SAM

Mar 29, 2008 - effect of SAM thickness on the electron-transfer rate has been studied, and ... group on the SAM causes the cytochrome c to bind electr...
4 downloads 0 Views 128KB Size
J. Phys. Chem. C 2008, 112, 6571-6576

6571

Electron-Transfer Kinetics of Covalently Attached Cytochrome c/SAM/Au Electrode Assemblies Kathryn L. Davis, Brianna J. Drews, Hongjun Yue, and David H. Waldeck* Department of Chemistry, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260

Kathryn Knorr and Rose A. Clark Department of Chemistry, Mathematics and Physical Sciences, Saint Francis UniVersity, Loretto, PennsylVania 15940 ReceiVed: December 17, 2007; In Final Form: February 5, 2008

The rotational motion of cytochrome c has been restricted by cross-linking it to mixed self-assembled monolayers (SAMs) with the compositions S-(CH2)mCOOH/S-(CH2)nOH on gold electrodes via the formation of amide bonds between lysine residues on the protein and terminal carboxylate groups of the SAM. The effect of SAM thickness on the electron-transfer rate has been studied, and two main observations are drawn. First, the electron-transfer rate displays the same qualitative dependence on SAM thickness that was previously reported for electrostatically adsorbed and pyridine-ligated assemblies, suggesting a tunneling mechanism at long distance and some other rate-limiting process at short distance. Second, a significant effect on the rate is observed for mixed SAMs having a hydroxyl-terminated alkanethiol diluent when the diluent is more than one methylene group shorter than the carboxylic acid alkanethiol. These conclusions suggest that largeamplitude protein motion (i.e., gating) is not rate-limiting at short distance, though smaller-amplitude motions cannot be ruled out.

Introduction Because it is small, easily purified, and well-characterized, cytochrome c is a favorite target for electron-transfer protein studies, both to clarify facets of its own electron-transfer behavior and as a model system for larger, less well-understood systems. Direct electrochemistry of cytochrome c immobilized at chemically modified electrodes provides an attractive model for probing electron transfer in supramolecular assemblies. Immobilization is often achieved by coating gold or silver electrodes with self-assembled monolayer (SAM) films that are composed of alkanethiols terminated with a headgroup that is capable of interacting with cytochrome c. Such SAM systems form compact, ordered, insulating films on the electrode surface. In addition to changing solution properties, such as pH, ionic strength, or viscosity, manipulation of the SAM chemical properties can provide more information about the nature of the electron-transfer process. For example, changing the length of the methylene chain can be used to change the electronic coupling between cytochrome c and the electrode surface. Also changing the headgroup can affect the properties of the resultant electron-transfer complex. Placing a carboxylic acid terminal group on the SAM causes the cytochrome c to bind electrostatically to the surface,1,2 whereas placement of a receptor ligand for cytochrome c’s heme iron can be used to directly wire the redox center to the electrode via displacement of the native Met80 ligand.3-5 Regardless of the nature of the cytochrome c/SAM headgroup interaction, cytochrome c’s electron-transfer rate has an unusual dependence on SAM thickness. Two distinct regions are * To whom correspondence should be addressed. Email: dave@ pitt.edu.

identifiable in a plot of ln(k0) versus the number of methylene groups in the SAM. For long methylene chain lengths, the rate decreases exponentially as the number of methylenes increases and can be attributed to a tunneling mechanism. On the other hand, at short methylene chain lengths, the rate becomes nearly distance-independent, here referred to as the plateau region. While the nature of the electron-transfer mechanism clearly changes from tunneling control for thick films to a different rate-limiting step in the plateau region, the detailed nature of the rate-limiting process (e.g., conformational gating,1 protoncoupled electron transfer modulated by electric field effects,6 or a friction control mechanism3,7,8) remains under investigation. Preparation of covalently bound cytochrome c electrontransfer complexes using carbodiimide cross-linking reagents is a well-established technique. Electron-transfer partners include a variety of biomolecules such as cytochrome b5,9 cytochrome c peroxidase,10-12 cytochrome c oxidase,13,14 and plastocyanin,15-18 as well as nonbiological moieties including N-acetyl cysteineterminated SAMs,19 -COOH-terminated SAMs,7,20-23 inorganic supports,24-29 and negatively charged succinate-terminated chromatography resin.30 In each case, cross-linking occurs via one or more amide bonds formed between lysine residues on the cytochrome c surface and a complementary carboxyl group on the partner molecule(s), as shown in Figure 1. The resultant complexes are stable, functional, and similar in structure to their electrostatic counterparts,9,12,21 providing a useful way to probe the effect of the protein’s surface flexibility on the electrontransfer rate. The aim of this study was to examine the electron-transfer behavior of cytochrome c that has been covalently bound to the -COOH termini of SAMs of varying length. Unlike

10.1021/jp711834t CCC: $40.75 © 2008 American Chemical Society Published on Web 03/29/2008

6572 J. Phys. Chem. C, Vol. 112, No. 16, 2008

Figure 1. A diagram of a covalent SAM/cytochrome c assembly resulting from the formation of multiple amide bonds between -COOH termini on the SAM and lysine residues on cytochrome c. The SAM shown here is a mixed monolayer of 2:3 (CH2)5COOH/(CH2)4OH.

electrostatic complexes, in which cytochrome c is immobilized at the -COOH headgroups via the formation of salt bridges with lysine groups on the protein, a cross-linking reagent is used to convert salt bridges into amide bonds (see Figure 1), resulting in a binding configuration in which the rotational diffusion/ rearrangement of cytochrome c at the SAM/solution interface is restricted. Thus, it is possible to “tie” the protein into a particular orientation, which may or may not be favorable for electron transfer. The distance dependence of the rate constant in covalent assemblies is therefore of great interest. If the distance dependence is qualitatively and/or quantitatively similar to that of electrostatic complexes, it indicates that largeamplitude rotational diffusion of cytochrome c on the SAM surface, that is, conformational gating, is likely not the ratelimiting process at short distances. The results of this study suggest that this is the case. Using the water-soluble carbodiimide cross-linking reagent N-cyclohexyl-N′-(2-morpholinoethyl)carbodiimide methyl-p-toluenesulfonate (CMC), cytochrome c was covalently attached to pure S-(CH2)mCOOH SAMs (henceforth abbreviated (CH2)mCOOH) or mixed S-(CH2)mCOOH/S-(CH2)nOH SAMs (henceforth (CH2)mCOOH/(CH2)nOH) of varying lengths. The data show that the protein is covalently bound and that its electron-transfer behavior is similar to that observed for electrostatically adsorbed cytochrome c. This finding indicates that rotational or largeamplitude rearrangement of the protein is not the limiting factor for electron transfer in the plateau region, and a different mechanism should be considered. Experimental Section Reagents and Materials. Water used in the experiments was purified using a Barnstead-Nanopure system and had a resistance of greater than 18 MΩ‚cm. N-Cyclohexyl-N′-(2-morpholinoethyl)carbodiimide methyl-p-toluenesulfonate (CMC), 16-mercaptohexadecanoic acid, 15-mercaptopentadecanoic acid, 12mercaptododecanoic acid, 11-mercaptoundecanoic acid, 6mercapto-1-hexanol, 4-mercapto-1-butanol, and 2-mercaptoethanol were purchased from Sigma-Aldrich. 7-Carboxy-1heptanethiol, 5-carboxy-1-pentanethiol, 3-carboxy-1-propanethiol, and 8-hydroxy-1-octanethiol were purchased from Dojindo. 14-Mercaptotetradecanol was synthesized according to a procedure reported elsewhere.31 Horse heart cytochrome c was purchased from Sigma-Aldrich (95% purity) and used without any further purification.

Davis et al. Electrode Preparation. A gold ball electrode was fabricated from gold wire. The gold wire was cleaned by 1-2 h reflux in nitric acid and rinsed with deionized water. The electrode was created by heating the tip of the wire in a flame until a small ball formed (ca. 0.1 mm in diameter). This surface was made smooth by heating the gold ball in a flame and then quickly cooling it in deionized water. This procedure was repeated 1020 times. The electrode was then sealed in a glass capillary tube, which was bonded to the gold wire, leaving the smooth working surface exposed. SAMs were formed by submerging the gold ball in a thiol solution for 12-16 h. The composition of the SAM solution was 50% HS(CH2)mCOOH/50% HS(CH2)nOH or 40% HS(CH2)mCOOH/60% HS(CH2)nOH; the concentration was 2 mM total thiol content when a diluent was used or 2 mM HS(CH2)mCOOH in the case of the pure SAM. Diluent hydroxylterminated thiol was used to form a more compact and stable SAM; since it is known that cytochrome c does not adsorb well to the hydroxyl moiety,32 the primary interaction remains between cytochrome c and the -COOH group. In preparation for forming covalent assemblies, the electrodes were removed from the SAM solution and rinsed with ethanol and then with deionized water. The rinsed electrodes were then placed into a solution of the cross-linking reagent N-cyclohexylN′-(2-morpholinoethyl)carbodiimide methyl-p-toluenesulfonate (CMC) in 100 mM, pH ) 7.0, potassium phosphate buffer in H2O (20 mg of CMC/10 mL of buffer) for 30 min. They were then transferred to 40 µL of a cytochrome c solution (30-40 µM) in 4.4 mM, pH)8.0, potassium phosphate buffer in H2O for 30-60 min. Electrochemical Measurements. Cyclic voltammetry was performed using a CH Instruments 618B electrochemical analyzer. The three-electrode cell contained a Ag/AgCl (1 M KCl or 3 M KCl) reference electrode, a platinum wire counter electrode, and the chemically modified gold wire electrode. Electrodes were rinsed in a high ionic strength buffer (I ) 200 mM) prior to voltammetric experiments to remove any weakly bound cytochrome c, after which they were transferred to the electrode holder and placed into 40 mM (pH)7, I ) 80 mM) potassium phosphate buffer in H2O. Voltammograms were collected at increasing scan rates from 0.1 to 1000 V/s for (CH2)5COOH/(CH2)nOH systems, 0.01 to 100 V/s for (CH2)10COOH/(CH2)nOH systems, and 0.005 to 10 V/s for (CH2)14COOH/(CH2)nOH and (CH2)15COOH/(CH2)nOH systems. Results Verification of Covalent Attachment. Covalent cytochrome c/SAM assemblies were rinsed vigorously with high ionic strength buffer (I ) 200 mM) prior to experimentation. Such a rinsing procedure is believed to remove electrostatically adsorbed and physisorbed cytochrome c from the surface.7 To ensure that this is the case and that no significant amount of strongly adsorbed electrostatic cytochrome c remained on the electrode, control experiments were performed. The electrostatic and covalent cytochrome c/(CH2)5COOH/(CH2)4OH SAM assemblies were rinsed first with I ≈10 mM buffer, followed by collection of a 10-1000 V/s voltammogram set, and then rinsed with an I ) 200 mM ionic strength buffer, and another voltammogram set was recorded. The surface coverage, formal potential, and rate were compared for each rinse. A full description of this experiment can be found in the Supporting Information. In brief, these experiments indicate that some electrostatically adsorbed protein remains on the covalent assemblies after high ionic strength rinsing but that the

Cytochrome c/Mixed SAM/Au Electrode Assemblies

Figure 2. (A) Cyclic voltammograms for cytochrome c covalently attached to a (CH2)5COOH/(CH2)6OH SAM collected at 10 (brown), 20 (blue), and 50 (red) V/s. (B) Fitting of Marcus theory (solid curves) to experimental peak separations (circles). The fitting corresponds to a rate constant of 5250 s-1.

voltammetric response is dominated by the covalent assembly. These studies show clearly that covalent attachment was achieved and covalently attached cytochrome c’s voltammetry can be distinguished from that of the electrostatic assemblies following rinsing with high ionic strength buffer. Voltammetry of Covalently Attached Cytochrome c/SAM Assemblies. Figures 2A and 3A show characteristic voltammograms for cytochrome c covalently attached to mixed (CH2)mCOOH/(CH2)nOH films; Table 1 details some of the properties of these voltammograms. The data in Figure 2A for the (CH2)5COOH/(CH2)6OH assembly correspond to the plateau region (n < 7) and display a small peak separation (∆Ep < 25 mV in all but the pure (CH2)5COOH case) and a peak width at halfmaximum (FWHM) of approximately 100 mV at 10 V/s scan rate, which is close to the fully reversible value of 90.6 mV.33 These data indicate quasireversible kinetics at slow scan rates. The data in Figure 3A for (CH2)15COOH/(CH2)14OH correspond to the tunneling region (n > 7). In this case, the FWHM increases and ∆Ep increases substantially with scan rate, indicating the increasing irreversibility of electron transfer for thicker SAMs. The observed formal potentials for the covalent assemblies ranged from -48 to +41 mV (vs 1 M Ag/AgCl) depending on the SAM chain length and composition, though on most SAMs E0′ is shifted negative of that observed for electrostatically bound cytochrome c.2,34 Collinson et al.21 reported E0′ ) -50 mV for cytochrome c covalently attached to a pure (CH2)15COOH SAM in 4.4 mM, pH ) 7.0 buffer. Previous work by this group7 showed that E0′ for cytochrome c covalently attached to a mixed (CH2)15COOH/(CH2)11OH film becomes more positive with increasing ionic strength. After correcting for the ionic strength difference (I ) 80 mM in this study), the value reported by Bowden would correspond to E0′ ≈ -20 mV at I ) 80 mM. This corrected value agrees well with most of the E0′ values reported in Table 1. Discrepancies for the assemblies on (CH2)10COOH/(CH2)8OH, (CH2)10COOH/(CH2)11OH, (CH2)14COOH/ (CH2)14OH, and (CH2)15COOH/(CH2)14OH are evident and may be caused by the presence and length of the diluent thiol, though no further studies were conducted to determine the origin of the E0′ differences in these cases. The standard electrochemical rate constant can be obtained by quantifying how the faradaic peak position shifts with scan

J. Phys. Chem. C, Vol. 112, No. 16, 2008 6573 rate (see Figures 2B and 3B). These data show that the peak separation increases with increasing scan rate. A curve generated by Marcus theory (Figures 2B and 3B, solid curves)35,36 was used to fit the anodic and cathodic waves (circles). The voltammetry is symmetric, and the cathodic and anodic rate constants differ by no more than 10-15%. Table 1 summarizes the rate constants obtained by this method. Comparison to Electrostatic Adsorption and Pyridine Ligation. As shown in Figure 4, the electron-transfer rate in covalent cytochrome c/SAM assemblies (red symbols) depends on SAM thickness in a manner similar to the electrostatic adsorption (blue circles) and pyridine-ligated (green stars) cases. Distinct tunneling and plateau regions are evident; however, the turnover between regions appears at n < 7 methylene groups rather than n < 6 for electrostatic assemblies and n < 12 for pyridine-ligated systems. This shift of the turnover point between the covalent data and the electrostatic data could be caused by stronger coupling of the iron center to the electrode through covalent bonds; the large shift in this onset between the pyridineligated system and the electrostatic system has been explained by a change in the tunneling pathway.3 In the tunneling region, the electron-transfer rate is slightly higher for the covalent system than for the electrostatically adsorbed cytochrome c, provided there is no diluent hydroxylterminated thiol or that it is close in length to the -COOHterminated thiol. This finding agrees with previous observations made by Bowden’s group.21 On the other hand, the electrontransfer rate in the plateau region appears to be higher for the covalent assemblies than for either the electrostatic or pyridineligated systems. Effect of SAM Composition on Rate. Recent studies by Dolidze et al.37 and Yue et al.31 demonstrated that the concentration and chain length of the diluent thiol in a mixed SAM have an effect on the electron-transfer rate for electrostatically adsorbed and pyridine-ligated cytochrome c, affecting both the electron-transfer rate constant and the formal potential. Dolidze et al. report that the presence of a significantly shorter hydroxyl-terminated diluent thiol (n ) m - 3) for a mixed SAM in the plateau region resulted in a slowing of the rate. Both studies conclude that the presence of a shorter, hydroxylterminated diluent thiol leads to rate enhancement in the tunneling region. The results of this study agree with the results of both Dolidze et al. and Yue et al. for the tunneling region, in that the presence of a diluent thiol can cause a dramatic increase in the rate constant. If the two components of the mixed (CH2)mCOOH/(CH2)nOH film have a comparable number of methylene groups (m - 1 e n e m + 1), the rate is similar to that on a pure (CH2)mCOOH film; however, as n becomes significantly smaller than m (n e m - 2), the rate becomes significantly faster (see Table 1). These data suggest that electron tunneling in covalent systems, as in electrostatic systems, proceeds through multiple pathways rather than exclusively through -COOH-terminated alkyl chains under such conditions. In the plateau region, however, the results reported here differ from those of Dolidze et al. for cytochrome c electrostatically immobilized on (CH2)5COOH/(CH2)2OH; in this case, the rate increased upon addition of a shorter diluent thiol, though, as long as n < m, the chain length of that thiol did not have a large effect on the observed rate constant. On the other hand, Dolidze et al. report a substantial slowing of the rate. This may be due to the much larger length disparity between the carboxylterminated thiol and the hydroxyl-terminated thiol. The electrontransfer rate constant for cytochrome c electrostatically adsorbed

6574 J. Phys. Chem. C, Vol. 112, No. 16, 2008

Davis et al.

TABLE 1: Rate Constants and Voltammetric Properties for Cytochrome c Covalently Attached to (CH2)mCOOH/(CH2)nOH SAMs m

n

m:n

average k0 (s-1)

3

2

1:1

3200 ( 800

5

0 4 6

1:0 1:1 2:3

7

6

10

0 8 11

ln(average k0)

∆Ep (mV)a

E0 (mV)a

FWHM (mV)b

trials

8.06 ( 0.25

22 ( 9

-24 ( 6

108 ( 32

7

1100 ( 700 3700 ( 800 4700 ( 700

7.04 ( 0.63 8.21 ( 0.22 8.46 ( 0.14

52 ( 21 22 ( 9 13 ( 9

-5 ( 11 -22 ( 5 -13 ( 4

118 ( 4 107 ( 4 102 ( 2

4 10 8

1:1

3300 ( 700

8.11 ( 0.21

36 ( 13

-28 ( 1

115 ( 1

5

1:0 2:3 2:3

120 ( 80 630 ( 200 160 ( 60

4.82 ( 1.60 6.44 ( 0.29 5.10 ( 0.35

64 ( 13 18 ( 12 76 ( 7

-26 ( 13 0(8 -14 ( 14

111 ( 10 114 ( 3 129 ( 11

4 19 7

11

11

2:3

53 ( 4

3.97 ( 0.08

15 ( 20

27 ( 17

98 ( 11

7

14

0 11 14

1:0 2:3 2:3

2.2 ( 2 15 ( 1 2.1 ( 1

0.77 ( 1.10 2.71 ( 0.06 0.72 ( 0.63

71 ( 11 20 ( 35 24 ( 20

-9 ( 41 -63 ( 0 21 ( 13

141 ( 49 109 ( 10 104 ( 4

4 3 4

15

14

2:3

0.39 ( 0.2

-0.88 ( 0.42

87 ( 36

23 ( 7

100 ( 15

9

a

∆Ep and FWHM are given at the following scan rates: 20 V/s for (CH2)3COOH/(CH2)nOH, (CH2)5COOH/(CH2)nOH, and (CH2)7OH/(CH2)nOH assemblies; 2 V/s for pure (CH2)10COOH and (CH2)10COOH/(CH2)8OH; 5 V/s for (CH2)10COOH/(CH2)11OH; 0.1 V/s for (CH2)11COOH/(CH2)11OH; 0.2 V/s for pure (CH2)14COOH; 0.5 V/s for (CH2)14COOH/(CH2)11COOH; 0.02 V/s for (CH2)14COOH/(CH2)14OH; and 0.01 V/s for (CH2)15COOH/ (CH2)14OH. b Reported with respect to Ag/AgCl reference electrode (1 M KCl).

Figure 3. (A) Cyclic voltammograms for cytochrome c covalently attached to a (CH2)15COOH/(CH2)14OH SAM collected at 25 (green), 50 (brown), 100 (blue), and 200 (red) mV/s. (B) Fitting of Marcus theory (solid curves) to experimental peak separations (circles). The fitting corresponds to a rate constant of 0.475 s-1.

on mixed (CH2)5COOH/(CH2)4OH films has also been measured, and it is close to the value reported here for the analogous covalent system, indicating that rate enhancement through the presence of a hydroxyl-terminated diluent thiol is also observed in electrostatic systems.38 Determination of the Tunneling Decay Coefficient. The exponential decay of the rate constant with chain length can be quantified as exp(-βm), where β is the tunneling decay coefficient and m is the methylene number in the -COOH chain. Since the rate constant measured at each SAM assembly has a unique uncertainty, a weighted linear regression method39 was employed to calculate β. The β values determined by this method are summarized in Table 2, along with a comparison to values obtained via standard (unweighted) regression analysis. When mixed SAMs having n e m - 2 are not included in the analysis, the tunneling decay coefficient determined by both methods is the same within error as for the case of electrostatic systems,34 where β ≈ 1.1, and close to that of the pyridineligated system,5 where β ≈ 1.2. The regression lines determined using the weighted method are provided in the Supporting Information.

Figure 4. Distance dependence of the rate constant in covalently attached cytochrome c/SAM assemblies (red symbols) and comparison to the electrostatic (blue circles) and pyridine-ligated (green stars) cases: ([) (CH2)mCOOH/(CH2)m-1OH SAMs, (+) pure (CH2)mCOOH film, (4) (CH2)mCOOH/(CH2)m+1OH, (2) (CH2)mCOOH/(CH2)mOH, (0) (CH2)10COOH/(CH2)8OH, and (9) (CH2)14COOH/(CH2)11OH. m denotes the number of methylene groups in the carboxyl-terminated SAM component, while n denotes the number of methylenes in the hydroxylterminated diluent. Electrostatic values for pure -COOH-terminated films were reported by Niki et al.34 and Bowden et al.;2 Py-ligated values were determined in earlier studies from this group.4,5 The dashed lines correspond to the rate constant predicted for a tunneling mechanism.

TABLE 2: Determining the Tunneling Decay Coefficient methylenes in diluent thiol (n)

number of points

β (weighted)

β (unweighted)

n ) 0 (pure -COOH) m-1