Identification and Characterization of a Secondary ... - ACS Publications

Article pubs.acs.org/molecularpharmaceutics

Identification and Characterization of a Secondary Sodium-Binding Site and the Main Selectivity Determinants in the Human Concentrative Nucleoside Transporter 3 C. Arimany-Nardi,†,‡ A. Claudio-Montero,†,‡ A. Viel-Oliva,†,‡ P. Schmidtke,§ C. Estarellas,⊥ X. Barril,§,∥ A. Bidon-Chanal,*,⊥ and M. Pastor-Anglada*,†,‡ †

Departament de Bioquímica i Biomedicina Molecular, Facultat de Biologia and Institute of Biomedicine (IBUB), Universitat de Barcelona, 08028 Barcelona, Spain ‡ Oncology Program, National Biomedical Research Institute on Liver and Gastrointestinal Diseases (CIBER EHD), Instituto de Salud Carlos III, 28029 Madrid, Spain § Departament de Farmàcia i Tecnologia Farmacèutica, i Fisicoquímica, Facultat de Farmàcia i Ciències de l′Alimentació and Institute of Biomedicine (IBUB), Universitat de Barcelona, 08028 Barcelona, Spain ∥ Institució Catalana de Recerca i Estudis Avançats (ICREA), 08010 Barcelona, Spain ⊥ Departament de Nutrició, Ciències de l′Alimentació i Gastronomia, Facultat de Farmàcia i Ciències de l′Alimentació and Institute of Biomedicine (IBUB), Campus de l′Alimentació de Torribera, Universitat de Barcelona, 08921 Santa Coloma de Gramenet, Spain S Supporting Information *

ABSTRACT: The family of concentrative Na+/nucleoside cotransporters in humans is constituted by three subtypes, namely, hCNT1, hCNT2, and hCNT3. Besides their different nucleoside selectivity, hCNT1 and hCNT2 have a Na+/nucleoside stoichiometry of 1:1, while for hCNT3 it is 2:1. This distinct stoichiometry of subtype 3 might hint the existence of a secondary sodium-binding site that is not present in the other two subtypes, but to date their three-dimensional structures remain unknown and the residues implicated in Na+ binding are unclear. In this work, we have identified and characterized the Na+ binding sites of hCNT3 by combining molecular modeling and mutagenesis studies. A model of the transporter was obtained by homology modeling, and key residues of two sodium-binding sites were identified and verified with a mutagenesis strategy. The structural model explains the altered sodium-binding properties of the hCNT3C602R polymorphic variant and supports previously generated data identifying the determinant residues of nucleoside selectivity, paving the way to understand how drugs can target this plasma membrane transporter. KEYWORDS: concentrative transporters, homology modeling, sodium-binding site, selectivity, nucleoside transport

■

INTRODUCTION Human Concentrative Nucleoside Transporter (hCNT) proteins are encoded by the SLC28 family genes and are responsible for the unidirectional, Na+-coupled uptake into cells of natural nucleosides and a broad variety of nucleoside-derived drugs used in anticancer and antiviral therapies.1−4 The three members of this family, hCNT1, hCNT2, and hCNT3, differ in their selectivity and specificity and, even more importantly, in their Na+/nucleoside stoichiometry, which is 1:1 for hCNT1 and hCNT2, but strikingly 2:1 for hCNT3.1,4 Indeed, the latter shows outstanding properties as a drug transporter, including broad substrate selectivity, accepting both purine and pyrimidine nucleosides, high affinity, and high concentrative transport capacity as a result of its 2:1 Na+/nucleoside stoichiometry.5−7 hCNTs are phylogenetically older than the other nucleoside transporter gene family (SLC29), known to encode human Equilibrative Nucleoside Transporter (hENT) proteins, because the latter do not have prokaryote orthologs, © XXXX American Chemical Society

whereas the former do. Among its three members, hCNT3 is more closely related to other more primitive orthologs.5 Indeed, the CNT protein expressed by the primitive pacific hagfish (Eptatetrus stoutii) has greater sequence similarities with hCNT3 than with the other two human subtypes. More importantly, at the functional level hfCNT shows identical stoichiometry to hCNT3 (i.e., 2:1, Na+/nucleoside) and is equally able to transport a broad variety of nucleosides, either purine or pyrimidine derived.8 Consistent with this view, SLC28 genes are poorly polymorphic in humans.7 This is even more evident for hCNT3, which anticipates that this transporter may be relevant to human fitness, with a function relevant for cell physiology.9,10 Nevertheless, we were able to Received: Revised: Accepted: Published: A

February April 17, April 25, April 25,

2, 2017 2017 2017 2017 DOI: 10.1021/acs.molpharmaceut.7b00085 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article

Molecular Pharmaceutics

μCi/ml) uptake was measured in HEK293 by incubating the cells either in a sodium-repleted (137 mM NaCl) or a sodiumfree (137 mM choline chloride) transport buffer, also containing 5.4 mM KCl, 1.8 mM CaCl2, 1.2 mM MgSO4, and 10 mM HEPES, pH 7.4. Uptake measurements were performed at 1 min, a routine time point in this cell system corresponding to initial velocity conditions in our hands. These assays were terminated by washing the cells off with an excess volume of chilled stop buffer (137 mM NaCl, 10 mM HEPES, pH 7.4). All assays were run in triplicate at room temperature (21−23 °C), and each assay corresponded to an independent cell culture and transfection experiment. Endogenous nucleoside transport activity of HEK293 cells was fully accounted for by hENT-type transporters (Na+independent). This is supported by the evidence that this cell line lacks Na+-coupled transport activity, being uptake rates measured in the presence of sodium similar to the ones determined in the absence of this cation. hENT-related transport represented a 20% of the hCNT-related activity induced after transfection of the wild type protein. Indeed, hCNT3 activity was calculated by subtracting uptake rates measured in the choline chloride medium (sodium-free medium) from uptake rates determined in the presence of sodium. Essentially, the hCNT signal in this cell system was very strong, repetitive, and suitable for this type of assays. For the determination of Hill coefficients, Na+-activation curves were generated by increasing the amounts of this cation in the media (from 0 to 100 mM), using identical amounts of choline to maintain their overall osmolarity in the absence of sodium. Western Blot. Twenty micrograms of total protein extracts obtained by lysing transfected HEK293 cells with NP40 buffer (50 mM Tris-HCl pH 7.5; 150 mM NaCl; 1% NP40; 5 mM sodium pyrophosphate; 50 mM NaF; 1 mM Na3VO4; and protease inhibitor) were separated on 10% polyacrylamide gels and transferred to nitrocellulose membranes (Merck Millipore; Billerica, MA). Membranes were incubated with anti-hCNT3 (Sigma-Aldrich) diluted 1:2000. Proteins were detected by using a secondary antibody (1:2000) conjugated to a horseradish peroxidase and a chemioluminiscence detection kit (Biological Industries; Kibbutz Beit-Haemek, Israel). Expression levels were normalized to those for α-actin (1:5000) (Sigma-Aldrich). Statistical Analysis. The unpaired t-Student’s test was used for the statistical comparison of experimental data. These analyses were carried out using GraphPad Prism 4.0 software. Homology Modeling. The model of the hCNT3 transporter was built using the crystal structure of vcCNT (PDB ID: 3TIJ).13 The sequence alignment between the crystal structure and the target hCNT3 sequence was performed using MOE (CCG).15 Next, the homology modeling capabilities within MOE were used to derive a series of homology models. For all of them, overall contact energies were calculated, and the best model was used throughout this study. Furthermore, phi−psi angle populations on the model were verified, by computing the Ramachandran plot (see Supplementary Figure S5), and the angle, distance, and contact energy properties throughout the structure were checked for abnormal values using as a reference the distributions of these parameters found in the high resolution entries of the Protein Data Bank (www.rcsb.org)16 following the protocol implemented in MOE (see Supplementary Table S5).

identify a novel hCNT3 polymorphic variant in Spanish population, hCNT3C602R, which showed an allelic frequency of 1%.11,12 This genetic variant retains function and shows affinity interaction constants unaltered for natural nucleosides, but variable for some nucleoside-derived drugs. Moreover, the Na+ binding properties of hCNT3C602R appear to be modified. In fact, analysis of the Na+-activation curves of the WT-protein and the C602R variant revealed a shift in the Hill coefficient from 2 to 1 in the polymorphic transporter.11 This anticipates that one of the two Na+ binding sites is somewhat hampered in this genetic variant. At the time we functionally characterized this transport variant, C602 was predicted to locate within the last transmembrane helix of hCNT3, TMD 13, but it was not evident how this residue could contribute to Na+ binding. Interestingly, the prokaryote orthologs of hCNT proteins show high homology with the mammalian, and human, CNTs.1−4 The Vibrio cholera CNT is 39% homologous to hCNT3 and has been crystallized in its substrate-bound (Na+ and uridine) occluded inward-facing conformation.13 In this study we have modeled the substrate binding pocket of hCNT3 as a way to better understand the structural determinants for Na+ binding and also for nucleoside docking to the transporter. The model has proven useful to understand how Na+ binds and fully explains the functional alteration of the hCNT3C602R genetic variant. Furthermore, it also provides a template on which to further analyze how this broad selectivity drug transporter may interact with a broad range of drugs covering a broad spectrum of diseases.

■

EXPERIMENTAL SECTION Reagents. Cytidine was obtained from Sigma-Aldrich (St. Louis, MO). [5-3H(N)]-Cytidine was obtained from Hartmann Analytic GmbH (Braunschweig, Germany). All other chemicals were of analytical grade. Molecular Biology. The hCNT3 cDNA (GenBank accession number AF305210) was cloned from human kidney.14 The polymorphic substitutions were introduced using the primers shown in Supplementary Table 1. All generated site-directed mutations and constructs were further confirmed by DNA sequencing (BigDye Terminator v3.1, Applied Biosystems, Foster City, CA). Cell Culture. Human embryonic kidney 293 cells (HEK293) were routinely maintained at 37 °C/5% CO2 in Dulbecco’s modified Eagle’s medium (Lonza Verviers SPRL, Verviers, Belgium) supplemented with 10% heat-inactivated fetal bovine serum (v/v), 2 mM glutamine, and a mixture of antibiotics (100 U penicillin, 0.1 mg/mL streptomycin, and 0.25 mg/mL fungizone) (Life Technologies, Paisley, UK). Transfection. hCNT3 wild type and the generated mutants were transiently expressed in HEK293 cells. HEK293 cells were chosen for this purpose because they lack endogenous hCNTtype activity (see below), thereby being an excellent background on which to express these transporter proteins. HEK293 cells were transiently transfected using calcium phosphate, as previously described.14 Proper expression has been verified by Western blot as summarized below. Nucleoside uptake experiments were carried out 40 h after transfection, as follows. Nucleoside Uptake Measurements. Twenty-four-multiwell plates were treated with a solution of poly-D-lysine in PBS during 4 h before seeding cells. Uptake experiments were performed 24 h after seeding. [5-3H(N)]-Cytidine (1 μM, 1 B

DOI: 10.1021/acs.molpharmaceut.7b00085 Mol. Pharmaceutics XXXX, XXX, XXX−XXX

Article


Figure 1. First sodium binding site activity. Panel (A) shows the measured activity of the WT hCNT3 and the mutants located at the first sodiumbinding site. Results are the mean of three independent experiments run in triplicate. Panel (B) shows the activity corrected by protein expression. Both graphs represent mean ± SEM of three single experiments. (C) First sodium binding site. Position occupied by the first sodium binding site with respect to the nucleoside binding pocket (in this case, with uridine bound). Residues from the X-ray crystal structure of vcCNT3 are shown in green, while that of hCNT3 are shown in white. Red spheres, labeled as WAT, represent water molecules present in the crystal structure. (D) N336A mutation. When asparagine 336 is mutated to alanine, one of the surrounding water molecules, represented as blue spheres, might fit in the space occupied by its side chain. *p < 0.05; **p < 0.01.

■

Molecular Docking. Binding mode prediction of hCNT3 substrates were obtained with the molecular docking programs GOLD17,18 and GLIDE.19,20 In both cases the binding site was defined as a 12 Å side cube, whose center was the centroid of residues Glu156 and Glu332. The cube enclosed the uridine molecule found in the crystal structure with PDB ID 3TIJ. The 3D structures of the ligands were obtained from the Zinc database21 with standard protonation at pH = 7.0, and the downloaded geometries were used as such. Options for free exploration of rotatable bonds and generation of different ring conformations were enabled in both, Glide and Gold docking experiments. Molecular Dynamics. The models obtained were used to run molecular dynamics (MD) simulations embedding the protein in lipid bilayer formed of 180 POPC lipid molecules using the CHARMM-GUI server (http://www.charmm-gui. org).22 The protein orientation with respect to the membrane was obtained by superposition to the OMP database oriented structure of vcCNT. The system was then solvated with 14,070 TIP3P water molecules, and seven calcium ions were added to neutralize the system. The Charmm3623 force field parameters were used to model the energetics of the system, and the simulations were run with the cuda version of the pmemd module of Amber14.24 Each modeled system contained around 73,000 atoms. The systems were minimized and equilibrated following the protocols taken from CHARMM-GUI, and the simulations were run to produce trajectories of 100 ns.

RESULTS AND DISCUSSION Identification of the Primary Sodium Binding Site. Taking advantage of the recently resolved crystal structure of a prokaryotic nucleoside concentrative transporter vcCNT,13 a model for the hCNT3 was built up by using sequence homology modeling and was used to explore the residues implicated in Na+ and nucleoside binding. The primary sodiumbinding site was defined by homology with its site in the vcCNT structure. It is worth noting that the percentage of identity is high around this site (>80%), which makes the model in this region particularly trustworthy (see Supplementary Figure S1). Four amino acids were identified as forming part of the coordination sphere of the ion, namely, N336, V339, T370, and I371 (N149, V152, S183, and I184 in V. cholerae) (see Figure 1C). The role of these residues in Na+ binding was confirmed by measuring the activity of three different mutants. On the first two, T370 and N336 were individually mutated to alanine. Since the ion coordinates with the hydroxyl (T370) and amide (N336) oxygen atoms, the mutation should abrogate the coordination position and destabilize the complex. The mutation I371E was aimed at disrupting the whole sodium-binding cavity by adding a negatively charged residue in an apolar environment adjacent to the ion binding-site cavity. Experimental results in HEK293 cells transiently transfected with the mutated transporters showed that mutants T370A and N336A presented uptake rates similar to WT, whereas the I371E mutant presented no transport activity (see Figure 1A). C


Article


Figure 2. Na+ stoichiometry measurements. Left upper panel: Cytidine uptake by hCNT3 WT or its mutants measured at increasing concentrations of sodium. Bottom panels: Linearization of the Na+-activation curves to calculate Hill coefficients. Right upper panel: Hill coefficients for each transporter protein. These results are the mean ± SEM of 3 (hCNT3 and N336A), 4 (T370A and S396A) and 5 (T605A) independent experiments run in triplicates. *p < 0.05.

Figure 3. Secondary binding site. Initial (left) and final (right) coordination spheres of the sodium-binding sites in hCNT3. The residues that coordinate both sodium ions are represented as sticks, and the water molecules participating in the coordination of the first and second sodium ions are represented as red spheres labeled WAT. Sodium ions appear as purple spheres labeled Na+.

Surprisingly, the value for N336A was 2 ± 0.03 (n = 3; mean ± SEM) (see Figure 2). In principle, this result would mean that N336 does not participate in the Na+ coordination, as it indeed does in vcCNT. However, analysis of the crystal structure shows that a chain of water molecules surrounds the coordination residue N149 (N336 in hCNT3) and that the void volume generated by the N336A mutation might be filled by a water molecule, which could instead participate in Na+ binding (Figure 1D). Interestingly, in molecular dynamics simulations run to relax the structure obtained with homology modeling, the side chain of N336 rotates and a water molecule enters the cavity to coordinate the sodium ion (see Figure 3). Assuming that the model is correct and that the mutation may not impact on the stoichiometry (i.e., concentrative capacity), the high degree of conservation of Asn at this location through evolution may be explained by the observed loss of uptake capacity, but also by additional factors such as providing structural stability or ion selectivity. This latter factor is particularly likely as it is known that selectivity in ionic channels and pumps results from a fairly rigid geometry of the binding site, closely matching the radius of the bound ion.26

To ensure that the transiently transfected HEK293 cells properly expressed all proteins, an antibody against the Nterminal tail was used to detect hCNT3 proteins by Western blot in cell extracts (see Supplementary Figure S2). Transient transfection routinely resulted in most of the expressed protein being localized at the plasma membrane, as deduced from membrane biotinylation assays (see Supplementary Figure S3 for a representative experiment), in accordance with previous work from our laboratory using confocal microscopy to monitor proper plasma membrane protein insertion of transiently transfected hCNT3.11,25 Expression levels were semiquantified by densitometry and uptake rates corrected by the protein expression levels of each hCNT3 protein normalized to its corresponding actin. Normalized activity of the mutants was approximately one-third of that of the WT protein (see Figures 1B and 2). As expected, the results for the Hill coefficient measurement, which is independent of the expression levels of each transporter protein, showed values of 2 ± 0.1 for hCNT3 WT and 0.9 ± 0.1 for the T370A mutant, in agreement with the loss of one coordination site upon mutation of T370. D


Article


Figure 4. Key residues of the binding site. Superposition of the three subtypes’ binding site region showing the most important residues for nucleoside binding in each of them. As a reference, the position occupied by uridine in the crystal structure of the vcCNT3 is shown as semitransparent in A. In B−D the main determinant substitutions to explain the different selectivity are shown highlighted with a yellow edge box for uridine (B), thymidine (C), and inosine (D).

0.05 for T605A (n = 4 and 5, respectively; mean ± SEM) (see Figure 2). This would be consistent with a shift in the Na+/ nucleoside stoichiometry from 2:1 to 1:1, supporting a perturbation in the sodium-binding capacity of the transporter. Interestingly, T605 is substituted by an asparagine in hCNT1 (N578) and by serine in hCNT2 (S583), two residues that are usually found within those that participate in the coordination of sodium.25 However, S396 is substituted by alanine in both hCNT1 (A369) and hCNT2 (A374) disrupting the possibility to bind sodium (see Supplementary Figure S4). When molecular dynamics simulations were run to relax the model, some side chains surrounding the secondary sodiumbinding site rotated, and consequently, the coordination sphere changed with respect to that obtained with homology modeling. Specifically, the side chain of S369, an amino acid that coordinated the sodium ion in the homology model with the oxygen atom of its backbone carbonyl group, rotated toward the sodium ion, and as a consequence of this rotation, the hydroxylic oxygen atom of the amino acid side chain became coordinated to the sodium ion. Interestingly, the mutation of the residue had a direct effect on the Hill coefficient. The mutation S369A, changed the Hill coefficient from 2:1 to 1:1, underlining the role of the side chain of S369

Identification of the Secondary Sodium-Binding Site. The substitution C602R is a polymorphism of hCNT3 likely to modify the Na+/nucleoside stoichiometry of the transporter from 2:1 to 1:1,11 presenting a Hill coefficient of 1 instead of 2. However, Cys602 does not participate in the primary sodiumbinding site. Instead, it is located in a small adjacent cavity shaped by its side chain, the side chains of S396 and T605, and the main-chain carbonyl units of Y558 and C561. The cavity is also present in the crystal structure of vcCNT and contains a water molecule inside; however, it is formed by residues A209, N405, and A408 instead of S396, C602, and T605 (see Figure 3). These substitutions increase the polarity of the cavity in hCNT3, supporting the possibility to host sodium inside. Coordination of Na+ with the S atom of Cys is not unheard of, but relatively rare,27 and the evolutionary preference for Cys over Ser (a much better Na+ coordination group) may hint at additional functional roles. To validate the hypothesis, two of the putative residues of the second sodium-binding site, T605 and S396, were mutated to alanine. Expression of both mutants was checked by Western blot, and it was observed to be similar to that of the wild type transporter (see Figure 2). Hill coefficients were calculated for these variants, and the results were 1 ± 0.1 for S396A and 1.1 ± E


Article


of this nucleoside is hydrogen-bonded to S546/541/568. Subtypes 1 and 3 can compensate this interaction with the hydrogen bond formed with Q319/341, but subtype 2 cannot, as this residue is a methionine instead. Furthermore, the interaction of the O4 carbonyl oxygen with S345 in thymidine is partially impeded by the methyl group attached to C5 (Figure 4C). Finally, clarification of the different preferences for pyrimidine and purine nucleosides regarding our structural models demands a closer look to the inosine−transporter complexes. Here, with the premise that the sugar position is the same as in uridine, the purine ring would occupy the same pocket as the pyrimidine ring. The results of the docking experiments show a clear coincidence between the position occupied by inosine in hCNT2 and hCNT3 that is not favorable for hCNT1. Thus, for subtypes 2 and 3, several poses within the 10 best scored have the sugar ring properly positioned, while in subtype 1 a proper position of the sugar is ranked as the 28th best scoring. Inspection of the predicted complexes clearly highlights that the presence of S318 in hCNT1 instead of the corresponding glycine in hCNT2 and hCNT3 represents a clear steric hindrance to the positioning of the purine ring and explains why mutation of this residue to glycine enables purine transport, albeit to a low degree, in hCNT1 (see Figure 4D). Overall, the results of the docking experiments show a consistency of our structural models with the available experimentally data and validate them.29 However, for a comprehensive understanding of the whole substrate translocation cycle of hCNT proteins, multiple crystals of the transporter in all possible intermediate states would be required. Unfortunately, only one conformation of the transporter, that of the Vibrio cholera nucleoside transporter in the inward facing occluded state, has been crystallized so far. This fact surely limits the validity of our model to explain the full translocation process, but not to understand the structural differences that modulate nucleoside affinity between subtypes.

as part of the coordination sphere of the sodium ion in the secondary sodium-binding site. However, the mutation T605A also changed the Hill coefficient from 2:1 to 1:1, but molecular dynamics simulations showed that a water molecule occupies the hydroxyl group position, and therefore, it indirectly participates in the coordination of the sodium ion by providing stability to the interacting water molecule. Nucleoside Binding Pocket: Model Agrees with Previously Published Data. The three hCNT subtypes present particular differences in the residues conforming the nucleoside binding pocket that are linked to their variable substrate selectivity. Loewen et al.28 identified a series of changes in the amino acid sequence between hCNT1 and hCNT2 that appeared to be responsible for their selectivity against pyrimidine and purine based nucleosides, respectively. Specifically, the sole mutation of S318 in hCNT1 to the glycine in hCNT2 enabled the transport of purine-like nucleosides in the former subtype. The double mutant S318G/Q319M increased the hCNT1 purine-like nucleoside transport capabilities and converted the transporter into a broad specificity concentrative nucleoside transporter. With a third mutation, S352T, the broad specificity of the hCNT1 double mutant was clearly diminished. Finally, the full transformation was achieved introducing a fourth mutation, L353V, converting the hCNT1 transporter into hCNT2-like. To put these findings on the perspective of our model, we generated homology models for hCNT1 and hCNT2, following the same protocol as for the hCNT3 transporter. On these models we performed docking experiments to find the most reliable nucleoside−protein complexes for uridine, thymidine, and inosine, which were used by Loewen et al. to measure the selectivity of hCNT transporters and their mutants.28 We took as a reference the position occupied by uridine in the structure of vcCNT3 (PDB ID: 3TIJ) and focused our attention to poses in which the superposition of the sugar ring between the docked ligand and uridine was maximized. A structural comparison of the binding site between the three subtypes (see Figure 4A) shows the position of the aforementioned residues and indicates its key role in substrate selectivity. Furthermore, it identifies other residues that have not been experimentally probed but might also contribute to nucleoside binding. As expected, S318 and Q319 in hCNT1 superpose with G313 and M314 in hCNT2 and with the G340/Q341 pair in hCNT3. Furthermore, S352 and L353 in hCNT1 superpose with T347 and V348 of hCNT2 and with S374 and V375 in hCNT3. Their distinct role in nucleoside binding can be envisaged from the docking experiments and clarifies, at least from a static structural point of view, the different selectivity of the three transporter subtypes. The three transporters bind and transport uridine efficiently; subtype 3 also transports both thymidine and inosine, while subtype 1 cannot transport inosine and subtype 2 cannot transport thymidine. For subtypes 1 and 3, a hydrogen bond interaction between the O2 carbonylic oxygen and the −NH2 group of the side chain of Q319/341 stabilizes the pyrimidine ring of uridine in the binding site. This interaction is absent in subtype 2 due to the Q319M substitution, but might be compensated by a favorable hydrogen-bond interaction between the O4 carbonyl oxygen and Ser345, which is an alanine in hCNT1 and hCNT3 (see Figure 4B). Regarding thymidine, the 2′-dehydroxylated position might represent an important interaction loss in the three subtypes with respect to uridine, as the hydroxyl group in the sugar ring

■

CONCLUSIONS The human Concentrative Nucleoside Transporter 3 (hCNT3) protein has been modeled based upon a single available prokaryotic hCNT ortholog, the one crystallized in its substrate loaded, inward facing, occluded conformation. The generated model clearly explains previous experimental evidence identifying key residues contributing to substrate selectivity and provides key information on the sodium binding sites, which has been experimentally validated in this contribution. Considering hCNT3 is expressed in most epithelial barriers and shows broad substrate selectivity for nucleoside-derived drugs, we anticipate this model represents the first step in the understanding of how already available drugs and others being developed can target this plasma membrane transporter.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.molpharmaceut.7b00085. Primers used for site-directed mutagenesis; multiple sequence alignment of the region surrounding the first sodium-binding; hCNT3 and hCNT3 mutants expression levels; Western blot analysis of the biotinylated F


Article


■

(7) Smith, K. M.; Slugoski, M. D.; Cass, C. E.; Baldwin, S. A.; Karpinski, E.; Young, J. D. Cation coupling properties of human concentrative nucleoside transporters hCNT1, hCNT2 and hCNT3. Mol. Membr. Biol. 2007, 24, 53−64. (8) Yao, S. Y.; Ng, A. M.; Loewen, S. K.; Cass, C. E.; Baldwin, S. A.; Young, J. D. An ancient prevertebrate Na+-nucleoside cotransporter (hfCNT) from the Pacific hagfish (Eptatretus stouti). Am. J. Physiol. 2002, 283, C155−C168. (9) Errasti-Murugarren, E.; Pastor-Anglada, M. Drug transporter pharmacogenetics in nucleoside-based therapies. Pharmacogenomics 2010, 11, 809−841. (10) Badagnani, I.; Chan, W.; Castro, R. A.; Brett, C. M.; Huang, C. C.; Stryke, D.; Kawamoto, M.; Johns, S. J.; Ferrin, T. E.; Carlson, E. J.; Burchard, E. G.; Giacomini, K. M. Functional analysis of genetic variants in the human concentrative nucleoside transporter 3 (CNT3; SLC28A3). Pharmacogenomics J. 2005, 5, 157−165. (11) Errasti-Murugarren, E.; Cano-Soldado, P.; Pastor-Anglada, M.; Casado, F. J. Functional characterization of a nucleoside-derived drug transporter variant (hCNT3C602R) showing altered sodium-binding capacity. Mol. Pharmacol. 2008, 73, 379−386. (12) Errasti-Murugarren, E.; Molina-Arcas, M.; Casado, F. J.; PastorAnglada, M. The human concentrative nucleoside transporter-3 C602R variant shows impaired sorting to lipid rafts and altered specificity for nucleoside-derived drugs. Mol. Pharmacol. 2010, 78, 157−165. (13) Johnson, Z. L.; Cheong, C. G.; Lee, S. Y. Crystal structure of a concentrative nucleoside transporter from Vibrio cholera. Nature 2012, 483, 489−493. (14) Errasti-Murugarren, E.; Pastor Anglada, M.; Casado, J. F. Role of CNT3 in the transepithelial flux of nucleosides and nucleoside-derived drugs. J. Physiol. 2007, 3, 1249−1260. (15) Molecular Operating Environment (MOE), 2013.08; Chemical Computing Group Inc.: Montreal, QC, Canada, 2016. (16) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data Bank. Nucleic Acids Res. 2000, 28, 235−242. (17) Jones, G.; Willett, P.; Glen, R. C. Molecular recognition of receptor sites using a genetic algorithm with a description of desolvation. J. Mol. Biol. 1995, 245, 43−53. (18) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. Development and validation of a genetic algorithm for flexible docking. J. Mol. Biol. 1997, 267, 727−748. (19) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shaw, D. E.; Shelley, M.; Perry, J. K.; Francis, P.; Shenkin, P. S. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment of Docking Accuracy. J. Med. Chem. 2004, 47, 1739−1749. (20) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W. T.; Banks, J. L. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 2. Enrichment Factors in Database Screening. J. Med. Chem. 2004, 47, 1750−1759. (21) Irwin, J. J.; Shoichet, B. K. ZINC−a free database of commercially available compounds for virtual screening. J. Chem. Inf. Model. 2005, 45, 177−182. (22) Jo, S.; Kim, T.; Iyer, V. G.; Im, W. CHARMM-GUI: A Webbased Graphical User Interface for CHARMM. J. Comput. Chem. 2008, 29, 1859−1865. (23) Best, R. B.; Zhu, X.; Shim, J.; Lopes, P. E. M.; Mittal, F.; Feig, M.; MacKerell, A. D., Jr. Optimization of the Additive CHARMM AllAtom Protein Force Field Targeting Improved Sampling of the Backbone φ, ψ and Side-Chain χ1 and χ2 Dihedral Angles. J. Chem. Theory Comput. 2012, 8, 3257−3273. (24) Case, A.; Babin, V.; Berryman, J. T.; Betz, R. M.; Cai, Q.; Cerutti, D. S.; Cheatham, T. E.; Darden, T. A.; Duke, R. E.; Gohlke, H.; Goetz, A. W.; Gusarov, S.; Homeyer, N.; Janowski, P.; Kaus, J.; I., Kolossváry, Kovalenko, A.; Lee, T. S.; LeGrand, S.; Luchko, T.; Luo, R.; Madej, B.; Merz, K. M.; Paesani, F.; Roe, D. R.; Roitberg, A.; Sagui, C.; R., Salomon-Ferrer, Seabra, G.; Simmerling, C. L.; Smith, W.;

membrane proteins; amino acid substitutions around the secondary sodium binding site; Ramachandran plot for the hCNT3 model; residues with nonexpected bond angles for the hCNT3 model (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Tel: +34 934033788. *E-mail: [email protected]. Tel: +34 934021543. ORCID

A. Bidon-Chanal: 0000-0002-1666-1490 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS X.B. and A.B.-C. thank the Spanish Ministerio de Economia (grants SAF2012-33481 and SAF2014-57094-R) and the Catalan government (grant 2014SGR1189) for financial support. The UB laboratory of the Faculty of Biology is a member of the Oncology Program of the National Biomedical Research Institute of Liver and Gastrointestinal Diseases (CIBER ehd). CIBER ehd is an initiative of Instituto de Salud Carlos III (Spain). This study was supported by research funding from Spanish Secretariat of Research (MINECO), grants SAF2011-23660 and SAF2014-52067-R to M.P.-A. C.A.N. was the recipient of predoctoral fellowships FPI from Ministerio de Ciencia e Innovación. The Barcelona Supercomputing Center is acknowledged for providing access to computational resources.

■

ABBREVIATIONS hCNT, human Concentrative Nucleoside Transporter; hCNT1, human Concentrative Nucleoside Transporter subtype 1; hCNT2, human Concentrative Nucleoside Transporter subtype 2; hCNT3, human Concentrative Nucleoside Transporter subtype 3; hfCNT, hagfish Concentrative Nucleoside Transporter; WT, wild type; vcCNT, Vibrio cholera Concentrative Nucleoside Transporter

■

REFERENCES

(1) Pastor-Anglada, M.; Cano-Soldado, P.; Errasti-Murugarren, E.; Casado, F. SLC28 genes and concentrative nucleoside transporter (CNT) proteins. Xenobiotica 2008, 338, 972−994. (2) Molina-Arcas, M.; Casado, F. J.; Pastor-Anglada, M. Nucleoside transporter proteins. Curr. Vasc. Pharmacol. 2009, 4, 426−434. (3) Cano-Soldado, P.; Pastor-Anglada, M. Transporters that translocate nucleosides and structural similar drugs: structural requirements for substrate recognition. Med. Res. Rev. 2012, 32, 428−457. (4) Young, J. D.; Yao, S. Y.; Baldwin, J. M.; Cass, C. E.; Baldwin, S. A. The human concentrative and equilibrative nucleoside transporter families, SLC28 and SLC29. Mol. Aspects Med. 2013, 34, 529−547. (5) Ritzel, M. W.; Ng, A. M.; Yao, S. Y.; Graham, K.; Loewen, S. K.; Smith, K. M.; Ritzel, R. G.; Mowles, D. A.; Carpenter, P.; Chen, X. Z.; Karpinski, E.; Hyde, R. J.; Baldwin, S. A.; Cass, C. E.; Young, J. D. Molecular identification and characterization of novel human and mouse concentrative Na+-nucleoside cotransporter proteins (hCNT3 and mCNT3) broadly selective for purine and pyrimidine nucleosides (system cib). J. Biol. Chem. 2001, 276, 2914−2927. (6) Smith, K. M.; Slugoski, M. D.; Loewen, S. K.; Ng, A. M.; Yao, S. Y.; Chen, X. Z.; Karpinski, E.; Cass, C. E.; Baldwin, S. A.; Young, J. D. The broadly selective human Na+/nucleoside cotransporter (hCNT3) exhibits novel cation-coupled nucleoside transport characteristics. J. Biol. Chem. 2005, 280, 25436−25449. G


Article

Molecular Pharmaceutics Swails, W. J., Wang, J.; Wolf, R. M.; Wu, X.; Kollman, P. A. Amber 14; University of California: San Francisco, 2014. (25) Errasti-Murugarren, E.; Casado, F. J.; Pastor-Anglada, M. Different N-terminal motifs determine plasma membrane targeting of the human concentrative nucleoside transporter 3 in polarized and nonpolarized cells. Mol. Pharmacol. 2010, 78, 795−803. (26) Page, M. J.; Di Cera, E. Role of Na+ and K+ in Enzyme Function. Physiol. Rev. 2006, 86, 1049−1092. (27) Zheng, H.; Chruszcz, M.; Lasota, P.; Lebioda, L.; Minor, W. Data mining of metal ion environments present in protein structures. J. Inorg. Biochem. 2008, 9, 1765−1776. (28) Loewen, S. K.; Ng, A. M.; Yao, S. Y.; Cass, C. E.; Baldwin, S. A.; Young, J. D. () Identification of amino acid residues responsible for the pyrimidine and purine nucleoside specificities of human concentrative Na+ nucleoside cotransporters hCNT1 and hCNT2. J. Biol. Chem. 1999, 274, 24475−24484. (29) Johnson, Z. L.; Lee, J. H.; Lee, K.; Lee, M.; Kwon, D. Y.; Hong, J.; Lee, S. Y. Structural basis of nucleoside and nucleoside drug selectivity by concentrative nucleoside transporters. eLife 2014, 3, e03604.

H


Identification and Characterization of a Secondary ... - ACS Publications

Recommend Documents