Photochemistry and Radiation Chemistry - ACS Publications

---

6

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

Study of Oxyferryl Heme Reactivity Using Both Radiation and Photochemical Techniques A. M. English , T. Fox , G. Tsaprailis , C. W. Fenwick , J. F. Wishart , J. T. Hazzard , and G. Tollin 1

1

1

3

1

2

3

Department of Chemistry and Biochemistry, Concordia University, Montreal, Quebec, Canada H3G 1M8 Department of Chemistry, Brookhaven National Laboratory, Upton, NY 11973 Department of Biochemistry, University of Arizona, Tucson, AZ 85721 1

2

3

Flash photolysis and pulse radiolysis were used to generate reductants in situ to study the electron transfer (ET) reactivity ofthe Fe =O heme centers in myoglobin and cytochrome c peroxidase. Reduction of a Ru groups covalently bound to surface histidines allowed intramolecular Ru -->Fe =O ET rates to be measured. Protonation of the oxene ligand was found to be largely rate determining in myoglobin, consistent with the lack of proton donors in its heme pocket. The large distance (21-23Å)between surface histidines and the heme in wild-type cyto­ chromec peroxidase prevented the determination of the rate-limiting step(s) involved in Fe =O reduction in this peroxidase, and strategies for attachment ofan artificial redox center closer to its heme are outlined. From the work performed to date, pulse radiolysis appears to be a more versatile technique than flash photolysis for the study of Fe =O heme reactivity in proteins. IV

III

5

II

IV

IV

IV

vJxyferryl heme centers {¥e =Q) are now believed to be reactive intermedi­ ates in all heme enzymes that undergo redox catalysis. F e = 0 species have been observed or predicted in heme peroxidases (1), catalases (2), oxygenases (3), and cytochrome c oxidase (4). Myoglobin (Mb), which normally functions as a reversible 0 -binding protein (5) and does not undergo redox catalysis, will, however, react with H 0 to generate an Fe ==0 center (6, 7). The detailed mechanisms of formation and decay of the transient Fe^^^^ intermediates in heme oxygenases and oxidases, such as cytochrome P and JW

I V

2

2

2

IV

4 5 0

©1998 American Chemical Society

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

81

82

PHOTOCHEMISTRY AND RADIATION CHEMISTRY

cytochrome c oxidase, respectively, are complex and the subjects of much con­ troversy {3,4). The mechanism of formation of the stable F e = 0 centers in small heme peroxidases such as cytochrome c peroxidase (CCP) and horseradish peroxidase (HRP) is much better understood (8, 9). The steps involved i n peroxidase catalysis can be summarized by the following scheme: I V

Fe ,P + H 0 m

2

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

+

Fe™=0 F 9

2

—Fe

I V

=0,F

+

+ H 0

(1)

2

+ le~ — F e ^ ^ P

(2)

F e = 0 , P + le~ + 2 H — F e , P + H 0 I V

+

m

(3)

2

The resting form ( F e , P ) reacts rapidly (>10 M " s" ) with H 0 to generate the two-electron oxidized intermediate termed compound I ( F e = 0 , P " ) , where P * is a cation radical that is located either on the porphyrin or protein (J). Compound I is generally reduced back to the resting form in one-electron reduction steps via compound II ( F e = 0 , P ) . The electron donors can be a large variety of species, including the macromolecule ferrocytochrome c, which is the physiological reducing substrate for C C P (JO), and small aromatic donors for H R P (11), Studies on mutant forms of C C P , where key catalytic residues around the heme (Figure la) have been mutated, reveal that the distal His52, and, to a lesser extent, the distal Arg48, control the rate of reaction 1 (12,13); similar studies on H R P mutants have confirmed the catalytic importance of the distal His and Arg (14). Although M b possesses both distal and proximal His residues like the peroxidases (Figure lb), metMb reacts with H 0 over 10 -fold more slowly than the peroxidases (15). Thus, the high reactivity of peroxidases with H 0 has been ascribed to the effective roles played by both the distal Arg and His. The Arg promotes ionization of H 0 in the heme cavity while the neighboring His accepts the proton, allowing the peroxy anion to bind to the F e heme. Heterolytic cleavage of the O - O bond of the peroxy ligand is promoted by back proton transfer (PT) from the distal His to the Op atom, and (at least in H R P ) (14) by stabilization of the transient negative charge on Op by the distal Arg. Ionization of H 0 in the apolar heme pocket of Mb, where Phe43 is a position equivalent to the distal Arg (Figure lb), is anticipated to be less favorable than in the peroxidases, accounting in part for the 10 -fold slower rate of reaction 1 in metMb. The efficiency of heme peroxidase catalysis also depends on the rates of reduction of compounds I and II (equations 2 and 3). For both C C P (16) and H R P (11), the rate-limiting step under optimal conditions is that of Fe =0 reduction. Details of how heme proteins control the reactivity of F e = 0 cata­ lytic intermediates are poorly understood. The F e = 0 catalytic intermediates of cytochrome P450 and cytochrome c oxidase are highly unstable (4, 17). In fact, substrate must already be bound close to the heme in cytochrome P 5o(eam) before the catalytic redox cycle begins to ensure that substrate hym

7

1

1

2

2

IY

+

+

IV

2

2

5

2

2

2

2

m

2

2

5

w

I V

I V

4

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

ENGLISH ET AL.

Oxyferryl Heme Reactivity

83

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

6.

Figure 1. (a) Heme pocket of CCP showing the key catalytic residues. The dashed lines represent H-honds. This diagram was generated using the X-ray coordinates for the 1.7-A structure of CCP (44). (b) Diagram of the heme pocket of Mb generated using the X-ray coordinates for the 1.9-A structure of horse heart metmyoglobin (51).

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

84

PHOTOCHEMISTRY AND RADIATION CHEMISTRY

droxylation is coupled to electron transfer (17). In contrast, the F e = 0 inter­ mediates in heme peroxidases, in particular the compound II species, exhibit half-lives on the order of hours (I). Hence, the reduction of the F e = 0 heme in peroxidases by a variety of redox reagents, including small inorganic complexes (18), ferrocytochrome c (1,19), and aromatic compounds (11, 20), has been studied in detail. For example, the rate of reduction of the F e = 0 heme in H R P by reagents such as F e ( C N ) ~ was observed to be p H dependent and increase at low p H (21), as expected for reductive protonation of the oxene ligand (equation 3). However, in the studies performed to date, p H dependence arising from association of the reagents cannot be distinguished from that due to protonation of the oxene ligand. To avoid the usual uncertainties associated with bimolecular electron trans­ fer (ET) reactions (22), a number of laboratories have covalently bound redox reagents such as a R u (a = N H ) to surface His residues of a variety of redox proteins (23-25). The same approach is under way in our laboratories to compare the E T reactivity of the F e = 0 heme in both C C P and M b . This comparison is of interest given that C C P is designed to rapidly turn over H 0 , which requires both rapid formation and decay of the Fe =0 center in the presence of reducing substrates (equations 1-3), whereas the 0 -storage func­ tion of M b requires that it reversibly bind dioxygen without O - O bond cleavage. Following surface His ruthenation, the intramolecular E T reaction of inter­ est is the following: I V

I V

I V

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

6

m

5

4

3

I V

2

2

w

2

+ 2 H -» a R u - F e

a Ru -Fe =0 5

ll

+

w

5

m

m

+ H 0

AG° ~ 1 eV (4)

2

Because of the highly favorable driving force for reaction 4, a R u —• F e = 0 heme E T was expected to be rapid (26, 27). Hence, the a R u reductant was generated in situ using both flash photolysis and pulse radiolysis techniques, as outlined in the following sections. An alternative approach to the generation of suitable protein-bound redox was also investigated. Nitration of surface Tyr residues in C C P was carried out to generate protein-bound reducing N 0 ' ~ - T y r radicals in situ (28), and our preliminary results are provided in the section "Pulse Radiolysis Studies of C C P . " Finally, the use of flash photolysis and pulse radiolysis techniques in the study of Fe ==0 heme systems is compared. n

5

5

I V

H

2

IV

Flash Photolysis Studies of CCP Modification of the three surface His residues of C C P (Figure 2a) was at­ tempted as described in detail (29). Derivatives containing asRu groups on His60 and His6 were isolated, and flash photolysis was used to photoinitiate E T (27, 30). Ru(bpy) (bpy = 2,2'-bipyridine) has been used extensively as a photoredox reagent, and its long-lived excited state efficiently reduces asRu bound to surface His residues of proteins (31). For example, *Ru(bpy) re111

3

2+

111

3

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

2+

ENGLISH ET AL.

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

Oxyferryl Heme Reactivity

Figure 2. (a) Surface histidine and tyrosine residues and heme of CCP superimposed on the C backbone. The cytochrome c binding domain is centered at the arrow, (b) Surface histidine residues and heme of horse heart myoglobin. a

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

86

PHOTOCHEMISTRY AND RADIATION CHEMISTRY

duces the R u and F e centers of a Ru(His33)cytochrome c with rates of ~7 X 10 and ~10 M s" , respectively, resulting in the formation of excess kinetic over thermodynamic product (32). The ground-state RuCbpy^ * complex is a strong oxidant (E° ~ 1.3 V) (33), so ethylenediaminetetraacetic acid (EDTA), or other sacrificial electron donors are added to prevent back E T . Unlike cyto­ chrome c, C C P is negatively charged at p H 7, so quenching of the negatively charged complexes Ru(DIPS) ~ (DIPS = 4,7-di(phenyl-4-sulfonate)-l,10phenanthroline) and Ru(DIC) ~ (DIC = 4,4'-dicarboxy-2,2'-bipyridine) by the surface-bound a R u center was anticipated to be electrostatically favored over quenching by the C C P heme. Studies using the * R u L photoredox reagents were performed with the xenon flash photolysis equipment at Concordia University described previously (30). C C P samples were titrated spectrophotometrically with H 0 to form C C P ( F e = 0 ) prior to use. For the purpose of this study, compound I of C C P will be designated as C C P ( F e = 0 ) because the fate of the protein radical ( F ) , which is located on Trpl91 (Figure la) (1), following flash photolysis of a Ru(His)CCP was not examined. However, Trpl91 is not on any direct E T pathway between the Ru and heme centers (Figures l a and 2a). Anaerobic solutions of 50-100 |xM R u L , 1-10 m M E D T A , and 5 JJLM C C P ( F e = 0 ) or 5 jxM ferricytochrome c in 0.1 M phosphate buffer, p H 7.0, were photoexcited with 25-|xs pulses from the xenon lamp. When ferricytochrome c solutions were flashed, rapid growth of absorbance at 550 nm due to ferrocytochrome c formation was observed on the millisecond time scale as reported previously (32). When C C P ( F e = 0 ) or a Ru(His60)CCP(Fe =O) solutions were flashed under identical conditions, reduction of the F e = 0 heme was observed at 564 nm, a maximum in the F e minus Fe ==0 difference spec­ trum, on the same time scale for both samples. Similar results were obtained with the negatively charged R u L complexes, suggesting that quenching of * R u L by the Fe =0 heme is more efficient than by a R u . Subsequently, direct k measurements revealed values of 1-2 X 10 M " s" for quenching of * R u L by C C P ( F e ) and C C P ( F e = 0 ) (A. English, Concordia University, unpublished results). A further problem encountered, even with 450-nm cutoff filters to eliminate U V light, was rapid photoinduced autoreduction of C C P ( F e = 0 ) in the absence of R u L / E D T A on exposure to the xenon flash. m

8

m

8

- 1

5

1

3

4

3

3

5

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

4

m

3

2

2

I V

I V

+

5

I V

3

I V

IV

5

I V

m

IV

3

w

3

m

3

m

5

9

q

1

1

I V

I V

3

Deazariboflavin semiquinone ( D R F L H ) had been shown to exhibit low reactivity (/; ~ 10 M " s" ) with the heme of C C P (34); thus, D R F L H * was next chosen to selectively reduce the a R u group on the surface of C C P . Photoexcitation of the flavin quinone (FL) by visible light generates a singlet which decays to the triplet ( F L ) in -10 ns. The triplet can abstract a hydrogen atom from donors such as E D T A to form the flavin semiquinone ( F L H ) in 10 M " s~\ respectively, in 40 m M phosphate buffer at p H 7.0 (26). Thus, most of the reduction occurred at the asRu center in the derivatized protein following the 60-ns electron pulse. Because the reduction potentials for the a R u ( H i s ) and F e heme cen­ ters are closely matched (ΔΕ°' = 19 mV), the observed rate constant for intra­ molecular E T over the 12.7 A from R u to the heme followed reversible firstorder kinetics (26): 11

I V

I V

m

2

Downloaded by UNIV MASSACHUSETTS AMHERST on July 31, 2012 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/ba-1998-0254.ch006

I V

IV

m

5

2

n

m

m

2

2

5

8

2

9

m

1

111

m

5

111

11

a5Ru (His48)HHMb(Fe -OH ) m

ni

C O "~

2

^

Fast

a Ru (His48)HHMb(Fe -OH ) n

5

in

(14)

2

a R u ( H i s 4 8 ) H H M b ( F e - O H )