Visual Overview
Abstract
Equilibrative nucleoside transporters (ENTs) mediate the transmembrane flux of endogenous nucleosides and nucleoside analogs used clinically. The predominant subtype, ENT1, has been well characterized. However, the other subtype, ENT2, has been less well characterized in its native milieu due to its relatively low expression and the confounding influence of coexpressed ENT1. We created a cell model where ENT1 was removed from human embryonic kidney (HEK293) cells using CRISPR/cas9 [ENT1 knockout (KO) cells]; this cell line has ENT2 as the only functional purine transporter. Transporter function was assessed through measurement of [3H]2-chloroadenosine uptake. ENT1 protein was quantified based on the binding of [3H]nitrobenzylthioinosine, and ENT1/ENT2 protein was detected by immunoblotting. Changes in expression of relevant transporters and enzymes involved in purine metabolism were examined by quantitative polymerase chain reaction. Wild-type HEK293 cells and ENT1KO cells had a similar expression of SLC29A2/ENT2 transcript/protein and ENT2-mediated [3H]2-chloroadenosine transport activity (Vmax values of 1.02 ± 0.06 and 1.50 ± 0.22 pmol/μl/s, respectively). Of the endogenous nucleosides/nucleobases tested, adenosine had the highest affinity (Ki) for ENT2 (2.6 μM), while hypoxanthine was the only nucleobase with a submillimolar affinity (320 μM). A range of nucleoside/nucleobase analogs were also tested for their affinity for ENT2 in this model, with affinities (Ki) ranging from 8.6 μM for ticagrelor to 2,300 μM for 6-mercaptopurine. Our data suggest that the removal of endogenous ENT1 from these cells does not change the expression or function of ENT2. This cell line should prove useful for the analysis of novel drugs acting via ENT2 and to study ENT2 regulation.
SIGNIFICANCE STATEMENT We have created a cell line whereby endogenous ENT2 can be studied in detail in the absence of the confounding influence of ENT1. Loss of ENT1 has no impact on the expression and function of ENT2. This novel cell line will provide an ideal model for studying drug interactions with ENT2 as well as the cellular regulation of ENT2 expression and function.
Introduction
The SLC29A-encoded family of equilibrative nucleoside transporters (ENTs) mediate the bidirectional flux of nucleosides (and some nucleobases) across cell membranes and are important for the maintenance of intracellular and extracellular nucleoside levels (Boswell-Casteel and Hays, 2017). There are four subtypes of ENT (ENT1–4) that vary in their substrate specificity, inhibitor sensitivity, and subcellular expression (Baldwin et al., 1999; Pastor-Anglada and Pérez-Torras, 2018). ENT1, 2, and 4 are targeted, at least in part, to the plasma membrane, while ENT3 is expressed only in intracellular compartments such as lysosomes (Baldwin et al., 2005). ENT1 and ENT2 have a broad substrate selectivity including both endogenous purine and pyrimidine nucleosides (Ward et al., 2000), and nucleoside analogs used as anticancer (Griffiths et al., 1997; Damaraju et al., 2003; Hioki et al., 2018) and antiviral drugs (Karbanova et al., 2017). ENT4, in contrast, is selective for adenosine and adenosine analogs (Tandio et al., 2019), in addition to a range of monoamines and organic cations, and is better known as the plasma membrane monoamine transporter (Wang, 2016).
Of the two broad-spectrum ENT subtypes (ENT1, ENT2), ENT1 has been studied the most due to its relative predominance and expression in most tissues (Löffler et al., 2007; Endres et al., 2009; Boswell-Casteel and Hays, 2017) and the availability of the highly selective ENT1 inhibitor, nitrobenzylthioinosine (NBMPR) (Cass et al., 1974; Paterson et al., 1977; Hammond, 1991b). In contrast, there is much less information on how drugs interact with endogenous ENT2 or factors that affect ENT2 expression or activity. ENT2 is expressed at low levels, relative to ENT1, and typically concomitantly with ENT1 (Cass et al., 2007). There is also no selective inhibitor available for ENT2. There is emerging evidence that ENT2 has its own physiological roles distinct from ENT1 (Naes et al., 2020). ENT2 transports antiretroviral drugs used in HIV therapies (Yao et al., 2001), and SLC29A2, the gene encoding ENT2, has been identified in HIV-1 target cells such as myeloid cells (Hoque et al., 2021). ENT2 is the key mediator of anti-DNA autoantibody (tumor-targeting) passage across the blood-brain barrier, thereby allowing brain tumor immunotherapy (Rattray et al., 2021). The intestinal epithelial ENT2 has been proposed as a target for inflammatory bowel disease in terms of limiting tissue inflammation (Morote-Garcia et al., 2009). ENT1 and ENT2 can also form homodimers and heterodimers (Jarvis et al., 1980; Grañé-Boladeras et al., 2019; https://era.library.ualberta.ca/items/dc2ddc23-208f-4362-ac9c-8d385e094d47), and interactions between these two subtypes have been shown to impact their plasma membrane trafficking and sensitivity to phosphorylation by protein kinase C (Grañé-Boladeras et al., 2019). It has also been shown that splice variants of ENT2 form hetero-oligomers at nuclear membranes and play important roles in cell proliferation (Grañé-Boladeras et al., 2016). Thus, a more defined understanding of ENT2 function and regulation is clearly warranted.
One of the issues with studying ENT2 has been the lack of a model system whereby it can be assessed in isolation in its endogenous milieu. To examine ENT2 function and pharmacology in situ it has been necessary to block ENT1-mediated activity either with the selective inhibitor NBMPR or via siRNA knockdown approaches (Guillén-Gómez et al., 2012) or by expressing recombinant ENT2 (SLC29A2) in heterologous models such as Xenopus oocytes (Kiss et al., 2000; Yao et al., 2002) and yeast (Vickers et al., 2002) or nucleoside transporter deficient cell lines (Ward et al., 2000).
To address this lack of a defined endogenous human ENT2 model, we have created a modified human embryonic kidney (HEK293) cell line where the ENT1 protein encoded by the gene SLC29A1 has been removed using CRISPR/cas9 gene editing. We report herein on the characterization of this ENT1 knockout (KO) cell model in terms of ENT2 expression levels (protein and gene), functional activities, and affinities for a range of known nucleoside/nucleobase compounds (both endogenous and therapeutic analogs). The loss and restoration of ENT1 activity in these cells was also assessed for its impact on the expression of other genes relevant to purine metabolism.
Materials and Methods
Materials
[8-3H]-2-chloroadenosine (37 MBQ) and [3H]-water (1 mCi/g) were obtained from Moravek Biochemicals (Brea, CA). Dipyridamole (DY), geneticin (G418), Dulbecco’s modified Eagle’s medium, FBS, penicillin-streptomycin, 2-mercaptoethanol, polybrene, guanosine, adenosine, thymidine, uridine, cytosine, adenine, hypoxanthine, cytarabine, gemcitabine, and ribavirin were purchased from Sigma-Aldrich (St. Louis, MO). Soluflazine and draflazine were from Janssen Research Foundation. The ECL Prime Western blotting system was purchased from Cytiva Life Sciences (Marlborough, MA). NBMPR, dilazep, zidovudine, abacavir, ticagrelor, 5-fluorouracil, and cladribine were from Tocris Bioscience (Toronto, ON). All primers, Alt-R crRNA, 20 nmol scale, Alt-R ATTO-550 tracrRNA, 20 nmol scale, and nuclease-Free Duplex Buffer were ordered through Integrated DNA Technologies (Coralville, IA). 2-Chloroadenosine, agarose, oligo (dT)12–18 primer, HEPES, sodium pyruvate, PowerUp SYBR Green, HALT protease inhibitor cocktail, TRIzol Reagent, and SuperScript III Reverse Transcriptase came from Thermo Fisher Scientific (Waltham, MA). The 100 bp DNA Ladder was supplied by Truin Science (Edmonton, AB). EnGen SpyCas9 (M0646T) was from New England Biolabs (Whitby, ON). Lipofectamine CRISPRMAX (CMAX00-001) and Opti-MEM were from Thermo Fisher Scientific (Burlington, ON) and the DNeasy Blood & Tissue Kit was from Qiagen Inc. (Toronto, ON). HEK293 cells were purchased from ATCC (Manassas, VA). The antibodies used for immunoblotting were mouse monoclonal IgG1 anti-Myc antibody (Clone 4A6, 05-724, Lot #3013479; EMD Millipore, Canada), monoclonal mouse anti-ENT1 (F-12:sc-377283-Santa Cruz), rabbit polyclonal anti-ENT2 (HPA030551, Lot #000001575; Sigma-Aldrich, St. Louis, MO), and mouse monoclonal antiβ-actin (C4, sc-47778, Lot #B0719; Santa Cruz Biotechnology Inc., Texas). Secondary antibodies were mouse antirabbit IgG-horseradish peroxidase (HRP) (sc-2357, Lot #2517) and m-IgGk BP-HRP (sc-516102, Lot #F1016) from Santa Cruz Biotechnology. Ecolite liquid scintillation cocktail was purchased from MP Biomedical (Irvine, CA). RedSafe nucleic acid staining solution was purchased from iNtRON Biotechnology (Gyeonggi-do, Republic of Korea).
Cell Culture
HEK293-wild-type (WT) and ENT1KO cells were cultured in Dulbecco’s modified Eagle’s medium with 10% FBS, penicillin (100 U/ml), streptomycin (100 μg/mL), and sodium pyruvate (1 mM). G418 was added (300 μg/ml) in the ENT1KO+ENT1 (ENT1KO cells stably transfected with MYC-ENT1) cell media to maintain selection pressure on the stable transfectants. The cells were washed with phosphate-buffered saline and then removed from flasks by exposure to 0.05% trypsin/0.18 mM EDTA for 10 minutes at 37°C and the suspended cells were washed in the appropriate buffer solution (without G418) immediately prior to use in subsequent assays. To assess the rate of cell proliferation, each cell line was seeded at 1 × 104 cells per well in a six-well plate. Cell numbers per well were then determined every 24 hours for six consecutive days using a BioRad TC10 automated cell counter.
Generation of KO and Recombinant ENT1 Cell Models
CRISPR/cas9 guide RNA (5′-CATTCTGGGACATGTCCAGG-3′) targeting human SLC29A1 was prepared as previously described (Cromwell et al., 2018) using Alt-R CRISPR/cas9 tracrRNA, ATTO 550. The genomic region targeted corresponded to bases 648 to 667 of the ENT1 mRNA (NM_001078177.2) which is 125 bases downstream of the translation initiation codon. mRNAEnGen Spy Cas9 NLS was assembled into ribonucleoprotein complexes and transfected into HEK293 cells using Lipofectamine CRISPRMAX as per the manufacturer’s instructions. Figure 1A illustrates the location of the CRISPR/cas9-induced modification in exon 3 of SLC29A1. Twenty-four hours following transfection, cells positive for ATTO 550 were sorted on a BD FACSAria III instrument by the Flow Cytometry Core at the University of Alberta. Positive cells were sorted into the wells of a 96-well plate containing growth media at a density of one cell per well. Clonal growth was monitored for the next 2 to 3 weeks until approximately 70% confluency was reached. Following clonal expansion, approximately 1 × 105 cells were lysed in DirectPCR Lysis Reagent (Viagen Biotech, Inc.) containing 1 mg/mL Proteinase K (Thermo Fisher) for 16 hours at 55°C. The remaining cells were transferred to a 24-well plate for continued growth. Following lysis, Proteinase K was heat inactivated at 85°C for 45 minutes. One μL of lysate was used as a template for polymerase chain reaction amplification of the corresponding target site. Genomic DNA was then isolated from the monoclonal populations and analyzed via Sanger sequencing to confirm Cas9-mediated gene disruption (Fig. 1B). These modified HEK293 cells with ENT1 genetically deleted are subsequently referred to as ENT1KO cells. To confirm the specificity of the changes in ENT1KO cells were due to loss of ENT1, these cells were subsequently transfected with recombinant N-terminal MYC-tagged SLC29A1 in pcDNA3.1 vector using the calcium phosphate precipitation method and stable clones selected with G418. These ENT1-reconstituted cells are referred to as ENT1KO+ENT1 cells.
Transport Assay
To assess transporter function, cell lines were examined for their ability to accumulate [3H]2-chloroadenosine, an established ENT1 substrate that is poorly metabolized by adenosine kinase and adenosine deaminase (Jarvis et al., 1985; Stolk et al., 2005). To eliminate the potential contribution of sodium-dependent concentrative nucleoside transporters (CNT), cells were suspended in sodium-free buffer (140 mM N-methylglucamine, 5 mM KCl, 4.2 mM KHCO3, 0.36 mM K2HPO4, 0.44 mM KH2PO4, 10 mM HEPES, 0.5 mM MgCl2, 1.3 mM CaCl2, pH 7.4) containing 0.1% dimethyl sulfoxide (solvent control; total uptake), 100 nM NBMPR (to selectively block ENT1), 5 mM adenosine (complete ENT1 and ENT2 block; nonmediated uptake), or 3 μM NBMPR/DY (dipyridamole) (alternative nonmediated uptake determination). This nonmediated component would reflect passive diffusion of [3H]2-chloroadenosine into the cells as well as extracellular [3H]2-chloroadenosine remaining in the cell pellet after processing the samples. Following a 15-minute incubation at room temperature, 250 μl of cell suspension was added to 250 μl of [3H]2-chloroadenosine (2.5–300 μM) layered over 21:4 (v:v) silicone:mineral oil (200 μl) in microcentrifuge tubes. The uptake reaction was terminated after 10 seconds for HEK293-WT, 15 seconds for ENT1KO, and 5 seconds for ENT1KO+ENT1 by centrifugation of the cells through the oil layer at 10,000 g. These incubation times were selected from preliminary time courses of 10 μM [3H]2-chloroadenosine uptake in each cell model as a time point that approximated the initial near-linear portion of the hyperbolic relationship while allowing a sufficient signal to noise ratio for analysis. After the reactions were terminated by centrifugation, the aqueous layer was aspirated and the tube above the oil was washed with approximately 1 ml of water. The water and the oil were then sequentially removed, leaving only the cell pellet. This pellet was digested in 1 M sodium hydroxide for ∼24 hours after which the radioactive content of each sample was measured using liquid scintillation techniques. Mediated uptake was defined as the difference between total and nonmediated uptake. Data were analyzed as initial rate of uptake (pmol/μl/s) against the concentration of 2-chloroadenosine used. Cell volume (μl) was measured by incubating cells with [3H]water for 2 minutes, centrifuging through 21:4 silicone:mineral oil, and then determining the [3H] content of the cell pellet and the supernatant. Total cellular volume was defined as the ratio of the dpm of the cell pellet to the dpm/μl of the supernatant and was used to define the Vmax values as pmol/μl/s. Curves were fitted using the Michaelis–Menten model in GraphPad Prism 10.1.1 to determine Km and Vmax of 2-chloroadenosine transport for each cell line. For inhibition studies, we used 10 μM [3H]2-chloroadenosine and a range of concentrations of the inhibitors/substrates with a 15-second incubation time. Ki values were calculated using the Cheng–Prusoff equation (Cheng and Prusoff, 1973) based on the IC50 values derived from the concentration-response curves and the Km of [3H]2-chloroadenosine for ENT2 in the particular cell model employed.
Radioligand Binding
Approximately 1 × 106 cells suspended in Dulbecco’s phosphate buffered saline were incubated at room temperature with a range of concentrations of [3H]NBMPR, in the presence or absence of 10 μM DY. When steady state binding was reached (∼45 minutes), cells were filtered through Whatman GF/B filters using a 24-port Brandel Cell Harvester. The [3H] content of each filter was then determined using liquid scintillation techniques. Nonspecific binding was defined as residual radioactivity associated with the filters after incubation of the cells with 10 μM DY. Specific binding was defined as the difference between total binding and nonspecific binding. Kd and Bmax values were determined by fitting hyperbolic curves to the site-specific [3H]NBMPR binding versus the equilibrium free concentration of [3H]NBMPR using GraphPad Prism 10.1.1.
Immunoblot
HEK293-WT, ENT1KO, and ENT1KO+ENT1 cell lysates were extracted in radioimmunoprecipitation buffer (150 mM NaCl, 50 mM Tris, 1% NP-40, 0.5% sodium deoxycholate, 1% SDS) containing HALT Protease Inhibitor Cocktail + EDTA. The concentrations of the protein extract were assessed using a bicinchoninic acid assay kit and accordingly diluted to 1.5 μg/μL protein and adjusted to 2% β-mercaptoethanol for reducing conditions. Thirty μg protein was resolved using SDS-PAGE on 12.5% acrylamide gels at 80 V for 15 minutes, followed by 150 V for 45 minutes. Proteins were then electro-transferred from gel onto Immobilon-P polyvinylidene fluoride membranes (Millipore Corporation, MA) for 1.5 hours at a constant current of 280 mA. Following transfer, membranes previously rinsed in Tris-buffered saline/Tween 20 (TBST; 150 mM NaCl, 50 mM Tris, pH 7.5 containing 0.1% v/v Tween-20) were incubated in 5% w/v skim milk powder in TBST for 1 hour at room temperature with gentle rocking to block nonspecific binding of the polyvinylidene fluoride membrane. Membranes were then incubated overnight (approximately 16 hours) at 4°C with gentle rocking in the presence of primary antibodies such as mouse anti-MYC, mouse anti-ENT1, rabbit anti-ENT2, or mouse antiβ-actin at 1:1000, 1:500, 1:1000, 1:500 dilution, respectively, in TBST containing 1% skim milk powder. The following day, the membranes were washed four to five times with TBST containing 1% skim milk powder, then incubated for 1 to 2 hours at room temperature with anti-rabbit IgG-HRP or antimouse IgG-HRP at 1:5000 or 1:3000 dilution, respectively, in TBST containing 1% skim milk powder. Following this incubation, membranes were further washed (four to five times) with TBST and protein detected using ECL Prime solution and visualized via chemiluminescence using an Amersham Imager 680 (GE Healthcare, Chicago, IL). Image J software was then used to conduct densitometry analyses.
Real-Time Semiquantitative Polymerase Chain Reaction
For extraction of RNA, cells from confluent 10 cm plates were suspended in 1 ml of TRIzol reagent and homogenized according to the manufacturer’s protocol (Thermo Fisher Scientific). Total RNA concentration and purity were determined using a Nanodrop 2000 spectrophotometer (Life Technologies Inc.). Two μg of total RNA was reverse transcribed to cDNA using Oligo (dT)12–18 primer and Superscript III Reverse Transcriptase. For semiquantitative reverse transcriptase polymerase chain reaction (RT-qPCR), cDNA (∼200 ng/well) was prepared with the primer sets specific for SLC29A1, SLC29A2, SLC29A3, SLC29A4, SLC28A1, SLC28A2, SLC28A3, SLC43A3, ADK, ADA, APRT, ADSL, ADSS, ABCC4, GART, HPRT, IMPDH2, MTAP2, NT5C, NT5E, PNP,and PRPS1 (Table 1) using Power Up SYBR Green fluorescence on a Roche Light Cycler 480 System (Cardiovascular Research Centre, Edmonton, Canada). Primer efficiency and melt curves were assessed prior to their use for gene expression analyses. RT-qPCR conditions were 2 minutes at 50°C (uracil-DNA glycosylase activation), 2 minutes at 95°C (denaturation), followed by 50 cycles of 15 seconds at 95°C, and 60 seconds at 60°C for amplification, with a final melt curve analysis. Gene expression was normalized to either glyceraldehyde-3-phosphate dehydrogenase (GAPDH) alone or the geographic mean of two separate reference genes, GAPDH and β-actin (ACTB), and the relative expression of the genes was determined using the ΔΔCt method and HEK293-WT cells as the comparator.
Data Analysis and Statistics
Sample size for the transport and binding assays was predetermined as N = 5 (five independent experiments with two to three internal replicates), which is the minimum needed to define statistical differences based on the known variability inherent in these types of studies. Nonlinear curves were fitted to data, and statistical analyses were done using GraphPad Prism 10.1.1 software. In all cases, if the P value determined from a statistical test was less than 0.05, the difference was considered significant and the null hypothesis (no difference between data sets) was rejected. Influx data were fit using a one-phase association for time course data and Michaelis–Menten curves were fit to the concentration-dependent uptake data for determination of Km and Vmax values. Statistical differences between Km and Vmax values were determined using the extra sum-of-squares F test. Significant differences between groups were assessed using a one-way or two-way ANOVA, as appropriate, corrected for multiple comparisons with the Holm–Sidak method.
Results
Characterization of Equilibrative Nucleoside Transport in HEK293 Cells
HEK293 cells were characterized for their equilibrative nucleoside transport activity based on their ability to accumulate [3H]2-chloroadenosine under sodium-free conditions and on their ability to bind the ENT1-selective probe [3H]NBMPR. The relative contribution of ENT1 and ENT2 to the substrate accumulation by these cells was defined by the proportion of the total uptake that was sensitive to inhibition by 100 nM NBMPR (to selectively inhibit ENT1) and the proportion of NBMPR-resistant uptake that could be subsequently inhibited by DY or supramaximal concentrations of the endogenous substrate adenosine (the ENT2-mediated component) (Fig. 2A). The cell associated [3H]2-chloroadenosine remaining after blocking ENT1 and ENT2 with adenosine or DY (nonmediated component) was higher than anticipated, but it was linear with concentration suggesting a nonsaturable process. Furthermore, there was no significant difference in the nonmediated component when it was determined using a combination of 3 μM NBMPR/DY or 5 mM adenosine, although data obtained using adenosine was more consistent between experiments. When the data were parsed (Fig. 2B) to define the Total, ENT1-, and ENT2-mediated components, it was determined that ENT1 transported 2-chloroadenosine with a Km of 34 ± 7 μM and a Vmax of 3.23 ± 0.20 pmol/μl/s. ENT2 had a similar affinity for 2-chloroadenosine (Km = 23 ± 5 μM) but a lower Vmax of 1.02 ± 0.06 pmol/μl/s, indicating that about 25% of the total 2-chloroadenosine uptake by HEK293 cells was mediated by ENT2 under these conditions. HEK293 cells also bound the ENT1-selective probe [3H]NBMPR with very high affinity (Kd = 0.063 ± 0.004 nM) to 53,400 ± 900 sites/cell (Fig. 2C). A limitation of these studies is that the transporter kinetics were assessed at room temperature. ENT1/2 Vmax values are likely substantially higher in vivo than those reported previously for this in vitro study. This work was done at room temperature to slow down the kinetics due to the equilibrative nature of these transporters; reasonable estimates of initial rates cannot be obtained at physiological temperatures (37°C) using this approach.
To assist with the interpretation of the functional data described earlier, the relative (to SLC29A1) level of expression of various nucleoside/nucleobase transporters (SLC29A2, SLC29A3, SLC29A4, SLC43A3, SLC28A1, SLC28A2,and SLC28A3) and metabolic enzymes involved in purine synthesis (de novo) or metabolism (ADK, ADA, APRT, ADSL, ADSS, ABCC4, GART, HPRT, IMPDH2, MTAP2, NT5C, NT5E, PNP, and PRPS1) were assessed by qRT-PCR) (Fig. 2D). Of note, SLC29A2 (ENT2) levels were about 20% of those seen for SLC29A1 (ENT1), which roughly parallels the functional ENT1/ENT2 ratio of 75:25. There was no detectable SLC29A4 (ENT4) transcript, nor was there any significant expression of the nucleobase transporter SLC43A3 (ENBT1). Likewise, the concentrative purine transporters SLC28A2 (CNT2) and SLC28A3 (CNT3) were expressed at very low levels. Interestingly, SLC28A1, which is selective for pyrimidine nucleosides, was expressed at relatively high levels in the HEK293 cells.
Creation and Characterization of the ENT1KO Cell Mode
The disrupting mutation in exon 3 of the SLC29A1 gene (see Fig. 1) in the ENT1KO cells was confirmed by Sanger sequencing and the loss of SLC29A1 transcript. The deletion of ENT1 in these cells was also confirmed by the loss of 1) ENT1 immunoreactivity, 2) NBMPR-sensitive [3H]2-chloroadenosine transport, and 3) high-affinity binding of [3H]NBMPR.
As expected, the level of SLC29A1 transcript (encoding ENT1) in the ENT1KO cells was negligible (Fig. 3A), but the SLC29A2 transcript (encoding ENT2) was unchanged from that seen in the HEK293-WT cells (Fig. 3B). Immunoreactivity corresponding to ENT1 and ENT2 was detected in the HEK293-WT cells using ENT-subtype selective antibodies at the molecular mass expected (∼55–65 kDa depending on degree of glycosylation) (Fig. 3, C and D). When normalized to β-actin expression, the ENT1KO cells showed negligible ENT1-like immunoreactivity (Fig. 3C) but had a similar level of ENT2 expression relative to the HEK293-WT cells (Fig. 3D).
[3H]2-chloroadenosine uptake by the ENT1KO cells in the absence and presence of 100 nM NBMPR or 5 mM adenosine is shown in Fig. 4A. Secondary analysis of these data to obtain transporter-mediated flux (Fig. 4B) indicated that there was no NBMPR-sensitive [3H]2-chloroadenosine uptake (ENT1-mediated), and the ENT2-mediated uptake had a Km of 72 ± 28 μM and a Vmax of 1.50 ± 0.22 pmol/μl/s. This Vmax is not significantly different from that determined for ENT2 in HEK293-WT cells. The loss of ENT1 in the ENT1KO cells was also confirmed by the lack of high affinity binding sites for the ENT1-probe [3H]NBMPR (Fig. 4C), confirming the complete knockout of SLC29A1/ENT1.
ENT2-mediated [3H]2-chloroadenosine uptake by the ENT1KO cells was inhibited by a range of nucleoside/nucleobase substrates (Fig. 5, A and B). Of the nucleosides tested, adenosine had the highest affinity for ENT2, followed by thymidine, inosine, and guanosine, and uridine had the lowest affinity. In terms of nucleobases, only hypoxanthine had an affinity similar to the nucleosides; adenine and cytosine had very low affinity (>1 mM Ki) for ENT2. Among the well-known inhibitors of ENTs, DY and soluflazine had the highest affinities for ENT2, followed by draflazine, NBMPR, and dilazep (Fig. 5C). We also examined the ENT2 affinity profile of some nucleoside analog drugs that are used to treat cancer, viral infections, and cardiovascular diseases (Fig. 5D). The cardiovascular drug ticagrelor had the highest affinity for ENT2 of all the analogs tested. This was followed by the anticancer drugs cladribine and gemcitabine. Among the antiviral drugs tested, abacavir had the highest affinity for ENT2 followed by zidovudine and ribavirin. Ki values derived from these data are shown in Table 2.
We then investigated potential compensatory changes in other transporters or enzymes involved in nucleoside/nucleobase metabolism (Fig. 6). Among the various transporters assessed, only the sodium-dependent concentrative purine/pyrimidine transporter, SLC28A3, and the nucleobase transporter, SLC43A3, were significantly decreased in the ENT1KO cells. With respect to the enzymes involved in purine metabolism, APRT expression was doubled in the ENT1KO cells, and the NT5E and ADSL transcripts decreased by about 50%.
Transfection of ENT1KO Cells with Recombinant Myc-Tagged ENT1
To examine whether changes noted previously in the ENT1KO cells were indeed due to the loss of ENT1, these cells were stably transfected with recombinant N-terminal Myc-tagged-ENT1 (coined as ENT1KO+ENT1 cells) and assessed in a manner similar to that described earlier. Note that the level of expression of the SLC29A1 transcript in these cells was about 40-fold higher than that seen endogenously in the HEK293-WT cells. In spite of the high level of SLC29A1 transcript in the ENT1KO+ENT1 cells, they only had about a threefold higher level of ENT1 protein expression relative to HEK293-WT cells (Fig. 3C), suggesting the cellular protein processing machinery was a limiting factor. The expression of ENT2 protein (Fig. 3D) and SLC29A2 transcript (Fig. 3B) were, however, not significantly different between the ENT1KO+ENT1 cells and HEK293-WT cells. Likewise, the ENT1KO+ENT1 cells accumulated [3H]2-chloroadenosine in an NBMPR-sensitive manner (ENT1-mediated) (Fig. 7A) at about double the rate of the HEK293-WT cells (Vmax of 5.07 ± 0.41 pmol/μl/s), and ENT2 mediated transport was similar to that in both the HEK293-WT cells and the ENT1KO cells (Vmax of 0.83 ± 0.14 pmol/μl/s) (Fig. 7B). These recombinant ENT1 expressing cells also had restored high-affinity [3H]NBMPR binding with a Kd of 0.20 ± 0.06 nM and a Bmax of 139,000 ± 14,000 sites/cell (again, double that seen in the HEK293-WT cells) (Fig. 7C).
Overexpression of SLC29A1 in these cells led to a significant increase in the expression of a number of other genes associated with purine/pyrimidine metabolism. Specifically, the transporters SLC28A2, SLC28A3, and ABCC4 were elevated two- to threefold (Fig. 6A), and the enzymes involved in purine metabolism APRT, HPRT, GART, NT5E, and PNP were elevated four- to sixfold (Fig. 6, B and C) over levels seen in untransfected HEK293 cells. The decrease in ADSL upon loss of SLC29A1 in the ENT1KO cells was also restored to levels similar to that of the HEK293-WT cells (Fig. 6B).
Cell Proliferation Rate
The ENT1KO cells had a significantly increased rate of proliferation with a rate constant k of 0.64 ± 0.06 (doubling time of 1.1 day) relative to the HEK293-WT cell (k = 0.40 ± 0.06; doubling time of 1.7 days) (Fig. 8). Transfection of the ENT1KO cells with recombinant ENT1 resulted in a cell doubling time of 1.6 days (k = 0.45 ± 0.06) that was not significantly different from that of the HEK293-WT cells.
Discussion
ENT2 has typically been studied in the presence of NBMPR to block ENT1 activity or by overexpressing recombinant ENT2 in Xenopus oocytes, yeast, or nucleoside transporter deficient cell models. For example, our research group and others have used the porcine kidney nucleoside transporter deficient cell line as a null background for expression of recombinant ENT proteins to study ENT function and regulation (Ward et al., 2000; Robillard et al., 2008; Park and Hammond, 2012; Hughes et al., 2015). While useful data has been obtained in these models, its relationship to the actual endogenous function of ENT2 must be interpreted with caution because 1) the high concentrations of NBMPR used to block ENT1 may also partially inhibit ENT2 and 2) expression of human ENT in a nonhuman cell line, or at supraphysiological levels, can lead to differences in protein processing and regulation. To address these issues, we have created and characterized a novel model to study endogenous ENT2 in isolation in human HEK293 cells where ENT1 has been deleted using CRISPR/cas9 gene editing. We chose the HEK293 cell line due to its ease of transfection, robust proliferation, and the fact that it does not express other purine transporters such as SLC43A3 (ENBT1) or SLC29A4 (ENT4). It also has a very low level of expression of the concentrative sodium-dependent purine nucleoside transporters SLC28A2 (CNT2) and SLC28A3 (CNT3) (see Fig. 2D). While this work was in progress, a similar study was published using HeLa cells (Miller et al., 2021). However, HeLa cells express high levels of SLC43A3 (Furukawa et al., 2015) and measurable SLC29A4 (Uhlén et al., 2015), making them less suitable for the subsequent development of a purine transporter deficient cell model (a future goal of our laboratory). We determined in the present study that HEK293 cells functionally express (based on the relative rates of [3H]2-chloroadenosine transport) both ENT1 and ENT2 in an approximate 3:1 ratio. We also established that these cells bind [3H]NBMPR to 53,400 sites per cell. [3H]NBMPR is known to have a single binding site on the ENT1 protein, so this translates to 53,400 ENT1 proteins per cell. Assuming similar translocation mechanisms between ENT1 and ENT2, one might therefore predict from this ratio that there were about 17,800 ENT2 transporters per cell.
The deletion of ENT1 from these cells to create the ENT1KO cell line was confirmed by Sanger sequencing, a decrease in ENT1-like immunoreactivity, the loss of ENT1-mediated transport, and the loss of high-affinity [3H]NBMPR binding. While it is possible that a truncated variant of ENT1 was expressed in these cells (the N-terminal 41 amino acids, based on the 125 bases between the ENT1 start codon and the location of the CRISPR/cas9-mediated gene disruption), this peptide would not be detected by the ENT1 antibodies used, and it is unlikely to be functional. The loss of ENT1 in these cells appeared to have no effect on the expression of ENT2. A similar lack of effect of SLC29A1 deletion on SLC29A2 expression was also noted in a recent study where the CRISPR/cas9 approach was used to delete SLC29A1 in HeLa cells (Miller et al., 2021); however, ENT1 and ENT2 protein levels were not measured in that study. This lack of effect of ENT1 knockout on ENT2 was unanticipated as other studies have suggested a potential role of oligomerization of ENT1 and ENT2 in ENT2 trafficking to the membrane and its regulation (Grañé-Boladeras et al., 2019). The only purine transporter transcripts that were impacted by the loss of ENT1 were SLC28A3 and SLC43A3, which were decreased. However, these transporters are expressed at very low levels already in the HEK293-WT cells such that this change would not be expected to confound interpretation of the functional studies. One might anticipate that the loss of ENT1 would decrease the ability of cells to salvage nucleosides from the extracellular media and thus result in an upregulation of enzymes involved in purine energy metabolism. This was indeed observed with a significant increase in APRT expression and a trend toward an increase in other ribotransferases such as GART and HPRT. Contrary to this theory, though, the gene that encodes for adenylosuccinate lyase, ADSL, which converts adenylosuccinate to AMP and fumarate as part of the purine nucleotide cycle, was downregulated in the ENT1KO cells. There was also a decrease in the expression of NT5E, which encodes for CD73, an ectonucleotidase, which would lead to decreased metabolism of AMP to adenosine extracellularly.
To determine if the loss of ENT1 (and potentially a disruption of hetero-oligomers) had an impact on the pharmacology of ENT2, we screened a wide range of nucleosides and nucleobases (both endogenous as well as analogs used therapeutically) and known ENT inhibitors for their ability to inhibit ENT2-mediated transport in the ENT1KO cell line. With some exceptions noted later, the results obtained in this study paralleled past studies with respect to the affinity of these compounds for ENT2. It must be noted, though, that comparison with previous literature is complicated due to the use of different recombinant expression models (oocytes, yeast, mammalian cells), and different experimental conditions (incubation times, radiolabeled substrate used). The only notable differences observed were for adenosine, guanosine, and thymidine, which all had significantly higher affinities for ENT2 relative to that seen in previous studies; adenosine in particular stood out with a Ki of 2.6 μM for inhibiting ENT2-mediated transport, which is an order of magnitude lower than what has been reported previously (57–140 μM) (Ward et al., 2000; Visser et al., 2005), including in a similar study on endogenous ENT2 in a HeLa cell line with SLC29A1 deleted via CRISPR/cas9 gene editing (Miller et al., 2021). It must also be noted that the latter study used uptake incubation times of 2 minutes or greater, so total cellular accumulation of substrate (including effects of intracellular metabolism) was measured rather than ENT transport kinetics. This may be why their Ki values were considerably higher than what is typically seen for ENT substrates. It is also noteworthy that out of the nucleobases tested in the present study, only hypoxanthine had a Ki of less than 1 mM for inhibiting ENT2. ENT2 has been described as both a nucleoside and nucleobase transporter (Yao et al., 2002; Naes et al., 2020), but based on our results ENT2 is unlikely to play a significant role in the transport of adenine and cytosine or the nucleobase analogs 5-fluorouracil or 6-mercaptopurine. Among the anticancer drugs tested, cladribine and gemcitabine had the highest affinity for ENT2, and among the antiviral drugs, abacavir showed a high affinity for ENT2. Our results also confirm that soluflazine is a reasonably potent inhibitor of ENT2, with an affinity similar to that seen in previous studies (Hammond, 1991a, 2000).
Finally, we found that the expression of ENT1 was inversely related to the rate of cell proliferation. This result concurs with data obtained using a mouse ENT1 knockout model, where intervertebral disk cells from the ENT1KO mice were seen to have a higher proliferation rate compared with those from the WT mice (Veras et al., 2019). This was associated with increased levels of intracellular adenosine resulting from the loss of ENT1 that was altering cell cycle processes by downregulating growth factor cytokines such as E2f1–8. Previous studies have also suggested that splice variants of ENT2 can form hetero-oligomers at nuclear membranes, which play important roles in cell proliferation (Grañé-Boladeras et al., 2016). Our novel ENT1-knockout cell model may be used in the future to explore the expression level of both ENT2 and its splice variants and their interactions, to gain further insights on their role in cell proliferation.
In conclusion, this is one of the first studies to conduct a detailed characterization of endogenous ENT2 in isolation. Our study demonstrates that the kinetics and pharmacology of ENT2 is independent of ENT1. This novel, well-characterized cell model will be useful for the examination of transcriptional and post-translational regulation of endogenous human ENT2 and for screening novel substrates and inhibitors of ENT2.
Acknowledgments
The authors wish to acknowledge the invaluable technical assistance provided by Monica Dabrowska, Deborah Sowsnoski, and Tierah Hinchliffe.
Data Availability
All experimental data are included in this manuscript.
Authorship Contributions
Participated in research design: Shahid, Cromwell, Hammond.
Conducted experiments: Shahid, Hammond.
Contributed new reagents or analytic tools: Cromwell, Hubbard.
Performed data analysis: Shahid, Hammond.
Wrote or contributed to the writing of the manuscript: Shahid, Hammond.
Footnotes
- Received May 28, 2024.
- Accepted July 15, 2024.
Funding for this project was provided to J.R.H. by the Natural Sciences and Engineering Research Council of Canada [Grant# RGPIN-2020-04046].
This study was conducted as part of the graduate thesis work of Nayiar Shahid. No author has an actual or perceived conflict of interest with the contents of this article.
Portions of this work have been presented at national and international conferences as follows:
Shahid N, Nguyen KH, Cromwell C, Hubbard BP, Hammond JR (2019) Development and assessment of novel HEK293 cell lines for the study of equilibrative nucleoside transporter regulation. 2019 Conference of the Canadian Society of Pharmacology and Therapeutics, Calgary, Canada.
Shahid N, Cromwell C, Hubbard BP, Hammond JR (2021) Characterization of a novel HEK293 cell line (HEK293-ENT1KO) to assess the role of equilibrative nucleoside transporter subtype-2. 2021 Experimental Biology Meeting (virtual).
Shahid N, Cromwell C, Hubbard BP, Hammond JR (2021) Assessing the role of equilibrative nucleoside transporter subtype-2 through development of a novel HEK293 cell line (HEK293-ENT1KO). Annual Meeting of the Canadian Society of Pharmacology and Therapeutics.
Shahid N, Hammond JR (2022) Impact of loss or regain of equilibrative nucleoside transporter (ENT) subtype-1 on ENT subtype-2 pharmacology in a novel CRISPR-Cas9 HEK293-ENT1KO cell model. Experimental Biology Conference, Philadelphia, PA, USA.
Abbreviations
- CNT
- concentrative nucleoside transporter
- DY
- dipyridamole
- ENT
- equilibrative nucleoside transporter
- G418
- geneticin
- HEK293
- human embryonic kidney
- HRP
- horseradish peroxidase
- KO
- knockout
- NBMPR
- nitrobenzylthioinosine
- RT-qPCR
- semiquantitative reverse transcriptase polymerase chain reaction
- TBST
- Tris-buffered saline/Tween 20
- WT
- wild-type
- Copyright © 2024 by The Author(s)
This is an open access article distributed under the CC BY-NC Attribution 4.0 International license.