| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Cellular Engineering and Biocatalysis, Biotechnology Institute, National University of Mexico, México; and
Department of Chemistry, University of Siena, Siena, Italy
1Correspondence: Instituto de Biotecnología, Universidad Nacional Autónoma de México. AP 510–3, Cuernavaca, Morelos, CP 62250, México. E-mail: brenda{at}ibt.unam.mx
SPECIFIC AIMS
The goals of this work were to 1) demonstrate that the oxidative inactivation of cytochrome c by hydrogen peroxide is mechanism-based and 2) achieve the oxidative stabilization of yeast iso-1-cytochrome c, without further compromise of its catalytic properties, through the rational alteration of intramolecular electron transfer networks.
PRINCIPAL FINDINGS
1. Cytochrome c oxidative inactivation occurs as a consequence of a naturally unproductive mechanism
We previously proposed that the oxidative stability of hemeperoxidases is a consequence of the competition between productive (the added reducing substrate) and unproductive electron sources (catalytically important enzyme components). This assumption suggests that the redox properties of the added substrate would influence the electron abstraction partition between sources controlling the stability of the enzyme under operation conditions. To demonstrate this we calculated the total turnover number of an iso-1-cytochrome c preparation using members from a set of diphenyl sulfur derivatives with different ionization potential values as substrates. We found a good correlation between the substrate ionization potential value and the total turnover number of the preparation with a 
30-fold extension of the operational life of the enzyme. This evidence indicates that the oxidative stability of iso-1-cytochrome c as a peroxidase can be significantly enhanced by including exogenous substrates with lower redox potential, and supports the existence of simultaneous and competing electron sources.
2. The partition of the oxidizing equivalents toward the exogenous substrate instead of the protein elements can be modulated by site-directed mutagenesis
Our experimental strategy was aimed at altering the intramolecular electron transfer pathways by site-directed mutagenesis. To achieve this, a rational design based on the redox properties of the amino acid side chain moieties was devised. Among the redox-sensitive amino acid side chains, Tyr residues present the lowest redox potential and emerged as the most probable electron donor. The involvement of all five Tyr residues of yeast iso-1-cytochrome c was tested by multiple replacements. This variant preserved 77% of the activity, enhanced the Soret band half-life by 2.6-fold but decreased the stability of the preparation to only 8.3% (Table 1
). Of the five mutated residues, only Tyr 67 interacts with the heme group through a well-delineated and flexible hydrogen bond network comprising the porphyrin moiety, the conserved water molecule Wat166, and other amino acid residues such as N52.
|
Disruption of the H-bond network in the N52I,Y67F double mutant abated its peroxidase activity below 0.1% and decreased the stability of the enzyme to 6.1% compared to the WT16 variant (Table 1)
. Stabilization of the Soret band was absolute even after extensive exposure to 1 mM H2O2. The reduced catalytic activity and the evident heme group stability suggest that the added substrate and the porphyrin were no longer used as electron donors, but only the protein scaffold, explaining the reduced stability. The fundamental principle underlying our approach to the oxidative inactivation phenomenon is that it is mechanism based. The poor catalytic performance of the N52I,Y67F double mutant raises the question of whether the apparent porphyrin stability could be due to its low turnover number and not to a distorted use of alternative electron sources. The only solution to this subject was to further modify the mutant iso-1-cytochrome c in order to restore the catalytic performance.
Aside from Tyr, other oxidation-sensitive residues were tested, including the single Trp residue at position 59. The N1H atom of the indole ring is H-bonded to the rear heme propionate and is likely to be important to the structural and functional properties of cytochromes c. To trim the H-bond network connecting the heme propionates with the protein backbone even further, we performed site-directed substitution of the single Trp residue for Phe in the N52I,Y67F background (Table 1)
. While the activity and heme stabilities were not significantly altered, the W59F substitution improved the catalytic activity of the double mutant by 900-fold, up to levels similar to the WT16. Significant restoration of the added substrate utilization, while preserving the Soret band integrity, indicates that the porphyrin was actually stabilized. Nevertheless, the stability was not improved, suggesting that other protein components were still being damaged. It is known that for a free radical to be transferred in a protein, both the receptor and the donor must be bonded. That means that in the N52I,W59F,Y67F triple mutant a residual interaction must still bind the porphyrin with the rest of the protein.
According to the 3-dimensional structure of yeast iso-1-cytochrome c, the phenyl ring of F82 is adjacent to porphyrin pyrrole rings B and C (within 5 Å) and is almost coplanar with them, leading to the proposal that the delocalized 
-electron systems of these groups are coupled, a property not shared by any other residue in the heme pocket. For this reason, we considered that F82, though apparently not bonded, may constitute the residual link between the heme and the protein backbone. The specific substitution of F82 for Gly in the N52I,W59F,Y67F triple mutant background resulted in the dramatic stabilization of the peroxidase activity, whereas the porphyrin integrity and the catalytic activity of the parental variant were kept untouched (Table 1)
.
Despite its substantial stabilization and higher affinity for hydrogen peroxide, the catalytic activity of the N52I,W59F,Y67F,F82G mutant variant was still below 75% of the WT16 value. This behavior could be explained by a reduced access of the substrate to the porphyrin moiety, probably caused by the accumulated cavity reorganization. In principle, disruption of the H-bond interaction between the side chains of K79 and S47 that kept the heme crevice occluded may allow a faster exchange rate between the active site and the surrounding medium. As can be seen in Table 1
, the catalytic performance of the N52I,W59F,Y67F,K79A,F82G variant improved to 90% of the WT16 value; also, the stability of the enzyme reflected the consequences of the substitution. Despite a discrete reduction of the Soret band stability, activity was completely stabilized. The affinity for hydrogen peroxide was slightly reduced compared to the parental variant, but the catalytic efficiency of the N52I,W59F,Y67F,K79A,F82G variant was restored to 85% of the WT16 value.
3. The macroscopic stability properties of the mutant cytochrome c variants were preserved at electronic levels as demonstrated by direct EPR
It has been concluded that the generation of amino acid-based radicals in iso-1-cytochrome c after exposure to hydrogen peroxide occurs by a peroxidase-type mechanism, although none of the catalytic intermediates have been characterized to date. Classical compound I consists in a high-valent oxo-iron(IV)porphyrin-based -free radical. In some cases, the second oxidation equivalent in compound I is delocalized into the protein. Identification of a protein-based radical with g value of 2.004 ± 0.0005 in WT16 can be compared to the formation of amino acid-based free radicals in other heme proteins. This signal appeared in WT16 after 10 s exposure to hydrogen peroxide (1:5, protein:H2O2) and increased even further after 2 min incubation, probably due to the progressive delocalization of the electron in the protein with the concomitant decrease in the iron signal (Fig. 1
A). Taking into consideration that this protein-based radical is present in the stable variants and that it accumulated even under noninactivating conditions, we suggest that in iso-1-cytochrome c, this radical is probably a catalytic intermediate mechanistically equivalent to the secondary species of compound I. Furthermore, the site of this radical seems to be a tyrosine residue. The selective substitution of all tyrosine positions for phenylalanine evidenced the transient formation of a classical compound I intermediate harboring a porphyrin-based radical (g||=2.52 and g
=2.0 values). After 2 min incubation, the g = 2.0 signal decreased as a very intense signal with g = 4.3 value accumulated. (Fig. 1B
). Since the EPR bands reflecting the porphyrin-bound iron status were not drastically affected, we suggest that in the absence of tyrosine residues an initiating radical would be transferred around the protein until reaching the porphyrin as a secondary site; nevertheless, this secondary site is not stable and eventually the electron determines some modification at the expense of iron.
|
CONCLUSIONS AND SIGNIFICANCE
We have described the oxidative inactivation of hemeproteins as an unusual case of enzymatic performance where the enzyme itself is used as an electron source and therefore should be considered in the stoichiometry of the reaction. We assumed the inactivation to be a consequence of multiple simultaneous oxidations susceptible to be observed and altered independently. Based on these premises, we proposed and developed a protein engineering strategy aimed at remodeling critical electron transfer contacts. This was accomplished by the selective disruption of redox-sensitive residues without further compromise of the activity, resulting in a variant with enhanced stability.
The mechanism-based oxidative inactivation of hemeperoxidases has been addressed from different viewpoints, including site-directed mutagenesis and directed evolution. In general, evolution strategies had not been successful in the isolating more stable peroxidases. A possible explanation for this shortfall of supportive results arises from the study of the evolution of physiological electron tunneling in oxidoreductases. Contrary to theoretical predictions, the view that the amino acid matrix has been optimized naturally to guide electron tunneling seems to lack experimental support. In contrast, the single characteristic that has been selected in the evolution of electron tunneling in oxidoreductases is the simple positioning of redox centers within reach. This position considers that almost every change of an amino acid will provide alternative bonds to form a comparable pathway, including water molecules. Conserved amino acids and structural motives in these regions may have many other biologically important functions with their own selective pressures not directly related to electron tunneling.
After these considerations, we support the point of view that most of the primary sequence of the protein is irrelevant for the electron transfer, since the chemical properties of the residues involved in the pathways are less critical that the residue packing between redox centers (45). As a consequence, only a redox-inspired strategy is likely to succeed in improving a complex circuitry such as the intramolecular electron transfer during the peroxidases inactivation, as demonstrated in this article.
In conclusion, the rational strategy applied on iso-1-cytochrome c yielded a more stable protein toward the oxidative inactivation by hydrogen peroxide. The phenotype of this variant was reflected in its electronic properties, allowing identification of a tyrosine residue as the site of a protein-based radical intermediate, mechanistically similar to compound I. Since the design relied on the oxidation sensitivity of the enzyme components more than in the structure, this strategy might be applied to any hemeprotein given the sequence is known.
|
FOOTNOTES
2 Present address: Department of Molecular Biology and Biotechnology, Biomedical Research Institute, National University of México, AP70228, México, D.F., México. 
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-4173fje
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= 9.390 GHz, 10 G modulation amplitude, 2 mW microwave power, and 100 kHz modulation frequency.