Crystal structure and characterization of a cytochrome c peroxidase-cytochrome c site-specific cross-link

Abstract

A specific covalently cross-linked complex between redox partners yeast cytochrome c peroxidase (CCP) and cytochrome c (cyt. c) has been made by engineering cysteines into CCP and cyt. c that form an intermolecular disulfide bond in high yield. The crystal structure of the cross-linked complex has been solved to 1.88-A resolution and closely resembles the structure of the noncovalent complex [Pellitier, H. & Kraut, J. (1992) Science 258, 1748-1755]. The higher resolution of the covalent complex has enabled the location of ordered water molecules at the peroxidase-cytochrome c interface that serve to bridge between the two proteins by hydrogen bonding. As in the noncovalent complex, direct electrostatic interactions between protein groups appear not to be critical in complex formation. UV-visible spectroscopic and stopped-flow studies indicate that CCP in the covalent complex reacts normally with H(2)O(2) to give compound I. Stopped-flow kinetic studies also show that intramolecular electron transfer between the cross-linked ferrocytochrome c and the Trp-191 cation radical site in CCP compound I occurs fast and is nearly complete within the dead time ( approximately 2 ms) of the instrument. These results indicate that the structure of the covalent complex closely mimics the physiological electron transfer complex. In addition, single-turnover and steady-state experiments reveal that CCP compound I in the covalent complex oxidizes exogenously added ferrocytochrome c at a slow rate (t(1/2) approximately 2 min), indicating that CCP does not have a second independent site for physiologically relevant electron transfer.

Fig. 1.

Electrophoretic analysis of cross-linked product. (A) SDS/PAGE under nonreducing conditions. Lanes 1 and 2 are the cross-linked product before and after purification, respectively, under nonreducing conditions. (B) Lane 1 is molecular weight standards, lane 2 is SDS/PAGE under nonreducing conditions, and lane 3 is SDS/PAGE of complex after boiling for 10 min in the presence of 150 mM DTT. (C) Native PAGE. Lane 1 is CCP alone, lane 2 is the complex not treated with DTT, and lane 3 is the complex treated with DTT (10 mM),4hat4°C.

Fig. 2.

2F_obs – F_calc composite omit electron density map contoured at 1σ showing the S–S cross-link between CCP and cyt. c. Figures were prepared withmolscript (34) and raster3d (35).

Fig. 3.

Superimposition of the covalent (green) and noncovalent (purple) CCP–cyt.c complexes. The CCP molecules were superimposed to show the slight difference in the orientation of cyt. c relative to CCP.

Fig. 4.

Stereo model showing the CCP–cyt.c interface. Note the ordered solvent molecules that bridge between polar groups in the two protein molecules. The thin lines represent the peptide backbone not in the interface, and dotted lines indicate hydrogen bonds.

Fig. 5.

The rate of oxidation of exogenously added ferrocyt. c by compound I in the covalently cross-linked complex (PCXL; 3.6 μM) and by three controls: 3.4 μM wild-type CCP (data not shown), 3.4 μM V197C/C128A (C197), and 3.2 μM inactive CCP mutant W191G (G191). For wild-type CCP or V197C/C128A CCP, the ferrocyt. c is fully oxidized immediately after mixing.