Intracellular Coupling via Limiting Calmodulin

Measurements of cellular Ca2+-calmodulin concentrations have suggested that competition for limiting calmodulin may couple calmodulin-dependent activities. Here we have directly tested this hypothesis. We have found that in endothelial cells the amount of calmodulin bound to nitric-oxide synthase and the catalytic activity of the enzyme both are increased ∼3-fold upon changes in the phosphorylation status of the enzyme. Quantitative immunoblotting indicates that the synthase can bind up to 25% of the total cellular calmodulin. Consistent with this, simultaneous determinations of the free Ca2+ and Ca2+-calmodulin concentrations in these cells performed using indo-1 and a fluorescent calmodulin biosensor (Kd = 2 nm) indicate that increased binding of calmodulin to the synthase is associated with substantial reductions in the Ca2+-calmodulin concentrations produced and an increase in the Ca2+50 for formation of the calmodulin-biosensor complex. The physiological significance of these effects is confirmed by a corresponding 40% reduction in calmodulin-dependent plasma membrane Ca2+ pump activity. An identical reduction in pump activity is produced by expression of a high affinity (Kd = 0.3 nm) calmodulin biosensor, and treatment to increase calmodulin binding to the synthase then has no further effect. This suggests that the observed reduction in pump activity is due specifically to reduced calmodulin availability. Increases in synthase activity thus appear to be coupled to decreases in the activities of other calmodulin targets through reductions in the size of a limiting pool of available calmodulin. This exemplifies what is likely to be a ubiquitous mechanism for coupling among diverse calmodulin-dependent activities. Measurements of cellular Ca2+-calmodulin concentrations have suggested that competition for limiting calmodulin may couple calmodulin-dependent activities. Here we have directly tested this hypothesis. We have found that in endothelial cells the amount of calmodulin bound to nitric-oxide synthase and the catalytic activity of the enzyme both are increased ∼3-fold upon changes in the phosphorylation status of the enzyme. Quantitative immunoblotting indicates that the synthase can bind up to 25% of the total cellular calmodulin. Consistent with this, simultaneous determinations of the free Ca2+ and Ca2+-calmodulin concentrations in these cells performed using indo-1 and a fluorescent calmodulin biosensor (Kd = 2 nm) indicate that increased binding of calmodulin to the synthase is associated with substantial reductions in the Ca2+-calmodulin concentrations produced and an increase in the Ca2+50 for formation of the calmodulin-biosensor complex. The physiological significance of these effects is confirmed by a corresponding 40% reduction in calmodulin-dependent plasma membrane Ca2+ pump activity. An identical reduction in pump activity is produced by expression of a high affinity (Kd = 0.3 nm) calmodulin biosensor, and treatment to increase calmodulin binding to the synthase then has no further effect. This suggests that the observed reduction in pump activity is due specifically to reduced calmodulin availability. Increases in synthase activity thus appear to be coupled to decreases in the activities of other calmodulin targets through reductions in the size of a limiting pool of available calmodulin. This exemplifies what is likely to be a ubiquitous mechanism for coupling among diverse calmodulin-dependent activities. The Ca2+-binding protein calmodulin (CaM) 1The abbreviations used are: CaM, calmodulin; eNOS, endothelial nitric-oxide synthase; PMCA, plasma membrane Ca2+ pump; BAEC, bovine aortic endothelial cell; FSK, forskolin; IBMX, 3-isobutyl-1-methylxanthine; l-NAME, Nω-nitro-l-arginine methyl ester; BSCaM2 and BSCaM0.3, fluorescent CaM biosensors with apparent Kd for CaM of 2 and 0.3 nm, respectively; BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid. is involved in essentially all aspects of cellular function through its many target proteins, which include adenylyl cyclases and phosphodiesterases (1Gu C. Cooper D.M. J. Biol. Chem. 1999; 274: 8012-8021Google Scholar), numerous protein kinases (2Nairn A.C. Picciotto M.R. Semin. Cancer Biol. 1994; 5: 295-303Google Scholar), the protein phosphatase calcineurin (3Aramburu J. Rao A. Klee C.B. Curr. Top Cell. Regul. 2000; 36: 237-295Google Scholar), nitric-oxide synthase (4Bredt D.S. Snyder S.H. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 682-685Google Scholar), the plasma membrane Ca2+ pump (5Jarrett H.W. Penniston J.T. Biochem. Biophys. Res. Commun. 1977; 77: 1210-1216Google Scholar, 6Vincenzi F.F. Larsen F.L. Fed. Proc. 1980; 39: 2427-2431Google Scholar), and several ion channels (7Levitan I.B. Neuron. 1999; 22: 645-648Google Scholar). Measurements of the Ca2+-CaM concentrations produced in living cells (8Romoser V.A. Hinkle P.M. Persechini A. J. Biol. Chem. 1997; 272: 13270-13274Google Scholar, 9Persechini A. Cronk B. J. Biol. Chem. 1999; 274: 6827-6830Google Scholar) have suggested that the intracellular pool of CaM is limiting, i.e. the concentration of available CaM in the cell is less than the concentration of CaM-binding sites (9Persechini A. Cronk B. J. Biol. Chem. 1999; 274: 6827-6830Google Scholar, 10Teruel M.N. Chen W. Persechini A. Meyer T. Curr. Biol. 2000; 10: 86-94Google Scholar). This has led us to propose that competition for a limiting pool of CaM likely constitutes a pervasive mechanism for coupling among CaM-dependent activities (11Persechini A. Stemmer P.M. Trends Cardiovasc. Med. 2002; 12: 32-37Google Scholar). In this study we have directly tested this hypothesis and have found that in endothelial cells increases in the CaM binding ability of nitricoxide synthase (eNOS) are correlated with significant reductions in the free Ca2+-CaM concentrations produced and in CaM-dependent activity of the plasma membrane Ca2+ pump (PMCA). Cell Culture and Transfection—Bovine aortic endothelial cells (BAECs) were purchased from Coriell Institute for Medical Research (Camden, NJ; repository number AG04762A) and cultured in Ham's F-12 medium containing 10% fetal bovine serum. Cells were used between 6 and 10 passages. Cells were plated onto glass coverslips and grown until subconfluence for transfection. pcDNA3.1 vectors encoding CaM biosensors were then transfected into the cells using Targefect F-1 (Targeting System, Santee, CA). Western Blotting and Immunoprecipitation—Subconfluent BAECs were stimulated with 1 μm ionomycin in the presence of 1 mm CaCl2 with or without pretreatment with 50 μm FSK and 0.5 mm 3-isobutyl-1-methylxanthine (IBMX). After 3 min, cells were lysed, and homogenates were processed as described elsewhere (12Fleming I. Fisslthaler B. Dimmeler S. Kemp B.E. Busse R. Circ. Res. 2001; 88: E68-E75Google Scholar). Anti-eNOS antibody and anti-CaM antibodies were obtained from Zymed Laboratories Inc. (San Francisco, CA); antibodies against phospho-Thr-497 and phospho-Ser-1179 (rabbit polyclonal IgG) were purchased from Upstate Biotechnology (Lake Placid, NY). Blotting was performed as per the manufacturer's instructions. Measurement of NO Production—Subconfluent BAECs were washed with normal HEPES-buffered saline (NHBS, composition in mm: 141 NaCl, 1 CaCl2, 1 MgSO4, 5 KCl, 10 HEPES, and 10 glucose) and incubated at room temperature for 30 min in the presence or absence of 50 μm forskolin and 0.5 mm IBMX with either 100 μm l-NAME or 100 μm l-arginine. This medium was then removed, and cells were stimulated by the addition of 0.5 ml of a buffer containing 1 μm ionomycin, in addition to the above concentrations of l-NAME or l-arginine, and forskolin/IBMX. After 5 min, the supernatant was collected and treated to reduce nitrate to nitrite using a procedure involving catalysis with cadmium metal (Nitralyzer II, World Precision Instruments). Nitrite was then converted to NO and measured using an NO-specific electrode as described by the manufacturer (ISO-NOP MARK II, World Precision Instruments). Simultaneous Measurement of Free Ca2+ and Ca2+-CaM Concentrations—Biosensor ECFP and EYFP fluorescence was collected using 480/40M and 535/30M emission filters, while indo-1 fluorescence was collected using 405/30M and 485/25M filters (Chroma Technology, Brattleboro, VT). Alternating 340 (indo-1) and 435 (biosensor) excitation light was provided by a DeltaRAM rapid-switching monochromator (PTI International) coupled with a custom-made 410/30M-460LP microscope polychroic (Chroma Technology). There is no detectable spillover between indo-1 and biosensor channels in this system. Alternating indo-1 (405/485) and biosensor (480/535) emission ratios were determined at 1-s intervals. Subconfluent BAECs transiently expressing CaM biosensors were incubated with indo-1/AM (6 μm) and the specified pharmacological agents or equal volumes of vehicle control (Me2SO) for 30 min. Free Ca2+ concentration was estimated from indo-1 emission ratios using a 460 nmKd value determined in BAECs by comparison with the response of a CaM-based Ca2+ biosensor (9Persechini A. Cronk B. J. Biol. Chem. 1999; 274: 6827-6830Google Scholar). Free Ca2+-CaM concentrations were determined from biosensor emission ratios as described previously and are “effective” values for free (Ca2+)4-CaM, as they are calculated based on biosensor Kd values for this fully liganded CaM species (9Persechini A. Cronk B. J. Biol. Chem. 1999; 274: 6827-6830Google Scholar). Thermodynamic coupling effects and differences in the mechanisms controlling CaM-binding to different targets make it difficult to precisely define the CaM species in equilibrium with a CaM-biosensor complex under all conditions in the cell, hence we prefer the generic “Ca2+-CaM” designation. Nevertheless, calculated effective values are useful for comparative purposes, and as a means of evaluating, based on its affinity for (Ca2+)4-CaM, whether a given target is likely to compete successfully for CaM in the cell. Competitive Binding Assay for eNOS and BSCaM2—Equimolar BSCaM2 and eNOS in a buffer containing 25 mm Tris-HCl, 3 mm BAPTA, 0.1 m KCl, 1 mm MgCl2, 0.1 mm CaCl2, and 0.1 mg/ml bovine serum albumin, pH 7.4, were titrated with CaM until the maximal BSCaM2 fluorescence response was obtained. Titration data were fit to a quadratic equation for competitive binding to extract the Kd value for CaM binding to eNOS. For several reasons, we chose to use endothelial cells for our initial investigations of coupling among calmodulin target activities. First, eNOS is a CaM-binding protein of undisputed physiological importance (13Huang P.L. Huang Z. Mashimo H. Bloch K.D. Moskowitz M.A. Bevan J.A. Fishman M.C. Nature. 1995; 377: 239-242Google Scholar), whose catalytic activity is increased up to 20-fold by Ca2+-CaM (4Bredt D.S. Snyder S.H. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 682-685Google Scholar). Second, CaM binding to eNOS can be manipulated experimentally, as it is known to be influenced by in vivo phosphorylation at one or more residues (14Fleming I. Busse R. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2003; 284: R1-R12Google Scholar). Thr-497 (Thr-495 in the human sequence) in the putative CaM-binding domain appears to be a particularly important determinant of CaM binding affinity. Dephosphorylation at this site is associated with a significantly increased CaM binding ability of eNOS, whereas phosphorylation occurs constitutively and is correlated with decreased binding (12Fleming I. Fisslthaler B. Dimmeler S. Kemp B.E. Busse R. Circ. Res. 2001; 88: E68-E75Google Scholar, 14Fleming I. Busse R. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2003; 284: R1-R12Google Scholar). Conditions that cause Thr-497 dephosphorylation have generally been found to also produce phosphorylation at Ser-1179, so this site may also play a role in controlling CaM binding to the synthase (15Michell B.J. Chen Z. Tiganis T. Stapleton D. Katsis F. Power D.A. Sim A.T. Kemp B.E. J. Biol. Chem. 2001; 276: 17625-17628Google Scholar). The focus of this study is not on eNOS phosphorylation per se, but is instead on how physiologically relevant changes in the phosphorylation status of eNOS affect CaM availability in the endothelial cell. To increase CaM binding to eNOS in BAECs, we used combined treatment with FSK and IBMX, which has previously been shown to mimic, albeit in a more sustained manner, agonist-evoked Thr-497 dephosphorylation and Ser-1179 phosphorylation (15Michell B.J. Chen Z. Tiganis T. Stapleton D. Katsis F. Power D.A. Sim A.T. Kemp B.E. J. Biol. Chem. 2001; 276: 17625-17628Google Scholar). In our hands, FSK/IBMX treatment produces a ∼3-fold increase in the amount of CaM bound to eNOS in cell homogenates (Fig. 1A). Western blotting indicates a similar 2.3-fold decrease in Thr-497 phosphorylation, with a concomitant 2.5-fold increase in Ser-1179 phosphorylation. To verify that FSK/IBMX treatment increases CaM-dependent eNOS activity, NO production was measured. This activity is also increased ∼3-fold and is completely inhibited by the NOS inhibitor l-NAME (Fig. 1B). Changes in CaM availability are seen as changes in the apparent free Ca2+-CaM concentrations produced at comparable free Ca2+ concentrations. Therefore, we have simultaneously measured both free Ca2+ and free Ca2+-CaM concentrations produced in BAECs in single cells using indo-1 and CaM biosensors. We have previously developed fluorescence biosensors to monitor dynamic changes in Ca2+-CaM concentrations in living cells (8Romoser V.A. Hinkle P.M. Persechini A. J. Biol. Chem. 1997; 272: 13270-13274Google Scholar, 9Persechini A. Cronk B. J. Biol. Chem. 1999; 274: 6827-6830Google Scholar). In this study, the ability to simultaneously monitor free Ca2+ and free Ca2+-CaM has allowed us for the first time to precisely assess their relationship under different experimental conditions. Fig. 2, A and B, contain pseudo-color fluorescence ratio images of cells transiently expressing a 2 nmKd CaM biosensor (BSCaM2) and also loaded with indo-1. All the cells take up indo-1 (Fig. 2A), and generally ∼20% express BSCaM2 (Fig. 2B). Under control conditions, ionomycin rapidly increases the apparent free Ca2+-CaM concentration from below 0.1 to ∼8 nm (Fig. 2C). Pretreatment with FSK/IBMX causes a ∼3-fold reduction in the peak free Ca2+-CaM concentrations produced in response to ionomycin, with a slight increase in the peak free Ca2+ concentrations (Fig. 2D). Hence, the observed reduction in free Ca2+-CaM is due to reduced CaM availability, not a decrease in the free Ca2+ concentration. The mean Ca2+ concentration producing 50% of the peak BSCaM2 fractional response (Ca2+50 value) is increased from 295 ± 18 nm to 502 ± 84 nm (p < 0.05, n = 6) by FSK/IBMX treatment (Fig. 2E). Blocking eNOS activity with 100 μm l-NAME does not alter the effects of FSK/IBMX, indicating that increased NO production is not a contributing factor. To verify that changes in CaM availability could account for the observed changes in both maximal biosensor response and Ca2+ sensitivity, we have determined in vitro how formation of the CaM-BSCaM2 complex at a fixed CaM concentration is affected by increasing amounts of a CaM-binding peptide (nPEP) (16Persechini A. White H.D. Gansz K.J. J. Biol. Chem. 1996; 271: 62-67Google Scholar) (Fig. 2F). At a low peptide concentration, the Ca2+ sensitivity of the biosensor response is reduced, followed at higher concentrations by decreases in both the Ca2+ sensitivity and magnitude of the response. Thus, changes in CaM availability are sufficient to explain the observed changes in both the magnitude and Ca2+ sensitivity of BSCaM2 response in BAECs (Fig. 2E). A potential difficulty with the results presented so far is that the biosensor may itself reduce CaM availability enough to exaggerate the effects of FSK/IBMX treatment. To address this potential problem, and to confirm the physiological significance of the observed changes in CaM availability, we have investigated the effect of FSK/IBMX and biosensor expression on PMCA activity. Among the major routes for cytosolic Ca2+ removal, only the PMCA has been shown to depend directly on Ca2+-CaM, which is bound with an apparent dissociation constant of 4 –10 nm and increases PMCA activity up to 10-fold (17Caride A.J. Elwess N.L. Verma A.K. Filoteo A.G. Enyedi A. Bajzer Z. Penniston J.T. J. Biol. Chem. 1999; 274: 35227-35232Google Scholar). To determine PMCA activity we first inhibited the SERCA pump with thapsigargin and the Na+-Ca2+ exchanger by replacing Na+ in cell buffers with an equimolar amount of N-methyl-d-glucamine. l-NAME was also applied as a precaution against NO-dependent effects. Under these conditions PMCA activity is directly proportional to the Ca2+ extrusion rate (18Sedova M. Blatter L.A. Cell Calcium. 1999; 25: 333-343Google Scholar, 19Wang X. Reznick S. Li P. Liang W. van Breemen C. Cell Calcium. 2002; 31: 265-277Google Scholar). Since PMCA activity is itself Ca2+-dependent, extrusion rates were determined for cells grouped on the free Ca2+ concentration (300 – 600 nm) at the start of each extrusion time course (Fig. 3A). Relaxation times (τ) for Ca2+ extrusion from individual cells were then estimated by fitting the first 50 s of extrusion time courses to a mono-exponential (Fig. 3A). A similar approach has been used elsewhere to assess PMCA activity in endothelial cells (18Sedova M. Blatter L.A. Cell Calcium. 1999; 25: 333-343Google Scholar, 19Wang X. Reznick S. Li P. Liang W. van Breemen C. Cell Calcium. 2002; 31: 265-277Google Scholar). As seen in Fig. 3, B and D, BSCaM2 expression does not alter PMCA activity, but FSK/IBMX treatment reduces it by ∼40% in both wild-type and BSCaM2-expressing BAECs. This effect may also partly explain the slight increase in peak free Ca2+ concentrations observed in cells treated with FSK/IBMX. We reasoned that if the effect of FSK/IBMX on PMCA activity is due specifically to reduced CaM availability, it should be reproduced by a CaM antagonist. We therefore expressed a CaM biosensor with a 0.3 nmKd for Ca2+-CaM (BSCaM0.3) in BAECs. This high affinity biosensor reduces PMCA activity to essentially the same extent as does FSK/IBMX treatment of wild-type BAECs (Fig. 3, C and D). Peak free Ca2+-CaM concentrations are reduced to ∼2 nm in cells expressing BSCaM0.3 (Fig. 4, A and B), compared with ∼8 nm in cells expressing BSCaM2. Most important, subsequent FSK/IBMX treatment has no further effect on PMCA activity (Fig. 3, C and D) or CaM availability (Fig. 4). Interestingly, FSK/IBMX treatment of cells expressing BSCaM2 also reduces free Ca2+-CaM to ∼2 nm (Fig. 2C), indicating that this is below what is needed for significant PMCA activation, and most likely accounting for the lack of any additional effect of FSK/IBMX in cells expressing BSCaM0.3. Consistent with this, published values for the apparent Kd value for CaM binding to the pump range from 4 to 10 nm (17Caride A.J. Elwess N.L. Verma A.K. Filoteo A.G. Enyedi A. Bajzer Z. Penniston J.T. J. Biol. Chem. 1999; 274: 35227-35232Google Scholar). Although a reduction in the amount of CaM bound to the pump due to reduced CaM availability appears to adequately explain our results, other factors, notably phosphorylation, also can significantly affect PMCA activity (20Caroni P. Carafoli E. J. Biol. Chem. 1981; 256: 9371-9373Google Scholar, 21Bruce J.I. Yule D.I. Shuttleworth T.J. J. Biol. Chem. 2002; 277: 48172-48181Google Scholar, 22James P.H. Pruschy M. Vorherr T.E. Penniston J.T. Carafoli E. Biochemistry. 1989; 28: 4253-4258Google Scholar, 23Smallwood J.I. Gugi B. Rasmussen H. J. Biol. Chem. 1988; 263: 2195-2202Google Scholar, 24Dean Chen D. J. Biol. Chem. 1997; 272: Scholar). for our indicate that FSK/IBMX treatment PMCA activity in BAECs by the free Ca2+-CaM concentrations that are This that the changes in CaM availability by BSCaM2 with physiologically significant changes in the activities of CaM targets in endothelial CaM binding to eNOS is correlated with significant reductions in both the Ca2+ sensitivity and of BSCaM2 response. changes coupling between increases in eNOS activity and decreases in the activities of other CaM as the In this it should be that the response of BSCaM2 itself is also of a CaM target as the response to a free Ca2+ concentration is A mechanism is in which additional CaM is bound to the synthase at the of other CaM-binding in the cell. Consistent with this immunoblotting indicates total eNOS and CaM concentrations in homogenates of ± and ± total not eNOS can bind up to 25% of total CaM in BAECs. a competitive binding we have determined an in vitro Kd value of nm for the complex between eNOS and This value indicates a major role for eNOS in controlling CaM availability, FSK/IBMX treatment reduces availability a range in the free Ca2+-CaM concentrations that binding to targets with Kd values nm, as the PMCA and but not the concentration range that binding to with dissociation to that of eNOS, as BSCaM0.3. CaM availability is a function of the Ca2+ and concentrations of all CaM-binding in the cell. A high affinity CaM-binding as BSCaM0.3, a pool of available CaM than does a affinity A complex with high Ca2+ sensitivity a pool than does one with a sensitivity, the can with CaM under conditions the In all an CaM-binding protein have a on CaM availability than one in The presented that through CaM is by the of cellular CaM-binding and is therefore by changes in their for the of other CaM-binding in the response to FSK/IBMX, our results that occurs and that physiologically relevant changes in the CaM-binding of eNOS to of CaM-binding is a among CaM is that this of can CaM availability, the of CaM targets and the effect of an initial phosphorylation or dephosphorylation a if involving CaM is under study it is important to both free Ca2+ and CaM availability, in biosensor or measured free We B. S. S. and J. of for the eNOS.