Caloxin 1b3: A novel plasma membrane Ca2+-pump isoform 1 selective inhibitor that increases cytosolic Ca2+ in endothelial cells

Cell Calcium 48 (2010) 352–357

Contents lists available at ScienceDirect

Cell Calcium journal homepage: www.elsevier.com/locate/ceca

Caloxin 1b3: A novel plasma membrane Ca2+ -pump isoform 1 selective inhibitor that increases cytosolic Ca2+ in endothelial cells Magdalena M. Szewczyk b , Jyoti Pande a , Gauri Akolkar b , Ashok K. Grover a,b,∗ a b

Department of Medicine, HSC 4N41 McMaster University, 1200 Main Street West Hamilton, Ontario, Canada L8N 3Z5 Department of Biology, HSC 4N41 McMaster University, 1200 Main Street West Hamilton, Ontario, Canada L8N 3Z5

a r t i c l e

i n f o

Article history: Received 31 August 2010 Received in revised form 19 October 2010 Accepted 21 October 2010 Available online 18 November 2010 Keywords: PMCA PMCA1 Coronary artery Vascular Smooth muscle Endothelium Phage display

a b s t r a c t The purpose of this study was to invent an extracellular inhibitor selective for the plasma membrane Ca2+ pump(s) (PMCA) isoform 1. PMCA extrude Ca2+ from cells during signalling and homeostasis. PMCA isoforms are encoded by 4 genes (PMCA1–4). Pig coronary artery endothelium and smooth muscle express the genes PMCA1 and 4. We showed that the endothelial cells contained mostly PMCA1 protein while smooth muscle cells had mostly PMCA4. A random peptide phage display library was screened for binding to synthetic extracellular domain 1 of PMCA1. The selected phage population was screened further by affinity chromatography using PMCA from rabbit duodenal mucosa which expressed mostly PMCA1. The peptide displayed by the selected phage was termed caloxin 1b3. Caloxin 1b3 inhibited PMCA Ca2+ –Mg2+ ATPase in the rabbit duodenal mucosa (PMCA1) with a greater affinity (inhibition constant = 17 ± 2 ␮M) than the PMCA in the human erythrocyte ghosts (PMCA4, inhibition constant = 45 ± 4 ␮M). The affinity of caloxin 1b3 was also higher for PMCA1 than for PMCA2 and 3 indicating its selectivity for PMCA1. Consistent with an inhibition of PMCA1, caloxin 1b3 addition to the medium increased cytosolic Ca2+ concentration in endothelial cells. Caloxin 1b3 is the first known PMCA1 selective inhibitor. We anticipate caloxin 1b3 to aid in understanding PMCA physiology in endothelium and other tissues. © 2010 Elsevier Ltd. All rights reserved.

1. Introduction The plasma membrane Ca2+ -pump(s) (PMCA) is the only high affinity Ca2+ extrusion system found in mammalian cells [1–3]. They are central to cell homeostasis and signal transduction. Defects in PMCA are associated with several cardiovascular and other disorders [1,3–7]. Mammalian cells also contain other systems that can lower cytosolic Ca2+ concentration ([Ca2+ ]i) [2,8–10]. The sarco/endoplasmic reticulum Ca2+ pumps can sequester Ca2+ into the lumen of the reticulum. Ca2+ can also be extruded from the cell by the Na+ –Ca2+ -exchanger but with a low affinity. The three Ca2+ lowering systems are regulated by diverse mechanisms. Their

Abbreviations: [Ca2+ ]i, cytosolic [Ca2+ ]; EDTA, ethylenediamine tetra-acetic acid; EGTA, ethylene glycol-bis(-aminoethyl ether)-N,N,N ,N -tetraacetic acid; exdom, extracellular domain; HEPES, N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid; Ki , inhibition constant; PMCA, plasma membrane Ca2+ pump(s); caloxin, a substance that binds the PMCA on its external surface and modulates its activity; PMCA isoforms, proteins encoded by different PMCA genes (the term does not refer to splice variants, the splicing does not affect the domains used in this study). ∗ Corresponding author at: Department of Medicine, HSC 4N41, McMaster University, 1200 Main Street West, Hamilton, Ontario, Canada L8N 3Z5. Tel.: +1 905 525 9140x22238; fax: +1 905 522 3114. E-mail address: [email protected] (A.K. Grover). 0143-4160/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ceca.2010.10.008

levels of expression and modes of regulation vary with cell types, development, functions and pathologies. PMCA are encoded by 4 genes, PMCA1–4. Alternative splicing of their primary transcripts yields a large variety of PMCA proteins which differ in their kinetics and regulation [3,11]. PMCA genes 1 and 4 are found in most tissues while genes 2 and 3 show tissuespecific expression. The unique distribution of PMCA isoforms in different regions of the brain and in different types of retinal neurons demonstrates cell-specific expression of PMCA isoforms that may reflect differences in individual Ca2+ handling requirements [12]. Even within the same cell, individual PMCA isoforms can be localized to different parts of the plasma membrane and be responsible for spatially unique functions, e.g. in neuronal cells, PMCA4 was localized exclusively in lipid rafts while PMCA1–3 were in other parts of the plasma membrane [13]. To delineate the specific physiological role of a PMCA isoform, it is important to determine its cellular and sub-cellular distribution, regulation and interacting protein partners. One of the methods currently used to study PMCA physiology involves the genetic and molecular modulation of various PMCA isoforms [10,14]. However, since PMCA is a low abundance protein, it is difficult to achieve high levels of overexpression with correct targeting to plasma membrane [15]. There also exists tight regulation of [Ca2+ ]i which is reflected in the plasticity and the adaptability of

M.M. Szewczyk et al. / Cell Calcium 48 (2010) 352–357

various Ca2+ transporters and sensors in the cell. For example, the overexpression of PMCA1 in rat aortic endothelial cells altered the expression of other Ca2+ regulating components like sarco/endoplasmic reticulum Ca2+ pump and inositol 1,4,5trisphosphate receptor [16]. Similarly, changes in the expression of these proteins as well as endogenous PMCA were also observed in the transgenic mice overexpressing PMCA4b under the control of arterial smooth muscle actin promoter [14]. Contrary to initial expectations, the transgenic mice showed enhanced arterial contractility. The ability of knockout mice to adapt by using alternative Ca2+ regulating pathways is shown by the phenotype observed in PMCA4 null mice. Despite the ubiquitous tissue distribution of PMCA4, the null mice show mainly a loss in sperm hypermotility as the major defect along with vascular impairment observed in some strains of PMCA4 null mice [7,10]. An alternative is to develop pharmacological agents which can allow local and immediate modulation of PMCA, thus avoiding the redundancy and compensatory effects associated with genetic manipulations. Therefore, we initiated the development of caloxins as specific extracellular inhibitors of PMCA [17,18]. We observed differences in the abundance of PMCA isoforms in the cells of pig coronary artery which is used as a biological model in our lab [19,20]. The endothelial cells are abundant in PMCA1 as compared to PMCA4, while the smooth muscle cells express higher levels of PMCA4 than PMCA1. Therefore, we have focussed on developing PMCA4 and PMCA1 selective caloxins to understand the PMCA physiology in the pig coronary arteries. The importance of targeting allosteric sites in proteins to obtain selective inhibitors is well established [21–23]. Allosteric sites provide unique targets as opposed to active and regulatory sites which may be conserved in a family of proteins with a similar function. For example, the currently used inhibitors of PMCA, vanadate and eosin, also inhibit other members of the ATPase family like the sarco/endoplasmic reticulum calcium pump and the sodium pump. Vanadate acts as a phosphate analog, while eosin interferes with binding of ATP to a site which is conserved in various ATPases [24–27]. In contrast, thapsigargin is a specific inhibitor of the sarco/endoplasmic reticulum Ca2+ pump which binds to an allosteric site on the luminal loops of the pump [28,29]. Similarly, ouabain is allosteric inhibitors of the Na+ -pump which binds to one of its extracellular domain (exdoms) [30,31]. We used exdoms of PMCA as allosteric sites to search for caloxins [17,18]. Exdoms are short loops connecting the transmembrane domains of PMCA on the extracellular surface. We have targeted different exdoms of PMCA to invent caloxins which are PMCA selective inhibitors. We are now exploiting exdom 1 of PMCA to develop isoform-selective caloxins. The amino acid sequence of exdom 1 differs significantly between the products of PMCA genes 1 and 4 (Swiss protein accession numbers P20020, Q01814, Q16720, P23634). Also, exdom 1 is conserved in all the splice variants of a PMCA gene and therefore exdom 1 based isoform-selective caloxin is expected to inhibit all the splice variants. Earlier, we reported PMCA4-selective caloxins based on exdom 1 of PMCA4, which showed higher affinity for PMCA in pig coronary artery smooth muscle [19,32]. However, PMCA1-selective caloxins are needed to study the role of PMCA in endothelium and its contribution to arterial contractility. Here, we report the invention of PMCA1-selective caloxin1b3. To our knowledge, this is the first report on any peptide or non-peptide inhibitor which has higher affinity for PMCA1 than for PMCA4.

2. Experimental procedures 2.1. Membrane isolation New Zealand white rabbits were purchased from Charles River Labs (St. Constant, QC), euthanized and dissected to remove the

353

duodenum. The duodenum was placed in chilled physiological saline solution containing the following in mM: 138 NaCl, 2 CaCl2 , 10 Glucose, 10 N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid (HEPES), 5 KCl, 1 MgCl2 , pH 6.4. The duodenal mucosa was homogenized for 3 × 15 s using a Polytron PT20 at a setting of 4.5 in a homogenization buffer containing 8% sucrose and the following in mM: 100 KCl, 1 phenylmethanesulfonylfluoride, 2 dithiothreitol, 0.5 ethylene glycol-bis(-aminoethyl ether)-N,N,N ,N -tetraacetic acid (EGTA), 0.002 calpain inhibitor MDL 28170. The homogenate was centrifuged at 10,000 × g for 10 min and the pellet was discarded. The supernatant was filtered through a cheese cloth and centrifuged at 388,000 × g for 30 min to obtain a microsomal pellet. This pellet was suspended in 8% sucrose and layered on 32% sucrose and centrifuged at 278,000 × g in a swinging bucket rotor for 2 h. The layer of turbid membranes at the interphase between 8% and 32% sucrose was collected and stored at −80 ◦ C until use. Leaky human erythrocyte ghosts were prepared as described previously [17]. To obtain vascular smooth muscle membranes, pig coronary artery was used for preparing microsomes as described previously [33]. Additionally, 1 mM EGTA was included in the homogenization buffer and in the buffer for suspending the microsomes to ensure that calmodulin was removed. The plasma membrane enriched fraction F2 was then isolated from the microsomes on a sucrose density gradient as described previously and aliquots were stored at −80 ◦ C [33]. The plasma membrane enriched fraction was also isolated from endothelial cells cultured from pig coronary artery. The culture procedures and phenotypic characterization of these cells has been described previously [33]. Microsomes from SF9 insect cells infected with bacculoviral vectors expressing PMCA2 and PMCA3 were prepared by a previously published method [34]. 2.2. Screening phage for binding to a synthetic sequence A peptide corresponding to the N-terminal half of the human PMCA1 exdom 1 (exdom 1X) consisting of residues 121-137 (SLGLSFYQPPEGDNALC, Protein bank P20020) was synthesized commercially (Dalton Chemical Labs., Toronto, Canada). The PMCA1 sequence is identical in rabbit, man, rat, mouse, pig and cow. The residue C was used to conjugate the peptide to keyhole limpet haemocyanin or ovalbumin (Biosynthesis Inc., USA). The M13 phage display library expressing random linear 12 amino acid peptides (Ph. D.12, New England Biolabs, USA) was panned using the exdom 1X target as described previously [17]. The eluted phage particles were amplified and used in the next panning cycle. The process was repeated two times. The selected phage pool was tested for its selectivity for binding to PMCA1 exdom 1X over PMCA4 exdom 1X. The protocol was the same as for panning but the wells were coated with keyhole limpet haemocyanin conjugates of either PMCA1 or PMCA4 exdom 1X and eluted with the corresponding ovalbumin conjugates. 2.3. Screening phage by PMCA1 affinity chromatography The duodenal mucosal plasma membranes (4 mg protein) were centrifuged at 500,000 × g for 15 min. The resulting pellet was resuspended to obtain a protein concentration of 8 mg/ml in a solubilization buffer containing the following in mM: 130 NaCl, 20 HEPES, 0.5 MgCl2 , 2 dithiothreitol, and 0.1 CaCl2 at pH 7.4 plus a cocktail of protease inhibitors (Complete Mini, EDTA-free, Roche, USA). To this suspension, an equal volume of a solubilization buffer containing 0.8% Triton X-100 was added slowly and the contents of the tube were mixed by inverting the tube 5–10 times over a period of 10 min. The suspension was centrifuged at 500,000 × g for 15 min and the supernatant was retained as the soluble fraction to be used in the subsequent steps. A bed volume of 500 ␮l agarose–calmodulin resin (Sigma–Aldrich,

354

M.M. Szewczyk et al. / Cell Calcium 48 (2010) 352–357

USA) was packed in a column (BioRad, USA) and washed in the wash buffer (0.4% Triton X-100, in mM 130 NaCl, 20 HEPES, 1 MgCl2 , 2 dithiothreitol, 0.1 CaCl2 , and 0.05% phosphatidyl choline). The soluble fraction (1 ml) was mixed with 26 ␮l of the stock phospholipid solution (2% phosphatidyl choline dissolved in 0.1% Triton X-100) by rocking for 10 min at 4 ◦ C, and mixed with the agarose–calmodulin slurry for 2 h. The unbound flow-through material from the column was discarded. The column was washed with 10 ml of the wash buffer. A similar column with the protein solubilized from erythrocyte ghosts was prepared the same way for negative screening. The phage was diluted in the wash buffer, applied to the erythrocyte ghost column to decrease the phage that would bind non-specifically or to PMCA4. The unbound phage suspension was collected and incubated with a column containing agarose–calmodulin–rabbit duodenal mucosal PMCA overnight at 4 ◦ C. The unbound phage particles were removed as flow through and by washing the column with 10 × 2.5 ml of the wash buffer. PMCA and the bound phage were eluted twice at 37 ◦ C, each time using 500 ␮l of Ca2+ -free elution buffer containing 0.4% Triton X100 and 0.05% phosphatidyl choline and the following in mM: 130 NaCl, 20 HEPES, 1 MgCl2 , 2 dithiothreitol and 5 EGTA. The eluted phage were pooled, precipitated and amplified. The affinity chromatography protocol was repeated two times. Phage titres were performed using Escherichia coli XL-2 blue cells (Strategene, Oakville). Individual clones of the eluted phage were amplified and their plasmid DNA were isolated. The plasmids were sequenced at the MOBIX facility at McMaster University using a reverse primer which was 96 bp downstream of the random library site. Next, a mixture containing equal titre of each unique phage clone was used for a competition experiment. The mixture was used for affinity chromatography as above with the difference that the column was washed extensively (50 ml/day for 4 days at 4 ◦ C) to remove the unbound phage. The phage were then eluted and the plasmid DNA of the eluted phage clones were sequenced. 2.4. Biochemical assays The analysis of PMCA isoforms present in membrane preparations from various tissues was conducted by Western blots. The following primary antibodies were used: 5F10 (recognizes all PMCA isoforms 1:1000, Affinity BioReagents), JA9 (anti-PMCA4, 1:1000, Sigma, St. Louis, MO, USA) and anti-PMCA1 (1:1000). The anti-PMCA1 polyclonal antibody was raised in rabbits against a PMCA1 specific sequence (8–23, SVAYGGVKNSLKEANH) (Protein bank P20020) (Pacific Immunology, Ramona, CA). The residue C was added at the N-terminus and used for linking to a carrier. The blots were washed and then incubated with the horseradish peroxidase conjugated anti-mouse IgG (1:20,000) or anti-rabbit IgG (1: 15,000, both from Sigma, St. Louis, MO, USA) antibodies. The peroxidase activity was visualized with a femto-kit (Pierce Chemical Company, Rockford) and a LAS3000 mini Luminiscent Image Analyzer (Fujifilm Life Science, Stamford, CT). Ca2+ –Mg2+ -ATPase and Mg2+ -ATPase assays were performed at ◦ 37 C by determining the hydrolysis of ATP in a coupled enzyme assay that monitored the rate of disappearance of absorbance of NADH as previously described but in 96-well UV-transparent microtiter plates (Corning, NY) using a TECAN Safire Infinite M1000 microplate reader [17]. The membranes were incubated with or without caloxin for 30 min at 0 ◦ C. The basal Mg2+ -ATPase activity was measured in an 140 ␮l assay solution containing 1 ␮g protein, excess pyruvate kinase-lactate dehydrogenase mix, 4 ng/ml calmodulin and 0.015% Triton X-100 and the following in mM: 0.2 ouabain, 1 sodium azide, 0.005 thapsigargin, 100 NaCl, 20 KCl, 6 MgCl2 , 30 imidazole-HCl (pH 7.0), 0.5 EDTA, 0.5 EGTA, 0.5 ATP, 0.2 NADH, 1 phosphoenol pyruvate. After 30 min, 10 ␮l CaCl2 was added to attain a final concentration of 0.55 mM and the disappear-

ance of NADH was monitored for another 30 min. The difference in the ATPase activity in saturating (6 mM) Mg2+ with and without Ca2+ was defined as the Ca2+ –Mg2+ -ATPase activity of PMCA. 2.5. Cytosolic Ca2+ determination in endothelial cells Endothelial cells cultured from pig coronary artery after passage 4 were seeded into 96-well microtitre plates and cultured until subconfluent. The cells were washed three times in a loading buffer at pH 7.4 containing the following in mM: 115 NaCl, 25 HEPES, 12 glucose, 5.8 KCl, 2.2 KH2 PO4 , 1 CaCl2 , 0.6 MgCl2 , 2 probenecid. They were then incubated in the loading buffer along with the Ca2+ fluorescence dye Fluo 4/AM (4 ␮M) and pluronic acid (0.02%) for 45 min at 22–24 ◦ C in the dark, washed two times in the loading buffer to remove the dye and incubated in it for another 25 min. The plates were washed with this buffer again and placed in TECAN Safire microplate reader at 37 ◦ C for 5 min. Background fluorescence (excitation 485 nm, emission 525 nm) was recorded for 2 min and then caloxin 1b3 was added. Caloxin 1b3 was dissolved in ethanol and loading buffer such that the final ethanol concentration would be 0.004%. Controls containing only ethanol were also tested. After another 20 min, calibration was carried out to determine Fmax and Fmin . Fmax was determined in the presence of 6 ␮M of the Ca2+ ionophore 4-bromo A23187 in the loading buffer and Fmin by addition of 2 mM EGTA to the above. Fluorescence values were expressed as percent of maximum (100 × (F − Fmin )/(Fmax − Fmin )). The values before the addition of caloxins were subtracted to determine the change in fluorescence with time. 2.6. Data analysis To compute values of the inhibition constant (Ki ), the data were analyzed for non-competitive inhibition, according to the equation: percent inhibition = 100 × [inhibitor]/(Ki + [inhibitor]) by nonlinear regression. Curve fitting was carried out using FigP software (Ancaster, Canada). Statistical significance was determined using Student’s t-test or ANOVA using the software GraphPad InStat (San Diego, CA) and values of p < 0.05 were considered to be significant. 3. Results 3.1. Distribution of PMCA1 and PMCA4 There were two main reasons for examining the distribution of PMCA isoforms 1 and 4 in the plasma membranes from pig coronary artery smooth muscle and endothelial cells, human erythrocytes and rabbit duodenal mucosa. One reason was to confirm the choice of rabbit duodenal mucosa as a rich and selective source of PMCA1 for phage screening by affinity chromatography and the second was to determine the relative abundance of PMCA isoforms 1 and 4 in the coronary artery tissues. The PMCA2 and PMCA3 expression was not analysed since rabbit duodenal mucosa, coronary artery smooth muscle or endothelium do not contain mRNA for these isoforms [19,20]. The human erythrocyte ghosts are known to express mainly PMCA4 [11]. Fig. 1 shows Western blots with 0.5, 1 and 2 ␮g plasma membrane protein of each tissue analysed with the following antibodies: an antibody that recognizes all PMCA isoforms, anti-PMCA1 or anti-PMCA4. The plasma membranes from rabbit duodenal mucosa were determined to be rich in PMCA1 but also contained very low levels of PMCA4. In contrast, human erythrocyte ghosts expressed mainly PMCA4 and very little PMCA1. Quantification of Western blot (Fig. 1) by densitometric analysis showed that the total PMCA protein in the plasma membrane of smooth muscle cells was 2.9 times that of endothelial cells. Based on the relative reactivity of each of the three antibodies in Fig. 1,

M.M. Szewczyk et al. / Cell Calcium 48 (2010) 352–357

355

Fig. 1. Western blots showing PMCA isoform distribution. The amount of plasma membrane proteins loaded, primary antibodies used and the tissues are labelled. All the lanes from different tissues for each antibody were compared together in one gel. Replicating the experiment gave similar results.

it was computed that the ratio of PMCA1 to PMCA4 was approximately 0.1 in the smooth muscle plasma membrane and 3.5 in the endothelial plasma membrane. Thus, in the pig coronary artery, the endothelial cells express mainly PMCA1 whereas the smooth muscle cells contain mostly PMCA4. 3.2. Screening phage library for binding to exdom 1 of PMCA1 Screening of phage library to select PMCA1-selective caloxin was conducted in two main steps: the first step used panning with synthetic exdom 1X peptide as a target and the second step used affinity chromatography with PMCA protein purified from duodenal mucosa as a target. Following three rounds of panning in the first step of screening, the resulting phage population contained a significantly larger number of phage particles that bound preferentially to the synthetic exdom 1X of PMCA1 as compared to synthetic exdom 1X of PMCA4 (Fig. 2). The non-selected phage library used as a negative control did not show this selectivity. Thus, the phage population was enriched in clones that bound to the synthetic exdom 1X of PMCA1 after the first step of screening. The phage population selected above was then screened by PMCA affinity chromatography as described in Section 2. Following the last round of screening, plaques were picked and their plasmids were sequenced. Of a total of 272 phage clones that were sequenced, there were 101 unique sequences. To rule out selection of phage due to copy-number bias that can arise during library construction, panning and phage amplification, we introduced a step of phage competition. In phage competition, a mixture of equal num-

Fig. 2. Selectivity of phage population obtained upon panning. The data are mean ± SEM of a total of 6 replicates from two experiments. Each experiment was conducted in triplicate. The number of phage particles applied/well was 9.6 × 105 in the first experiment and 2.6 × 106 in the second. The counts for the eluted phage in each experiment were converted to percent applied and pooled. A multiway ANOVA test showed that the group PMCA1-selected phage differed significantly from all the other groups (p < 0.05). The three remaining groups did not differ significantly from each other (p > 0.05).

Fig. 3. Caloxin 1b3 concentration dependence of inhibition of PMCA1 and 4. The values are mean ± SEM of a total of 6–12 replicates from experiments on two separate days. The values at each of the caloxin 1b3 concentrations were significantly greater for PMCA1 than for PMCA4 (p < 0.05).

bers of phage particles of each of the selected clones was allowed to compete for binding to PMCA during affinity chromatography. Following competition, 184 plaques were picked and their plasmid DNA was sequenced. The dominant phage clone had 94 copies and encoded for a variable peptide sequence TIPKWISIIQALR. The variable sequence along with a spacer that connects it to the coat protein in phage was chemically synthesized and designated as caloxin 1b3 (TIPKWISIIQALRGGGSK-amide). 3.3. Inhibition of PMCA1 and PMCA4 Ca2+ –Mg2+ -ATPases by caloxin 1b3 As shown in Fig. 1, PMCA1 is the predominant isoform present in the plasma membrane of duodenal mucosa and PMCA4 is the major isoform in erythrocyte ghosts. Attempts were then made to monitor PMCA activity in the endothelial cell membranes that also have PMCA1 as the dominant isoform. However, due to higher basal Mg2+ -ATPase activity, it was difficult to monitor Ca2+ –Mg2+ ATPase activity of PMCA in endothelial membranes. Therefore, only the mucosal plasma membrane (PMCA1) and the erythrocyte ghosts (PMCA4) were used to study the effect of caloxin 1b3 on PMCA isoforms 1 and 4. Fig. 3 shows the caloxin 1b3 concentration dependence of the inhibition of PMCA1 and PMCA4. Even though it inhibited both PMCA1 and PMCA4, the inhibition was greater for PMCA1 than for PMCA4 at 15, 25, 50 or 100 ␮M of this peptide. Caloxin 1b3 was PMCA selective in that it inhibited the Ca2+ –Mg2+ -ATPase in both the membranes, but it had no effect on the Mg2+ -ATPase activity at concentrations up to 100 ␮M. Based on the data in Fig. 3 and similar experiments on other days, the inhibition constants of caloxin 1b3 for PMCA1 and PMCA4 were 17 ± 2, and 45 ± 4 ␮M, respectively. The effect of 25 ␮M caloxin 1b3 on PMCA1, 2, 3 and 4 is compared in Fig. 4. The inhibition was greater for PMCA1 than for any of the other isoforms (Fig. 4). The inhibition was also examined using 15 ␮M caloxin 1b3 and similar results were observed (data not shown). Thus, caloxin 1b3 had a higher affinity for PMCA1 than for PMCA2, 3 or 4. 3.4. Effects of caloxin 1b3 on cytosolic Ca2+ in endothelium As caloxin 1b3 showed some PMCA1 isoform specificity, we determined its effects on [Ca2+ ]i in the pig coronary artery endothelial cells (predominantly PMCA1) (Fig. 5). As caloxin 1b3 was designed to work extracellularly, we added 20 and 50 ␮M of this

356

M.M. Szewczyk et al. / Cell Calcium 48 (2010) 352–357

Fig. 4. Effect of caloxin 1b3 on all PMCA isoforms. Caloxin 1b3 concentration = 25 ␮M. The values are mean ± SEM of a total of 6–12 replicates from experiments on two separate days. A multiway ANOVA test showed that all the other groups differed from each other (p < 0.05) except groups PMCA2 and PMCA4.

Fig. 5. Effect of caloxins 1b3 and 1b1 on [Ca2+ ]i in endothelial cells. F = change in fluorescence. Arrow indicates the time of addition of caloxin. All the values obtained 20 s after adding the caloxins differed significantly from the baseline values (p < 0.05). The values obtained with 20 and 50 ␮M 1b3 were significantly larger than the corresponding values obtained with caloxin 1b1 (p < 0.05).

peptide to the medium containing the cells. There was an increase in [Ca2+ ]i at both the concentrations with the increase being much greater at 50 than at 20 ␮M. In other experiments, the effects of 0, 3, 10, 50 and 100 ␮M caloxin 1b3 were also examined (data not shown). Whereas 0, 3, 10 ␮M had no effect, the effects of 50 and 100 ␮M were similar. For comparison, Fig. 5 also shows the effects of 20 and 50 ␮M caloxin 1b1 which has preference for PMCA4 but also inhibits PMCA1 with a relatively lower affinity. Caloxin 1b1 had a much smaller effect on [Ca2+ ]i in endothelial cells as compared to caloxin 1b3.

mainly PMCA1 and addition of caloxin 1b3 to medium caused an increase in [Ca2+ ]i in these cells. The discussion focuses on the invention of caloxin 1b3, its PMCA isoform preference, and its potential in understanding the role of PMCA in endothelium and other tissues. We observed that the plasma membrane of rabbit duodenal mucosa expresses mainly PMCA1, while human erythrocyte ghosts express mainly PMCA4 (Fig. 1). These two membranes could, therefore, be used as relatively rich sources of PMCA isoforms 1 and 4, respectively, for phage screening by affinity chromatography. As shown in Fig. 1, the pig coronary artery smooth muscle cells contain more PMCA4, whereas the endothelial cells are rich in PMCA1. These results are consistent with the previous observations on the relative abundance of the respective mRNA in the two tissues [19,20]. Smooth muscle showed multiple bands with the antibodies 5F10 (all PMCA) and JA9 (PMCA4). This result also confirms the previous observation that smooth muscle expresses multiple splice variants of the PMCA4 gene. The method of phage display screening to obtain PMCA1selective caloxin 1b3 was very similar to that developed for PMCA4-selective caloxin 1b1 [19]. The main difference lies in the nature of the target used: synthetic exdom 1X of PMCA1 (instead of PMCA4) for panning and PMCA protein purified from plasma membrane of rabbit duodenal mucosa (instead of erythrocyte ghosts) for affinity chromatography. Thus, this study shows that the methods developed for inventing PMCA4-selective caloxin can be similarly used to obtain caloxins selective for other isoforms of PMCA. Upon comparing the inhibitory effect of caloxin 1b3 on the Ca2+ –Mg2+ ATPase activity of various PMCA isoforms, it is observed that caloxin 1b3 had a higher affinity for PMCA1 as compared to PMCA2, 3 or 4. Thus, caloxin 1b3 is the first ever reported inhibitor which is selective for PMCA1 over all the other PMCA isoforms. This isoform selectivity of caloxin 1b3 can be improved in the future by limited mutagenesis and other methods described in a recent review [23]. Caloxins have been used by various labs to understand PMCA physiology. Human bone marrow-derived mesenchymal stem cells show spontaneous [Ca2+ ]i oscillations [35]. An application of 2 mM caloxin 2a1 produced a [Ca2+ ]i transient followed by complete block of the oscillations with the return of [Ca2+ ]i to basal level. In contrast, eosin caused a large sustained increase in basal [Ca2+ ]i before completely stopping the oscillations in response similar to that observed with removal of extracellular Na+ to inhibit sodium calcium exchanger. This showed that caloxin 2a1 (a PMCA-selective inhibitor) gave different results than eosin (inhibitor of PMCA and Na+ –K+ -ATPase). The genetic diversity that exists in PMCA emphasizes the need for isoform-selective caloxins to understand PMCA physiology and pathophysiology [3,5]. Changes in the levels of expression or activity of various PMCA isoforms have been associated with several pathologies. Differences are also observed in the distribution of PMCA isoforms at tissue, cell and subcellular levels. Higher affinity and isoform-selective caloxins will allow better understanding of the functional significance of differential distribution of PMCA isoforms and the interactions between different Ca2+ transport mechanisms in a given cell. The main impetus for the invention of a PMCA1-selective caloxin was to study endothelial function in the coronary artery. However, PMCA1 selective caloxins will find wider applicability. For instance, PMCA1 is the main isoform in the rabbit intestinal mucosa and hence the PMCA1-selective caloxins may also prove useful in understanding gastrointestinal functions.

4. Discussion Caloxin 1b3 was selected for binding to an extracellular epitope of PMCA1. The results show that caloxin 1b3 inhibited PMCA1 with higher affinity as compared to PMCA4. Endothelial cells express

Conflict of interest There is no conflict of interest.

M.M. Szewczyk et al. / Cell Calcium 48 (2010) 352–357

Acknowledgements The microsomes expressing PMCA2 and 3 were a gift from Drs. A. Filoteo and E.E. Strehler of Mayo Clinic College of Medicine (Rochester, Minnesota). This work was supported by a grant-in-aid from the Heart & Stroke Foundation of Ontario (T 6355) and a Doctoral Fellowship from the Heart and Stroke Foundation of Canada to MMS. Observations in this manuscript form the basis of a patent application in progress. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ceca.2010.10.008. References [1] E. Carafoli, Calcium signaling: a tale for all seasons, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 1115–1122. [2] D.E. Clapham, Calcium signaling, Cell 131 (2007) 1047–1058. [3] E.E. Strehler, A.J. Caride, A.G. Filoteo, Y. Xiong, J.T. Penniston, A. Enyedi, Plasma membrane Ca2+ ATPases as dynamic regulators of cellular calcium handling, Ann. N. Y. Acad. Sci. 1099 (2007) 226–236. [4] D. Guerini, The significance of the isoforms of plasma membrane calcium ATPase, Cell Tissue Res. 292 (1998) 191–197. [5] J. Lehotsky, P. Kaplan, R. Murin, L. Raeymaekers, The role of plasma membrane Ca2+ pumps (PMCAs) in pathologies of mammalian cells, Front. Biosci. 7 (2002) d53–d84. [6] D. Oceandy, M.H. Buch, E.J. Cartwright, L. Neyses, The emergence of plasma membrane calcium pump as a novel therapeutic target for heart disease, Mini. Rev. Med. Chem. 6 (2006) 583–588. [7] G.W. Okunade, M.L. Miller, G.J. Pyne, R.L. Sutliff, K.T. O’Connor, J.C. Neumann, A. Andringa, D.A. Miller, V. Prasad, T. Doetschman, R.J. Paul, G.E. Shull, Targeted ablation of plasma membrane Ca2+ -ATPase (PMCA) 1 and 4 indicates a major housekeeping function for PMCA1 and a critical role in hyperactivated sperm motility and male fertility for PMCA4, J. Biol. Chem. 279 (2004) 33742–33750. [8] C.M. Misquitta, D.P. Mack, A.K. Grover, Sarco/endoplasmic reticulum Ca2+ (SERCA)-pumps: link to heart beats and calcium waves, Cell Calcium 25 (1999) 277–290. [9] K.D. Philipson, D.A. Nicoll, M. Ottolia, B.D. Quednau, H. Reuter, S. John, Z. Qiu, The Na+ /Ca2+ exchange molecule: an overview, Ann. N. Y. Acad. Sci. 976 (2002) 1–10. [10] G.E. Shull, G. Okunade, L.H. Liu, P. Kozel, M. Periasamy, J.N. Lorenz, V. Prasad, Physiological functions of plasma membrane and intracellular Ca2+ pumps revealed by analysis of null mutants, Ann. N. Y. Acad. Sci. 986 (2003) 453–460. [11] T.P. Stauffer, D. Guerini, E. Carafoli, Tissue distribution of the four gene products of the plasma membrane Ca2+ pump. A study using specific antibodies, J. Biol. Chem. 270 (1995) 12184–12190. [12] D. Krizaj, S.J. Demarco, J. Johnson, E.E. Strehler, D.R. Copenhagen, Cell-specific expression of plasma membrane calcium ATPase isoforms in retinal neurons, J. Comp. Neurol. 451 (2002) 1–21. [13] M.R. Sepulveda, M. Berrocal-Carrillo, M. Gasset, A.M. Mata, The plasma membrane Ca2+ -ATPase isoform 4 is localized in lipid rafts of cerebellum synaptic plasma membranes, J. Biol. Chem. 281 (2006) 447–453. [14] K. Schuh, T. Quaschning, S. Knauer, K. Hu, S. Kocak, N. Roethlein, L. Neyses, Regulation of vascular tone in animals overexpressing the sarcolemmal calcium pump, J. Biol. Chem. 278 (2003) 41246–41252.

357

[15] H.P. Adamo, A.K. Verma, M.A. Sanders, R. Heim, J.L. Salisbury, E.D. Wieben, J.T. Penniston, Overexpression of the erythrocyte plasma membrane Ca2+ pump in COS-1 cells, Biochem. J. 285 (Pt 3) (1992) 791–797. [16] B.F. Liu, X. Xu, R. Fridman, S. Muallem, T.H. Kuo, Consequences of functional expression of the plasma membrane Ca2+ pump isoform 1a, J. Biol. Chem. 271 (1996) 5536–5544. [17] J. Chaudhary, M. Walia, J. Matharu, E. Escher, A.K. Grover, Caloxin: a novel plasma membrane Ca2+ pump inhibitor, Am. J. Physiol. Cell Physiol. 280 (2001) C1027–C1030. [18] M.M. Szewczyk, J. Pande, A.K. Grover, Caloxins: a novel class of selective plasma membrane Ca(2+) pump inhibitors obtained using biotechnology, Pflugers Arch. 456 (2008) 255–266. [19] J. Pande, K.K. Mallhi, A. Sawh, M.M. Szewczyk, F. Simpson, A.K. Grover, Aortic smooth muscle and endothelial plasma membrane Ca2+ pump isoforms are inhibited differently by the extracellular inhibitor caloxin 1b1, Am. J. Physiol. Cell Physiol. 290 (2006) C1341–C1349. [20] M.M. Szewczyk, K.A. Davis, S.E. Samson, F. Simpson, P.K. Rangachari, A.K. Grover, Ca2+ -pumps and Na2+ –Ca2+ -exchangers in coronary artery endothelium versus smooth muscle, J. Cell Mol. Med. 11 (2007) 129–138. [21] J.P. Changeux, Allosteric receptors: from electric organ to cognition, Annu. Rev. Pharmacol. Toxicol. 50 (2010) 1–38. [22] T. Kenakin, L.J. Miller, Seven transmembrane receptors as shapeshifting proteins: the impact of allosteric modulation and functional selectivity on new drug discovery, Pharmacol. Rev. 62 (2010) 265–304. [23] J. Pande, M.M. Szewczyk, A.K. Grover, Phage display: concept, innovations, applications and future, Biotechnol. Adv. 28 (2010) 849–858. [24] L.C. Cantley Jr., L Josephson, R. Warner, M. Yanagisawa, C. Lechene, G. Guidotti, Vanadate is a potent (Na,K)-ATPase inhibitor found in ATP derived from muscle, J. Biol. Chem. 252 (1977) 7421–7423. [25] N.B. Klaven, H.A. Pershadsingh, G.V. Henius, P.C. Laris, J.W. Long Jr., J.M. McDonald, A high-affinity, calmodulin-sensitive (Ca2+ + Mg2+ )-ATPase and associated calcium-transport pump in the Ehrlich ascites tumor cell plasma membrane, Arch. Biochem. Biophys. 226 (1983) 618–628. [26] J.C Skou, M. Esmann, Eosin, a fluorescent probe of ATP binding to the (Na+ + K+ )ATPase, Biochim. Biophys. Acta 647 (1981) 232–240. [27] T. Tiffert, V.L. Lew, Kinetics of inhibition of the plasma membrane calcium pump by vanadate in intact human red cells, Cell Calcium 30 (2001) 337–342. [28] C. Toyoshima, H. Nomura, Structural changes in the calcium pump accompanying the dissociation of calcium, Nature 418 (2002) 605–611. [29] H.S. Young, C. Xu, P. Zhang, D.L. Stokes, Locating the thapsigargin-binding site on Ca(2+ )-ATPase by cryoelectron microscopy, J. Mol. Biol. 308 (2001) 231–240. [30] P. Artigas, D.C. Gadsby, Ouabain affinity determining residues lie close to the Na/K pump ion pathway, Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 12613–12618. [31] L.Y. Qiu, E. Krieger, G. Schaftenaar, H.G. Swarts, P.H. Willems, J.J. De Pont, J.B. Koenderink, Reconstruction of the complete ouabain-binding pocket of Na,KATPase in gastric H,K-ATPase by substitution of only seven amino acids, J. Biol. Chem. 280 (2005) 32349–32355. [32] J. Pande, M.M. Szewczyk, I. Kuszczak, S. Grover, E. Escher, A.K. Grover, Functional effects of caloxin 1c2, a novel engineered selective inhibitor of plasma membrane Ca(2+)-pump isoform 4, on coronary artery, J. Cell Mol. Med. 12 (2008) 1049–1060. [33] A.K Grover, S.E. Samson, Peroxide resistance of ER Ca2+ pump in endothelium: implications to coronary artery function, Am. J. Physiol. 273 (1997) C1250–C1258. [34] A.J. Caride, A.R. Penheiter, A.G. Filoteo, Z. Bajzer, A. Enyedi, J.T. Penniston, The plasma membrane calcium pump displays memory of past calcium spikes. Differences between isoforms 2b and 4b, J. Biol. Chem. 276 (2001) 39797– 39804. [35] S. Kawano, K. Otsu, S. Shoji, K. Yamagata, M. Hiraoka, Ca(2+) oscillations regulated by Na(+)–Ca(2+) exchanger and plasma membrane Ca(2+) pump induce fluctuations of membrane currents and potentials in human mesenchymal stem cells, Cell Calcium 34 (2003) 145–156.