and Temperature-Dependent 13C NMR Studies of the EDTA–Zn2+

Article pubs.acs.org/JPCB

Field- and Temperature-Dependent 13C NMR Studies of the EDTA− Zn2+ Complex: Insight into Structure and Dynamics via Relaxation Measurements Sriramya Garapati, Colin S. Burns,‡ and A. A. Rodriguez*,‡ Department of Chemistry, East Carolina University, Greenville, North Carolina 27858, United States ABSTRACT: The relaxation rates for the three different carbon types in EDTA (carbonyl, CH2 central, and CH2 lateral) were measured with and without Zn2+ as a function of field strength and temperature. The use of different field strengths in combination with NOE measurements allowed for the contribution of each relaxation mechanism (chemical shift anisotropy; spin rotation; dipole−dipole) to the total relaxation rate for each carbon to be determined. Temperature studies allowed for determination of the activation energy (Ea) for the motions of each carbon type. The most surprising result was the observation that the τc decreases significantly for the lateral carbon upon addition of Zn2+ at neutral pH, going from 54 to 8.6 ps at 298 K. This appears to be a pH-dependent phenomenon as other reports indicate that τc increases for the lateral carbon upon addition of Zn2+ under strongly basic conditions.

■

INTRODUCTION Ethylenediaminetetraacetic acid (EDTA) is a widely used complexing and sequestering agent. Its ability to act as a multidentate ligand enables it to bind a wide range of cations with high affinity. Here we turn our attention to the measurement of the relaxation rates (R1) of the carbon atoms in EDTA as a function of complexation with divalent zinc (Zn2+). The structure of the EDTA−Zn complex is know from crystallographic studies, where the Zn2+ ion is coordinated in a hexadentate manner resulting in a structure with octahedral geometry.1 However, in solution, the number of ligands provided by EDTA may change (e.g., acting as a pentadentate chelator), and rapid exchange of ligands may occur leading to interconversion between configurations (e.g., Δ and Λ configurations for hexacoordinate EDTA isomers).2 The degree of ligation and rate of configurational interconversion will depend on the solution conditions and most especially on the pH.3 Both 1H- and 13C NMR provide powerful means of gaining insight into the structure of such complexes and intramolecular processes occurring in them. Although the nuclides detected by these methods do not directly coordinate Zn2+, they are close enough to the metal-ligating atoms (i.e., N and O) as to experience measurable changes in their chemical shifts (δ) and R1 relaxation rates in the metal-bound versus unbound forms. Thus, these values reveal key features about the structure and dynamics of the EDTA−Zn complex. Most solution studies of EDTA−metal complexes are performed under considerably basic conditions so as to work with the fully deprotonated form of EDTA.4 In this study, 13C NMR measurements were performed near neutral pH (pH = © 2014 American Chemical Society

7.2), as this condition is more relevant to biological systems where EDTA may be used to remove ions from metalloproteins or in chelation therapy. Further, NMR spectra were collected at different field strengths and at different temperatures. By performing measurements at different magnetic field strengths, the contribution of each relaxation mechanism (chemical shift anisotropy, CSA; spin rotation, SR; dipole−dipole, DD) to the total relaxation rate for each carbon type (carbonyl, CH2 central, and CH2 lateral) was determined. The temperature study allowed for determination of the activation energy (Ea) of the motions for each carbon type. The results are discussed and compared to those collected under very basic (pH = 12) conditions.4 Surprisingly, our results reveal that the lateral methylene groups increase their reorientation rate upon the introduction of Zn2+ to the EDTA solution.

■

EXPERIMENTAL SECTION Ethylenediaminetetraacetic acid (≥99%), deuterium oxide (99.9% atom D), and zinc dichloride (99.999%) were purchased from Sigma-Aldrich (St. Louis, MO); sodium phosphate monobasic (>98%) and sodium phosphate dibasic (>98%) were purchased from Fisher Scientific (Pittsburgh, PA). All compounds were used as received without further purification. All NMR samples contained 60 mM EDTA dissolved in 50 mM phosphate buffer at pH = 7.2. For the Zn2+ containing samples, ZnCl2 was added to a concentration of 90 Received: July 30, 2014 Revised: October 17, 2014 Published: October 17, 2014 12960

dx.doi.org/10.1021/jp507674a | J. Phys. Chem. B 2014, 118, 12960−12964

The Journal of Physical Chemistry B

Article

Table 1. Spin−Lattice Relaxation Rates, CSA and SR Contributions, and Correlation Times of Carbonyl Carbon at Different Field Strengths and as a Function of Temperature of EDTAa ν (MHz)

T (K) 298

125.2 100.6 125.2 100.6 125.2 100.6

313 323 a

R1CSA (1/s)

R1 (1/s) 0.175 0.104 0.113 0.081 0.087 0.068

−4

(5.8 × 10 ) (0.002) (0.0) (0.002) (2.7 × 10−3) (0.0)

0.150 0.082 0.088 0.056 0.053 0.034

R1SR (1/s)

(0.006) (0.006) (0.006) (0.006) (0.002) (0.002)

0.025 0.022 0.025 0.025 0.034 0.034

(0.006) (0.006) (0.006) (0.006) (0.005) (0.005)

% CSA

% SR

τc (ps)

85.7 78.8 77.9 69.1 60.9 50.0

14.3 21.2 22.1 30.9 39.1 50.0

105 (4) 61 (4) 36 (2)

The numbers in parentheses represent one standard deviation of a minimum of three separate measurements.

Table 2. Spin−Lattice Relaxation Rates, CSA and SR Contributions, and Correlation Times of Carbonyl Carbon at Different Field Strengths and as a Function of Temperature of EDTA with Zn2+a T (K)

ν (MHz)

298

125.2 100.6 125.2 100.6 125.2 100.6

313 323 a

R1 (1/s) 0.214 0.130 0.135 0.089 0.094 0.076

(0.005) (0.0) (0.007) (0.0) (6.0 × 10−4) (6.0 × 10−4)

R1CSA (1/s) 0.195 0.111 0.129 0.083 0.051 0.033

R1SR (1/s)

(0.042) (0.042) (0.006) (0.006) (0.002) (0.002)

0.019 0.019 0.129 0.006 0.043 0.043

(0.019) (0.046) (0.009) (0.009) (0.002) (0.002)

% CSA

% SR

τc (ps)

91.1 85.4 95.6 93.3 54.3 43.4

8.9 14.6 4.4 6.7 45.7 56.6

165 (4) 110 (5) 44 (1)

The numbers in parentheses represent one standard deviation of a minimum of three separate measurements.

reorientational correlation time from the slope. The SR contribution can be deduced from either the intercept and/or from the difference between R1 and R1CSA. The various relaxation contributions to the carbonyl carbon along with the reorientational times for EDTA with and without Zn2+ are given on Tables 1 and 2. The lateral and central carbons possess two hydrogens each, which will lead to dipole−dipole interactions. The rate of spin− lattice relaxation of these two carbons proceeds primarily via the dipole−dipole (R1DD) and spin rotation (R1SR) mechanisms:

mM. The NMR samples were degassed by three successive free-thaw cycles and the NMR tubes flame-sealed. NMR spectra were collected on a 100.6 MHz (B0 = 9.4 T, 400 MHz 1H frequency) spectrometer (Bruker, Billerica, MA), and a 125.2 MHz (B0 = 11.7 T, 500 MHz MHz 1H frequency) spectrometer (Varian/Agilent Technologies, Santa Clara, CA). Data were collected using the inversion recovery method, and nonlinear least-squares procedures were used for the analysis of T1 measurements. The NOE factors were determined at 125.2 MHz by setting the decoupler on-resonance and far offresonance in successive experiments. Temperature control was provided by an XR401 Air-Jet Crystal Cooler (FTS Systems, Stone Ridge, NY).

R1 = R1DD + R1SR

■

The dipole−dipole contribution can be experimentally isolated through the use of NOE (ηNOE) measurements and eq 5:5,7 η R1DD = NOE R1 (5) 1.987

THEORY The carbonyl carbon’s spin−lattice relaxation mechanism is known to involve the chemical shift anisotropy (R1CSA) and spin rotation pathways (R1SR), as given in eq 1: R1 = R1CSA + R1SR

(1)

The spin rotation contribution is obtained from the difference between R1 and R1DD. The various relaxation contributions to the lateral and central carbons along with the reorientational correlation times for EDTA with and without Zn2+ are given in Tables 4 and 5. The reorientational correlation times, τc, are extracted from the dipole−dipole contribution, R1DD, to the overall relaxation rate using eq 6:5,7

Under extreme narrowing conditions and assuming a symmetric shielding tensor, the chemical shift anisotropy is given by eq 2:5−7 R1CSA =

⎛2 ⎞ 2 2 ⎜ ⎟γ B Δσ τ C ⎝ 15 ⎠ O

(2)

In eq 2, γ is the carbon gyromagnetic ratio, Bo is the field strength, Δσ is the shielding anisotropy (132 ppm in EDTA and 119 ppm in EDTA/Zn2+),4 and τc is the reorientational correlation time. By combining eqs 1 and 2, one sees an experimental method of separating the CSA contribution from the overall relaxation via the employment of multiple instruments operating at different field strengths, ΔBo: ΔR1 =

⎛2 ⎞ 2 2 SR ⎜ ⎟γ ΔB Δσ τ + R O C 1 ⎝ 15 ⎠

(4)

R1DD =

2 2 ⎛ μo ⎞2 ⎛ h ⎞2 ⎛ γC γH ⎞ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟N τ 6 ⎟ H C ⎝ 4π ⎠ ⎝ 2π ⎠ ⎜⎝ rCH ⎠

(6)

where γC and γH are the gyromagnetic ratios for C and 1H nuclei, respectively, NH is the number of hydrogens bonded to each carbon atom, while rCH is the average carbon−hydrogen bond length (1.09 × 10−10 m).4 13

■

(3)

RESULTS AND DISCUSSION Relaxation Rates of the Carbonyl Carbon at Different Field Strengths and as a Function of Temperature in

As can be seen from eq 3, a linear fit of the relaxation rates as a function of two field strengths, ΔR1 versus ΔBo, will yield the 12961

dx.doi.org/10.1021/jp507674a | J. Phys. Chem. B 2014, 118, 12960−12964

The Journal of Physical Chemistry B

Article

EDTA without and with the Presence of Zn2+ in D2O. EDTA is a simple carbon-13 spin system with three carbon resonances appearing in the 13C NMR spectrum at 175.7 ppm (carbonyl), 57.5 ppm (lateral), and 51.2 ppm (central), while in the presence of Zn2+, the resonances shifted downfield to 178.1, 61.1, and 55.9 ppm, respectively. The fact that three different resonances are observed for the EDTA−Zn2+ complex indicates that even if a hexadentate complex is forming with octahedral symmetry, the acetate groups (which contain the lateral methylene units) must be rapidly exchanging so as to give an average chemical shift value on the NMR timescale. Figure 1 shows the structures of EDTA and EDTA−Zn expected at pH 7.2.1,8

Table 3. Activation Energies of the Carbonyl, Lateral, and Central Carbons of EDTA without and with Zn2+ at 125.2 MHz (11.7 T) and 298 K activation energy (kJ/mol) carbon

without Zn2+

with Zn2+

carbonyl lateral central

33.7 22.1 9.7

40.3 20.4 11.4

the NOE values (ηNOE), the dipole−dipole (R1DD), and spin rotation (R1SR) contributions as well as the correlation times at these two sites. This table illustrates the overall relaxation rates decrease only slightly with rising temperature with the dipole− dipole mechanism dominating at all temperatures. As anticipated, the spin rotation contribution gradually increases with rising temperature. The final column in this table lists the correlation times which provide a glimpse of the dynamics at these two sites. One sees a larger increase in the reorientational motion of the lateral carbon with rising temperature as compared to the central carbon. An Arrhenius fit of the correlation times versus T−1, yielded energies of activation of 22.1 and 9.7 kJ/mol for the lateral and central carbons, respectively. These activation energies indicate greater rotational freedom at the central versus the lateral carbon site. The experimental spin−lattice relaxation rates (R1), the NOE values (ηNOE), the dipole−dipole (R1DD), and spin rotation (R1SR) contributions as well as the correlation times of the lateral and central carbons upon the introduction of Zn2+ to the EDTA solution are found in Table 5. The data in this table illustrate a more dramatic change in all the experimental observables with the introduction of Zn2+ to the solution. In particular, the overall relaxation rates of the lateral carbon experience larger changes with increasing temperature. In fact, and surprisingly, the dipole−dipole mechanism becomes less efficient than spin rotation at all temperatures, suggesting an increase in the reorientational motion at this carbon site which is supported by the reduction of τc values with the introduction of Zn2+. An Arrhenius fit of this carbon’s correlation times versus T−1 yielded an energy of activation of 20.4 kJ/mol, approximately 1.7 kJ/mol less than without the presence of Zn2+. This observation is conceivable if EDTA and Zn2+ interactions are simultaneously occurring at the carbonyl and −CH2−N− (central site), resulting in enhanced rotational freedom (i.e., “quasi” free spinning) at the lateral methylene carbon site. The overall relaxation rate of the central methylene group decreases incrementally with rising temperature with the dipole−dipole mechanism dominating at all temperatures. A comparison of the correlation times, without (Table 4) and with Zn2+ (Table 5), shows a slowing of the reorientational motion of this methylene group upon the introduction of the Zn2+ ion. The τc data are congruous with coordination of Zn2+ by the −CH2−N− sites. When complexed with Zn2+, the Arrhenius fit of the central carbon’s correlation times versus T−1 yielded an energy of activation of 11.4 kJ/mol. This reflects an increase of 1.7 kJ/mol in the activation energy for the central carbon compared to the EDTA sample without Zn2+. The most surprising result is the observation that the τc decreases significantly for the lateral carbon upon addition of Zn2+, going from 54 to 8.6 ps at 298 K, whereas that of the central carbon increases modestly, going from 42 to 56 ps. Interestingly, the contribution of R1SR for the lateral carbon

Figure 1. Structures of (a) EDTA and the (b) EDTA−Zn complex expected in aqueous solution at pH = 7.2. The three different types of carbons in EDTA are indicated (A, B, or C) on the structure.

Shown in Table 1 are the variable temperature relaxation rates (R1), the chemical shift anisotropy (R1CSA) and spin rotation (R1SR) contributions, as well as the correlation times (τc) calculated via eqs 1, 3, and 6 of the carbonyl carbon of EDTA in the absence of Zn2+. As expected, these rates are seen to decrease with rising temperature, indicating a decrease in the effectiveness of the dominating mechanism, CSA. The SR mechanism displays the opposite trend, increasing with rising temperature. The last column of this table contains the values for the correlation times obtained via the CSA contribution and eq 2. The correlation times are seen to decrease with rising temperature, indicating faster molecular dynamics at the carbonyl site with increasing temperature. Table 2 lists the same variable temperature data but with the addition of Zn2+ to the solution. A comparison of the relaxation rates from these two tables reveals that the introduction of Zn2+ to the solution does affect the relaxation rates of the carbonyl carbon appreciatively. One notices an increase in the overall relaxation rate and a similar tendency for the CSA contribution. At the lower field strength, the data show the SR mechanism dominates at 323 K. A comparison of the correlation times in Tables 1 and 2 indicate that the addition of Zn2+ does lead to noticeable reduction of the molecular dynamics at the carbonyl site congruous with coordination of Zn2+ via the carboxylate oxygen. A complementary change in the activation energies at this site are also observed. An Arrhenius fit of τc versus T−1 yielded energies of activation of 33.7 and 40.3 kJ/mol without and with Zn2+, respectively; thus, the activation energy for reorientation of the carbonyl carbon increases by 6.6 kJ/mol upon Zn2+ coordination by EDTA (see Table 3). Relaxation Rates of the Lateral and Central Carbons as a Function of Temperature in EDTA without and with the Presence of Zn2+ in D2O. Table 4 lists the overall spin− lattice relaxation rates (R1) of the lateral and central carbons, 12962

dx.doi.org/10.1021/jp507674a | J. Phys. Chem. B 2014, 118, 12960−12964

The Journal of Physical Chemistry B

Article

Table 4. Spin−Lattice Relaxation Rates, NOE, Dipole−Dipole and Spin Rotation Contributions, and Correlation Times of Lateral and Central Carbons as a Function of Temperature of EDTA at 125.2 MHz (11.7 T)a T (K)

carbon

298

lateral central lateral central lateral central

313 323 a

R1 (1/s) 2.94 2.94 2.04 2.04 1.61 2.00

(0.05) (0.05) (0.02) (0.02) (0.08) (0.08)

ηNOE 1.57 1.22 1.51 1.49 1.44 1.33

(0.01) (0.02) (0.01) (0.02) (0.01) (0.02)

R1DD (1/s)

R1SR (1/s)

2.32 1.80 1.55 1.53 1.17 1.34

0.62 1.14 0.49 0.51 0.44 0.66

(0.05) (0.2) (0.07) (0.16) (0.07) (0.19)

(0.05) (0.2) (0.07) (0.16) (0.07) (0.19)

% DD

% SR

τc (ps)

78.9 61.2 76.0 75.0 72.7 67.0

21.1 38.8 24.0 25.0 27.3 33.0

54 42 36 35 27 31

(2) (4) (2) (4) (2) (5)

Number in parentheses represent one standard deviation of a minimum of three separate measurements.

Table 5. Spin−Lattice Relaxation Rates, NOE, Dipole−Dipole and Spin Rotation Contributions, and Correlation Times of Lateral and Central Carbons as a Function of Temperature of EDTA with Zn2+ at 125.2 MHz (11.7 T)a T (K)

carbon

298

lateral central lateral central lateral central

313 323 a

R1 (1/s) 3.13 4.00 1.30 3.23 0.73 1.96

(0.34) (0.34) (0.10) (0.12) (0.02) (0.02)

ηNOE 0.24 1.19 0.35 1.21 0.54 1.71

(0.01) (0.03) (0.02) (0.02) (0.03) (0.03)

R1DD (1/s)

R1SR (1/s)

0.37 2.40 0.23 1.97 0.20 1.68

2.76 1.60 1.07 1.26 0.53 0.28

(0.21) (0.35) (0.11) (0.51) (0.05) (0.29)

(0.21) (0.35) (0.11) (0.51) (0.05) (0.29)

% DD

% SR

τc (ps)

11.8 60.0 17.7 61.0 27.4 84.8

88.2 40.0 82.3 39.0 72.6 15.2

8.6 (5) 56 (8) 5.4 (3) 46 (9) 4.6 (1) 39 (6)

Number in parentheses represent one standard deviation of a minimum of three separate measurements.

as a function of field strength and temperature. Each carbon type (carbonyl, central, lateral) showed a measurable change in correlation time upon the complexation of Zn2+ by EDTA. Even though the carbon atoms in EDTA do not directly coordinate Zn2+, they are close enough to the metal-ligating atoms (i.e., N and O) to experience measurable changes in their relaxation times and can report on the molecular dynamics. These measurements showed that the chemical shift anisotropy mechanism is the dominating mechanism in the carbonyl carbon relaxation process. Surprisingly, our results show that the lateral methylene groups increase their reorientational motion upon the introduction of Zn2+ to the solution, suggesting the anchoring of the carbonyl and the central methylene sites give rise to increased reorientational freedom at the lateral methylene sites.

increases dramatically, going from 21.1% to 88.2% upon introduction of Zn2+. In contrast, the contribution of R1SR for the central carbon remains essentially constant, going from 38.8% to 40.0%. At pH = 12 and 298 K, the τc of both the lateral and central carbons decreases modestly upon addition of Zn2+, with the lateral carbon dropping from 64 to 57 ps and the central carbon dropping from 75 to 58 ps (it should be noted these experiments were run at 100.6 MHz). Here, the separation of the different relaxation pathways may be providing deeper insight into the internal motions of the complex at different pH values. Upon complexation of Zn2+ by EDTA at pH = 7.2, the lateral carbons appear to gain the ability to rotate more freely, as revealed by the large decrease in τc and concomitant increase in the R1SR contribution. This increase in motion for the lateral carbon is difficult to envision for the EDTA−Zn2+ complex if the carboxylates and nitrogen atoms remain coordinated to Zn2+, thus forming five-membered chelate rings. Studies employing radioisotopes have shown that Zn2+ bound by EDTA readily exchanges with any unchelated Zn2+.9 As the exchange process must at some point necessarily involve breaking of the bonds joining EDTA to Zn2+ (i.e., N− Zn and O−Zn), this may afford an opportunity for a carboxylate arm, for instance, to dissociate from one Zn2+ ion and coordinate to another Zn2+ ion, thus forming a bridged complex. This situation would lead to the temporary opening of a five-membered chelate ring, which could account for an increase in the rotational motion of the lateral carbon. In turn, this leads to an increase in the R1SR contribution, as the methylene protons have more freedom to rotate. As exchange processes are pH-dependent, with rates enhanced by lower pH, this would explain the differing results between our studies at pH = 7.2 and the other performed at pH = 12.4,9 However, at this time we are unable to account for why the lateral carbons of EDTA at pH 7.2 have a significantly smaller τc value in the EDTA−Zn2+ complex compared to the free ligand.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: 252-328-9804. Author Contributions ‡

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. C.S.B. and A.A.R. contributed equally. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge support for this work by a graduate teaching assistantship from the East Carolina University Division of Research and Graduate Studies (S.G.).

■

REFERENCES

(1) Solans, X.; Font-Altaba, M. Crystal Structures of Ethylenediaminetetraacetato Metal Complexes. I. A Comparison of Crystal Stuctures Containing Hexacoordinated Metal Ions, [(H2O)4X(C10H12N2O8)Y]n 2nH2O. Acta Crystallogr. 1983, C39, 435−438. (2) Gennaro, M. C.; Mirti, P.; Casalino, C. NMR Study of Intramolecular Processes in EDTA Metal Complexes. Polyhedron 1983, 2, 13−18.

■

CONCLUSIONS Spin−lattice relaxation measurements were conducted on EDTA/D2O solutions as well as EDTA/Zn2+/D2O solutions 12963

dx.doi.org/10.1021/jp507674a | J. Phys. Chem. B 2014, 118, 12960−12964

The Journal of Physical Chemistry B

Article

(3) Kula, R. J.; Sawyer, D. T.; Chan, S. I.; Finley, C. M. Nuclear Magnetic Resonance Studies of Metal-Ethylenediaminetetraacetic Acid Complexes. J. Am. Chem. Soc. 1963, 85, 2930−2936. (4) Abdallaoui, H. E. A. E; Champmartin, D.; Rubini, P. Complexes of EDTA in Aqueous Solutions. Structural Aspects from a 13C NMR Relaxation Study. J. Chem. Soc., Dalton Trans. 2001, 2153−2156. (5) Becker, E. D. In High Resolution NMR: Theory and Applications, 2nd ed.; Academic Press: New York, 1980. (6) Canet, D.; Robert, J. B. In NMR Basic Principles and Progress; Springer: Berlin, Germany, 1990. (7) Abragam, A. In Principles of Nuclear Magnetism; Oxford: U.K., 1961. (8) NIST Standard Reference Database 46. NIST Critically Selected Stability Constants of Metal Complexes Database, version 8.0; Martell, A. E., Smith, R. M., Eds.; National Institute of Standards and Technology: Gaithersburg, MD, 2004. (9) Jervis, R. E.; Krishnan, S. S. Kinetic Isotopic Exchange Studies of Metal Ion Substitution in EDTA Chelates II: Zinc(II)-Zinc-EDTA Exchange. J. Inorg. Nucl. Chem. 1967, 29, 97−103.

12964

dx.doi.org/10.1021/jp507674a | J. Phys. Chem. B 2014, 118, 12960−12964