Mechanism of Indium (III) Exchange between NTA and Transferrin: A

Aug 30, 2008 - (log β12 ) 24.4) at 20 °C and I ) 0.1 M45 and 2 orders of magnitude compared to the Harris et al. value (log β12 ) 23.7).35. The lar...
0 downloads 0 Views 122KB Size
Download PDF
12168

J. Phys. Chem. B 2008, 112, 12168–12173

Mechanism of Indium(III) Exchange between NTA and Transferrin: A Kinetic Approach Tarita Biver, Rossella Friani, Chiara Gattai, Fernando Secco,* Maria Rosaria Tine´, and Marcella Venturini Dipartimento di Chimica e Chimica Industriale, UniVersity of Pisa, Via Risorgimento 35, 56100 Pisa, Italy ReceiVed: May 21, 2008; ReVised Manuscript ReceiVed: July 9, 2008

The equilibria and kinetics for the process of In3+ exchange between nitrilotriacetic acid (NTA) and bovine serum transferrin (T) have been investigated in aqueous solution containing sodium bicarbonate. The metal exchange equilibria have been measured by difference ultraviolet spectroscopy at 25 °C, pH ) 7.4, and I ) 0.2 M (NaClO4). The acid dissociation constants of NTA and the binding constants of In(III) to NTA have also been measured. Kinetic experiments revealed that the process of In3+ uptake by transferrin from [In(NTA)2]3- is biphasic, the fast phase being completed in a few seconds, the slow phase lasting for hours. The fast phase has been investigated by the stopped-flow method and results in monoexponential kinetics. It involves rapid interaction of the 1:1 complex ML (M ) In, L ) NTA) with TB (T ) transferrin, B ) CO32-) to give a quaternary intermediate MLTB which then evolves to an “open” MTB* ternary complex complex with expulsion of L. In turn, this complex interconverts to a “closed”, more stable, form MTB. Neither the prevailing complex M2L nor the TB2 form of transferrin are directly involved in the exchange process but act as metal and protein reservoirs. The pH dependence of the reaction has been also investigated. The slow phase has not been investigated in detail; it takes several hours to go to the completeness, its slowness being ascribed to metal redistribution between the C-site and N-site of the protein, and/or metal release from polynuclear In(III) species. Introduction Serum transferrin is the most important iron-transport system in vertebrates. It binds iron strongly enough to prevent hydrolysis and to deny it to potentially harmful bacteria; at the same time it is able to deliver iron from the blood stream across the plasma membrane to the cytosol by means of receptor mediated endocytosis.1 Transferrin is a bilobate monomeric glycoprotein and each lobe possesses one metal binding site, the C-site in the C-terminal lobe, and the N-site in the N-terminal lobe. In each of these sites iron(III) ion is probably coordinated by four protein residues, namely two tyrosines, one histidine and one aspartate residue.2 It was recognized for long time3 that the metal binding can occur only in the presence of a synergistic anion (in ViVo carbonate), that takes part to the creation of the metal site.3 The synergistic anion coordinates to Fe3+ ion as a bidentate ligand and also interact by hydrogen bonding with the charged and polar groups of the protein forming a ternary complex.2 The peculiarity of the transferrin interaction with iron(III) has prompted a variety of investigations on this system, including kinetic approaches to the mechanism of iron uptake4-7 and release.8-27 In most cases the features of the metal release process are interpreted in terms of the Bates mechanism28 which involves a conformational equilibrium between a “closed” and an “open” form of transferrin, only the latter being able to donate Fe3+ to the accepting ligand. The ability of transferrin to accommodate metal ions other than Fe3+ depends on interplay of charge, size, coordination and geometry preferences of the coordinated metal ion. An important role seems to be played by the structure of the six coordinated (approximately octahedral) site provided by the protein. This feature reflects in a strong preference of transferrin * To whom correspondence should be addressed. Telephone: +39-0502219259. Fax: +39-050-2219260. E-mail: [email protected].

Figure 1. Titrations curves of transferrin with [In(NTA)2]3- solutions of the following compositions: CNTA/CIn ) 35 (triangles); CNTA/CIn ) 10 (squares); CNTA/CIn ) 3 (circles). CT ) 1.2 × 10-5 M, CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), λ ) 240 nm, and T ) 25 °C.

for cations with high positive charge and octahedral geometry29 as Fe3+, Al3+, Ga3+, and In3+. The group XIII metals have never been classified as essential substances for life, but are of particular interest because several of them are medically important.30 Indium-113 and Indium-111 are used in radiodiagnostic medicine as imaging agents for liver, spleen, and placenta tumors.31 The metal is administered as a complex with citrate or NTA and then is transferred to transferrin that introduces it into the cells. The metal exchange between two strong ligands excludes precipitation of indium as hydroxide at physiological pH values. In Vitro, NTA is widely used as a competitive ligand to study the metal-transferrin interaction32-34 and the equilibria of indium exchange between nitrilotriacetic acid (NTA) and human transferrin have been investigated by spectrophotometric methods.35 More recently the kinetics of

10.1021/jp8045033 CCC: $40.75  2008 American Chemical Society Published on Web 08/30/2008

Mechanism of Indium(III) Exchange

J. Phys. Chem. B, Vol. 112, No. 38, 2008 12169

TABLE 1: Acid Dissociation Constants of NTA (pKAi), of In3+ ion (pKHi) and Complex Formation Constants of the In(III)-NTA System (log β1i) at T ) 25°C, I ) 0.2 M (NaClO4) pKA1

pKA2

pKA3

pKH1

pKH2

pKH3

pKH4

log β11

log β12

9.31 ( 0.08

2.28 ( 0.1

1.91 ( 0.1

3.7a

3.9a

4.7a

9.5a

16.1 ( 0.1

25.7 ( 0.1

a

Data from ref 46 reduced to I ) 0.2 M.

Figure 2. Analysis according to eq 3 of a titration made under the conditions CT ) 1.13 × 10-5 M, CNTA/CIn ) 35. CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M, T ) 25 °C. Difference UV titration has been performed at 240 nm under batch-wise conditions.

Figure 4. Stopped-flow curve recorded at 240 nm after mixing a transferrin solution with a In-NTA solution. CT ) 8.65 × 10-6 M, CIn ) 1.74 × 10-4 M, CNTA ) 10CIn, CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), T ) 25 °C.

TABLE 2: Apparent Equilibrium Constants for Reactions I and II and Related Extinction Coefficient Changesa ∆1 (M-1 s-1) 1.4 × a

104

∆2 (M-1 s-1) 3.5 ×

104

log KEX1

log kEX2

-1.07

-2.63

I ) 0.2M (NaClO4), pH ) 7.44, and T ) 25°C.

Figure 3. Analysis according to eq 4 of titration curves with CNTA/CIn ) 3 and CNTA/CIn ) 10. CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), T ) 25 °C.

SCHEME 1

metal exchange processes between transferrin and citrate involving Al3+, Ga3+, In3+, and Tb3+ have been studied.36 Concerning In(III), the time course of the interaction of the In-NTA complex with transferrin has been investigated only qualitatively and it has been concluded that indium uptake by the protein is a very slow process.35 On the other hand, In(III) binds ligands very fastly,37 so inducing the suspect that at least the initial steps of the process of indium binding to transferrin could be rapid even in case of a metal exchange reaction, unless a complex multistep mechanisms would be operative. In this paper the results are presented of a thermodynamic and kinetic study of the reaction of indium exchange between NTA and transferrin. Despite the fact that the most universally accepted symbol for transferrin is Tf, in this paper we found convenient to replace

Figure 5. Plot of 1/τ vs CIn for the In(III)/transferrin system. CT ) 8.65 × 10-6 M, CNTA ) 4.33 × 10-3 M, CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), and T ) 25 °C. Continuous line is based on eq 6.

Tf by T in order to leave out any confusion with Tf that in the paper will denote the whole of free and carbonate bound transferrin. Materials and Methods Materials. Water to be used for preparation of stock solutions and as a reaction medium was obtained through a four-bowl Millipore purification system, and thoroughly degassed. The resulting carbon dioxide-free solvent was then kept under argon atmosphere.

12170 J. Phys. Chem. B, Vol. 112, No. 38, 2008

Biver et al. TABLE 3: Reaction Parameters for the Reaction of Indium Exchange between Nitrilotriacetate Ion and Transferrin at I ) 0.2 M and T )25°C, Evaluated According to Scheme 1 (Eq 6) k-2 k1KA (M-1 s-1) (s-1) 1.8 × 106 a

0.3

k-1/k2 (M-1)

K′ML (M-1)

1.2 × 103

20

K′ML2 (M-1)

Kb1a (M-1)

2.0 × 103 2.3 × 102

Kb2a (M-1) 22

Literature values introduced as fixed parameters.

Figure 6. Plot of 1/τ vs CNTA for the In(III)/transferrin system. CT ) 8.65 × 10-6 M, CIn ) 1.73 × 10-4 M, CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), and T ) 25 °C. Continuous line is based on eq 6.

Figure 9. Plot of 1/τ vs pH for the In(III)/transferrin system. CT ) 1.73 × 10-5 M, CIn ) 4.33 × 10-4 M, CNTA ) 4.33 × 10-3 M, CNaHCO3 ) 5.0 × 10-3 M, I ) 0.2 M (NaClO4), and T ) 25 °C.

Figure 7. Plot of 1/τ vs CNaHCO3 for the In(III)/transferrin system. CT ) 1.73 × 10-5 M, CIn ) 4.33 × 10-4 M, CNTA ) 4.33 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), and T ) 25 °C. Continuous line is based on eq 6.

Figure 8. Plot of 1/τ vs CIn for the In(III)/transferrin system under the conditions CT ) 1.73 × 10-5 M, CNTA/CIn ) 10, CNaHCO3 ) 5.0 × 10-3 M, pH ) 7.44, I ) 0.2 M (NaClO4), and T ) 25 °C. The continuous line is based on eq 6.

Bovine serum transferrin (T) was purchased from Sigma and made free from traces of low molecular weight impurities by dissolving the appropriate amount of solid protein in a small volume of water and dialysing it against water for 24 h at 4 °C.4 The protein concentration has been obtained from the UV spectrum using a molecular extinction coefficient of 83600 M-1 cm-1 at 280 nm.38 Stock solutions of indium perchlorate have been prepared by weighing appropriate amounts of the pure metal into a beaker, adding concentrated HClO4, covering and gently heating until complete dissolution of the metal. Then an aliquot of the solution has been titrated with EDTA ((ethylenedinitrilo)tetracetic acid) at

pH ) 2.4 using PAN (1-(2-pyridyl-azo)-2-naphtol) as an indicator.39 The solution of In(ClO4)3 contains same perchloric acid to prevent precipitation of indium hydroxide. NaOH titration of a sample of the stock solution, previously passed through an exchange resin loaded with H+ ions, enabled us to evaluate the acid content. Solutions of indium nitrilotriacetate ([In(NTA)2]3-) have been prepared by mixing indium perchlorate and sodium nitrilotriacetate solutions in the stoichiometric ratio 1:2. Methods. pH measurements have been made at 25 °C with a PHM 84 (Radiometer Copenhagen). Potentiometric titrations have been performed under argon atmosphere using a computercontrolled autotitrator (Crison) able to deliver calibrated amounts of NaOH, and to monitor pH versus time until equilibrium achievement. The glass electrode, connected to the reference electrode by a 3 M NaCl solution, was calibrated according to Gran’s method.40 The potentiometric curves have been evaluated using the Hyperquad 2000 least-squares program.41 Spectrophotometric measurements have been performed on a Perkin-Elmer Lambda 35 spectrophotometer. Solutions containing identical amounts of transferrin and different excess amounts of In(NTA2)3- have been left to equilibrate; then the equilibria of the indium exchange between NTA and T have been measured by difference spectrophotometric batch-wise titrations. Kinetic measurements have been performed with a stopped-flow apparatus constructed in our laboratory.42 The signal revealing the course of the reaction was acquired by a Tektronix 2212 digital oscilloscope with a storage capability of 2500 data points; Then it was transferred to a personal computer, via a GPIB interface, using the WaveStar 2.0 program, and analyzed by a nonlinear least-squares procedure.43 The kinetic curves were monoexponential in most cases. Occasionally, biexponential fits gave better results but the amplitude of the additional exponential was small; hence, this minor effect was neglected. The time constants used to evaluate the kinetic parameters are the average of six repeated experiments at least and display a maximum spread of 10%. All experiments have been performed at 25 ( 0.1 °C.

Mechanism of Indium(III) Exchange Results Acid Dissociation of NTA and Stability Constants of the In(III)-NTA Complexes. The literature data concerning H+ and In3+ binding to NTA are rather discord.35,44,45 Hence, the acid dissociation constants of NTA and the complex formation constants between indium and NTA were measured under the experimental conditions (I ) 0.2 M (NaClO4) and T ) 25 °C) used to investigate the metal-transferrin interactions. First, NTA solutions in the absence of indium were titrated with NaOH in order to determine the pKAi values of the unbound ligand (Figure 1S(a) of the Supporting Information). Then, potentiometric titrations with NaOH of a solution containing a mixture of In(III) and NTA in the ratio 1:3 (Figure 1S(b) of the Supporting Information) allowed the formation constants of the complexes InNTA and [In(NTA)2]3- to be evaluated. The equilibrium parameter values are reported in Table 1, where the literature values of the acid dissociation constants of In3+ ion are also given;46 species distribution plots of NTA and InNTA complexes are given in the Supporting Information (Figures 2S and 3S). Equilibria of Indium Exchange between NTA and Transferrin. The metal exchange process is revealed by changes of the difference spectra of transferrin produced by reaction of different amounts of In(NTA)23- with the same transferrin amount at constant pH (Figure 4S of the Supporting Information). Figure 1 shows the binding curves derived from difference batch-wise titrations performed with CNTA/CIn ) 35 (triangles), CNTA/CIn ) 10 (squares) and CNTA/CIn ) 3 (circles). All titrations were performed at pH ) 7.4 and CNaHCO3 ) 5 × 10-3. The data points of the first curve reach a plateau just after the concentration ratio of indium to transferrin, CIn/CT, has reached the value of 1, thus indicating that for large excess of NTA (CNTA ) 35CIn) the 1:1 indium-transferrin complex is prevailing. The third curve (CNTA ) 3CIn) levels off for CIn/CT values between 1 and 2, and the plateau value is remarkably higher compared to that of the first curve, thus suggesting that the 2:1 complex gives an important contribution to the overall exchange process.35 The process of In3+ exchange between NTA and transferrin can be thus represented by the apparent reactions (I) and (II) (charges omitted)

In(NTA)2 + Tf a InT + 2NTA

(I)

In(NTA)2 + InT a In2T + 2NTA

(II)

where the unbound NTA represents the sum of the differently protonated forms of nitrilotriacetic acid while Tf represents the sum of free and carbonate bound transferrin ([Tf] ) [T] + [TB] + [TB2] (Scheme 1)). The corresponding apparent equilibrium constants are expressed by eqs 1 and 2

KEX1 ) KEX2 )

[InT][NTA]2 [In(NTA)2][Tf]

[In2T][NTA]2 [In(NTA)2][InT]

(1) (2)

The reaction parameters of reaction I, ∆1 ) InT - T and KEX1, have been evaluated for CNTA/CIn ) 35 by applying an iterative procedure to eq 3 whose derivation is given in the Supporting Information.

(

J. Phys. Chem. B, Vol. 112, No. 38, 2008 12171

)

2 CInCT/CNTA2 ∆A/CNTA2 (CIn + CT)/CNTA 1 ) + + ∆A ∆ε1KEX1 ∆ε1 ∆ε 2 1

(3) The analysis of the data shown in Figure 2 yields ∆1 ) (1.43 ( 0.03) × 104 M-1 cm-1 and KEX1 ) (8.6 ( 2.9) × 10-2 M. The equilibrium constant of reaction II, KEX2, and ∆2 ) In2T - InT have been evaluated according to eq 4, whose derivation is shown in the Supporting Information

∆A ) CT

∆ε1 + ∆ε2KEX2 1+

1 KEX1

×

[L]2

[ML2]

[ML2] [L]2

+ KEX2 ×

[ML2]

(4)

[L]2

where [L] ) [NTA] and [ML2] ) [In(NTA)2]. The value of ∆1 and that of KEX1 obtained from titration at CNTA/CIn ) 35 have been introduced in eq 4 as constant parameters. The nonlinear least-squares treatment of the data requires an iterative procedure, as shown in the Supporting Information, in order to evaluate the concentration variable [ML2]/[L].2 The reaction parameters are reported in Table 2 and the species distribution evaluated by the above parameters is given in Figure 5S (Supporting Information). The trend calculated according to eq 4 is shown in Figure 3. Kinetic Measurements. Stopped-flow experiments have been carried out by mixing transferrin with In(NTA)23- under the conditions CNTA g 10CIn g CT and using reactants concentrations such that reaction II is suppressed and the metal exchange is represented by the overall reaction I. Preliminary experiments revealed that the exchange reaction is biphasic. The first phase, completed in a few seconds, contains information about the initial steps of the interaction between the In-NTA complex and transferrin. The second phase, of smaller amplitude, persists for some hours after the mixing of the reactants and corresponds to a kinetic process already detected by other authors.35,36 Note that it is because of this slow phase that we had to resort to batch-wise titrations for the study of the equilibria. This slow phase has not been investigated while the rapid phase of the metal exchange reaction has been investigated by the stoppedflow technique and a typical experiment is shown in Figure 4. Since the reactant concentrations are such that the reacting system tends to an equilibrium position, the time constant, 1/τ, includes the contributions of both the forward and the reverse steps. Hence, this investigation provides simultaneous information about the feature of metal uptake and metal removal from transferrin. Rate Dependence on the Indium(III) Concentration. The dependence of 1/τ on the total concentration of indium(III), CIn, has been investigated at constant concentrations of all the remaining species and it was found to be linear (Figure 5). Rate Dependence on NTA Concentration. Keeping constant the concentrations of all reactants except that of nitrilotriacetate, the time constant was found to change with CNTA as shown in Figure 6. Granted that NTA was always in excess over the other reactants, no data could be collected for CNTA values larger than 2.4 × 10-2 M because a large excess of NTA inhibits the protein complex formation, thus making the amplitudes of the kinetic curves too small to allow a reliable analysis of the experiments to be made. Rate Dependence on the NaHCO3 Concentration. Keeping all concentrations constant except that of bicarbonate the dependence of 1/τ on CNaHCO3 was obtained as shown in Figure 7.

12172 J. Phys. Chem. B, Vol. 112, No. 38, 2008 The kinetics behavior could be rationalized on the basis of reaction Scheme 1, where Lf and Bf represent respectively the differently protonated forms of NTA and carbonate. The first step of the reaction corresponds to the fast formation of the quaternary intermediate MLTB. The second step leads to formation of an “open” ternary complex MTB* (minority) which in the third step converts to a “closed” complex MTB (majority), in agreement with the Bates argument.28 Assuming that MTB* is steady state, the concentration dependence of the time constant at pH ) 7.4 is obtained in the form of eq 5)

1/τ ) k1KARMLRTBCIn/(1 + (k-1/k2)CL)(1 + KARMLRTBCIn) + (k-2k-1/k2)CL/(1 + (k-1⁄k2)CL) (5) where CL ) CNTA, CB ) CNaHCO3, RML ) K′MLCL/(1 + K′MLCL + K′MLK′ML2CL2) and RTB ) Kb1CB/(1 + Kb1CB + Kb1Kb2CB2). K′ML, K′ML2, Kb1, and Kb2 are apparent equilibrium constants defined in the Supporting Information. The linear dependence of 1/τ on CIn, shown in Figure 5, indicates that KARMLRTBCIn , 1, which is reasonable, since the concentration of the quaternary intermediate In(NTA)TB is expected to be small. Hence eq 5 is reduced to eq 6.

1/τ ) k1KAK′MLKb1CBCLCIn/[(1 + (k-1/k2)CL) ×

(1 + K′MLCL + K′MLK′ML2CL2) × (1 + Kb1CB + Kb1Kb2CB2)] + (k-2k-1/k2)CL/[1 + (k-1/k2)CL] (6) Figure 8 shows the dependence of 1/τ on the indium concentration for a set of experiments were CIn and CNTA were simultaneously changed under the condition CNTA ) 10 × CIn. The good agreement between this data set and the trend calculated with the parameters of Table 3 and eq 6 confirms the validity of the proposed reaction scheme. The species distribution obtained for the investigated system is shown in Figure 6S of the Supporting Information. Rate Dependence on the Hydrogen Ion Concentration. Figure 9 shows the dependence of 1/τ on pH in the range between 5.5 e pH e 8.5. Acidity levels higher than pH ) 5.5 could not be investigated since under such conditions the amplitude of the signals were too small to allow reliable results to be obtained. Discussion The values of the acid dissociation constants of NTA (Table 1) agree with those measured by other authors.35,44 Concerning the stability constants of the In/NTA system our potentiometric value of log β11 (16.1) at 25 °C and I ) 0.2 M (NaClO4) largely differs from the value of 13.8 obtained by Harris35 by spectrophotometric measurements at 25 °C and I ) 0.1 M but approaches the value of 16.9 measured by potentiometry45 at 20 °C and I ) 0.1 M. Our value of log β12 (25.7) is higher by 1 order of magnitude compared to the value reported by Tuck (log β12 ) 24.4) at 20 °C and I ) 0.1 M45 and 2 orders of magnitude compared to the Harris et al. value (log β12 ) 23.7).35 The large spread in the log βij values coming from different sources are only partially due to difference in temperature and ionic strength. Most likely, the differences come from the large spread of the hydrolysis constants of In3+ ion used in the calculations. Further discrepancies can derive from the fact that polynuclear species of indium(III)46,47 have been neglected in all calculations. The values of the equilibrium constants for the exchange reactions KEX1 and KEX2 here measured could compare with log KEX1 ) -3.15 and log KEX2 ) -5.04, obtained by Lurie et

Biver et al. al.48 The first value, once corrected for partial protonation of competitive ligands used as binding agents lead to log KEX1 ) -1.1,35 in good agreement with our result (Table 2). Even if the precision of the methods currently employed for equilibrium measurements ranges within (0.15 logarithmic units, the spread between the values of binding constants reported for Intransferrin is much larger. Actually, temperature, ionic strength, nature of the added salt (which could cause conformational changes during binding), and origin of transferrin could be responsible for the observed discrepancies. The kinetic behavior of the system reveals that the metal exchange between NTA and transferrin occurs in the fast phase and involves a reaction between the 1:1 indium-NTA complex (ML) and the 1:1 transferrin-carbonate complex (TB). Neither the prevailing complex ML2 nor the TB2 form of transferrin are directly involved in the exchange process, but simply act as metal and protein reservoirs. Which of the two C- and N-terminal lobes of transferrin will allocate the metal ion in this phase is not known.36 It is most probable that in the fast phase both the C- and N-site will accept the metal ion with similar avidity, while metal redistribution, possibly in favor of the C-site, will take place in the slow phase. Actually, if metal redistribution between sites does occur only through a dissociation-reassociation process, a slow phase should be expected. Another phenomenon that can explain the slow phase could be related to slow donation of In3+ from polynuclear In(III) species to transferrin.49 In the first of the binding steps shown in Scheme 1 (step KA) the water molecules still present in the metal coordination shell of InNTA are replaced by CO32ion present in TB. This process, which leads to the formation of a quaternary adduct, is fast owing to the lability of In(H2O)63+ ion. Moreover, it could be argued that the presence of NTA in the metal coordination shell will enhance the lability of the residual water molecules in InL.50 Recently, a quaternary complex has been obtained by reaction of Fe-NTA with a 18KDa fragment of ovotransferrin.51 The X-ray structure of this complex shows that CO32- is bound to the Fe3+ coordination shell in a bidentate manner and that the NTA moiety appears to be neither fully bound to the cation nor fully released. Similar representation could hold for the quaternary adduct MLTB of Scheme 1; if the ligands in this complex are not fully coordinated to In3+, as in the case of Fe3+, then its fast formation could be justified. The reverse step, corresponding to dissociation of MLTB to reform ML and TB should also be fast as deduced from the small value of KA. That the latter parameter should be small can be argued from the fact that noticeable amounts of the quaternary complex have never been observed in solution.28 The evolution of MLTB toward the ternary complex MTB involves the full breaking of the metal-NTA bonds and full formation of metal-protein bonds accompanied by a deprotonation process probably involving the tyrosine residues. This is a rather complex process that could be partially responsible for the activation energy of the total process. The value of k1KA (Table 3) provides the limits of the k1 values. Assuming for k1 a maximum value of about 109 s-1, corresponding to diffusional limit,52 then a lower limit of about 10-3 M-1 should be estimated for KA. On the other hand, KA should not exceed the value of 104 M-1, otherwise a downward deviation from linearity larger than 20% would be observed in the plot of Figure 5 for CIn ) 4 × 10-4 M, contrary to experiment; hence, it turns out that k1 g 102 s-1. The reverse step displays saturation kinetics with respect to ligand; a path firstorder with respect to ligand, and parallel to the saturation path, observed by Harris and his school53 in many cases, seems to be absent in the present system. The values of k-2 for InT reaction with NTA should be compared with the values of kmax (min-1) reported by Harris et al.36 for donation of In3+ from transferrin to

Mechanism of Indium(III) Exchange citrate. This parameter is related to k-2 by the relationship kmax/60 ) k-2; hence, in the case of citrate it turns out that k-2 ) 1.1 × 10-2 s-1. In principle, the kinetics of the process MTB* a MTB should not depend on ligand nature and concentration, so similar values of k-2 should be expected. Actually, the value of k-2 for NTA is about 30 times higher than that measured for citrate. However, the ionic strength values are different, being I ) 0.2 M (NaClO4) for NTA and 0.05 M (HEPES buffer) for citrate. The difference in ionic strength and in the nature of the background electrolyte could be the cause of our higher value of k-2. The MTB* a MTB step plays a crucial role in the control of the rate of the exchange process. Actually, the rate of indium release at high NTA loading is limited by the k-2 value, while the rate of indium uptake decreases since increase of CNTA results in increasing amount of the unreactive species ML2. At low NTA loading, on the other hand, step MTB* a MTB intervenes in the metal release process as a pre-equilibrium, while the rate of reaction is controlled by CNTA in both directions. Under these circumstances the key step of the overall exchange reaction is the interconversion MLTB/ MTB*. The dependence of 1/τ on pH can be explained taking into account, on one hand, the role of the proton in regulating the concentrations of active species and, on the other hand, the necessity of making weaker, through protonation, the bonds to be broken. Since the prevailing form of unbound NTA and bicarbonate are HNTA2- and HCO3-, while the corresponding bound forms are NTA3- and CO32-, a slight reduction of the rate of the forward reaction would result on increasing [H+], as observed between pH 8.5 and 8.0. A flat minimum (between pH 7.5 and 7) is displayed by the plot of Figure 9, followed by a rate increase as the acidity is raised up to pH ) 6. This behavior suggests that the effect exerted by [H+] on the forward reaction is counterbalanced by an opposite effect to be ascribed to the step of indium release by transferrin. Tyrosine protonation, with consequent weakening of the Tyro-In bonds, could be responsible for the observed rate increase on lowering pH. At pH values lower than 5.5 the amplitudes of the kinetic process became very small, revealing that no binding of TB to InNTA can occur at these acidity levels. Similar behavior has been observed also in the case of iron(III), and it is known that metal release from transferrin, triggered by a pH jump from 7.4 to 5.5, plays a fundamental role in the cycle of iron transport and uptake.54 Supporting Information Available: Figures showing potentiometric titration curves of NTA and In-NTA, species distribution plots for NTA, indium-NTA complexes, and indium-NTA-transferrin systems, and the difference spectra of transferrin in the presence of different amounts of [In(NTA)2]3-, text giving the mathematical derivation of eqs 3, 4, and 5 and a scheme showing the reactions used in the derivation of eq 5. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Bridges, K. R. Transferrin and Iron Transport Physiology; Harvard Medical School: Cambridge, MA, 2001, on-line article. (2) Baker, E. N. AdV. Inorg. Chem., 1994, 41, 389–463. (3) Chasteen, N. D. Coord. Chem. ReV. 1977, 22, 1–36. (4) El Hage Chahine, J. M.; Fain, D. J. Chem. Soc., Dalton Trans. 1993, 3137–3143. (5) Pakdaman, R.; El Hage Chahine, J. M. Eur. J. Biochem. 1996, 236, 922–931. (6) Abdallah, F. B.; El Hage Chahine, J. M. Eur. J. Biochem. 1998, 258, 1022–1031.

J. Phys. Chem. B, Vol. 112, No. 38, 2008 12173 (7) Pakdaman, R.; Abdallah, F. B.; El Hage Chahine, J. M. J. Mol. Biol. 1999, 293, 1273–1284. (8) Brook, C. E.; Harris, W. R.; Spilling, C. D.; Wang, P.; Harburn, J. J.; Srisung, S. Inorg. Chem. 2005, 44 (14), 5183–5191. (9) Halbrooks, P. J.; Mason, A. B.; Adams, T. E.; Briggs, S. K.; Everse, S. J. J. Mol. Biol. 2004, 339, 217–226. (10) Halbrooks, P. J.; He, Q. Y.; Briggs, S. K.; Everse, S. J.; Smith, V. C.; MacGillivray, R. T.; Mason, A. B. Biochemistry 2003, 42, 3701– 3707. (11) Boukhalfa, H.; Anderson, D. S.; Mietzner, T. A.; Crumbliss, A. L. J. Biol. Inorg. Chem. 2003, 8, 881–892. (12) He, Q. Y.; Woodwarth, R. C.; Chasteen, N. D. J. Biol. Inorg. Chem. 2003, 8, 635–643. (13) He, Q. Y.; Mason, A. B.; Nguyen, V.; MacGillivray, R. T. A.; Woodworth, R. C. Biochem. J. 2000, 350 (3), 909–915. (14) Harris, W. R.; Bao, G. Polyhedron 1997, 16, 1069–1079. (15) El Hage Chahine, J. M.; Pakdaman, R. Eur. J. Biochem. 1995, 236, 1102–1110. (16) El Hage Chahine, J. M.; Fain, D. Eur. J. Biochem. 1994, 223, 581– 587. (17) Egan, T. J.; Zak, O.; Aisen, P. Biophys. Biochem. 1993, 32, 8162– 8167. (18) Kretchmar, S. A.; Raymond, K. N. J. Am. Chem. Soc. 1993, 115, 6212–6218. (19) Kretchmar Nguyen, S. A.; Craig, A.; Raymond, K. N. J. Am. Chem. Soc. 1993, 115, 6758–6764. (20) Egan, T. J.; Ross, D. C.; Purves, L, R.; Adams, P. A. Inorg. Chem. 1992, 31, 1994–1998. (21) Marquez, H. M.; Watson, D. L.; Egan, T. J. Inorg. Chem. 1991, 30, 3758–3762. (22) Thompson, C. P.; Gradi, J. K.; Chasteen, N. D. J. Biol. Chem. 1986, 26128, 13128–13134. (23) Baldwin, D. A.; De Sousa, D. M. R.; Von Wandruszka, R. M. Biochim. Biophys. Acta 1982, 719 (1), 140–146. (24) Baldwin, D. A.; De Sousa, D. M. R. Biochem. Biophys. Res. Commun. 1981, 99, 1101–1107. (25) Baldwin, D. A. Biochim. Biophys. Acta 1980, 623 (1), 183–198. (26) Morgan, E. H. Biochim. Biophys. Acta 1977, 499 (1), 169–177. (27) Johnson, D. O. Biochim. Biophys. Acta 1969, 177 (1), 113–117. (28) Cowart, R. E.; Kojima, N.; Bates, G. W. J. Biol. Chem. 1982, 257, 7560–7565. (29) Sun, H.; Li, H.; Sadler, P. Chem. ReV. 1999, 99, 2817–2842. (30) Hayes R. L.; Hu¨bner K. F. Metal Ions in Biol. System; Sigel, H., Ed.; Dekker, Inc.: New York, 1983; Vol. 16. (31) Van Hulle, M.; De Cremer, K.; Cornelis, R. Fres. J. Anal. Chem. 2000, 368, 293–296. (32) Harris, W. R.; Pecoraro, V. L. Biochem. 1983, 22, 292–299. (33) Harris, W. R. Biochemistry 1983, 22, 3920–3926. (34) Harris, W. R.; Sheldon, J. Inorg. Chem. 1990, 29, 119–124. (35) Harris, W. R.; Chen, Y.; Wein, K. Inorg. Chem. 1994, 33, 4991– 4998. (36) Harris, W. R.; Wang, Z.; Brook, C.; Yang, B.; Islam, A. Inorg. Chem. 2003, 42, 5880–5889. (37) Ricciu, A.; Secco, F.; Venturini, M.; Garcia, B.; Leal, M. J. Chem. Eur. J. 2001, 7, 4613–4620. (38) Harris, W. R.; Bali, P. K.; Crouley, M. M. Inorg. Chem. 1992, 31, 2700–2705. (39) Cheng, K. L. Anal. Chem. 1955, 27, 1582–1583. (40) Gran, G. Analyst 1952, 77, 661–671. (41) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739–1753. (42) Biver, T.; Cavazza, C.; Secco, F.; Venturini, M. J. Inorg. Biochem. 2007, 101, 461–469. (43) Provencher, S. W. J. Chem. Phys. 1976, 64, 2772–2777. (44) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum Press: New York, 1974,Vol. 1. (45) Tuck, D. G. Pure Appl. Chem. 1983, 55, 1477–1528. (46) Baes, C. F.; Mesmer, R. E., The hydrolysis of cations; John Wiley and Sons: New York, 1976. (47) Eyring, E. M.; Owen, J. D. J. Phys. Chem. 1970, 74, 1825–1828. (48) Lurie, D. J.; Smith, F. A.; Shukri, A. Int. J. Appl. Radiat. Isot. 1985, 36, 57–62. (49) Harris, W. R. Struct. Bonding (Berlin) 1998, 92, 121–162. (50) Burgess, J. Metal Ions in Solution; Horwood, Chichester, 1978. (51) Kruser, P.; Hall, D. R.; Haw, M. L.; Neu, M.; Evans, R. W.; Lindley, P. F. Acta Crystallogr. 2002, D58, 777–783. (52) Eigen, M.; Kruse, W.; Maass, G.; De Maeyer, L. Prog. React. Kinet. 1964, 2, 286–318. (53) Harris, W. R.; Messori, L. Coord. Chem. ReV. 2002, 228, 237– 262. (54) May, W. S.; Cuatrecasas, P. J. Membr. Biol. 1985, 88, 205–215.

JP8045033