Electron Paramagnetic Resonance Spectroscopy of Nitroxide-Labeled Calmodulin

5

Supplement

Electron Paramagnetic Resonance Spectroscopy of Nitroxide-Labeled Calmodulin

Paula B. Bowmana,* and David Puettb,c,**

aDepartment of Biochemistry, Vanderbilt University School of Medicine, Nashville, TN 37232; bDepartment of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602; cDepartment of Biochemistry and Biophysics, University of North Carolina School of Medicine, Chapel Hill, NC 27599

*Formerly Paula B. Hewgley.

**To whom correspondence should be addressed. Email:

1 Methods

1.1 Determination of Free Ca2+ Concentrations from EGTA/CaCl2 Buffer Solutions

Free Ca2+ concentrations were obtained by using a 2 mM EGTA/CaCl2 system buffered with 50 mM Tris-HCl, pH 7.5. The association constants for EGTA equilibria and different Ca2+/EGTA complexes were taken from the early literature [1-3]. Although EGTA4- is the major species that binds Ca2+, we also included HEGTA3- as it also binds Ca2+. Beginning with 10 unknowns: H4E, H3E-, H2E-2, HE-3, E-4, H+, C, CE-2, CHE-, and CH2E, where E = EGTA and C = Ca2+, one can develop 10 equations that are simplified somewhat by taking (H4E) and (CH2E) = 0 and (H+) = constant. With appropriate substitutions the following relationship (equation 1) is obtained where (Ca2+)t = total Ca2+ concentration, (Ca2+)f = free Ca2+ concentration, and (EGTA)t = total EGTA concentration.

Equation 1: (Ca2+)t = (Ca2+)f {1 + A/B)}, where A and B are as follows.

A = {[100.98/(H+)] + 105.32}(EGTA)t

B = [104.53 (H+)] + [108.85 (H+)] + 1 + [10-9.43/ (H+)] + (Ca2+)f [{100.98/(H+)} + 105.32]

One can solve for (Ca2+)f using an appropriate program or an iterative or graphical procedure.

1.2 Chemical and Physicochemical Characterization of CaM and SL-CaM

Standard techniques were used for polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, isoelectric focusing, and ultracentrifugal analysis. Amino acid compositions were determined using acid-hydrolyzed protein and Beckman and Durrum analyzers with appropriate standards, including the unmodified amino acids and 1-carboxyl-methylhistidine, 3-carboxylmethylhistidine, dicarboxymethylhistidine, homoserine, homoserine lactone, S-carboxymethylhomo-cysteine (the major decomposition product of carboxymethyl-methionine when heated in acid), tricarboxylmethylhomocysteine, and methionine sulfone.

The molecular weight and diffusion coefficient (D) of CaM (2 mg/mL) in 20 mM Tris-HCl, pH 7.0 at 20oC was determined with established methods on a Beckman Model E analytical ultracentrifuge operated at 22,000 and 36,000 rpm using scanner optics and an An-F rotor. The Mr was calculated from equation 2 below, where Cr is the protein concentration at radial distance r, ω is the angular velocity, υ the partial specific volume, and ρ the density. D was measured under the same conditions used for determining Mr (rotor speed 22,000 rpm) and was determined from the slope of a plot of {ln [(ΔCt - ΔCeq)/ΔCeq]} vs. time.

Equation 2: Mr = 2RT {d[ln(Cr)/dr2]/[ω2 (1-υρ)]}

1.3 X-band EPR Spectroscopy

In order to limit molecular motion, spectra were also collected on SL-CaM precipitated with (NH4)2SO4. SL-CaM (0.44 mg) was dissolved in each of the following buffers: 1 mL of 100 mM Tris-HCl, pH 7.5, and 50 mM CaCl2, and in 1 mL of 100 mM Tris-HCl, pH 7.5, and 100 mM EGTA. Each of the samples was gradually saturated at 0oC with (NH4)2SO4 (3.88 M) and shielded from light by wrapping in aluminum foil. The samples were centrifuged at 3,000 x g, and 0.25 mL of each of the buffers saturated with (NH4)2SO4 was used to suspend the pellets and introduce them into the flat cells.

2 Results and Discussion

2.1 Characterization of CaM

The purified calmodulin was found to exhibit a single band on SDS-PAGE and had an isoelectric pH of 4.2. Since gel filtration and SDS-PAGE are not reliable methods for obtaining the molecular weight (Mr) of calmodulin, sedimentation equilibrium was done. The partial specific volume, υ, was 0.72 cc/g based on the amino acid composition and 0.73 cc/g as determined by the H2O/D2O technique. The measured Mr was 16,724 ± 114 (υ = 0.72 cc/g) and 17,484 ± 119 (υ = 0.73 cc/g), in excellent agreement with the expected value of 16,680 minus Ca2+ or 16,840 with four Ca2+ ions bound. D was found to be 1.3 x 10-6 cm2/sec, and the extinction coefficient was determined as 0.205 at 280 nm (1 mg/mL, 1 cm; minor corrections were made for light scattering). These values are in good agreement with the many reports in the literature and demonstrate, along with the amino acid composition, the homogeneity of the purified protein.

2.2 Chemical Characterization of SL-CaM

Under the conditions used for spin labeling, one would expect that histidine and/or methionine would be preferentially alkylated. Using a Durrum single-column analyzer for acid-hydrolyzed SL-CaM, the mole % of histidine was found to be 0.7%, in excellent agreement with that expected for the single histidine. Thus, it is concluded that the histidine is not modified. SL-CaM was performic acid-oxidized to convert methionines to methionine sulfone, a derivative stable to acid hydrolysis. A combination of single and double-column analyzers was used to show that, under the conditions used, an average of 2.7 methionyl residues were modified by the spin-labeling reagent. Whether the degree of modification reflects three nitroxides incorporated or a combination of derivatives, some with two nitroxides and more with three nitroxides, is not known at this time. Clearly, however, the degree of modification under the conditions used exceeds one per protein molecule.

2.3 CD Spectroscopy of CaM and SL-CaM

The CD spectra of CaM and SL-CaM ± Ca2+ are presented in Fig. 1, along with spectra of the denatured protein (see main paper for a description of the methods and instrument used). As documented in many reports, the secondary structure of CaM increases with Ca2+ binding, and this is true for SL-CaM as well, this being reflected in the minima of the n-π* and lower

energy π-π* bands at 222 and 208.5 nm, respectively.

Fig. 1 CD spectra of CaM (A) and SL-CaM (B) in 50 mM Tris-HCl, pH 7.5 ± Ca2+. A. 2.0 mM EGTA (0 mM Ca2+), ----; 1.2 mM Ca2+, ____ . B. 2.0 mM EGTA (0 mM Ca2+), ----; 0.21 μM Ca2+, __ __ ; and 1.2 mM Ca2+, ____ . Spectra are also shown for CaM and SL-CaM in 7 M guanidinium chloride, 4 mM Tris-HCl, and 2.0 mM EGTA, pH 7.5 ( ). Protein solutions were between 0.24 – 0.26 mg/mL, and replicate scans gave [Θ]222 nm values that did not exceed 500 deg cm2/dmol.

2.4 EPR Spectroscopy of SL-CaM

Fig. 2 shows the relative spectral intensity of the

sum of the three nitroxide bands as a function of SL-CaM concentration in the presence of EGTA and Ca2+. The linearity of the data shows that the decrease in intensity concomitant with Ca2+ binding represents an intramolecular and not an intermolecular effect associated with the suggested Heisenberg spin exchange and possible dipole-dipole coupling.

Fig. 2 Spectral intensity of SL-CaM ± Ca2+ at various protein concentrations. The intensity (area) of each Lorentzian peak was determined from the relationship, (π/√3)hw2, where h is the peak-to-peak height and w the band width, the total intensity being the sum of the three areas. The samples were in 20 mM Tris-HCl, pH 7.5, containing 2 mg/mL of bovine serum albumin and either 1 mM Ca2+ (open squares) or 10 mM EGTA (open circles).

To ensure that we were collecting data in a linear region of power, the high field peak height was plotted as a function of the square root of the instrument power for CaM ± Ca2+ (Fig. 3). This appears to be the case for the power (5 mW) used in the present study. A normalization of the data to the peak height at 150 mW results in identical responses for SL-CaM ± Ca2+. These results indicate that the apparent spin-lattice relaxation time, T1, is not appreciably affected when Ca2+ binds to CaM.

The peak height ratios of low to center field and high to center field are given in Fig. 4 for free spin label and SL-CaM as a function of free pCa2+. For free spin label the peak height ratios are independent of Ca2+. With SL-CaM, however, there is some evidence of two transitions in h+1/ho vs. Ca2+; in contrast, the h-1/ho ratio exhibits linearity and slightly increases with higher Ca2+ concentrations.

Fig. 5 shows the widths of the low, center, and high field peaks of SL-CaM at various concentrations of Ca2+. The low field (+1) and center field (o) peak widths have no discernable deviation from linearity, both exhibiting a very small negative slope with increasing Ca2+ concentrations over a broad range, ca. 0 – 8.1 mM. The high field peak width, in contrast, is linear from 0 Ca2+ to a p(Ca2+)f of about 6.7 - 7; at higher concentrations the high field peak width decreases in a linear manner with increasing concentrations of Ca2+. The peak widths of the free spin label are invariant to Ca2+ over the broad concentration range investigated.

Fig. 3 Plots of the height of the high field nitroxide EPR peak as a function of the square root of power for SL-CaM ± Ca2+. The open circles refer to spin-labeled CaM in the presence of 10 mM EGTA; the open squares denote 1 mM Ca2+ (0 mM EGTA). All samples were in 20 mM Tris-HCl, pH 7.5, and 2 mg/mL bovine serum albumin. The inset shows the data normalized to the respective peak height at 150 mW.

Fig. 4 EPR peak height ratios as a function of free Ca2+ concentration. Peak height ratios of the low field to center field peak (A) and the high field to center field peak (B) for SL-CaM (closed circles) and free spin label (open circles) are shown as function of Ca2+ concentration. The samples were in 50 mM Tris-HCl, pH 7.5 with 2 mM EGTA (squares) or EGTA/Ca2+ (circles). The arrows refer to the transition midpoint determined by [Θ]222 nm; the p(Ca2+)f values corresponding to the minor changes in the (h+1/h0) ratio are 6.19 and 6.49. The abscissa is –log10 (Ca2+)f with Ca2+ in M.

The EPR spectra of (NH4)2SO4-precipitated SL-CaM ± Ca2+ are shown in Fig. 6 where it can be seen that the low and high field resonance peaks are broader in the presence of Ca2+.

Fig. 5 EPR resonance peak widths as a function of free Ca2+ concentration. Widths of the low field (A), center field (B), and high field (C) peaks are plotted vs. concentration of free Ca2+ for SL-CaM (closed circles) and free spin label (open circles). The samples were in 50 mM Tris-HCl, pH 7.5, and the square symbols denote values in 2 mM EGTA. The transition midpoint as determined by CD spectroscopy (222 nm) is indicated by the arrows. The X-axis is –log10 (Ca2+)f with Ca2+ in M.

Fig. 6 EPR spectra of (NH4)2SO4-precipitated SL-CaM ± Ca2+ (see above). We thank Prof. Albert Beth (Vanderbilt University) for obtaining these spectra.

2.5 Distances Between Methionines in CaM

Table 1 shows the distances between the methionyl sulfurs in apo-CaM and (Ca2+)4-CaM. Since there are nine methionines, there are 36 pairwise distances.

References

1. Godt RE (1974) Calcium-activated tension of skinned muscle fibers of the frog: Dependence on magnesium adenosine triphosphate concentration. J Gen Physiol 63: 722-739.

2. Owen JD (1976) The stability constant for calcium EGTA. Biochim Biophys Acta 451: 321-325.

3. Sillen LG, Martell AE (1991) Stability constants of metal-ion complexes. Royal Society of Chemistry, Great Britain, London (2nd edition and earlier editions).

4. Kuboniwa H, Tjandra N, Grzesiek S, Ren H, Klee CB, Bax A (1995) Solution structure of calcium-free calmodulin. Nat Struct Biol 2, 768-776.

5. Chattopadhyaya R, Meador WE, Means AR, Quiocho FA (1992) Calmodulin structure refined at 1.7 Å resolution. J Mol Biol 228, 1177-1192.

Table 1 Pairwise distances between the methionyl sulfurs in

apo-CaM and (Ca2+)4-CaMa

Methionines / Distance (Å) / Methionines / Distance (Å)
36 - 51 / 6.4 → 5.0 / 71 - 124 / 33.2 → 37.6
36 - 71 / 6.5 → 10.9 / 71 - 144 / 34.1 → 30.3
36 - 72 / 4.0 → 11.7 / 71 - 145 / 30.6 → 26.7
36 - 76 / 9.4 → 15.5 / 72 - 76 / 6.8 → 5.5
36 - 109 / 31.5 → 40.1 / 72 - 109 / 29.1 → 35.7
36 - 124 / 30.7 → 40.9 / 72 - 124 / 27.8 → 37.3
36 - 144 / 32.5 → 33.8 / 72 - 144 / 29.5 → 30.9
36 - 145 / 27.5 → 30.8 / 72 - 145 / 25.0 → 26.4
51 - 71 / 3.6 → 9.9 / 76 - 109 / 22.3 → 31.3
51 - 72 / 8.1 → 12.5 / 76 - 124 / 21.3 → 33.4
51 - 76 / 12.7 → 16.6 / 76 - 144 / 23.2 → 28.1
51 - 109 / 34.1 → 39.7 / 76 - 145 / 18.3 → 22.7
51 - 124 / 33.3 → 40.0 / 109 - 124 / 5.9 → 5.2
51 - 144 / 34.1 → 32.1 / 109 - 144 / 8.5 → 12.3
51 - 145 / 30.3 → 30.0 / 109 - 145 / 4.3 → 10.3
71 - 72 / 6.9 → 4.6 / 124 - 144 / 5.6 → 9.6
71 - 76 / 12.7 → 9.1 / 124 - 145 / 5.1 → 11.0
71 – 109 / 34.6 → 36.5 / 144 - 145 / 8.6 → 7.2

aThe structures were from the Protein Data Bank (1CFD for apo-CaM and 1CLL for

(Ca2+)4-CaM), and distances were determined using Chimera. The NMR (apo-CaM)

and X-ray crystallographic [(Ca2+)4-CaM] structures were from [4] and [5], respectively.

In the two `Distance` columns the first and second numbers give the separation

distances of the two methionines in apo-CaM and (Ca2+)4-CaM, respectively.