Sodium/proton exchanger isoform 1 regulates intracellular pH and cell proliferation in human ovarian cancer

Sodium/proton exchanger isoform 1 regulates intracellular pH and cell proliferation in human ovarian cancer Carlos Sanhueza a, ⁎, Joaquín Araos a , Luciano Naranjo a , Fernando Toledo a,b , Ana R Beltrán c,d , Marco A Ramírez a,d , Jaime Gutiérrez a,e ,FabiánPardo a,f ,AndreaLeiva a , Luis Sobrevia a,g,h, ⁎⁎ a Cellular and Molecular Physiology Laboratory (CMPL), Division of Obstetrics and Gynaecology, School of Medicine, Faculty of Medicine, Pontificia Universidad Católica de Chile, Santiago 8330024, Chile b Department of Basic Sciences, Faculty of Sciences, Universidad del Bío-Bío, Chillán 3780000, Chile c Department of Education, Faculty of Education, Universidad de Antofagasta, Antofagasta 1270300, Chile d Biomedical Department, Faculty of Health Sciences, Universidad de Antofagasta, Antofagasta 1270300, Chile e Cellular Signalling Differentiation and Regeneration Laboratory (CSDRL), Health Sciences Faculty, Universidad San Sebastian, Santiago 7510157,Chile f Metabolic Diseases Research Laboratory, Center of Research, Development and Innovation in Health - Aconcagua Valley, San Felipe Campus, School of Medicine, Faculty of Medicine, Universidad de Valparaíso, San Felipe 2172972, Chile g Department of Physiology, Faculty of Pharmacy, Universidad de Sevilla, Seville E-41012, Spain h University of Queensland Centre for Clinical Research (UQCCR), Faculty of Medicine and Biomedical Sciences, University of Queensland, Herston, QLD 4029, Queensland, Australia abstractarticle info Article history: Received 16 August 2016 Received in revised form 5 October 2016 Accepted 18 October 2016 Available online 21 October 2016 Cancer cells generate protons (H + ) that are extruded to the extracellular medium mainly via the Na + /H + exchanger 1 (NHE1), which regulates intracellular pH (pHi) and cell proliferation. In primary cultures of human ascites-derived ovarian cancer cells (haOC) we assayed whether NHE1 was required for pHi modulation and cell proliferation. Human ovary expresses NHE1, which is higher in haOC and A2780 (ovarian cancer cells) compared with HOSE cells (normal ovarian cells). Basal pHi and pHi recovery (following a NH 4 Cl pulse) was higher in haOC and A2780, compared with HOSE cells. Zoniporide (NHE1 inhibitor) caused intracellular acidification and pHi recovery was independent of intracellular buffer capacity, but reduced in NHE1 knockdown A2780 cells. Zoniporide reduced the maximal proliferation capacity, cell number, thymidine incorporation, and ki67 (marker of proliferation) fluorescence in haOC cells. SLC9A1 (for NHE1) amplification associated with lower overall patient survival. In conclusion, NHE1 is expressed in human ovarian cancer where it has a pro-proliferative role. Increased NHE1 expression and activity constitute an unfavourable prognostic factor in these patients. © 2016 Elsevier B.V. All rights reserved. Keywords: NHE1 pHi Human Ovarian cancer Cell proliferation 1. Introduction Ovarian cancer is frequently diagnosed at advanced stages [1,2]. Cancer cells generate protons (H + )[3], which are released by membrane transport systems resulting in low extracellular pH (pHo), i.e., acidity, but high intracellular pH (pHi), i.e., alkalization [4].The membrane transport systems involved in pHi control include the sodium (Na + )/H + exchangers (NHEs) [5–8]. NHE isoform 1 (NHE1) predominates and plays critical roles in cancer [4,8–10]. NHE1 activity causes intracellular alkalization and extracellular acidification [4,11]. Since NHE1 inactivation reduces proliferation of human tumour gastric [12] and small lung cancer [13] cells, NHE1 may act as a pro-proliferative factor in human tumour cells. Mechanisms of pHi control in the human ovary or ascites-cancer cells are unknown [8]. Since non-human ovary cells exhibit NHEs activity [14–18], NHE1 may control the pHi in the human ovary. However, there are not reports addressing expression or activity of NHE1 in human ovarian cancer cells [8]. In this study, we show that functional NHE1 is expressed in human ovarian cancer cells playing a critical role in pHi recovery and cell proliferation. Thus, NHE1 expression and activity are factors that could result in pHi-dependent modulation of human ovarian cancer cells proliferation. The latter could be relevant to the reduced survival of patients suffering from this disease. Biochimica et Biophysica Acta 1863 (2017) 81–91 Abbreviations: NHEs, Na + /H + exchangers; NHE1, NHE isoform 1; haOC, human ascites-derived ovarian cancer cells; K, cell growth rate; D t , doubling time; K/D t , maximal proliferation capacity; HMA, 5-N,N-hexamethylene amiloride; V-ATPases, V type ATPases; 1/ pHi E, pHi modulation efficiency; TCGA, The Cancer Genome Atlas; CNA, DNA copynumber alterations. ⁎Corresponding author. ⁎⁎ Correspondence to: L. Sobrevia Cellular and Molecular Physiology Laboratory (CMPL), Division of Obstetrics and Gynaecology, School of Medicine, Faculty of Medicine, Pontificia Universidad Católica de Chile, P.O. Box 114-D, Santiago 8330024, Chile. E-mail addresses: [email protected] (C. Sanhueza), [email protected] (L. Sobrevia). http://dx.doi.org/10.1016/j.bbadis.2016.10.013 0925-4439/© 2016 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbadis 2. Materials and methods 2.1. Study groups Patients included in this study (3 patients) coursed with ovarian cancer (high degree of ovarian serous papillary carcinoma stage IIIC or IV, and endometrioid mix type carcinoma with high degree serous differentiation) and were all subjected to primary surgery. Ovarian ascites samples were obtained from the Hospital Clínico UC-CHRISTUS in Santiago de Chile (collected by Dr Mauricio Cuello from Division of Obstetrics and Gynaecology, School of Medicine, Pontificia Universidad Católica de Chile). Ethnicity of patients included in this study was A B 0.0 0.5 1.0 1.5 2.0 2.5 Immunosignal for NHE1 (CTCF x106) Tumour epithelium Stroma * C NHE1 CK7 S Te NHE1 Fig. 1. NHE1 expression in human ovary. A. NHE1 and cytokeratin-7 (CK7) proteins (arrows) detected by immunofluorescence in human ovarian cancer tissue sections. B. NHE1 protein at the epithelial origin (Te) and stroma non-epithelial (S) regions (dotted line limits Te and S). C. NHE1 immunosignal in Te and S from B. Bars: 50 μm in A and 200 μminB.*Pb0.05. Mean ± S.E.M. (n = 3). 82 C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 Hispanic. The investigation conforms to the principles outlined in the Declaration of Helsinki. Ethics Committee approval from the Faculty of Medicine of the Pontificia Universidad Católica de Chile and informed consent of patients were obtained. 2.2. Cell culture Primary cultures of human ascites-derived ovarian cancer cells (haOC) were established by mixing equal amounts of freshly isolated ascites with MCDB 105 (Sigma-Aldrich, St Louis, MO, USA)/medium 199 (M199) (Gibco, Grand Island, NY, USA) (1:1 v/v) culture media containing 10% fetal bovine serum (FBS), 12.5 mmol/L NaHCO 3 , 4.5 mmol/L D-glucose, 12.5 mmol/L (4-(2-hydroxyethyl)- 1-piperazineethanesulfonic acid (HEPES), 100 IU/mL penicillin, and 100 μg/mL streptomycin (37°C, pH 7.4) (MCDB 105/M199 culture medium). Cells at 70% confluence were harvested with trypsin/ethylenediamine tetraacetic acid (EDTA) (0.05/0.2%, 3 minutes, 37°C) and grew up to passage 2[19]. The human ovary cell lines HOSE (normal ovarian surface epithelium) and A2780 (ovary cancer cells) (kindly provided by Dr Carmen Romero from the Hospital Clínico Universidad de Chile, Santiago, Chile) were maintained in Dulbecco’smodified Eagle’s F12 medium (DMEM-F12) (Gibco) containing high (5 mmol/L) D-glucose and supplemented with 10% FBS, 16 mmol/L NaHCO 3 , 15 mmol/L HEPES, 100 IU/mL penicillin and 100 μg/mL streptomycin (DMEM-F12 culture medium). 2.3. Cell proliferation assay HOSE (6.5 x10 3 cells/cm 2 ), A2780 (2.0 x10 4 cells/cm 2 ), and primary cultures of haOC (1.0 x10 4 cells/cm 2 ) were seeded in 24 well plates and cultured for 0-48 hours in MCDB 105/M199 for haOC or DMEM-F12 culture media for HOSE and A2780 in the absence (Control) or presence of 100 nmol/L zoniporide (NHE1 inhibitor) [20]. Cells were resuspended following trypsin/EDTA (0.05/0.2%, 3 minutes, 37°C) digestion and counted in a haemocytometer as described [21]. Cell growth rates (K) were derived from the exponential growth equation: K¼ln Ct ðÞ−ln Ci ðÞ t•ln eðÞ where tis time in culture, C t is number of cells at a given time in culture, C i is number of cells at the beginning of the experiment (i.e., t= 0 hour), and eis 2.7182. The Kvalues were expressed as number of cells x10 3 per cm 2 of cell culture surface per hour [21].Doublingtime(D t )forcell growth was derived from 0.6932/Kand expressed in hours. Corrected growth rates by the corresponding doubling times at any given time (i.e., maximal proliferation capacity) in culture was estimatedfrom K/D t . 2.4. Thymidine incorporation Cells in the logarithmic phase of cell growth (i.e., 40-45% confluence) were incubated with [ 3 H]thymidine (2 μCi/mL, 6-[ 3 H]thymidine, 6.7 Ci/mmol (NEN, Dreieich, Germany), 48 hours, 37°C) in MCDB105/M199 for haOC or DMEM-F12 culture media for HOSE and A2780 cells in the absence or presence of zoniporide. After this incubation period the cells were rinsed with phosphate-buffered salt (PBS) solution ((mmol/L) NaCl 130, KCl 2.7, Na 2 HPO 4 0.8, KH 2 PO 4 1.4 (pH 7.4, 4°C)) and exposed to 5% trichloroacetic acid (TCA, 200 mL, 10 minutes). TCA was removed and cells rinsed with 99% methanol (200 mL) and digested with 25 mmol/L formic acid for radioactivity determination as described [21]. A B haOC HOSE A2780 0 5 10 15 20 NHE1 (arbitrary units) HOSE A2780 haOC * * HOSE A2780 haOC NHE1 -Actin -actin/ Fig. 2. NHE1 protein abundance in human ovarian tumour cells. A. Western blot for NHE1 and β-actin (internal reference) protein abundance in HOSE, A2780, and haOC cells. Lower panel: NHE1/β-actin ratio densitometries normalized to 1 in HOSE. B. Immunofluorescence for NHE1 (arrows show NHE1). Bars: 10 μm. *Pb0.01. Mean ± S.E.M. (n = 3-7). 83C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 2.5. Measurement of pHi Cell monolayers in 96 well plates (70% confluence) were incubated for 10 minutes at 37°C with the fluorescent pH sensitive probe 2,7- bicarboxyethyl-5,6-carboxyfluorescein acetoxymethyl ester (BCECFAM, 12 μmol/L) (Sigma-Aldrich), as described [22]. Some experiments were also performed in cells incubated for 1 minute with BCECF-AM, and the results for pHi were not significantly different from 10 minutes incubation (not shown). The excess of probe was removed by rinse three times with control solution (CS) ((mmol/L) NaCl 145, KCl 5, NaH 2 PO 4 1Na 2 SO 4 1, CaCl 2 1.8, MgCl 2 1, HEPES 30, D-glucose 5 (pH 7.4, 37°C). The fluorescence ratios was registered in a fluorimeter Tecan M200Pro (Untersbergstr, Austria) after alternate cell excitation at 440 and 490 nm, and emission at 530 nm registered every 2 seconds interval for a period of 150 seconds. pHi was estimated by interpolation of emission ratios using standard calibration curves in cells incubated with 10 μmol/L nigericin in a calibrating solution ((mmol/L) KCl 130, NaCl 20, CaCl 2 1, MgCl 2 1, HEPES 5 (pH 6.2, 7.2, or 8.2)) as described [21,22]. Cells were incubated without or with 5 μmol/L 5-N,N- hexamethylene amiloride (HMA, NHEs general inhibitor) [23],100 nmol/L zoniporide [20],0.1μmol/L concanamycin A (V type ATPases (V-ATPases) inhibitor) [24],and10μmol/L Schering 28080 (H + /K + ATPase inhibitor) [25]. 2.6. Recovery of pH i The pH i recovery was examined by applying the NH 4 Cl pulse technique [21,22]. In brief, BCECF-AM loaded cells were incubated in CS until the basal pH i was stabilized (~3 minutes). Cells were exposed (2 minutes) to CS supplemented with NH 4 Cl (NH 4 Cl/CS solution) ((mmol/L) NaCl 121, KCl 5.4, CaCl 2 1, KH 2 PO 4 0.4, MgCl 2 0.5, MgSO 4 0.4, Na 2 HPO 4 0.3, HEPES 10, D-glucose 0.6, NH 4 Cl 20 (pH 7.4, 37°C)). After this incubation period the NH 4 Cl/CS solution was replaced by rinsing the cells with CS free of NH 4 Cl ( 0 Na + /CS) ((mmol/L) N-methyl-D-glucamine (NMDG) 120, KCl 5, CaCl 2 1.8, MgCl 2 1, HEPES 30, D-glucose 5 (pH 7.4, 37°C)), without or with HMA, zoniporide, concanamycin A and/or Schering 28080 as above. Recovery rates of pH i (dpHi/dt) were calculated from data collected for the first 60 seconds of the recovery (i.e., after removing the NH 4 Cl load) and fitted by a first order lineal regression as described [21,22]. The results were expressed in pH i units/minute. The NHEs-mediated component of dpHi/dt ( NHEs f) was estimated from: NHEs f¼ HMAα−δ HMAþCAþSchα−δ•100 where HMA αis dpHi/dt inhibited by HMA, HMA+CA+Sch αis dpHi/dt inhibited by HMA + concanamycin A + Schering, and δrepresents the corresponding remaining dpHi/dt (background) on each condition. The involvement of NHE1 ( NHE1 f)inthe NHEs fcomponent was estimated by NHE1 f= NHEs f– Z α,where Z αis dpHi/dt inhibited by zoniporide. The involvement of V-ATPase ( V f)andH + /K + ATPase ( H/K f) components were estimated as for NHEs fconsidering CA α(dpHi/dt inhibited by concanamycin A) or Sch α(dpHi/dt inhibited by Schering) and HMA+Z+Sch α(dpHi/dt inhibited by HMA + zoniporide + Schering) or HMA+Z+CA α(dpHi/dt inhibited by HMA + zoniporide + concanamycin A), respectively. Initial velocity for pHi recovery was derived from slope of linear phase of dpHi/dt adjusted to the one phase exponential association equation considering the least squares fit[26]: vi¼dpHi dt 1−e−ktðÞ  where v i is initial velocity, dpHi/dt is mayor recovery rate at a given time (t) (in this study t= 0.03 seconds) and eand kare constants. The capacity of cells to change the pHi value in certain units of pHi as a result in the dpHi/dt modifications (i.e., efficiency of pHi modulation (1/ pHi E)) was estimated from: 1=pHiE¼ΔdpHi=dtðÞ ΔpHi •α where ΔdpHi/dt is the change in dpHi/dt from control, ΔpHi is the change in pHi from control, and αis a given pHi unit. 2.7. Intrinsic buffering capacity The ability of intrinsic cellular components to buffer changes in pHi, i.e., intracellular buffer capacity (βi), was measured as described [21]. After determining the basal pHi the cells were incubated in a 0.5 mmol/L KCl-containing 0 Na + /CS plus Schering 28080 + concanamycin A (pH 7.4, 37°C) until the pHi was stabilized under this experimental condition (~3 minutes). Cells were then incubated in the latter solution containing decreasing concentrations of NH 4 Cl (50, 20, 10, 5, 2.5 or 1 mmol/L). To assay the effect of each of the concentrations of NH 4 Cl the cells were rinsed three times with the corresponding lower NH 4 Cl concentration used in this study. The βiwas calculated from the expression: βi¼ΔNHþ 4  i ΔpHi where the intracellular NH 4 + concentration ([NH 4 + ] i ) was obtained from the Henderson-Hasselbalch equation on the assumption that [NH 3 ] i Table 1 pHi and pHi recovery rates in human ovary cells. HOSE A2780 haOC pHi Control 7.391 ± 0.034 7.621 ± 0.044 * 7.840 ± 0.053 * HMA 7.262 ± 0.112 †7.574 ± 0.145 *†7.648 ± 0.095 *† Zoniporide (Z) 7.270 ± 0.079 †7.383 ± 0.161 †7.656 ± 0.069 *† Concanamycin A (CA) 7.601 ± 0.183 7.546 ± 0.031 7.858 ± 0.196 Schering 28080 (Sch) 7.692 ± 0.132 7.607 ± 0.043 7.782 ± 0.163 dpHi/dt (pHi units/minute) Control 0.114 ± 0.081 0.152 ± 0.025 * 0.194 ± 0.050 * HMA 0.018 ± 0.007 * 0.026 ± 0.017 * 0.004 ± 0.001 * Z 0.051 ± 0.006 *†0.026 ± 0.008 * 0.049 ± 0.006 *† CA 0.110 ± 0.010 ‡0.062 ± 0.018 *‡0.196 ± 0.086 ‡ Sch 0.106 ± 0.013 ‡0.080 ± 0.011 *‡0.157 ± 0.082 ‡ HMA + Z 0.025 ± 0.002 *§ 0.025 ± 0.017 *§ 0.001 ± 0.001 *§ HMA + CA 0.012 ± 0.015 *§ 0.034 ± 0.008 *§ 0.001 ± 0.005 *§ HMA + Sch 0.014 ± 0.009 *§ 0.037 ± 0.011 *§ 0.001 ± 0.030 *§ HMA + Z + CA 0.012 ± 0.029 *§ 0.039 ± 0.016 *§ 0.009 ± 0.001 *§ HMA + Z + Sch 0.025 ± 0.004 *§ 0.033 ± 0.018 *§ 0.005 ± 0.014 *§ HMA + CA + Sch 0.011 ± 0.002 *§ 0.012 ± 0.007 *§ 0.011 ± 0.002 *§ HMA + Z + CA + Sch 0.026 ± 0.016 *§ 0.006 ± 0.011 *§ 0.001 ± 0.023 *§ Z + CA 0.053 ± 0.032 * 0.031 ± 0.010 *# 0.047 ± 0.011 * Z + Sch 0.061 ± 0.057 * 0.022 ± 0.015 *# 0.037 ± 0.009 * Z + CA + Sch 0.051 ± 0.055 * 0.011 ± 0.015 *# 0.037 ± 0.009 * CA + Sch 0.112 ± 0.013 $ 0.033 ± 0.025 *£ 0.120 ± 0.051 $ The intracellular pH (pHi) was measured in BCECF-AM–preloaded HOSE, A2780 and human ascites ovarian cancer cells (haOC) in the absence (Control) or presence of 5 μmol/L 5-N,N-hexamethylene amiloride (HMA) (NHEs general inhibitor), 100 nmol/L zoniporide (Z) (NHE1 inhibitor), 0.1 μmol/L concanamycin A (CA) (V type ATPases inhibitor), or 10 μmol/L Schering 28080 (Sch) (H + /K + ATPase inhibitor) as described in Materials and Methods. Initial rates of pH i recovery (dpHi/dt) were determined in cells subjected to an acid pulse (NH 4 Cl assay) as above. In pHi section: *Pb0.05 versus Control in HOSE. †Pb0.05 versus corresponding Control values. In dpHi/dt section: *Pb0.05 versus Control in HOSE or in the corresponding cell type. †Pb0.05 versus all other corresponding values except Z + CA, Z + Sch, and Z + Ca + Sch. ‡Pb0.05 versus corresponding HMA and Z. §Pb0.05 versus corresponding values in Z, CA, or Sch. ¶Pb0.05 versus corresponding values in HMA (alone or in mix with other inhibitors), CA, and Sch. #Pb0.05 versus corresponding values in CA or Sch. $Pb0.05 versus corresponding values in HMA and Z (alone or in mix with other inhibitors), CA or Sch. £Pb0.05 versus all other corresponding values except CA and Sch. Values are mean ± S.E.M. (n = 3-22). 84 C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 (intracellular NH 3 ) was equivalent to [NH 3 ] o (extracellular NH 3 ), and Δ pHi is the fraction of change in units of pHi value. Knowing the dpHi/dt and the βivalues (at similar pHi values in each cell type), the rate of overall transmembrane H + flux (J H+ ) was calculated from the following expression: JHþ¼dpHi dt •βi 2.8. Western blotting Total protein was obtained from confluent cells rinsed (x2) with ice-cold PBS and harvested in 100 μL of lysis buffer (100 mmol/L NaCl, 10 mmol/L Na 4 P2O 7 x10-H 2 O, 10 mmol/L NaF, 1% Triton X100, 1 mmol/L sodium orthovanadate, 50 mg/mL leupeptin, 20 mmol/L HEPES (pH 7.4, 4°C) as described [22]. Cells were sonicated (6 cycles, 5 seconds, 100 W, 4°C) and total protein was separated by centrifugation (13500 g, 15 minutes, 4°C). Proteins (70 μg) were separated by polyacrylamide gel (8%) electrophoresis, transferred to Immobilon-P polyvinylidene difluoride membranes (BioRad Laboratories, Hertfordshire, UK) and probed with primary monoclonal rabbit anti- NHE1 (1:500 dilution, 12 hours, 4°C) or monoclonal mouse anti-β-actin (1:2500 dilution, internal reference) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) antibodies. The membranes were rinsed in PBS/0.1% Tween 20 (PBS-T) and further incubated (1 hour) in PBS-T/0.2% bovine serum albumin (BSA) containing secondary horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse antibodies (Thermo Scientific, Rockford, IL, USA). Proteins were detected by enhanced chemiluminescence in a ChemiDoc-It 510 Imagen System (UVP, LCC Upland, CA, USA) and quantified by densitometry [22]. 2.9. Immunofluorescence microscopy Paraformaldehyde-fixed, paraffin-embedded ovarian cancer serial sections (4 μm) were dewaxed and subsequently rehydrated with serial incubations using 100, 95, or 70% ethanol solutions. Histological slices were permeabilized (3% Tween 20, 10 minutes) and blocked with defatted-cow milk (5% in PBS, 1 hour, 22°C). NHE1, cytokeratin-7 (CK-7) (marker for epithelial cells) and (marker for proliferation) were immunolocalized following incubation (overnight at 4°C) with primary policlonal rabbit anti-NHE1 (1:200 dilution) (Santa Cruz Biotechnology), monoclonal mouse anti-CK-7 (1:200 dilution) (Dako, Glostrup, Denmark) or anti-ki67 (1:50 dilution) (Thermo Scientific, Rockford, IL, USA) in PBS containing 2% BSA (PBS/2% BSA) and followed by incubation (1 hour, 22°C) with the secondary antibody Alexa Fluor 568 goat anti-rabbit or anti-mouse IgG (H+L, λexc/λem: 578/603 nm, 1:750) (Life Technologies). 024681012 6.5 7.0 7.5 8.0 8.5 Time (minutes) pHi haOC HOSE A2780 + NH 4 Cl - NH 4 Cl 0.0 0.5 1.0 1.5 2.0 0.0 0.1 0.2 0.3 0.4 Time (seconds) pHi units haOC A2780 HOSE 6.0 6.5 7.0 7.5 8.0 0 10 20 30 40 50 pHi i ((mmol/L) /pHi units) HOSE A2780 haOC 0.0 0.5 1.0 1.5 2.0 2.5 JH+ ((mmol/L)/second) HOSE A2780 haOC * * 0 2 4 6 8 10 12 i ((mmol /L)/pHi units) HOSE A2780 haOC A ED C B Fig. 3. pHi recovery. A. HOSE, A2780, and haOC cells were exposed to 20 mmol/L NH 4 Cl (+NH 4 Cl). NH 4 Cl was removed (–NH 4 Cl) and initial rates of pHi recovery were calculated (see Methods). B. Initial velocity for pHi recovery as in A. C. Intracellular buffering capacity (βi) determined as described in Methods. D.βifor corresponding pHi from data in C. E.H + flux rates (J H+ ) (see Methods). *Pb0.05 versus all other values. Values are mean ± S.E.M. (n = 3-10). 85C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 Cells were grown on microscope cover glasses (6 x10 4 cells per slide) (Sail Brand, Shangai, China) in corresponding culture media to 70% confluence as described [27].Inbrief,cellswerefixed (4% paraformaldehyde, 15 minutes), permeabilized (0.1% Triton X-100, 10 minutes) and blocked (PBS/2% BSA, 1 hour, 22°C). NHE1 and ki67 were immunolocalized following incubation (overnight at 4°C) with primary policlonal rabbit anti-NHE1 (1:200 dilution) or monoclonal mouse anti-ki67 (1:50 dilution) in PBS/2% BSA and followed by incubation (1 hour, 22°C) with the secondary antibody Alexa Fluor 488 goat anti-mouse IgG (H+L, λexc/λem: 488/519 nm, 1:750 dilution) or Alexa Fluor 568 goat anti-rabbit IgG (H+L, λexc/λem: 578/603 nm, 1:750 dilution) (Life Technologies). Cell nuclei were stained with Hoescht 33342 (4 μmol/L, 10 minutes, 22°C) (Thermo Scientific). Images were obtained under an EVOS FL Imaging System (AMF 4300) (Life Technologies). Fluorescence quantification was performed by the method of Corrected Total Fluorescence Intensity (CTCF) as described [28] in regions of optical interest (ROI) in serial slices. ROI was defined as a section of the tissue that was positive (epithelial) or negative (stroma) for the epithelial marker CK-7 where NHE1 and ki67 staining was determined. Images were processed with Image J version 1.48v (Wayne Rusband NIH, USA). 2.10. NHE1 suppression Suppression of NHE1 expression was done using the commercially available short interference RNAs (siRNA) NHE-1 siRNA (h) or control sequence (siC) (Santa Cruz Biotechnology) following manufacturer’s instructions (http://datasheets.scbt.com/siRNA_ protocol.pdf). HOSE and A2780 cells knockdown for NHE1 ( KD-NHE1 HOSE and KD-NHE1 A2780, respectively) were generated. haOC cells were not viable in vitro under this approach, thus KD-NHE1 haOC cells was not possible. 2.11. TCGA data analysis Genomic, mRNA expression and clinical information of high-grade serous cancer were obtained in silico from The Cancer Genome Atlas Ovarian Cancer (https://confluence.broadinstitute.org/display/GDAC/ Home). Analyses of patient’s survival and NHE1 signalling pathway were performed in cBioPortal platform for Cancer Genomics (http:// www.cbioportal.org/study.do?cancer_study_id=ov_tcga) according to previously published protocols [29,30]. 2.12. Statistical analysis The values are mean ± S.E.M., where nindicates the number of different biological samples (3 ascites from patients with ovarian cancer) and corresponding cell cultures with 3-4 replicates per experiment. Comparisons between two groups were performed by means of Student’s unpaired t-test. The difference betweenmore than two groups was performed by ANOVA (one or two ways). If the ANOVA demonstrated a significant interaction between variables, post hoc analyses were performed by the multiple-comparison Bonferroni test. To estimate the correlation between NHE1 and ki67 expression in tissue samples, the correlation analysis was performed using Pearson’s correlation coefficient (r). Patient survival analyses were addressed by Kaplan Meier methodology using log-rank, Wilcoxon parameters. AC BD Fig. 4. NHE1 involvement in pHi recovery. A. pHi recovery rates (dpHi/dt) in HOSE, A2780, and haOC cells without (Control, 100%) or with 5 μmol/L 5-N,N-hexamethylene amiloride (HMA), 100 nmol/L zoniporide (Z), 0.1 μmol/L concanamycin A (CA), and/or 10 μmol/L Schering 28080 (Sch). B.dpHi/dt fraction mediated by NHEs ( NHEs f), V type ATPases ( V f), or H + /K + ATPase ( H/K f) components. Arrows indicate the dpHi/dt fraction mediated by NHE1 ( NHE1 f)of NHEs fcomponent (see Methods). ‘All’is all components of dpH/dt.C.WesternblotforNHE1andβ-actin (internal reference) protein abundance in HOSE and A2780 expressing NHE1 ( NHE1 HOSE, NHE1 A2780) or NHE1 knockdown ( KD-NHE1 HOSE, KD-NHE1 A2780) cells. Lower panel:NHE1/β-actin ratio densitometries normalized to 1 in NHE1 HOSE cells. D.dpH/dt in cells as in C. In A: *Pb0.04 versus Control, †Pb0.05 versus all other corresponding values, ‡Pb0.05 versus all other values except for Z in HOSE cells. In B: *Pb0.05 versus all other values, †Pb0.04 and ‡Pb0.05 versus corresponding values in HOSE and haOC cells. In C and D: *Pb0.05 versus corresponding values in NHE1 HOSE or NHE1 A2780. In D: †Pb0.05 versus all other values. Values are mean ± S.E.M. (n = 3-10). 86 C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 Pb0.05 value was considered statistically different. The statistical software GraphPad InStat 3.0b and GraphPad Prism 7.0b (GraphPad Software Inc., San Diego, CA, USA) were used for data analysis. Pb0.05 was considered statistically significant. 3. Results 3.1. NHE1 expression in human ovary The first goal of this study was to determine whether NHE1 was expressed in human ovarian cancer tissue. Ovarian cancer tissue was positive for NHE1 and CK7 (Fig. 1A). NHE1 was detected in tumour epithelium and stroma (Fig. 1B), with higher fluorescence at the epithelium, compared with stroma (Fig. 1C). We then assayed whether NHE1 was expressed in cultured human ovarian cancer cells. NHE1 protein was detected in all cell types (Fig. 2A) and was higher in haOC and A2780 compared with HOSE cells. NHE1 protein was detected in the cells and in the cell periphery (Fig. 2B). 3.2. Basal pHi and dpHi/dt Since NHE1 expression associates with pHi modulation in several tumour cells [4–10], we determined basal pHi and the potential functional involvement of this and other membrane transport mechanisms to regulate intracellular H + content. Basal pHi in haOC and A2780 was higher than in HOSE cells (Table 1). Zoniporide ABC D Fig. 5. NHE1 involvement in cell proliferation. A. HOSE, A2780, and haOC cells cultured in the absence (Control) or presence of 100 nmol/L zoniporide. Cell number was determined in a haemocytometer (see Methods). B. Cell number, thymidine incorporation, and ki67 positive cells in the absence (Control) or presence of zoniporide (+Zoniporide) as in A. C. Immunofluorescence for NHE1 and ki67 protein in human ovarian cancer tissue sections (bars: 50 μm). The graph represents NHE1 versus ki67 immunosignal adjusted to a first-order equation (Pearson’s correlation coefficient (r) = 0.83, goodness-of-fitR 2 =0.672,Pb0.0001, n = 32 ROIs in 3 patients). (D) Cell number in HOSE and A2780 expressing NHE1 ( NHE1 HOSE, NHE1 A2780) or NHE1 knockdown cells ( KD-NHE1 HOSE, KD-NHE1 A2780). Values are normalized to 1 in NHE1 HOSE or NHE1 A2780 cells. In B: *Pb0.05 versus Control. In D, *Pb0.04 versus corresponding NHE1 HOSE or NHE1 A2780. Values are mean ± S.E.M. (n = 3-10). Table 2 Effect of zoniporide on the parameters for human ovary cells proliferation. K(1/hour) D t (hour) K/D t (1/(hour) 2 ) HOSE Control 0.01702 ± 0.00477 41 ± 11 0.00042 ± 0.00011 Zoniporide 0.01131 ± 0.00402 61 ± 22 0.00019 ± 0.00007 * A2780 Control 0.03949 ± 0.00304 †17 ± 1 †0.00232 ± 0.00016 † Zoniporide 0.03328 ± 0.00401 †21 ± 3 †0.00158 ± 0.00023 *† haOC Control 0.01357 ± 0.00131 51 ± 1 0.00027 ± 0.00003 Zoniporide 0.00791 ± 0.00074 * 88 ± 1 * 0.00009 ± 0.00001 * HOSE, A2780, and human ascites ovarian cancer cells (haOC) cells were seeded in 24-well plates (2.05 cm 2 surface) and cultured for 48 hours in primary culture medium without (Control) or with 100 nmol/L zoniporide. Cell number was estimated by counting at different periods of time and cell growth rates (K) and doubling time (D t ) for cell growth was defined as described in Materials and Methods. K/D t represents corrected cell growth rate by doubling times (maximal proliferation capacity) of cells in culture. *Pb0.05 versus corresponding Control values. †Pb0.05 versus corresponding values in HOSE and haOC cells. Values are mean ± S.E.M. (n = 3-22). 87C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 and HMA, but not concanamycin A or Schering 28080 caused intracellular acidificationinallcelltypes.Toestimatethe capacity of cells to recover from an acidic pHi cells were subjected to the NH 4 Cl pulse assay [21–23]. The acidic pHi caused by the NH 4 Cl pulse was fully restored after ~3 minutes (Fig. 3A). Basal dpHi/dt (Table 1)andv i for dpHi/dt (Fig. 3B) in haOC (v i = 10.2 ± 0.001 pHi units/minute) and A2780 cells (v i = 9.8 ± 0.002 pHi units/minute) were higher compared with HOSE cells (v i =7.8± 0.001 pHi units/minute). 3.3. ßi and J H+ Since a change in basal pHi and dpHi/dt depends on the ßi capacity of cells [21], this parameter was estimated and used to determine J H+ . The results show that ßi increased as pHi value decreased (Fig. 3C) and ßi at equal pHi after a 20 mmol/L NH 4 Cl pulse (pHi HOSE = 6.95, pHi haOC = 6.93, pHi A2780 = 6.95) were similar in all cell types (Fig. 3D). In addition, J H + was higher in haOC and A2780 compared with HOSE cells (Fig. 3E). 3.4. NHE1, V-ATPase, and H + /K + ATPase involvement on dpHi/dt To know the relative contribution of the membrane transport mechanisms involved in the regulation of pHi, the dpH/dt was measured in the absence or presence of specific inhibitors for NHE1, V-ATPase, and H + /K + ATPase. The results show that HMA reduced dpHi/dt in all cell types (Table 1). Partial inhibition was detected in HOSE and A2780 cells; however, HMA abolished dpHi/dt in haOC cells (Fig. 4A). Zoniporide caused a partial reduction of dpHi/dt in HOSE and haOC cells, which was similar to HMA effect in A2780 cells. Concanamycin A and Schering 28080 caused partial reduction (47-59%) of dpHi/dt only in A2780 cells. The NHEs fcomponent accounted for a fraction of dpHi/dt similar in HOSE and haOC cells (Fig. 4B). However, NHEs faccounted for almost all dpHi/dt in A2780 cells. Involvement of NHE1 fcomponent in NHEs fresulted in a fraction of dpHi/dt that was similar in HOSE and haOC cells; however, in A2780 cells accounted for all the NHEs fcomponent. The V fand H/K f components were detected only in A2780 cells. With the aim to corroborate that NHE1 expression is required for a pHi recovery, cells knockdown for this protein were used to 0 1×10 7 2×10 7 3×10 7 SLC9A1, mRNA expression (RNA Seq V2 RSEM) (log2) Homologous deletion Shallow deletion Diploidism Gain Amplification 050100150200 0 25 50 75 100 Months Percent survival Amplification Homologous deletion Diploid Gain Shallow deletion A B CD Fig. 6. NHE1 expression and survival in patients with ovarian cancer. A. Analysis of the DNA copy-number alterations (CNA) for SLC9A1 in human ovarian cancer (The Cancer Genome Atlas Ovarian Cancer (TCGA) database) (the number of cases with the indicated genetic alterations (Genetic Alteration) is given). B. Relation between mRNA number of copies and alterations in CNA of SLC9A1 in patients with high-grade serous ovarian cancer. C. Representation of NHE1 pathway as in C. Green arrows depicts direct-modulatory pathway of SLC9A1 by endothelin-1 (END1)/endothelin receptor A (ENDRA) and epidermal growth factor (EGF)/EGF receptor (EGFR). D. Overall survival of patients with ovarian cancer according to the indicated CNA alterations (by cBioPortal platform for Cancer Genomics). 88 C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91 estimate dpHi/dt. The NHE1 protein abundance was lower in KD-NHE1 HOSE and KD-NHE1 A2780 cells compared with cells expressing NHE1 (Fig. 4C). The KD-NHE1 HOSE and KD-NHE1 A2780 cells also show lower dpHi/dt compared with cells expressing NHE1 (Fig. 4D). 3.5. Cell proliferation and NHE1 and ki67 correlation Since an acidic pHo due to increased activity of NHE1 is associated with increased tumour cells proliferation [12,13], cell proliferation was assayed in the presence of a NHE1 inhibitor. The results show that zoniporide reduced proliferation (Fig. 5A) and the K/D t in all cell types (Table 2). The reduction in the K/D t was less pronounced in A2780 compared with HOSE or haOC cells. Cell growth rate was higher, but doubling times were lower in A2780 cells compared with HOSE or haOC cells. Zoniporide decreased growth rate and increased the doubling time only in haOC cells. Results in the kinetics of cell growth were complemented with an estimation of the changes in the actual cell number changes and nucleic acid turnover. The cell number and [ 3 H]thymidine incorporation were reduced by zoniporide (Fig. 5B). Additionally, the number of ki67 positive cells (marker for cell proliferation) in the presence of zoniporide was lower in A2780 and haOC, but unaltered in HOSE cells. To confirm a parallel increase in NHE1 and ki67 protein expression in human ovarian tissue, a potential correlation between these proteins expression was analysed. The results show that NHE1 and ki67 proteins were detected in the same regions in ovarian cancer tissues, with a positive correlation for NHE1 and ki67 protein abundances (Fig. 5C). Additionally, the cell number counted in KD-NHE1 HOSE and KD-NHE1 A2780 cells was lower compared with these cells expressing NHE1 (Fig. 5D). 3.6. SLC9A1 and overall survival In the aim of knowing whether NHE1 overexpression associates with adverse prognostic in patients with high-grade serous ovarian cancer, we analysed in silico the DNA copy-number alterations (CAN) for SLC9A1. Analysis of the CNA for this gene in human ovarian cancer reveals that 4% of cases will present with genetic changes including amplification (nine patients), deep deletions (one patients), and upregulation (six patients) or downregulation (two patients) of mRNA expression (Fig. 6A). SLC9A1 mRNA expression increases according to the type of CNA alteration in patients with high-serous ovarian cancer (Fig. 6B). The analysis in silico reveals that SLC9A1 is in direct-pathway with endothelin-1 (END1)/endothelin receptor A (ENDRA) and with epidermal growth factor (EGF)/EGF receptor (EGFR) in patients with ovarian cancer (Fig. 6C). SLC9A1 amplification associated with reduced overall survival (~50%) compared with patients without changes in SLC9A1 (diploid (98 patients) (41.4 versus 20.5 months in nine patients with SLC9A1 amplification, respectively) (Log-rank = 0.0005, Wilcoxon = 0.0013) or with mild CNA alteration (shallow deletion (108 patients) or gain (63 patients)) (Fig. 6D). There was not a difference between amplification and homologous deletions (5 patients, Log-rank = 0.36, Wilcoxon = 0.68), or homologous deletion and overall survival when compared with unaltered or mild CNA alteration. 4. Discussion This study shows that NHE1 is expressed in the human ovarian cancer and primary cultures of human ascites ovarian cancer cells (haOC). Tumour cells show intracellular alkalization compared with normal tissues [4,31,32]. This phenomenon depends on the expression of membrane transport systems for H + removal [8,33]. haOC cells show pHi values higher (~0.35 pHi units) than in non-tumour cells, suggesting that pHi regulation is altered by this pathology. This condition may be responsible for cell malignancy since an increase in 0.2-0.4 pHi units associates with higher cancer cell proliferation [4,21,33]. NHE1, 2, 3, and 4 are critical regulators of pHi in tumour tissues [8,34]. Our results show that NHE1 protein is expressed in CK7 positive tumour epithelium in the human ovary. Additionally, haOC cells show increased NHE1 expression (~12 fold) compared with non-tumour cells, suggesting its potential role in this type of cancer. haOC cells also show increased dpHi/dt and pHi that was comparable to that in non-tumour cells, suggesting that intracellular alkalization may result from a higher dpHi/dt.Indeed, increased v i of dpHi/dt in haOC agrees with other cancer cells [4,8,31, 32]. Since haOC and non-tumour cells intrinsic buffer capacity (ßi) were similar, increased dpHi/dt and J H + are not due to abnormal cell capacity to buffer a change in pHi. Interestingly, the ßi values determined in HOSE, haOC, and A2780 cells are in the same range as for other tumour cells such as T 84 cells (~31 (mmol/L)/pHi units) [22,35], and lower than in non-tumour MDCK cells (~110 (mmol/L)/pHi units) [21]. The differences in the βivalues could be due to specie differences (human versus dog) since the experimental conditions to determine this parameter were similar in these studies. HMA reduced the dpHi/dt and pHi, suggesting a role for NHEs in these phenomena. Since HMA-reduced pHi was higher (~1.5 fold) in haOC than non-tumour cells, haOC cells are more sensitive to NHEs activity-dependent pHi regulation. Same results were found in cells treated with zoniporide. Thus, haOC cells require NHE1 activity by alargerfraction( NHE1 f/ NHEs f~0.8), in this phenomenon. NHE1 f involvement to increase pHi is supported by the lower dpHi/dt detected in NHE1-knockdown ovarian tumour cells. The maximal proliferative capacity (K/D t ) of haOC was similar to non-tumour cells, suggesting that ovarian cancer does not alter ovarian cells proliferation. NHE1’s activity plays a preferential role in haOC cells proliferation since zoniporide-reduced K/D t was ~1.3 fold higher than non-tumour cells. Since a positive correlation between NHE1 and ki67 protein abundance was found (i.e., NHE1-expression dependent proliferation), and zoniporide reduces K/D t (i.e., NHE1-activity dependent proliferation), NHE1 expression and activity are required for haOC cells proliferation. The latter is supported by results showing zoniporide-reduced haOC cell number and DNA turnover. haOC cells show pHi modulation efficiency (1/ pHi E~4.74 pHi units/minute) lower than in non-tumour cells. Thus, haOC cells may count with a less efficient mechanism protecting for intracellular acidification than in non-tumour cells. Zoniporide-reduced K/D t in haOC cells may depend on a lower 1/ pHi Ein haOC (1/ pHi E~0.17 pHi units/minute in the presence of zoniporide) compared with non-tumour cells (1/ pHi E~0.11 pHi units/minute in the presence of zoniporide) [(K/D t )/1/ pHi E~0.31 in haOC and ~0.11 in HOSE cells]. This mechanism will favour tumour cells survival since intracellular acidification causes cell death [36–38]. Interestingly, a high pHi favours the transition from phase S to G2/M in the cell cycle and DNA/RNA synthesis [6,39]. Our results show that zoniporide reduced [ 3 H]thymidine incorporation. Thus, blocking NHE1 activity leads to acidic pHi-dependent reduced proliferation likely due to reduced cell cycle progression of haOC cells. These findings are supported by the reduced dpHi/dt detected in A2780 and non-tumour NHE knockdown cells. Thus, NHE1 expression and activity are pro-proliferative factors in human ovarian cells. Parallel analysis of CNA in SLC9A1 (for NHE1) revealed that a fraction of patients showed with CNA including higher amplification. Since SLC9A1 mRNA expression is increased depending on CNA alterations, this change in SLC9A1 could determine human ovarian cancer stage. CNA alterations may modulate the SLC9A1 expression by endothelin-1 and epidermal growth factor, complementing reports involving endothelin-1 and endothelin receptor A in this type of cancer [40] and endothelin-1 in combination with β-arrestin-1 activation increasing chemoresistance [41]. Equally, activation of epidermal growth factor/epidermal growth factor receptor signalling is involved in higher angiogenesis [42] and resistance to paclitaxel [43] in human ovarian cancer. Higher serum epidermal growth factor [44] and endothelin-1 in the peritoneal fluid in patients with malignant ascites [45,46] are reported in human ovarian cancer. Thus, epidermal growth factor and 89C. Sanhueza et al. / Biochimica et Biophysica Acta 1863 (2017) 81–91