Dimers of Azurin as model systems for electron transfer
Jongh, Thyra Estrid de
Citation
Jongh, T. E. de. (2006, September 12). Dimers of Azurin as model systems for electron
transfer. Retrieved from https://hdl.handle.net/1887/4554
Version:
Corrected Publisher’s Version
License:
Licence agreement concerning inclusion of doctoral thesis in the
Institutional Repository of the University of Leiden
Downloaded from:
https://hdl.handle.net/1887/4554
Chapter
7
H otw iring redox proteins by ligand
reconstitution using conducting
m olecular w ires
Thyra E. de Jongh, Sergey M ilikysiants,
M artina H uber, Adrianus M .C.H . van den
N ieuw endijk, M ark O verhand, M arcellus
U bbink and G erard W . Canters
132 C h a p te r 7
Abstract
133 C h a p te r 7
Introduction
Redox proteins may be able to play an important role in the development of nanoscale biosensors or bioelectronic components.[1-3] One of the main challenges
in bioelectrochemistry, however, remains the need to establish good electronic communication between a protein and an electrode. As the protein matrix is a natural insulator, direct electron transfer (ET) between a protein encapsulated redox centre and an electrode can only be achieved if there exists an electronic path along which the electrons can enter and exit the redox centre with sufficient ease. W hen the coupling becomes too weak, ET may require the involvement of mediatory relay stations that can be attached to the protein.[4-6] An attractive alternative can be
to substitute the native redox-cofactor, such as flavins or NADP, for functionalized derivatives that act as molecular ‘hotwires’.[7-13] Similarly, residues that ligate the
redox centre can be removed by site-directed mutagenesis to create mutants capable of binding exogenous ligands.[14]
The Type-1 blue copper protein azurin from Pseudom onas aeruginosa is a potential candidate for hotwiring. It allows direct interaction between its copper centre and the surroundings when a solvent exposed aperture is created by removal of the copper ligand His117. This H117G mutant is able to accommodate a variety of exogenous ligands such as Cl-, imidazole or pyridine with restoration of the W
T-like spectroscopic properties, such as an intense absorption at 630nm.[15-17] In the
absence of additional ligands the aperture is filled by solvent molecules and the protein exhibits characteristics of a Type-2 copper centre with a strong absorption maximum at 420 nm.[15-17] It was demonstrated by van Pouderoyen et al. that
bifunctional linkers with copper coordinating endgroups can promote non-covalent dimerization of H117G azurin.[18;19] By analogy, we have developed a series of
134 C h a p te r 7
The use of molecular wires in nanoelectronics has attracted intense interest and many of the parameters that affect the efficiency of such donor-bridge-acceptor (DBA) systems have been explored.[20-23] By variation of the electronic and structural
properties of the bridge structure some general insights into the viability of the hotwiring approach for development of copper protein-based biosensors are here explored.
Materials & Methods
Synthesis of linkersThe linkers 2-6 were kindly provided by A.M .C.H. van den Nieuwendijk and were synthesized as described below.
T etraethyl [1,4-phenylenebis(methylene)]bis(phosphonate)[24]
A mixture of D,D’-dibromo-p-xylene (3.00 g, 11.4 mmol) and triethyl phosphite (5.0 P/ PPRO ZDVKHDWHGWRÜ&DQGVWLUUHGDWWKLVWHPSHUDWXUHIRUKRXUV Upon cooling the mixture solidified. It was dissolved in a small volume of CHCl3 and purified by silicagel column chromatography. Eluting with a mixture of PE/
Figure 7.1: ligands and ligand wires for coordination to H117G azurin: [1] pyridine [2]4,4’-(1,4-phenylenediethane-2,1-diyl)dipyridine [3]4-4’-[1,4-phenylenedi(E)ethene-2,1-diyl]dipyridine [4]3,3’-[1,4-phenylenedi(E)ethene-2,1-diyl]dipyridine [5] 4,4’-(1,4-phenylenediethyne-2,1-diyl)dipyridine [6] 4,4’-(1E,3E,5E)-hexa-1,3,5-triene-1,6-diyldipyridine [7] 4,4’-(2,2’-bithiene-5,5’-diyl)dipyridine [8] 1-methyl-4-(5’-pyridin-4-yl-2,2’-bithien-5-yl)pyridinium
135 C h a p te r 7
EtOAc = 1/1 removed impurities. Eluting with DCM/MeOH = 9/1 afforded 4.20 g (97%) of the target compound as a white solid. [1H-NMR (CDCl
3, G, ppm) 7.25
(s, 4H, aromatic); 4.00 (m, 8H, OCH2); 3.13 (d, 2H, JH ,P = 19.7 Hz); 1.24 (t, 12H, CH3).31P-NMR (CDCl
3, G, ppm), 26.8 (s)]
Tetraethyl (2E)-but-2-ene-1,4-diylbis(phosphonate)[24]
A mixture of 1,4-dibromo-2-butene (5.60 g, 26.2 mmol) and triethyl phosphite (14.0 P/PPRO ZDVKHDWHGWRÜ&DQGVWLUUHGDWWKLVWHPSHUDWXUHIRUKRXUV Subsequent distillation in vacuo yielded the target bisphosphonate (7.42 g, 86%) as a FRORXUOHVVOLTXLG ESÜ&DW PEDU >1H-NMR (CDCl 3, G, ppm) 5.61 (m, 2H, vinyl); 4.11 (m, 8H, OCH2); 2.61 (m, 4H, CH2P); 1.32 (t, 12H, CH3).31P-NMR (CDCl3, G, ppm), 27.4 (s)] L inker 2: 4,4’-(1,4-phenylenediethane-2,1-diyl)dipyridine[25] N Br Br + N 58% BuLi, THF N 2
To a solution of 4-picoline (1.40g, 15.0 mmol) in THF (100 mL) MeLi (1.6 M in hexane ,8.0 mL, 12.8 mmol) was dropwise added. The reaction was stirred for 15 minutes followed by refluxing for two hours. At room temperature a solution of D,D’-dibromo-p-xylene (1.32 g, 5.00 mmol) in THF (10 mL) was dropwise added. The reaction was refluxed overnight. After cooling to room temperature the solvents were evaporated and the residue was taken up in diethyl ether (100 mL). The ether layer was washed with water (20 mL) and brine (20 mL), dried (MgSO4) and filtered. Evaporation of the solvent and column chromatography (PE : EtOAc = 3:1) yielded 0.84 g (58%) of a white solid. [1H-NMR ((CD
3)2CO, G, ppm) 8.42 (d, 4H, J = 4.4 Hz,
136 C h a p te r 7 Linker 3: 4,4’-[1,4-phenylenedi(E)ethene-2,1-diyl]dipyridine[26] P(OEt)2 P(OEt)2 O Cl Cl (EtO)3P 1500C N N KOtBu / THF 20% O N CHO
Tetraethyl [1,4-phenylenebis(methylene)]bis(phosphonate) (300 mg, 0.79 mmol) was dissolved in THF (25 mL) and potassium tert-butoxide (270 mg, 2.41mmol) was added. The mixture immediately turned dark red and after 20 minutes pyridine-4-carboxaldehyde (200 µL, 2.13 mmol) was added dropwise and stirring was continued at room temperature overnight. The reaction mixture was diluted with EtOAc (40 mL), washed with water (20 mL) and brine (20 mL), dried (Na2SO4) and filtered. Evaporation of the solvents afforded a yellow solid. Trituration with MeOH (2x) and diethyl ether (2x) followed by drying afforded the pure target compound as a yellow solid (44 mg, 20%). [1H-NMR (D 2O/DCl, G, ppm) 8.27 (d, 4H, J = 5.1 Hz, py); 7.65 (d, 4H, J = 5.1 Hz, py); 7.18 (s, 4H, Ph); 6.97 (d, 2H, J = 16.1 Hz, =CH); 6.60 (d, 2H, J = 16.1 Hz, =CH). ESI-MS [M+H]+ = 285.1; [M 2+H] + = 143.2] Linker 4: 3,3’-[1,4-phenylenedi(E)ethene-2,1-diyl]dipyridine[26] P(OEt)2 P(OEt)2 O Cl Cl (EtO)3P 1500C N N KOtBu / THF 31% O N CHO
The compound was prepared as described above from tetraethyl [1,4-phenylenebis (methylene)]bis(phosphonate) (309 mg, 0.82mmol) and pyridine-3-carboxaldehyde (2.22 mmol) in 31% yield. [1H-NMR (D
2O/DCl, G, ppm) 8.67 (s, 2H, py); 8.54 (d,
2H, J = 8.0 Hz, py); 8.40 (d, 2H, J = 5.8 Hz, py); 7.87 (dd, 2H, J = 8.0, 5.1 Hz, py); 7.39 (s, 4H, Ph); 7.16 (d, 2H, J = 16.1 Hz, =CH); 6.96 (d, 2H, J = 16.1 Hz, =CH). ESI-MS [M+H]+ = 285.1; [M
2+H]
137 C h a p te r 7 Linker 5: 4,4’-(1,4-phenylenediethyne-2,1-diyl)dipyridine[27] N N N I I (Ph3P)2PdCl2/ CuBr Et3N, 60 -90oC 55% + 2
Under a dry nitrogen atmosphere, a mixture of 1,4-diiodobenzen (496 mg, 1.50 mmol), 4-ethynylpyridine hydrochloride (435 mg, 3.11 mmol), copper(I) bromide (10 mg, 0.070 mmol), bis(triphenylphosphine)-palladium(II) dichloride (24 mg, PPRO DQGWULHWK\ODPLQH P/ ZDVVWLUUHGDWÜ&IRUKRXUV$IWHU WKLVSHULRGWKHWHPSHUDWXUHZDVUDLVHGWR Ü&DWDUDWHRIÜ&SHUKRXU Triethylamine was evaporated and the residue taken up in DCM. The DCM layer was washed with a 10% Na2CO3 solution, dried (Na2SO4), filtered and concentrated. The residue was purified by column chromatography (DCM : MeOH = 98: 2) to afford the target compound as a colourless solid (160 mg, 38%). [1H-NMR (CDCl
3, G, ppm) 8.63 (d, 4H, J = 4.4 Hz, py); 7.56 (s, 4H, Ph); 7.38 (d, 4H, J = 4.4 Hz, py). ESI-MS [M+H]+ = 281.0]. Linker 6: 4-[(1E,3E)-6-pyridin-4-ylhexa-1,3,5-trien-1-yl]pyridine[28] P(OEt)2 (EtO)2P O O Cl Cl (EtO)3P 1500C N N KOtBu / THF 9% N CHO
138 C h a p te r 7
to yield 70 mg (9%) of a yellow solid. [1H-NMR MeOD, G, ppm) 8.44 (m, 4H,
py); 7.46 (m, 4H, py); 7.24-7.37 (m, 2H, =CH); 6.66-6.80 (m, 4H, =CH). ESI-MS [M+H]+ = 234.8].
Linkers 7 and 8 were prepared by Albers et al.[29;30] Linker 8 was initially synthesized
as an iodine salt (I--8). The presence of iodine in the protein solution, however,
caused reduction of Cu(II)-H117G azurin. Addition of a stoichiometric amount of AgNO3 to a solution of I--8 dissolved in N-methyl-pyrrolidone (NMP) resulted in
precipitation of AgI, which was removed from the solution by centrifugation, and the formation of a soluble and redox-inactive NO3- salt of linker 8. Stock solutions
(10 and 1 mM) of all linkers were prepared in NMP.
Protein sample preparation
H117G azurin was produced and isolated in the apo-form as described by Jeuken et al.[31;32] Shortly before use, the protein solution was incubated with 1.2 molar
equivalents of Cu(NO3)2 at RT until full reconstitution was achieved, indicated by a ratio of A420/A280 = 0.22. The excess of Cu(NO3)2 was removed on a PD10 gel filtration column that was eluted with 20 mM MES pH 6.0.
Extinction coefficients
Extinction coefficients for the individual linkers in NMP and in 20 mM MES, pH 6.0 were determined from the measured absorptions of solutions of known concentration. These concentrations were determined by carefully weighing out the dried powder on a high precision balance. The extinction coefficients for the absorption maxima between 340 and 420 nm for each of the H117G-linker complexes were obtained from the slope of a plot of absorbance against the concentration of linker at a known concentration of azurin. Since free H117G azurin has an absorption at 630 nm (H = 1.2 mM-1cm-1) and because the concentrations of the
139 C h a p te r 7
Ligand coordination to H117G azurin
All UV-Vis titration experiments were performed on solutions of 70-100 PM Cu(II)-H117G azurin in 20 mM MES, pH 6.0 at RT on a Perkin Elmer lambda 800 spectrophotometer. The reaction stoichiometry of coordination of the linkers to H117G azurin was determined by Job's method:[33] a series of solutions was
prepared in which the combined concentration of protein and linker was kept constant at 100 PM whilst the molar fraction of each was varied from 0 to 100 PM. After mixing of the reactants, the samples were incubated for several minutes at RT. Before measurement of the absorption spectra, all samples were centrifuged for 5 mins at 13,000 rpm to remove any precipitation. Plots of A630nm versus the molar fraction of linker were constructed from which the reaction stoichiometry was determined.
Binding of the linkers was analyzed by titration of Cu(II)-H117G azurin (A) with small volumes of the linkers (L) dissolved in NMP that were rapidly mixed into the protein solution to prevent precipitation. Coordination of either pyridine or linker 8 involves only a single binding event and the corresponding data were therefore fitted to a model as proposed by Jeuken et al. derived for binding of imidazole.[32] The data
for titration with bifunctional linkers were fitted numerically (Gepasi software[34])
against a model that describes linker coordination as a two-step process;
140 C h a p te r 7
No further correction for the protonation equilibria was therefore applied to Kd,1 and Kd,2.
Reduction of (Cu(II)-H117G azurin)2-3
The effects of -partial- reduction on the dimer complex of Cu(II)-H117G azurin with linker 3 (25 PM in 20 mM MES, pH 6.0) were probed by titration with sodium ascorbate. Reduction of both copper centres within the dimer (A2L) is expected to result in the release of the linker and the concomitant dissociation of the complex.[31]
Partial reduction of the dimer, however, can result in either one of two possibilities: 1) the reduced protein dissociates from the complex leaving behind the oxidized form mono-coordinated by the linker (ALO) or 2) a rapid electron self-exchange (e.s.e.) between the connected copper centres prevents dissociation and a stable mixed ‘OR’ dimer is formed (A2LOR). In the former case, the ALO complex is anticipated to exhibit an extinction coefficient at 385 nm (i.e. the absorption maximum of the linker when coordinated to H117G azurin) that is the average between that of the fully coordinated form (A2LOO) and of the linker free in aqueous solution (Lf). If a fraction x of the total amount of protein is reduced, given by A630nm relative to the fully oxidized sample, and provided that the protein and linker are present in a stoichiometric ratio of 2:1, the concentrations of each of the possible species correspond to:
[A2LOO] = (1-x)2* [Azurin] Eq. 1a
[A2LRR] = 0.5*[AR] = [Lf] = x2* [Azurin] Eq. 1b
[A2LO(R)] = (2x -2x2) * [Azurin] Eq. 1c
The species denoted A2LO(R) can represent either A2LOR or ALO depending on whether the semi-reduced complex dissociates. Since AR does not absorb at 385 nm it will not contribute to the total A385nm which can thus be expressed in terms of the components formed in the course of a reductive titration by:
141 C h a p te r 7 EPR spectroscopy
Samples for X-band EPR were prepared in 20 mM MES pH 6.0, 30% glycerol with a protein concentration of 0.7 mM Cu(II)-H117G azurin to which the linkers were titrated in a 1 : 0.55 ratio. X-band (9.48 GHz) cw-EPR spectra were recorded on an ELEXSYS E680 spectrometer (Bruker BioSpin, Rheinstetten) equipped with a TE102 resonator. Sample solutions were contained in quartz tubes of 3 mm inner diameter. The measurements were performed at 40 K using an Oxford ESR 900H cryostat. The microwave power supplied to the resonator was about 0.2 mW. A 100 kHz magnetic field modulation with a peak-to-peak amplitude of 0.5 mT was used. All samples contained small amounts of uncomplexed Cu(II)-H117G azurin that gave rise to additional Type-2 like signals in the EPR spectra. These signals were removed from the experimental spectra by subtraction of a spectrum of pure uncomplexed Cu(II)-H117G azurin such that upon visual inspection the Type-2 like signals could no longer be discerned. No absolute calibration of g-values was performed. The g- and hyperfine parameters were determined from simulations of the spectra by EasySpin-2.2.0 software assuming a Gaussian shape of the individual lines.[35] The linewidth was defined as the full width at
142 C h a p te r 7
Results
U V -V is characterization of linkers free in solution
When dissolved in NMP, the linkers 3, 5, 6, 7 and 8 are characterized by strong absorption maxima around 350-400 nm that arise from extensive electron delocalization [Figure 7.2, Table 7.1]. In aqueous solution the absorptions are shifted towards lower wavelengths and the absorptivities are significantly lowered. Both pyridine and linker 2 show no absorption in the visible region of the spectrum whereas linker 4 could not be solubilized in NMP.
Table 7.1: Absorption maxima and extinction coefficients for the various linkers dissolved in N MP, in 20 mM MES, pH 6.0 and complexed w ith Cu(II)-H117G azurin [Figure 7.2, Figure 7.4] Omax (nm) H (mM-1cm-1) Linker 1 2 3 5 6 7 8 NM P --- -- 358 (63 ± 1) 320 (60 ± 1) 358 (51 ± 0.5) 387 (42 ± 0.5) 402 (20 ± 0.5) aqueous (M ES) --- --332 (23 ± 0.5) 315 (26 ± 1) 374 (35 ± 0.1) 354 (15 ± 0.6) 407-418 (13.6 ± 0.1) H117G -L --- -- 385 (70 ± 1) 343 (35 ± 1) 397 (64 ± 2) 418 (47 ± 1) 422 (29 ± 1) H117G -L* 630 (3.1 ± 0.1) 630 (4.4 ± 0.1) 630 (5.8 ± 0.1) 629 (3.5 ± 0.1) 638 (5.9 ± 0.1) 636 (5.8 ± 0.1) 630 (5.4 ± 0.1) *for better comparison w ith W T and H117G-1 azurin, these extinction coefficients are
143 C h a p te r 7
144 C h a p te r 7
Linker coordination to H117G azurin
Analysis of the reaction of Cu(II)-H117G azurin with linkers 2, 3, 5 and 7 by Job’s method [Figure 7.3] shows a binding stoichiometry of 2:1, demonstrating the formation of linked protein dimers. Titration of Cu(II)-H117G azurin with linkers 2, 3, 5, 6 and 7 indicates strong coordination to the protein with Kd,1§PM and Kd,2§PM [Figure 7.4, Table 7.2]. For most linkers their poor solubility in aqueous solutions at concentrations where the linker is present in excess results in sample turbidity so that the samples could not be titrated much beyond the point where full binding is reached. The problem was most pronounced for 7 [Figure 7.4E] where a correction of the absorption data for Rayleigh scattering was required prior to data fitting.[36] In all cases binding of the linkers causes a shift towards higher
wavelengths of the 350-400 nm band and gives rise to a classic Type-1 absorption at 630 nm. With the exception of the H117G-5 complex, this latter absorption is characterized by extinction coefficients comparable to WT azurin (H628 = 5.7 mM-1
cm-1) [Table 7.1].
145 C h a p te r 7
In linker 8 one of the pyridine N-atoms is blocked by a methyl group which prevents it from coordinating to the protein and accordingly, titration of H117G azurin with 8 indicates a 1:1 stoichiometry. The extinction at 420 nm of the monomeric H117G-8 complex corresponds to the average extinction of the dimeric H117G-7 complex and the extinction of 8 at 420 nm when dissolved in 20 mM MES (H420,app=14 mM-1
cm-1), consistent with one free and one protein-coordinated pyridine moiety. Linker
4 was found unable to coordinate to the protein and immediately precipitated from the protein solution.
Table 7.2: Binding affinities for coordination to H117G azurin [Figure 7.4]
Ligand Kd,1(PM) Kd,2 (PM) 1 140 ± 3 (A630), 80 ± 9 (A420) --2 1.1 (A420),11 ± 2* (A 630) 3.4 (A420), 0.2 ± 0.1 (A630) 3 11 ± 2* 0.6 ± 0.2 5 11 ± 2* 0.2 ± 0.1 6 11 ± 2* 0.9 ± 0.2 7 11 ± 2* 0.2 ± 0.1 8 11 ± 2
--* the values for K
d,1 for binding of linkers 2, 3, 5, 6 and 7 were constrained during
the fitting procedure at 11 ± 2 PM.
14
147 C h a p te r 7
Formation of semi-reduced protein-wire dimers
The H117G-3 complex ‘averages’ the different structural, chemical and electronic characteristics of all the linkers described and was therefore chosen for further analysis. Dimers of Cu(II)-H117G-3 azurin were stepwise reduced by titration with sodium ascorbate. A630nm is a measure of the amount of oxidized protein present in solution, while A385nm is considered an indicator of the fraction of linker complexed with the protein. The inset in Figure 7.5A shows a non-linear dependency of A385 on the level of reduction. Comparison of the extinction coefficients of the H117G-7 and H11H117G-7G-8 complexes suggests that the intensity of the absorption band at 420 nm is directly related to the number of protein molecules coordinated to the linker. Assuming the same is true for linker 3 release of the reduced protein from the complex, which results in the linker becoming mono-coordinated, will lead to a linear dependency of A385 on the fraction of reduced azurin. On the other hand, if the complex does not dissociate and a stable A2LOR species is formed, the observed value for H385 is more likely to be close to that of the A2LOO complex. If HA2LOR= HA2LOO, A385 would follow the behaviour indicated with (–+–) in the inset of Figure 7.5A. Fitting of the data to Eq. 2 produces HA2LOR = 61.0 mM-1cm-1. The difference
148 C h a p te r 7 EPR
The cw-EPR spectra of the various H117G azurin dimers and the spectrum of H117G azurin with pyridine (H117G-1) are shown in Figure 7.6. The spectrum of H117G-1 shows the typical features of a Type-1 copper site. The principal values of the g-tensor and parallel component (Azz) of the copper hyperfine tensor are: gzz = 2.255, gxx = 2.041, gyy = 2.065 and Azz = 196 ± 5 MHz. A contribution of the hyperfine interaction appears in the perpendicular region of the spectrum (at gxx, gyy) but could not be sufficiently resolved for analysis. The spectrum of the H117G-2 dimer is very similar to that of H117G-1. Lineshape simulation results in the parameters gzz = 2.259, gxx = 2.041, gyy = 2.065 and Azz = 191 ± 5 MHz. The individual lines of the dimer are broadened relative to H117G-1.
Figure 7.5: A) UV-Vis absorption spectra of Cu(II)-H117G-3 azurin dimers titrated with sodium ascorbate (25 PM azurin in 20 mM MES, pH 6.0). Inset: A385nm as a function of the fraction of reduced azurin given by A630 relative to the initial, fully oxidized sample. Dashed lines show the theoretical values for A385nm expected when H
A2LOR = HA2LOO (+) or HA2LOR = (HA2LOO + HL)/2 (X). The solid line represents a
least squared fit against Eq. 2 yielding H
A2LOR = 61.0 mM
-1cm-1. B) Spectra of A 2LOO
149 C h a p te r 7
Table 7.3: Line position of gxx,yy relative to the centre of the spectrum and EPR simulation parameters Complex 'B1 + 'B2 (Gauss) J0 (MHz) VJ (MHz) E0 (o) V E( o) H117G-1 190 - - - -H117G-2 190 - - - -H117G-3 100 780 ± 50 290 60 ± 5 12 H117G-5 170 300 ± 20 80 30 ± 5 8 H117G-7 90 850 ± 50 400 60 ± 5 15
150 C h a p te r 7
To estimate the magnitude of the broadening the spectra were fit with different component linewidths, paying particular attention to the best resolved line in the spectra, namely the lowest field line. For H117G-1 the best fit was achieved at FW HH = 3.3 m T and for H117G-2 at 3.8 m T. The spectra of the dim ers with linkers 3, 5, and 7 differ m ore strongly from the spectrum of H117G-1. The signals corresponding to gzz and gxx, gyy are shifted by various am ounts com pared to H117G-1. The signal at g
zz is shifted towards higher fields and the splitting due to Azz is no
longer resolved. The signal at gxx, gyy is shifted towards lower fields and is broadened relative to H117G-1. The EPR spectra recorded of H117G-6 showed indications of heterogeneity in the sam ple and were found to be irreproducible.
Discussion
The H117G mutant of azurin has attracted particular notice due to its ability to bind exogenous ligands such as im idazole or pyridine as well as ligands that protrude from the protein m atrix to facilitate interaction between the copper centre and the outside environm ent. The use of bifunctional wires that induce dim erization allows the properties of the protein-wire constructs to be studied directly in solution without need for surface im m obilisation. To this end, a series of wires with varying bridge structures, on both ends term inated by pyridine groups, was constructed. All of these were found to be of sufficient length to allow simultaneous coordination of two azurin m olecules, while the inability of linker 4 to coordinate shows that the ligand is able to bind only when the bridge structure is attached at the para-position but not at the m eta-position of the pyridine ring.
U V -V is characterization of linkers and H 117G azurin dim ers
The linkers with conjugated bridge structures display intense absorption bands with a pronounced vibrational fine structure between 350 and 400 nm , the energies of which are related to the extent of conjugation and provide an indication of the expected coupling strength. Coordination to the copper centres induces a bathochrom ic shift, im plying an increase of electron delocalization. The observation that com plexes form ed with linkers 2, 3, 6, 7 and 8 exhibit S-S bands at ~630 nm with extinction
151 C h a p te r 7
coordination of free pyridine. Given that H117G azurin and linker 5 were shown to react with high affinity to form dimers, it is unlikely that this is due to only partial ligation of the available copper sites. M ore likely, coordination of linker 5 induces a small geometric change that weakens the Cu-Cys112 bond responsible for the S-S
transition connected to the blue colour.
At increasing concentrations of linker ([Linker]>[H117G azurin]) all samples experience substantial bleaching. This has previously been encountered at high concentrations of imidazole and was attributed to autoreduction of the copper centre.[32] Although accurate fitting of the titration data to the proposed two-step
binding model was complicated by the effects of sample turbidity and bleaching at high concentrations of ligand, the obtained dissociation constants Kd,1 and Kd,2 are all about an order of magnitude or more (0.2 - 11 PM ) smaller than that found for pyridine (140 PM ). Complexation of the hydrophobic wires most likely helps to reduce their solvent exposure and the linkers will consequently bind with high affinity. The data given in Table 7.2 indicate that Kd,2 < Kd,1 for all investigated complexes suggesting cooperative binding. However, these Kd,2values were obtained by imposing rigorous constraints on Kd,1 under the assumption that Kd,1 for all these complexes is similar to the Kd for binding of ligand 8. This does, however, not necessarily need to be the case and in order to obtain a more accurate description of the complexation kinetics further experiments are required. For binding of linker 2 to H117G azurin, fitting of the absorbances at 420 nm and at 630 nm produces significantly different Kd values, eventhough the data should correspond to the same process. Possibly an additional process, unaccounted for in the fitting model, also influences A630 but this can not be ascertained based on the data presented here.
EPR: electronic coupling of the complexes
152 C h a p te r 7
change of the FWHH for a Gaussian line from 3.3 mT to 3.8 mT corresponds to an increase of the spectral second moment by 64*10-8 T2. The interspin distance (r) can
be estimated using the equation :
r = 2.32 * ('BD2)-1/6 nm Eq. 3
where 'BD2 is the spectral second moment contribution due to dipole-dipole
interaction expressed in T2.[37] This equation yields r | 1.2 nm, which is somewhat
shorter than the Cu-to-Cu distance of 1.8 nm estimated for linker 2 by summation over the C-C and C-N bond lengths. This suggests that the larger width of the lines in H117G-2 is mainly due to the dipolar interaction, but the shorter distance (larger linewidth) found for the second moment analysis could point towards additional line broadening , e.g. due to exchange interaction.
153 C h a p te r 7
coupling, the fits are not unique and depend strongly on the model used as input. For a proper analysis of the spectra, information on the g-tensor orientation would be needed. Furthermore, experiments at other EPR fields and frequencies would give additional information on the magnitude of J. Such experiments are in progress right now.
Electron self-exchange in non-covalent dimers of H117G azurin
An important obstacle to the use of H117G azurin in bioelectrochemistry is the low ligand affinity of Cu(I)-H117G azurin. It was shown by Jeuken et al. that upon dissociation of the ligand from the Cu(I) centre the redox potential sharply increases, rendering the system redox-inactive under normal solution conditions. It was, however, demonstrated by fast scan protein-film voltammetry (FS-PFV) that reversible reduction of an H117G-imidazole complex can be achieved provided that the rate of reoxidation is faster than the rate of ligand dissociation.[31] It was
anticipated that in dimers where the copper centres are sufficiently coupled, electron self-exchange rates are attainable that exceed that of ligand dissociation so that redox-active dimers can be formed. A reductive titration of the complex formed by coordination of linker 3 to H117G azurin indeed suggests the existence of a stable semi-reduced state of the dimer, implying that the e.s.e. between the associated copper centres is more rapid than the –partial– dissociation of the complex. Although the latter rate was not determined for any of the complexes studied, an estimate can be made based on the dissociation kinetics for Cu(I)-H117G-imidazole azurin determined by Jeuken et al.[31] Using FS-PFV, a value for k
off was found between 500
and 2000 s-1suggesting that k
ese must be faster than this.
Polyethene and polyethenyl type wires have been extensively studied and electronic decay constants for ET along them have been determined ranging between 0.6 and 0.1 Å-1.[38;39] Over a distance of approximately 18 Å, such as the Cu-to-Cu distances
expected in these dimers, intramolecular tunnelling rates of up to 1010 s-1 should
therefore be considered feasible when assuming that:
154 C h a p te r 7
serve multiple purposes: they provide general examples of a more widely applicable strategy for direct ‘hotwiring’ of redox proteins using conductive wires whilst at the same time they may find more direct applications in biosensors based on catalytically active blue copper proteins such as copper containing nitrite reductase (NiR). Analogous cavity mutants of NiR, capable of binding a variety of ligands, have already been reported.[40] Albers et al. have previously shown that conductive
monolayers can be formed by self-insertion of thienoviologens such as linkers 7 and 8 into a layer of octadecylmercaptan.[30] Further steps towards surface-immobilized electrochemistry on the described wires and complexes are currently being pursued in our lab.
Conclusions
The electrical contacting between proteins and electrodes is one of the major challenges in the construction of biosensors. Based on replacement of the native H117 ligand of azurin with conductive wires, dimeric protein-wire complexes were constructed in which a relatively strong electronic coupling between the copper centres was identified by the effects of spin-exchange exerted on their EPR spectra. The coupling strength appears to be correlated with the extent of conjugation of the system as indicated by the energy of the absorption band of the wire. Results from a reductive titration of one such complex, monitored by UV-Vis spectroscopy, indicate the formation of a stable semi-reduced redox state consistent with fast intramolecular electron self-exchange. It was demonstrated that insertion of conductive wires directly into a redox centre represents a promising strategy for establishing efficient electrical contact between proteins and electrodes.
A cknow ledgements
Dr. H.J. Wijma and Dr. A.W.J.W. Tepper are kindly acknowledged for their help with the scatter correction of UV-Vis spectra and on usage of the Gepasi software for data fitting respectively.
References
[1.] Y. Lu, S. M. Berry, T. D. Pfister, Chemical Review s 2001,101 3047-3080. [2.] G. Gilardi, A. Fantuzzi, S. J. Sadeghi, Current O pinion in Structural Biology
155 C h a p te r 7
[3.] G. Gilardi, A. Fantuzzi, Trends in Biotechnology 2001, 19 468-476.
[4.] A. E. G. Cass, G. Davis, G. D. Francis, H. A. O. Hill, W. J. Aston, I. J. Higgins, E. V. Plotkin, L. D. L. Scott, A. P. F. Turner, Analytical Chemistry 1984, 56 667-671.
[5.] A. Heller, Accounts of Chemical Research 1990, 23 128-134.
[6.] E. Katz, V. Heleg-Shabtai, A. Bardea, I. Willner, H. K. Rau, W. Haehnel, Biosensors & Bioelectronics 1998, 13 741-756.
[7.] A. Riklin, E. Katz, I. Willner, A. Stocker, A. F. Buckmann, N ature 1995, 376 672-675.
[8.] I. Willner, V. Heleg-Shabtai, R. Blonder, E. Katz, G. L. Tao, Journal of the American Chemical Society 1996, 118 10321-10322.
[9.] A. Bardea, E. Katz, A. F. Buckmann, I. Willner, Journal of the American Chemical Society 1997, 119 9114-9119.
[10.] I. Hamachi, S. Shinkai, European Journal of Organic Chemistry 1999, 539-549.
[11.] E. Katz, A. Riklin, V. Heleg-Shabtai, I. Willner, A. F. Buckmann, Analytica Chimica Acta 1999, 385 45-58.
[12.] Y. Xiao, F. Patolsky, E. Katz, J. F. Hainfeld, I. Willner, Science 2003, 299 1877-1881.
[13.] M. Zayats, E. Katz, I. Willner, Journal of the American Chemical Society 2002, 124 14724-14735.
[14.] D. Barrick, Current Opinion in Biotechnology 1995, 6 411-418.
[15.] T. den Blaauwen, G. W. Canters, Journal of the American Chemical Society 1993, 115 1121-1129.
[16.] T. den Blaauwen, C. W. G. Hoitink, G. W. Canters, J. Han, T. M. Loehr, J. Sanders-Loehr, Biochemistry 1993, 32 12455-12464.
[17.] T. den Blaauwen, M. van de Kamp, G. W. Canters, Journal of the American Chemical Society 1991, 113 5050-5052.
[18.] G. van Pouderoyen, T. den Blaauwen, J. Reedijk, G. W. Canters, Biochemistry 1996, 35 13205-13211.
[19.] G. van Pouderoyen, T. den Blaauwen, J. Reedijk, G. W. Canters, Biochemistry 1998, 37 7656.
842-156 C h a p te r 7 851.
[21.] S. Creager, C. J. Yu, C. Bamdad, S. O’Connor, T. MacLean, E. Lam, Y. Chong, G. T. Olsen, J. Y. Luo, M. Gozin, J. F. Kayyem, Journal of the American Chemical Society 1999, 121 1059-1064.
[22.] E. A. Weiss, M. J. Ahrens, L. E. Sinks, A. V. Gusev, M. A. Ratner, M. R. Wasielewski, Journal of the American Chemical Society 2004, 126 5577-5584. [23.] F. Giacalone, J. L. Segura, N. Martin, J. Ramey, D. M. Guldi, Chemistry-A
European Journal 2005, 11 4819-4834.
[24.] J. M. Kauffman, G. Moyna, Journal of Organic Chemistry 2003, 68 839-853.
[25.] V. W. W. Yam, V. C. Y. Lan, L. X. Wu, Journal of the Chemical
Society-Dalton Transactions 1998, 1461-1468.
[26.] H. Detert, L. Sadovshi, E. Sugiono, Journal of Physical Organic
Chemistry 2004, 17 1046-1050.
[27.] N. R. Champness, A. N. Khlobystov, A. G. Majuga, M. Schroder,
N. V. Zyk, Tetrahedron Letters 1999, 40 5413-5416.
[28.] S. Woitellier, J. P. Launay, C. W. Spangler, Inorganic Chemistry 1989,
28 758-762.
[29.] W. M. Albers, G. W. Canters, J. Reedijk, Tetrahedron 1995, 51
3895-3904.
[30.] W. M. Albers, J. O. Lekkala, L. Jeuken, G. W. Canters, A. P. F.
Turner, Bioelectrochemistry and Bioenergetics 1997, 42 25-33.
[31.] L. J. C. Jeuken, P. van Vliet, M. P. Verbeet, R. Camba, J. P. McEvoy,
F. K. Armstrong, G. W. Canters, Journal of the American Chemical
Society 2000, 122 12186-12194.
[32.] L. J. C. Jeuken, M. Ubbink, J. H. Bitter, P. van Vliet, W.
Meyer-Klaucke, G. W. Canters, Journal of M olecular Biology 2000, 299
737-755.
[33.] M. M. Jones, Elementary Coordination Chemistry, Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1964.
[34.] P. Mendes, Computer Applications in the Biosciences 1993, 9 563-571.
[35.] S. Stoll, A. Schweiger, Journal of M agnetic Resonance 2006, 178
42-55.
157 C h a p te r 7
[37.] H. J. Steinhoff, Frontiers in Bioscience 2002, 7 C97-C110.
[38.] W. B. Davis, W. A. Svec, M. A. Ratner, M. R. Wasielewski, Nature
1998, 396 60-63.
[39.] C. Joachim, J. P. Launay, S. Woitellier, Chemical Physics 1990, 147
131-141.
15
