View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Biochimica et Biophysica Acta 1852 (2015) 2086–2095

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbadis

Equilibrative transporter 3 depletion in β-cells impairs mitochondrial function and promotes apoptosis: Relationship to pigmented hypertrichotic dermatosis with insulin-dependent diabetes

B. Liu a,A.Czajkaa,A.N.Malika, K. Hussain b,P.M.Jonesa,S.J.Persauda,⁎

a Diabetes Research Group, Division of Diabetes and Nutritional Sciences, Faculty of Life and Medical Sciences, King's College London, London SE1 1UL, United Kingdom b Institute of Child Health, London WC1N 1EH, United Kingdom

article info abstract

Article history: Loss of function recessive mutations in the SLC29A3 gene that encodes human equilibrative nucleoside transporter 3 Received 24 February 2015 (ENT3) have been identified in patients with pigmented hypertrichotic dermatosis with insulin-dependent diabetes Received in revised form 17 June 2015 (PHID). ENT3 is a member of the equilibrative nucleoside transporter (ENT) family whose primary function is Accepted 7 July 2015 mediating transport of and nucleobases. The aims of this study were to characterise ENT3 expression Available online 9 July 2015 in islet β-cells and identify the effects of its depletion on β-cell mitochondrial activity and apoptosis. RT-PCR fi fi β Keywords: ampli cation identi ed ENT3 expression in human and mouse islets and exocrine pancreas, and in MIN6 -cells. Islets of Langerhans Immunohistochemistry using human and mouse pancreas sections exhibited extensive ENT3 immunostaining of β-Cells β-cells, which was confirmed by co-staining with an anti-insulin antibody. In addition, exposure of dispersed Equilibrative nucleoside transporter 3 human islet cells and MIN6 β-cells to MitoTracker and an ENT3 antibody showed co-localisation of ENT3 to β-cell Diabetes mitochondria. Consistent with this, Western blot analysis confirmed enhanced ENT3 immunoreactivity in β-cell Mitochondria mitochondria-enriched fractions. Furthermore, ENT3 depletion in β-cells increased mitochondrial DNA content Apoptosis and promoted an energy crisis characterised by enhanced ATP-linked respiration and proton leak. Finally, inhibition of ENT3 activity by dypridamole and depletion of ENT3 by siRNA-induced knockdown resulted in increased caspase 3/7 activities in β-cells. These observations demonstrate that ENT3 is predominantly expressed by islet β-cells where it co-localises with mitochondria. Depletion of ENT3 causes mitochondrial dysfunction which is associated with enhanced β-cell apoptosis. Thus, apoptotic loss of islet β-cells may contribute to the occurrence of autoantibody- negative insulin-dependent diabetes in individuals with non-functional ENT3 mutations. © 2015 Elsevier B.V. All rights reserved.

1. Introduction ENTs are a widely distributed family of mammalian transporters whose primary function is transporting synthesis precursors, Pigmented hypertrichotic dermatosis with insulin-dependent diabe- such as nucleosides and nucleobases within cells [4,5]. There are four tes (PHID) is a rare autosomal recessive disorder that is characterised ENTs, designated ENT1-4 [4,6],thatinfluence diverse physiological predominantly by autoantibody negative, insulin-dependent diabetes processes such as cardiovascular function and neurotransmission by that is associated with hyperpigmentation and hypertrichosis [1,2]. regulating intracellular trafficking. The cellular uptake of Homozygosity mapping from six patients with PHID in five unrelated hydrophilic nucleoside analogues used in the treatment of cancers and families identified five loss of function mutations in the SLC29A3 gene viral diseases (e.g. gencitabine and zidovidine) is also facilitated by that encodes human equilibrative nucleoside transporter 3 (ENT3), a ENTs. ENT1 and 2 are ubiquitously distributed in human and rodent member of the equilibrative nucleoside transporter (ENT) family [3]. tissues and their functions in nucleoside transport are relatively well- characterised. Much less is known about the tissue distributions and functions of ENT3, although it is reported to be expressed by several mouse and human tissues such as placenta, where the cDNA was origi-

Abbreviations: ENT, equilibrative nucleoside transporter; MIN6 cells, mouse insulinoma nally cloned, heart, where it functions as a pH-dependent lysosomal 6 cells; MtDNA, mitochondrial DNA; OCR, oxygen consumption rate; PHID, pigmented transporter [7,8] and most recently neurons and astrocytes, where it is hypertrichotic dermatosis with insulin-dependent diabetes; VDAC, voltage-dependent thought to mediate adenosine transport [9]. anion channel Human ENT3 (hENT3) is a 475-residue protein that is 74% identical ⁎ Corresponding author at: Division of Diabetes & Nutritional Sciences, Faculty of Life & in amino acid sequence to its mouse homologue mENT3 [7,8]. Unlike Medical Sciences, King's College London, 2.9N Hodgkin Building, Guy's Campus, London SE1 1UL, United Kingdom. ENT1 and ENT2, the nucleoside transport function of ENT3 is maximal E-mail address: [email protected] (S.J. Persaud). at pH 5.5, suggesting that it may be localised to acidic intracellular

http://dx.doi.org/10.1016/j.bbadis.2015.07.002 0925-4439/© 2015 Elsevier B.V. All rights reserved. B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 2087 compartments [6]. In support of this, human and mouse ENT3 se- 2.2. Isolation of mouse and human islets quences possess a 51 residue hydrophilic N-terminal region containing two dileucine motifs that are characteristic of endosomal/lysosomal Islets of Langerhans were isolated from male ICR mice by collagenase targeting sequences [6,10], so it has been suggested that ENT3 is localised digestion of the exocrine pancreas, essentially as described previously to lysosomes rather than sharing the cell surface distribution of ENT1 and [13]. Human islets of Langerhans and acinar cells were aseptically isolat- ENT2 [7]. However, a more recent study has identified a putative mito- ed from pancreases from non-diabetic, heart beating cadaver organ chondria targeting signal at the N-terminus of hENT3 and demonstrated donors at King's College Hospital Islet Transplantation Unit with appro- that hENT3 is predominantly expressed by mitochondria in human priate ethical approval [14]. hepatocytes, where it regulates transport of native nucleosides and nucleoside drugs [11]. A mitochondrial adenosine transport sys- 2.3. RT-PCR tem, distinct from the mitochondrial nucleotide carrier, has previously been identified in rat testis [12] supporting the observations RNAs were extracted from MIN6 β-cells, mouse and human islets and [11] that adenosine transport by mitochondria is of physiological mouse and human acinar cells according to the RNeasy kit manufacturer's relevance. protocol and reverse-transcribed into cDNAs as described previously [15]. Given the essential role of mitochondria in regulating cellular cDNAs were then amplified over 40 cycles using synthetic oligonucleo- metabolism and cell survival, it is possible that the development of tide primers specificformouseENT1(sense:5′-ccagtggttctgagctgtca-3′; diabetes in individuals with PHID arises, at least in part, through antisense: 5′-ctgttggtgggtggagagtt-3′; 241 bp product), ENT2 (sense: 5′- impaired β-cell mitochondrial ENT3 activity resulting in metabolic gctgggtaccatgccttcta-3′;antisense:5′-ccacacagggtgtgatgaag-3′;152bp defects and apoptotic loss of β-cells. To date, there have been no product), ENT3 (sense: 5′-ttgggctctgtatgggactc-3′; antisense: 5′- experimental evaluations of mammalian ENT3 expression and sub- ttcttcaggatgggtccaag-3′; 190 bp product) and human ENT3 (sense: cellular localisation in islets, so this study was designed to charac- 5′-atgaccggctcctttcctat-3′;antisense:5′-atgcagagcacgaggaagat-3′; terise ENT3 expression by human and mouse β-cells and identify 170 bp product). For some analyses, quantitative PCR was performed the effects on mitochondrial bioenergetics and β-cell apoptosis using the same primers that were used in the standard PCR. Relative following ENT3 depletion. expression of mRNAs was determined after normalisation against β- actin as an internal reference and calculated by the 2−ΔΔCt method 2. Materials and methods [16].

2.1. Materials 2.4. Immunohistochemistry

Dulbecco's modified eagle's medium (DMEM), Research Park Memo- Mouse and human pancreases were fixed in 4% paraformaldehyde, rial Institute (RPMI) medium, penicillin/streptomycin, L-glutamine, colla- wax embedded, and 5 μm sections were cut onto microscope slides. genase type XI, cell dissociation solution, DAPI (4′,6-Diamidino-2- Dewaxed and rehydrated mouse pancreas sections were incubated over- phenylindole), oligomycin, rotenone, antimycin A, mitochondria isola- night at 4 °C with ENT3 antibody alone (1/50 dilution) and human pan- tion kit (MITOISO2), anti-β-actin antibody and anti-glucagon antibody creas sections were co-incubated with antibodies directed against ENT3 were purchased from Sigma-Aldrich (Dorset, UK). was and insulin (both at 1/50 dilution). Immunoreactive ENT3 in mouse obtained from Tocris Biosciences (Bristol, UK). MitoTracker red CMXRos, sections was detected by incubation with streptavidin–horseradish foetal bovine serum (FBS) and 10% polyacrylamide bis–tris gels were peroxidase and diaminobenzidine, and with a FITC-conjugated second- from Invitrogen (Paisley, UK). ECL Western blotting detection reagents ary antibody (1/50 dilution) in human sections. Insulin immunoreactivity and Rainbow™ molecular weight markers were supplied by GE in human pancreas was detected by incubation with a Texas Red- Healthcare (Buckinghamshire, UK). RNeasy mini kits, DNeasy blood conjugated secondary antibody (1/50 dilution). Immunostained pancre- and tissue kits, RNase-free DNase sets and QuantiTect SYBR® Green as sections were visualised under a Nikon TE2000-U microscope, images PCR Kits were obtained from Qiagen (Manchester, UK). Primers that were acquired with a Nikon Digital Sight-Qi1Mc camera and analysed were used for both standard PCR and quantitative PCR were from Operon with the NIS-Elements Br 3.0 software. Biotech (Cologne, Germany). Standard PCR was carried out using a Px2 Thermal RT-PCR cycler (Thermo Scientific, Epsom, UK) and quantitative 2.5. Immunocytochemistry PCR was performed using a LightCycler480 (Roche Diagnostics, West Sussex, UK). The goat polyclonal antibody, raised against a synthetic MIN6 β-cells were retrieved by trypsinisation and isolated human peptide sequence near the N terminus of SLC29A3 (ENT3) that is islets were dispersed into cell suspensions by incubating for 10 min at not present within ENT2 or ENT1, was purchased from Santa Cruz 37 °C with cell dissociation solution. These cell suspensions were seeded (Cambridgeshire, UK), the anti-insulin antibody was from Dako UK onto glass coverslips (30,000 cells/coverslip) and exposed to 200 nM Ltd. (Cambridgeshire, UK), the voltage-dependent anion channel MitoTracker red CMXRos for 30 min, fixed with 4% formaldehyde and antibody was from Cell Signaling Technology, Inc. (Massachusetts, permeabilised with 0.1% (v/v) Triton X-100 in PBS. ENT3 immunoreac- USA), Pierce bicinchoninic acid (BCA) protein assay kits, horseradish tivity was detected by incubating overnight with ENT3 antibody followed peroxidase-linked anti-rabbit secondary antibody and Texas Red- by incubation with a FITC-conjugated secondary antibody (both at 1/50 and FITC-conjugated secondary antibodies were from Thermo Scientific dilution). Nuclei were visualised by staining cells with DAPI (0.5 μg/ml). (Surrey, UK), as were scrambled non-silencing siRNAs, ON-TARGETplus Immunostained cells were visualised using a Nikon TE2000-U micro- SMARTpool Slc29a3 siRNAs, Accell green fluorescent nontargeting scope (MIN6 cells, 40× objective under oil immersion) or a Leica TCS control siRNAs, and Accell SMARTpool Slc29a3 siRNAs. Nucleofector SP2 laser scanning spectral confocal microscope (human islet cells, 63× Kit R for MIN6 cell transfection was from Amaxa Biosystems (Cologne, objective under oil immersion). Germany). Caspase-Glo 3/7 and CellTiter-Glo assay kits were purchased from Promega (Southampton, UK). Recombinant murine tumour necro- 2.6. Preparation of mitochondria-enriched fractions from MIN6 β-cells sis factor α (TNFα), interferonγ (IFNγ) and interleukin-1β (IL-1β)were from PeproTech EC (London, UK). All materials and reagents for the Mitochondria-enriched fractions were prepared using a commercial extracellular flux (XF) assays were obtained from Seahorse Biosciences mitochondria isolation kit. Briefly, 2 × 107 MIN6 cells were washed with (Copenhagen, Denmark). ICR mice were purchased from Harlan PBS and resuspended in 1 ml cell lysis buffer for 5 min. An aliquot of this (Oxfordshire, UK). lysate was removed and used as the “crude lysate” fraction. Thereafter, 2088 B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095

1 ml extraction buffer (10 mM HEPES, pH 7.5, containing 200 mM man- 2.10. Measurements of ATP generation nitol, 70 mM sucrose, 1 mM EGTA, 2 mg/ml BSA) supplemented with protease inhibitors (104 mM AEBSF, 80 μM aprotinin, 4 mM bestatin, MIN6 β-cells were transiently transfected with scrambled 1.4 mM E-64, 2 mM leupeptin and 1.5 mM pepstatin A, all provided as siRNAs or ENT3 siRNAs and ATP generation was measured after a single reagent in the mitochondria isolation kit) was added to the 40 h using the CellTiter-Glo luminescent assay following the remaining cell lysate followed by centrifugation (1000 rpm, 10 min) to manufacturer's instructions. pellet unbroken cells and nuclei. The supernatant was then centrifuged at 11,000 rpm for 10 min and the pellet, containing mitochondria, was 2.11. Cellular Bioenergetics resuspended in 100 μl CellLytic M lysis reagent (provided in the mito- chondria isolation kit). MIN6 β-cells that had been transiently transfected with non- silencing siRNAs or ENT3 siRNAs were seeded onto 96-well flux plates at a density of 40,000 cells/well for 40 h then washed and equilibrated 2.7. SDS-PAGE/Western blotting for 1 h with an assay medium supplemented with 5 mM glucose and 1 mM sodium pyruvate. Cellular bioenergetics were then determined Protein extracts obtained from whole MIN6 β-cells and MIN6 β-cell using the Seahorse extracellular flux analyser (XFe96) with the XF cell crude lysate and mitochondria-enriched fractions were quantified by Mito stress kit. Briefly, basal oxygen consumption rate (OCR) was mea- BCA protein assay. They were then fractionated by electrophoresis on sured to establish the resting levels of mitochondrial respiration. Cells 10% polyacrylamide gels and transferred to polyvinylidene fluoride were then treated with an ATP synthase blocker (oligomycin; 1 μM) membranes. ENT3 expression was detected by Western blotting with a to determine respiration used for ATP turnover and this was followed goat anti-ENT3 antibody (1/200 dilution) and a horseradish peroxidase- by treatment with a mixture of rotenone and antimycin A (both at linked anti-goat secondary antibody (1/7500 dilution). β-actin expression 1 μM) to inhibit mitochondrial respiration and allow measurement was determined in the same samples (1/250 and 1/5000 dilutions of the of non-mitochondrial respiration. OCR values were normalised to total primary antibody and a horseradish peroxidase-linked anti-mouse protein/well using the BCA protein assay and corrected for non- secondary antibody, respectively) and ENT3 expression relative to β- mitochondrial respiration [18,19]. actin was quantified by densitometric scanning using the Syngene GeneTools image analysis software. Enrichment of MIN6 β-cell mitochon- 2.12. Apoptosis dria was determined by immunoprobing protein extracts with an anti- body directed against the mitochondrial voltage-dependent anion MIN6 β-cells growing on 96-well plates (30,000 cells/well) were channel, VDAC (1/500 dilution) and a horseradish peroxidase-linked maintained for 24 h at pH 5.5, the optimal pH for ENT3 function, in anti-rabbit secondary antibody (1/5000 dilution). the absence or presence of a cytokine cocktail (0.05 U/μl IL-1β,1U/μl TNFα and 1 U/μlIFNγ). Some experiments were carried out in the presence of a non-selective ENT inhibitor dipyridamole (10 μM), 2.8. SiRNA-mediated ENT3 knockdown and in others ENT3 was depleted from MIN6 β-cells or mouse islets for 16 h by transient transfection with ENT3 siRNAs. Cells and islets MIN6 cells were transiently transfected with the ON-TARGETplus were then incubated in the absence or presence of the cytokine cocktail SMARTpool Slc29a3 siRNAs, which are a mixture of 4 individual siRNAs for a further 24 h. Apoptosis was determined by measuring caspase 3/7 targeting ENT3 (each at 100 nM), using the AMAXA nucleofection activities using the Caspase-Glo 3/7 luminescent assay following the system according to the manufacturer's instructions. For ENT3 knock- manufacturer's instructions. down in islets, groups of 250 islets were briefly exposed to cell dissoci- ation solution to allow access of the siRNAs to cells within the islets, 2.13. Insulin secretion then exposed to 1 μM of the Accell SMARTpool Slc29a3 siRNAs. Control MIN6 β-cells and mouse islets were transiently transfected with 100 nM Groups of three islets that had been exposed to ENT3 siRNAs or green scrambled non-silencing siRNAs and 1 μM Accell green fluorescent fluorescent nontargeting control siRNAs for 40 h were pre-incubated in nontargeting control siRNAs respectively and maintained under the a physiological salt solution [20] containing 2 mM glucose for 2 h at same conditions as the ENT3 siRNAs-treated cells and islets. The 37 °C. Islets were then incubated in the presence of 2 mM glucose, transfected cells and islets were transferred to appropriate culture vessels 20 mM glucose or 20 mM glucose plus 500 nM of 4β phorbol myristate followed by maintenance in a humidified incubator (37 °C, 95% air/5% acetate (PMA) for 1 h, after which samples of the supernatant were

CO2) before being harvested for protein analysis by western blotting removed for quantification of insulin by radioimmunoassay [21]. (24–36 h of incubation), or for ATP measurements, monitoring of cellular bioenergetics, caspase 3/7 assays, mitochondrial DNA (mtDNA) quantifi- 2.14. Statistical analyses cation and static insulin secretion experiments (40 h of incubation). Numerical data are expressed as means ± SEM of multiple experi- ments. All statistical comparisons were made by Student's t tests, two- 2.9. Mitochondrial DNA content way ANOVA or Mann–Whitney tests, as appropriate. Statistical signifi- cance was assumed at p b 0.05. Total genomic DNA from MIN6 β-cells that had been transfected with scrambled siRNAs or ENT3 siRNAs for 40 h was extracted using 3. Results the DNeasy blood and tissue kit and pre-treated by sonication for 10 min to remove dilution bias [17]. MtDNA content was assessed as 3.1. Expression of ENT3 by mouse and human β-cells the mitochondrial to nuclear genome ratio by real-time quantitative PCR using absolute quantification of a unique mitochondrial region It can be seen from Fig. 1A (upper panel, left) that PCR amplifications (sense: 5′-CTAGAAACCCCGAAACCAAA-3′; antisense: 5′-CCAGCTATCA with primers for mouse ENT3 produced a product of the expected size CCAAGCTCGT-3′; 125 bp product) relative to the nuclear gene Beta 2 (190 bp) using cDNA templates derived from MIN6 β-cells (lane 1), microglobulin (B2M, sense: 5′-ATGGGAAGCCGAACATACTG-3′;anti- mouse islets (lane 2) and mouse acinar cells (lane 4). However, as sense: 5′-CAGTCTCAGTGGGGGTGAAT-3′; 177 bp product) as described expected, no products were amplified using non reverse-transcribed elsewhere [17]. RNAs extracted from mouse islets (lane 3) and acinar cells (lane 5). B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 2089

Fig. 1. A: Products of RT-PCR using primers for mouse (upper panel) and human (lower panel) ENT3 (left) or ENT1 (right, mouse heart lane 1; MIN6 β-cells lane 2) were electrophoretically separated on 1.8% agarose gels. ENT3 amplicons of the correct sizes were amplified from cDNAs obtained from MIN6 β-cells (upper, lane 1), mouse islets (upper, lane 2), mouse acinar cells (upper, lane 4), human heart (lower, lane 1), human islets (lower, lane 2), β-cell enriched human islets (lower, lane 3) and human acinar cells (lower, lane 4). Non reverse-transcribed mouse islet (upper, lane 3) and acinar cell (upper, lane 5) RNAs or H2O (lower, lane 5) were used as negative controls. B: Wax-embedded sections of mouse pancreas were immunoprobed with an antibody directed against ENT3 and immunoreactivity was detected using an HRP-conjugated secondary antibody in the presence of diaminobenzidine. The left panel demonstrates that ENT3 is predominantly expressed by islet cells, with lower expression in the exocrine pancreas. The right panel shows the absence of immunostaining in a pancreas section incubated only with the secondary antibody. C: Wax-embedded sections of human pancreas were immunostained with antibodies directed against ENT3 and insulin, and immunopositive cells were detected using secondary antibodies conjugated to FITC (ENT3) or Texas Red (insulin). The merged image indicates the localisation of ENT3 in insulin-expressing β-cells in human islets.

ENT1 mRNA was also detected in MIN6 β-cells and mouse heart, which insulin-negative (green) islet cells are evident in the merged image, was used as a positive control (Fig. 1A upper panel, right), but ENT2 indicating that ENT3 is not exclusively expressed by islet β-cells. In mRNA was not detected in MIN6 cells. As shown in Fig. 1A(lower addition, there were some ENT3 immunoreactive cells in the exocrine panel), primers for human ENT3 produced single PCR products of pancreas, although ENT3 expression by acinar cells was considerably 170 bp with cDNAs from human heart (lane 1), human islets (lane 2), less abundant than in the endocrine pancreas. a β-cell-enriched human islet preparation (lane 3), and a low level of expression was also evident when human acinar cell cDNA was used 3.2. Co-localisation of ENT3 with mitochondria in β-cells as the template (lane 4). In addition, no products were amplified using H2O as a negative control (lane 5). Sequencing of amplicons con- Following identification of ENT3 expression by islet β-cells, as firmed that there was 100% homology between predicted sequences described above, further experiments were designed to investigate the and those of the amplified mouse and human products. subcellular localisation of ENT3 in β-cells. MIN6 β-cellsgrownonglass To determine whether ENT3 RNAs were translated into protein in coverslips were incubated with MitoTracker CMXRos (MitoTracker), a mouse and human pancreas and to examine the cellular localisation of mitochondria-specific fluorescent dye, to identify mitochondria in ENT3, immunohistochemical analysis of pancreas sections was carried these cells. After formaldehyde-fixation and acetone-permeabilisation, out. Fig. 1B (left panel) shows immunostaining of ENT3 in mouse MIN6 β-cells were then incubated with the ENT3 antibody and also pancreas, where it can be seen that it is abundantly expressed by islet with DAPI to visualise nuclei. It can be seen that MitoTracker labelling cells, with much less immunoreactivity identified in the exocrine was observed in the cytoplasm surrounding DAPI-stained nuclei, as pancreas, indicating a predominant localisation of ENT3 to mouse islets. expected (Fig. 2A). ENT3 immunoreactivity was observed in the cyto- A section stained only with an HRP-conjugated secondary antibody plasm, and an overlay of the images revealed co-localisation of ENT3 (right panel) indicated no islet immunostaining, as expected. To further immunoreactivity and MitoTracker staining in MIN6 β-cells (Fig. 2A investigate the localisation of ENT3 immunoreactivity in the mixed cell and arrows in panels i and iii), indicating ENT3 localisation to β-cell populations of whole islets, wax-embedded sections of human pancreas mitochondria. This was confirmed by detection of ENT3 immunoreac- were double-stained with antibodies directed against ENT3 (Fig. 1C; tivity in mitochondria-enriched fractions prepared from MIN6 β-cells, green) and insulin (Fig. 1C; red), which demonstrated that ENT3 was as shown in Fig. 2B. In these experiments equivalent amounts of protein present in insulin-expressing human β-cells. Some ENT3-positive, (50 μg) from MIN6 cell crude lysate and a MIN6 cell mitochondria- 2090 B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095

Fig. 2. A: MIN6 β-cells grown on glass coverslips were incubated with MitoTracker before being fixed and incubated with an ENT3 antibody. The panels show MIN6 β-cells stained with MitoTracker (i, mitochondria), DAPI (ii, nuclei) and ENT3 (iii). Arrows in panels i and iii, and the merged image (iv) demonstrate co-localisation of ENT3 to β-cell mitochondria. B: 50 μgof MIN6 β-cell crude lysate (lane 1) and 50 μgofMIN6β-cell mitochondria-enriched fraction (lane 2) were separated on a 10% polyacrylamide gel and ENT3 and VDAC expression in these samples was detected by Western blot analysis. C: Dispersed human islet cells adherent on glass coverslips were incubated with MitoTracker (i) before being fixed and incubated with an insulin antibody (ii), to identify β-cells, and with an ENT3 antibody (iii). The merged images (iv) demonstrate co-localisation of ENT3 to mitochondria in insulin-expressing human β-cells. enriched fraction were separated by polyacrylamide gel electrophoresis for the mitochondrial voltage-dependent anion channel, VDAC, shown and ENT3 expression in these preparations was measured by Western in the lower panel of Fig. 2B. With 50 μg of total MIN6 cell crude lysate blot analysis using the same ENT3 antibody that had been used for there was little expression of ENT3 (lane 1, upper panel), and VDAC immunohistochemistry. In addition, membranes were immunoprobed immunoreactivity was not detectable in this lysate (lane 1, lower panel). B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 2091

However, both ENT3 and VDAC were readily detectable in 50 μgofthe mitochondria-enriched fraction (lane 2), consistent with the mitochon- drial localisation of ENT3 identified by the immunohistochemical staining shown in Fig. 2A. Dispersed human islet cells were also stained with MitoTracker, then incubated with antibodies directed against insulin and ENT3 to determine whether ENT3 was also present in human islet β-cell mito- chondria. Consistent with the observations in mouse MIN6 β-cells, ENT3 immunoreactivity was found closely associated with mitochon- dria labelled with MitoTracker in insulin-expressing human islet β- cells (Fig. 2C).

3.3. Depletion of β-cell ENT3 expression by siRNAs

The specific role of ENT3 in β-cells was investigated by transiently transfecting MIN6 β-cells with synthetic ON-TARGETplus SMARTpool Slc29a3 siRNAs, which are a mixture of four individual siRNAs targeting ENT3, provided as a single reagent. Control cells were transfected with non-silencing siRNAs. The extent of depletion was determined by examining ENT3 protein expression in these cells relative to β-actin expression in the same samples. It can be seen from the western blots in Fig. 3A and densitometric quantifications in Fig. 3B that exposure to siRNAs for 24 h did not affect ENT3 protein expression, but there was a 59% decrease in its expression 36 h after transfection. The conse- quences of ENT3 depletion were therefore examined 40 h after transfec- tion. Under the conditions where ENT3 was depleted from β-cells by treatment with siRNAs there were no changes in expression of the relat- ed nucleoside transporter, ENT1, as shown in Fig. 3C.

3.4. Depletion of ENT3 increases β-cell mtDNA content and impairs mitochondrial function

Mitochondria contain their own DNA genome and changes in mtDNA content have been proposed as a biomarker of mitochondrial dysfunction [17,22,23]. Mitochondrial nucleoside transport via ENTs may be important in mitochondrial DNA synthesis and repair so exper- iments were performed to investigate if ENT3 depletion affected mtDNA content. It can be seen from Fig. 4AthatMIN6β-cells in which ENT3 had been depleted by siRNA-induced knockdown for 40 h showed a signifi- Fig. 3. A: Western blotting for ENT3 (upper panel) and β-actin (lower panel) expression cant increase in mtDNA content, as evidenced by a 1.7-fold increase in by MIN6 β-cells 24 h and 36 h after transfection with scrambled siRNAs (control) or fi levels of mitochondrial DNA relative to the nuclear gene B2M [17,24,25]. siRNAs directed against ENT3 (ENT3 siRNA). B: Densitometric quanti cation of ENT3 expression relative to β-actin following exposure of MIN6 cells for 24 h or 36 h to scrambled MIN6 cell oxygen consumption rate (OCR) was measured using the siRNAs (black bars) or siRNAs directed against ENT3 (grey bars). C: Quantification of ENT1 Seahorse XF analyser to assess β-cell mitochondrial function following mRNA relative to β-actin following exposure of MIN6 cells for 36 h to scrambled siRNAs ENT3 depletion. MIN6 β-cells in which ENT3 expression was inhibited (black bar) or siRNAs directed against ENT3 (grey bar). showed a 1.5-fold increase in basal OCR. An ATP synthase blocker, oligomycin, was used to block mitochondrial respiration used for ATP production (ATP-linked OCR) and this parameter was increased by 1.4-fold when ENT3 was depleted, indicating that there was a higher 3.5. Effects of inhibition and depletion of ENT3 on β-cell apoptosis and energy demand in these cells (Fig. 4B). In addition, significantly higher insulin secretion levels of proton leak (1.3-fold) were observed following ENT3 deple- tion, suggesting reduced mitochondrial efficiency (Fig. 4B). Oxygen In addition to energy production, mitochondria play a crucial role in consumed in other intracellular processes that are not associated regulating cell death pathways, and dysfunction of mitochondria causes with mitochondrial respiration was determined by exposing MIN6 apoptotic cell death. Experiments were thus performed to investigate cells to a mixture of rotenone (inhibitor of complex I) and antimycin if inhibition of ENT3 affects β-cell apoptosis, and since ENT3 shows A (blocker of complex III) to completely block mitochondrial respira- maximal activity under acidic conditions [6] these experiments were tion. It was found that non-mitochondrial respiration was also performed at pH 5.5. Maintaining the MIN6 β-cells at pH 5.5 did not increased significantly following ENT3 depletion (Fig. 4B), suggest- affect basal or cytokine-induced apoptosis, with comparable caspase ing an increase in the activity of cellular oxidases that are linked to 3/7 activities to those seen in our earlier experiments with MIN6 β- cytosolic reactive oxygen species (ROS) production [26,27].Consis- cells performed at pH 7.4 [28]. In addition, western blotting of MIN6 cell tent with impaired mitochondrial function following ENT3 deple- extracts after exposure to mixed cytokines for 24 h indicated that cyto- tion, MIN6 β-cells exposed to ENT3 siRNAs for 40 h showed a small kine treatment did not induce changes in ENT3 expression (optical densi- but significant reduction in ATP synthesis (control: 828 ± 26 × 103 ty ratio of ENT3/β-actin, control: 0.068; cytokine-treated: 0.066). MIN6 β- luminescence units; ENT3 siRNA: 751 ± 20 × 103 luminescence cells that had been incubated for 24 h with dipyridamole, an ENT inhibi- units; n = 8; P b 0.05). tor, demonstrated a small yet significant (P b 0.05) increase in the low, 2092 B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095

Fig. 4. A: Mitochondrial DNA content of MIN6 β-cells that had been transfected with non-silencing scrambled siRNAs (black bar) or ENT3 siRNAs (grey bar) was determined by qPCR amplification of mitochondrial DNA which was then normalised to the nuclear gene B2M. Significantly higher mitochondrial DNA content was observed following depletion of ENT3. Mean + SEM, n = 6, ***P b 0.001. B: A seahorse extracellular flux analyser (XFe96) and XF cell Mito stress kit was used to quantify basal and ATP-linked respiration, as well as proton leak and non-mitochondrial respiration, in MIN6 β-cells that had been transfected with non-silencing scrambled siRNAs (black bars) or ENT3 siRNAs (grey bars). Mean + SEM, n = 15, ***P b 0.001. basal level of β-cell apoptosis. In addition, the pro-apoptotic effects of syndromes was due to disrupted apoptotic cell clearance, lysosomal mixed cytokines on MIN6 β-cell caspase-3/7 activities were also signif- defects and macrophage accumulation [35]. The mechanisms underly- icantly enhanced by the presence of dipyridamole (Fig. 5A). These ing the diabetes phenotype in individuals with PHID have not been observations are consistent with ENT3 playing an anti-apoptotic role established, and the reduced lifespan of ENT3 knockout mice [35] in β-cells, but should be interpreted with caution because dipyridamole means that they are not an ideal model for defining how ENT3 disruption does not fully inhibit ENT3 activity, and it also inhibits other members of contributes to β-cell loss and hypoinsulinaemia. The current study there- the ENT family [29–31], including ENT1, which was also detected in fore used in vitro techniques to characterise ENT3 expression by β-cells, MIN6 β-cells (Fig. 1A). and identify the effects of ENT3 depletion on β-cell mitochondrial func- To investigate the specificroleofENT3inβ-cell apoptosis its expres- tion and apoptosis. sion was down-regulated by transiently transfecting MIN6 β-cells with We have demonstrated, for the first time, that ENT3 is expressed by ENT3 siRNAs. It can be seen from Fig. 5B that MIN6 β-cells in which both mouse and human pancreas, with a higher abundance in islets ENT3 had been depleted by siRNA-induced knockdown for 40 h showed than in pancreatic exocrine tissue. In the endocrine pancreas ENT3 co- increased apoptosis in response to cytokines, similar to the data obtained localises with insulin-producing β-cells, which is consistent with the in the presence of dipyridamole (Fig. 5A). Similar results were obtained clinical features of ENT3 mutations in PHID where patients require insu- when ENT3 was down-regulated in mouse islets by siRNAs treatment, lin therapy due to β-cell failure [2,3]. Expression of ENT3 by pancreatic which produced a significant increase in cytokine-induced islet cell acinar cells, as observed here, may play a role in the exocrine pancreatic caspase 3/7 activities (Fig. 5C), further confirming a regulatory role deficiency observed in some PHID cases [2] but this is not a consistent of ENT3 in β-cell apoptosis. feature of PHID, whereas insulin-dependent diabetes has been observed The effects of knocking down ENT3 in mouse islets (Fig. 6A) on insulin in over 80% of children with PHID [1]. secretory responses to glucose and the PKC activator PMA were also PHID shares features of the clinical manifestations encountered in investigated. It can be seen from Fig. 6B that depletion of ENT3 was mitochondrial and lysosomal disorders [33], and ENT3 possesses an without significant effect on basal or glucose-stimulated insulin secretion, endosomal/lysosomal targeting dileucine motif [6,7] and a putative nor on the potentiation of insulin release by PMA. mitochondria targeting signal [11]. An intracellular localisation is supported by observations that mutation of the dileucine to alanine 4. Discussion causes ENT3 to relocate to the cell surface [7], although a cell surface localisation of native ENT3 has been demonstrated in human placental Loss of function mutations in SLC29A3, the gene encoding ENT3, are tissues, suggesting that cellular localisation of ENT3 may be cell-type linked to pigmented hypertrichotic dermatosis with insulin-dependent dependent [11]. In the current study incubation of mouse and human diabetes (PHID), H syndrome, Faisalabad histiocytosis and Rosai– β-cells with a mitochondria-specific fluorescent dye and an anti-ENT3 Dorfman disease, all of which are autosomal recessive disorders with antibody indicated that ENT3 is localised to these organelles in β-cells, histiocytic infiltration [1,32–34]. Recent studies using mice in which an observation that was confirmed by Western blotting of a β-cell ENT3 was globally deleted suggested that the histiocytosis in these mitochondria-enriched fraction. B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 2093

Fig. 6. A: Mouse islets were exposed to green fluorescent nontargeting control siRNAs (black bar) or ENT3 siRNAs (grey bar) for 36 h and ENT3 mRNA levels were quantified by qPCR. Mean + SEM, n = 3, *P b 0.05. B: Groups of three mouse islets that had been transfected with green fluorescent nontargeting control siRNAs (black bars) or ENT3 siRNAs (grey bars) were incubated in media supplemented with 2 mM glucose, 20 mM glucose or 20 mM glucose + 500 nM PMA for 1 h and insulin secretion was quantified by radioimmunoassay. Mean + SEM, n = 10.

Fig. 5. A: MIN6 β-cells were maintained for 24 h in the absence or presence of a cytokine excessive build up of nucleosides that are required for the synthesis of cocktail, and exposed to 10 μM of the non-selective ENT inhibitor dipyridamole (grey bars) fl or the appropriate concentration of vehicle (black bars). Mean + SEM, n = 8, *P b 0.05. mtDNA,whichinturncausesoxidativestressandin ammation associat- BandC:MIN6β-cells (B) and mouse islets (C) that had been transfected with non-silencing ed with increased mtDNA content, leading to mtDNA damage/mutation scrambled siRNAs (black bars) or ENT3 siRNAs (grey bars) for 16 h were maintained for and mitochondrial dysfunction [22,23,43–45]. This theory is supported another 24 h in the absence or presence of a cytokine cocktail and caspase 3/7 activities by our experimental observations of a 1.7-fold increase in mitochondrial were measured by a luminescence assay. Mean + SEM, n = 8, ***P b 0.001. copy number following siRNAs-mediated knockdown of β-cell ENT3 expression. This mitochondrial localisation of ENT3 in β-cells is likely to be of To investigate whether perturbation of β-cell ENT3 expression and physiological relevance given the essential role of mitochondria in cellu- the consequent elevation in mtDNA content affected mitochondrial lar metabolism and the regulation of apoptosis [36–39]. Mitochondria function, MIN6 β-cell bioenergetics were measured following ENT3 are known to produce large quantities of adenosine from adenine and depletion. MIN6 cells have been used in earlier studies for measurements pyridine [40] and a mitochondrial equilibrative adenosine of cellular bioenergetics [46,47], which indicated that these cells are an transport system has been identified that can export adenosine formed appropriate model for these experiments. Our studies indicated that in the mitochondria across the mitochondrial membrane [12].It MIN6 β-cells in which ENT3 had been knocked down by transient trans- appears that accumulation of mitochondrial adenosine may have dele- fection with siRNAs exhibited significantly higher levels of basal respira- terious effects on cell function since adenosine A1 receptors have been tion, which was caused both by enhanced ATP-linked respiration and identified in mitochondria and A1 deletion is associated with increased proton leak. In addition, oxygen consumption in intracellular reductions in apoptotic cell death [41]. These observations suggest processes that are not associated with mitochondrial respiration (non- that blocking adenosine's actions via mitochondrial A1 receptors can mitochondrial respiration) was also significantly elevated following protect cells, and it has been proposed that this reduction in apoptosis ENT3 depletion. These findings suggest that depletion of ENT3 in β-cells was secondary to inhibition of mitochondrial Ca2+ overload and can promote an energy crisis due to increased energy demand and spill- decreased ROS generation [41]. age of electrons, which is likely to be caused by damage to the mitochon- It has been reported that PHID hENT3 mutations cause mis-targeting drial membrane and mitochondrial complexes. Sustained, high energy of mutant ENT3 away from mitochondria and these mutants show demand could lead to the loss of bioenergetic control, and compromised reduced adenosine transport activity [42]. Thus, by influencing the energy production in mitochondria will trigger apoptotic cell death [48]. concentration of adenosine that is present within mitochondria, ENT3 In parallel, dysfunctional mitochondria release proteins such as cyto- can regulate mitochondrial function and modifications in the intra- chrome C, Smac, EndoG and AIF that activate caspases or induce DNA mitochondrial nucleoside pool through inactivating mutations may fragmentation and chromatin condensation, leading to induction of culminate in deleterious consequences on mitochondria function. It is apoptosis [36–39]. Furthermore, increased non-mitochondrial respiration therefore possible that depletion of ENT3 in β-cell mitochondria causes causes additional cytosolic ROS production, which may further contribute 2094 B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 to oxidative stress, leading to apoptotic cell death through both the Author contribution extrinsic and intrinsic pathways [26,27]. The possibility that the alterations in cellular bioenergetics following B Liu, A Czajka, A Malik, K Hussain, PM Jones and SJ Persaud were ENT3 disruption is associated with enhanced β-cell apoptosis was involved in the study concept and design; B Liu and A Czajka were investigated using dipyridamole, a global ENT inhibitor that reduces responsible for performing the experiments and carrying out data analy- hENT3-mediated adenosine transport [7,11]. In our experiments the sis. B Liu and SJ Persaud drafted the manuscript; all authors read and presence of dipyridamole enhanced β-cell caspase 3/7 activities, suggest- approved the final paper. ing that impairing nucleoside transport via ENTs provokes apoptosis. However, we found that β-cells also express mRNA encoding ENT1, but Funding not ENT2, so this increased apoptosis could not be definitively ascribed to ENT3 inhibition. ENT3 was therefore specifically depleted from β- We are grateful to Diabetes UK for the grant support [grant number: cells using siRNAs, under the same conditions that were used for the 06/0003387]. bioenergetics studies. Under conditions where ENT3 expression was reduced in MIN6 β-cells there was a significant increase in caspase 3/7 activities and experiments performed on mouse islets indicated that Acknowledgements ENT3 knockdown also resulted in enhanced apoptosis, demonstrating that the cell line data were replicated in primary tissue. We are grateful to Professor Junichi Miyazaki (University of Osaka, Interestingly, however, knockdown of ENT3 in primary islets was not Osaka, Japan) for the provision of MIN6 β-cells. We thank the families associated with impaired basal or glucose-stimulated insulin release, of organ donors for making available the human pancreases that were nor was secretion in response to the PKC activator PMA compromised. used in some of the experiments. The authors declare that there is no Although these insulin secretion data are apparently inconsistent with duality of interest associated with this manuscript. the observations of mitochondrial dysfunction and enhanced apoptosis following ENT3 knockdown, there are several possible explanations for References this. First, the extent of knockdown in primary islets was less than in MIN6 β-cells (36% vs 59%), so the majority of islet cells still had active [1] J. Prendiville, M. Rogers, A. Kan, et al., Pigmented hypertrichotic dermatosis and ENT3. Second, in the apoptosis experiments caspase 3/7 activities were insulin dependent diabetes: manifestations of a unique genetic disorder? Pediatr. Dermatol. 24 (2007) 101–107. only elevated in ENT3-depleted cells after the stress of cytokine expo- [2] K. Hussain, R. Padidela, R.R. Kapoor, et al., Diabetes mellitus, exocrine pancreatic sure, while basal apoptosis was unaffected. Third, it was evident that deficiency, hypertrichosis, hyperpigmentation, and chronic inflammation: confirma- – some islets that had been exposed to ENT3 siRNAs for 40 h had tion of a syndrome, Pediatr. Diabetes 10 (2009) 193 197. [3] S.T. Cliffe, J.M. Kramer, K. Hussain, et al., SLC29A3 gene is mutated in pigmented fragmented and only whole islets are picked for the secretion experi- hypertrichosis with insulin-dependent diabetes mellitus syndrome and interacts ments, which will not capture potential secretory dysfunction in β- with the insulin signaling pathway, Hum. Mol. Genet. 18 (2009) 2257–2265. cells of disrupted islets. It therefore seems likely that reductions in [4] S.A. Baldwin, P.R. Beal, S.Y. Yao, A.E. King, C.E. Cass, J.D. Young, The equilibrative nucle- fl — – β oside transporter family, SLC29, P ugers Arch. Eur. J. Physiol. 447 (2004) 735 743. ENT3 prime the -cells to enter a cascade of mitochondrial dysfunction [5] F.E. Parkinson, V.L. Damaraju, K. Graham, et al., Molecular biology of nucleoside and cellular loss, which will ultimately cause hypoinsulinaemia. transporters and their distributions and functions in the brain, Curr. Top. Med. Thus, our data are consistent with the occurrence of diabetes in indi- Chem. 11 (2011) 948–972. [6] J.D. Young, S.Y. Yao, L. Sun, C.E. Cass, S.A. Baldwin, Human equilibrative nucleo- viduals exhibiting loss of function ENT3 mutations being a consequence side transporter (ENT) family of nucleoside and nucleobase transporter proteins, of impaired mitochondrial nucleoside transport [42], leading to increased Xenobiotica 38 (2008) 995–1021. mtDNA expression and mitochondrial dysfunction, which ultimately [7] S.A. Baldwin, S.Y. Yao, R.J. Hyde, et al., Functional characterization of novel human β and mouse equilibrative nucleoside transporters (hENT3 and mENT3) located in leads to the apoptotic loss of -cells. These experimental observations intracellular membranes, J. Biol. Chem. 280 (2005) 15880–15887. are supported by the phenotype of PHID patients and of previously pub- [8] R.J. Hyde, C.E. Cass, J.D. Young, S.A. Baldwin, The ENT family of eukaryote nucleoside and lished mitochondrial diabetes syndromes. Thus, PHID is an unusual condi- nucleobase transporters: recent advances in the investigation of structure/function fi tion in which insulin-dependent diabetes is autoantibody-negative and is relationships and the identi cation of novel isoforms, Mol. Membr. Biol. 18 (2001) 53–63. therefore not secondary to autoimmune destruction of β-cells, as occurs [9] B. Li, L. Gu, L. Hertz, L. Peng, Expression of nucleoside transporter in freshly isolated in archetypal type 1 diabetes. In addition, the age of onset of diabetes in neurons and astrocytes from mouse brain, Neurochem. Res. 38 (2013) 2351–2358. PHID(median12years)isconsistentwithagraduallossofβ-cell mito- [10] I.V. Sandoval, S. Martinez-Arca, J. Valdueza, S. Palacios, G.D. Holman, Distinct reading β of different structural determinants modulates the dileucine-mediated transport chondrial homeostasis, ultimately resulting in -cell apoptosis. Further- steps of the lysosomal LIMPII and the insulin-sensitive glucose more, some mtDNA mutations show strong associations with diabetes transporter GLUT4, J. Biol. Chem. 275 (2000) 39874–39885. [49] and there are certain parallels between diabetes in PHID and mito- [11] R. Govindarajan, G.P. Leung, M. Zhou, C.M. Tse, J. Wang, J.D. Unadkat, Facilitated mitochondrial import of antiviral and anticancer nucleoside drugs by human equili- chondrial diabetes, which most commonly occurs through mitochondrial brative nucleoside transporter-3, Am. J. Physiol. Gastrointest. Liver Physiol. 296 DNA-encoded tRNA gene mutations. In these mutant mitochondria there (2009) G910–G922. is disruption of ATP synthesis and progressive reductions in β-cell func- [12] A. Jimenez, D. Pubill, M. Pallas, et al., Further characterization of an adenosine transport system in the mitochondrial fraction of rat testis, Eur. J. Pharmacol. 398 (2000) 31–39. tion, which ultimately leads to diabetes [50]. [13] J.E.Bowe,A.J.King,J.S.Kinsey-Jones,et al., Kisspeptin stimulation of insulin In summary, the data presented here, of increased mtDNA copy secretion: mechanisms of action in mouse islets and rats, Diabetologia 52 (2009) number, mitochondrial dysfunction and enhanced apoptosis following 855–862. β [14] G.C. Huang, M. Zhao, P. Jones, et al., The development of new density gradient media ENT3 depletion, support an important role of ENT3 in maintaining - for purifying human islets and islet-quality assessments, Transplantation 77 (2004) cell number that is lost in PHID. 143–145. [15] S.J. Persaud, H.M. Roderigo-Milne, P.E. Squires, et al., A key role for beta-cell cytosolic phospholipase A(2) in the maintenance of insulin stores but not in the initiation of insulin secretion, Diabetes 51 (2002) 98–104. 5. Conclusion [16] B. Liu, Z. Hassan, S. Amisten, et al., The novel chemokine receptor, G-protein-coupled receptor 75, is expressed by islets and is coupled to stimulation of insulin secretion – We have shown that the equilibrative nucleoside transporter 3 and improved glucose homeostasis, Diabetologia 56 (2013) 2467 2476. β [17] A.N. Malik, R. Shahni, A. Rodriguez-de-Ledesma, A. Laftah, P. Cunningham, Mitochondri- (ENT3) is expressed in -cell mitochondria and that its depletion al DNA as a non-invasive biomarker: accurate quantification using real time quanti- increases mitochondrial DNA content, promotes mitochondrial dysfunc- tative PCR without co-amplification of pseudogenes and dilution bias, Biochem. tion and induces apoptosis. These effects of ENT3 loss in β-cells may con- Biophys. Res. Commun. 412 (2011) 1–7. [18] B.P. Dranka, G.A. Benavides, A.R. Diers, et al., Assessing bioenergetic function in tribute to the aetiology of pigmented hypertrichotic dermatosis with response to oxidative stress by metabolic profiling, Free Radic. Biol. Med. 51 (2011) insulin-dependent diabetes. 1621–1635. B. Liu et al. / Biochimica et Biophysica Acta 1852 (2015) 2086–2095 2095

[19] M.D. Brand, D.G. Nicholls, Assessing mitochondrial dysfunction in cells, Biochem. J. [34] I. Huber-Ruano, E. Errasti-Murugarren, V. Godoy, et al., Functional outcome of a 435 (2011) 297–312. novel SLC29A3 mutation identified in a patient with H syndrome, Biochem. Biophys. [20] G.O. Gey, M.K. Gey, The maintenance of human normal cells and tumor cells in Res. Commun. 428 (2012) 532–537. continuous culture: I. Preliminary report: cultivation of mesoblastic tumors and [35] C.L. Hsu, W. Lin, D. Seshasayee, et al., Equilibrative nucleoside transporter 3 normal tissue and notes on methods of cultivation, Am. J. Cancer 27 (1936) 45–76. deficiency perturbs lysosome function and macrophage homeostasis, Science [21] P.M. Jones, D.M. Salmon, S.L. Howell, Protein phosphorylation in electrically perme- 335 (2012) 89–92. abilized islets of Langerhans. Effects of Ca2+,cyclicAMP,aphorbolesterandnoradren- [36] Y.T. Hsu, K.G. Wolter, R.J. Youle, Cytosol-to-membrane redistribution of Bax and Bcl- aline, Biochem. J. 254 (1988) 397–403. X(L) during apoptosis, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 3668–3672. [22] A.N. Malik, R. Shahni, M.M. Iqbal, Increased peripheral blood mitochondrial DNA in [37] D.D. Newmeyer, D.M. Farschon, J.C. Reed, Cell-free apoptosis in Xenopus egg type 2 diabetic patients with nephropathy, Diabetes Res. Clin. Pract. 86 (2009) extracts: inhibition by Bcl-2 and requirement for an organelle fraction enriched in e22–e24. mitochondria, Cell 79 (1994) 353–364. [23] A.N. Malik, A. Czajka, Is mitochondrial DNA content a potential biomarker of mitochon- [38] C. Wang, R.J. Youle, The role of mitochondria in apoptosis*, Annu. Rev. Genet. 43 drial dysfunction? Mitochondrion 13 (2013) 481–492. (2009) 95–118. [24] A.L. Andreu, R. Martinez, R. Marti, E. Garcia-Arumi, Quantification of mitochondrial [39] X. Wang, The expanding role of mitochondria in apoptosis, Genes Dev. 15 (2001) DNA copy number: pre-analytical factors, Mitochondrion 9 (2009) 242–246. 2922–2933. [25] K. Gourlain, B. Amellal, Z. Ait Arkoub, N. Dupin, C. Katlama, V. Calvez, Quantitative [40] W. Henke, M. Ziegler, W. Dubiel, K. Jung, Adenosine formation by isolated rat kidney analysis of human mitochondrial DNA using a real-time PCR assay, HIV Med. 4 mitochondria, FEBS Lett. 254 (1989) 5–7. (2003) 287–292. [41] M.Alba,G.Silvia,C.Lanfranco,etal.,Adenosine A(1) receptors contribute to mitochon- [26] E.J. Moon, P. Sonveaux, P.E. Porporato, et al., NADPH oxidase-mediated reactive dria vulnerability to pro-oxidant stressors, Mitochondrion 10 (2010) 369–379. oxygen species production activates hypoxia-inducible factor-1 (HIF-1) via the [42] N.Kang,A.H.Jun,Y.D.Bhutia,N.Kannan,J.D.Unadkat,R.Govindarajan,Human ERK pathway after hyperthermia treatment, Proc. Natl. Acad. Sci. U. S. A. 107 equilibrative nucleoside transporter-3 (hENT3) spectrum disorder mutations (2010) 20477–20482. impair nucleoside transport, protein localization, and stability, J. Biol. Chem. [27] C. Piccoli, R. Ria, R. Scrima, et al., Characterization of mitochondrial and extra- 285 (2010) 28343–28352. mitochondrial oxygen consuming reactions in human hematopoietic stem [43] P. Maechler, C.B. Wollheim, Mitochondrial function in normal and diabetic beta-cells, cells. Novel evidence of the occurrence of NAD(P)H oxidase activity, J. Biol. Chem. Nature 414 (2001) 807–812. 280 (2005) 26467–26476. [44] R. Barbalat, S.E. Ewald, M.L. Mouchess, G.M. Barton, Nucleic acid recognition by the [28] B. Liu, H. Barbosa-Sampaio, P.M. Jones, S.J. Persaud, D.S. Muller, The CaMK4/CREB/ innate immune system, Annu. Rev. Immunol. 29 (2011) 185–214. IRS-2 cascade stimulates proliferation and inhibits apoptosis of beta-cells, PLoS [45] L.V. Collins, S. Hajizadeh, E. Holme, I.M. Jonsson, A. Tarkowski, Endogenously One 7 (2012) e45711. oxidized mitochondrial DNA induces in vivo and in vitro inflammatory responses, [29] J.R. Hammond, Interaction of a series of draflazine analogues with equilibrative J. Leukoc. Biol. 75 (2004) 995–1000. nucleoside transporters: species differences and transporter subtype selectivity, [46] R.J. Mailloux, A. Fu, C. Robson-Doucette, et al., Glutathionylation state of uncoupling Naunyn Schmiedeberg's Arch. Pharmacol. 361 (2000) 373–382. protein-2 and the control of glucose-stimulated insulin secretion, J. Biol. Chem. 287 [30] M.A. Zimmerman, E. Tak, S.F. Ehrentraut, et al., Equilibrative nucleoside (2012) 39673–39685. transporter (ENT)-1-dependent elevation of extracellular adenosine protects [47] M.R. Fernandez, M.D. Henry, R.E. Lewis, Kinase suppressor of Ras 2 (KSR2) regulates the liver during ischemia and reperfusion, Hepatology (Baltimore, Md) 58 tumor cell transformation via AMPK, Mol. Cell. Biol. 32 (2012) 3718–3731. (2013) 1766–1778. [48] C. Ruggiero, M. Ehrenshaft, E. Cleland, K. Stadler, High-fat diet induces an initial [31] T. Eckle, K. Hughes, H. Ehrentraut, et al., Crosstalk between the equilibrative nucleoside adaptation of mitochondrial bioenergetics in the kidney despite evident oxidative transporter ENT2 and alveolar Adora2b adenosine receptors dampens acute lung stress and mitochondrial ROS production, Am. J. Physiol. Endocrinol. Metab. 300 injury, FASEB J. 27 (2013) 3078–3089. (2011) E1047–E1058. [32] V. Molho-Pessach, I. Lerer, D. Abeliovich, et al., The H syndrome is caused by mutations [49] K.Majamaa,J.S.Moilanen,S.Uimonen,etal.,EpidemiologyofA3243G,the in the nucleoside transporter hENT3, Am. J. Hum. Genet. 83 (2008) 529–534. mutation for mitochondrial encephalomyopathy, lactic acidosis, and strokelike [33] N.V. Morgan, M.R. Morris, H. Cangul, et al., Mutations in SLC29A3, encoding an episodes: prevalence of the mutation in an adult population, Am. J. Hum. Genet. equilibrative nucleoside transporter ENT3, cause a familial histiocytosis syndrome 63 (1998) 447–454. (Faisalabad histiocytosis) and familial Rosai–Dorfman disease, PLoS Genet. 6 (2010) [50] J.A. Maassen, LM TH, E. Van Essen, et al., Mitochondrial diabetes: molecular mechanisms e1000833. and clinical presentation, Diabetes 53 (Suppl. 1) (2004) S103–S109.