Nuclear magnetic resonance investigation of cadmium 113 substituted pea and lentil lectins

The lentil (LcH) and pea (PSA) lectins, which are members of the class of D-glucose/D-mannose binding lectins, are Ca2+ X Mn2+ metalloproteins that require the metal ions for their saccharide binding and biological activities. We have prepared a variety of Cd2+ derivatives of PSA and LcH, with Cd2+ in either the transition metal (S1) or calcium (S2) sites, or in both. Thus, Cd2+ X Zn2+, Cd2+ X Mn2+, and Ca2+ X Cd2+ derivatives were prepared, in addition to the Cd2+ X Cd2+ derivatives which we have recently reported. This is the first report of stable mixed metal Cd2+ complexes of lectins. The physical and saccharide binding properties of the Cd2+ derivatives of both lectins were characterized by a variety of physiochemical techniques and found to be the same as those of the corresponding native proteins. 113Cd NMR spectra of mono- and disubstituted 113Cd2+ complexes of LcH and PSA were recorded and compared with 113Cd NMR data for concanavalin A (ConA) (Palmer, A.R., Bailey, D.B., Behnke, W.D., Cardin, A.D., Yang, P.P., and Ellis, P.D. (1980) Biochemistry 19, 5063-5070). The data for the PSA and LcH derivatives were found to be very similar, indicating close homology of their metal ion binding sites. 113Cd resonances at 44.6 ppm and -129.4 ppm for 113Cd2+ X 113Cd2+ X LcH, and at 46.6 and -130.4 for the corresponding PSA derivative, are chemical shifts very similar to those observed for 113Cd2+ X 113Cd2+ X ConA. Assignment of the resonances to the transition metal (S1) and calcium (S2) sites were unambiguous since the Ca2+ X 113Cd2+ and 113Cd2+ X Zn2+ derivatives of both lectins showed single resonances characteristic of the S1 and S2 sites, respectively. The results indicate that, unlike ConA, 113Cd2+ binds tightly to PSA and LcH. Binding of monosaccharide to both lectins induce small (2 ppm) upfield shifts in their S2 113Cd resonances, in contrast to the larger shift (8 ppm) observed in ConA. The 113Cd2+ X Mn2+ complexes of PSA and LcH fail to show a 113Cd resonance characteristic of these derivatives, which provides evidence for the close proximity of the metal ions in the two proteins. The present findings indicate that the coordinating ligand atoms to the metal ions at the S1 and S2 sites in LcH, PSA, and ConA are the same.     

Dynamic and structural properties of the calcium binding site of bovine serine proteases and their zymogens. A multinuclear nuclear magnetic resonance and stopped-flow study

The combined use of 43Ca and 113Cd nuclear magnetic resonance (n.m.r.) has provided information on the structural and dynamic properties of the calcium binding site located in homologous positions in bovine beta-trypsin, alpha-chymotrypsin and their zymogens. The 43Ca and 113Cd n.m.r. chemical shifts are consistent with an octahedral symmetry of the binding site and with the substitution of one of the two carboxylate ligands present in trypsin(ogen) with a neutral ligand in chymotrypsin(ogen). The constancy of the 113Cd n.m.r. chemical shifts upon binding of the pancreatic trypsin inhibitor and/or the dipeptide Ile-Val to trypsinogen confirms that structural changes in the activation domain do not affect the calcium binding site. The exchange between bound and "free" (solvated) Ca2+ is slow on the 43Ca n.m.r. time-scale for trypsin(ogen), but falls in the intermediate exchange region for chymotrypsin(ogen). In trypsin, the Ca2+ off-rate was measured by stopped-flow making use of the calcium indicator 1,2-bis(o-aminophenoxy)-ethane-N,N,N',N'-tetraacetic acid and was found to be 3(+/- 1) s-1. In chymotrypsin(ogen) the off-rates calculated from the 43Ca n.m.r. data are 70 s-1 and 350 s-1, respectively. The dynamic properties of the calcium binding site of serine (pro)enzymes have been related to the flexibility of the binding site itself and have been compared to those of other extracellular and intracellular calcium binding proteins.     

Probing different conformational states of bovine alpha-lactalbumin: fluorescence studies with 4,4'-bis[1-(phenylamino)-8-naphthalenesulfonate]

The binding of the fluorescent probe 4,4'-bis[1-(phenylamino)-8-naphthalenesulfonate] (bis-ANS) to bovine alpha-lactalbumin (alpha-LA) was investigated. A strong dependence of the Kd value with the bound calcium stoichiometry was found, with Kd values ranging from 6.2 +/- 0.4 to 64.6 +/- 5.9 microM for apo-alpha-LA and 1:1 Ca(II)-alpha-LA, respectively. A 350-fold enhancement of the bis-ANS emission was observed in the protein-bis-ANS complex, along with an approximately 33-nm blue shift. Both appeared to be related to the hydrophobicity of the binding site and were independent of the Ca(II) ion content. From the difference in bis-ANS affinity between apo-alpha-LA and Ca(II)-alpha-LA, we demonstrated that Zn(II) and Al(III) were able to "lock" the protein into a new "apo-like" conformation, which was similar to, but not identical with, the apo conformation. The protein could be interconverted between all three conformations in a Mn(II) titration. The first Mn(II) shifted the apoprotein to the Ca(II) conformation; at higher Mn(II) levels, binding to the second site shifted the protein toward the apo-like conformation. The same behavior was observed with calcium in large excess. The evidence supported a model for the mutually nonexclusive binding of metals both to site I ("calcium site") and to site II ("zinc site") simultaneously. The results suggest that alpha-lactalbumin possesses a hydrophobic surface that becomes somewhat less accessible upon 1:1 calcium binding in the absence of metals that also bind to the zinc site.     

Cadmium-113 NMR studies of the DNA binding domain of the mammalian glucocorticoid receptor

The DNA binding domain of the mammalian glucocorticoid hormone receptor (GR) contains nine highly conserved cysteine residues, a conservation shared by the superfamily of steroid and thyroid hormone receptors. A fragment [150 amino acids (AA) in length] consisting of GR residues 407-556, containing within it the entire DNA binding domain (residues 440-525), has been overexpressed and purified from Escherichia coli previously. This fragment has been shown to contain 2.3 +/- 0.2 mol of Zn(II) per mole of protein [Freedman, L. P., Luisi, B. F., Korszun, Z. R., Basavappa, R., Sigler, P. B., & Yamamoto, K. R. (1988) Nature 334, 543]. Zn(II) [or Cd(II) substitution] has been shown to be essential for specific DNA binding. 113Cd NMR of a cloned construct containing the minimal DNA binding domain of 86 AA residues [denoted GR(440-525)] with 113Cd(II) substituted for Zn(II) identifies 2 Cd(II) binding sites by the presence of 2 113Cd NMR signals each of which integrates to 1 113Cd nucleus. The chemical shifts of these two sites, 704 and 710 ppm, suggest that each 113Cd(II) is coordinated to four isolated -S- ligands. Shared -S- ligands connecting the two 113Cd(II) ions do not appear to be present, since their T1s differ by 10-fold, 0.2 and 2.0 s, respectively. Addition of a third 113Cd(II) or Zn(II) to 113Cd2GR(440-525) results in occupancy of a third site, which introduces exchange modulation of the two original 113Cd NMR signals causing them to disappear. Addition of EDTA to the protein restores the original two signals.(ABSTRACT TRUNCATED AT 250 WORDS)     

Competitive ion binding to low density lipoproteins: an electron spin resonance study

The electron spin resonance (ESR) technique was used to evaluate binding constants for Ca(II) and Mg(II) in interaction with low density lipoprotein (LDL). The Ca(II) or Mg(II) ions competed with the paramagnetic Mn(II) ions for the same binding sites of two different classes on the LDL surface. For each ion competing with Mn(II), the solutions of eight non-linear competition equations were fit to the experimental titration curves, with two adjustable parameters, the two binding constants. The derived "intrinsic" values (the values corrected for the electrolyte-induced change of the surface potential) for "strong" binding sites for Ca(II) (170 +/- 85 M-1) and Mg(II) (60 +/- 30 M-1) differ significantly from the respective value for Mn(II) (760 M-1). The values for the "weak" binding sites (18 M-1, 15 M-1 and 10 M-1 for Mn(II), Ca(II) and Mg(II), respectively are in the range of the binding constants for these ions in interaction with model membranes.     

113Cd nuclear magnetic resonance of Cd(II) alkaline phosphatases

113Cd NMR spectra of 113Cd(II)-substituted Escherichia coli alkaline phosphatase have been recorded over a range of pH values, levels of metal site occupancy, and states of phosphorylation. Under all conditions resonances attributable to cadmium specifically bound at one or more of the three pairs of metal-binding sites (A, B, and C sites) are detected. By following changes in both the 113Cd and 31P NMR spectra of 113Cd(II)2 alkaline phosphatase during and after phosphorylation, it has been possible to assign the cadmium resonance that occurs between 140 and 170 ppm to Cd(II) bound to the A or catalytic site of the enzyme and the resonance occurring between 51 and 76 ppm to Cd(II) bound to B site, which from x-ray data is located 3.9 A from the A site. The kinetics of phosphorylation show that cadmium migration from the A site of one subunit to the B site of the second subunit follows and is a consequence of phosphate binding, thus precluding the migration as a sufficient explanation for half-of-the-sites reactivity. Rather, there is evidence for subunit-subunit interaction rendering the phosphate binding sites inequivalent. Although one metal ion, at A site, is sufficient for phosphate binding and phosphorylation, the presence of a second metal ion at B site greatly enhances the rate of phosphorylation. In the absence of phosphate, occupation of the lower affinity B and C sites produces exchange broadening of the cadmium resonances. Phosphorylation abolishes this exchange modulation. Magnesium at high concentration broadens the resonances to the point of undetectability. The chemical shift of 113Cd(II) in both A and B sites (but not C site) is different depending on the state of the bound phosphate (whether covalently or noncovalently bound) and gives separate resonances for each form. Care must be taken in attributing the initial distribution of cadmium or phosphate in the reconstituted enzyme to that of the equilibrium species in samples reconstituted from apoenzyme. Both 113Cd NMR and 31P NMR show that some conformational changes consequent to metal ion or phosphate binding require several days before the final equilibrium species is formed.     

Electronic structure of the Mn-cofactor of modified bacterial reaction centers measured by electron paramagnetic resonance and electron spin echo envelope modulation spectroscopies

The electronic structure of a Mn(II) ion bound to highly oxidizing reaction centers of Rhodobacter sphaeroides was studied in a mutant modified to possess a metal binding site at a location comparable to the Mn₄Ca cluster of photosystem II. The Mn-binding site of the previously described mutant, M2, contains three carboxylates and one His at the binding site (Thielges et al., Biochemistry 44:389–7394, 2005). The redox-active Mn-cofactor was characterized using electron paramagnetic resonance (EPR) and electron spin echo envelope modulation (ESEEM) spectroscopies. In the light without bound metal, the Mn-binding mutants showed an EPR spectrum characteristic of the oxidized bacteriochlorophyll dimer and reduced quinone whose intensity was significantly reduced due to the diminished quantum yield of charge separation in the mutant compared to wild type. In the presence of the metal and in the dark, the EPR spectrum measured at the X-band frequency of 9.4 GHz showed a distinctive spin 5/2 Mn(II) signal consisting of 16 lines associated with both allowed and forbidden transitions. Upon illumination, the amplitude of the spectrum is decreased by over 80 % due to oxidation of the metal upon electron transfer to the oxidized bacteriochlorophyll dimer. The EPR spectrum of the Mn-cofactor was also measured at the Q-band frequency of 34 GHz and was better resolved as the signal was composed of the six allowed electronic transitions with only minor contributions from other transitions. A fit of the Q-band EPR spectrum shows that the Mn-cofactor is a high spin Mn(II) species (S = 5/2) that is six-coordinated with an isotropic g-value of 2.0006, a weak zero-field splitting and E/D ratio of approximately 1/3. The ESEEM experiments showed the presence of one ¹⁴N coordinating the Mn-cofactor. The nitrogen atom is assigned to a His by comparing our ESEEM results to those previously reported for Mn(II) ions bound to other proteins and on the basis of the X-ray structure of the M2 mutant that shows the presence of only one His, residue M193, that can coordinate the Mn-cofactor. Together, the data allow the electronic structure and coordination environment of the designed Mn-cofactor in the modified reaction centers to be characterized in detail and compared to those observed in other proteins with Mn-cofactors.     

Protein dynamics and monomer-monomer interactions in AntR activation by electron paramagnetic resonance and double electron-electron resonance

The Anthracis repressor (AntR) is a Mn(II)-activated DNA binding protein that is involved in the regulation of Mn(II) homeostasis in Bacillus anthracis. AntR is structurally and functionally homologous to Mn(II)-activated repressor from Bacillus subtillis (MntR). Our studies on AntR focus on metal-regulated activation of the protein. Line shape analysis of continuous wave electron paramagnetic resonance (EPR) spectra showed that metal binding resulted in a general reduction of backbone dynamics and that there were no further changes in backbone motion upon DNA binding. Double electron-electron resonance (DEER) pulsed EPR spectroscopy was used to measure distances between nitroxide spin labels strategically placed in dimeric AntR. The DEER data were analyzed assuming Gaussian distributions for discrete populations of spins. A structural model for AntR was built from homology to MntR, and the experimentally measured distances were simulated to distinguish between spin label and backbone motions. Together with the computational analysis, the DEER results for apo-AntR indicated relatively narrow conformational distributions for backbone residues at the dimer interface and near the metal binding site. No significant changes were observed on these sites in the presence of metal or DNA. On the other hand, the distribution of the conformers and the distances between the putative DNA binding helices decreased upon metal binding. These results suggest that the DNA binding region of AntR shows large amplitude backbone motions in the absence of metal, which may preclude sequence-specific binding to promoter sites. Metal binding narrows the range of conformations accessible in this region and shortens the mean distance between the DNA binding helices, probably resulting in alignment that optimizes promoter recognition and binding.     

On experimental artifacts in the use of metal ion chelators for the determination of the cation binding constants of alpha-lactalbumin. A reply

The binding constant of Ca2+ to the strong cation site of bovine alpha-lactalbumin has been measured directly by monitoring the free calcium concentration by Quin 2 fluorescence. A dissociation constant of 1-4 nM was calculated, which confirms the strong calcium binding properties of this protein. In order to examine whether the metal ion chelators EDTA or EGTA affect the cation binding equilbria by binding to bovine alpha-lactalbumin, calcium binding equilibria were carefully measured under highly stabilized pH and temperature conditions. Within the concentration ranges required for competitive binding by these ligands (EDTA or EGTA) (less than 1-3 mM) these chelators produced no artifacts, in contradiction to the data of Kronman and Bratcher (Kronman, M. J., and Bratcher, S. C. (1983) J. Biol. Chem. 258, 5707-5709).     

Intramolecular distance measurements in alpha-lactalbumin

The distance between the calcium site (site I) and the zinc site (site II) in alpha-lactalbumin was estimated from Forster energy-transfer measurements between donor Eu(III) [or Tb(III)] at site I and acceptor Co(II) at site II to be 11.5 +/- 1.5 A. Intersite distances were also measured between the bis-ANS [4,4'-bis[1-(phenylamino)-8-naphthalenesulfonate]] binding locus and cobalt at site II (13.6 +/- 1.0 A), between bis-ANS and a fluorescein moiety covalently bound to Met-90 (33.5 +/- 3.0 A), and between Met-90 (fluorescein) and cobalt at site II (16.7 +/- 1.0 A). The apparent Kd for cobalt binding to site II agreed well with the value measured previously by intrinsic fluorescence [Murakami, K., & Berliner, L. J. (1983) Biochemistry 22, 3370-3374]. A Zn(II) titration of Eu(III)-alpha-lactalbumin reconfirmed that both sites I and II can be occupied simultaneously [Musci, G., & Berliner, L. J. (1985) Biochemistry 24, 3852-3856], since the lanthanide fluorescence was unaffected.     

Cadmium-substituted skeletal troponin C metal binding investigations and sequence assignment of the cadmium-113 resonances

The binding of cadmium to the calcium binding subunit of skeletal troponin (STnC) has been reinvestigated using direct binding methods and fluorescent derivatives. These data provide straightforward explanations of the observed titration behavior in the 113Cd NMR (Ellis, P.D., Strang, P., and Potter, J.D. (1984) J. Biol. Chem. 259, 10348-10356). Further, fluorescent derivatives of skeletal troponin C provide an excellent means of establishing a sequence assignment for the resonances observed in the 113Cd NMR. The results of these experiments demonstrate that sites I and II, the Ca2+ regulatory sites, can be assigned to resonances at -108.5 and -101.5 ppm, respectively. Sites III and IV, the structural sites, are assigned to resonances -112.8 and -106.8 ppm, respectively. These data are discussed in terms of recent structural findings and speculations.     

Comparative Cd-113 nuclear magnetic resonance studies of Cd(II)-substituted blue copper proteins

The 113Cd NMR spectra of plastocyanin (Spinacea), stellacyanin (Rhus vernicifera), and two azurins (Pseudomonas aeruginosa and Alcaligenes faecalis) have been measured after introducing Cd(II) into the blue copper-binding sites. Relative to Cd(C1O4)2 the chemical shifts are 432, 380, 372, and 379 ppm, respectively, all of which are found to be reasonable values for binding sites containing a cysteine thiolate ligand. The 113Cd resonances of the cadmium derivatives of stellacyanin and the azurins are so near the same that the proteins must present very similar metal-binding sites. In contrast the plastocyanin derivative resonates about 50 ppm further downfield which may signal a change in coordination number. The spin lattice relaxation times of the 113Cd resonances are of the order of 0.1 s, and a major portion of the relaxation apparently occurs through the chemical shift anisotropy mechanism. At 13 degrees C the 113Cd resonance of Psuedomonas azurin shifts slightly downfield with increasing pH. This is explained by a small change in the environment about cadmium which occurs as a result of the conformational change that attends the titration of His-35.     

Effect of metal ion binding on the secondary structure of bovine alpha-lactalbumin as examined by infrared spectroscopy

We have examined the influence of monovalent and divalent cations on the secondary structure of bovine alpha-lactalbumin at neutral pH using Fourier-transform infrared spectroscopy. Our present studies are based on previously reported amide I' component band assignments for this protein [Prestrelski, S. J., Byler, D. M., & Thompson, M. P. (1991) Int. J. Pept. Protein Res. 37, 508-512]. The results indicate that upon dissolution, alpha-lactalbumin undergoes a small, but significant, time-dependent conformational change, regardless of the ions present. Additionally, these studies provide the first quantitative measure of the well-known secondary structural change which accompanies calcium binding. Results indicate that removal of Ca2+ from holo alpha-lactalbumin results in local unfolding of the Ca(2+)-binding loop; the spectra indicate that approximately 16% of the backbone chain changes from a rigid coordination complex to an unordered loop. We have also examined the effects of binding of several other metal ions. Our studies have revealed that binding of Mn2+ to apo alpha-lactalbumin (Ca(2+)-free), while inducing a small, but significant, conformational change, does not cause the alpha-lactalbumin backbone conformation to change to that of the holo (Ca(2+)-bound) form as characterized by infrared spectroscopy. Similar changes to those induced by Mn2+ are observed upon binding of Na+ to apo alpha-lactalbumin, and furthermore, even at very high concentrations (0.2 M), Na+ does not stabilize a structure similar to the holo form. Binding of Zn2+ to the apo form of alpha-lactalbumin does not result in significant backbone conformational changes, suggesting a rigid Zn(2+)-binding site.(ABSTRACT TRUNCATED AT 250 WORDS)     

Metal ion binding to alpha-lactalbumin species

A strong cation (calcium) binding site has been demonstrated to exist in several alpha-lactalbumin species; bovine, goat, human, and guinea pig. A metal ion induced conformational change occurs, resulting in a unique (10-14-nm) blue shift and relative quenching of Trp fluorescence for all species. Calcium ion binding to the alpha-lactalbumins yielded dissociation constants (Kdiss consistently in the 10(-10)--10(-12) M range, while Mn(II) binding was in the 20-30 microM range. Independent determinations of these cation binding equilibria were made by ESR measurements of free unliganded Mn(II) in titrations with the bovine species. One strong site (Kdiss = 30.5 microM) was found, which correlated directly with the fluorescence-associated cation binding, plus three weaker sites (Kdiss = 1.1, 5.0, and 5.0 mM, respectively). Several lanthanides as well as Mg(II) were found to displace Mn(II) from the strong site on bovine alpha-lactalbumin (as monitored by ESR) and to cause the identical fluorescence changes as found for Ca(II) and Mn(II) above. The importance of measuring these equilibria by both fluorescence and ESR was borne out by demonstrating the potential errors in estimating dissociation equilibria by the fluorescence method alone. Also, the errors in estimating Kdiss for samples containing partially metal bound apo-alpha-lactalbumin are described as well as rapid, sensitive methods for estimating the extent of metal-free protein and correctly accounting for residual bound metal in equilibrium calculations.     

Manganese (II) And spin-labeled uridine 5'-diphosphate binding to bovine galactosyltransferase.

The kinetically observed Mn(II) activation as well as inhibition has been clarified for bovine galactosyltransferase. An electron spin resonance (ESR) titration of MnCl2 with galactosyltransferase alone at pH 8.0 clearly shows the existence of at least two metal ion binding sites with microscopic dissociation constants of 0.84 +/- 0.1 and 9.0 +/- 1.0 mM, respectively. The second site corresponds with either published kinetic constant for Mn(II) of 8.5 mM (inhibition) or 3.40 mM (activation). The contribution of the binary complex Mn(II)-UDPGal is of lesser significance, as concluded by its ESR measured Kdiss of 14.5 +/- 1.1 mM at pH 8.0. A spin-labeled inhibitor analog of UDPgalactose, UDP-4-O-(2,2,6,6-tetramethyl-4-piperidinyl-1-oxy), or UDP-R, was synthesized as a competitive inhibitor for UDPGal. It was shown from inhibition kinetics to be almost as potent an inhibitor as UDPGlu. The Ki values at pH 8.0 in the N-acetyllactosamine and lactose reactions were 0.38 +/- 0.04 and 0.63 +/- 0.06 mM, respectively, as compared with 0.10 +/- 0.01 and 0.094 +/- 0.009 mM for UDPGlu. An ESR titration of UDP-R with galactosyltransferase at pH 8.0 yielded direct physical dissociation constants of 0.40 +/- 0.07 and 0.53 +/- 0.08 mM in the absence and presence of alpha-lactalbumin, respectively. No other substrates (glucose of N-acetylglucosamine) nor Mn(II) were present.     

113Cd-NMR investigation of a cadmium-substituted copper, zinc-containing superoxide dismutase from yeast

113Cd nuclear magnetic resonance spectroscopy has been used to investigate the metal binding sites of cadmium-substituted copper, zinc-containing superoxide dismutase from baker's yeast. NMR signals were obtained for 113Cd(II) at the Cu site as well as for 113Cd(II) at the Zn site. The two subunits in the dimeric enzyme were found to have identical coordination properties towards 113Cd(II) at the Zn site when no copper is coordinated at the Cu site, and when Cu(I) or Cd(II) is coordinated, were found to be very small indicating that 113Cd(II) must be bound to the same number and type of ligands in both cases. Furthermore, the spectra show that the rate of exchange of protein-bound 113Cd(II) and free 113Cd2+ is slow on the NMR time scale also at the Cu site. The present study suggests an explanation for the discrepancy in the literature regarding 113Cd-NMR investigations of bovine superoxide dismutase.     

Methionine-90-spin-labeled bovine alpha-lactalbumin: electron spin resonance and NMR distance measurements

The unique methionine residue of bovine alpha-lactalbumin was modified by irreversible alkylation with the bromoacetamido nitroxide spin-label 4-(2-bromoacetamido)-2,2,6,6-tetramethylpiperidine-N-oxyl. The line shape of the electron spin resonance (ESR) spectrum was indicative of a fairly mobile spin-label and was sensitive to the calcium-induced conformational change. Paramagnetic broadening of the spin-label ESR lines by a Gd(III) ion substituted at the high-affinity calcium site of the protein yielded a distance between the spin-label and the metal-binding site of 8.0 +/- 1.0 A. The extent of the paramagnetic line broadening by the covalently attached nitroxide spin-label on the proton resonances of several amino acid residues of the protein at 500 MHz allowed estimation of intramolecular distances between the methionine-90 residue and several resolvable protons.     

Structural differences in the two calcium binding sites of the porcine intestinal calcium binding protein: a multinuclear NMR study

Cadmium-113 and calcium-43 NMR spectra of Cd2+ and Ca2+ bound to the porcine intestinal calcium binding protein (ICaBP; Mr 9000) contain two resonances. The first resonance is characterized by NMR parameters resembling those found for these cations bound to proteins containing the typical helix-loop-helix calcium binding domains of parvalbumin, calmodulin, and troponin C, which are defined as EF-hands by Kretsinger [Kretsinger, R. H. (1976) Annu. Rev. Biochem. 45, 239]. The second resonance in both spectra has a unique chemical shift and is consequently assigned to the metal ion bound in the N-terminal site of ICaBP. This site is characterized by an insertion of a proline in the loop of the helix-loop-helix domain and will be called the pseudo-EF-hand site. The binding of Cd2+ to the apo form of ICaBP is sequential. The EF-hand site is filled first. Both binding sites have similar, but not identical, affinities for Ca2+: at a Ca2+ to protein ratio of 1:1, 65% of the ion is bound in the EF-hand site and 35% in the pseudo-EF-hand site. The two sites do not appear to act independently; thus, replacement of Ca2+ or Cd2+ by La3+ in the EF-hand site causes changes in the environment of the ions in the pseudo-EF-hand site. In addition, the chemical shift of Cd2+ bound to the EF-hand site is dependent on the presence or absence of Ca2+ or Cd2+ in the pseudo-EF-hand site.(ABSTRACT TRUNCATED AT 250 WORDS)