TY - JOUR
AU - Hirota, Hiroshi
AB - Abstract Aequorin, which is a calcium-sensitive photoprotein and a member of the EF-hand superfamily, binds to Mg2+ under physiological conditions, which modulates its light emission. The Mg2+ binding site and its stabilizing influence were examined by NMR spectroscopy. The binding of Mg2+ to aequorin prevented the molecule from aggregating and stabilized it in the monomeric form. To determine the structural differences between Mg2+-bound and free aequorin, we have performed backbone NMR assignments of aequorin in the Mg2+-free state. Mg2+ binding induces conformational changes that are localized in the EF-hand loops. The chemical shift difference data indicated that there are two Mg2+-binding sites, EF-hands I and III. The Mg2+ titration experiment revealed that EF-hand III binds to Mg2+ with higher affinity than EF-hand I, and that only EF-hand III seems to be occupied by Mg2+ under physiological conditions. bioluminescence, EF-hand, magnesium, NMR, photoprotein DLS, dynamic light-scattering, MW, calculated molecular weight, RH, hydrodynamic radius The photoprotein aequorin, isolated from the jellyfish Aequoreaaequorea (synonyms A. victoria or A. forskalea), emits blue light by an intramolecular reaction upon Ca2+ binding (1–4), and it has been used as a probe to monitor Ca2+ concentrations in living cells (5, 6). Aequorin consists of apoaequorin (apoprotein) and 2-hydroperoxycoelenterazine, which is formed from coelenterazine and molecular oxygen (7, 8). When mixed with Ca2+, aequorin emits light (λmax = ∼465 nm), and decomposes into apoaequorin, coelenteramide and CO2. Apoaequorin comprises 189 amino acid residues in a single polypeptide chain (9, 10). Homology searches revealed that aequorin has three EF-hand motifs characteristic of Ca2+-binding sites (Fig. 1), and thus it has been classified as a member of the EF-hand superfamily, which includes calmodulin and troponin C. The EF-hand motif comprises two helices that flank a loop of 12 contiguous residues (11, 12). Recent crystal structure analysis revealed that aequorin has four helix-loop-helix structures for the EF-hand domains (I, II, III and IV), and that the second EF-hand (II) domain cannot bind Ca2+ (8) due to a lack of Ca2+-binding amino acid residues (Fig. 1). Fig. 1. View largeDownload slide Ribbon representation, amino acid sequence and helix-loop-helix structure of the EF-hands in aequorin. (A) Ribbon representation showing the secondary structure elements in the protein (PDB ID: 1EJ3). The loops of EF-hands are colored green (EF-hand I), yellow (EF-hand II), purple (EF-hand III), and pink (EF-hand IV). Coelenterazine is shown as a CPK representation. (B) The amino acid sequence of aequorin, and the helix regions of the EF-hands identified from the crystal structure study (8). The loops of EF-hands I-IV are shown in boxes and the helices of the EF-hand motifs are indicated with bars below the sequence. (C) The positions of each EF-hand loop are numbered 1–12, corresponding to amino acid residues 24–35 (EF-hand I), 69–80 (EF-hand II), 117–128 (EF-hand III), and 153–164 (EF-hand IV) in aequorin. The amino acid residues interacting with Ca2+ are circled, and ±X, ±Y, and ±Z are indicated as the coordinating positions in the pentagonal bipyramidal arrangement of Ca2+. The –Y coordinating position is assumed to be the backbone carbonyl oxygen at the position of the 7th amino acid residue. The glycine residues conserved in the canonical EF-hand loops are boxed. Fig. 1. View largeDownload slide Ribbon representation, amino acid sequence and helix-loop-helix structure of the EF-hands in aequorin. (A) Ribbon representation showing the secondary structure elements in the protein (PDB ID: 1EJ3). The loops of EF-hands are colored green (EF-hand I), yellow (EF-hand II), purple (EF-hand III), and pink (EF-hand IV). Coelenterazine is shown as a CPK representation. (B) The amino acid sequence of aequorin, and the helix regions of the EF-hands identified from the crystal structure study (8). The loops of EF-hands I-IV are shown in boxes and the helices of the EF-hand motifs are indicated with bars below the sequence. (C) The positions of each EF-hand loop are numbered 1–12, corresponding to amino acid residues 24–35 (EF-hand I), 69–80 (EF-hand II), 117–128 (EF-hand III), and 153–164 (EF-hand IV) in aequorin. The amino acid residues interacting with Ca2+ are circled, and ±X, ±Y, and ±Z are indicated as the coordinating positions in the pentagonal bipyramidal arrangement of Ca2+. The –Y coordinating position is assumed to be the backbone carbonyl oxygen at the position of the 7th amino acid residue. The glycine residues conserved in the canonical EF-hand loops are boxed. As previously reported, results obtained on luminometric titration of native aequorin (13) and recombinant aequorin (14) with Ca2+ led to the conclusion that the number of Ca2+ ions required for the luminescence of aequorin is two, not three, per aequorin molecule. On the other hand, the titration of Ca2+ with a Ca2+-sensitive electrode indicated that aequorin could bind more than two, probably three, Ca2+ (14). The luminescence of aequorin is triggered by two Ca2+, and the binding-affinity of the first Ca2+ to aequorin is about 20 times higher than that of the third Ca2+, which may be unrelated to light emission. However, whether two Ca2+-binding sites exist among the three EF-hands for luminescence remains unclear. The metal specificities of aequorin have been investigated (15, 16), and Ca2+ and Sr2+ are suitable ions for light emission. Although Mg2+ and Ca2+ have similar chemical properties, Mg2+ shows different effects on the luminescence of aequorin: it decreases the calcium-independent light emission (17, 18), the Ca2+ sensitivity (19–22) and the rate of luminescence. The EF-hands have been classified into Ca2+/Mg2+-type and Ca2+-specific type on the basis of their selectivity and affinity for Mg2+ and Ca2+ (23). The Ca2+/Mg2+-type EF-hand binds Ca2+ with high affinity and Mg2+ with moderate affinity. We studied the effect of Mg2+ on the structure of aequorin. The dynamic light scattering method was used for assessment of the aggregation of molecules and the stability of the monomeric state. Recently, we reported the NMR assignment of Mg2+-bound aequorin (24). By comparing the NMR signals of Mg2+-free aequorin with those of Mg2+-bound aequorin, the binding sites of Mg2+ in aequorin were identified and the difference in the Mg2+ binding affinity was investigated by means of the Mg2+ titration method. MATERIALS AND METHODS Materials Tris(hydroxymethyl)aminomethane (Tris), 2-(N-morpholino)ethanesulfonic acid (MES), KCl, FeCl3 and MgCl2 were purchased from Nacalai Tesque Inc. (Kyoto, Japan). LeMaster's medium, (NH4)2SO4, ethylenediamine-N,N,N′,N′-tetraacetic acid (disodium salt: EDTA·2Na), and dithiothreitol were obtained from Wako Pure Chemicals (Osaka, Japan), and MgSO4 was from Junsei Chemical (Tokyo, Japan). 15NH4Cl was purchased from Shoko Co. (Tokyo, Japan), d-[U-13C]glucose and dl-1,4-[U-2H]dithiothreitol were obtained from Isotec Inc. (Miamisburg, OH, USA), and [U-2H]MES was from Cambridge Isotope Laboratories Inc. (Andover, MA, USA). All stable isotope enriched chemicals were >98 atom %. Preparation of 15N-Labeled and 13C-, 15N-Labeled Aequorin Coelenterazine was chemically synthesized as previously reported (25). The recombinant aequorin used in this experiment consists of 191 amino acids, with an artificial modification at the N-terminus with the amino acid sequence of ANS instead of V in native aequorin (9, 26, 27). The procedures for the purification of recombinant aequorin were as follows (28). E. coli strain BL21 (Amersham Biosciences Corp., NJ, USA) was transformed with the expression vector piP-HE (26), and a single colony was grown in 10 ml of Luria-Bertani medium containing ampicillin (50 µg/ml) at 25°C for 15 h. The cultured cells were harvested by centrifugation at 5,000 × g for 5 min, suspended in 10 ml of M9-medium, and then added to 400 ml of M9-medium (13C-glucose, 0.8 g; 15NH4Cl, 0.4 g) and 0.01% Disfoam CE457 (a foam suppressant; NOF, Tokyo, Japan) in a 2 liter Erlenmeyer flask. After incubation at 37°C for 20 h on a rotary shaker (160 rpm/min), apoaequorin from the cells and the culture medium was used for aequorin preparation. To recover apoaequorin from the culture medium, the supernatant obtained by centrifugation at 5000 × g for 5 min was acidified with 1 M acetic acid to pH 4.6 and then the mixture was allowed to stand for 1 h at 4°C. A white precipitate formed and was collected by centrifugation at 12,000 × g for 10 min at 4°C. The precipitate containing apoaequorin was dissolved in 4 ml of 1 M Tris-HCl (pH 8.0). To obtain apoaequorin in cells, the cells were disrupted by sonication with a Branson model 200 sonifier in an ice bath. The suspension was centrifuged at 12,000 × g for 10 min at 4°C, and the pellet was discarded. Apoaequorin obtained from the culture medium and the cells was converted to aequorin and purified as previously described (28). The yield of 13C-, 15N-labeled aequorin was 10 mg from 400 ml of culture medium and cells. Dynamic Light Scattering Dynamic light scattering measurements were performed with a DynaPro molecular sizing instrument (DynaPro International Ltd., Milton Keynes, England) at 20°C, using a quartz scattering cuvette (12 µl). The instrument has a laser wavelength of 828.7 nm with a fixed scattering angle of 90°. The sample solutions, containing various concentrations of (NH4)2SO4, KCl and MgCl 2 in 10 mM Tris-HCl (pH 7.0) with 3 mM EDTA, were passed through a filter (0.22 µm, Whatman) and then used for measurements. Averaged data were obtained for twenty points and were analyzed with Dynamics 6.3.40 software (DynaPro International Ltd.). NMR Spectroscopy All NMR spectra were measured at 20°C on Bruker AVANCE500, Bruker AVANCE600 and Bruker AVANCE800 spectrometers, equipped with pulsed field gradients and triple resonance probes. The resonance assignments for Mg2+-bound aequorin were reported previously (24). The resonance assignments for the Mg2+-free state of aequorin were obtained through HNCO, HNCA, HNCOCA, and 15N NOESY-HSQC experiments with a 0.5 mM aequorin sample (uniformly labeled with 13C and 15N), in a buffer consisting of 10 mM Tris-d11 (pH 7.0), 3 mM EDTA, 100 mM KCl, and 10 mM dithiothreitol-d10 in 90% H2O/10% D2O. The samples used for the magnesium titration initially contained 0.2 mM 15N-labeled aequorin, in a buffer consisting of 10 mM MES-d13, 0.1 mM EDTA, 100 mM KCl, and 10 mM dithiothreitol-d10 in 90% H2O/10% D2O, pH 6.6. Small aliquots of a 1 M MgCl2 solution were added to the protein solution, and then two-dimensional 15N HSQC spectra were recorded. The sample used for the Ca2+-loaded aequorin contained 0.8 mM 15N-labeled aequorin, in a buffer consisting of 20 mM MES-d13, 100 mM KCl, 10 mM CaCl2 and 10 mM dithiothreitol-d10 in 90% H2O/10% D2O, pH 6.6. All spectra were processed using NMRPipe (29), and were analyzed using NMRview (30) with a home built module, Kujira. RESULTS Stability of Aequorin in Solution In our preliminary studies involving heteronuclear 1H-15N correlation NMR (HSQC) analysis of 15N-labeled aequorin at 20°C, aequorin dissolved in 10 mM MES buffer (pH 6.5) containing 100 mM KCl showed good dispersion. However, after 10 days, broad signals appeared in the middle region and signals exhibiting heterogeneous intensity were found in the 1H-15N HSQC spectra, suggesting that aequorin in solution formed aggregates and/or was denatured. For 2D and 3D NMR analyses, it is necessary to prevent protein aggregation and to stabilize the molecule. Dynamic light-scattering (DLS) analysis was employed to monitor the aggregation of aequorin in the presence of various salts. The results of the DLS analysis under various conditions are summarized in Table 1. The effect of (NH4)2SO4, which is known as a suitable stabilizer for aequorin, was examined. In the presence of 1.2 M (NH4)2SO4, the molecular weight of 1 mM aequorin was estimated to be 56 kDa, indicating that aequorin was present in an oligomeric state. A lower concentration of (NH4)2SO4 (10 mM) initially gave a monomer of 24 kDa, but the hydrodynamic radius and the calculated molecular mass increased after a week at 20°C, indicating that the protein became aggregated. Table 1. Effects of various salts on aequorin stability and aggregation. Calculated and measured values for the hydrodynamic radius (RH) and the molecular weight (MW) of aequorin under various ionic conditions after 7 days. Data with a monomodal distribution were analyzed using Dynamics 6.3.40 software. No.  Salt  Salt concentration  0 days     7 days           RH (nm)   MW (kDa)   RH (nm)   MW (kDa)   1  (NH4)2SO4  1.2 M  3.3  56  3.3  56  2  (NH4)2SO4  10 mM  2.3  24  Multicomponents    3  KCl  100 mM  2.5  28  Multicomponents    4  KCl MgCl2  100 mM 10 mM  2.3  21  2.3  22  No.  Salt  Salt concentration  0 days     7 days           RH (nm)   MW (kDa)   RH (nm)   MW (kDa)   1  (NH4)2SO4  1.2 M  3.3  56  3.3  56  2  (NH4)2SO4  10 mM  2.3  24  Multicomponents    3  KCl  100 mM  2.5  28  Multicomponents    4  KCl MgCl2  100 mM 10 mM  2.3  21  2.3  22  View Large The addition of 100 mM KCl to a 1 mM aequorin solution in 10 mM Tris buffer (pH 7.0) containing 3 mM EDTA and 10 mM DTT initially generated a monodisperse size distribution around a molecular weight of 28 kDa, which is similar to that of the monomeric aequorin (Table 1, No. 3; 0 days). However, after a week at 20°C, a broad size distribution was observed and the molecular size had increased. These data revealed that the molecular state of aequorin was heterogeneous and that it became aggregated (Table 1, No. 3; 7 days), suggesting that KCl is not effective in preventing aequorin aggregation. The addition of MgCl2 to the above buffer improved the situation. After a week, the size distribution of aequorin observed on DLS analysis was narrow and monomodal, indicating the absence of aggregation. The hydrodynamic radius was 2.3 nm and the molecular weight estimated from the hydrodynamic radius was 22 kDa (Table 1, No. 4). This result clearly indicated that MgCl2 has a profound stabilizing effect on monomeric aequorin. Mg2+-Bound and Free States Since the Mg2+-bound state of aequorin is more preferable for nuclear magnetic resonance (NMR) measurement than the Mg2+-free state, we applied NMR spectroscopy to backbone assignment of the Mg2+-bound state, as previously reported (24). To investigate the Mg2+ binding sites and the conformational changes induced by Mg2+, backbone resonance assignment of the Mg2+-free state of aequorin was performed. Due to the instability of the Mg2+-free state, each sample solution of the Mg2+-free aequorin was used within a week for the NMR measurements. Using the reported assignment results for the Mg2+-bound state (24), the assignment of the Mg2+-free state was performed using HNCA, HNCOCA, HNCO and 15N NOESY-HSQC spectra. The slightly poor quality of the spectra of Mg2+-free aequorin necessitated the use of a NOESY spectrum for the Mg2+-free assignment. The 1H-15N HSQC spectra of aequorin at pH 6.6 are shown in Fig. 2, the Mg2+-free aequorin exhibiting well-resolved peaks (Fig. 2, red). The HSQC spectra of the Mg2+-bound and Mg2+-free states of aequorin shared signal dispersion characteristics, suggesting that Mg2+ binding does not alter the global structure of aequorin. The glycine residues at G29, G74, G122 and G158 are located at the sixth position of the 12-residue loop in EF-hands I, II, III and IV, respectively. In a typical EF-hand motif, chemical shifts of the sixth glycine residue appear at low fields, as a consequence of the hydrogen bonding between the first amino acid residue and the sixth glycine residue of the EF-hand loop (31–33). The backbone assignments indicated that the 1H signals of G29, G74 and G122 appeared in the low-field region, which are a typical feature of an EF-hand motif. On the other hand, G158 was observed in the middle region of the 1H chemical shift, suggesting that the features of EF-hand IV differ from those of a typical EF-hand motif. The signals that appeared at different positions for the Mg2+-bound and Mg2+-free spectra were mainly assigned to the loop regions of EF-hands I, II, III and IV. The amide protons of residues I31 (loop I), D117 (loop III), I124 (loop III), T125 (loop III), L126 (loop III), D153 (loop IV), and I154 (loop IV) remain unassigned. The chemical shift difference values for the Mg2+-bound and free states were plotted against the residue number (Fig. 3). Large chemical shift changes were observed for the Ca2+ binding loops of EF-hands I and III. In these EF-hand loops, N26, N28, D119, N121 and G122 exhibited large chemical shift changes. An EF-hand motif binds Mg2+ and Ca2+, with the chelating loop residues at 1(+X), 3(+Y), 5(+Z), 7(–Y), 9(–X), and 12(–Z) arranged in an octahedral geometry with six oxygen atoms and a pentagonal bipyramidal geometry with seven oxygen atoms, respectively. In aequorin, N26 and D119 are located at the +Y position, and N28 and N121 at the +Z position in loops I and III, respectively. These data showed that Mg2+ can bind to EF-hands I and III, and that the amino acid residues positioned at +Y and +Z play an important role in Mg2+ binding. Although small chemical shift changes were observed in the loops of EF-hands II (E76) and IV (V162), other residues positioned at X, Y and Z that are involved in Ca2+ or Mg2+ coordination showed little or no chemical shift changes, indicating that Mg2+ did not bind to EF-hands II and IV. These data showed that Mg2+-binding to EF-hands I and III did not change the global structure of aequorin, in spite of the alteration of the local conformation. Fig. 2. View largeDownload slide Superimposed 1H-15N HSQC spectra of 15N-labeled aequorin in the presence and absence of Mg2+. The Mg2+ concentrations were 0 mM (red) and 10 mM (black), respectively. Labels indicate the glycines (G29, G74, G122, G158) located at position 6 of each EF-hand loop, N28 and N121 located at the +Z position of EF-hands I and III, respectively, and V162 located in the EF-hand IV loop. Fig. 2. View largeDownload slide Superimposed 1H-15N HSQC spectra of 15N-labeled aequorin in the presence and absence of Mg2+. The Mg2+ concentrations were 0 mM (red) and 10 mM (black), respectively. Labels indicate the glycines (G29, G74, G122, G158) located at position 6 of each EF-hand loop, N28 and N121 located at the +Z position of EF-hands I and III, respectively, and V162 located in the EF-hand IV loop. Fig. 3. View largeDownload slide Absolute value differences of amide proton chemical shifts between Mg2+-free and Mg2+-bound aequorin. The lines represent the average chemical shift differences. The loop regions of EF-hands I–IV are indicated by lines. Proline residues are indicated by plus (+) marks. The asterisks (*) represent uncalculated residues. Fig. 3. View largeDownload slide Absolute value differences of amide proton chemical shifts between Mg2+-free and Mg2+-bound aequorin. The lines represent the average chemical shift differences. The loop regions of EF-hands I–IV are indicated by lines. Proline residues are indicated by plus (+) marks. The asterisks (*) represent uncalculated residues. Mg2+ Titration To obtain more detailed information on the effect of Mg2+ on aequorin, including the stoichiometry of Mg2+, the site preference of Mg2+-binding and the Mg2+-induced conformational change, Mg2+ titration with uniformly 15N-labeled aequorin was performed while 1H-15N HSQC spectra were recorded. This technique, which monitors the backbone amide chemical shifts, is useful for investigating the protein structure and the global conformational changes. When Mg2+ was added to saturate EDTA in the aequorin solution, the intensities of the resonances corresponding to the EF-hand III and IV loops decreased, whereas the resonance intensities for the loops of EF-hands I and II remained unchanged (Fig. 4, yellow). When the concentration of Mg2+ reached 4 molar equivalents of aequorin (aequorin 0.2 mM, EDTA 0.1 mM and Mg2+ 0.9 mM; Fig. 4, blue), the resonance intensities for EF-hands I and II decreased and the signal positions did not change. On the other hand, the resonances corresponding to the loops of EF-hands III and IV, including G122 and G158, disappeared with substantial line broadenings and the signal positions shifted slightly. Such line broadening arises in the case when the rate of exchange between the Mg2+-bound and Mg2+-free states of aequorin is similar to the difference between the corresponding resonance frequencies. No further significant changes of the residues in EF-hands I and II were observed when 20 molar equivalents of Mg2+ were added to the sample (Fig. 4, pale green and green). In the presence of an excess amount of Mg2+ relative to the protein (aequorin 0.2 mM, EDTA 0.1 mM and Mg2+ 10 mM), the resonance frequencies of EF-hands I and II shifted, indicating that the affinity of EF-hand I is much lower than that of EF-hand III (Fig. 4, black). In the course of the titration, most of the resonances corresponding to the helical regions throughout the protein did not undergo appreciable chemical shift changes, suggesting that the conformational changes induced by Mg2+ binding were localized to only the loops of the EF-hands. Fig. 4. View largeDownload slide Selected regions of the 1H-15N HSQC spectra of aequorin, showing the G29, G74, G122 and G158 peaks with different concentrations of Mg2+. The inset shows the region of G158 located in the EF-hand IV loop. The glycine residues located at the 6 position of each EF-hand is numbered. The molar ratios of Mg2+ to aequorin are shown in colors: red, 0 (Mg2+-free); yellow, 0.5 molar equiv.; cyan, 4 molar equiv.; pale green, 10 molar equiv.; green, 20 molar equiv.; black, 50 molar equiv. (10 mM). Peaks of G122 for 0.5, 4, 10 and 20 molar equiv. were not detected. Fig. 4. View largeDownload slide Selected regions of the 1H-15N HSQC spectra of aequorin, showing the G29, G74, G122 and G158 peaks with different concentrations of Mg2+. The inset shows the region of G158 located in the EF-hand IV loop. The glycine residues located at the 6 position of each EF-hand is numbered. The molar ratios of Mg2+ to aequorin are shown in colors: red, 0 (Mg2+-free); yellow, 0.5 molar equiv.; cyan, 4 molar equiv.; pale green, 10 molar equiv.; green, 20 molar equiv.; black, 50 molar equiv. (10 mM). Peaks of G122 for 0.5, 4, 10 and 20 molar equiv. were not detected. DISCUSSION In this study, we have used NMR to identify the Mg2+ binding sites in aequorin and to determine the preference for Mg2+. It is known that Mg2+ decreases the Ca2+-dependent light emission of aequorin (17, 35). Since the intracellular concentration of Mg2+ (10−3 M) is approximately 102–104-fold higher than that of Ca2+, intracellular Mg2+ might have some effect on aequorin, in terms of the Ca2+-triggered luminescent reaction in living cells. Aequorin Stabilization on Mg2+ Binding Our characterization of the Mg2+-bound and metal-free states of aequorin has suggested that Mg2+ binding provides structural stability to the protein. The dynamic light scattering data clearly showed that Mg2+ stabilizes aequorin in the monomeric state for a week. Furthermore, the NMR and CD spectroscopic data demonstrated that the binding of Mg2+ to the EF-hand loops increases the structural stability of aequorin. In the metal-free state, temperature and time-dependent spectral changes were observed (data not shown). The CD spectrum of metal-free aequorin (10 µM) at 45°C shows a substantial reduction of the α-helix component, as compared to that obtained at 20°C. A smaller spectral change than that of the metal-free state is observed for aequorin with 0.3 mM Mg2+. The contribution of Mg2+ to structural stability has also been observed with troponin C, and Mg2+ binding to the C-terminal EF-hands increases the structural stability (36, 37). Similarly, the binding of Mg2+ to EF-hand III in aequorin may represent the most important contribution to the structural stability of aequorin. EF-Hand II and IV, to Which Mg2+ Does Not Bind Due to the lack of Ca2+-binding amino acid residues, EF-hand II cannot bind to Ca2+. The present NMR results show EF-hand II does not bind to Mg2+. Although the chemical shift of the loop region of EF-hand II including E76 changed slightly, this is not due to the binding of Mg2+. In the crystal structure of the metal-free form of aequorin, the loop of EF-hand II is located close to the loop of EF-hand I, and a β-sheet structure is formed between E76–D78 of EF-hand II and K30–S32 of EF-hand I. Therefore, the structural change of EF-hand I loop that is induced by the binding of Mg2+ can lead to a slight change in the chemical shift of the EF-hand II loop including E76. The NMR data showed that Mg2+ does not bind to EF-hand IV. This result was rather unexpected, since the loop of EF-hand IV is composed of canonical EF-hand sequences and possesses Ca2+ binding ability. The chemical shift of V162, located in the EF-hand IV loop, changed slightly, but this does not indicate Mg2+-binding to this region. In the crystal structure of aequorin in the Ca2+-free state, the amide proton of V162 forms a hydrogen bond with the carbonyl oxygen of G122, located in the loop of EF-hand III (8). It is reasonable to conclude that the resonance of V162 changes as a consequence of the rearrangement of the hydrogen bond network of Mg2+-bound EF-hand III. In our study, a hydrogen bond between the residues at the first and sixth positions of the EF-hand IV loop was not detected, and the chemical shifts of V151, D153 and I154 located in the loop of EF-hand IV remained unassigned. In the crystal structure of the Ca2+-free state of aequorin, the hydrogen bond network is disrupted and the loop structure in the loop of EF-hand IV is deformed, as compared to in other EF-hand loops (8, 38). The NMR and X-ray data consistently show that the loop conformation of EF-hand IV differs from the canonical one and that it exhibits structural flexibility. This may be a reason why EF-hand IV is not suitable for Mg2+ binding. The affinities of EF-hands I, III and IV for Ca2+ have been investigated by using synthetic peptide fragments (20–22 amino acid residues) of the EF-hand loops, and the dissociation constants for Ca2+ showed the binding affinity order of III, I and IV (39). This order agrees well with that found in our Mg2+ binding experiments. These results imply that aequorin has the same binding preference for Ca2+ and Mg2+. Mg2+ Binding to Aequorin under Physiological Conditions The Mg2+ titration experiments on aequorin involving NMR revealed that EF-hand III is a high Mg2+ affinity site and that EF-hand I is a low affinity site. This is the first report of the metal-binding preferences of the EF-hands in aequorin. Titration experiments showed that the Mg2+ affinities of EF-hand III and EF-hand I are 0.2–1 mM and 2–10 mM (KMg), respectively. Since no data were available regarding the direct interaction of Mg2+ with each of the binding sites in aequorin, these results provide a new insight into the interaction of Mg2+ with aequorin. As the intracellular concentration of Mg2+ is in the 10−3 M range, this level would only facilitate Mg2+ binding to EF-hand III. Therefore, our results suggest that in the absence of an intracellular Ca2+ stimulus, the predominant form of aequorin within the cell would be Mg1-aequorin (Fig. 5). The affinity (KD) of Ca2+ for aequorin was reported to be 13 µM (34), and that the affinity of Mg2+ for aequorin is much weaker than that of Ca2+. When Ca2+ binds to aequorin in the presence of excess Mg2+, such as under intracellular conditions, EF-hand III occupied by Mg2+ would interfere with Ca2+-binding to this site, which would reduce the light emission. The presence of 1 mM Mg2+ decreases the Ca2+-dependent luminescence (21). Our results suggest that the inhibition is caused by the binding of Mg2+ to EF-hand III. Since EF-hand I exhibits low affinity for Mg2+ and is considered to exist in the free form with a physiological Mg2+ concentration, the Mg2+ effects on EF-hand I in terms of Ca2+-binding and luminescence would probably be weaker than those on EF-hand III. The conformational change induced by Mg2+ is only localized in the loops of the EF-hands of aequorin, the other part remaining unchanged. Only the peaks assigned to the EF-hand loop regions exhibit different chemical shifts in the Mg2+-bound and free HSQC spectra. A local conformational change induced by Mg2+ has also been observed for other EF-hand proteins. In both calmodulin and troponin C, the binding of Ca2+ to the EF-hand induces conformational changes including interhelical angle rearrangements (32, 33), although the proteins also undergo local structural changes induced by Mg2+ (40, 41). The 1H-15N HSQC spectrum of Ca2+-bound aequorin showed good dispersion, suggesting that Ca2+-bound aequorin is folded (data not shown). However, there are some different features from the Mg2+-bound state: the number of signals increases, and the signal intensity becomes heterogeneous and distorted, indicating that Ca2+-bound aequorin assumes several conformations. It is proposed that the binding of Ca2+ to aequorin induces large conformational changes, as observed for other EF-hand proteins. Fig. 5. View largeDownload slide Model of Mg2+ and/or Ca2+ binding to aequorin. (A) The aequorin binding equilibria with various concentrations of Mg2+ and Ca2+. (B) The Mg2+/Ca2+ binding types of EF-hands and the order of affinity for Mg2+ binding to aequorin. Fig. 5. View largeDownload slide Model of Mg2+ and/or Ca2+ binding to aequorin. (A) The aequorin binding equilibria with various concentrations of Mg2+ and Ca2+. (B) The Mg2+/Ca2+ binding types of EF-hands and the order of affinity for Mg2+ binding to aequorin. In conclusion, our results demonstrate that aequorin has two Mg2+ binding sites located within the EF-hands. Among the three functional EF-hands (I, III and IV), only EF-hands I and III serve as Mg2+ binding sites. The affinity of EF-hand III for Mg2+ is stronger than that of EF-hand I. EF-hand III exhibits physiologically relevant affinity for Mg2+. The Mg2+-induced conformational change occurs locally within the EF-hand loops, the rest of the protein structure remaining unchanged. The binding of Mg2+ to the EF-hand stabilizes the aequorin molecule in the monomeric state. The stabilization by Mg2+ facilitated the collection of two-dimensional and three-dimensional NMR spectroscopic data, and the backbone assignment of the Mg2+-bound state. The backbone assignment of the Mg2+-bound state of aequorin provides information about the backbone assignment of the metal-free state and the Mg2+-binding properties of aequorin. This work was supported in part by the National Project on Protein Structural and Functional Analyses of the Ministry of Education, Culture, Sports, Science and Technology of Japan, and in part by the Program for Promotion of Fundamental Studies in Health Science of the Pharmaceutical and Medical Devices Agency (PMDA) of Japan (to S.I.). REFERENCES 1. Shimomura, O., Johnson, F.H., and Saiga, Y. ( 1962) Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan. Aequorea. J. Cell. Comp. Physiol.  59, 223–240 Google Scholar 2. Shimomura, O. and Johnson, F.H. ( 1979) Chemistry of the calcium-sensitive photoprotein aequorin. In Detection and Measurement of Free Calcium Ions in Cells (Ashley, C.C. and Campbell, A.K., eds.) pp. 73–83, Elsevier, North Amsterdam Google Scholar 3. Shimomura, O. ( 1983) Bioluminescence. Photochem. Photobiol.  38, 773–779 Google Scholar 4. Shimomura, O. and Johnson, F.H. ( 1976) Calcium-triggered luminescence of the photoprotein aequorin. Symp. Soc. Exptl. Biol.  30, 41–54 Google Scholar 5. Blinks, J.R., Prendergast, F.G., and Allen, D.G. ( 1976) Photoproteins as biological calcium indicators. Pharmacol. Rev.  28, 1–93 Google Scholar 6. Blinks, J.R. ( 1985) In Bioluminescence and Chemiluminescence: Instrumental Applications (K. Van Dyke, ed.) Vol. 2. pp. 185–226, CRC, Boca Raton, FL Google Scholar 7. Shimomura, O. and Johnson, F.H. ( 1978) Peroxidized coelenterazine, the active group in the photoprotein aequorin. Proc. Natl. Acad. Sci. USA  75, 2611–2615 Google Scholar 8. Head, J.F., Inouye, S., Teranishi, K., and Shimomura, O. ( 2000) The crystal structure of the photoprotein aequorin at 2.3 Å resolution. Nature  405, 372–376 Google Scholar 9. Inouye, S., Noguchi, M., Sakaki, Y., Takagi, Y., Miyata, T., Iwanaga, S. Miyata, T., and Tsuji, F.I. ( 1985) Cloning and sequence analysis of cDNA for the luminescent protein aequorin. Proc. Natl. Acad. Sci. USA  82, 3154–3158 Google Scholar 10. Charbonneau, H., Walsh, A.K., McCann, R.O., Prendergast, F.G., Cormier, M.J., and Vanaman, T.C. ( 1985) Amino acid sequence of the calcium-dependent photoprotein aequorin. Biochemistry  24, 6762–6771 Google Scholar 11. Kretsinger, R.H. and Nockolds, C.E. ( 1973) Carp muscle calcium-binding protein. II. Structure determination and general description. J. Biol. Chem.  248, 3313–3326 Google Scholar 12. Kawasaki, H., Nakayama, S., and Kretsinger, R.H. ( 1998) Classification and evolution of EF-hand proteins. Biometals  4, 277–295 Google Scholar 13. Shimomura, O. ( 1995) Luminescence of aequorin is triggered by the binding of two calcium ions. Biochem. Biophys. Res. Commun.  211, 359–363 Google Scholar 14. Shimomura, O. and Inouye, S. ( 1996) Titration of recombinant aequorin with calcium chloride. Biochem. Biophys. Res. Commun.  221, 77–81 Google Scholar 15. Izutsu, K.T., Felton, S.P., Siegel, I.A., Yoda, W.T., and Chen, A.C.N. ( 1972) Aequorin: its ionic specificity. Biochem. Biophys. Res. Commun.  49, 1034–1039 Google Scholar 16. Shimomura, O. and Johnson, F.H. ( 1973) Further data on the specificity of aequorin luminescence to calcium. Biochem. Biophys. Res. Commun.  53, 490–494 Google Scholar 17. Allen, D.G., Blinks, J.R., and Prendergast, F.G. ( 1977) Aequorin luminescence: relation of light emission to calcium concentration - a calcium-independent component. Science  196, 996–998 Google Scholar 18. Ray, B.C., Ho, S., Kemple, M.D., Prendergast, F.G., and Nageswara Rao, B.D.N. ( 1985) Proton NMR of aequorin. Structural changes concomitant with calcium-independent light emission. Biochemistry  24, 4280–4287 Google Scholar 19. Moisecu, D.G., Ashley, C.C., and Campbell, A.K. ( 1975) Comparative aspects of the calcium-sensitive photoproteins aequorin and obelin. Biochim. Biophys. Acta  369, 133–140 Google Scholar 20. Illarionov, B.A., Frank, L.A., Illarionova, V.A., Bondar, V. S, Vysotski, E.S., and Blinks, J.R. ( 2000) Recombinant obelin: Cloning and expression of cDNA, purification, and characterization as a calcium indicator. Methods Enzymol.  305, 223–249 Google Scholar 21. Moisescu, D.G. and Ashley, C.C. ( 1977) The effect of physiologically occurring cations upon aequorin light emission. Determination of the binding constants. Biochim. Biophys. Acta  460, 189–205 Google Scholar 22. Hastings, J.W., Mitchell, G., Mattingly, P.H., Blinks, J.R., and van Leeuwen, M. ( 1969) Response of aequorin bioluminescence to rapid changes in calcium concentration. Nature  222, 1047–1050 Google Scholar 23. Potter, J.F. and Gergely, J. ( 1975) The calcium and magnesium binding sites on troponin and their role in the regulation of myofibrillar adenosine triphosphatase. J. Biol. Chem.  250, 4628–4633 Google Scholar 24. Ohashi, W., Inouye, S., Yamazaki, T., Doi-Katayama, Y., Yokoyama, S., and Hirota, H. ( 2005) Backbone 1H, 13C and 15N resonance assignments for the Mg2+-bound form of the Ca2+-binding photoprotein aequorin. J. Biomol. NMR  31, 375–376 Google Scholar 25. Inoue, S., Sugiura, S., Kakoi, H., Hasuzume, K., Goto, T., and Iio, H. ( 1975) Squid bioluminescence II. Isolation from Watasenia scintillans and synthesis of 2-(p-hydroxybenzyl)-6-(p-hydroxyphenyl)-3,7-dihydroimidazo[1,2-a]pyrazin-3-one. Chem. Lett. 141–144 Google Scholar 26. Inouye, S., Aoyama, S., Miyata, T., Tsuji, F.I., and Sakaki, Y. ( 1989) Overexpression and purification of the recombinant Ca2+-binding protein, apoaequorin. J. Biochem.  105, 473–477 Google Scholar 27. Inouye, S., Zenno, S., Sakaki, Y., and Tsuji, F.I. ( 1991) High-level expression and purification of apoaequorin. Protein Expres. Purif.  2, 122–126 Google Scholar 28. Shimomura, O. and Inouye, S. ( 1999) The in situ regeneration and extraction of recombinant aequorin from Escherichia coli cells and the purification of extracted aequorin. Protein Expres. Purif.  16, 91–95 Google Scholar 29. Delaglio, F., Grzesiek, S., Vuister, G.W., Zhu, G., Pfeifer, J., and Bax, A. ( 1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR  6, 277–293 Google Scholar 30. Johnson, B. and Blevins, R. ( 1994) NMRView: a computer program for the visualization and analysis of NMR data. J. Biomol. NMR  4, 603–614 Google Scholar 31. Ikura, M., Minowa, O., Yazawa, M., Yagi, K., and Hikichi, K. ( 1987) Sequence-specific assignments of downfield-shifted amide proton resonances of calmodulin. FEBS Lett.  219, 17–21 Google Scholar 32. Spyracopoulos, L., Li, M.X., Sia, S.K., Gagne, S.M., Chandra, M., Solaro, R.J., and Sykes, B.D. ( 1997) Calcium-induced structural transition in the regulatory domain of human cardiac troponin C. Biochemistry  36, 12138–12146 Google Scholar 33. Zhang, M., Tanaka, T., and Ikura, M. ( 1995) Calcium-induced conformational transition revealed by the solution structure of apo calmodulin. Nat. Struct. Biol.  2, 758–767 Google Scholar 34. Kendall, J.M., Sala-Newby, G., Ghalaut, V., Dormer, R.L., and Campbell, A.K. ( 1992) Engineering the Ca2+-activated photoprotein aequorin with reduced affinity for calcium. Biochem. Biophys. Res. Commun.  187, 1091–1097 Google Scholar 35. Allen, D.G. and Blinks, J.R. ( 1978) Calcium transients in aequorin-injected frog cardiac muscle. Nature  273, 509–513 Google Scholar 36. Szczesna, D., Guzman, G., Miller, T., Zhao, J., Farokhi, K., Ellemberger, H., and Potter, J.D. ( 1996) The role of the four Ca2+ binding sites of troponin C in the regulation of skeletal muscle contraction. J. Biol. Chem.  271, 8381–8386 Google Scholar 37. Zot, H.G. and Potter, J.D. ( 1982) A structural role for the Ca2+-Mg2+ sites on troponin C in the regulation of muscle contraction. Preparation and properties of troponin C depleted myofibrils. J. Biol. Chem.  257, 7678–7683 Google Scholar 38. Toma, S., Chong, K, T., Nakagawa, A., Teranishi, K., Inouye, S., and Shimomura, O. ( 2005) The crystal structures of semi-synthetic aequorins. Protein Sci.  14, 409–416 Google Scholar 39. Oishi, O., Nagatomo, A., Kohzuma, T., Oda, N., Miyazima, T., Ohno, M., and Sakaki, Y. ( 1992) Validity of putative calcium binding loops of photoprotein aequorin. FEBS Lett.  307, 272–274 Google Scholar 40. Finley, N.L., Howarth, J.W., and Rosevear, P.R. ( 2004) Structure of the Mg2+-loaded C-lobe of cardiac troponin C bound to the N-domain of cardiac troponin I: comparison with the Ca2+-loaded structure. Biochemistry  43, 11371–11379 Google Scholar 41. Ohki, S., Ikura, M., and Zhang, M. ( 1997) Identification of Mg2+-binding sites and the role of Mg2+ on target recognition by calmodulin. Biochemistry  36, 4309–4316 Google Scholar Author notes 1RIKEN Genomic Sciences Center, 1-7-22, Suehiro, Tsurumi, Yokohama 230-0045; and 2Yokohama Research Center, Chisso Corporation, 5-1 Okawa, Kanazawa, Yokohama 236-8605 © 2005 The Japanese Biochemical Society.
TI - NMR Analysis of the Mg2+-Binding Properties of Aequorin, a Ca2+-Binding Photoprotein
JF - The Journal of Biochemistry
DO - 10.1093/jb/mvi164
DA - 2005-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/nmr-analysis-of-the-mg2-binding-properties-of-aequorin-a-ca2-binding-nCGdvHiRiK
SP - 613
EP - 620
VL - 138
IS - 5
DP - DeepDyve
ER -