Thermodynamic and Structural Investigation of ... - ACS Publications

Jan 4, 2016 - Institut de Physique Nucléaire d'Orsay and ... Institut de Chimie de Nice, Université de Nice Sophia Antipolis, 28 Avenue Valrose, 061...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/IC

Thermodynamic and Structural Investigation of Synthetic Actinide− Peptide Scaffolds Samir Safi,† Aurélie Jeanson,† Jérome Roques,† Pier Lorenzo Solari,‡ Florence Charnay-Pouget,§ Christophe Den Auwer,∥ Gael̈ le Creff,∥ David J. Aitken,§ and Eric Simoni*,† †

Institut de Physique Nucléaire d’Orsay and §ICMMO, CNRS, Univ. Paris-Sud, Université Paris-Saclay, 91405 Orsay cedex, France MARS Beamline, Synchrotron SOLEIL, L’Orme des Merisiers, 91192 Gif-sur-Yvette, France ∥ Institut de Chimie de Nice, Université de Nice Sophia Antipolis, 28 Avenue Valrose, 06108 Nice, France ‡

S Supporting Information *

ABSTRACT: The complexation of uranium and europium, in oxidation states +VI and +III, respectively, was investigated with pertinent bio-inorganic systems. Three aspartate-rich pentapeptides with different structural properties were selected for study to rationalize the structure−affinity relationships. Thermodynamic results, crosschecked by both isothermal titration calorimetry and time-resolved laser fluorescence spectroscopy, showed different affinity depending on the peptide for both Eu(III) and U(VI). The thermodynamic aspects were correlated to structural predictions, which were acquired by density functional theory quantum chemical calculations and from IR and extended X-ray absorption fine structure experiments. The combination of these microscopic properties revealed that carbonyl−metal interactions affected the entropy in the case of europium, while the larger uranyl cation was mostly affected by preorganization and steric effects, so that the affinity was enhanced through enthalpy. The approach described here revealed various microscopic aspects governing peptide actinide affinity. Highlighting these mechanisms should certainly contribute to the rational synthesis of higher affinity biomimetic aspartic ligands.



INTRODUCTION

In this context, we have undertaken an investigation of aspartate-rich pentapeptides bestowed with different geometrical constraints. These peptides are relevant partners when studying the factors, which govern An complexation since they are model fragments of protein primary structures. In this work we focus on trivalent and hexavalent An; An(IV) coordination has been covered elsewhere.6 The use of various aspartate scaffolds to relate their microscopic structural and thermodynamic properties may provide valuable leads for a better understanding of the biomolecular interactions as well as for the development of new complexes implicating these ions, for use in various applications (anticancer drugs, decorporating reagents, and so forth). Furthermore, we examine the influence of the intrinsic cation geometry by comparing the complexation of the linear uranyl(VI) oxo-cation with that of the “spherical” europium(III), which we adopt as a non-radioactive surrogate of trivalent actinides. Indeed spherical hard metals such as lanthanides or actinides may coordinate with peptides in a similar manner to amino acids. The carbonyl group originating from the amide function interacts with such cations,7 which could not be the case with UO22+oxo-cation due to the constrained pentagonal ion structure. We selected here Eu(III) as a representative of all

In the wake of the Fukushima accident, many radioactive elements were released into the environment. Plutonium and non-natural uranium were indeed found as far as 120 km from the explosion site.1 Although lighter elements are the major source of radioactive contamination,2 from a public health and societal standpoint, the lack of understanding of actinide (An) migration pathways in the biosphere is of particular concern. Earlier studies emphasized that exogenic actinides, and more specifically uranium, can contaminate aspartate side chains and, to a lesser extent, amide oxygens in native proteins.3 This observation is significant in the light of the Pearson hard acid/ hard base relationship. The description of the structural and thermodynamic aspects of such complexes is therefore imperative for an understanding of the different contamination pathways, which are governed by both structure (steric hindrance, affinity) and stability.4 Most bio-inorganic studies have concentrated on amino acid transition metal complexes and have been reviewed extensively by Sigel and Martin;5 in contrast, descriptions of biologically relevant actinide and lanthanide complexes remain scarce. This is partly due to both the highly soluble nature of the complexes involved, which makes their crystallization difficult, and the complex speciation of actinides, which may undergo hydrolysis at biologically relevant pH. © XXXX American Chemical Society

Received: October 15, 2015

A

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

series of TRLFS spectra were measured at pH 3.5 for U(VI) and pH 6 for Eu(III) since, under these conditions, the formation of hydrolytic species is minimal and therefore, according to the meal speciation diagram, only the free aquo metals must be taken into account in solution.11,12 Solutions were prepared at peptide concentrations ranging from 0 to 3 × 10−3 M. Experimental Setup. TRLFS experiments were performed using a Continuum R Nd:YAG laser (frequency 10 Hz, pulse duration of ∼7 ns) as the excitation source coupled with a Panther OPO. Detection was achieved using a Spectra-Pro-300 monochromator (Acton Research corporation) coupled with a CCD camera (Princeton Instruments). The excitation wavelength was chosen as 430 nm for uranium and 466 nm for europium, and laser power was ∼2 mJ. Emission spectra were recorded using WINSPEC software (Princeton Instruments). The fluorescence spectra had better than 0.2 nm resolution. The error in measured fluorescence intensity was less than 5%. All fluorescence intensity measurements were performed on nonaerated freshly prepared solutions in a quartz cuvette at temperature 25 ± 1 °C. Every measurement was repeated at least three times. The fluorescence spectra integration (from 450 to 650 nm) and fluorescence decay curves were calculated with the Origin Pro 8.0 software program. The stability constants and spectral deconvolutions were refined using the least-squares fitting program HypSpec. Isothermal Titration Calorimetry. The overall procedure for ITC experiments was similar to that described by Lin et al.13 The VPITC from Microcal (Northampton, MA) was used in all experiments. Briefly, peptide solutions (1.4 mL) were prepared at 0.02 to 0.05 mM concentrations in buffer (25 mM HEPES, pH 3.5, and 100 mM NaCl) and were placed in the titrating cell, while solutions at 0.4 mM (Eu(III)) and 10 mM (U(VI), at this pH value (according to the metal speciation diagrams, only the free aquo metals are present in solution) prepared in the same buffer were placed in the syringe. The peptide solutions were then titrated at 25 °C with 20 μL injections (spaced at 4 min intervals). The blank injections of the metal solution into the corresponding buffer were used to account for the heat of mixing and dilution. In each experiment 15 injections were made, so that the final molar ratio of U(VI) and Eu(III) to peptide was 4 to 5. The heat of the reaction, Q, was obtained by integrating the peak after each injection of peptide ligand using ORIGIN software provided by the manufacturer (Microcal, Northampton, MA). The heat of the reaction at each injection is related to the calorimetric enthalpy of binding; in our case the total cumulative heat was composed of contributions from a single binding site. The model equations for fitting the binding isotherms are included with the Microcal software and are detailed elsewhere.14 ATR-FTIR. Peptide solutions (10 mM) were mixed at pH 4 and 6 for uranium and europium, respectively, with increasing metal-topeptide molar ratios (0.5 to 5). The freshly prepared solutions were then deposited on a diamond crystal surface (Pike Technologies). Absorption spectra were recorded in the range of 2400−600 cm−1 (resolution: 2 cm−1, 50 scans per spectrum) on a Bruker IFS 66 V spectrometer with an ATR module and an MCT detector. Absorption spectra of the peptides were obtained after subtracting the absorption spectra of the buffers, recorded in the same conditions. Under these conditions and using H2O and D2O as solvents the difference spectra were found to be accurate in the 1650−900 cm−1 frequency domain. Density Functional Theory Computational Details. All structures were optimized using the DFT approach as implemented in the Gaussian09 package.15 Calculations were performed using the hybrid B3LYP functional.16 MWB60 Stuttgart relativistic effective core potentials were used to describe uranium atoms, while the 6-31+G* basis set was used for others. To take into account part of the solvent effect, solvation was introduced using a dielectric continuum model of permittivity ε = 80. The conductor-like polarizable continuum model implemented in Gaussian09 was used. Vibrational harmonic frequencies, corresponding to the optimized structures, were also calculated and compared to the experimental ones. EXAFS. EXAFS Sample Preparation. Peptide-Eu(III) and peptideU(VI) complexes were prepared in a 10 to 1 molar ratio at a U(VI)/

the +III elements of the f families (lanthanides, heavy actinides) and U(VI) ({UO22+}) as the ubiquitous form of uranium in natural conditions but also as the representative of the actinyl species (uranyl, neptunyl, plutonyl). To investigate these characteristics, three N- and C-terminal protected pentapeptide sequences were selected: DGDGD, ADPDA, and DPDPD (A = alanine, G = glycine, D = aspartate, P = proline). The latter two contain proline (P) residues that induce an angular strain, while the first peptide contains two flexible glycine (G) residues. The alanine (A) residues in the second peptide introduce partial conformational restrictions. In all cases, the expected complexing functions are the side-chain carboxylates of the aspartate (D) residues. The number of carboxylate side chains was also varied between 2 and 3, as were the positions of the coordinating residues in the pentapeptide sequences. Among the structural investigation techniques for bioinorganic systems, X-ray absorption spectroscopy (XAS) was found suitable since it is an element-specific local structural and electronic probe that is increasingly used to study actinide interaction in biological systems.8 This technique, coupled with other spectroscopic methods such as infrared spectroscopy and time-resolved laser fluorescence spectroscopy (TRLFS), enabled us to achieve a full understanding of the metal environment and the affinity for the coordination partners. Additionally, density functional theory (DFT) was used to model the actinide complex stability. This approach was already applied successfully in our group to study the coordination of uranyl with proteins.9 Affinity studies were used to investigate the thermodynamics of peptide complexes with both Eu(III) and U(VI) by isothermal titration calorimetry (ITC). Lately, this method emerged as the key technique for quantifying the thermodynamics of metal−biomolecule interactions. Furthermore, it is a true solution-phase method, since it does not require labeling or immobilization.10 Finally, the stoichiometry and the affinity were independently cross checked by both ITC and TRLFS, which gives further information about the species present in solution, through fluorescence lifetime measurement.



EXPERIMENTAL PROCEDURES

Materials and Stock Solutions. A stock of UO2 solution (0.1 M in 0.63 M HClO4) was obtained by dissolving UO2(NO3)2·6H2O (99% Fluka puriss) in 12 M HClO4 (Merck Suprapur) and evaporation. This operation (adding HClO4 and evaporation) was repeated three times. An analogous protocol was followed to prepare a stock solution of Eu(III) (0.1 M in 0.63 M HClO4). The pH values of the test solutions were measured using a pH microelectrode (semi micro, Mettler Toledo, France) and pH meter (seven multi, Mettler Toledo, France). The electrode was calibrated using standard buffers. The pH measurements were made at 20 °C with a precision of 0.05 pH units. The pH values of the test solutions were adjusted with NaOH (p.a., CO2-free, Merck) and HClO4 (Suprapur, Merck). Peptide Synthesis. N- and C-protected derivatives of the two peptides ADPDA and DGDGD were synthesized in-house, using solution-state peptide coupling procedures; full details are provided in the Supporting Information document. A commercially available sample of DPDPD-NH2 (Polypeptide, France) was found to be of sufficient purity for the experiments conducted in this work. Time-Resolved Laser Fluorescence Spectroscopy. Sample Preparation. The U(VI) an Eu(III) concentrations were adjusted by diluting the stock solution (the concentration of which is regularly measured using liquid scintillation counting) to 0.01 mM. The ionic strength of the experimental solutions was kept constant at 0.15 M, using NaClO4 (p.a., Merck) under an argon atmosphere. Multiple B

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Peak deconvolution of a normalized fluorescence spectrum of a solution containing 1 × 10−5 M U(VI) or Eu(III) and 1 × 10−4 M peptides at pH 3.5 for U(VI) and pH 6 for Eu(III) and I = 0.1 M. Eu(III) HCl solution concentration of 0.5 mM. Higher stoichiometries were tested without any significant effect on the EXAFS spectrum. Final pH was adjusted either with NaOH or HCl (provided by Sigma−Aldrich). The U(VI)/Eu(III) solutions were added slowly to an acidic peptide solution to avoid hydrolysis, then the pH was raised to the desired value. The solutions were found to be stable at pH values as high as 6. Extended X-ray Absorption Fine Structure Data Acquisition. EXAFS experiments, at the Eu and U LIII edge, were performed on the MARS beamline at the SOLEIL synchrotron facility (a ring operated at 2.75 GeV with 400 mA), which is the French bending magnet beamline dedicated to the study of radioactive materials.17 The optics of the beamline essentially consist of a water-cooled double-crystal monochromator, which is used to select the incident energy of the Xray beam and for horizontal focalization, as well as two large watercooled reflecting mirrors that are used for high-energy rejection (harmonic part) and vertical collimation and focalization.18 In this case of uranium samples, the monochromator was set with the Si(220) crystals and the mirrors with the Pt strips at 3.1 mrad, whereas, in the case of europium the monochromator was set with the Si(111) crystals

and the mirrors with the Si strips at the same angle. The spectra were acquired in fluorescence mode using a multielement Ge detector from ORTEC. No metal reduction was observed in the X-ray absorption near-edge structure region, and thus no beam attenuation was necessary. All measurements were recorded in double-layered 200 μL solution cells specifically designed for radioactive samples, at room temperature. Data were processed using the ATHENA code.19 Background removal was performed using a pre-edge linear function. Atomic absorption was simulated with a square spline function. Phases and amplitudes were calculated with Feff82 code20 with the model clusters obtained by the above computational methods. The EXAFS signal extracted was fitted in R space without any additional filtering, using the ARTEMIS code.19 In all the fits, only one global amplitude factor and one energy threshold factor were considered for all the contributions. Hanning windows with k2 [3 ; 11.5] Å−1 for uranium and [1.5 ; 9.5] Å−1 for europium and a fitting range with R [1 ; 4.0] Å for both were used. The R factor (r) and quality factor χi2/n are both provided as an indication of fit quality in the R space. C

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry



RESULTS AND DISCUSSION Time-Resolved Laser Fluorescence Spectroscopy Analysis. The stoichiometry and pseudo formation constant were determined by varying the concentration of the peptide and using the Hypspec deconvolution program to determine each of the species concentrations. Stoichiometries larger than unity were rejected by Hypspec for all the studied systems. These data were used to calculate the corresponding concentrations of the complex and of the non-complexed ligand. This well-established method for analyzing fluorescence spectra is detailed in the literature. 21,22 All spectral deconvolutions are presented in Figure 1. The observed spectral edge at 570 nm for U (VI) is due to spectrometer cutoff. In the case of uranyl complexes small spectral shifts from 2 to 5 nm toward the higher wavelengths and a decrease of fluorescence intensity were observed upon addition of the peptides to the solution (Figure 1). This decrease is also related to the increased contribution from a species with a lower lifetime, of ∼1 μs, that we assign to the complex. The DGDGD uranium complex presents the lowest lifetime value (0.5 μs), which is consistent with its relative flexibility compared to the relative rigidity induced to the proline group in the two others systems. In the case of europium an increase of the fluorescence yield of the 5D0 → 7F2 (at 615 nm) transition marks the coordination process in solution, as this band is hypersensitive to the coordination environment. Higher lifetimes are also detected to mark the replacement of the water molecules by coordinating amino acids in the peptide’s sequences. Noninteger values of determined q numbers (q = number of water molecules in the coordination sphere) are relatively common. This phenomenon was investigated by Supkowski et al.23 and was attributed to the mixed nature of europium coordination (eight to nine oxygens) or, as in the present case, to the contributions from the N−H oscillators (which affect the lifetime) in which the amide carbonyl oxygen is in the first coordination sphere. Both the formation constants and the determined fluorescence lifetimes are presented in Table 1.

one found for the neptunyl ion, which is an oxo-cation as well, but with only one positive charge (corresponding to a pentavalent neptunium).24 Considering the chemical similarities and therefore the analogies between trivalent lanthanides and actinides, the obtained formation constant values for europium could be used to get an order of magnitude for the values that could be expected for trivalent americium and curium. This is corroborated by the values associated with the interactions between these to trivalent actinides and amino acids such as aspartate or glutamate.24 Isothermal Titration Calorimetry Analysis of the Peptide-U(VI) Complex. ITC was used to investigate the thermodynamics of U(VI) and Eu(III) with all three peptides. To increase the reliability of the thermodynamics model, the constants determined with TRLFS were used as constraints in the fitting process, considering that the metal speciation conditions are the same (only free aquo ions are present in solution) for both techniques. The stoichiometry, which is determined from the inflection point of the titration curve, was fitted between 0.85 and 0.95, increasing the confidence interval of our values. These obtained values, under unity, are common in literature due to molecular diffusion between the syringe and the calorimeter cell. Moreover, as the K values were held constant in the fitting procedure, the stoichiometry numbers obtained are quite reliable and thus allow us to conclude on the 1:1 complex formation. Consequently, this result (and the unique lifetime value found for each system) rules out the possibility of the 2:1 complex formation. The error in the determined thermodynamic values was negligible, of the order of ±20 cal. Figure 2 shows an example of raw data and the fit obtained from the U(VI)/ADPDA titration. The complexation of the peptides under study with U(VI) and Eu(III) ions is driven predominantly by the favorable entropic change (TΔS = 7003−13112 cal/mol) since enthalpy changes were positive (ΔH = 193−8279 cal/mol). A control run with TRIS instead of HEPES gave the same results, indicating that no heat contribution from buffer ionization is detected. One possible explanation for the interaction’s large entropy drive is that both dissociated metal ions and free peptides are heavily solvated by both ion−dipole and hydrogen-bonding interactions. Therefore, the cation binding mechanisms demands the extensive desolvation of both host and guest, thus affording the highly positive entropy change upon complexation. This phenomenon was also reported in the case of transition metals by Trapaidze et al.25 Enthalpy changes in the case of biomolecules, and more specifically peptides, are often related to the formation or rupture of noncovalent interactions (complexation, hydrogen bonds, van der Waals, and London interactions). Lopez et al.26 used metal binding to determine the α-helix formation enthalpy, which is in part due to hydrogen/van der Waals bonds formation. Furthermore, they reported the dependence of this enthalpy contribution on the peptide backbone and specifically in the case of glycine and proline as compared to alanine. This is not surprising, considering that glycine provides more flexibility to the backbone, while proline limits the degrees for freedom. Considering our results, the smaller global endothermic enthalpies reported for DGDGD for both uranyl and europium can thus be explained as follows: DGDGD contains probably less hydrogen intramolecular bonds than the two peptides with proline residue (due to the cis-amide bonds next to the proline, which gives an intramolecular structura-

Table 1. Formation Constants and Fluorescence Lifetime Attributions uranium sample free metal ADPDAmetal DPDPDmetal DGDGDmetal

europium

log kU ± 0.2

lifetime (μs) ± 0.1

log kEu ± 0.2

lifetime (μs) ± 0.1

q

3.2

1.7 1.0

3.1

112 212

8.9 4.5

3.8

0.9

3.5

196

4.8

5.0

0.5

3.8

224

4.2

We observed that in both cases the formation constants present the highest value with the peptide containing no proline residue. The thermodynamics of this affinity enhancement are explored by ITC in the next paragraph. Concerning the uranyl ion, the obtained formation constants values are the same order of magnitude than the ones found with different kind of amino acids such as, for instance, aspartate, glutamate, or alanine, which correspond to the typical values associated with uranyl−carboxylate interactions.3 In contrast, these formation constants values are about twice more than the D

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. (left) Calorimetric titrations with 15 injections of U(VI) (black) and blank buffer (gray) into ADPDA peptide in a 1.4 mL reaction cell (see Experimental Section). An exothermic reaction is shown by a negative heat effect, whereas an endothermic reaction is shown by a positive heat effect. (right) The blank dilution heat values are subtracted from the sample integrated heat values from the left panel and plotted as a function of the molar ratio of U(VI) to ADPDA in the reaction cell (◆), and the corresponding fit is represented by the straight line.

of uranium it is the enthalpy contribution that affects the Gibbs free energy (Table 2). To understand this phenomenon we characterized the structure of the formed complexes by theoretical calculation coupled to FTIR and EXAFS. Fourier Transform Infrared Analysis. To observe structural changes induced by metal addition, FTIR difference spectra were recorded. The structural changes revealed themselves via the appearance of negative and positive peaks in the spectra. To eliminate artificial negative peaks due to solvent subtraction the experiment was performed in both D2O and H2O for all peptides (Figure 3). In each of the spectra we noted the presence of the negative peaks at 1575−1581 and 1463−1486 cm−1, which were attributed, respectively, to the deprotonated ionic carboxylate’s asymmetric (ν as) and symmetric (νsym) stretch modes. This depicts the carboxylate’s contribution to the coordination of Eu(III) and U(VI). Furthermore, the positive peaks obtained by metal addition reveal complex formation through carboxylate symmetric and asymmetric stretch modes,27 recapitulated in Table 3. The type of coordination of carboxylate groups to metals ions can be derived from vibrational spectra by studying the splitting (Δν) between the νas and νs (COO) modes. It has been shown that Δν monodentate > Δν ionic = Δν bridging > Δν bidentate.28 In our case the Δν values of the deprotonated peptides serve as a reference and are equal to 100 ± 15 cm−1,

tion). Therefore, to break these hydrogen bonds, the enthalpy values needed are higher for the ADPDA and DPDPD than for the DGDGD peptide. This calorimetric study also reveals that both the entropy and enthalpy changes show, for both metals, a higher affinity for the peptide without proline residue (Table 2). However, in the case Table 2. Thermodynamic Parametersa Determined with a 1:1 Binding Isotherm Model

The ΔG was calculated from the affinity constants determined by TRLFS. All values in the table are reported in calories per mole.

a

Figure 3. FTIR difference (peptide-(U(VI) or Eu(III)) minus peptide). The difference spectra were obtained as detailed in the Methods Section. E

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Table 3. Experimental and Theoretical Infrared Carboxylate Energy Gap Calculated by DFT Band Position for “Free” Peptide and Complexes peptides bands

a

free peptide υ (cm−1) υas COO



υsym COO



Eu(III) peptide υ (cm−1) υgap

υamide(I)

ADPDA

1583

1477

106

1626

DPDPD

1587

1484

103

1620

DGDGD

1578

1465

113

1643

υas COO a

1536 1560b 1521a 1556b 1522a 1541b



υsym COO



a

1449 1426b 1439a 1398b 1451a 1422b

Δυ 87 134 82 158 71 120

U(VI) peptide υ (cm−1) DFT-Δυ 89 151 55 161 68 142

υassUO2 937 936 937

υas COO− a

1434 1532b 1438a 1528b 1459a 1521b

υsym COO−

Δυ

DFT-Δυ

1379a 1352b 1376a 1351b 1389a 1349b

55 180 62 177 70 172

42 180 56 189 51 128

Bidentate carboxylate coordination. bMonodentate carboxylate coordination.

Figure 4. Theoretical models calculated using DFT for the three Eu(III) (blue) and the three U(VI) (yellow) interaction sites. Hydrogen atoms are omitted for clarity.

depending on the peptide (cf. Table 3). In general, bidentate coordination of the carboxylate groups to the Eu(III) or U(VI) metal ion is expected to show a significantly lower spectral splitting than monodentate binding. For europium, Δν values of 71−87 cm−1 and 120−134 cm−1 reveal, respectively, the presence of both monodentate and bidentate complexes. These values are in agreement with those we determined from vibrational DFT calculations, using the structural models described in the latter section. The same conclusions can be drawn for uranium. However, the band splits for U(VI)

complexes were more pronounced compared to the free peptide (Δνbi 55−62 and Δνmono 172−180 cm−1), probably due to a stronger covalent character of the interaction. In the spectra of Eu(III) complexes we noticed an positive band increasing in the 1620−1643 cm−1 region, blue-shifting from the negative band at 1585 cm−1. This corresponds to the trans-amide I (primarily CO stretch) mode that appears for many peptide−metal cation complexes29 and is consistent with strong coordination of carbonyl oxygen atoms to trivalent metal cations, resulting in greater partial charge transfer.30,31 Finally, F

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 4. Description of EXAFS Parameters Used for the Fit to Both Europium and Uranium Complexes europium complex

uranium complex

Ndegenb

ΔRc

σd

patha

Ndegenb

ΔRc

σd

Eu−Ofirst‑sphere Eu−Cbi Eu−Cmono

9 1 1

ΔR1 ΔR2 ΔR3

σ12 σ22 σ32

Eu→ C→Cbi Eu→ O →Ccarb

2 2

ΔR2 ΔR4

σ42 σ52

U−Oyle U−O1mono U−O2bi U−Cbi U−Cmono U→Oyle1→Oyle2 U→Oyle1→U→Oyle2 U→Oyle1→U→Oyle1

2 1 4 1 1 2 2 2

ΔR1 ΔR2 ΔR3 ΔR4 ΔR5 2 × ΔR1 2 × ΔR1 2 × ΔR1

σ12 σ22 σ32 σ42 σ52 2 × σ12 2 × σ12 4 × σ12

patha

a Atom types in the scattering path. bDegeneracy from the theoretical models. cDeviation from the theoretical bond length. dDebye−Waller Factor (bi: bidentate; mono: monodentate caboxylate; carb: amide carbonyl carbon).

Table 5. Fitted Values (R, σ2, AMP, ΔE0) Determined for Each Peptide−Actinide Complex structure ADPDA

DPDPD

DGDGD

a

AMP 1.0 ΔE0 2.1 ev R factor 0.002 ChiR(88)

AMP 1.0 ΔE0 1.8 ev R factor 0.003 ChiR(90)

AMP 1.0 ΔE0 3.6 ev R factor 0.002 ChiR(78)

uranium bond

N

DFT bond Å

fitted length Å (σ2 in Å2)

UOyl

2

1.80

1.77 (0.003)

UO1 UO2 UCbi UCmono UOyl

1 4 1 1 2

2.20 2.48 2.92 3.53 1.80

2.25 (0.002) 2.43 (0.007) 2.92 (0.001) 3.38 (0.004) 1.78(0.003)

UO1 UO2 UCbi UCmono UOyl

1 4 1 1 2

2.20 2.49 2.92 3.52 1.80

2.29 2.43 2.93 3.50 1.78

(0.004) (0.009) (0.001) (0.005) (0.003)

UO1 UO2 UCbi UCmono

1 4 1 1

2.33 2.48 2.92 3.46

2.29 2.45 2.88 3.46

(0.003) (0.009) (0.001) (0.004)

AMP 1.0 ΔE0 6.7 ev R factor 0.02 ChiR(63)

AMP 1.0 ΔE0 6.88 ev R factor 0.01 ChiR(55)

AMP 1.0 ΔE0 6.6 ev R factor 0.02 ChiR(63)

europium bond

N

DFT bond Å

fitted length Å (σ2 in Å2)

EuOfirst sphere

9

2.39−2.60

2.44 (0.008)

EuCbi EuCbi-C EuCmono EuCcarb (C‑o) EuOfirst sphere

1 2 1 1 9

2.94 4.44 3.36 3.70 2.39−2.89

2.87 4.37 3.45 3.63 2.44

(0.01) (0.004) (0.004) (0.001) (0.009)

EuCbi EuCbi-C EuCmono EuCcarb (C‑o) EuO

1 2 1 1 9

2.89 4.39 3.37 4.02 2.30−2.58

2.84 4.35 3.53 4.04 2.44

(0.01) (0.001) (0.004) (0.005) (0.008)

EuCbi EuCbi-C EuCmono EuCcarb (C‑o)

1 2 1 1

2.90 4.39 3.36 3.59

2.80 4.29 3.40 3.51

(0.01) (0.002) (0.006) (0.003)

All parameters were already defined in Table 4

in the case of uranyl, the υas UO2 stretch mode is sensitive to the nature and number of uranyl equatorial ligands.32 Since all three uranyl complex bands are shifted from 954 (free uranyl) to 937 cm−1, we can presume a similar environment for all three complexes. Density Functional Theory Peptide-U(VI) and PeptideEu(III) Model Complex Calculations. In the absence of crystallographic data for the peptide complexes, indications regarding the coordination scheme come from the knowledge of the peptide amino acid sequence and the preliminary IR and TRLFS data. We therefore tested in silico various geometrical combinations around the U(VI) and Eu(III) that were in agreement with IR and TRLFS, following a similar procedure described elsewhere.9 This consisted of defining plausible geometries for a first coarse-grain refinement, using initial interatomic distances obtained from previously reported work.33 The next step consisted of refining all the interatomic distances in the presence of the peptide sequence strain and steric effects using DFT calculations. The structures were then cross-checked by comparing theoretical and experimental vibrational spectra (Table 3). The structures that were found to be most stable and that fitted the criteria are presented in Figure 4. For the three peptides, the coordination site consists in one bidentate carboxylate, one monodentate carboxylate, one carbonyl, and four water molecules for europium. For uranium, the coordination site is composed of one monodentate

carboxylate, one bidentate carboxylate, and two water molecules, whatever the peptide. The DFT bond distances are also presented in Table 5. Extended X-ray Absorption Fine Structure Results. The experimental europium and uranium peptide EXAFS data were fitted using the independent structural parameters summarized in Table 4. Only one first layer coordination sphere (Ofirst‑sphere) was necessary to fit the europium data, while three O layers were necessary in the case of uranyl: dioxo uranyl atoms (Oyl), monodentate carboxylate (O1), bidentate carboxylate, and water oxygen atoms (O2). Extra shells of second sphere carbon atoms were added to the fits for both coordinating carboxylates (monodentate Cmono and bidentate Cbi) and an extra C shell in the europium case for the amide peptidic bond carbonyl coordination. Multiple scattering contributions, based on amplitude Feff calculation, were also added to the fitting process. The second sphere carbon MS paths in the case of europium (carbonyl, carboxylate bidentate) were thus selected, and the dioxo MS paths were selected in the case of uranium. We note that for the latter cation second sphere carbon MS was minimal due to distorted angles. The fitted values and the experimental and fitted spectra are summarized in Table 5 and in Figure 5, respectively. The fitted amplitude error was found to be of the order of 0.05, which emphasizes fit quality. Further relevance of G

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. K3-weighted U and Eu LIII-edges EXAFS spectra of peptide complexes and the corresponding Fourier transforms. Black lines are the experimental data, and dotted lines are corresponding fits.

Uranium Peptide Complexes. The contribution of axial oxygens was fitted from 1.77 to 1.78 Å with a coherent Debye

structural values was analyzed separately for uranium and europium. H

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Weller factor of 0.003 Å2. This confirms the U(VI) speciation and is in complete agreement with literature.34,35 The absence of a short distance contribution at 2.20 Å in the equatorial range confirmed the absence of hydrolysis. The first equatorial oxygen shell was fitted at 2.25−2.29 Å in accordance with a monodentate carboxylate. The bidentate carboxylate was confirmed by U−O distances at 2.48−2.49 Å and U···C distances at 2.92 Å for all the peptide uranyl complexes, similar to literature.36,37 These results are in line with the spectral IR analysis. We note that no structural contribution for amide peptidic bond carbonyl function in the uranyl coordination sphere was needed to fit the data. This absence of the carbonyl interaction, also suggested by IR, can be explained by the steric hindrance due to the pentagonal bipyramidal coordination geometry of uranyl. Since all three peptides have the same coordination site we can conclude that the enhanced enthalpy through the series relates mostly to the bond strength. The more flexible DGDGD structure generates thus a more stable complex. However, it is interesting to note that is has been found, using EXAFS and NMR experiments, a structure radically different with a same kind of linear pentapeptide (Asp-Asp-Pro-Asp-Asp-NH2) interacting with Th(IV), Np(IV), and Pu(IV) ions. Indeed, the results show the presence of an μhydroxo oxygen atoms and a metal−metal contribution, which could be due to a complex nuclearity higher than two.38 Consequently, in that case, the absence of the oxo-cation structure could probably play the most important role in the interactions between these tetravalent actinides and this pentapeptide. Europium Peptide Complexes. The inner coordination sphere of europium was fitted with nine oxygens to determine a mean value, for all three peptides, at 2.44 Å in agreement with literature.39,40 As for carboxylate monodentate (3.36 Å) and bidentate (2.89−2.94 Å) carbons, distances were fitted in accordance with reported data.41 Moreover the amide carbonyl multiple scattering contributes to the EXAFS signal. The fitted amide carbonyl bond lengths were in agreement with our calculated DFT models. Furthermore, we notice that the carbonyl bond length follows the same trend as the entropy and enthalpy determined by ITC. We can thus suggest that the carbonyl bond makes an exothermic contribution and generates a rigid complex, which is illustrated by decreased entropy. These two factors contribute to a higher affinity. The discussion of the EXAFS data in combination with thermodynamic figures determined by ITC have revealed that DGDGD flexibility enables structural rearrangement with a thermodynamic positive effect in both europium and uranyl cases. However, the origin of the affinity increase is cationdependent: - bond strength for uranium. - amide carbonyl interaction and skeleton rigidity in the case of europium.

the bond formation and a rigidified, ordered complex structure. The directional coordination due to uranyl is mostly governed by the enthalpy and thus by the bond stability. The high affinity of DGDGD compared to the other two peptides is due to its high degrees of freedom (no intramolecular structuration) enabling it to easily reorganize itself to fit the coordination sphere. Finally, for the three peptides, the strain caused by the short backbone of the peptide sequence confined the carboxylate to an exclusively monodentate coordination next to the neighbor amide carbonyl function. This observation was verified by DFT theoretical calculations that demonstrated the instability of a bidentate carboxylate next to a neighboring carbonyl. Finally, the rather strong interactions between these metals and the studied peptides show that, in case of an internal contamination, these elements could take the place of the natural elements in larger biological system such as proteins.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.5b02379. Discussion of experimental procedures and spectroscopic characterization of Boc-DGDGD-OMe and MocADPDA-NH2. Schematic descriptions of syntheses. Numbered structures of organic synthesized products. (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors wish to thank B. Sitaud, S. Schultig, and I. Llorens for beamline assistance on the MARS beamline of the SOLEIL synchrotron. We would also like to thank M. Desmadril from IBBMC, Paris Sud Univ. and L. Assairi from the Curie Institute in Orsay for providing access to the ITC apparatus. We thank C. Diarra for his precious help in managing the IPNO cluster: GRIF (http://www.grif.fr). Finally we would like to acknowledge P. Roy and J.-B. Brubach from the AILES beamline of the SOLEIL synchrotron and R. Barillon from the Univ. of Strasbourg for providing access to the IR apparatus.



REFERENCES

(1) Shinonaga, T.; Steier, P.; Lagos, M.; Ohkura, T. Environ. Sci. Technol. 2014, 48, 3808−3814. (2) Garnier-Laplace, J.; Beaugelin-Seiller, K.; Hinton, T. G. Environ. Sci. Technol. 2011, 45, 5077−5078. (3) Van Horn, J. D.; Huang, H. Coord. Chem. Rev. 2006, 250, 765− 775. (4) Eidson, A. F. Health Phys. 1994, 67, 1−14. (5) Sigel, H.; Martin, R. B. Chem. Rev. 1982, 82, 385−426. (6) Jeanson, A.; Berthon, C.; Coantic, S.; Den Auwer, C.; Floquet, N.; Funke, H.; Guillaneux, D.; Hennig, C.; Martinez, J.; Moisy, P.; Petit, S.; Proux, O.; Quemeneur, E.; Lorenzo Solari, P.; Subra, G. New J. Chem. 2009, 33, 976−985. (7) Kremer, C.; Torres, J.; Domínguez, S.; Mederos, A. Coord. Chem. Rev. 2005, 249, 567−590. (8) Ascone, I.; Fourme, R.; Hasnain, S.; Hodgson, K. J. Synchrotron Radiat. 2005, 12, 1−3.



CONCLUSIONS In this study we have described europium and uranyl peptidic complexes both structurally using EXAFS and IR and thermodynamically using ITC and TRLFS. This structure affinity relationship approach revealed molecular scale aspects affecting the thermodynamic stability of peptide europium/ uranium complexes. The contribution of the carbonyl function to the affinity of the interaction with spherical europium provides a positive effect on both entropy and enthalpy due to I

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (9) Safi, S.; Creff, G.; Jeanson, A.; Qi, L.; Basset, C.; Roques, J.; Solari, P. L.; Simoni, E.; Vidaud, C.; Den Auwer, C. Chem. - Eur. J. 2013, 19, 11261−11269. (10) Wilcox, D. E. Inorg. Chim. Acta 2008, 361, 857−867. (11) Krestou, A.; Panias, D. Eur. J. Miner. Process. Environ. Prot. 2004, 4, 113−129. (12) Sastri, V. R.; Perumareddi, J. R.; Ramachandra Rao, V.; Rayudu, G. V. S.; Bünzli, J.-C. G. Modern Aspects of Rare Earths and their complexes, 1st ed.; Elsevier: Netherlands, 2003. (13) Lin, L. N.; Mason, A. B.; Woodworth, R. C.; Brandts, J. F. Biochemistry 1991, 30, 11660−11669. (14) Freire, E.; Mayorga, O. L.; Straume, M. Anal. Chem. 1990, 62, 950A−959A. (15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian; Gaussian, Inc: Wallingford, CT, 2009. (16) Becke, A. J. Chem. Phys. 1993, 98, 5648−5652. (17) Sitaud, B.; Solari, P. L.; Schlutig, S.; Llorens, I.; Hermange, H. J. Nucl. Mater. 2012, 425, 238−243. (18) Solari, P. L.; Schlutig, S.; Hermange, H.; Sitaud, B. J. Phys.: Conf. Ser. 2009, 190, 012042. (19) Ravel, B.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537− 541. (20) Rehr, J. J.; Albers, R. C. Rev. Mod. Phys. 2000, 72, 621−654. (21) Koban, A.; Geipel, G.; Bernhard, G. Radiochim. Acta 2003, 91, 393−396. (22) Brachmann, A.; Geipel, G.; Bernhard, G.; Nitsche, H. Radiochim. Acta 2002, 90, 147−153. (23) Supkowski, R. M.; Horrocks, W. D., Jr Inorg. Chim. Acta 2002, 340, 44−48. (24) Ansoborlo, E.; Prat, O.; Moisy, Ph.; Den Auwer, Ch.; Guilbaud, Ph.; Carriere, M.; Gouget, B.; Duffield, J.; Doizi, D.; Vercouter, Th.; Moulin, Ch.; Moulin, V. Biochimie 2006, 8, 1605−1618. (25) Trapaidze, A.; Hureau, C.; Bal, W.; Winterhalter, M.; Faller, P. JBIC, J. Biol. Inorg. Chem. 2012, 17, 37−47. (26) Lopez, M. M.; Chin, D.-H.; Baldwin, R. L.; Makhatadze, G. I. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 1298−1302. (27) Groenewold, G. S.; de Jong, W. A.; Oomens, J.; Van Stipdonk, M. J. J. Am. Soc. Mass Spectrom. 2010, 21, 719−727. (28) Kakihana, M.; Nagumo, T.; Okamoto, M.; Kakihana, H. J. Phys. Chem. 1987, 91, 6128−6136. (29) Prell, J. S.; Flick, T. G.; Oomens, J.; Berden, G.; Williams, E. R. J. Phys. Chem. A 2010, 114, 854−860. (30) Velasquez, J.; Pillai, E. D.; Carnegie, P. D.; Duncan, M. A. J. Phys. Chem. A 2006, 110, 2325−2330. (31) Zhang, G.; Wang, W.; Chen, D. Chem. Phys. 2009, 359, 40−44. (32) Pardoux, R.; Sauge-Merle, S.; Lemaire, D.; Delangle, P.; Guilloreau, L.; Adriano, J.-M.; Berthomieu, C. PLoS One 2012, 7, e41922. (33) Safi, S.; Charbonnel, M. C.; Creff, G.; Jeanson, A.; Mostapha, S.; Roques, J.; Simoni, E.; Solari, P. L.; Vidaud, C.; Den Auwer, C. MRS Online Proc. Libr. 2012, 1444, 199−209. (34) Thompson, H. A.; Brown, G. E.; Parks, G. A. Am. Mineral. 1997, 82, 483−496.

(35) Kelly, S. D.; Kemner, K. M.; Fein, J. B.; Fowle, D. A.; Boyanov, M. I.; Bunker, B. A.; Yee, N. Geochim. Cosmochim. Acta 2002, 66, 3855−3871. (36) Francis, A. J.; Gillow, J. B.; Dodge, C. J.; Harris, R.; Beveridge, T. J.; Papenguth, H. W. Radiochim. Acta 2004, 92, 481−488. (37) Jeanson, A.; Ferrand, M.; Funke, H.; Hennig, C.; Moisy, P.; Solari, P. L.; Vidaud, C.; Den Auwer, C. Chem. - Eur. J. 2010, 16, 1378−1387. (38) Jeanson, A.; Dahou, S.; Guillaumont, D.; Moisy, Ph.; Den Auwer1, Ch.; Scheinost, A.; Hennig, Ch.; Vidaud, C.; Subra, G.; Solari, P. L. J. Phys.: Conf. Ser. 2009, 190. (39) Suib, S. L.; Zerger, R. P.; Stucky, G. D.; Morrison, T. I.; Shenoy, G. K. J. Chem. Phys. 1984, 80, 2203−2207. (40) Runde, W.; Van Pelt, C.; Allen, P. G. J. Alloys Compd. 2000, 303−304, 182−190. (41) Mathur, J. N.; Thakur, P.; Dodge, C. J.; Francis, A. J.; Choppin, G. R. Inorg. Chem. 2006, 45, 8026−8035.

J

DOI: 10.1021/acs.inorgchem.5b02379 Inorg. Chem. XXXX, XXX, XXX−XXX