Membranous adenylyl cyclase 1 activation is regulated by oxidation of N- and C-terminal methionine residues in calmodulin

Biochemical Pharmacology 93 (2015) 196–209

Contents lists available at ScienceDirect

Biochemical Pharmacology journal homepage: www.elsevier.com/locate/biochempharm

Membranous adenylyl cyclase 1 activation is regulated by oxidation of N- and C-terminal methionine residues in calmodulin Carolin Lu¨bker a, Ramona J. Bieber Urbauer b, Jackob Moskovitz c, Stefan Dove d, Jasmin Weisemann e, Maria Fedorova f, Jeffrey L. Urbauer b, Roland Seifert a,* a

Institute of Pharmacology, Hannover Medical School, Carl-Neuberg-Str.1, D-30625 Hannover, Germany Department of Chemistry, University of Georgia, 140 Cedar Street, Athens, GA 30602-2556, USA Department of Pharmacology and Toxicology, University of Kansas, 1251 Wescoe Hall Drive, Lawrence, KS 66045, USA d Institute of Pharmacy, University of Regensburg, Universita¨tsstr. 31, D-93053 Regensburg, Germany e Institute of Toxicology, Hannover Medical School, Carl-Neuberg-Str.1, D-30625 Hannover, Germany f Institute of Bioanalytical Chemistry, University of Leipzig, Deutscher Platz 5, D-04103 Leipzig, Germany b c

A R T I C L E I N F O

A B S T R A C T

Article history: Received 27 August 2014 Accepted 17 November 2014 Available online 25 November 2014

Membranous adenylyl cyclase 1 (AC1) is associated with memory and learning. AC1 is activated by the eukaryotic Ca2+-sensor calmodulin (CaM), which contains nine methionine residues (Met) important for CaM-target interactions. During ageing, Met residues are oxidized to (S)- and (R)-methionine sulfoxide (MetSO) by reactive oxygen species arising from an age-related oxidative stress. We examined how oxidation by H2O2 of Met in CaM regulates CaM activation of AC1. We employed a series of thirteen mutant CaM proteins never assessed before in a single study, where leucine is substituted for Met, in order to analyze the effects of oxidation of specific Met. CaM activation of AC1 is regulated by oxidation of all of the C-terminal Met in CaM, and by two N-terminal Met, M36 and M51. CaM with all Met oxidized is unable to activate AC1. Activity is fully restored by the combined catalytic activities of methionine sulfoxide reductases A and B (MsrA and B), which catalyze reduction of the (S)- and (R)-MetSO stereoisomers. A small change in secondary structure is observed in wild-type CaM upon oxidation of all nine Met, but no significant secondary structure changes occur in the mutant proteins when Met residues are oxidized by H2O2, suggesting that localized polarity, flexibility and structural changes promote the functional changes accompanying oxidation. The results signify that AC1 catalytic activity can be delicately adjusted by mediating CaM activation of AC1 by reversible Met oxidation in CaM. The results are important for memory, learning and possible therapeutic routes for regulating AC1. ß 2014 Elsevier Inc. All rights reserved.

Keywords: Adenylyl cyclase Calmodulin Methionine oxidation Reactive oxygen species Ageing

1. Introduction The conversion of ATP into the second messenger cAMP, which regulates numerous physiological functions including cardiac

Abbreviations: AC, adenylyl cyclase; AD, Alzheimer’s disease; CaM, calmodulin; CaM-mut, mutant calmodulin species; CaM-wt, wild-type calmodulin; CD, circular dichroism; DTT, dithiothreitol; EGTA, ethylene glycol-bis(2-aminoethylether)N,N,N0 ,N0 -tetraacetic acid; H2O2, hydrogen peroxide; L, leucine residue; Leu, leucine; M, methionine residue; Met, methionine; MetSO, methionine sulfoxide; MsrA, methionine sulfoxide reductase A; MsrB, methionine sulfoxide reductase B; NADPH, nicotinamide adenine dinucleotide phosphate; eNOS, endothelial nitric oxide synthase; nNOS, neuronal nitric oxide synthase; PDE, cyclic nucleotide phosphodiesterase; PMCA, plasma membrane Ca2+-ATPase; ROS, reactive oxygen species; Sf9, Spodoptera frugiperda insect cells. * Corresponding author. E-mail address: [email protected] (R. Seifert). http://dx.doi.org/10.1016/j.bcp.2014.11.007 0006-2952/ß 2014 Elsevier Inc. All rights reserved.

contractility [1,2], smooth muscle relaxation [3] and olfaction [4], is catalyzed by adenylyl cyclases (ACs) [5]. The family of membranous ACs consists of nine isoforms (AC1-AC9), which are classified into four subfamilies based on their regulation by stimulatory and inhibitory molecules [5]. ACs 1 and 8 are activated by the highly conserved Ca2+-sensor calmodulin (CaM) [6,7]. Two CaM-binding sites on AC1 are known: a Ca2+-independent binding site of 28 amino acids in the C1b domain [8] and a Ca2+-dependent binding site of 14 amino acids in the C2a-region [9]. AC1 is expressed in specific brain regions (hippocampus, neocortex, entorhinal cortex and cerebellar cortex) associated with memory and learning [10]. Studies with transgenic and knockout animal models provide strong evidence that AC1 plays important roles in these processes [11,12]. Oxidative stress, resulting from increased cellular levels of reactive oxygen species (ROS), is associated with ageing and many

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

disease states. Under conditions of oxidative stress, amino acids in proteins susceptible to oxidation, such as cysteine and methionine (Met), can be readily oxidized. Oxidation can therefore result in significant structural or functional changes to the protein [13–16]. This is particularly relevant for critical regulatory proteins, such as CaM, which interact with many different cellular target proteins [13,17]. There are nine Met in CaM, and all of these are sensitive to oxidation. These Met residues also interact directly with target proteins [18]. They are arranged into two N-terminal Met patches (M36, M51 and M71, M72, M76) and two C-terminal Met patches (M109, M124 and M144, M145) (Fig. 1). Oxidation of the hydrophobic and flexible Met results in a hydrophilic and slightly more rigid methionine sulfoxide (MetSO), as a mixture of (S)- and (R)-diastereomers [19,20]. Under conditions of oxidative stress associate with ageing, oxidized CaM is known to accumulate. For instance, in CaM from the brains of old rats, a significantly higher percentage of the Met are oxidized compared to CaM from the brains of young rats [21]. Oxidation of Met in CaM has likewise been shown to affect the ability of CaM to activate target proteins. For example, compared to wild-type CaM (CaM-wt), oxidized CaM is unable to fully activate the plasma membrane Ca2+-ATPase (PMCA) [21]. In this case, it was demonstrated that even CaM with all Met residues oxidized was still able to bind to the PMCA, but not always productively [22]. In the case of CaM activation of the nitric oxide synthetases (NOS), CaM with all Met residues oxidized was unable to activate nitric oxide production by either neuronal (nNOS) or endothelial (eNOS) [23]. Interestingly, this oxidized CaM was able to partially activate other activities of nNOS and eNOS, for instance cytochrome c reduction [23]. In proteins with multiple Met residues, such as CaM, there is no facile means for selective oxidation of individual Met residues. Thus, determining how CaM function is regulated by oxidation of specific Met residues is challenging. One route that has proven successful is substitution of Met for leucine (Leu) by site-directed mutagenesis. Leu is comparable to Met in terms of volume, hydrophobicity and a-helical propensity [24]. Moreover, the

197

tertiary structure of CaM is only nominally perturbed by Met to Leu substitutions, and CaM-mutants (CaM-mut) bind to, and activate, CaM targets similarly or with marginal differences compared to CaM-wt [23,25–27]. This strategy was used to demonstrate that oxidation of just M144 and M145 in CaM was responsible for the inability of CaM to fully activate the PMCA; oxidation of the remaining Met residues had no affect [26]. In contrast, oxidizing M144 and M145 in CaM had virtually no effect on the ability of CaM to fully activate eNOS and nNOS, but it did increase the concentration of CaM necessary for half-maximal activation [23]. This suggests that regulating activation of eNOS and nNOS by Met oxidation in CaM involves other Met residues in the C-terminal domain or residues in the N-terminal domain [23]. This approach was also used recently to determine that oxidation of N-terminal Met in CaM is necessary for targeting oxidized CaM for degradation by the 20S proteasome [28]. Oxidation of Met in proteins can be reversed (‘‘repaired’’) in cells by the catalytic activities of the methionine sulfoxide reductases (Msr), and in the absence of irreversible consequences of the oxidation (unfolding, aggregation), the native functions of the proteins can be recovered [29–32]. In eukaryotes, there typically are two isoforms, MsrA and MsrB, that catalyze the stereospecific reduction of the (S)- and (R)-enantiomers of MetSO, respectively [33,34]. Thus, the Msr enzymes have critical regulatory functions, both rescuing proteins targeted for degradation and reversing functional consequences of oxidation. These enzymes are also very useful tools for reducing MetSO in vitro, for examining the reversibility of functional changes accompanying oxidation, and potentially for determining the stereospecificity of oxidation. Given the age induced oxidative stress in the brain, the facile oxidation of Met in CaM by ROS, and the regulation of target activation by CaM via Met oxidation in CaM, it is prudent to examine the consequences of oxidizing specific Met residues in CaM for AC1 activation. It is conceivable that a decline in AC1 activity caused by an impaired activation of AC1 by oxidized CaM is involved in neurodegenerative diseases, such as Alzheimer’s

Fig. 1. Localization of the nine Met residues in Ca2+-saturated CaM. The N-terminal domain is connected with the C-terminal domain via a flexible a-helical linker. Each domain contains two Ca2+-binding motifs (EF hands). The backbone is shown as stick model (a-helices and b-sheets – thick, turns and loops – thin). The nine Met residues are represented as ball and stick models (light-coloured sulfur atoms). The Met residues are arranged into four patches: two N-terminal patches with M36, M51 and M71, M72, M76 and two C-terminal patches with M109, M124 and M144, M145. The model was generated with the modelling suite Sybyl-X (Certara, L.P., St. Louis, MO) from the highresolution crystal structure of human CaM [76] (Brookhaven Protein Data Bank ID 1CLL).

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

198

disease [35,36]. For these reasons alone, it is important to assess the impact of CaM oxidation on AC1 activation. Here we analyzed in detail how oxidation by H2O2 of specific Met residues in CaM regulates AC1 activation by CaM, a question never studied before although AC1 was cloned more than 20 years ago [6]. We examined a comprehensive panel of thirteen CaM-mut with distinct Met to Leu substitutions in the C-terminal as well as in the N-terminal domains of CaM (Table 1). Five of these CaM-mut (native and/or oxidized L9, M144/L8, M145/L8 and M144, M145/ L7) were studied before with regard to activation of CaM-targets such as NOS [23], PMCA [26] and CyaA [27], but the remaining eight CaM-mut (Table 1) were exclusively cloned, expressed, purified and characterized for our present AC1 studies. CaM-mut L9, M144/L8, M145/L8 and M144,M145/L7 were already investigated with regard to AC1 activation in a previous study by our group, but only in the native and not in the oxidized state [27]. Unlike simpler cases, such as CaM activation of the PMCA, where regulation by Met oxidation in CaM results from oxidation of only two of the nine Met residues, for AC1 regulation is rich and complex, with oxidation of many of the Met residues contributing to the overall regulation by oxidation. The results lay the groundwork for understanding how oxidative stress controls AC1 function. 2. Materials and methods 2.1. Materials Amicon centrifugal filter devices (10 K) and Coomassie Brilliant Blue G250 were purchased from Merck Millipore (Darmstadt, Germany). Baculovirus encoding bovine AC1 was provided by Dr. Alfred G. Gilman (University of Texas, Dallas, TX) and Dr. Roger Sunahara (University of Michigan, Ann Arbor, MI) [6,37]. Spodoptera frugiperda (Sf9) insect cells for expression of ACs were from the American Type Culture Collection (Rockville, MD). InsectXPRESSTM Media and gentamicin sulfate for Sf9 cell culture were acquired from Lonza (Basel, Switzerland). [a-32P]ATP (3000 Ci/ mmol) was purchased from Hartmann (Braunschweig, Germany). All other materials used for AC activity assays were from sources as described previously [38]. H2O2 was purchased from Sigma– Aldrich (St. Louis, MO). HisPurTM Ni-NTA resins, restriction enzymes BamHI and HindIII (Fermentas), dialysis membrane Slide-A-Lyzer1 devices (10 K, 0.5–3 ml, Pierce), and the Page RulerTM Plus Prestained Protein #SM1811 and #0679 protein standards (both Fermentas) were acquired from Thermo Scientific (Rockford, IL). Bio-Rad DC protein assay kit was purchased from Bio-Rad (Hercules, CA).

2.2. Expression of membranous AC1 in Sf9 insect cells and preparation of cell membranes Infection of Sf9 insect cells with baculovirus harbouring the gene encoding membranous AC1, and preparation of AC1 in cell membranes were performed essentially as described previously [39]. Sf9 cells were cultured in Insect-XPRESSTM media supplemented with 100 mg/ml gentamicin and 5% (v/v) foetal bovine serum at 28 8C. For expression of membranous AC1, 3  106 cells/ml were infected with a dilution of 1:100 of high-titre baculovirus. The cell suspension was incubated for 48 h at 28 8C. For preparation of membranes containing AC1, harvested cells were washed with 1 Dulbecco’s phosphate buffered saline without Ca2+ and Mg2+ (PAA laboratories, Pasching, Austria). After centrifugation at 1000  g for 10 min at 4 8C, the cell pellet was resuspended in lysis buffer, pH 7.4 (10 mM Tris–HCl, 1 mM EDTA and 10 mg/ml benzamide, 10 mg/ml leupeptin and 200 mM phenylmethanesulfonyl fluoride as protease inhibitors). Cells were lysed with 25 strokes in a Dounce homogenizer. A centrifugation step at 500  g for 5 min at 4 8C was conducted to sediment the nuclei. The supernatant suspension containing the cell membranes was centrifuged at 40 000  g for 20 min at 4 8C. The supernatant fluid was discarded. At this point CaM was removed from the membranes by Ca2+ chelation using EGTA. The cell pellet was resuspended and incubated for 10 min at 4 8C in HEED buffer, pH 7.4 [6] (20 mM HEPES, 1 mM EDTA, 1 mM EGTA, 2 mM DTT and 10 mg/ml benzamide, 10 mg/ml leupeptin and 200 mM phenylmethanesulfonyl fluoride as protease inhibitors) as described previously [6,9,40]. Following incubation, membranes were centrifuged at 40 000  g for 20 min at 4 8C. The supernatant fluid was discarded and the cell pellet was resuspended again in lysis buffer. A final centrifugation step was performed at 40 000  g for 20 min at 4 8C. The supernatant fluid was discarded and the cell pellet was resuspended in binding buffer, pH 7.4 (75 mM Tris–HCl, 12.5 mM MgCl2 and 1 mM EDTA). The protein concentration of membrane preparations was determined by using the Lowry method [41] with the Bio-Rad DC protein assay kit using bovine serum albumin as the standard. Membranes were stored at 80 8C. 2.3. CaM mutagenesis, expression and purification The CaM used for these studies was expressed in E. coli and purified as described previously [26]. The CaM gene subcloned into the E. coli expression vector is the chicken gene (Uni-Prot entry P62149), which codes for a protein with an amino acid sequence identical to the human protein (Uni-Prot entry P62158). The Nterminal Met is cleaved during expression, and the amino acid sequence numbering used here neglects this Met (A1-K148).

Table 1 Nomenclature of CaM-muta CaM

Position in CaM 36

51

71

72

76

109

124

144

145

wt L9 M36/L8 M51/L8 M71/L8 M72/L8 M76/L8 M144/L8 M145/L8 M36,M51/L7 M71,M72,M76/L6 M109,M124/L7 M144,M145/L7 L2

M L M L L L L L L M L L L M

M L L M L L L L L M L L L M

M L L L M L L L L L M L L M

M L L L L M L L L L M L L M

M L L L L L M L L L M L L M

M L L L L L L L L L L M L M

M L L L L L L L L L L M L M

M L L L L L L M L L L L M L

M L L L L L L L M L L L M L

a

Positions of methionines (M) and leucine (L) substitutions in CaM-wt and analyzed CaM-mut.

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

Site-directed mutagenesis of CaM to substitute Leu for Met was performed precisely as described previously [26]. The masses of all CaM-mut were confirmed by mass spectrometry. CaM was stored in a solution composed of 1 mM imidazole, 0.1 mM KCl and 10 mM CaCl2, pH 6.5. Before use, it was dialyzed exhaustively against deionized water. CaM concentration was determined by using the Lowry method with the Bio-Rad DC protein assay kit using bovine serum albumin as the standard. Aliquots of 60 mM CaM were stored at 80 8C. Positions of Met to Leu substitutions are shown in Table 1.

199

2.7. Expression and purification of MsrA Yeast MsrA was expressed and purified using immobilized nickel affinity chromatography as described [44]. The purity of the protein was analyzed by 12.5% (w/v) SDS-PAGE. After staining the proteins with Coomassie Brilliant Blue G250, a single band at 24 kDa was observed, demonstrating the purity of the MsrA protein (data not shown). 2.8. Cloning, expression and purification of MsrB3A

2.4. Oxidation of Met in CaM Met residues in CaM were oxidized in vitro using H2O2. Aliquots containing 60 mM CaM-wt or CaM-mut in deionized water were incubated with 0.1 mM CaCl2 and the desired concentration of H2O2 for 24 h at 25 8C. For experiments to examine how the electrophoretic mobility of CaM-wt and CaM activation of AC1 is regulated by the degree of Met oxidation in CaM, various concentrations of H2O2 (0.05 mM, 0.5 mM, 5 mM and 50 mM) were used. For experiments requiring complete oxidation of all Met residues in CaM-wt and CaM-mut, 50 mM H2O2 was used. After the 24 h incubation, H2O2 and excess Ca2+ were removed by solvent exchange using an Amicon centrifugal ultrafiltration device (10 K cutoff, five centrifugation steps, 14 000  g for 10 min each at 4 8C). For circular dichroism (CD) measurements, the solvent was exchanged for deionized water. For AC assays and assays of Msr activity, the solvent was exchanged for a 30 mM Tris– HCl buffer solution, pH 7.5. The concentrations of oxidized CaM-wt and CaM-mut were determined by using the Lowry method with the Bio-Rad DC protein assay kit using bovine serum albumin as the standard. Preparations of 60 mM oxidized CaM-wt and CaMmut were stored at 80 8C for 2 days without observing changes in stability of proteins.

Three human MsrB genes are known (MsrB1, MsrB2 and MsrB3) [45,46]. MsrB3A, in addition to MsrB3B, is one of the two existing isoforms of MsrB3 [46]. The cloning, expression and purification of human MsrB3A has been described previously by Kim and Gladyshev [46]. We followed the same procedures for production and purification of MsrB3A. The protein construct consists of residues 32–192 fused to a His-tag at the C-terminus. The signal peptide encoded residues 1–31 is not expressed. 2.9. Repair of oxidized CaM with MsrA/MsrB3A MsrA and MsrB were used to catalyze the DTT-dependent reduction of MetSO in CaM-wt and CaM-mut in order to assess the reversibility of the functional changes caused by Met oxidation [29,47,48]. Reactions contained 60 mM oxidized CaM-wt or oxidized CaM-mut, 10 mM DTT, 30 mM Tris–HCl, and Msr enzyme(s), at a final pH of 7.5. For reactions with only MsrA or MsrB, a concentration of 4 mM Msr enzyme was used. For reactions with both MsrA and MsrB, a concentration of 2 mM of each Msr enzyme was used. The high concentration of Msr enzymes was chosen to ensure a complete reduction of MetSO to Met. The reactions were incubated for 1 h at 37 8C and stopped by freezing the samples at 80 8C.

2.5. SDS-PAGE analyses 2.10. AC activity assay Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was used for a visual, qualitative assessment of the degree of Met oxidation in CaM proteins. The experiments were performed essentially as described previously [42,43]. A 4.5% (w/v) acrylamide stacking gel and a 15% (w/v) acrylamide resolving gel were used. CaM samples were dissolved in sample buffer (48% (w/ v) urea, 2.5% (w/v) SDS, 3% (w/v) DTT, 5% (v/v) glycerol, 0.01% (w/v) bromophenol blue and 25 mM Tris–HCl, pH 8.0) and were heated for 10 min at 95 8C before loading onto the gel. Either 1 mg or 2 mg CaM samples were loaded onto the gel. The electrophoresis was performed at 15 mA. The proteins in the gels were visualized by staining with 0.1% (w/v) Coomassie Brilliant Blue G 250 in 20% (v/v) methanol and 10% acetic acid (v/v) for 1 h at 70 8C. The gels were destained with 7.5% (v/v) acetic acid and 5% (v/v) methanol for 6 h at 70 8C. Page RulerTM Plus Prestained Protein #SM1811 and #0679, both Fermentas as used as standard. 2.6. Mass spectrometry analysis Proteins (50 pmol/ml, 20 ml, acetonitrile/methanol/water/ acetic acid mixture; 50:25:24:1, v/v) were electrosprayed by a robotic nanoflow ion source TriVersa NanoMate (Advion BioSciences, Ithaca, NY) using nanoelectrospray chips (1.5 kV ionization voltage, 1.3 psi backpressure) in to a LTQ Orbitrap XL ETD mass spectrometer (Thermo Fischer Scientific GmbH, Bremen, Germany). The temperature of the transfer capillary was set to 200 8C and the tube lens voltage to 115 V. Mass spectra were recorded from m/z 400 to 2000 in the orbitrap mass analyzer at a mass resolution of 100 000 at m/z 400. Acquired data was analyzed by using Xcalibur software (version 2.0.7).

The adenylyl cyclase (AC) catalytic activity of AC1 in Sf9 insect cell membranes was determined essentially as previously described [27,38]. Sf9 membranes expressing AC1 were resuspended using syringes at 4 8C and diluted with binding buffer (75 mM Tris– HCl, 12.5 mM MgCl2 and 1 mM EDTA) to a final protein concentration of 1 mg/ml and a final pH of 7.4. Samples containing 20 mg of protein were pre-incubated for 2 min at 30 8C with 10 ml of 30 mM Tris–HCl, pH 7.5 (Figs. 4 and 5) or 2 mM DTT (Figs. 6 and 7) for determining the basal activity of AC1 or AC activity, which was activated by CaM (10 ml of native, oxidized or MsrA/MsrB3Atreated oxidized CaM-wt or CaM-mut). Msr enzymes and DTT were left in CaM samples following their reaction with oxidized CaM samples and prior to the performance of the AC activity assay. Because of the increase of basal AC1 activity by DTT (data not shown), all CaM samples used for concentration–response curves of native and oxidized CaM-wt and CaM-mut proteins were complemented with DTT with a final concentration of 2 mM (this DTT addition allowed a valid comparison of the resulting AC activities to the Msr-treated samples). The AC activities in Figs. 2 and 3 were determined in the absence of DTT because a comparison to the Msr-treated samples was not necessary. Reactions were initiated by adding 20 ml of reaction mixture consisting of 100 mM cAMP, 40 mM ATP, 100 mM 3-isobutyl-1methylxanthine, 9 mM phosphocreatine, 0.4 mg/ml creatine kinase, 100 mM EGTA, and 0.25 mCi [a-32P]ATP to the samples. Additionally, samples contained 5 mM MgCl2, 0.4 mM EDTA, 30 mM Tris–HCl, pH 7.4, and various concentrations of free Ca2+. In general, a concentration of 10 mM of free Ca2+ was used, except for the assays shown in Fig. 3 that were performed either

200

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

100 millidegrees (mdeg), a bandwidth of 1 nm and five accumulations/spectrum at 25 8C. Samples contained 5 mM (oxidized) CaMwt or (oxidized) CaM-mut dissolved in deionized water and 0.1 mM CaCl2. Deionized water was chosen as the solvent because of its low absorbance and minimal interference with the spectrum of the protein in the examined measurement range. For these experiments, for the oxidized CaM-wt and CaM-mut samples, all Met were oxidized by using 50 mM H2O2 as described above. The a-helical content of the proteins was calculated by using the algorithm of the neural network K2D3 (http://k2d3.ogic.ca) [49]. The molar ellipticities [(u)  103 in degrees 2 1 (deg) cm dmol ] of the far-UV CD spectra were calculated after subtracting the spectrum of the solvent. The spectra were plotted and displayed using Prism 5.04 (GraphPad Prism, San Diego, CA). 2.12. Statistical analysis

Fig. 2. Analysis of the oxidation state of methionines in CaM-wt oxidized with different concentrations of H2O2 by the electrophoretic mobilities. Oxidation of CaM-wt with 0.05 mM, 0.5 mM, 5 mM or 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and the analysis by SDS-PAGE [15% (w/v) acrylamide] were performed as described in Sections 2.4 and 2.5. A concentration of 1 mg per CaM sample was loaded on the gel. The gel was stained with Coomassie Brilliant Blue G250 to detect the samples. The protein standard (Page RulerTM Plus Prestained Protein #SM1811, Fermentas, Thermo Scientific) shows bands at 11, 17 and 28 kDa in the area where the CaM samples were detected. The other lanes show CaM-wt and CaM-wt oxidized with 0.05 mM, 0.5 mM, 5 mM and 50 mM H2O2. Kilodalton, kDa.

without free Ca2+ or with free Ca2+-concentrations of 10 mM, 25 mM, 50 mM, 75 mM and 100 mM. EGTA was used to keep the concentration of free Ca2+ constant. Free Ca2+-concentrations, with consideration of buffer components, pH and temperature, were calculated using WebMax C standard (http://www.stanford.edu/ cpatton/webmaxcE.htm). Reactions were conducted for 20 min at 30 8C and terminated by adding 20 ml of 2.2 N HCl. Denatured protein was sedimented using a centrifugation for 1 min at 12 000  g. Separation of the product [32P]cAMP from the educt [a-32P]ATP was performed by transferring the samples onto columns filled with 1.3 g of aluminium oxide (MP Alumina N Super I, MP Biomedicals, Eschwege, Germany). [32P]cAMP was eluted from columns with 4 ml of 0.1 M ammonium acetate, pH 7.0. The ˇ erenkov radiation. concentration of [32P]cAMP was measured by C The AC activities were calculated after subtracting the blank value, which was determined by transferring only the amount of total added [a-32P]ATP on the columns with reference to the amount of total added [a-32P]ATP. Blank values were 0.03% and the turnover of substrate was 5% of the amount of total added [a-32P]ATP. 2.11. Circular dichroism spectroscopy CD spectra were obtained using a Jasco J-810 spectropolarimeter (Jasco, Tokyo, Japan) and a 1 mm path length cuvette. Far-UV spectra were collected from 190 to 240 nm with a scanning speed of 100 nm/min, a response of 1 s, a sensitivity standard of

The concentration–response curves of AC activity following stimulation of ACs with CaM were analyzed by non-linear regression (variable slope) using Prism 5.04 (GraphPad Prism, San Diego, CA). The statistical analyses were also performed by using Prism 5.04. The effects of different Ca2+-concentrations on AC1 activity stimulated by 10 mM native or oxidized CaM-wt were compared with one-way analysis of variances with Dunnett’s multiple comparison post-test using AC1 activity with 10 mM free Ca2+ as control. In order to determine significant differences of AC1 activity stimulated by native CaM-mut in comparison to AC1 activity stimulated by native CaM-wt, the mean AC1 activity with native CaM-wt was set to 100% and with 2 mM DTT to 0% and the absolute AC1 activities stimulated by native CaM-mut were normalized to these values in each experiment. The stimulatory effects of native CaM-wt and CaM-mut on the AC activity of AC1 at 10 mM were compared with one-way analysis of variances with Dunnett’s multiple comparison post-test using pooled normalized data of at least three independent experiments performed in duplicates and native CaM-wt as control. In order to determine significant differences of AC1 activity stimulated by oxidized or MsrA-treated oxidized CaM in comparison to AC1 activity stimulated by native CaM), the mean AC1 activity with each native CaM was set to 100% and with 2 mM DTT to 0% and the absolute AC1 activities stimulated by oxidized or MsrA-treated oxidized CaM were normalized to these values in each experiment. The normalization on each native CaM was necessary because of the differences in AC1 stimulation by each native CaM-mut compared to AC1 stimulation by native CaM-wt. Thus, the normalization of AC1 activities on each native CaM allows interpretation of the effects of site-specific Met oxidation and reduction on AC1 activity. The stimulatory effects of native, oxidized and MsrA-treated oxidized CaM on the AC activity of AC1 at 10 mM were compared with one-way analysis of variances with Dunnett’s multiple comparison post-test using pooled normalized data of at least three independent experiments performed in duplicates and each native CaM as control. The a-helical content of (oxidized) CaM-wt or (oxidized) CaM-mut was compared with one-way analysis of variances with Bonferroni’s multiple comparison post-test. 3. Results 3.1. Analysis of the oxidation state of native and oxidized CaM samples by the electrophoretic mobility on a polyacrylamide gel and by mass spectrometry SDS-PAGE is a simple and convenient method to establish qualitatively the oxidation state of CaM. Oxidation of CaM results in a loss of electrophoretic mobility, as demonstrated previously

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

201

Fig. 3. Mass spectrometry analysis of native and oxidized CaM-wt. The oxidation of CaM-wt with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and mass spectrometry analysis were performed as described in Sections 2.4 and 2.6, A, native CaM-wt; B, oxidized CaM-wt.

[21,26,50]. SDS-PAGE analysis of the degree of oxidation of CaM-wt oxidized with different concentrations of H2O2 (Fig. 2) shows identical bands for native CaM-wt and CaM-wt oxidized with 0.05 mM H2O2 demonstrating no measurable Met oxidation by this concentration of H2O2. It should be noted that, in each case, two bands are observed. These are due to the well-know tendency of CaM to lose bound Ca2+ during electrophoresis, with the Ca2+ free form displaying a reduced electrophoretic mobility [51]. This

behaviour is also observed in the results for native and oxidized CaM-mut. Under conditions (50 mM H2O2) known to promote complete oxidation of all Met in CaM, a single band, at a much higher apparent mass, is observed. The same result is obtained when oxidation is performed with 5 mM H2O2. For CaM-wt oxidized with 0.5 mM H2O2 several bands with different electrophoretic mobilities, indicating different oxiforms of CaM, appear between bands of native CaM-wt and CaM-wt oxidized with

202

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

50 mM H2O2. This indicates that, under these conditions, a heterogeneous population of partially oxidized CaM species is produced. The mass spectrometry analysis shows a major peak at 17 585 Da for native CaM-wt and at 17 729 Da for CaM-wt oxidized with 50 mM H2O2 (Fig. 3). The mass difference corresponds exactly to the addition of one oxygen, with a molecular mass of 16 Da, to each of the nine Met in CaM-wt. Thus, all Met were completely oxidized under these conditions, as was demonstrated previously [52]. The results also indicate that no modifications other than Met oxidation were generated by the conditions used for the Met oxidation. As result of a 13C- and 15N-labelling of CaM-wt, the molecular mass of native CaM-wt (17 585 Da) was higher than the molecular mass of CaM at natural isotopic abundance (16 723 Da). Thus, for our studies requiring complete oxidation of all Met in CaM-wt and CaM-mut, we used 50 mM H2O2. SDS-PAGE analysis of the oxidation state of native and oxidized CaM-wt and CaM-mut shows the expected reduction of the electrophoretic mobility after oxidation for CaM-wt and all CaMmut, except for L9-CaM (data not shown). As demonstrated previously, the degree of electrophoretic mobility loss is dependent on the number of oxidized Met [21,26,50]. Thus, the greatest loss of electrophoretic mobility was observed for oxidized CaM-wt and L2-CaM because of the oxidation of nine or seven Met, respectively. Accordingly, the loss of electrophoretic mobility was smaller for CaM-mut with one, two or three oxidized Met. L9-CaM, containing no oxidizable Met, showed identical electrophoretic mobilities before and after oxidation, as expected. This finding also corroborates the notion that the observed changes of the electrophoretic mobility are due to Met oxidation. 3.2. CaM activation of AC1 is regulated by the degree of Met oxidation in CaM Oxidation of Met in CaM decreases the ability of CaM to activate the AC activity of AC1 (Fig. 4). The response curves show the strong activation of AC1 with increasing CaM-wt concentration. No saturation of AC1 activity was reached with CaM-wt up to 10 mM, also observed in a previous study analyzing CaM-stimulated AC activity in rat cerebellar membranes [53]. These data are indicative for a low-affinity protein–protein interaction. Because a CaM

concentration of 10 mM is in accordance with the physiological CaM concentration in the vicinity of cell membranes [54], we defined AC1 activities determined at a CaM concentration of 10 mM as reference values for subsequent mutant and oxidation studies, although Vmax was not obtained at this concentration. Thus, with respect to protein concentrations, we analyzed CaM-wt and CaM-mut under clearly defined and comparable experimental conditions. When CaM-wt is oxidized with 0.05 mM and 0.5 mM H2O2, the changes in AC1 activation are nominal. When CaM-wt is oxidized by 5 mM H2O2, there is a substantial decrease in the ability of CaM to activate AC1; at the highest concentration of CaM (10 mM), the activation is only half that observed for activation by CaM-wt (not oxidized). No stimulation of AC1 by CaM-wt oxidized with 50 mM H2O2 was obtained at any CaM concentration. It is worthwhile to note that CaM-wt oxidized with 0.5 mM H2O2 activated AC1 as well as CaM-wt that had not been oxidized, even though partial oxidation was observed by SDS-PAGE analysis (Fig. 2). This suggests that some Met, when oxidized, do not affect significantly CaM activation of AC1. In addition, although the SDS-PAGE analysis indicated similar degrees of CaM-wt oxidation using 5 mM and 50 mM H2O2, CaM-wt oxidized with 5 mM H2O2 was still able to activate AC1, whereas CaM-wt oxidized with 50 mM H2O2 was not. These results demonstrate the inherent limits in the sensitivity of the SDS-PAGE method for analyzing the degree of oxidation, as low concentrations of some oxiforms are not observed. 3.3. Ca2+ dependence of AC1 activation by CaM The results in Fig. 5 demonstrate the dependence of AC1 activation by CaM and oxidized CaM on the concentration of Ca2+. The native CaM-stimulated AC1 activity without added Ca2+ is very low. Adding just 10 mM free Ca2+ resulted in a four-fold increase of AC1 activity. Increasing the free Ca2+ concentration above 10 mM resulted in a gradual decrease in activation. The results with fully oxidized CaM were similar, except that the increase in activation with 10 mM free Ca2+ relative to the activity with no added Ca2+ was nominal. These findings suggest that the decreased AC1 activation observed with oxidized CaM-wt is not related to a decreased affinity of CaM to Ca2+ upon oxidation [55], but rather to a direct inhibition of AC1 by Ca2+ due to the experimentally unavoidable competition between Mg2+ (the cation essential for the catalytic activity of AC1) and Ca2+ at the catalytic site [38,56]. Based on these results, we used a free Ca2+concentration of 10 mM in the AC activity assays for subsequent studies analyzing CaM activation of AC1 to eliminate the direct inhibition of AC1 by higher Ca2+ concentrations and to evaluate oxidation effects on AC1 activation. Thus, also with respect to Ca2+ concentration, we analyzed CaM-wt and CaM-mut under clearly defined and comparable experimental conditions. 3.4. Activation of AC1 by CaM mutants with Leu for Met substitutions

Fig. 4. Regulation of AC1 by the degree of Met oxidation in CaM. Met oxidation with 0.05 mM, 0.5 mM, 5 mM and 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and the AC activity assay was performed as described in Sections 2.4 and 2.10. Oxidized samples of CaM-wt were dissolved in deionized water without DTT. The concentrations of (oxidized) CaM-wt were varied from 10 nM to 10 mM. Concentration–response curves of native CaM-wt (*) and oxidized CaM-wt (0.05 mM H2O2 (&), 0.5 mM H2O2 (~), 5 mM H2O2 (!) and 50 mM H2O2 (^)) were analyzed by nonlinear regression (variable slope) using Prism 5.04. The AC activities of AC1 with 30 mM Tris–HCl, pH 7.5 was set to 0% and with 10 mM native CaM-wt to 100%. The AC activities show the means  SD of three independent experiments performed in duplicates.

In order to evaluate how oxidation of specific Met residues in CaM regulates CaM activation of AC1, a series of thirteen mutant CaM proteins with Met replaced by Leu were produced and employed (Table 1). It was therefore necessary first to analyze the impact of Met to Leu substitutions in CaM-mut on AC1 activation. Concentration–response curves of CaM-wt and CaM-mut were recorded (data not shown). Table 2 shows the activation of AC1 by these CaM-mut relative to CaM-wt. For AC1 activation by CaM-mut L9, M36/L8, M144/L8, M145/L8, M71,M72,M76/L6, M109,M124/L7 and L2 significantly higher AC activities (20–70%) were observed relative to activation by CaM-wt. The highest AC activities, 60–70% higher than activation with CaM-wt, were obtained with activation by L9-CaM, M144/L8-CaM and M145/L8-CaM. For CaM-mut

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

203

Fig. 5. Effect of various Ca2+-concentrations on AC1 activity stimulated by native or oxidized CaM-wt. The calculation of free Ca2+-concentrations, the oxidation of CaM-wt with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and the AC activity assay were performed as described in Sections 2.4 and 2.10. A concentration of 10 mM of native CaMwt (A) or oxidized CaM-wt (B) and free Ca2+-concentrations of 0 mM (Ø), 25 mM, 50 mM, 75 mM and 100 mM were used. The AC activities show the means  SD of three independent experiments performed in duplicates. A one-way analysis of variances with a Dunnett’s multiple comparison post-test with native or oxidized CaM-stimulated AC activity and 10 mM Ca2+ as control was performed to detect significant differences of native or oxidized CaM-stimulated AC activities with various Ca2+-concentrations (**0.001 < pvalue < 0.01; ***p-value < 0.001). All calculations were performed using Prism 5.04.

M51/L8, M71/L8, M72/L8, M76/L8, and M36,M51/L7, no significant differences in AC1 activation relative to CaM-wt were obtained. The only mutant with a decreased ability to activate AC1 was the M144,M145/L7 mutant. With this exception, these results are consistent with our previous results analyzing AC1 activation by CaM-mutants L9, M144/L8, M145/L8 and M144,M145/L7 [27]. We suspect this is because, in the previous work, the Sf9 cell membranes were not washed with EGTA during membrane preparation to remove endogenous Ca2+-bound CaM, which could influence the stimulation of AC1 by exogenous CaM [40]. In this project, we washed Sf9 cell membranes with EGTA during membrane preparation following a standard protocol described in Section 2.2. Despite washing with EGTA, it is possible that some endogenous CaM remains at the Sf9 cell membranes expressing AC1. Using the same washing method for all membrane preparations, we observed highly reproducible AC1 activities and stimulations in the AC activity Table 2 Activity of AC1 with CaM-wt/CaM-mut.a CaM

AC activity with 10 mM CaM-wt/CaM-mut [%]

wt L9 M36/L8 M51/L8 M71/L8 M72/L8 M76/L8 M144/L8 M145/L8 M36,M51/L7 M71,M72,M76/L6 M109,M124/L7 M144,M145/L7 L2

100.0  0.0 168.9  27.2*** 131.2  14.4** 98.5  4.9 98.9  10.1 97.5  10.0 111.6  2.7 161.0  14.3*** 168.1  4.4*** 97.7  11.0 126.2  7.3* 122.7  15.0* 81.1  4.8 145.1  21.9***

a AC activities with 10 mM CaM-wt/CaM-mut were collected from corresponding concentration–response curves (data not shown) and were determined as described in Section 2.10. Concentration–response curves were analyzed by nonlinear regression (variable slope). The mean of AC activities of AC1 with 2 mM DTT was set to 0% and with 10 mM CaM-wt to 100%. Absolute AC1 activities were normalized on these values. The AC activities show the means  SD of at least three independent experiments performed in duplicates. A one-way analysis of variances with a Dunnett’s multiple comparison post-test using the normalized data with CaM-wt as control was performed to detect significant differences of AC activities with 10 mM CaM-mut in comparison to CaM-wt (no *: p-value > 0.05; *: 0.01 < pvalue < 0.05; **: 0.001 < p-value < 0.01; ***: p-value < 0.001). All calculations were performed using Prism 5.04.

assays, documented in the figures and tables of this report. Based on these data, we conclude that the amount of endogenous CaM in the various membrane preparations was probably similar and that a valid comparison of data following stimulation of AC1 with exogenous CaM is possible. 3.5. Selective oxidation of N- and C-terminal Met in CaM regulates AC1 activation The CaM-mut with Leu substituted for Met were used to examine how oxidation of specific Met in CaM affects CaM activation of AC1. The summarized results of the determination of concentration–response curves for these experiments are shown in Table 3 and Fig. 6. In order to evaluate the effect of site-specific Met oxidation, AC1 activity determined with 10 mM of each native CaM-mut was set to 100% to eliminate increases of AC1 activity evoked by Met to Leu substitutions. For these experiments, the reaction solutions included 2 mM DTT. DTT stimulates AC1 (data not shown), but the mechanism of this stimulation is not known. DTT was included to allow the results of these experiments to be compared directly to results of subsequent experiments (see below) using DTT as the reductant for Msr-catalyzed reduction of MetSO in CaM. As shown in Table 3 and Fig. 6, oxidation of all Met in CaM-wt eliminates all AC1 activation by CaM, and in fact, decreases the AC1 activity below 0%. Most likely, there is a very small amount of endogenous CaM bound to AC1 in the membranes that is displaced by the much larger amount of added oxidized CaM. This is not unlikely, as similar findings were made for PMCA, where oxidized CaM is also able to bind to PMCA, but in a non-productive manner [57]. To clarify this point for AC1 more detailed, further affinity and binding studies are required. No activity decrease is observed, as expected, for the control CaM (L9) where all Met are replaced by Leu. The remaining results indicate that AC1 activation by CaM is regulated by oxidation of both N- and C-terminal residues. In the N-terminal domain, oxidation of M36 (in M36/L8) or M51 (in M51/ L8) results in decreases in activation of 15% and 13%, respectively. When both are oxidized (in M36,M51/L7), a decrease of about 37% is observed. No significant decrease in activation is observed for oxidation of M71 or M72, but a small decrease is observed for oxidation of M76 (7%). When all three are oxidized (in M71,M72, M76/L6), a 37% decrease in activation results. In the C-terminal domain, oxidation of M144 (in M144/L8) or M145 (in M145/L8) results in decreases in activation of 30% and 22%, respectively.

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

204

Table 3 Activity of AC1 with native, oxidized and MsrA-treated oxidized CaM-wt/CaMmut.a CaM

wt L9 M36/L8 M51/L8 M71/L8 M72/L8 M76/L8 M144/L8 M145/L8 M36,M51/L7 M71,M72,M76/L6 M109,M124/L7 M144,M145/L7 L2

AC activity with 10 mM CaM-wt/CaM-mut [%] Native

Oxidized

MsrA-treated

100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0 100.0  0.0

8.0  4.0*** 101.5  9.2 84.6  4.1*** 87.0  9.8 98.1  17.4 96.3  5.2 92.5  6.4 70.3  4.5*** 78.2  8.5** 62.9  7.5*** 78.9  2.1*** 74.8  13.4* 87.5  13.9 25.1  5.0***

61.8  9.3*** 94.1  5.1 88.9  6.6** 94.0  3.5 101.8  15.7 103.3  10.0 92.3  14.1 82.8  3.0*** 93.2  3.6 76.5  8.8** 95.0  2.6* 97.8  6.0 110.2  23.4 76.4  4.4***

a AC activities with 10 mM CaM-sample were collected from the corresponding concentration–response curves (data not shown). Met oxidation with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C, the reaction of oxidized CaM-wt and CaM-mut with MsrA and the AC activity assay were performed as described in Sections 2.4, 2.9 and 2.10. Concentration–response curves were analyzed by nonlinear regression (variable slope). The mean of AC activities of AC1 with 2 mM DTT was set to 0% and with 10 mM native CaM-wt or each native CaM-mut to 100%. Absolute AC1 activities were normalized on these values. The AC activities show the means  SD of at least three independent experiments performed in duplicates. A oneway analysis of variances with a Dunnett’s multiple comparison post-test using the normalized data with native CaM-wt or each corresponding native CaM-mut as control was performed to detect significant differences of AC activities with 10 mM oxidized or with MsrA-treated oxidized CaM-wt/CaM-mut in comparison to native CaM-wt/CaMmut (no *: p-value > 0.05; *: 0.01 < p-value < 0.05; **: 0.001 < p-value < 0.01; ***: pvalue < 0.001). All calculations were performed using Prism 5.04.

Oxidizing both (in M144,M145/L7) results in a surprising smaller decrease of only 13%. Oxidation of M109 and M124 (in M109,M124/L7) gives a decrease of 25%. Finally, oxidation of a mutant protein (L2) with Leu substitutions for only M144 and

M145 results in a large 75% activation decrease, as would be expected based on the aforementioned results. 3.6. Recovery of CaM activation of AC1 by Msr catalyzed MetSO reduction The catalytic MetSO reduction activities of MsrA and MsrB provide an avenue for reversible redox regulation of CaM activation of AC1. As shown in Table 3 and Fig. 7, although oxidation of all Met in CaM eliminates its ability to activate AC1, the combined activities of MsrA and MsrB can restore the ability to native levels. Neither MsrA nor MsrB alone can restore all activity, indicating oxidation produces both MetSO stereoisomers. The catalytic activities of MsrA and MsrB do not change the ability of L9-CaM to activate AC1, as expected. The oxidized mutant CaM species (Table 3) show varying degrees of recovery of activity when reduced with DTT catalyzed by MsrA. It is possible that some irreversible changes in the structure of a particular mutant protein accompany oxidation, and these cannot be reversed by reduction. However, the results in Fig. 7 with CaM-wt suggest that the failure to regain all lost activity is most likely due to a stereoselectivity in the initial oxidation for a particular site. This is not unprecendented or unexpected, given the chiral nature of proteins. 3.7. Secondary structure changes in CaM accompanying Met oxidation and Leu substitution CD spectroscopy was used to investigate the impact of Met to Leu substitutions and the oxidation of Met on the secondary structure of CaM-wt and CaM-mut. Table 4 shows the a-helical content of native and oxidized CaM-wt and CaM-mut calculated from the far-UV CD spectra. The spectra of CaM-wt and M36/L8CaM are shown in Fig. 8. The spectra of the other CaM-mut are not shown. The CD spectra of L2-CaM are similar to the CD spectra of CaM-wt and the CD spectra of the remaining CaM-mut are similar

Fig. 6. Effect of oxidation of specific Met in CaM on CaM activation of AC1. AC activities with 10 mM CaM-wt or CaM-mut were collected from corresponding concentration– response curves (data not shown). Met oxidation with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and the AC activity assay were performed as described in Sections 2.4 and 2.10. Concentration–response curves were analyzed by nonlinear regression (variable slope). The mean of AC activities of AC1 with 2 mM DTT was set to 0% and with 10 mM native CaM-wt or each native CaM-mut to 100%. The AC activities show the means  SD of at least three independent experiments performed in duplicates. A one-way analysis of variances with a Dunnett’s multiple comparison post-test using the normalized data with native CaM-wt or each corresponding native CaM-mut as control was performed to detect significant differences of AC activities with 10 mM oxidized CaM-wt or CaM-mut in comparison to native CaM-wt or CaM-mut (no *p-value > 0.05; *0.01 < pvalue < 0.05; **0.001 < p-value < 0.01; ***p-value < 0.001). All calculations were performed using Prism 5.04.

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

205

Fig. 7. Restoration of CaM activation of AC1 by reduction of MetSO in oxidized CaM catalyzed by MsrA and MsrB. Met oxidation with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C, the treatment of oxidized CaM-wt and L9-CaM with MsrA or/and MsrB3A and the AC activity assay were performed as described in Sections 2.4, 2.9 and 2.10. The concentrations of different forms of CaM-wt and L9-CaM were varied from 0.3 mM to 10 mM. The concentration–response curves of native (*), oxidized (&), MsrA- (~), MsrB3A- (!) and MsrA/MsrB3A-treated (^) CaM-wt (A) and L9-CaM (B) were analyzed by nonlinear regression (variable slope) using Prism 5.04. The mean of AC activity of AC1 with 2 mM DTT was set to 0% and with 10 mM native CaM-wt/native L9-CaM to 100%. The AC activities show the means  SD of at least three independent experiments performed in duplicates.

to the CD spectra of M36/L8-CaM. In general, the CD spectra show minima at 206 nm and 222 nm and a maximum at 193 nm, which are typical for proteins with a high a-helical content (Fig. 8) [58]. The a-helical content of native CaM-wt was 57  8% and for CaM-mut it ranged from 51  4% to 62  8%. The a-helical content of oxidized CaM-wt was 45  2% and the content of oxidized CaM-mut ranged from 45  4% to 57  1%. There were significant differences in the a-helical content of oxidized CaM in comparison to native CaM only in the case of CaM-wt and L2-CaM. No significant differences between native and oxidized forms were obtained for the other CaMmut as well as for the a-helical content of native CaM-mut in comparison to CaM-wt (Table 4).

Table 4 a-helical content of native and oxidized CaM-wt and CaM-mut.a CaM

Native [%]

Oxidized [%]

wt L9 M36/L8 M51/L8 M71/L8 M72/L8 M76/L8 M144/L8 M145/L8 M36,M51/L7 M71,M72,M76/L6 M109,M124/L7 M144,M145/L7 L2

57  8 61  8 54  10 53  1 58  4 51  3 53  2 58  6 62  8 52  4 51  4 53  2 57  7 59  9

45  2** 55  6 55  5 53  6 56  6 53  2 57  1 56  6 54  6 47  3 49  1 49  2 52  5 45  4*

a The a-helical content of native and oxidized CaM-wt and CaM-mut was calculated using the algorithm of K2D3 (http://k2d3.ogic.ca) as described in Section 2.11 from the corresponding far-UV CD spectra shown in Fig. 6 (CaM-wt and M36/ L8-CaM). Corresponding far-UV CD spectra of other CaM-mut are not shown. Met oxidation with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C and the measurements of CD spectra were performed as described in Sections 2.4 and 2.11. The data for a-helical contents are the means  SD of three independent measurements (five accumulations/measurement) of three independent samples of native or oxidized CaM-wt or CaM-mut. A one-way analysis of variances with a Bonferroni’s multiple comparison post-test was performed to detect significant differences of a-helical content on the one hand of CaM-mut in comparison to CaM-wt and on the other hand of oxidized form in comparison to the native form of CaM-wt and each CaM-mut (native form of CaM-wt/each CaM-mut as control: no *: p-value > 0.05; *: 0.01 < p-value < 0.05; **: 0.001 < p-value < 0.01). Generation of CD spectra, calculation of the means  SD and the statistical analysis were performed using the Jasco J-810 spectropolarimeter software and Prism 5.04.

4. Discussion Since the first detailed analysis of the CaM-AC1 interaction at the beginning of the 1990s [6], little progress has been made in our understanding of the regulation of AC1 by CaM [8,9,27,40,59,60]. This is due to formidable challenges regarding the analysis of AC1 expressed in cell membranes. First, CaM-AC1 interaction is a low-affinity protein–protein interaction because no saturation of AC1 activation is reached at physiological CaM concentrations (up to 10 mM) [54]. Second, an increase of the Ca2+ concentration to examine effects of changes in the affinity of oxidized CaM for Ca2+ on AC1 activation is highly problematic due to the competition of Mg2+ and Ca2+ at the catalytic site of AC1 and associated inhibitory effects of Ca2+ directly on the catalytic activity of the enzyme [38,56]. Third, endogenous CaM very tightly bound to cell membranes expressing AC1 may influence the activation by exogenous CaM [40]. Despite these very substantial above technical limitations, we took up the challenge to analyze AC1-CaM interaction biochemically because of the high (patho)physiological relevance of AC1 [35,36] and defined our assay conditions as far as possible to obtain a valid interpretation of our results. The regulation of AC1 activation by CaM with regard to Met residues important for this interaction was never studied before. An additional layer of regulation, and complexity, is added when the posttranslational modification of CaM, including reversible oxidation of Met in CaM, is considered. Our goal was to investigate for the very first time the impact of Met oxidation in CaM on the ability of CaM to activate AC1. We analyzed systematically the changes in AC1 activation using a large selection of thirteen CaM-mut with Met to Leu substitutions to permit site-specific Met oxidations by using H2O2 (Table 1). Eight of these CaM-mut proteins were never studied before at any CaM target. As evidenced by our results, oxidation of selected Met residues in CaM is able to finely tune the activation of AC1 by CaM. Oxidation of multiple Met in CaM is able to alter AC1 activation more dramatically. Much of this is summarized in Table 5, along with a comparison to results of similar studies with a CaMstimulated AC toxin from Bordetella pertussis, CyaA. 4.1. Conservative amino acid substitutions for Met in CaM The strategy of using Leu substitutions for Met in CaM to examine the structural, functional and physiological ramifications of oxidation of remaining Met residues is not new. These studies

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

206

Fig. 8. CD analysis of secondary structure changes in CaM accompanying Met oxidation. CaM-wt and M36/L8-CaM were oxidized with 50 mM H2O2 and 0.1 mM CaCl2 for 24 h at 25 8C. The far-UV CD spectra of native (—) and oxidized (- - -) CaM-wt (A) and M36/L8-CaM (B) were collected as described in Section 2.11. Briefly, the spectra for 5 mM (oxidized) CaM-wt and (oxidized) M36/L8-CaM with 0.1 mM CaCl2 were collected from 190 to 240 nm at 25 8C. The CD spectra are the results of three independent measurements (five accumulations/measurement) of three independent samples of native or oxidized CaM-wt or M36/L8-CaM. The CD spectra were generated using the Jasco J-810 spectropolarimeter software and Prism 5.04 as described in Section 2. The CD spectra of the other CaM-mut are not shown.

also lend themselves to examining the structural and functional consequences of the Leu substitutions themselves [23,26–28,61]. Studies specifically addressing the effects of Leu substitutions, and even conservative substitutions of unnatural amino acids including norleucine, ethionine and selenomethionine, for Met in CaM have been performed [25,62–64]. The overarching conclusions of these studies are largely consistent with our conclusions regarding CaM activation of AC1. Structurally, substitution of Leu, or other conservative substitutions, for Met in CaM does not result in global structural perturbations. Even when all Met are substituted for Leu the secondary structure is largely unchanged (Table 4). The same is true of the tertiary structure, as indicated by NMR spectra [26]. These mutant proteins purify normally using Ca2+ dependent binding to phenyl Sepharose and selective elution with a Ca2+ chelator, indicating substantially normal binding to Ca2+ and

phenyl Sepharose. The rationale for the structural resiliency is that the Met residues are largely solvent exposed, especially in the Ca2+loaded form, so conservative mutations are less likely to promote global structural changes, relative to mutations of residues in the hydrophobic cores of proteins [63]. Functionally, the substitution of Leu for Met in CaM results in unpredictable and highly site- and target protein-specific effects. CaM with all Met replaced with Leu (L9) activates AC1 nearly 70% better that CaM-wt (Table 2). However, when all Met except M144 and M145 are replaced with Leu (M144,M145/L7), a 20% decrease in AC1 activation results. This is in contrast to CaM activation of the PMCA, where activation by L9 CaM is 30% higher than by wild type CaM, but activation by M144,M145/L7 CaM is 50% higher [26]. Moreover, for activation of the AC toxin CyaA from B. pertussis, neither L9 nor M144,M145/L7 CaM activate as well as wild type (15% and 10% poorer activation,

Table 5 Relevance of specific Met in CaM for regulation of membranous AC1 and Bordetella pertussis AC toxin CyaA by oxidation to MetSO. Ma

Membranous AC1

Bordetella pertussis AC toxin CyaAb

36

High relevance: single Met in the N-terminal region of CaM enhances AC1 activation and MetSO impairs AC1 activation

No relevance: Met is accessible to oxidation within the CaM-CyaA complex; oxidation does not prevent binding to CyaA

51

High relevance: MetSO impairs AC1 activation

No relevance: Met is accessible to oxidation within the CaM-CyaA complex; oxidation does not prevent binding to CyaA

71

Only relevant in combination with oxidized M72 and M76: AC1 activity decreases significantly

No relevance: Met is accessible to oxidation within the CaM-CyaA complex; oxidation does not prevent binding to CyaA

72

Only relevant in combination with oxidized M71 and M76: AC1 activity decreases significantly

No relevance: Met is accessible to oxidation within the CaM-CyaA complex; oxidation does not prevent binding to CyaA

76

Only relevant in combination with oxidized M71 and M72: AC1 activity decreases significantly

No relevance: Met is accessible to oxidation within the CaM-CyaA complex; oxidation does not prevent binding to CyaA

109

High relevance in combination with oxidized M124: AC1 activity decreases significantly

High relevance: crystallographic studies reveal important hydrophobic interactions with CyaA; preservation of this Met from oxidation within the CaM-CyaA complex and prevention of binding to CyaA in oxidized state suggest an involvement in the formation of CaM-CyaA complex

124

High relevance in combination with oxidized M109: AC1 activity decreases significantly

High relevance: crystallographic studies reveal important hydrophobic interactions with CyaA; preservation of this Met from oxidation within the CaM-CyaA complex and prevention of binding to CyaA in oxidized state suggest an involvement in the formation of CaM-CyaA complex

144

High relevance: Met enhances AC1 activation and MetSO impairs AC1 activation; no altered AC1 activation in combination with a Met or MetSO at position 145

Moderate relevance: crystallographic studies reveal important hydrophobic interactions with CyaA; Met is accessible to oxidation within the CaM-CyaA complex; site-specific oxidation does not prevent binding to CyaA but in combination with other C-terminal Met oxidation prevents binding to CyaA

145

High relevance: Met enhances AC1 activation and MetSO impairs AC1 activation; no altered AC1 activation in combination with Met or MetSO at position 144

High relevance: crystallographic studies reveal important hydrophobic interactions with CyaA; preservation of this Met from oxidation within the CaM-CyaA complex and prevention of binding to CyaA in oxidized state suggest an involvement in the formation of CaM-CyaA complex; important for CyaA activation

a b

M, methionine. Based on experimental data shown in Refs. [27,50,75].

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

respectively) [27]. The situation is again different for phosphodiesterase (PDE) activation [25]. Single Leu substitutions for Met have no effect, except for M36L. This mutant has a very low affinity for PDE and is a very poor activator. Replacing all C-terminal Met with Leu also results in poor affinity, although maximal PDE activation is still achieved. These results collectively serve to illustrate the site- and target-dependence of the substitutions. Presumably, because Leu is somewhat bulkier and more hydrophobic than Met, and somewhat less flexible, these physical properties conspire to promote local changes in structure and dynamics that promote a specific effect for a given target. 4.2. Site-specific Met oxidation in CaM MetSO substitution for Met in CaM, by oxidation of existing Met residues, is less conservative than substitution of Leu or Met analogues. Nevertheless, neither oxidation of specific Met residues nor simultaneous oxidation of all nine Met residues promotes largescale changes in the secondary structure of CaM (Fig. 8 and Table 4), in accord with previous observations [21,57,61,65–67]. Again, this can be rationalized by the relatively high solvent exposure of most of the Met in calcium-loaded CaM; the more polar and hydrophilic MetSO is thus well tolerated, and the potential for structural destabilization is reduced. However, oxidation is not without structural consequence, as fluorescence, fluorescence quenching and NMR studies indicate significant tertiary structural perturbations result from Met oxidation that alter the structures of the globular domains of CaM and promote irregular and nonproductive interactions with target domains [21,57,61]. Met oxidation destabilizes helices, and in CaM oxidation of M144 and M145 is known to destabilize the C-terminal helix [21,57,61,68]. Met oxidation in CaM also alters coordination of the Ca2+ ions and can potentially reduce the affinity of CaM for Ca2+ [24,55,69]. Similar to conservative amino acid substitutions for Met in CaM, the functional effects of Met oxidation in CaM are highly site- and target protein-specific. For individual Met oxidations in CaM, oxidation of M144 caused the largest decrease in the ability of CaM to activate AC1 (Table 3). This is interesting, as the same effect was observed for CaM activation of the PMCA and for CaM activation of nNOS [23,26]. Because oxidation of M144 causes significant perturbations in the tertiary structure of the C-terminal domain, as discussed above, it would be tempting to suspect that this functional effect might be general. However, oxidation of M144 in CaM has no effect on the ability of CaM to activate eNOS [23]. Oxidation of the other C-terminal Met in CaM also causes significant decreases in the ability of CaM to activate AC1. This is in contrast with the results for PMCA activation, where nearly the entire effect is due to oxidation of M144 alone [26]. Also in contrast with PMCA activation is that AC1 activation by CaM is also regulated by oxidation of N-terminal Met. Oxidation of M36 causes a significant decrease in CaM activation of AC1. Oxidation of M51 also causes a substantial, if smaller, decrease. Oxidation of both results in a nearly 30% reduction in activation. Oxidation of the individual remaining N-terminal Met (M71, M72, M76) has little effect, but oxidation of all three decreased CaM activation of AC1 by more than 20%. These results also contrast the results of studies of CaM binding to the AC toxin CyaA from B. pertussis which showed that oxidation of N-terminal CaM Met did not alter the affinity of CaM for CyaA, but oxidation of C-terminal Met (M109, M124 and M145) substantially decreased the affinity [50]. Overall, for AC1, the effects of Met oxidation in CaM are cumulative, if not strictly additive, as CaM activation of AC1 is reduced by 75% when all the Met in the CaM-mut L2 are oxidized, and all ability to activate AC1 is lost when all nine Met are oxidized. Evidence suggests that oxidation of Met in CaM can promote changes in the ligands that coordinate the Ca2+ ions [55], changes

207

in cooperativity of Ca2+ binding, and, when many or all Met are oxidized, significant changes in Ca2+ binding affinity [21,69]. For CaM isolated from senescent rat brain, the observed changes in the Ca2+-dependence of PMCA activation by CaM were shown to be due to a reduced cooperativity between the high affinity Ca2+ binding sites, and not an overall decrease in Ca2+ affinity for CaM [21]. This CaM is a heterogeneous mixture of CaM species with one (65%), two (30%) or at most three (<10%) oxidized Met. The results therefore suggest that oxidation of a limited number of Met in CaM is unlikely to promote significant changes in Ca2+ affinity, and therefore observed functional effects are most likely due to altered interactions between the oxidized CaM and the target. This is important for our studies of AC1 activation, because the experiments cannot be performed with a large excess of Ca2+ that guarantees saturation (Fig. 4). This is due to inhibition of AC1 catalytic activity by competition from Ca2+ for the essential Mg2+ catalytic ion binding site [38,56]. The experiments to assess the ability of oxidized CaM species to activate AC1 were performed with 10 mM free CaCl2 concentrations. This is roughly the concentration of free Ca2+ in the cell attained during typical Ca2+ spikes [70]. However, given the results noted above, it is very unlikely that the decreased ability of CaM species with singly, or perhaps doubly or triply, oxidized Met results simply from a reduced Ca2+ affinity and correspondingly reduced ability to bind to AC1. Most likely, the results indicate that structural changes in CaM accompanying Met oxidation promote suboptimal interactions with AC1 that are less likely to induce activation. 4.3. Repair of oxidized Met in CaM by MsrA and MsrB The Msr enzymes are critical components for regulation of protein function by Met oxidation, as they reverse the regulatory restraints imposed by oxidation. Reversible regulation by this mechanism also requires that the structural effects of oxidation are not irreversible, and that the oxidized Met are accessible to the enzymes. Interestingly, Msr catalyzed, stereospecific Met oxidation in proteins (including CaM) has also been demonstrated recently, suggesting perhaps a dual regulatory role for Msr enzymes [71]. Many previous studies have examined reduction of MetSO in CaM catalyzed by Msr enzymes [19,34,50,52,72,73]. Most notably, for CaM with all nine Met oxidized, using eukaryotic Msr enzymes and at physiological temperature (37 8C), a mixture of MsrA and MsrB enzymes readily catalyzes complete reduction of all MetSO to Met, as confirmed by mass spectrometry and SDS-PAGE analyses [50]. Under these same conditions, our results demonstrate complete restoration of CaM activation of AC1, as expected (Fig. 7). Using only either MsrA or MsrB, activation returned to approximately one-half the level achieved with native (fully reduced) CaM, which might be expected given the diastereoselectivity of the Msr enzymes. The results for reduction of MetSO by MsrA in the oxidized mutant proteins essentially mirror those with the fully oxidized CaM-wt. Where the error limits permit interpretation, essentially MsrA treatment resulted in partial, but not full, restoration of CaM activation of AC1. This again is consistent with the diastereoselectivity of MsrA reduction. Given the sensitivity of AC1 activation to oxidation of many of the Met in CaM, in both of the globular domains of CaM, and given that these can be reduced by MsrA and MsrB, then this suggests a rich, intricate regulation of AC1 by reversible Met oxidation. 4.4. Pathological and therapeutical implications of results The impact of (oxidized) Met in CaM on AC1 activation is of great importance in the field of neurodegenerative diseases such as Alzheimer’s disease (AD). A significant decrease in the protein level

208

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209

and the AC activity of AC1 in parietal cortex membranes and in the hippocampus of brains isolated from patients with AD were verified suggesting an involvement of the AC1-cAMP system in impaired processes of memory and learning in AD [35,36]. Indeed, to date it is unknown which signalling pathways are responsible for decreased AC1 activity [36]. Our data suggest that prevention of AC1 activation by oxidized CaM, resulting from increased oxidative stress by ROS during ageing, could be associated with impaired memory and learning in AD. Furthermore, our results demonstrate that MsrA and MsrB can reduce MetSO in CaM to restore AC1 activation (Fig. 7). In combination with previous findings describing the induction of Msr activity by pergolide, pergolide sulfoxide and (S)-adenosyl-Met in neuronal cells [74], this could represent a possible therapeutical approach to enhance AC1 activity. Future studies will investigate AC1 and Msr activity in oxidatively stressed neuronal cells. It will also be important to study the impact of thioredoxin and thioredoxin reductase on AC1. Thus, the ability of Met oxidation to regulate CaM activation of AC1, and the ability to finely tune regulation by Msr activity, represents not only an elegant mechanism but also a therapeutic avenue for addressing declining cognitive function associated with neurodegenerative diseases. Although systematic Met oxidation mapping allows for analysis of CaM-AC1 interaction, a crucial next step to evaluate the CaMAC1 interaction more detailed is to crystallize the complex of CaM and AC1.

Source of funding This work was supported by a grant from the Deutsche Forschungsgemeinschaft (Se 529/5-2) to R. S. and by funds from the Vice President for Research and from the Office of the Provost at the University of Georgia to J.U. and R.B.U. Acknowledgments The authors thank Dr. A. Rummel (Institute of Toxicology, Hannover Medical School) for advice with the CD measurements, Prof. Dr. R. Gerhard (Institute of Toxicology, Hannover Medical School) for helpful discussions and Mrs. Juliane von der Ohe for expert technical assistance. We thank Henry Niedermaier for assistance purifying of many of the mutant calmodulin proteins. We thank Prof. Vadim Gladyshev for the kind gift of the expression vector for human MsrB3A. The research utilized the resources of the Proteomics an Mass Spectrometry Core facility at the University of Georgia and the expertise of its director and staff. We thank Prof. Ralf Hoffmann (Institute of Bioanalytical Chemistry, University of Leipzig) for providing access to ESI-LTQ-Orbitrap mass spectrometer. Lastly, thanks are due to the reviewers for their helpful suggestions. References [1] Okumura S, Kawabe J, Yatani A, Takagi G, Lee MC, Hong C, et al. Type 5 adenylyl cyclase disruption alters not only sympathetic but also parasympathetic and calcium-mediated cardiac regulation. Circ Res 2003;93:364–71. [2] Post SR, Hammond HK, Insel PA. b-Adrenergic receptors and receptor signaling in heart failure. Annu Rev Pharmacol Toxicol 1999;39:343–60. [3] Andersson R, Nilsson K. Cyclic AMP and calcium in relaxation in intestinal smooth muscle. Nat New Biol 1972;238:119–20. [4] Wong ST, Trinh K, Hacker B, Chan GC, Lowe G, Gaggar A, et al. Disruption of the type III adenylyl cyclase gene leads to peripheral and behavioral anosmia in transgenic mice. Neuron 2000;27:487–97. [5] Sadana R, Dessauer CW. Physiological roles for G protein-regulated adenylyl cyclase isoforms: insights from knockout and overexpression studies. Neurosignals 2009;17:5–22. [6] Tang W-J, Krupinski J, Gilman AG. Expression and characterization of calmodulin-activated (type I) adenylylcyclase. J Biol Chem 1991;266:8595–603.

[7] Cali JJ, Zwaagstra JC, Mons N, Cooper DM, Krupinski J. Type VIII adenylyl cyclase. A Ca2+/calmodulin-stimulated enzyme expressed in discrete regions of rat brain. J Biol Chem 1994;269:12190–95. [8] Vorherr T, Kno¨pfel L, Hofmann F, Mollner S, Pfeuffer T, Carafoli E. The calmodulin binding domain of nitric oxide synthase and adenylyl cyclase. Biochemistry 1993;32:6081–8. [9] Diel S, Beyermann M, Llorens JM, Wittig B, Kleuss C. Two interaction sites on mammalian adenylyl cyclase type I and II: modulation by calmodulin and Gbg. Biochem J 2008;411:449–56. [10] Xia ZG, Refsdal CD, Merchant KM, Dorsa DM, Storm DR. Distribution of mRNA for the calmodulin-sensitive adenylate cyclase in rat brain: expression in areas associated with learning and memory. Neuron 1991;6:431–43. [11] Wu ZL, Thomas SA, Villacres EC, Xia Z, Simmons ML, Chavkin C, et al. Altered behavior and long-term potentiation in type I adenylyl cyclase mutant mice. Proc Natl Acad Sci U S A 1995;92:220–4. [12] Wang H, Ferguson GD, Pineda VV, Cundiff PE, Storm DR. Overexpression of type-1 adenylyl cyclase in mouse forebrain enhances recognition memory and LTP. Nat Neurosci 2004;7:635–42. [13] Squier TC, Bigelow DJ. Protein oxidation and age-dependent alterations in calcium homeostasis. Front Biosci 2000;5:D504–26. [14] Squier TC. Oxidative stress and protein aggregation during biological aging. Exp Gerontol 2001;36:1539–50. [15] Rasmussen HH, Hamilton EJ, Liu CC, Figtree GA. Reversible oxidative modification: implications for cardiovascular physiology and pathophysiology. Trends Cardiovasc Med 2010;20:85–90. [16] Agbas A, Moskovitz J. The role of methionine oxidation/reduction in the regulation of immune response. Curr Signal Transduct Ther 2009;4:46–50. [17] Stadtman ER. Protein oxidation in aging and age-related diseases. Ann N Y Acad Sci 2001;928:22–38. [18] O’Neil KT, DeGrado WF. How calmodulin binds its targets: sequence independent recognition of amphiphilic a-helices. Trends Biochem Sci 1990;15:59–64. [19] Sharov VS, Ferrington DA, Squier TC, Scho¨neich C. Diastereoselective reduction of protein-bound methionine sulfoxide by methionine sulfoxide reductase. FEBS Lett 1999;455:247–50. [20] Bigelow DJ, Squier TC. Redox modulation of cellular signaling and metabolism through reversible oxidation of methionine sensors in calcium regulatory proteins. Biochim Biophys Acta 2005;1703:121–34. [21] Gao J, Yin D, Yao Y, Williams TD, Squier TC. Progressive decline in the ability of calmodulin isolated from aged brain to activate the plasma membrane CaATPase. Biochemistry 1998;37:9536–48. [22] Osborn KD, Bartlett RK, Mandal A, Zaidi A, Urbauer RJ, Urbauer JL, et al. Singlemolecule dynamics reveal an altered conformation for the autoinhibitory domain of plasma membrane Ca2+-ATPase bound to oxidatively modified calmodulin. Biochemistry 2004;43:12937–44. [23] Montgomery HJ, Bartlett R, Perdicakis B, Jervis E, Squier TC, Guillemette JG. Activation of constitutive nitric oxide synthases by oxidized calmodulin mutants. Biochemistry 2003;42:7759–68. [24] Richardson JS, Richardson DC. Principles and patterns of protein conformation. In: Fasman GD, editor. Prediction of protein structure and the principles of protein conformation. Boston, MA: Springer; 1989. p. 1–98. [25] Zhang M, Li M, Wang JH, Vogel HJ. The effect of Met!Leu mutations on calmodulin’s ability to activate cyclic nucleotide phosphodiesterase. J Biol Chem 1994;269:15546–52. [26] Bartlett RK, Bieber Urbauer RJ, Anbanandam A, Smallwood HS, Urbauer JL, Squier TC. Oxidation of Met144 and Met145 in calmodulin blocks calmodulin dependent activation of the plasma membrane Ca-ATPase. Biochemistry 2003;42:3231–8. [27] Schuler D, Lu¨bker C, Lushington GH, Tang WJ, Shen Y, Richter M, et al. Interactions of Bordetella pertussis adenylyl cyclase toxin CyaA with calmodulin mutants and calmodulin antagonists: comparison with membranous adenylyl cyclase I. Biochem Pharmacol 2012;83:839–48. [28] Balog EM, Lockamy EL, Thomas DD, Ferrington DA. Site-specific methionine oxidation initiates calmodulin degradation by the 20S proteasome. Biochemistry 2009;48:3005–16. [29] Brot N, Weissbach L, Werth J, Weissbach H. Enzymatic reduction of proteinbound methionine sulfoxide. Proc Natl Acad Sci U S A 1981;78:2155–8. [30] Weissbach H, Etienne F, Hoshi T, Heinemann SH, Lowther WT, Matthews B, et al. Peptide methionine sulfoxide reductase: structure, mechanism of action, and biological function. Arch Biochem Biophys 2002;397:172–8. [31] Moskovitz J. Roles of methionine sulfoxide reductases in antioxidant defense, protein regulation and survival. Curr Pharm Des 2005;11:1451–7. [32] Oien DB, Moskovitz J. Substrates of the methionine sulfoxide reductase system and their physiological relevance. Curr Top Dev Biol 2008;80:93–133. [33] Moskovitz J, Poston JM, Berlett BS, Nosworthy NJ, Szczepanowski R, Stadtman ER. Identification and characterization of a putative active site for peptide methionine sulfoxide reductase (MsrA) and its substrate stereospecificity. J Biol Chem 2000;275:14167–72. [34] Sharov VS, Scho¨neich C. Diastereoselective protein methionine oxidation by reactive oxygen species and diastereoselective repair by methionine sulfoxide reductase. Free Radic Biol Med 2000;29:986–94. [35] Yamamoto M, Ozawa H, Saito T, Hatta S, Riederer P, Takahata N. Ca2+/CaMsensitive adenylyl cyclase activity is decreased in the Alzheimer’s brain: possible relation to type I adenylyl cyclase. J Neural Transm 1997;104:721–32. [36] Yamamoto M, Go¨tz ME, Ozawa H, Luckhaus C, Saito T, Rosler M, et al. Hippocampal level of neural specific adenylyl cyclase type I is decreased in Alzheimer’s disease. Biochim Biophys Acta 2000;1535:60–8.

C. Lu¨bker et al. / Biochemical Pharmacology 93 (2015) 196–209 [37] Mou TC, Gille A, Fancy DA, Seifert R, Sprang SR. Structural basis for the inhibition of mammalian membrane adenylyl cyclase by 20 (30 )-O-(N-methylanthraniloyl)-guanosine 50 -triphosphate. J Biol Chem 2005;280:7253–61. [38] Erdorf M, Seifert R. Pharmacological characterization of adenylyl cyclase isoforms in rabbit kidney membranes. Naunyn-Schmiedeberg’s Arch Pharmacol 2011;383:357–72. [39] Seifert R, Lee TW, Lam VT, Kobilka BK. Reconstitution of b2-adrenoceptor-GTPbinding-protein interaction in Sf9 cells – high coupling efficiency in a b2adrenoceptor-Gsa fusion protein. Eur J Biochem 1998;255:369–82. [40] Masada N, Ciruela A, Macdougall DA, Cooper DM. Distinct mechanisms of regulation by Ca2+/calmodulin of type 1 and 8 adenylyl cyclases support their different physiological roles. J Biol Chem 2009;284:4451–63. [41] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193:265–75. [42] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. [43] Trybus KM, Lowey S. Mechanism of smooth muscle myosin phosphorylation. J Biol Chem 1985;260:15988–95. [44] Moskovitz J, Berlett BS, Poston JM, Stadtman ER. The yeast peptide-methionine sulfoxide reductase functions as an antioxidant in vivo. Proc Natl Acad Sci U S A 1997;94:9585–9. [45] Moskovitz J, Singh VK, Requena J, Wilkinson BJ, Jayaswal RK, Stadtman ER. Purification and characterization of methionine sulfoxide reductases from mouse and Staphylococcus aureus and their substrate stereospecificity. Biochem Biophys Res Commun 2002;290:62–5. [46] Kim HY, Gladyshev VN. Methionine sulfoxide reduction in mammals: characterization of methionine-R-sulfoxide reductases. Mol Biol Cell 2004;15: 1055–64. [47] Brot N, Weissbach H. Biochemistry of methionine sulfoxide residues in proteins. Biofactors 1991;3:91–6. [48] Ejiri SI, Weissbach H, Brot N. The purification of methionine sulfoxide reductase from Escherichia coli. Anal Biochem 1980;102:393–8. [49] Louis-Jeune C, Andrade-Navarro MA, Perez-Iratxeta C. Prediction of protein secondary structure from circular dichroism using theoretically derived spectra. Proteins 2011;80:374–81. [50] Vougier S, Mary J, Dautin N, Vinh J, Friguet B, Ladant D. Essential role of methionine residues in calmodulin binding to Bordetella pertussis adenylate cyclase, as probed by selective oxidation and repair by the peptide methionine sulfoxide reductases. J Biol Chem 2004;279:30210–18. [51] Klee CB, Crouch TH, Krinks MH. Calcineurin: a calcium- and calmodulinbinding protein of the nervous system. Proc Natl Acad Sci U S A 1979;76: 6270–3. [52] Sun H, Gao J, Ferrington DA, Biesiada H, Williams TD, Squier TC. Repair of oxidized calmodulin by methionine sulfoxide reductase restores ability to activate the plasma membrane Ca-ATPase. Biochemistry 1999;38:105–12. [53] Ahlijanian MK, Cooper DM. Antagonism of calmodulin-stimulated adenylate cyclase by trifluoperazine, calmidazolium and W-7 in rat cerebellar membranes. J Pharmacol Exp Ther 1987;241:407–14. [54] Carafoli E. Intracellular calcium homeostasis. Annu Rev Biochem 1987;56: 395–433. [55] Jones EM, Squier TC, Sacksteder CA. An altered mode of calcium coordination in methionine-oxidized calmodulin. Biophys J 2008;95:5268–80. [56] Hu B, Nakata H, Gu C, De Beer T, Cooper DM. A critical interplay between Ca2+ inhibition and activation by Mg2+ of AC5 revealed by mutants and chimeric constructs. J Biol Chem 2002;277:33139–47.

209

[57] Gao J, Yao Y, Squier TC. Oxidatively modified calmodulin binds to the plasma membrane Ca-ATPase in a nonproductive and conformationally disordered complex. Biophys J 2001;80:1791–801. [58] Holzwarth G, Doty P. The ultraviolet circular dichroism of polypeptides. J Am Chem Soc 1965;87:218–28. [59] Masada N, Schaks S, Jackson SE, Sinz A, Cooper DM. Distinct mechanisms of calmodulin binding and regulation of adenylyl cyclases 1 and 8. Biochemistry 2012;51:7917–29. [60] Gao ZH, Krebs J, VanBerkum MF, Tang WJ, Maune JF, Means AR, et al. Activation of four enzymes by two series of calmodulin mutants with point mutations in individual Ca2+ binding sites. J Biol Chem 1993;268:20096–104. [61] Anbanandam A, Bieber Urbauer RJ, Bartlett RK, Smallwood HS, Squier TC, Urbauer JL. Mediating molecular recognition by methionine oxidation: conformational switching by oxidation of methionine in the carboxyl-terminal domain of calmodulin. Biochemistry 2005;44:9486–96. [62] Zhang M, Vogel HJ. Two-dimensional NMR studies of selenomethionyl calmodulin. J Mol Biol 1994;239:545–54. [63] Yuan T, Vogel HJ. Substitution of the methionine residues of calmodulin with the unnatural amino acid analogs ethionine and norleucine: biochemical and spectroscopic studies. Protein Sci 1999;8:113–21. [64] Edwards RA, Walsh MP, Sutherland C, Vogel HJ. Activation of calcineurin and smooth muscle myosin light chain kinase by met-to-leu mutants of calmodulin. Biochem J 1998;331(Pt 1):149–52. [65] Gao J, Yin DH, Yao Y, Sun H, Qin Z, Scho¨neich C, et al. Loss of conformational stability in calmodulin upon methionine oxidation. Biophys J 1998;74:1115–34. [66] Sacksteder CA, Whittier JE, Xiong Y, Li J, Galeva NA, Jacoby ME, et al. Tertiary structural rearrangements upon oxidation of Methionine145 in calmodulin promotes targeted proteasomal degradation. Biophys J 2006;91:1480–93. [67] Sharp JS, Tomer KB. Analysis of the oxidative damage-induced conformational changes of apo- and holocalmodulin by dose-dependent protein oxidative surface mapping. Biophys J 2007;92:1682–92. [68] Yin D, Kuczera K, Squier TC. The sensitivity of carboxyl-terminal methionines in calmodulin isoforms to oxidation by H2O2 modulates the ability to activate the plasma membrane Ca-ATPase. Chem Res Toxicol 2000;13:103–10. [69] Lafitte D, Tsvetkov PO, Devred F, Toci R, Barras F, Briand C, et al. Cation binding mode of fully oxidised calmodulin explained by the unfolding of the apostate. Biochim Biophys Acta 2002;1600:105–10. [70] Bootman MD, Berridge MJ. The elemental principles of calcium signaling. Cell 1995;83:675–8. [71] Lim JC, Kim G, Levine RL. Stereospecific oxidation of calmodulin by methionine sulfoxide reductase A. Free Radic Biol Med 2013;61C:257–64. [72] Chen B, Markillie LM, Xiong Y, Mayer MU, Squier TC. Increased catalytic efficiency following gene fusion of bifunctional methionine sulfoxide reductase enzymes from Shewanella oneidensis. Biochemistry 2007;46:14153–61. [73] Ferrington DA, Sun H, Murray KK, Costa J, Williams TD, Bigelow DJ, et al. Selective degradation of oxidized calmodulin by the 20S proteasome. J Biol Chem 2001;276:937–43. [74] Franklin JM, Carrasco GA, Moskovitz J. Induction of methionine sulfoxide reductase activity by pergolide, pergolide sulfoxide, and S-adenosyl-methionine in neuronal cells. Neurosci Lett 2013;533:86–9. [75] Guo Q, Shen Y, Lee YS, Gibbs CS, Mrksich M, Tang W-J. Structural basis for the interaction of Bordetella pertussis adenylyl cyclase toxin with calmodulin. EMBO J 2005;24:3190–201. [76] Chattopadhyaya R, Meador WE, Means AR, Quiocho FA. Calmodulin structure refined at 1.7 A˚ resolution. J Mol Biol 1992;228:1177–92.