(Received for publication, March 6, 1997, and in revised form, April 10, 1997)
From Unité de Bioénergétique et
Ingéniérie des Protéines, IFR1-Centre National de la
Recherche Scientifique, 13402 Marseille Cedex 20, France and the
§ Department of Biochemistry, University of Missouri,
Columbia, Missouri 65211
The gene encoding Desulfovibrio desulfuricans Norway cytochrome c3 (Mr 26,000), a dimeric octaheme cytochrome belonging to the polyheme cytochrome c3 superfamily, has been cloned and successfully expressed in another sulfate reducing bacteria, D. desulfuricans G201. The gene, named cycD, is monocistronic and encodes a cytochrome precursor of 135 amino acids with an extension at the NH2 terminus of 24 amino acids. This extension acts as a signal sequence which allows export across the cytoplasmic membrane into the periplasmic space. Tyrosine 73, which is in a close contact with the histidine sixth axial ligand to the heme 4 iron atom, has been replaced by a glutamate residue using site-directed mutagenesis. The cytochrome mutant when expressed in D. desulfuricans G201, is correctly folded and matured. A global increase of the oxidoreduction potentials of about 50 mV is measured for the Y73E cytochrome. The mutation also has a strong influence on the interaction of the cytochrome with its redox partner, the hydrogenase. This suggests, like the tetraheme cytochrome c3 (Mr 13,000), heme 4 is the interactive heme in the cytochrome-hydrogenase complex and that alteration of the heme 4 environment can greatly affect the electron transfer reaction with its redox partner.
Multi-redox center proteins are involved in several reactions of energy transduction phenomena. The complexity of these proteins has hindered determination of their precise physicochemical mechanisms and the indispensable cross-assignment between each measurable parameters. Among these, multiheme cytochromes have an important role and have been extensively investigated. Discovery of a large number of multiheme cytochromes in sulfur and sulfate reducing bacteria led to the description of a cytochrome c3 superfamily (1). It includes multiheme cytochromes all belonging to the class III of c-type cytochrome as described by Ambler (2). Members of this superfamily are cytochrome c3 (Mr 13,000), cytochrome c3 (Mr 26,000), cytochrome Hmc as well as cytochrome c7 isolated from the sulfur-reducing bacterium Desulfuromonas acetoxidans (3). Three-dimensional structures of four different cytochromes c3 (Mr 13,000) are available (Ref. 4 and references therein), as well as that of cytochrome c3 (Mr 26,000) (5) and recently of cytochrome c7 (3). These analyses reveal that cytochrome c3 (Mr 13,000) is the basic unit of this superfamily and that all these cytochromes may have a common ancestral origin (1).
Cytochrome c3 (Mr 13000) is a periplasmic protein uniformly present in Desulfovibrio. To date, cytochromes c3 from Desulfovibrio vulgaris Miyazaki, Desulfovibrio vulgaris Hildenborough, Desulfovibrio desulfuricans Norway, and Desulfovibrio gigas have been extensively studied in terms of both structural (4, 6, 7) and physicochemical properties (8-12). Despite low sequence identity, their three-dimensional structures are very well conserved, especially in terms of position and orientation of the heme core.
Cytochrome c3 (Mr 26,000)
was isolated from D. desulfuricans Norway (13), D. gigas (14), and more recently from Desulfovibrio africanus (15). It is a homodimer with four hemes per subunit bound to the polypeptide chain via two thioether bonds involving cysteine residues from the consensus sequence
C-X2 or 4-C-H and two vinyl groups of the
porphyrin. All four heme iron atoms in each subunit are coordinated by
two imidazole nitrogen atoms of histidine residues, adopting a
hexacoordinated low-spin form in both the oxidized and reduced states
of the protein. The four heme groups have very low oxidoreduction
midpoint potentials ranging from 220 to
370 mV (16), enabling
reduction of relatively poor oxidants in the sulfate reduction pathway.
The three-dimensional structure of cytochrome c3
(Mr 26,000) from D. desulfuricans
Norway has been solved at 2.16-Å resolution (5). Each subunit displays the cytochrome c3 fold with significant amino
acid substitutions, relative to the tetraheme cytochrome
c3, in the interface region. The orientation and
relative arrangement of the heme core within each subunit is very
similar to those of all tetraheme cytochromes c3 (Mr 13,000). The
dimer contact is relatively small (6% of the total accessible surface)
involving only 19 residues in each subunit. No salt-bridges are located
at the interface but 16 hydrogen bonds are formed. In the case of
D. gigas, two disulfide bridges are present between the two
subunits of the dimeric cytochrome c3 (Mr 26,000) (14). The heme 1 of each subunit are
in close contact, the shortest distance between the heme edges being 5 Å. As it is the case for the cytochromes
c3 (Mr 13,000), the
heme 4 crevice is surrounded by several basic residues; 7 out of 11 lysine and arginine residues present in the cytochrome
c3 (Mr 26,000) are located near heme 4 (17).
The three-dimensional structures of these polyheme cytochromes provide a basis for understanding the factors controlling the redox potential and therefore the intra- and inter-molecular electron transfer. Sequence alignments of the various cytochromes c3 on the basis of the three-dimensional structures pointed out several highly conserved residues (17) in the heme environment that might play an important role in establishing the redox properties of the molecules. Aromatic residues might be of great importance as they are generally involved in influencing the redox potential, the electron transfer as well as the stability of the molecule. Tyrosine 73 is a highly conserved aromatic residue, positioned parallel to the plane of the sixth axial ligand to heme 4. Numerous studies have revealed the importance of this heme group in the interaction of the cytochrome c3 (Mr 13,000) with various redox partners (18) and that slight alteration of its environment has a drastic influence on the interaction and electron transfer processes (19). However, tyrosine 73 is replaced by a glutamate residue in D. desulfuricans Norway cytochrome c3 (Mr 13,000). This might be responsible for a totally different orientation of the axial ligand histidine planes that are almost perpendicular in the case of the D. desulfuricans Norway tetraheme cytochrome c3 (Mr 13,000) and approximately parallel in the case of the octaheme cytochrome c3 (Mr 26,000) (5).
In D. desulfuricans Norway cells, both cytochrome c3 (Mr 13,000) and cytochrome c3 (Mr 26,000) are present and they have only 34.6% sequence identity (14). The physiological roles for multiple multiheme cytochromes in the same bacteria are unknown. In the case of the dimeric cytochrome, the role of heme 4 as well as the importance of the dimer formation in the molecule activity are unresolved.
For insights into the physicochemical properties of cytochrome c3 (Mr 26,000) from D. desulfuricans Norway and the importance of the highly conserved residues, we cloned its gene and performed site-directed mutagenesis. Here, we report the cloning of the cycD gene encoding cytochrome c3 (Mr 26,000) and its expression in a closely related sulfate reducing bacterium, D. desulfuricans G201. The influence of the replacement by site-directed mutagenesis of tyrosine 73 which is located in the heme 4 environment on the properties and activity of the molecule has also been studied.
The bacterial strains and
plasmids used are described in Table I. Growth of
Escherichia coli strains was carried out in TY medium (20),
supplemented with the appropriate antibiotic concentrated at 0.27 mM for ampicillin and 0.17 mM for kanamycin.
E. coli DH5 and D. desulfuricans G201,
required for conjugational transfer of broad-host range vectors, were
grown as described previously (21). D. desulfuricans Norway
(NCIMB 8310) was grown at 37 °C in Starkey's medium. Large-scale
growth of D. desulfuricans G201 recombinant strains was
carried out in Postgate C medium (22) supplemented with 0.28 mM kanamycin.
|
All enzymes were obtained from
Eurogentec, Life Technologies, Inc., or Boehringer Mannheim, France.
The radiochemical [-35S]dATP (400 Ci
mmol
1) was purchased from Amersham Corp. and used for
dideoxynucleotide sequencing. Cloning and sequencing oligonucleotides
were obtained from Genset (France).
A degenerate probe 33-44D
(5-ca(ct)atgga(ct)at(act)gcitg(ct)ca(ag)ca(ag)tg(ct)ca(ct)ca(ct)ac-3
),
encoding the amino acid sequence
His33-Met-Asp-Ile-Ala-Cys-Gln-Gln-Cys-His-His44
of D. desulfuricans Norway cytochrome
c3 (Mr 26,000) (23), was
3
end-labeled nonisotopically with DIG-11-dUTP (Boehringer Mannheim
DIG Oligonucleotide Tailing Kit, catalog number 1417231) and used to
detect a 1.5-kb1 BamHI DNA
fragment from the chromosome of D. desulfuricans Norway. BamHI-digested chromosomal DNA of 1-2 kb was cloned into
pUC18 and transformed into E. coli TG1. Potential
recombinants were tested by polymerase chain reaction amplification
using forward and reverse pUC oligonucleotides (Promega). Amplified
products of about 1.5 kb were probed with the 33-44D degenerate
oligonucleotide by Southern analysis to identify the cycD
gene encoding cytochrome c3
(Mr 26,000). One positive clone was called
pUCC19.
The oligonucleotide 48-55 (5-gcagctttcaatggtatagg-3
) complementary
to the sequence encoding 48TYTIESC55 of
cytochrome c3 (Mr 26,000)
was designed and used to clone genomic DNA downstream from the
cytochrome c3 (Mr 26,000)
encoding gene. Selection and cloning of a suitable fragment was
performed as described above. A 1.5-kb fragment obtained from a double
EcoRV-SacI digestion of the genomic DNA was
selected and subsequently cloned into pUCBM21 (Boehringer Mannheim) cut
with the same enzymes to obtain pUCC20. Both the 1.5-kb
BamHI fragment from pUCC19 and the 1.5-kb
EcoRI-SacI fragment from pUCC20 were subcloned
into the replicative form of M13mp18 cut with the corresponding enzymes to give mp18C19 and mp18C20, respectively. Sequences from both strands
were determined using the method by Sanger et al. (24). A
877-bp BamHI-SalI fragment, including the
cycD gene, was isolated from pUCC19 and subcloned into
M13mp18 cut with the same enzymes to get mp18CACC3.
Total cellular RNA from D. desulfuricans Norway, D. desulfuricans G201 (pJRD215), and D. desulfuricans G201 (pCACC3) were prepared by the hot phenol method (25). The RNA was denatured by incubation 15 min at 68 °C in 65% (v/v) formamide, 2.7 M formaldehyde, 1 × MOPS buffer (pH 7.0). RNA (15 µg) was electrophoretically separated using 1.2% (w/v) agarose gels containing 2.2 M formaldehyde in 1 × MOPS buffer (pH 7.0) run in the same buffer. RNA was then transferred to a positively charged nylon membrane by capillarity blotting. Specific transcripts from the cycD gene were probed with an intragenic cycD 397-bp fragment nonisotopically labeled with DIG-11-dUTP by Northern analysis.
Site-directed MutagenesisThe site-specific change in
mp18CACC3 employed the Eckstein method (Amersham code RPN 1523). The
mutagenic oligonucleotide P73 (5-ggaaatttcctcggtcgaacggactttcc-3
) was
designed to change the TAC codon at nucleotides 557-559 encoding
Tyr73 within the cycD gene into GAA encoding E. The change was confirmed by dideoxynucleotide sequencing. The
replicative form of mp18CACC3 carrying the Y73E mutation was called
mp18CC3Y73E.
The replicative forms of mp18CACC3 and mp18CC3Y73E
were digested with both EcoRI and HindIII and the
resulting 895-bp fragments were ligated to the pJRD215 vector
previously digested with the same enzymes to give pCACC3 and pCC3Y73E,
respectively. The resulting plasmids were transformed into E. coli DH5 and subsequently transferred to D. desulfuricans G201 as described previously (21).
D. desulfuricans G201(pCACC3) and D. desulfuricans G201(pCC3Y73E) cells were obtained from 300-liter fermentations in Postgate C medium supplemented with 0.28 mM kanamycin. Cells were harvested, resuspended in 600 ml of 100 mM Tris-HCl, 100 mM EDTA (pH 9) buffer, and stirred for 30 min at 37 °C in a water bath. The mixture was then centrifuged (7700 × g, 1 h, 4 °C) and the resulting supernatant was dialyzed overnight against distilled water at 4 °C. This periplasmic fraction was centrifuged for 1 h at 140,000 × g and the supernatant was then loaded onto a column of DEAE-cellulose (Whatman DE52) equilibrated with 10 mM Tris-HCl (pH 7.6). Elution with 400 mM Tris-HCl buffer (pH 7.6) yielded the cytochrome fraction containing both cytochromes from D. desulfuricans G201 and expressed cytochrome c3 (Mr 26,000) which were then applied to a hydroxyapatite (Bio-Rad) column equilibrated with 400 mM Tris-HCl (pH 7.6). The cytochrome c3 (Mr 26,000) containing fraction was eluted with 200 mM phosphate buffer (pH 7.6). After overnight dialysis against distilled water, this fraction was absorbed on a DEAE-cellulose column and eluted using a gradient from 50 mM Tris-HCl, 50 mM NaCl buffer (pH 7.6) to 50 mM Tris-HCl, 200 mM NaCl buffer (pH 7.6). Pure wild-type or mutant cytochrome c3 (Mr 26,000) was obtained in the fraction eluting at 50 mM Tris-HCl, 200 mM NaCl (pH 7.6). The purity of the samples was analyzed by SDS-polyacrylamide gel electrophoresis (PhastSystem, Pharmacia) and by amino acid analyses (Beckman amino acid analyzer system 6300).
Protein SequencingNH2 terminus sequence determinations were performed with an Applied Biosystem A470 gas-phase sequencer. Quantitative determination of phenylthiohydantoin-derivatives was done by high pressure liquid chromatography (Waters) monitored by a data and chromatography control station (Waters Model 840).
ElectrochemistryDetermination of the individual oxidoreduction potentials of the heme groups of the wild-type and mutant cytochromes was performed as described previously (19).
Isoelectric Point DeterminationThe isoelectric point of the protein was measured by isoelectric focusing using a PhastSystem apparatus from Pharmacia Biotech Inc. Phast gel IEF 3-9, which operates in the 3-9 pH range, and morpholine polyacrylamide gel plates from Pharmacia (pH ranging from 3.5 to 9.5) were used together with a Pharmacia broad-range pI calibration kit.
Analysis of the Interaction between Both Wild-type and Y73E Cytochromes c3 (Mr 26,000) and HydrogenaseThe [NiFeSe] hydrogenase was purified from D. desulfuricans Norway cells as described previously (26). Interaction between periplasmic [NiFeSe] hydrogenase from D. desulfuricans Norway and either wild-type or Y73E cytochromes was investigated with a biomolecular interaction analysis biosensor-based analytical system (BIAcore, Pharmacia). All experiments were performed at 25 °C. Hydrogenase was immobilized on a Sensor chip CM5 (Pharmacia biosensor) through amine coupling. Hydrogenase was injected for 7 min (10 µl/min) resulting in approximately 3000 resonance units of immobilized protein. A continuous flow cell was equilibrated with HBS buffer (10 mM Hepes (pH 7.4), 150 mM NaCl, 3.4 mM EDTA, 0.005% surfactant P20). Both cytochromes were diluted in 50 mM Hepes buffer (pH 7.5) and injected using a flow rate of 30 µl/min (150 µl).
The complete
amino acid sequence of cytochrome c3
(Mr 26,000) from D. desulfuricans
Norway (23) enabled design of several degenerate oligonucleotides to
clone its structural gene. Among these, probe 33-44D, corresponding to
amino acid sequence 33 to 44 including the first heme-binding site, was
the most suitable. Genomic DNA was digested with several restriction
enzymes and blotted onto positively charged membrane. A 1.5-kb
BamHI fragment which hybridized with the labeled 33-44D
probe was selected for cloning. The same probe was used to screen
recombinant plasmids for the presence of the gene. The screening of
approximately 200 recombinant plasmids yielded one positive clone
called pUCC19, a restriction map of which is shown in Fig.
1. Further hybridization analyses and sequencing of this
fragment revealed that the structural gene of cytochrome
c3 (Mr 26,000), named
cycD, was present in a 877-bp
SalI-BamHI fragment. Its nucleotide sequence was
obtained from both strands and translated into protein (Fig. 1). The
mature cytochrome sequence, deduced from the nucleotide sequence 371 to
703, perfectly matches the amino acid sequence previously published (23). A putative promoter (the 35 sequence TTGCCC at nucleotides 171 and the
10 sequence TATAAA at nucleotides 192) has been selected as
the best fit to the E. coli
70 consensus
promoter sequences (27). A possible ribosome-binding site (AGGAGG) is
also present at nucleotide 291. A potential hairpin loop transcription
termination signal, 8 nucleotides downstream from the stop codon,
suggests that the cycD gene is not co-transcribed with other
open reading frames. To confirm this hypothesis, a 1.5-kb
EcoRV-SacI DNA fragment downstream from the
cycD gene was cloned by hybridization with the
oligonucleotide 48-55COMP, derived from nucleotides 513-532 (Fig. 1).
Sequence analyses of this fragment revealed no open reading frames
downstream from the cycD gene, further indicating the
cycD gene is monocistronic. Moreover, Northern blot analyses
revealed only one hybridization band when an intragenic cycD
fragment was used as probe against total RNA isolated either from
D. desulfuricans Norway (Fig. 2, lane
B) or from D. desulfuricans G201(pCACC3) (Fig. 2,
lane A). No hybridization signal was visible when total RNA
from D. desulfuricans G201(pJRD215) was probed in the same
conditions (Fig. 2, lane C). The length of the specific cycD mRNA was determined to be about 550 bp, in complete
agreement with the sequence data, confirming that the cycD
gene is monocistronic.
The nucleotide sequence clearly shows that the mature cytochrome sequence is preceded by an NH2 terminus sequence of 24 residues that could initiate at nucleotide 299. Computer analysis predicts a signal peptide with a cleavage site between residues 24 and 25, which is in agreement with the NH2 terminus sequence of the mature protein. This signal sequence participates in the export of the cytochrome from the cytoplasm to the periplasm across the cytoplasmic membrane. Comparison of signal sequence of cytochrome c3 (Mr 26,000) with those of four other Desulfovibrio cytochromes whose genes have been cloned (31, 34, 39, 40) revealed no sequence homologies (not shown). However, they all exhibit typical features of a signal peptide with a positively charged NH2 terminus followed by a hydrophobic region.
Expression of the cycD Gene in D. desulfuricans G201As attempts to express polyheme cytochromes in E. coli have been unsuccessful, the cycD gene in the broad-host range vector pJRD215 was introduced into another sulfate-reducing bacterium, D. desulfuricans G201, by transconjugation with E. coli. By expressing the plasmid-encoded cycD gene in D. desulfuricans G201(pCACC3), it was possible to purify about 90 mg of cytochrome c3 (Mr 26,000) per kg of wet cells, which is four times more than the amount obtained from the endogenous D. desulfuricans Norway chromosomal cycD gene. The cytochrome was purified from the periplasmic extract of D. desulfuricans G201(pCACC3). No mature cytochrome c3 (Mr 26,000) was detected in the cytoplasmic fraction, indicating that its signal effectively exports the molecule across the cytoplasmic membrane. The NH2 terminus sequence was identical to that of the protein purified from D. desulfuricans Norway, showing that the signal sequence was correctly cleaved in D. desulfuricans G201 during the transport of the protein to the periplasm (Table II).
|
The midpoint redox potentials of the four hemes were determined by electrochemistry; no significant differences were observed between the values measured in the case of the cytochrome c3 (Mr 26,000) expressed in D. desulfuricans G201 and that of the cytochrome c3 (Mr 26,000) isolated from D. desulfuricans Norway. In the same way, the isoelectric point of the molecules was the same (Table II) as were the molar extinction coefficients at 553 nm in the reduced state (data not shown). Therefore, cytochrome c3 (Mr 26,000) overexpressed in D. desulfuricans G201 is indistinguishable from the molecule produced by D. desulfuricans Norway with regards to several biochemical and biophysical properties, suggesting that the folding of the two proteins is the same.
Characteristic of the Y73E Cytochrome MutantThe Y73E cytochrome c3 (Mr 26,000) was overexpressed and purified from the D. desulfuricans G201 (pCC3Y73E) cells as was done for the wild-type cytochrome. The NH2 terminus sequence of the Y73E cytochrome was identical to that of the wild-type molecule, indicating then that the mutant also was correctly cleaved during its export in the periplasm. Reconstructed mass spectra of Y73E cytochrome gave only one peak corresponding to the molecular mass of 14,950 (±2) Da for the subunit. This was consistent with the tyrosine/glutamate replacement and the insertion of four heme groups per subunit. No differences in the isoelectric point were noticed between the wild-type and Y73E cytochromes c3, each exhibiting an acidic character (Table II). Both proteins (wild-type and mutant) were purified as dimers; the wild-type molecule is also isolated as a homodimer when it is purified from D. desulfuricans Norway. Thus, the Y73E mutation did not affect the dimerization process.
The macroscopic oxidoreduction potentials, deduced from the simulation of the experimental cyclic voltamograms are summarized in Table II. It is apparent that all redox potential values were affected by the single mutation, yielding a global increase of about 50 mV.
Tyrosine 73 in the cytochrome molecule is in close contact with the sixth axial ligand to the heme 4 iron with its plane parallel to the histidine 77 plane (5). Effects of the mutation on the relative orientation of the histidine ligands to heme 4 iron was investigated by EPR. A ligand-field model has been proposed to correlate the gz values to the dihedral angles between the planes of the two histidine ligands (28). As gz values increase when the ligand field becomes more axial, the largest values are expected when the dihedral angle tends toward 90°. However, no significant differences are observed when the EPR spectra of both wild-type and mutant cytochromes are compared (data not shown), and especially in the region of the highest gz values. This suggests that the histidine axial ligands to the heme 4 iron did not adopt an almost perpendicular orientation as in the case of D. desulfuricans Norway tetraheme cytochrome c3 (Mr 13,000). We cannot, however, rule out that the mutation induces a change in the dihedral angle between the planes of the two histidines as dihedral angles up to 41° do not give rise to a strong gz signal in the EPR spectrum (28).
Interaction between Hydrogenase and Either Wild-type or Y73E Cytochrome c3It has been reported that periplasmic
[NiFeSe] hydrogenase from D. desulfuricans Norway was able
to reduce cytochrome c3
(Mr 26,000) (29) as well as cytochrome
c3 (Mr 13,000) (30). We investigated the influence of the Y73E replacement on the interaction with the [NiFeSe] hydrogenase using the BIAcore technology. Fig. 3 shows typical response curves of association and
dissociation of the two proteins: wild-type cytochrome
c3 (Mr 26,000) binds to
the immobilized hydrogenase and reaches a stable level but when only
buffer was injected, dissociation of wild-type cytochrome with
hydrogenase was detected (Fig. 3A). In contrast, no binding of the Y73E cytochrome with immobilized hydrogenase was observed on the
sensorgram using the same experimental conditions (Fig. 3B).
As a control, no binding was detected between either wild-type or
mutant cytochrome and dextran surface at pH 7.4 in the experimental conditions described above, which is consistent with the acidic property of the cytochromes.
Among the multiheme proteins, the homodimer cytochrome c3 (Mr 26,000) is the least thoroughly studied but possesses several structural features that might provide insights into the structure-function relationships of this class of metalloproteins. Here we have reported the influence of replacement of an invariant aromatic residue on the physicochemical properties and activity of cytochrome c3 (Mr 26,000) from D. desulfuricans Norway.
Thus, the D. desulfuricans Norway cytochrome c3 (Mr 26,000) gene, named cycD gene, was cloned and overexpressed in D. desulfuricans G201. Nucleotide sequencing and Northern analysis revealed that the gene is monocistronic as it is the case for the cyc gene encoding cytochrome c3 (Mr 13,000) from D. vulgaris Hildenborough (31). This situation is, however, opposite to that found for the hmc gene from D. vulgaris Hildenborough which belongs to an operon encoding seven other open reading frames (32).
Sequence analysis of the gene also shows that the protein is synthesized as a precursor of 135 amino acids. This form differs from the purified cytochrome c3 (Mr 26,000) in the NH2 terminus sequence, which included 24 additional amino acids. The 877-bp SalI-BamHI fragment, including the cycD gene, inserted into the broad-host range vector pJRD215, allows the heterologous expression of the cytochrome c3 (Mr 26,000) in D. desulfuricans G201. The mature form of the protein is found in the periplasm indicating that this extra NH2 terminus sequence acts as a signal sequence, targeting the protein to the periplasmic space through the cytoplasmic membrane before being removed by a signal peptidase. Although it has the typical features of a signal peptide used for the export of proteins to the periplasm, the signal peptide cleavage site of cytochrome c3 (Mr 26,000) as in Hmc (32) was not of the traditional form A/A. The unusual cleavage site may explain why Le Gall and Peck (33), on the basis of the NH2 terminus sequence, postulated that cytochrome c3 (Mr 26,000) of D. desulfuricans Norway would be a cytoplasmic protein. Their hypothesis was supported by the observation that cytochrome c3 (Mr 26,000) from D. gigas reacts, in crude extract, with a cytoplasmic thiosulfate reductase (1). Our data show, however, that cytochrome c3 (Mr 26,000) from D. desulfuricans Norway is obviously a periplasmic protein.
Our studies confirm that D. desulfuricans G201 is an appropriate organism to express polyheme cytochromes from sulfate reducing bacteria as tetraheme (21), octaheme (this work), and decahexaheme (34) cytochromes are correctly folded and matured in this bacterium. The relatively high rate of expression allows protein engineering studies. Thus, the cytochrome c3 (Mr 26,000) tyrosine 73 has been then replaced by a glutamate residue. Expression in D. desulfuricans G201 allowed us to obtain a mutant molecule in which the four heme groups per subunit still are bound to the apoprotein. All of the physicochemical parameters that we measured suggest the Y73E cytochrome c3 folds like wild-type cytochrome and that the mutation does not induce drastic structural changes.
When the relative orientation of the planes of the histidine axial ligands to the heme iron are compared within several cytochromes c3, stronger similarities were found with cytochromes c3 (Mr 13,000) from D. gigas, D. vulgaris Hildenborough, Miyazaki and cytochrome c3 (Mr 26,000) from D. desulfuricans Norway than with cytochrome c3 (Mr 13,000) from D. desulfuricans Norway. In the former case, the plane angles of the histidine axial ligands to the heme 4 iron are close to the parallel orientation while 77.2° is found in the latter case. Tyrosine 73 is present in almost all cytochromes c3 and is parallel to the histidine plane of the sixth axial ligand to the heme 4 iron atom. In the D. desulfuricans Norway cytochrome c3 (Mr 13,000), it is, however, replaced by a glutamate residue. It has been proposed that this replacement was responsible for the different orientation of the axial ligands to the heme 4 iron (17). Our data suggest that the Y73E replacement in the cytochrome c3 (Mr 26,000) does not induce a perpendicular rearrangement of the histidine axial ligands as is observed in the D. desulfuricans Norway tetraheme cytochrome c3. Other factors must then be necessary to maintain the relative orientation of the two histidines. This can be related to experiments on chemical model compounds showing that coplanar orientation of the histidine axial ligands to the porphyrin iron is the most thermodynamically favorable case (28). We cannot, however, rule out the possibility that the mutation has induced a smaller rearrangement of the axial ligands to the heme 4 iron. A structural analysis is therefore necessary to accurately determine the orientation of the histidine planes in the cytochrome mutant.
Because of the location of tyrosine 73 in the close proximity of heme 4, a change of the redox potential of that heme group was expected in the cytochrome mutant. The dielectric constant of the heme crevice is a primary determinant in setting the redox potential of the heme group (35). Two essential factors affecting the heme crevice dielectric constant are the solvent exposure of the heme moiety and the polarity of the amino acids in close proximity to this group. In this regard, replacement of the aromatic residue Tyr73 by a charged one is expected to modify the polarity of the heme crevice. In addition to that effect, the Y73E mutation should modify the hydrogen bond network around the axial ligand to heme 4 and then disturb the electronic distribution of the heme group. All these effects may account for the observed modification in the redox potential values measured in the cytochrome mutant. By examining a wide range of side chain replacement for Phe82, It has been proposed that the contribution of this aromatic residue to the redox potential of the wild-type yeast iso-1 cytochrome c is of +40 mV (36). The global increase of about 50 mV, measured in the case of the cytochrome c3 mutant is in the same order of magnitude. Because of the close redox potential values and electronic interactions between the heme groups, the macroscopic redox potentials of the cytochrome c3 mutant cannot, however, be assigned to those of the wild-type molecule. It appears that even if the Y73E mutation is closer to heme 4 than to the three other hemes, all redox potential values are affected by the mutation. One explaining factor is the very compact structure of the heme core, the largest iron-iron distance being 17.7 Å (17). Perturbation of heme 4, because of the Y73E replacement, might have repercussions on the redox properties of the three other heme groups.
Numerous studies on cytochromes c3 (Mr 13,000) indicate that heme 4 is most likely to be the heme involved in the electron exchange with the various redox partners. Our results show that cytochrome c3 (Mr 26,000) is able to bind specifically to the [NiFeSe] hydrogenase from the same organism to form a complex. The Y73E mutation clearly affects the interaction process with the hydrogenase indicating that heme 4 is the electron transfer heme group in the dimeric cytochrome c3 (Mr 26,000) with regards to hydrogenase. Our data also show that the single Y73E replacement has a drastic influence on the interaction process with the redox partner. The same kind of observation was made in the case of D. vulgaris Hildenborough cytochrome c3 (Mr 13,000). Replacement of the histidine, sixth axial ligand to the heme 4 iron by a methionine hindered the interaction with the [Fe] hydrogenase from the same organism while lower effects were measured when other heme groups were altered (19). Slight alteration of the heme 4 environment has then a strong influence on the interaction with the redox partner and on the electron transfer properties.
More experiments must be done to determine the thermodynamic parameters of the complex formation using either wild-type or mutant cytochromes c3 (Mr 26,000). Kinetic studies should quantify the influence of the mutation on the electron transfer rate. The data obtained will be analyzed on the basis of three-dimensional structure of the Y73E cytochrome (in progress), which will reveal fine structure changes induced by the mutation. The influence of the replacement of this aromatic residue in the stability of the molecule will also be analyzed.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U68390[GenBank].
We thank Drs. J. Haladjian and P. Bianco for the electrochemistry experiments, Dr. E. Forest for the MS Analysis, Dr. B. Guigliarelli for recording the EPR spectra, Dr. M. T. Giudici-Orticoni for the BIAcore experiments and fruitful discussions as well as Dr. M. Czjzek for the three-dimensional analyses.