Chapter 2
Electron Transfer in M121E Azurin

Measuring electron transfer rates in proteins

Several methods have been used to obtain electron transfer rates within protein media. One may study the natural systems plants and animals use to obtain energy, for example the oxidase systems of the mitochondria or the photosynthetic reaction center. Studying most oxidase systems is complicated by the fact that electron transfer is coupled to proton transfer, adding another layer of complexity. Several bacterial photosynthetic reaction centers have been widely studied. They are large but well defined; X-ray structures are available for Rhodopseudomonas viridis (Allen et al. 1987; Yeates et al. 1987) and Rhodobacter sphaeroides (Deisenhofer et al. 1984). Electron transfers between various of the multiple chromophores can be followed spectroscopically and yield rates for ET over various distances and with different driving forces (Boxer 1990).

However, one is limited in these natural systems to the distances, driving forces, and intervening media provided by the particular reaction center. The Gray group, among others, have chosen to study electron transfer reactions in model systems consisting of a metalloprotein that has been covalently modified to add a second metal center at various places on the protein surface (Winkler and Gray 1992). The metal modification used in this study is a ruthenium bisbipyridyl imidazole (Ru(bpy)2Im-) bound to surface histidines introduced on the surface of azurin through site-directed mutagenesis. Excitation with 480 nm laser light creates a long-lived Ru(bpy)2ImHis excited state (66 ns for the Ru(bpy)2Im2 model compound (Wuttke 1994), 100 ns for Ru(bpy)2Im-labeled azurin (Di Bilio et al. 1997; Kiser 1997)) which is a good reductant. The reducing potential of the Ru(bpy)2Im label has been exploited for the direct reduction of the Cu2+ center of azurin, in the photoinduced ET scheme, and, for the indirect oxidation of a Cu1+ center after the reduction of an exogenous quencher in the flash/quench scheme.

Photoinduced electron transfer

In the photoinduced electron transfer scheme (see figure 1), the Ru(bpy)2ImHis label is excited at 480 nm, creating an excited electron on one of the bipyridyl ligands. This can decay by phosphorescence (emitting in a broad band around 670 nm), energy transfer, and non-radiative decay (collectively kd), and by electron transfer (kET). When the Ru label is attached at points on the surface of azurin close to the copper center, the excited electron may be transferred to the half filled Cu2+ HOMO. This reduction causes a bleach in the characteristic azurin ~600 nm LMCT band. The concomitant creation of the Ru3+ center can most easily be monitored as a bleach in the Ru2+/Ru3+ couple at 430 nm, the isobestic point for the Ru(bpy)2Im22+/2+*. (See figure 2 (Sigfridsson et al. 1996).) The Cu1+/Ru3+ couple that is created is thermodynamically unstable and returns to the Cu2+/Ru2+ ground state through a second electron transfer reaction (kBET). Since we are mainly interested in metal to metal, ground state electron transfers, this back electron transfer is the reaction of interest. This is fortunate, since the forward, excited state electron transfer reaction is on time scales too fast to be resolved from the 25 ns laser pulse used to create the initial Ru2+ excited state. Additional strong absorbance changes in the Ru2+/Ru3+ difference spectrum occur at 310 and 500 nm. At these wavelengths, there are also strong absorbance changes due to the Ru2+* excited state. Rates for both the excited state decay and the back electron transfer can be identified using biexponential fits of the data because they occur on sufficiently different time scales as to be distinguishable.

Flash/quench methodology

The photoinduced electron transfer scheme only works in systems where the initial electron transfer from the Ru label competes favorably with the other processes by which the excited state electron may relax. In general, this requires an acceptor with a fairly high reduction potential located close to the point of attachment of the label. To study electron transfer at slower rates (longer distances), the flash/quench scheme was developed (Chang et al. 1991). (See figure 3.) With this methodology, one starts with reduced azurin. The excited state of the Ru2+ label is rapidly quenched using an exogenously added [Ru(NH3)6]3+ quencher. The Ru3+ label can then be reduced intramolecularly by electron transfer from a reduced Cu1+ center in protein. Again this ground state electron transfer can be monitored at wavelengths characteristic of the Cu2+/1+ and Ru3+/2+ couples (600, 430, 310, and 500 nm). On longer time scales (0.5 ms), the reduced [Ru(NH3)6]2+ quencher rereduces the azurin Cu center. Because both the photoinduced reaction and the flash/quench reaction using [Ru(NH3)6]3+ are reversible, the sample can be excited repeatedly and the resulting absorption transients averaged to increase the signal-to-noise ratio.

Experimental design

In this project I have attempted to alter the electron transfer properties of the blue copper protein, azurin, by altering one of its ligand residues. Using site-directed mutagenesis, M121E mutants of azurin containing a single surface-exposed histidine at either position 83 or 122 were constructed. They were labeled with Ru(bpy)2Im and, using the photoinduced and flash/quench schemes described above, ET rates from Cu1+ to Ru3+ were measured at pH 4.3 and 8.1. These rates were compared to rates of ET to labels at the same places on the wild-type protein (Langen 1995). In chapter 3 the observed alterations in ET rates are discussed within the framework of semiclassical Marcus theory and attempts are made to correlate changes in the ET function of the M121E mutant to other spectroscopic and structural changes at the Cu site.

Material and methods

General

Unless otherwise stated, all chemicals were reagent grade. Restriction enzymes and T4 DNA ligase were purchased from a variety of suppliers (New England Biolabs, Beverly MA; Boehringer Mannheim, Indianapolis, IN) and used according to the manufacturer's instructions. PD10 columns, disposable columns prepacked with Sephadex G25, were purchased from Pharmacia Biotech (Uppsala, Sweden) and used for most buffer exchanges. Protein concentration was done by ultrafiltration using either an Amicon YM10 membrane or a Centricon10 spin concentrator (Amicon, Beverly, MA). Oligonucleotide synthesis was done at the Caltech polymer synthesis facility. Sequencing was originally done manually using a Sequenase Version 2.0 kit from US Biochemical and, later in the project, by the Caltech DNA sequencing facility. Protein purifications used a Pharmacia FPLC with a MonoQ 10/10 column and protein elution was monitored by absorption at 280 nm. Absorption spectra were taken using a Hewlett Packard 8452A diode array UV/Vis spectrophotometer. The pET expression system (vectors and the E. coli expression strain BL21(DE3)) were purchased from Novagen (Madison, WI).

Inorganic reagents

Ruthenium(II) bis-bipyridine carbonate was prepared according to the method of Johnson and coworkers (Johnson et al. 1978). 1.01grams of dichloro bis(2, 2' bipyridine)ruthenium(II) dihydrate (Strem Chemicals, Inc., Newburyport, MA) was refluxed for 25 minutes in 100 ml water under a stream of nitrogen gas. To this hot solution was added 3.3 g sodium carbonate (Na2CO3); the solution turned a purple/red almost immediately. This solution was refluxed for a further two hours, still flushing the flask with nitrogen gas. The solution was allowed to cool, then filtered through Whatman No. 5 filter paper to collect the crystals. UV/Vis absorption spectra show the expected bis(2, 2' bipyridine)ruthenium(II)carbonate product.

Hexamine ruthenium(III) trichloride, used as an oxidative quencher in laser experiments, was purchased from Strem Chemicals, Inc. Before use, it was recrystalized by first dissolving 1 g in 15 ml water, then precipitating it by addition of excess acetone (75 ml). The solid filtered out of this water/acetone mixture was then redissolved in 20-25 ml 1 M HCl, filtered, and the solvent and acid were removed by rotary evaporation. The spectrum of the purified hexamine ruthenium(III) trichloride shows a strong absorbance at 276 nm and a greatly reduced shoulder at 322 nm.

Mutant construction

Our expression system for wild type azurin is the synthetic gene for Pseudomonas aeruginosa azurin constructed by Dr. Thomas K.-Y. Chang (Chang et al. 1991) cloned into a T7 promoter expression vector (pET, Novagen, Madison, WI) (Germanas et al. 1993). The construct I currently use is cloned into pET9a and is a gift from Dr. Jy-Ye Luo. The M121E mutant was constructed using the Kunkel method (Kunkel et al. 1991) of mutagenesis (Muta-gene kit from Bio-Rad) on a single stranded pTZ18U/azurin template with subsequent subcloning of the mutants into the pET9a expression vector; M121E azurin in pET9a was a gift from Dr. T. Jack Mizogouchi. The M121E/K122H/H83Q mutant was constructed using the Kunkel mutagenesis method (bottom strand mutagenic oligo: 5' CAG AGT CAG GGT ACC GTG CTC CAG TGC GGA GTG on a single-stranded template, pTZ18U/azurin containing the H83Q mutation, a kind gift from Dr. Ralf Langen) and was sub-cloned into pET9a. The M121E/T124H/H83Q gene was prepared by ligating a BamHI/KpnI fragment containing the M121E mutation into a pET3a construct containing the H83Q/T124H azurin described by Dr. Ralf Langen (Langen 1995). See appendix B for maps and sequences of these constructs.

Protein production

The expression vector is transformed into chemically competent BL21(DE3) E. coli (Novagen). Single colonies are used to inoculate starter cultures of LB (1% tryptone, 0.5% yeast extract, 0.5% NaCl) supplemented with the appropriate antibiotic (50 ┤g/ml ampicillin for pET3a constructs or 50 ┤g/ml kanamycin for pET9a constructs). These are grown overnight shaking at 37ŚC and used to inoculate three 3 liter flasks of LB containing antibiotic. Cultures are grown to an OD600 of 0.6-1.0. Protein production is initiated by addition of IPTG to a final concentration of 0.4 mM. After 4-8 hours at 37ŚC, the cells are harvested by centrifugation for 10 minutes, 4,000 x g. The cell paste is resuspended in 1/10th volume of a high osmolarity solution (20% sucrose, 30 mM Tris pH 8.0, 1 mM EDTA) and shaken for 10-30 minutes. The cells were repelleted (20 minutes, 8,000 x g) then resuspended in 1/10th volume distilled water and left shaking at 4ŚC for 4-12 hours. The cells are then pelleted (20 minutes, 8,000 x g) and the periplasmic extrudate is decanted from the cell debris, taking care to minimize the amount of lysed cell material carried into the next step. The protein solution is made acidic by the addition of sodium acetate to 20-100 mM and the pH adjusted to 4.5; this precipitates most of the other periplasmic proteins as well as the DNA from lysed cells, leaving mainly azurin in solution. After sitting at room temperature for an hour or more, the precipitate is removed by centrifugation (30 minutes 8,000 x g). The solution is brought to 10 mM CuSO4 and left at room temperature for up to a week to allow the protein to incorporate Cu. Then the protein solution is concentrated by ultrafiltration with an Amicon YM10 membrane. The protein is generally stored at 4ŚC in sodium acetate buffer, pH 4.3-4.5, with 5-10 mM CuSO4. Upon storage at concentrations of 1-6 mg/ml, protein, mainly azurin, tends to precipitate.

Protein purification

For the labeling reactions, the proteins were used without further purification. For studies requiring pure holo-azurin, this crude protein preparation is purified by FPLC on a MonoQ anion exchange column using 20 mM diethanolamine, pH 8.8, and eluting with a salt gradient of 0-30 mM NaCl; holo-azurin elutes at ~20 mM NaCl. The apo-azurin can be made by exhaustive washing (concentration and dilution using a Centricon10) with buffer consisting of 100 mM thiourea, 10 mM EDTA, 100 mM NaOAc pH 4.5. The apo-protein is purified on a MonoS cation exchange column, loading with 25 mM NaOAc, pH 4.5, 1 mM EDTA and eluting with a gradient of 300 mM NaOAc, pH 4.5, 1 mM EDTA; apoM121EAz elutes at 8-9% B. Apo-azurin may also be purified with a MonoQ anion exchange column, loading with 20 mM diethanolamine, pH 8.8, and eluting in the same buffer containing 200 mM NaCl; apo-azurin elutes at ~17% B.

Ruthenation and purification - M121E/K122H/H83Q azurin

Approximately 100 mg of crude azurin (quantitated using an e=3000 M-1 cm-1 for the 610 nm band) is washed by repeated concentration and dilution to remove the excess CuSO4 in which the protein is usually stored. Then, again by repeated concentration and dilution, the protein is exchanged into fresh 300 mM NaHCO3, pH 7.3. Ru(bpy)2CO3.4H2O is added to a final azurin:Ru ratio of 20mg:1mg (3:4 azurin:Ru molar ratio). The total volume of the reaction is adjusted to give final protein concentration of approximately 3 mg/ml and the reaction is left at room temp in the dark overnight (generally 12-16 hours). To check extent of the reaction, a small amount is exchanged into 25 mM NaOAc pH 4.5, using a PD10 column; the ruthenated azurin migrates as a forest green band, while the excess inorganic Ru migrates as a slower-moving red band. Initially there is a broad symmetric absorption near 465 nm which gradually shifts toward 486 nm, becoming sharper and less symmetric, leaning toward the red. The reaction is stopped when the inorganic Ru is consumed or when the Ru absorption band has shifted to 486 nm. The ratio of the extinction coefficients of the Ru2+ 486 nm band and M121E azurin Cu2+ band near 600 nm is ~ 3.

The 'mono-aquo' species azurinHis122Ru(bpy)2(H2O) is concentrated and exchanged into 300 mM imidazole, pH 8.0. The imidazole reaction is again monitored by exchanging small aliquots into 25 mM NaOAc pH 4.5 using a PD10 column. Again one generally sees a trailing band of inorganic Ru on the PD10 column, indicating that some non-specifically bound Ru dissociates from the protein. As the reaction progresses the visible spectrum shows a shoulder growing in near 440 nm. One also generally sees an increase in the Ru:Az 486/600 ratio, perhaps indicating Cu2+ loss due to imidazole chelation. After 3-6 days, there are generally no further changes in the spectrum.

The ruthenium modified azurin is purified by two successive FPLC separations using a Mono Q 10/10 column. The protein is loaded with 20 mM ethanolamine, pH 9.2, and eluted with 20 mM ethanolamine, pH 9.2, containing 200 mM NaCl. The first separation uses a linear gradient from 0-30% buffer B. The constellation of peaks eluting between 0 and 10% B are thought to be multiply ruthenated azurins because the ratios of their 486 to 600 nm bands are greater than 3. The next fraction, the largest of the peaks (at ~14% B), is Ru(bpy)2ImHis122Az. This is followed by the unmodified azurin (which still contains Cu2+) and then another peak with a spectrum nearly identical to the main peak (as yet unidentified). The pooled main peak is re-chromatographed using the same column and buffers but with a much shallower gradient, 10-12% B. Typical elution profiles are shown in figure 4. Pooled main peak fractions are stored at 4ŚC in 25 mM NaOAc, pH 4.5, with or without addition of excess CuSO4 (~5 mM). Typical yields of Ru(bpy)2ImHis122Az are around 60%.

Ruthenation and purification - M121E(H83) azurin

The protein is exchanged into 300 mM NaHCO3 buffer, pH 7.40, using a Centricon10. The extinction coefficient of a stock solution of Ru(bpy)2CO3 .4H2O in 300 mM NaHCO3 (e510 = 9,200 M-1cm-1) is used to calculate the amount needed to achieve a 1:1 or 1:2 molar ratio of azurin:Ru. The final protein concentration is adjusted to 3 mg/ml and the reaction is allowed to proceed for 12 hours at room temperature in the dark. The protein is exchanged into 300 mM imidazole, pH 8.0, and left at room temperature in the dark for 1-3 days.

Two rounds of FPLC purification are performed using a MonoQ 10/10 column, loading in a buffer of 20 mM ethanolamine, pH 9.4, and eluting with a buffer containing 20 mM ethanolamine, pH 9.4, and 200 mM NaCl (see figure 5). For the initial purification, a gradient of 0-20% buffer B was used. M121EH83Ru(bpy)2Im elutes as the main peak at ~5% B. For the second round of purification, the sample is again loaded onto the column in 20 mM ethanolamine, pH 9.4, and washed with buffer A for 1/2 column volume before initiating a gradient of 0-10% buffer B. Sodium borate buffers (20 mM, pH 9.6) have also been used to purify M121EH83Ru(bpy)2Im. At this pH, the protein interacts more strongly with the column and elutes with ~30 mM NaCl.

Laser sample preparation

For photoinduced electron transfer experiments, purified protein was exchanged into either 100 mM sodium phosphate, pH 8.1, or 25 mM sodium acetate, pH 4.3, using a PD10 column. 1-2 ml samples with protein concentrations ranging between 30 and 115 mM were placed in a 1 cm path-length laser cuvette and degassed by repeated cycles of vacuum pumping and flushing with N2 gas. Luminescence and absorption transients were taken using the Beckman Institute Laser Research Center (BILRC) nanosecond dye laser, exciting with 25 nanosecond, 1.5 mJ pulses of 480 nm light. Luminescence decay was monitored at 670 nm. Transient species were monitored at 600 (Cu1+/2+), 430 (Ru3+/2+), 500 and 310 (Ru2+/3+ and Ru2+*/2+) nm. The BILRC laser and transient absorption detection apparatus have been described previously (Wuttke 1994).

For flash/quench laser experiments, initial data were taken using a scheme allowing in situ reduction of oxidized azurin. Laser samples were prepared using the photoinduced methodology above except that during the sample preparation, 1/10th volume of 60 mM [Ru(NH3)6]3+ in the appropriate buffer was placed in the other side of a two chambered laser cuvette - along with enough sodium ascorbate to be in 1-3 fold molar excess over the protein when the contents of the two chambers were mixed together. The sample was degassed and photoinduced laser data was taken. A UV/Vis spectrum of the sample verified that the photoinduced laser experiment did not degrade the labeled protein sample. The ascorbate and quencher were then mixed into the protein sample and reduction of the Cu center was verified by UV/Vis spectroscopy. Luminescence decay and transient absorption data were then taken in the same manner as the photoinduced data was taken. In samples containing only a 1:1 molar ratio of ascorbate to protein (which is a 2 fold excess of electrons), the spectra indicated that the initial protein reduction was not complete. In the electron transfer data taken on in situ reduced samples, absorption transients did not return to baseline, even on time scales 10 times longer that than those required for the decay of the initial signal. In addition, the rate of the slower reoxidation of Ru2+ depended on the ascorbate concentration - higher ascorbate leading to faster bimolecular reoxidation.

To remove the effects of ascorbate, in later experiments purified protein was reduced with a large excess of sodium dithionite then immediately loaded on a PD10 column. The column served both to remove the excess dithionite and to exchange the protein into the desired phosphate or acetate buffer. Sample was loaded into the cuvette side of a 2 chambered laser cuvette. One volume of 12 mM [Ru(NH3)6]3+ in the appropriate buffer (or 1/10th volume of 60 mM [Ru(NH3)6]3+) was placed in the other side and the sample degassed. Luminescence decay measurements were taken on the reduced sample alone then the quencher was mixed in and luminescence decay and transient absorption data were taken. Initially samples were prepared on the lab bench and degassed with normal N2 gas. The M121E Cu1+ center appears to be very susceptible to reoxidation (see below) so in the final experiments for this thesis, flash/quench samples were prepared in a glove box under a nitrogen atmosphere and degassed using repeated cycles of vacuum pumping and flushing with N2 gas passed over a manganese-based oxygen-scrubbing catalyst to remove any residual impurities.

Results and discussion

Oxygen lability of reduced M121E azurin

In flash/quench experiments using samples prepared on the lab bench, absorption signals did not return to baseline, even on time scales 10x the rate of the initial decay (see figure 7f). This was believed to indicate heterogeneity in the sample, either lack of copper occupancy (with possible Zn contamination of the sample (Nar et al. 1992)) or reoxidation of the Cu center, which prevents electron transfer to the quenched ruthenium label. At the concentrations used, the [Ru(NH3)6]3+ quencher will oxidize less than 1% of the reduced protein. Experiments described below indicated that the M121E mutant copper site was far more susceptible to reoxidation in air than the wild type center.

UV/Vis spectra taken after shooting WT azurin did not show signs of steady-state reoxidation of the copper center during the course of flash/quench laser experiments. However, spectra taken after flash/quench experiments with the M121E mutant showed some signs of reoxidation of the Cu site. The 600 nm Cu2+ absorption seen at low pH increased with time even after shooting ceased. This raised the possibility that the absorption transients not returning to baseline was due to reoxidation of the Cu1+ center but in amounts that were hard to detect by steady state spectroscopy. The extinction coefficient of the 610 nm Cu2+ band is about 3000 M-1cm-1, while the 486 Ru2+ band, which has a considerable tail to lower wavelengths, has an extinction coefficient of 9,200 M-1cm-1 (Regan et al. 1995). In high pH experiments, no immediate information is available on the reoxidation of the sample because at pH 8.1 the copper peaks at 413 and 570 nm are obscured by the tails of the much larger Ru(bpy)2ImHis absorption.

Steady state reoxidation rates were studied using unlabeled WT and labeled and unlabeled M121E azurin. Samples were prepared by reduction with excess dithionite which was then removed using a PD10 column. When the experiment was done on the lab bench, extensive reoxidation, up to 61% of the M121E sample, was seen on the time that it takes to run the PD10 column, transfer the sample to the laser cuvette, and degas (10-15 minutes). Minimal reoxidation of the WT protein (14%) was seen on this time scale. Once degassed, the M121E does not reoxidize further; data for an experiment using M121E/H83Q/K122H Ru(bpy)2(Im) is shown in figure 2.6. The reoxidation of the M121E Cu1+ center while on the PD10 column indicates that, while the reduction potential of the M121E mutant at pH 4 is substantially the same as that of the WT protein (0.37 vs. 0.35 mV (Pascher et al. 1993)), the center is far more kinetically labile and so is susceptible to attack by atmospheric O2. Preparation of laser samples in a glove box with a N2 atmosphere solved the problem of long-lived Ru2+(bpy)2ImHis signals and the O2 lability of the M121E Cu2+ center was not explored further.

Electron transfer rates

The first step in the photoinduced electron transfer scheme is injection of the excited electron from the Ru2+*(bpy)2Im label into the oxidized blue copper site. The estimated driving force for this reaction is -1.39 eV at pH 4.0 (see chapter 3). With this relatively strong driving force, electron transfer is able to compete effectively with other relaxation processes when the electron acceptor is close to the label. Thus for labels placed at position 122 (metal to metal distance of 15.9╩) and at the native His83 (metal to metal distance of 16.9╩ (Faham et al. 1998)), considerable photoinduced electron injection occurs. The initial injection rate in these experiments was not measured but it could be determined from the change in emission lifetime between the Zn2+ and Cu2+ proteins if the quantum yield for the reaction were measured. The absorbance transients that can be monitored on the time scale of our instrument are for the second step of the photoinduced electron transfer - the back electron transfer from the transiently reduced Cu center to the Ru3+ label.

Photoinduced electron transfer experiments performed at low pH (25 mM NaOAc, pH 4.3) show good absorbance transients when the label was placed at position 122 and 83. At high pH (pH 8.1), no rate associated with electron transfer can be seen in photoinduced experiments with either of the labeled proteins. Absorption transients at all observed wavelengths for both the H83 and H122 samples decay with the same rate, 2 x107 s-1. The driving force for electron injection is lower at high pH but it is still considerably exothermic, -1.20 eV. The major difference between the high and low pH experiments is the major CT absorption band is considerably blue shifted at high pH. This has a major impact on the ability of energy transfer to compete with electron transfer and fluorescence as a means of relaxing the Ru2+ excited state. Energy transfer is promoted by close contact and good spectral overlap between the donor and acceptor states. In the oxidized protein at low pH, the absorption spectra of the Cu center and Ru(bpy)2Im label overlap somewhat (see figure 8). At high pH, this overlap is further enhanced by the shift of the azurin LMCT band from 610 to 570 nm.

While we do not have enough information to calculate the individual contributions of energy transfer, forward electron transfer, radiative and non-radiative decay, some idea of the magnitude of the energy transfer reaction can be obtained by comparing the emission rates of the Ru2+* in different experiments. The observed emissions rate is the sum of all processes that lead to the decay of the Ru(bpy)2ImHis excited state (Connors 1990). When the photoinduced experiment is done using the reduced Cu1+ azurin, the protein's metal center is d10 so there is no acceptor state for the excited electron. In addition the site is colorless so there is no spectral overlap to facilitate energy transfer. The decay of the Ru2+ excited state (t = 111(7) or 105(9) ns at positions 83 and 122 respectively) is thus only due to fluorescence and non-radiative relaxation processes. The excited-state lifetimes of the oxidized proteins are all much shorter than this. At high pH, where no photoinduced electron transfer could be seen, the Ru2+* lifetimes are 76(6) and 50(7) ns for positions 83 and 122 respectively. At low pH, where the Ru2+* excited state is quenched by both energy and electron transfer, the lifetimes are shorter still, t = 65(10) and 37(11) ns for positions 83 and 122.

Electron transfer rates for each protein at high and low pH are given in the table in figure 9. Confirmatory photoinduced rates are included where available. All of the electron transfer rates from the M121E Cu center are much slower than their WT counterparts. In addition, contrary to expectations we do not see a consistent increase in rate with increasing pH; ET to position 83 is faster at high pH but the 122 rate decreases with increasing pH. The interpretation of these rates will be taken up in chapter 3. In all cases, the rates observed were independent of protein concentration.

Figure 2.1 Photoinduced electron transfer scheme.

Figure 2.2 Difference spectra for the Ru2+*-Ru2+ excited state and the Ru3+-Ru2+ ground state couples of the model complex Ru(bpy)2Im2. The difference spectra were taken by Morten Bjerrum and described in (Sigfridsson et al. 1996).

Figure 2.3 Flash/quench electron transfer scheme.

Need to scan in data

Figure 2.4 FPLC purification of M121E/K122H Ru(bpy)2Im azurin.Mono Q 10/10 Buffer A: 20 mM ethanolamine, pH 9.2 Buffer B: 20 mM ethanolamine, pH 9.2 with 200 mM NaCl. (a) Early peaks are multiply ruthenated protein. The main peak, eluting at 14% B, is the desired M121E/K122H Ru(bpy)2Im azurin. The peak at 20% B is ruthenated but lacks the type I Cu center. And the peak at 23% B is unmodified azurin. (b) The main peak from (a) rerun with a shallower gradient.

Need to scan in data

Figure 2.5 FPLC purification of M121E/H83 Ru(bpy)2Im azurin.Mono Q 10/10 Buffer A: 20 mM ethanolamine, pH 9.4 Buffer B: 20 mM ethanolamine, pH 9.4 with 200 mM NaCl. (a) The early peak is multiply ruthenated protein. The main peak, eluting at 6% B, is the desired M121E/H83 Ru(bpy)2Im azurin. (b) The main peak from (a) rerun with a shallower gradient; multiply ruthenated protein elutes early, at 0% B.

Figure 2.6 Reoxidation experiment using with Ru(bpy)2Im-labeled H83Q/M121E/K122H azurin. The first data point gives the A486/A600 ratio of the protein sample before reduction. Time zero is the reduced sample before it is introduced into the laser cuvette. The first time point for the sample with quencher (6 mM [Ru(NH3)6]) is immediately after degassing and mixing with the quencher. The final data point on each curve is the sample after removal from the laser cuvette and reoxidation using excess K3Fe(CN)6.

Need to scan in data

Figure 2.7 Laser data: (a-c) Photoinduced ET monitored at 430 nm. At pH 4.3, we can see the Ru3+ signal disappear as the electron is transferred back out of the transiently reduced Cu center ((a) M121E/H83Ru(bpy)2Im (c) M121E/H83Q/K122HRu(bpy)2Im. At pH 8.0, although plenty of Ru2+* excited state is seen at other wavelengths, no Ru3+ is made ((b) M121E/H83Ru(bpy)2Im).

(d-f) ET measured using the flash/quench technique with 6 mM [Ru(NH3)6] quencher. Data taken at pH 4.3 with reduced M121E/H83Ru(bpy)2Im show that monitoring at characteristic Cu (590 nm (d)) and Ru wavelengths (430 nm (e)) give the same rate. (f) Data taken with samples prepared on the bench top show evidence of reoxidation. Data taken at high pH on reduced M121E/H83Q/K122HRu(bpy)2Im show rapid electron transfer but only about 1/3 of the sample is still reduced and so able to undergo intramolecular ET to the Ru3+ label.

Figure 2.8 Overlaid absorption spectra of the Ru(bpy)2Im label and M121E azurin at pH 4.5 and 8.0 showing the shift of the Cu LMCT up to wavelengths where the Ru label absorbs strongly.

  Wild Type M121E
Acceptor site pH 7.0 pH 4.3 pH 8.1
H83 FQ 1.2(1) x 106 s-1 (a) 3.9(2) x 105 s-1 5.2(6) x 105 s-1
H83 PI   4.4(5) x 105 s-1  
H122 FQ 7.1(4) x 106 s-1 (b) 1.9(2) x 106 s-1 1.3(1) x 106 s-1
 
E˚ (vs. NHE) 326 mV (a) 350 mV (c) 184 mV (c)
DG˚ (est.) 756 mV (a) 712 mV (c) 898 mV (c)
 
FQ = rates determined by the flash/quench technique
PI = rates determined by the photoinduced technique
a. (Di Bilio et al. 1997)
b. (Langen et al. 1995)
c. (Karlsson et al. 1997)

Figure 2.9 Electron transfer rates for intramolecular electron transfer from ruthenium labels at positions 83 and 122 to the WT and M121E azurin Cu sites. ET rates to the M121E center could not be determined with the photoinduced methodology at pH 8.0; see text for interpretation.

References

Allen, J. P., Feher, G., Yeates, T. O., Koniya, H. and Rees, D. (1987). ˘Structure of the reaction center from Rhodobacter sphaeroides R-26 - the protein subunits.÷ Proc. Natl. Acad. Sci. USA 84: 6162-6166.

Boxer, S. G. (1990). ˘Mechanisms of long-distance electron transfer in proteins: lessons from photosynthetic reaction centers.÷ Annu. Rev. Biopys. BIophys. Chem. 19: 267-299.

Chang, I.-J., Gray, H. B. and Winkler, J. R. (1991). ˘High-driving force electron transfer in metalloproteins: Intramolecular oxidation of ferrocytochrome c by Ru(2,2'-bpy)2(im)(His-33)3+J. Am. Chem. Soc. 113: 7056-7057.

Chang, T. K., Iverson, S. A., Rodrigues, C. G., Kiser, C. N., Lew, A. Y. C., Germanas, J. P. and Richards, J. H. (1991). ˘Gene synthesis, expression, and mutagenesis of the blue copper proteins azurin and plastocyanin.÷ Proc. Natl. Acad. Sci. USA 88: 1325-1329.

Connors, K. A. (1990). Chemical kinetics: the study of reactions rates in solution. New York, VCH Publishers, Inc.

Deisenhofer, J., Epp, O., Miki, K., Huber, R. and Michel, H. (1984). ˘X-ray structure-analysis of a membrane-protein complex - electron-density map at 3╩ resolution and a model of the chromophores of the photosynthetic reaction center from Rhodopseudomonas viridisJ. Mol. Biol. 180: 385-398.

Di Bilio, A. J., Hill, M. G., Bonander, N., Karlsson, B. G., Villahermosa, R. M., Malmstr¸m, B. G., Winkler, J. R. and Gray, H. B. (1997). ˘Reorganization energy of blue copper: effects of temperature and driving force on the rates of electron transfer in ruthenium- and osmium-modified azurins.÷ J. Am. Chem. Soc. 119: 9921-9922.

Faham, S., Day, M. W., Connick, W. B., Crane, B. R., Di Bilio, A. J., Schaefer, W. P., Rees, D. C. and Gray, H. B. (1998). ˘Structures of ruthenium-modified Pseudomonas aeruginosa azurin and [Ru(2, 2'-bipyridine)2(imidazole)2]SO4 Ş10H2O.÷ submitted.

Germanas, J. P., Di Bilio, A. J., Gray, H. B. and Richards, J. H. (1993). ˘Site saturation of the histidine-46 position in Pseudomonas aeruginosa azurin: Characterization of the His46Asp copper and cobalt proteins.÷ Biochem. 32: 7698-7702.

Johnson, E. C., Sullivan, B. P., Salmon, D. J., Adeyemi, S. A. and Meyer, T. J. (1978). ˘Synthesis and properties of the chloro-bridged dimer [(bpy)2RuCl]22+ and its transient 3+ mixed-valence ion.÷ Inorg. Chem. 17(8): 2211-2215.

Karlsson, B. G., Tsai, L.-C., Nar, H., Sanders-Loehr, J., Bonander, N., Langer, V. and Sj¸lin, L. (1997). ˘X-ray structure determination and characterization of the Pseudomonas aeruginosa azurin mutant Met121Glu.÷ Biochem. 36: 4089-4095.

Kiser, C. N. (1997). Unpublished observations.

Kunkel, T. A., Bebenek, K. and McClary, J. (1991). ˘Efficient site-directed mutagenesis using uracil-containing DNA.÷ Methods Enzymol. 204: 125-139.

Langen, R. (1995). Electron transfer in proteins: theory and experiments. Pasadena, California Institute of Technology.

Langen, R., Chang, I.-J., Germanas, J. P., Richards, J. H., Winkler, J. R. and Gray, H. B. (1995). ˘Electron tunneling in proteins: coupling through a b-strand.÷ Science 268: 1733-1735.

Nar, H., Huber, R., Messerschmidt, A., Filippou, A. C., Barth, M., Jaquinod, M., van de Kemp, M. and Canters, G. W. (1992). ˘Characterization and crystal structure of zinc azurin, a by-product of heterologous expression in Escherichia coli of Pseudomonas aeruginosa copper azurin.÷ Eur. J. Biochem. 205: 1123-1129.

Pascher, T., Karlsson, B. G., Nordling, M., Malmstr¸m, B. G. and Vännůrd, T. (1993). ˘Reduction potentials and their pH dependence in site-directed mutant forms of azurin from Pseudomonas aeruginosa.÷ Eur. J. Biochem. 212: 289-296.

Regan, J. J., Di Bilio, A. J., Langen, R., Skov, L. K., Winkler, J. R., Gray, H. B. and Onuchic, J. N. (1995). ˘Electron tunneling in azurin: the coupling across a b-sheet.÷ Chem. & Biol. 2: 489-496.

Sigfridsson, K., Sundahl, M., Bjerrum, M. J. and Hansson, ┘. (1996). ˘Intraprotein electron transfer in a rutheinum-modified Tyr83-His plastocyanin mutant: evidence for strong electronic coupling.÷ JBIC 1: 405-414.

Winkler, J. R. and Gray, H. B. (1992). ˘Electron transfer in ruthenium-modified proteins.÷ Chem. Rev. 92: 369-379.

Wuttke, D. S. (1994). Preparation, characterization, and intramolecular electron-transfer studies of ruthenium-modified cytochromes c. California Institute of Technology.

Yeates, T. O., Koniya, H., Rees, D., Allen, J. P. and Feher, G. (1987). ˘Structure of the reaction center from Rhodobacter sphaeroides R-26 - membrane-protein interactions.÷ Proc. Natl. Acad. Sci. USA 84(18): 6438-6442.