fig. 2. Transport of [3H]adenosine by AT-B in S. cerevisiae strain MG887-1. Uptake of 0.25 µM[3H]adenosine was linear over 120 s (: r2 = 0.96, linear regression). In the presence of 1 mM unlabeled adenosine () or the control (pDR195 vector, ), uptake was not significantly different from zero (P > 0.4 and P > 0.2, respectively; F test).
Fig. 3. Transport of 0.05 µM [3H]adenosine (A), 0.1 µM [3H]inosine (B), 2.0 µM [3H]tubercidin (C), and 1.0 µM [3H]3-deazaadenosine by S. cerevisiae transformed with pDR195:NT10 (D). Representative experiments are shown, each performed in triplicate, out of a total of three to four independent Km determinations. Label concentrations were chosen well below anticipated Km values. Transport was expressed as picomoles per 107 cells per second.
Characterization of NT10/AT-B by Heterologous Expression in S. cerevisiae and the Corresponding Transport Activity in T. b. brucei. Expression of NT10/AT-B in S. cerevisiae strain MG887-1 enabled these cells to rapidly accumulate 0.25 µM [3H]adenosine (Fig. 2), with a rate of 0.080 ± 0.011 pmol/107 cells/s, and linear for at least 120 s. This accumulation, which was absent in cells transformed with the same vector without the NT10/AT-B insert, seemed to be mediated by a high-affinity transporter, in that it was completely inhibited by coincubation with 1 mM unlabeled adenosine (Fig. 2). The Km (adenosine) for this transporter was determined as 0.41 ± 0.03 µM (n = 5), using a very low concentration of 0.015 µM[3H]adenosine supplemented with increasing concentrations of unlabeled adenosine (Fig. 3A). Km values were further determined for [3H]inosine, [3H]tubercidin, and [3H]3-deazaadenosine (Fig. 3, B-D; Table 1). For each of these permeants, Ki values for adenosine were also determined as a positive control, yielding values close to the [3H]adenosine Km value. The Ki value for guanosine was 2.4 ± 0.28 µM (n = 3).
TABLE 1 Km and Vmax values for P1-type transporters
P1 activity, traditionally ascribed to the TbNT2 gene product and related genes on chromosome 2, and possibly made up of the combined flux through several different P1-type transporters (Sanchez et al., 2002), was measured in T. brucei brucei bloodstream forms as described previously (de Koning and Jarvis, 1999). NT9/AT-D and NT10/AT-B were expressed in S. cerevisiae, and transport of the various permeants measured as described under Materials and Methods. Results are the average and S.E. of three to five experiments, each performed in triplicate.
The results show that although NT10/AT-B has a high affinity for purine nucleosides, this is much reduced when the H-bond-accepting ring nitrogen residues, N3 and N7, are replaced by carbon atoms. We further determined that the high-affinity binding of TbNT10/AT-B is to a large extent dependent on a strong interaction with the 3'-hydroxyl group [Ki (3'-deoxyadenosine) = 31 ± 6.0 µM; n = 4], whereas the 2'-deoxy and 5'-deoxy groups were much less important for binding [Ki (2'-deoxyadenosine) = 1.0 ± 0.3 µM (n = 4); Ki (5'-deoxyadenosine) = 0.9 ± 0.1 µM (n = 3)]. Pyrimidine nucleosides and xanthosine had no effect on NT10/AT-B-mediated transport, and affinity for adenine and hypoxanthine was very low, with mean Ki values of 150 ± 42 and 320 ± 80 µM, respectively (n = 3).
Procyclic Trypanosomes Express Multiple Adenosine Transport Activities. Because NT10/AT-B is expressed in procyclic trypanosomes and pharmacologically distinct from the P1 activity of LS T. brucei brucei (Table 1), the question arises whether PC P1 is encoded by NT10/AT-B. A reinvestigation of adenosine transport in PC T. brucei brucei found that PC P1 is highly sensitive to 5'-deoxyadenosine (Fig. 4A) and displayed a Km (adenosine) value of 0.15 ± 0.02 µM (n = 4), significantly lower than for NT10/AT-B (P < 0.001, Student t test). The observed Km value could be the net result of adenosine uptake through multiple transporters with different Km values. In four independent experiments, dipyridamole inhibited adenosine in procyclic T. brucei brucei by 50% (Fig. 4A), with a Ki value of 0.64 ± 0.09 µM. Transport of 0.015 µM [3H]adenosine was inhibited by unlabeled adenosine and 2'-deoxyadenosine in a monophasic way, with a Ki value for 2'-deoxyadenosine of 0.42 ± 0.04 µM (Fig. 4B).
Fig. 4. Inhibition of [3H]adenosine transport by dipyridamole and 5'-deoxyadenosine in procyclic and long-slender bloodstream forms. A, uptake of 0.05 µM [3H]adenosine over 10 s in procyclic T. brucei brucei s427 in the presence or absence of various concentrations of dipyridamole (), 5'-deoxyadenosine () and adenosine (). B, identical experiment using LS bloodstream forms from T. brucei brucei s427. Uptake was expressed as picomoles per 107 cells per second.
NT9/AT-D Encodes a Purine Nucleoside Transporter with Exceptionally High Affinity for Adenosine. No transport of [3H]uracil, [3H]hypoxanthine, or [3H]adenine could be detected in yeast MG887-1 transformed with pDR195:AT-D (time course over 5 min; label concentration, 0.5 µM). Likewise, no transport of 2.5 µM [3H]pentamidine was observed, even over 180 min (data not shown). However, transport of 0.25 µM [3H]adenosine was shown to be linear over 120 s, with a rate of 0.0079 ± 0.0005 pmol/107 cells/s (linear regression, r2 > 0.99), which was >95% inhibited by 1 mM unlabeled adenosine. As it was becoming clear that AT-D encodes an extremely high-affinity transporter, the experiment was repeated with 0.015 µM [3H]adenosine, a concentration required to obtain valid Km and Ki values for this transporter. Figure 5A shows that transport was linear for at least 4 min (r2 = 0.97), and although it was only 65% inhibited by 1 mM adenine, it was fully inhibited by 1 mM unlabeled adenosine. Likewise, transport of 0.25 µM [3H]inosine by AT-D-transformed yeast was fully saturable (Fig. 5B), with a rate of 0.0096 ± 0.0005 pmol/107 cells/s.
Fig. 5. Transport of purine nucleosides by NT9/AT-D expressed in Saccharomyces cerevisiae. A, transport of 0.015 µM [3H]adenosine in the presence of 1 mM adenine (), 1 mM adenosine (), or without inhibitor (). B, transport of 0.25 µM [3H]inosine in the presence () or absence () of 1 mM unlabeled adenine. Both experiments were performed in triplicate. Error bars are S.E. and, when not shown, fall within the symbols. Transport in the presence of excess permeant was not significantly different from zero (P > 0.05; F test). Transport units are picomoles per 107 cells.
NT9/AT-D displayed extremely high affinity for adenosine, approximately 40-fold higher than for inosine, but the Vmax was much higher for inosine (Fig. 6, Table 1). The transporter displayed very low affinity to nucleobases, including adenine (Table 2).
Fig. 6. Determination of adenosine and inosine Km values for NT9/AT-D expressed in yeast. A, uptake of 0.015 µM [3H]adenosine in the presence of various concentrations adenosine () or inosine (). Inset, conversion of the adenosine inhibition data to a Michaelis-Menten curve. B, uptake of 0.05 µM [3H]inosine in the presence of various concentrations of adenosine () or inosine (). Inset, conversion of the inosine inhibition data to a Michaelis-Menten curve. Both experiments were performed in triplicate; error bars are S.E. Transport units are picomoles per 107 cells per second.
TABLE 2 Kinetic parameters and Gibbs free energies of the interactions of potential substrates and inhibitors with NT9/AT-D
Substrate Binding by NT9/AT-D. The model for substrate binding by P1-type transporters (de Koning and Jarvis, 1999; de Koning, 2001) predicts a similar binding energy for inosine and adenosine. We therefore undertook to investigate the molecular basis of the uniquely high adenosine binding energy (G0) of 42.5 kJ/mol. Whereas the ribose group of adenosine is known not to be involved in binding of the P2 transporter (de Koning and Jarvis, 1999), the ribose moiety made an important contribution to binding NT9/AT-D. 3'-Deoxyadenosine (cordycepin) and 5'-deoxyadenosine both displayed significantly reduced affinity for this transporter compared with adenosine (P < 0.02; Student t test) (Fig. 7A), corresponding with a reduced Gibbs free energy (G0) of 11.5 and 4.0 kJ/mol, respectively. In contrast, the 2'-deoxy group was not required for binding, with the Ki value for 2'-deoxyadenosine not significantly different from the Km value for adenosine (Table 2; Fig. 7A).
Fig. 7. Inhibition of adenosine transport by NT9/AT-D expressed in S. cerevisiae. Uptake of 0.05 µM [3H]adenosine in yeast strain MG887-1 transformed with pDR195:AT-D was determined over 60 s in the presence of various purine analogs. Transport was expressed as percentage of control value, being transport in the absence of inhibitor. A, 2'-deoxyadenosine (), 3'-deoxyadenosine (), 5'-deoxyadenosine (). B, 1-deazaadenosine (), tubercidin (), nebularine (). Data shown are combined from representative single experiments performed in triplicate.
Of the functional groups of the purine ring itself, N3 and N7 contributed significantly to binding (P < 0.05), with (G0) values of 3.95 and 9.8 kJ/mol, respectively, relative to adenosine, calculated from the Ki values of 3-deazaadenosine and tubercidin (7-deazaadenosine)(Fig. 7B). Any contribution of N1 to adenosine binding was found to be small: 1-deazaadenosine displayed a Ki value of 0.21 ± 0.07 µM, which was not statistically different from the adenosine Km. The calculated (G0) for 1-deazaadenosine was 2.9 kJ/mol, possibly attributable to a weakening of the N3 H-bond acceptor in 1-deazaadenosine rather than disruption of a weak H-bond at N1 itself. In contrast, the 6-position amine group seemed to make a major contribution, with nebularine (purine riboside) and 6-chloropurine riboside displaying significantly higher Ki values (P < 0.02 and P < 0.01, respectively) than the adenosine Km (Fig. 7B), resulting in (G0) values of 10.5 and 13.5 kJ/mol. It is possible that the partial negative charge of the chloride residue in 6-chloroadenosine has an energetically negative effect on the binding energy, in that the 6-amine group presumably functions as a H-bond donor to a partially negative charge in the binding pocket. Figure 8 summarizes the model for adenosine binding by NT9/AT-D, using the mean (G0) value of nebularine and 6-chloroadenosine for the interaction energy at position 6. According to this model, the sum of the Gibbs free energy for the individual interactions is 41.3 kJ/mol, which is within 3% of the G0 calculated from the adenosine Km (Table 2).
Fig. 8. Model for the binding of adenosine (top) and inosine (bottom) by the P1 transport activity measured in long-slender bloodstream forms or by NT9/AT-D expressed in S. cerevisiae MG887-1. Estimated Gibbs free energy of interaction, G0, for the indicated functional groups is given in kilojoules per mole.
It could further be noted from the results listed in Table 2 that substitutions at position 2 of the purine ring are generally unfavorable. Guanosine displayed a significantly lower Ki value than inosine Km (P < 0.05), and the addition of a nitro group at position 2 reduced the binding energy of adenosine by more than 18 kJ/mol.
The purine salvage system, prominently including the primary step, transmembrane transport, has received much attention in many pathogenic organisms, both as drug target and as a conduit for the delivery of novel therapeutic nucleoside analogs (reviewed by de Koning et al., 2005; Lüscher et al., 2007). Similar reasoning has created much interest in the nucleoside transporters in neoplastic tissues (Huang et al., 2004) and in viral infections (Pastor-Anglada et al., 2005). The Trypanosoma brucei nucleoside transporters have thus far been divided in P1- and P2-type transporters, according to the classification introduced by Carter and Fairlamb (1993). Both P1 and P2 transporters efficiently transport numerous purine nucleoside analogs, whereas P2 additionally transports the first-line sleeping sickness drugs pentamidine and melarsoprol, which make the T. brucei nucleoside transporters of great pharmacological significance (Geiser et al., 2005; Lüscher et al., 2006). The work on T. brucei brucei purine transporters has contributed very significantly to our understanding of the ENT family (de Koning et al., 2005). We now report the cloning and characterization of an ENT nucleoside transporter with a uniquely high affinity for adenosine and the elucidation of the mechanism of binding responsible for this phenomenon.
Biochemical Basis of Unique Substrate-Binding Profile by NT9/AT-D. We found that NT9/AT-D, when expressed in S. cerevisiae, displayed an extraordinarily high affinity for adenosine and 2'-deoxyadenosine. We are not aware of any other nucleoside transporter with as high an affinity for its substrate(s). It also displays 40-fold lower affinity for inosine, which sets it apart in the P1-class of ENT transporters, although phylogenetic analysis clearly placed it in this category (de Koning et al., 2005). The strong selectivity of adenosine over inosine, in particular, could not be explained on the basis of the binding model for P1 transporters based on kinetic analysis of the P1-activity in LS trypanosomes (de Koning and Jarvis, 1999; de Koning, 2001). We thus examined which functional groups contribute to adenosine and inosine binding to NT9/AT-D. We found that in addition to the established four interactions between P1 transporters and purine nucleosides, 3'-hydroxy, 5'-hydroxy, N3 and N7-which do not differentiate between oxypurine and aminopurine nucleosides-NT9/AT-D also interacts strongly with the 6-position amine group of adenosine, presumably through one or more hydrogen bonds. This interaction explains the much higher affinity for adenosine than other P1 transporters, which apparently lack this interaction. In addition, a possible interaction with the pyrimidine N1 of adenosine was identified, though this did not reach statistical significance. The binding of the N1-C6-NH2 motif is the characteristic signature of P2 that enables it to bind diamidines and melaminophenyl arsenicals as well as aminopurines (de Koning and Jarvis, 1999; Barrett and Fairlamb, 1999).
Assuming an identical orientation for oxypurines and aminopurines in the NT9/AT-D binding pocket, inosine interactions would involve the 3'-hydroxy, 5'-hydroxy, N3, and N7 functional groups with energies similar to those for adenosine. The 6-keto group does not seem to be involved in binding, judging from the near-identical binding energies for 6-mercaptopurine riboside and inosine (Table 2). It is also not likely that the lactam hydrogen N1(H) engages in hydrogen bonding, because nebularine and inosine have almost the same calculated binding energy (Table 2). The difference between the adenosine and inosine binding energies, 9.6 kJ/mol, can thus be explained by the existence of a strong H-bond between the 6-amino group and the NT9/AT-D binding pocket, and a possible weak H-bond with N1. This effectively constitutes a fusion of the P1 and P2 binding motifs.
Demonstration of Multiple Distinct P1-type Transport Activities in T. brucei. A P1-type adenosine and inosine transport activity was first described in T. brucei brucei bloodstream forms (Carter and Fairlamb, 1993) and later in procyclic trypanosomes (de Koning et al., 1998). To date, it has been unclear how many adenosine transporters are functionally expressed in the various life cycle forms, particularly because in trypanosomes, which express most genes constitutively in large polycistrons, the presence of mRNA does not necessarily indicate the presence of the protein product (Clayton, 2002). The number of distinct adenosine transporters is relevant from a pharmacological perspective, and certainly in the context of induction of drug resistance, but is difficult to assess on a protein level, although efforts to establish a T. brucei brucei plasma membrane subproteome for bloodstream forms are in progress (D. J. Bridges, A. R. Pitt, H. P. Voorheis, H. P. de Koning, and R. J. S. Burchmore, unpublished). Here, we show from both the inhibitor profiles of LS P1, PC P1, and NT10/AT-B expressed in yeast that 1) whereas NT10/AT-B encodes a P1-type transporter, it is not the P1 activity expressed in LS forms and 2) that the P1 transport activity in PC forms (de Koning et al., 1998) is subtly different from the corresponding transporter in LS forms.
This follows an earlier observation that substrate binding by LS P1, but not PC P1, is negatively affected by substitutions at position 2 of the purine ring (2-chloroadenosine; de Koning et al., 1998; de Koning and Jarvis, 1999) and that only PC P1 is sensitive to inhibition by the nucleoside transport inhibitor dipyridamole (de Koning et al., 1998). We now show that adenosine transport in procyclics is inhibited by 5'-deoxyadenosine in a monophasic manner, with a Ki value consistent with this transporter being NT10/AT-B (but very different from the 5'-dAdo Ki value of LS P1) but is only partially inhibited by dipyridamole. These observations strongly suggest that procyclic trypanosomes express at least two highly similar purine nucleoside transporters, including NT10, distinguished by their sensitivity to dipyridamole and different from the P1 adenosine transporter of T. brucei brucei LS bloodstream forms.
These conclusions are consistent with the results of the RT-PCR analysis, which indicate that NT10/AT-B is not expressed in LS forms, but that NT10 mRNA is present in procyclic and short stumpy forms, and in agreement with the observations of Sanchez et al. (2004). In unpublished work, Roditi and coworkers constructed an NT10-/- T. brucei brucei line, which was only partly resistant to the nucleoside antimetabolites tubercidin and cordycepin (I. Roditi, personal communication). This observation is further evidence that transport-related resistance against purine antimetabolites will be very hard to induce in Trypanosoma brucei spp. after unsuccessful attempts to induce allopurinol resistance in bloodstream form T. brucei brucei (Natto et al., 2005).
NT9/AT-D mRNA was also found in procyclic trypanosomes, but we were unable to demonstrate the activity of this transporter in PC T. brucei brucei, even when using very low concentrations of labeled adenosine to favor uptake through this carrier. This may be because the expression of NT9/AT-D is relatively low under standard culture conditions, coupled with the low capacity associated with such high substrate affinity. However, it is likely that NT9/AT-D is of physiological value to the procyclic parasite, which, in its environment of the tsetse fly midgut, must periodically experience very low purine concentrations. We have previously demonstrated that a high-affinity hypoxanthine transporter of procyclic T. brucei brucei is up-regulated as a result of purine-depleted conditions (de Koning et al., 2000). Similar up-regulation of purine salvage proteins has been demonstrated in several other kinetoplastids (Gero et al., 1997; Seyfang and Landfear, 1999; Liu et al., 2005).
We were unable to definitively identify the second P1-type transport activity observed in procyclic cells. Phylogenetic analysis of ENT family genes in T. brucei brucei suggests the existence of eight such genes (de Koning et al., 2005): NT9/AT-D, NT10/AT-B, and the cluster of six highly similar copies, NT2 to NT7, on chromosome 2, of which NT3 and NT4 are reportedly nonfunctional and NT5 to NT7 are mostly expressed in bloodstream forms rather than procyclic forms (Sanchez et al., 2002). By a process of elimination, we thus arrive at the hypothesis that NT2 and NT10/AT-B together constitute adenosine transport in procyclic T. brucei brucei under purine replete conditions.
In conclusion, we have cloned and characterized two nucleoside transporters from African trypanosomes and shown the biochemical basis of their unique substrate binding profiles. We have further shown that these transporters mediate the uptake of pharmacologically relevant antimetabolites and provided new insights in the complexities of nucleoside transporter usage by this protozoan parasite.
Acknowledgements
We are grateful to Dr. Richard J. S. Burchmore for expert technical advice and helpful discussions.
ABBREVIATIONS: ENT, equilibrative nucleoside transporter; AT, adenosine transporter; RT, reverse transcriptase; ORF, open reading frame; LS, long slender; SS, short stumpy; PC, procyclic; kb, kilobase(s).
The online version of this article (available at http://molpharm.aspetjournals.org) contains supplemental material.
【参考文献】
Baliani A, Bueno GJ, Stewart ML, Yardley V, Brun R, Barrett MP, and Gilbert IH (2005) Design and synthesis of a series of melamine-based nitroheterocycles with activity against trypanosomatid parasites. J Med Chem 48: 5570-5579.
Barrett MP and Fairlamb AH (1999) The biochemical basis of arsenical-diamidine cross-resistance in African trypanosomes. Parasitol Today 15: 136-140.
Berens RL, Krug EC, and Marr JJ (1995) Purine and pyrimidine metabolism, in Biochemistry and Molecular Biology of Parasites (Marr JJ and Muller M eds) pp 89-117, Academic Press, London.
Brun R and Sch?nenberger M (1979) Cultivation and in vitro cloning of procyclic culture forms in Trypanosoma brucei in a semi-defined medium. Acta Trop 36: 289-292.
Burchmore R, Wallace LJM, Candlish D, Al-Salabi MI, Beal P, Barrett MP, Baldwin SA, and de Koning HP (2003) Cloning, heterologous expression and in situ characterization of the first high affinity nucleobase transporter from a protozoan. J Biol Chem 278: 23502-23507.[Abstract/Free Full Text]
Carter NS and Fairlamb AH (1993) Arsenical-resistant trypanosomes lack an unusual adenosine transporter. Nature (Lond) 361: 173-176.
Clayton CE (2002) Life without transcriptional control? From fly to man and back again. EMBO (Eur Mol Biol Organ) J 21: 1881-1888.
de Koning HP (2001) Transporters in African trypanosomes: role in drug action and resistance. Int J Parasitol 31: 512-522.
de Koning HP and Jarvis SM (1997) Purine nucleobase transport in bloodstream forms of Trypanosoma brucei brucei is mediated by two novel transporters. Mol Biochem Parasitol 89: 245-258.
de Koning HP and Jarvis SM (1999) Adenosine transporters in bloodstream forms of Trypanosoma brucei brucei: substrate recognition motifs and affinity for trypanocidal drugs. Mol Pharmacol 56: 1162-1170.[Abstract/Free Full Text]
de Koning HP, Bridges DJ, and Burchmore RJS (2005) Purine transporters of protozoa: from biology to therapy. FEMS Microbiol Rev 29: 987-1020.
de Koning HP, Watson CJ, and Jarvis SM (1998) Characterisation of a nucleoside/proton symporter in procyclic Trypanosoma brucei brucei. J Biol Chem 273: 9486-9494.[Abstract/Free Full Text]
de Koning HP, Watson CJ, Sutcliffe L, and Jarvis SM (2000) Differential regulation of nucleoside and nucleobase transport in Crithidia fasciculata and Trypanosoma brucei in response to purine stress. Mol Biochem Parasitol 106: 93-107.
Delespaux V and de Koning HP (2007) Drugs and drug resistance in African trypanosomiasis. Drug Resist Updat, in press.
Geiser F, Lüscher A, de Koning HP, Seebeck T, and M?ser P (2005) Molecular pharmacology of adenosine transport in Trypanosoma brucei: P1/P2 revisited. Mol Pharmacol 68: 589-595.[Abstract/Free Full Text]
Gero AM, Day RE, and Hall ST (1997) Stimulated transport of adenosine, guanosine and hypoxanthine in Crithidia luciliae: metabolic machinery in which the parasite has a distinct advantage over the host. Int J Parasitol 27: 241-249.
Gillissen B, Bürkle L, André B, Kühn C, Rentsch D, Brandl B, and Frommer WF (2000) A new family of high-affinity transporters for adenine, cytosine, and purine derivatives in Arabidopsis. Plant Cell 12: 291-300.[Abstract/Free Full Text]
Huang Y, Anderle P, Bussey KJ, Barbacioru C, Shankavaram U, Dai Z, Reinhold WC, Papp A, Weinstein JN, and Sadée W (2004) Membrane transporters and channels: role of the transportome in cancer chemosensitivity and chemoresistance. Cancer Res 64: 4294-4301.[Abstract/Free Full Text]
Landfear SM, Ullman B, Carter NS, and Sanchez MA (2004) Nucleoside and nucleobase transporters in parasitic protozoa. Eukaryot Cell 3: 245-254.[Free Full Text]
Liu W, Arendt CS, Gessford SK, Ntaba D, Carter NS, and Ullman B (2005) identification and characterization of purine nucleoside transporters from Crithidia fasciculata. Mol Biochem Parasitol 140: 1-12.
Lüscher A, de Koning HP, and M?ser P (2007) Chemotherapeutic strategies against Trypanosoma brucei: drug targets vs. drug targeting. Curr Pharm Des 13: 555-567.
M?ser P, Sütterlin C, Kralli A, and Kaminsky R (1999) A nucleoside transporter from Trypanosoma brucei involved in drug resistance. Science (Wash DC) 285: 242-244.[Abstract/Free Full Text]
Natto M, Wallace LJM, Candlish D, Al Salabi MI, Coutts SE, and de Koning HP (2005) Trypanosoma brucei: expression of multiple purine transporters prevents the development of allopurinol resistance. Exp Parasitol 109: 80-86.
Pastor-Anglada M, Cano-Solado P, Molina-Arcas M, Lostao MP, Larráyoz I, Martínez-Picado J, and Casado FJ (2005) Virus Res 107: 151-164.
Rentsch D, Laloi M, Rouhara I, Schmeltzer E, Delrot S, and Frommer WB (1995) NTR1 encodes a high affinity oligopeptide transporter in Arabidopsis. FEBS Lett 370: 264-268.
Sanchez MA, Drutman S, Van Ampting M, Matthews K, and Landfear SM (2004) A novel purine nucleoside transporter whose expression is up-regulated in the shortstumpy form of the Trypanosoma brucei life cycle. Mol Biochem Parasitol 136: 265-272.
Sanchez MA, Tryon R, Green J, Boor I, and Landfear SM (2002) Six related nucleoside/nucleobase transporters from Trypanosoma brucei exhibit distinct biochemical functions. J Biol Chem 277: 21499-21504.[Abstract/Free Full Text]
Sanchez MA, Ullman B, Landfear SM, and Carter NS (1999) Cloning and functional expression of a gene encoding a P1 type nucleoside transporter from Trypanosoma brucei. J Biol Chem 274: 30244-30249.[Abstract/Free Full Text]
Sankar N, Machado J, Abdulla P, Hilliker AJ, and Coe IR (2002) Comparative genomic analysis of equilibrative nucleoside transporters suggests conserved protein structure despite limited sequence identity. Nucleic Acids Res 30: 4339-4350.[Abstract/Free Full Text]
Seyfang A and Landfear SM (1999) Substrate depletion upregulates uptake of myo-inositol, glucose and adenosine in Leishmania. Mol Biochem Parasitol 104: 121-130.
Soulère L, Hoffmann P, Bringaud F, and Périé J (2000) Synthesis and uptake of nitric oxide-releasing drugs by the P2 nucleoside transporter in Trypanosoma equiperdum. Bioorg Med Chem Lett 10: 1247-1350.
Tye C, Kasinathan G, Barrett MP, Brun R, Doyle VE, Fairlamb AH, Weaver R, and Gibert IH (1998) An approach to use an unusual adenosine transporter to selectively deliver polyamine analogues to trypanosomes. Bioorg Med Chem Lett 8: 811-816.
Vickerman K (1985) Developmental cycles and biology of pathogenic trypanosomes. Br Med Bull 41: 105-114.[Abstract/Free Full Text]
Wallace LJ, Candlish D, and de Koning HP (2002) Different substrate recognition motifs of human and trypanosome nucleobase transporters. Selective uptake of purine antimetabolites. J Biol Chem 277: 26149-26156.[Abstract/Free Full Text]
作者单位：Institute of Biomedical and Life Sciences, Division of Infection and Immunity, Glasgow Biomedical Research Centre, University of Glasgow, Glasgow, United Kingdom (M.I.A., L.J.M.W., D.C., B.R., M.K.G., I.J., S.N.A., H.P.d.K.); and Institute of Cell Biology, University of Bern, Switzerland (A.L., P.M.