Photochemistry and Radiation Chemistry - American Chemical Society

---

5

Downloaded by PENNSYLVANIA STATE UNIV on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch005

Pulse Radiolysis: A Tool for Investigating Long-Range Electron Transfer in Proteins I. Pecht and O. Farver 1

2

Department of Immunology, Weizmann Institute of Science, Rehovot 76100, Israel Department of Chemistry, Royal Danish School of Pharmacy, Copenhagen, Denmark 1

2

One of the fundamental processes in biological energy-conversion sys­ tems is that ofelectron transfer. It takes place among and within proteins over considerable distances between active sites containing transition metal ions or organic cofactors and is generally characterized by rela­ tively weak electronic interactions among them. Considerable research efforts have and are being invested in resolving the factors that control the rates of long-range electron transfer in proteins. Different exper­ mental methods have been employed in these studies, ranging from fast mixing (stoppedflow,rapidfreeze EPR) and T-jump chemical relaxation to flash photolysis and pulseradiolysis. The latter two methods employ introduction of very short electromagnetic radiation pulses, absorbed by solutes in the former and by solvent in the latter. These in turn produce the electron donors or acceptors initiating the reaction of inter­ est.The application of pulse radiolysis to studies of electron transfer within proteins is briefly reviewed to indicate its advantages. Results of its application to two types of copper-containing electron mediating proteins are presented and discussed.

Pulse radiolysis is a method first introduced in the 1960s, and it has found a broad range of important applications in chemistry and biochemistry, extending far beyond the scope of free radicals and radiation chemistry to which it was first applied (1). One application that is of considerable significance is the investigation of electron transfer processes in proteins. The method is based on the excitation and ionization of solvent molecules by short pulses of highenergy electrons. Thus, although pulse radiolysis is the electron analog of the ©1998 American Chemical Society In Photochemistry and Radiation Chemistry; Wishart, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1998.

65

66

PHOTOCHEMISTRY AND RADIATION CHEMISTRY

flash photolysis method, photoexcitation of specific solutes rather than of the bulk solvent molecules distinguishes these two methods and provides the for­ mer with some concrete advantages. Introducing high-energy electrons (e.g., 5-10 MeV) into dilute aqueous solutions of a given solute causes primary changes only in the solvent. Thus, water molecules undergo conversion into O H radicals and hydrated electrons, and to a lesser extent, H atoms, H , and H 0 are also produced. (The yields are usually presented as G values, that is, the number of entities produced per 100 eV of absorbed energy: e" = 2.9, O H = 2.8, H = 0.55, H 0 = 0.75, and H = 0.45.) The hydrated electrons and O H radicals are exceptionally reactive and present thermodynamic extremes of reducing and oxidizing poten­ tials, respectively. Hence, they provide the possibility of initiating a wide range of electron transfer processes. As will be detailed later in this chapter, these highly reactive agents, though having their own applications, are usually converted to less aggressive agents via protocols devised by radiation chemists. This is illustrated by one useful procedure, that of converting the e^q (with a reduction potential of -2.8 V) to a considerably milder reductant, the C 0 ~ radical (E° = -1.8 V). First the former is converted, in N 0-saturated solutions, into an additional equivalent of O H radicals by the following reaction: 2

2

2

Downloaded by PENNSYLVANIA STATE UNIV on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch005

q

2

2

2

2

2

e^q + N 0 —• N + O H + O H " 2

2

The two equivalents of O H radicals then react with formate ions, also present in the solution, to produce two equivalents of the C 0 " radical: 2

HC0 ~ + OH — H 0 + C 0 2

2

2

By analogy, other reducing and oxidizing agents can be produced and employed (2). An additional, technically important advantage of pulse radiolysis over flash photolysis is noteworthy, namely, the whole range of optical spectrum is usually available for monitoring the induced reactions by the former method. This is the case because the reactive species is derived from the solvent rather than from the excitation of a given solute, as is the case in flash photolysis. Taken together, the wide range of chemical reactivity of the produced reagents, com­ bined with a time resolution that usually extends from nanoseconds to minutes and the convenience of spectrophotometrie monitoring of the reactions, has made pulse radiolysis the method of choice for investigation of a wide range of chemical processes. The potential of pulse radiolysis for studying biological redox processes, particularly of macromoleeules, was recognized rather early. It was initially employed for investigating radiation-induced damage and, later on, as an effec­ tive tool for resolving electron transfer processes to and within proteins. Cyto­ chrome c, a well-characterized electron-mediating protein, was the first to be

In Photochemistry and Radiation Chemistry; Wishart, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1998.

5.

PECHT & FARVER

67

Pulse Radiolysis

examined for its reactivity with hydrated electrons by this method. Two groups pioneered this study (3, 4) and found the reduction of the F e site to be a diffusion-controlled process (5 X 1 0 M s ) . These two and several other groups (e.g., 5-7) then extended the investigation of this protein. Soon there­ after, other electron-mediating proteins with distinct redox centers like copper or nonheme iron (8,9) were also examined for their reactivity with e^q, yielding similar, rather high, diffusion-limited rate constants. These results illustrated the applicability of the method, yet at the same time they indicated the draw­ backs of the hydrated electron as a reductant with excessive driving force and therefore limited specificity, which leads to additional side reactions. Hence, new studies increasingly employed milder reducing or oxidizing agents than the e^q or O H radicals, respectively (cf. ref. 10). Two main interests were guiding studies that applied pulse radiolysis to redox proteins: the pursuit of reaction mechanisms of these proteins, and the more general problem of resolving the parameters that determine specific rates of electron transfer within proteins (10-12). Obviously, these two motives over­ lap and complement each other, as we shall see. The fast progress attained in the last two decades in resolving three-dimensional structures of an increasing number of redox-active proteins provided the insights essential for a meaningful analysis and interpretation of the kinetic results. 3 +

Downloaded by PENNSYLVANIA STATE UNIV on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch005

10

_ 1

_ 1

Azurin as a Model System for Intramolecular Electron Transfer Azurins are "blue" single-copper proteins that mediate electrons in the energy conversion systems of several bacterial strains (13, 14). Although azurins iso­ lated from distinct bacteria are highly homologous in their sequences, differ­ ences do exist, conferring upon these proteins variation in reactivity and redox potentials (13). A l l azurins sequenced to date contain a disulfide bridge (Cys3-Cys26) at one end of their P-sandwich structure, 2.6 nm from the copper binding site present at the opposite end of the barrel-shaped protein (Figure 1) (13,14). Using pulse-radiolytically produced C 0 " radicals, this disulfide is reduced to the RSSR~ radical, and this transient species was found to decay by an intramolecular electron transfer process to the copper(II) center (15, 16). This process has been investigated in greater detail in both wild type (wt) and single-site mutated azurins (15-19). The former were isolated from differ­ ent bacteria, exhibited a range of differences in their properties (15, 16), yet were less amenable to analysis than the latter. Hence, changes in specific param­ eters of the protein seen in single-site mutants of Pseudomonas aeruginosa (Pae) azurin constitute a main topic of these studies. For example, mutants that have redox potentials close to that of the wild type (wt) ( £ ° = 304 mV) but differ in the residues proximal to the copper center were investigated. In one of these, the Met64 residue of the wild type protein is substituted by Glu 2

In Photochemistry and Radiation Chemistry; Wishart, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1998.

Downloaded by PENNSYLVANIA STATE UNIV on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch005

68

PHOTOCHEMISTRY AND RADIATION CHEMISTRY

Figure 1. Three-dimensional structure of the polypeptide backbone ojf Pseudomonas aeruginosa azurin, with some amino acid residues of particular interest included. Coordinates were obtained from reference 21.

([M64E]azurin) with a Cu(II)-Cu(I) redox potential of 278 mV (18). In two other mutants, Ser has been introduced instead of either Ue7 ([I7S]azurin) or P h e l l O ([F110S]azurin) with redox potentials at p H 7.0 of 301 and 314 mV, respectively (17, 19). Pulse-radiolytic reductions are routinely performed in N 0-saturated aqueous solutions containing 10 m M phosphate and 100 m M formate ions. Under these conditions, the primary products of water decomposition by the pulse of accelerated electrons are practically all converted into C 0 ~ radical ions (10) that react rapidly with the two redox-active sites present in azurin: the Cu(II) ion and the disulfide bridge (8,15). This is illustrated by absorption changes observed solutions of oxidized, single-site mutant (F100S) of Pae azurin following the pulse-radiolytically produced C 0 " radicals (Figure 2). When the reaction is monitored at the main absorption band of the Cu(II) (e625 = 5700 M " c m ) , an initial, relatively fast phase of decrease in absorption is noted. This decay was found to be a second-order process corresponding to a direct, diffusion-controlled bimolecular reduction of the Cu(II) center by C 0 ~ radi­ cals. However, as is illustrated by Figure 2a, a slower phase of Cu(II) reduction 2

2

2

1

-1

2

In Photochemistry and Radiation Chemistry; Wishart, J., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1998.

5.

P E C H T & FARVER

69

Pulse Radiolysis

Ps. Azurin Cu(ll) reduction at 625 nm

Downloaded by PENNSYLVANIA STATE UNIV on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch005

0.01 -rrr

-0.03

I i

1

I '

: ; : i i i I i i i

Time scale:10 us

a

I

i

I

Time scale.i 0 ms

Ps. Azurin RSSR* formation/decay at 410 nm