High-Molecular-Weight Cytochrome c from Desulfovibrio - Journal of ...

3 downloads 0 Views 2MB Size Report
Desulfovibrio species: D. vulgaris Brockhurst Hill, Hilden- borough, Monticello ..... Miyazaki and Hildenborough Hmcs are further compared below. Hmc protein .... Pollock, W. B. R., P. J. Chemerika, M. E. Forrest, J. T. Beatty, and G. Voordouw.
JOURNAL OF BACTERIOLOGY, Jan. 1991, p. 220-228

Vol. 173, No. 1

0021-9193/91/010220-09$02.00/0 Copyright ©D 1991, American Society for Microbiology

Cloning, Sequencing, and Expression of the Gene Encoding the High-Molecular-Weight Cytochrome c from Desulfovibrio vulgaris Hildenborough W. BRENT R. POLLOCK,1 MOHAMED LOUTFI,2 MIREILLE BRUSCHI,2 BARBARA J. RAPP-GILES,3 JUDY D. WALL,3 AND GERRIT VOORDOUW1*

Division of Biochemistry, Department of Biological Sciences, The University of Calgary, Calgary, Alberta, Canada T2N IN41; Laboratoire de Chimie Bacterienne, Centre National de la Recherche Scientifique, 13277 Marseille Cedex 9, France2; and Departmcnt of Biochemistry, University of Missouri, Columbia, Missouri3 Received 26 July 1990/Accepted 16 October 1990

By using a synthetic deoxyoligonucleotide probe designed to recognize the structural gene for cytochrome cc3 from Desulfovibrio vulgaris Hildenborough, a 3.7-kb Xhol genomic DNA fragment containing the cc3 gene was isolated. The gene encodes a precursor polypeptide of 58.9 kDa, with an NH2-terminal signal sequence of 31 residues. The mature polypeptide (55.7 kDa) has 16 heme binding sites of the form C-X-X-C-H. Covalent binding of heme to these 16 sites gives a holoprotein of 65.5 kDa with properties similar to those of the high-molecular-weight cytochrome c (Hmc) isolated from the same strain by Higuchi et al. (Y. Higuchi, K. Inaka, N. Yasuoka, and T. Yagi, Biochim. Biophys. Acta 911:341-348, 1987). Since the data indicate that and Hmc are the same protein, the gene has been named hmc. The Hmc polypeptide contains cytochrome 31 histidinyl residues, 16 of which are integral to heme binding sites. Thus, only 15 of the 16 hemes can have bis-histidinyl coordination. A comparison of the arrangement of heme binding sites and coordinated histidines in the amino acid sequences of cytochrome c3 and Hmc from D. vulgaris Hildenborough suggests that the latter contains three cytochrome c3-like domains. Cloning of the D. vulgaris Hildenborough hmc gene into the broad-host-range vector pJRD215 and subsequent conjugational transfer of the recombinant plasmid into D. desulfuricans G200 led to expression of a periplasmic Hmc gene product with covalently bound hemes. cc3

c-type cytochromes are distinct from other classes of cytochrome in that they are covalently bound to their hemes by ac-thioether bonds (7). These bonds arise from the Markovnikov addition of sulfhydryl groups to the vinyl groups of the heme. The sulfhydryl groups are derived from the two cysteinyl residues present in the heme binding sequence C-X-X-C-H. The histidinyl residue in this sequence occupies the fifth coordination position of the heme Fe atom, while the sixth coordination position is occupied by either a histidinyl or a methionyl residue. The nature of the ligand occupying this position has a direct effect on the redox potential of the heme (21). Desulfovibrio vulgaris contains at least four different c-type cytochromes: cytochrome C553 (23, 32), cytochrome C3 (14, 35), cytochrome CC3 (18), and high-molecular-weight cytochrome c (Hmc) (13, 38). Cytochrome C553 is the only Desulfovibrio cytochrome reported to be a monoheme and to have a His-Met-coordinated heme Fe atom; the other three cytochromes have more than one heme, all of which are thought to have His-His coordination. Complete amino acid sequence data and three-dimensional structure information are required to ascertain the exact number and organization of the c-type hemes in each of these cytochromes. A determination of the cellular localization of these cytochromes is also important, since Desulfovibrio spp. contain both a periplasmic and a cytoplasmic chain of redox carriers (16, 27). This information can be obtained by sequencing the structural genes for these cytochromes. At present, only the D. vulgaris subsp. vulgaris Hildenborough (hereafter referred to as D. vulgaris Hildenborough) cy-

*

tochromes C553 (32) and C3 (35) structural genes have been sequenced. For the two remaining cytochromes, partial amino acid sequence data are available for cytochrome CC3 (18), and the isolation and amino acid composition for Hmc have been reported (13). This work has now been extended to D. vulgaris Hildenborough cytochrome CC3, the gene for which has been cloned and sequenced. Perhaps surprisingly, it appears that cytochrome CC3 and Hmc are the same protein, and therefore this work completes the characterization of genes for known periplasmic c-type cytochromes in D. vulgaris. MATERIALS AND METHODS

Strains, vectors, and media. The bacterial strains and plasmids used are described in Table 1. Desulfovibrio strains were cultured for 18 h at 37°C in stoppered bottles containing 100 ml of Postgate C medium (27) made anaerobic by flushing with N2. The inoculum consisted of 3 ml of a Postgate B (27) culture that had been grown under the same conditions. Growth of Escherichia coli TG2 in TY medium was carried out as previously described (26). E. coli DH5ao and Desulfovibrio desulfuricans G200, used for conjugational transfer of broad-host-range plasmids, were cultured as described elsewhere (35a). Biochemical reagents. All enzymes were obtained from either Pharmacia, Inc., or Boehringer Mannheim Biochemicals. The radiochemicals [a-35S]dATP (400 Ci mmol-1; 10 mCi ml-1), [k-32P]dATP (3,000 Ci mmol-P; 10 mCi ml-'), and [y-32P]dATP (3,000 Ci mmol-'; 10 mCi ml-') were purchased from Amersham Corp. and were used for dideoxynucleotide sequencing, nick translation, and 5' end labeling, respectively. Ficoll 400 was purchased from Pharmacia.

Corresponding author. 220

D. VULGARIS HILDENBOROUGH hmc GENE

VOL. 173 1991

221

TABLE 1. Bacterial strains and DNA vectors Genotype, comments, and reference

Strain or vector

NCIMB 8303; isolated from clay soil near Hildenborough, Kent, United Kingdom (27); source of the hmc gene, encoding high-molecular-weight cytochrome c Spontaneous Nalr mutant of D. desulfuricans G100A (37) D. desulfuricans G200 A(lac-pro) supE thi hsdM hsdR recA(F' traD36 proAB+ lacZAM15P); from T. J. Gibson E. coli TG2a +80d lacZAM15 endAI recAl hsdRJ7 (rK- MK+) supE44 thi-J X- gyrA96 relA1[F' A (lacZAE. coli DH5ot argF)U169] (12) 33 pUC8 Contains the hmc gene of D. vulgaris Hildenborough on a 3.7-kb XhoI insert in pUC8; this study pP6A IncQ group, broad-host-range cloning vector; Kmr Smr (5) pJRD215 Contains the D. vulgaris Hildenborough hmc gene on the 2.2-kb SmaI fragment of pP6A in pBPHMC-1 pJRD215; this study. a Constructed from E. coli JM101 by T. J. Gibson and M. D. Biggin at the Laboratory of Molecular Biology, MRC Centre, Cambridge, United Kingdom.

D. vulgaris Hildenborough

Polyvinylpyrrolidone (molecular weight, 40,000), bovine serum albumin (fraction V), molecular biology-grade sodium dodecyl sulfate (SDS), and salmon sperm DNA (sodium salt) were purchased from Sigma Chemical Co. Low- and highgelling-temperature (LGT and HGT) agarose were obtained from Bethesda Research Laboratories, Inc. Nitrocellulose and Hybond-N hybridization transfer membranes were acquired from Schleicher & Schuell, Inc., and Amersham, respectively. Acrylamide and N,N'-methylenebisacrylamide were purchased from BDH Ltd. Low-molecular-weight protein electrophoresis markers were obtained from Pharmacia. Dialysis tubing (molecular size cutoff, 6 to 8 kDa) was obtained from Spectrum Medical Industries, Inc. All other reagent-grade chemicals were obtained from either Sigma or Fisher Scientific Co. Purification and peptide sequencing of cytochrome CC3. Cytochrome CC3 was purified from French-pressed D. vulgaris Hildenborough cells essentially as described by Loutfi et al. (18), who referred to this protein as cytochrome C3 (Mr, 26,000). In addition, a new purification procedure was used (not shown) in which this cytochrome was isolated from the periplasmic fraction of D. vulgaris Hildenborough prepared as described elsewhere (31). The yields of both procedures were approximately 1 mg/100 g of wet cells. These isolated proteins were pure as judged by SDS-polyacrylamide gel electrophoresis (PAGE) (not shown), which indicated molecular sizes of 62 and 56 kDa for the holoprotein and apoprotein, respectively. The 553 of the reduced holoprotein was found to be 424 mM-' cm-', a value similar to that of Higuchi et al. (13), who showed the presence of 16.3 hemes per polypeptide. The carboxymethylated apoprotein was prepared as previously described (18) and treated variously with CNBr, Asp-N endopeptidase, Arg-C endopeptidase, or Glu-C endopeptidase. Automated amino acid sequencing (18) of isolated peptides yielded 47 peptides which could be organized into 12 polypeptides as indicated in the legend to Fig. 2. The

l

H

1

K

H

hmc

phmc

stretches of Hmc amino acid sequence covered by these peptide groups are indicated in Fig. 2. Gene cloning. DNA manipulations were carried out as described by Maniatis et al. (19). Samples (10 ,ug) of D. vulgaris Hildenborough genomic DNA were digested with various restriction endonucleases, electrophoresed on a 0.7% (wt/vol) HGT agarose gel as described elsewhere (36), and blotted onto a nitrocellulose membrane. Following prehybridization for 15 min at 68°C in 0.2% SDS-10x Denhardt solution (19) and 6x SSC (19), the blot was incubated with the 5'-end-labeled probe P43 for 16 h at 68°C in 6x SSC. The sequence of the deoxyoligonucleotide probe P43, a 47-mer, was 5'-AAGATCGA(AG)AAGCC(CG)GC(CG)AACAC(CG) GC(CG)TGCGT(CG)GACTGCCACAAGGA. This probe was designed to recognize the COOH-terminal region of the structural gene encoding the sequence K-I-E-K-P-A-N-T-AC-V-D-C-H-K-E. The bias toward G or C in the third position of each codon was chosen on the basis of the known codon usage in the nucleotide sequences of the hydA,B (34) and cyc (35) genes from D. vulgaris Hildenborough. After incubation, the blot was washed thrice with 2x SSC, dried, and autoradiographed. Autoradiography (not shown) revealed a 3.7-kb XhoI fragment of chromosomal DNA to be the one most suitable for cloning. This was accomplished by size fractionating chromosomal DNA digested with XhoI on 1% (wt/vol) LGT agarose and ligating the 3- to 4.5-kb fraction into the vector pUC8, which had been digested with calf alkaline phosphatase and the restriction endonuclease Sall. E. coli TG2 cells, made competent by CaCl2 treatment, were transformed with the ligation mixture and spread onto TY plates containing isopropylindolyl-,-D-thiogalactopyranoside (IPTG), 5-bromo-4-chloro-3-indolyl-,3-D-galactopyranoside (X-Gal), and ampicillin. Southern blots of plasmids isolated from white recombinant colonies and digested with both EcoRI and HindlIl were probed with P43 as described above. A restriction map of the 3.7-kb insert of the only positive plasmid obtained (pP6A) is shown in Fig. 1.

T

s

L

500 bp

x

I

FIG. 1. Restriction map of the 3.7-kb XhoI fragment of D. vulgaris Hildenborough chromosomal DNA containing the hmc gene (boxed region). The restriction sites for XhoI (X), HindlIl (H), KpnI (K), and SmaI (S) are noted. The position of an E. coli-like u70 promoter sequence is shown (Phmc); the arrow indicates the direction of transcription.

222

POLLOCK ET AL.

Shotgun nucleotide sequencing. The procedure followed for shotgun nucleotide sequencing was essentially that outlined by Bankier and Barrell (1). Plasmid pP6A (1 pRg RI-') was sonicated, and the resulting fragments were end repaired and size fractionated as described above. The 0.3- to 0.6-kb fraction was isolated from the gel and ligated into the replicative form of M13mp8 digested with SmaI and calf alkaline phosphatase. Competent E. coli TG2 cells were transfected with the ligation mixture and spread onto TY plates with top agar containing IPTG and X-Gal. Singlestranded DNA was isolated from white recombinant phage plaques by using a 1.5-ml minipreparation procedure (1). Purified phage DNAs (2 Rd) were dot blotted on a Hybond-N hybridization transfer membrane and fixed by exposure to UV light. Following digestion with EcoRI and PstI, the 3.7-kb insert of plasmid pP6A was isolated by 1% (wt/vol) LGT agarose gel electrophoresis, radiolabeled by nick translation (19), and denatured by boiling. The filter was prehybridized in 6x SSC-lOx Denhardt solution-0.5% (wt/vol) SDS for 1 h at 68°C. The 3.7-kb DNA probe was then added to the prehybridization solution, and hybridization was continued for 16 h at 68°C. After hybridization, the filter was washed twice in 2x SSC for 1 min at room temperature, followed by a 1-h wash at 68°C in 6x SSC-0.5% (wt/vol) SDS. The dried filter was autoradiographed, and positive clones were sequenced by the method of Sanger et al. (28) as detailed by Bankier and Barrell (1). The sequencing data were processed and analyzed by using the programs of Staden and McLachlan (29, 30). Broad-host-range plasmid construction and bacterial conjugation. Plasmid pP6A was digested with SmaI and size fractionated as described above. The 2.2-kb fragment was isolated from the gel and ligated to the vector pJRD215, which had been digested with Stul and calf alkaline phosphatase. E. coli TG2 was transformed with the ligation mixture and spread onto TY-kanamycin plates. Recombinant plasmids were purified and analyzed by restriction digestion with BamHI, resulting in excision of the hmc gene as a 2.6-kb BamHI fragment (not shown). The recombinant plasmid and the vector alone were each transformed into E. coli DH5a and subsequently transferred to D. desulfuricans G200 as described elsewhere (35a). Total (chromosomal and plasmid) DNA of the recombinant exconjugant was prepared as described by Marmur (20), digested with BamHI, and subjected to Southern analysis (not shown). The 2.2-kb SmaI fragment derived from plasmid pP6A was radiolabeled by nick translation and used as the probe to verify the presence of plasmid pBPHMC-1 within the recombinant exconjugant D. desulfuricans G200

(pBPHMC-1). SDS-PAGE and fluorophotography. Desulfovibrio periplasmic proteins were isolated essentially as described by van der Westen et al. (31). Cultures of D. desulfuricans G200(pJRD215), D. desulfuricans G200(pBPHMC-1), and D. vulgaris Hildenborough grown in 100 ml of Postgate C medium were harvested by centrifugation at 10,000 x g for 10 min at 4°C. Each pellet was resuspended in 2 ml of H20 and transferred to three separate 15-ml glass centrifuge tubes, to each of which was added 2 ml of 0.1 M Tris hydrochloride-0.1 M EDTA (pH 9.0). These suspensions were heated to 30°C in a 60°C water bath and then shaken for 15 min at room temperature, after which they were centrifuged for 10 min at 10,000 x g. The supernatant, which contained the periplasmic proteins, was dialyzed for 20 h against two 1,000-ml volumes of cold deionized H20. For SDS-PAGE, 1-ml aliquots of each dialysate were

J. BACTERIOL.

placed into microfuge tubes and dried by vacuum centrifugation, resuspended in 100 IL of H20, 100 p,l of SDS incubation buffer (26), and 20 ,ul of 2-mercaptoethanol, and placed in a boiling water bath for 10 min. SDS-PAGE and fluorophotography were carried out as previously described (26) except that the acrylamide gradient was 10 to 15% and the slab gel was photographed by using Polaroid type 55 film and an exposure time of 10 min. Nucleotide sequence accession number. The sequence data presented in this paper have been submitted to GenBank and assigned accession number M34607. RESULTS

Cloning. Deoxyoligonucleotide P43, which was designed recognize the COOH-terminal sequence of cytochrome CC3, was used to screen purified recombinant plasmids for the presence of this gene. Screening of approximately 140 plasmids yielded one positive plasmid, pP6A. A restriction map for the 3.7-kb XhoI insert of pP6A is shown in Fig. 1. hmc coding region. The sequence of the first 2,160 nucleotides (nt) of the pP6A insert is shown in Fig. 2. The region containing the structural gene for the cytochrome (nt 472 to 2106) was obtained on both strands and has been translated into protein. The NH2-terminal sequence found for cytochrome CC3 (18) by protein sequencing (peptide P1: KALP...) is present at nt 565. It is clear from the nucleotide sequence (Fig. 2) that this sequence is preceded by an NH2-terminal signal sequence of 31 residues which initiates at nt 472. The gene-translated protein sequence confirmed the directly determined sequences for the 12 polypeptides, which together span 85% of the mature amino acid sequence (Fig. 2). The region of the structural gene corresponding to the mature cytochrome sequence (nt 565 to 2106) encodes an apoprotein of 55.7 kDa. This sequence includes 16 C-X-Xto

C-H heme binding sites, and the holoprotein, with 16 covalently bound hemes, has a calculated molecular size of 65.5 kDa. As discussed in detail below, the properties of the cytochrome defined by the nucleotide sequence depicted in Fig. 2 are reminiscent of those of the Hmc first correctly described by Higuchi et al. (13). Therefore, this gene will be referred to as the hmc gene, and its product will be referred to as Hmc.

The Hmc polypeptide contains 31 His residues, 16 of which are integral to the 16 C-X-X-C-H heme binding sites. Thus, only 15 of the 16 hemes can have bis-histidinyl coordination. The heme binding sites and histidinyl residues of the Hmc polypeptide are spatially arranged into three complete cytochrome c3-like domains (II, III, and IV) and one incomplete domain (I) that lacks both the second heme binding site and its corresponding sixth-coordinate-position histidine (Fig. 3). Visible spectroscopic data on Hmc (13) do not disallow a Met residue at this position. The E695 value for His-Met-coordinated c-type cytochromes is typically on the order of 0.9 mM-1 cm-', as determined for cytochromes c (9) and C553 (23, 38). Higuchi et al. (13) determined an E553.2 value of 428 mM-1 cm-' for Hmc based on a molecular size of 75 kDa, which corresponds to an E553 2 value of 374 mM-1 cm-1 when adjusted to the correct molecular size of 65.5 kDa. Applying this adjusted E553.2 value to the visible spectrum of D. vulgaris Hildenborough Hmc shown by Higuchi et al. (A5532 = 0.14; concentration of Hmc, 370 nM) gives an A695 on the order of 0.0003. Thus, if there is a His-Metcoordinated heme in Hmc, it has remained undetected because the A695 is only 0.2% of the heme A5532 in Hmc. If

VOL. 173, 1991

D. VULGARIS HILDENBOROUGH hmc GENE

223

Xb.QI

CTCGAGCGATCCATACTTAACGGCAAGTTTACGAACATAGCATTTCCCGCGCCACCGGTCGGTGACCATTGAAGGTTGCACAGCAGAGGCCCGCGCCCCTTTCACCGCATCGACGTTCAT 10

20

50

40

30

60

70

80

90

100

110

120

-35

-10

GCCCGCAACCACACGTCCCCAAGGGAAAAGGGGGCAATTGCCACATGCGCTGCAATTGCATCTGCGCGTGTTTTTTCACTTGACGCTCZTTGACACAATAGACTACCCCCGCAGTAGGAlTT 130

140

180

170

160

150

190

200

210

220

230

240

TATTCCATACTATACAGGTAGGGTATGCGATGCTCCCCACACCACCGCCCGCCCGCCTGTCGTATCACGAGGACAGCCACGAAGGAATCAGGTTCCCCTCAGGAGGGGGCGACCGGGAAA 250 260 270 280 290 300 310 320 330 340 350 360 r b s

M

R

N

CCACCGTGGCACGTCCGTCTGGCGACGCGCTCCCCGCGGGGAAACGGGGGCGACGTCCGGCTGGGGATGTCTGAGGATGTGTATCACCCATTCTTCCAAAC-Z-&CTGCCACTATGAGGAAC 370 G

R

T

390

380 L

L

R

W

A

G

V

400 L

A

A

410 T

A

420 G

I

I

G

V

430 G

F

440 S

W

Q

G

T

450 T

K

A

460 L

P

E

470 G

P

G

E

480 K

R

A

GGAAGGACACTGCTGCGATGGGCGGGAGTGTTGGCTGCCACCGCCATCATCGGCGTCGGCGGTTTCTGGTCACAGGGGACGACCAAGGCCCTCCCCGAAGGGCCCGGCGAAAAACGCGCC 490

500

530

520

510

550

540

560

570

580

590

600

HindTIl

D

L

I

E

I

G

A

M

E

R

F

G

K

L

D

L

K

P

A

V

F

R

D

H

Q

H

T

T

A

V

T

G

M

G

K

D

C

A

A

C

GACCTGATCGAGATCGGGGCCATGGAACGCTTCGGGAAGCTTGATCTTCCCAAGGTGGCGTTCCGCCACGACCAGCACACCACCGCCGTGACGGGCATGGGCAAGGACTGCGCCGCCTGT 610 H

K

S

630

620 K

D

G

K

M

S

L

640 K

F

M

650 R

L

D

660

D

N

S

A

670 A

E

680

L

K

E

I

Y

690 H

A

N

700 C

I

G

710 C

H

T

D

720 L

A

K

CACAAGAGCAAGGACGGCAAGATGTCGCTGAAGTTCATGCGGCTTGACGACAACAGCGCCGCTGAACTCAAGGAAATCTACCACGCCAACTGTATCGGCTGTCATACCGACCTCGCCAAG 730

A

G

740

K

K

T

G

P

750 Q

D

A

760 E

C

R

770 S

C

H

780

N

P

K

P

790 S

A

800

A

S

S

K

W

810 E

G

I

820 F

D

K

830 S

L

H

Y

840 R

H

V

GCCGGTAAGAAGACCGGCCCGCAGGACGCCGAATGCCGCTCGTGCCACAACCCGAAGCCCTCTGCCGCCTCGTCGTGGAAGGAGATAGGCTTCGACAAGTCGCTGCACTACCGTCACGTG 850

A

S

K

860 A

I

K

P

880

870 V

G

D

P

Q

K

90v

890 N

C

G

A

C

H

H

910 V

Y

D

920 IiindIIII E A S

K

930 K

L

V

;40 W

G

K

950 N

K

E

D

960 S

C

R

GCGTCCAAGGCCATCAAGCCCGTGGGCGACCCGCAGAAGAACTGCGGCGCTTGCCACCACGTCTATGACGAAGCTTCCAAGAAGCTCGTATGGGGTAAGAACAAGGAAGACTCGTGCCGC 970 A

C

H

980 G

E

K

P

990 V

D

K

P

A

1020

1010

1000 R

L

D

T

A

A

H

T

1040

1030 A

C

I

S

C

H

M

1050 D

A

V

1060 K

T

K

1070 A

E

T

G

1080 P

V

N

GCCTGTCACGGCGAGAAGCCCGTGGACAAGCGTCCCGCGCTCGACACCGCCGCCCACACCGCGTGCATCAGCTGCCACATGGACGTCGCCAAGACCAAGGCCGAGACCGGCCCGGTCAAC 1090 C

A

G

1100 C

H

A

P

1110 E

Q

A

1130

1120 A

K

F

K

V

V

R

1140 E

V

P

1160

1150 R

L

D

R

G

Q

P

1170 D

A

A

1180 L

I

L

1190 P

V

P

G

1200 K

D

A

TGCGCCGGCTGCCACGCCCCCGAGGCGCAGGCCAAGTTCAAGGTGGTACGCGAAGTGCCCCGCCTCGACCGCGGTCAGCCCGACGCCGCCCTCATCCTGCCCGTACCCGGCAAGGATGCC 1210 P

R

E

1220 M

K

G

T

1230 M

K

P

1240 V

A

F

1260

1250 D

H

K

A

H

E

A

1270 K

A

N

1280 D

C

R

T

1290 C

H

H

1300 V

R

I

1320

1310 D

T

C

T

A

C

H

CCGCGCGAGATGAAGGGCACCATGAAGCCGGTCGCCTTCGACCACAAGGCCCACGAAGCCAAGGCCAACGACTGCCGCACCTGCCACCATGTGCGCATCGACACCTGCACCGCCTGCCAT 1330 T

V

1340

__Kpn I N G T

A

D

1360

1350 S

K

F

V

Q

L

1370 E

K

A

M

1380 H

Q

P

1390 D

S

M

1400 R

S

C

V

1410 G

C

H

1420 N

T

R

1430 V

Q

Q

P

1440 T

C

A

ACCGTGAACGGTACCGCAGACAGCAAGTTCGTCCAGCTTGAAAAGGCCATGCACCAGCCCGACTCCATGCGCAGCTGCGTGGGTTGTCACAACACCCGCGTCCAGCAGCCCACC.TGTGCC 1450 G

C

H

1460 G

F

I

K

1470 P

T

K

1480 S

D

A

1490 Q

C

G

V

1500 C

H

V

1510 A

A

P

1530

1520 G

F

D

A

K

Q V

1540 E

A

G

1550 A

L

L

N

1560 L

K

A

GGTTGCCACGGCTTCATCAAGCCCACGAAGAGCGATGCCCAGTGCGGCGTGTGCCACGTGGCCGCCCCCGGCTTCGACGCCAAGCAGGTCGAAGCGGGCGCATTGCTGAACCTCAAGGCA 1570 E

Q

R

1580 S

Q

V

A

1600

1590 A

S

M

L

S

A

1610 R

P

Q P

1630

1620 K

G

T

F

D

L

1640 N

D

I

P

1650 E

K

V

1660 V

I

G

1670 S

I

A

K

1680 E

Y

Q

GAGCAGCGTTCGCAGGTTGCCGCCTCCATGCTGTCGGCCCGTCCGCAGCCCAAGGGCACCTTCGACCTCAATGACATCCCCGAAAAGGTCGTCATCGGCTCCATCGCCAAGGAATACCAG 1700

1690 P

S

E

F

P

H

R

1720

1710 K

I

V

K

T

L

1730 I

A

G

I

1740 G

E

D

1760

1750 K

L

A

A

T

F

H

1770 I

E

K

1780 G

T

L

1790 C

Q

G

C

1800 H

H

N

CCCAGCGAGTTCCCGCACCGCAAGATCGTGAAGACGCTCATCGCCGGTATCGGTGAAGACAAGCTGGCCGCCACCTTCCACATCGAGAAGGGCACGCTGTGCCAGGGCTGCCACCACAAC 1820

1810

1840

1830

1850

1870

1860

1880

1890

1900

1910

1920

S P A S L T P P K C A S C H G K P F D A D R G D R P G L K A A Y H Q Q C M G C H AGCCCCGCCAGCCTCACCCCGCCCAAATGCGCGAGCTGCCACGGCAAGCCCTTCGACGCCGACAGGGGCGACCGTCCCGGTCTCAAGGCCGCCTACCACCAGCAGTGCATGGGCTGCCAC

1930 D

R

M

1940 K

I

E

K

1960

1950 P

A

N

T

A

C

1970 V

D

C

H

1980 K

E

R

1990 A

2000

2010

2020

2030

2040

K

GACCGGATGAAGATCGAAAAGCCGGCGAACACCGCCTGCGTCGATTGCCACAAGGAACGCGCGAAP ATAGGACGGTGAGGAGTACGACATGGATCGCAGAAGATTCCTGACCCTGCTGGGA 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160

p4 3

FIG. 2. Nucleic acid sequence of the first 2,160 nt of the 3.7-kb XhoI fragment of D. vulgaris Hildenborough chromosomal DNA depicted in Fig. 1. The -35 and -10 regions of a putative promoter sequence are indicated, as is the ribosome binding site (rbs). The hmc structural gene spans the region from nt 472 (initiator methionine residue) to nt 2109 (termination codon). A vertical arrow marks the signal peptide cleavage site. The overlined regions represent 436 of the 514 amino acid residues of the mature Hmc polypeptide which have been sequenced at the peptide level in polypeptides Pl to P12. The corresponding positions of these polypeptides in the gene-translated sequence are Pl, nt 565 to 921; P2, nt 1636 to 1971; P3, nt 1375 to 1530; P4, nt 955 to 1026; P5, nt 1162 to 1224; P6, nt 1315 to 1374; P7, nt 922 to 954; P8, nt 1069 to 1104; Pg, nt 1105 to 1119; Plo, nt 1255 to 1293; Pll, nt 1594 to 1635; and P12, nt 2008 to 2106. The region corresponding to the deoxyoligonucleotide probe P43 is underlined, and the asterisk in that region denotes the one nucleotide that is mismatched in the P43 sequence. Restriction sites for enzymes XhoI, HindIll, and KpnI are also noted.

J. BACTERIOL.

POLLOCK ET AL.

224

3

2

1

13 2

4

4

DvH cytochrome C3

I

I| I iI I 11

12

DvH Hmc

II 2 ni

3

I I

n-

ii

lI4l

Il

4

5

ri

n

5

6

I 7

7

n

8

9

nn

11 1 8 10 9

IV 13 I I I lI

I I i I I

i

I

I

I

I

13

14

l_i I

.

15 14

rI

I

15

16

r-i

rF]

IF I

L

16

100 A.A.

FIG. 3. Linear arrangement of the heme binding sites (boxes; in cytochrome c3, this denotes both C-X-X-C-H and C-X-X-X-X-C-H) and sixth-position His residues (short bars) as they occur in the cytochrome c3 and Hmc polypeptides from D. vulgaris Hildenborough. The heme binding sites are numbered as they occur proceeding from NH2 to COOH termini. The numbers of the His residues correspond to the heme binding site with which they are supposedly associated; for the Hmc polypeptide, these numbers are suggested by comparing each of domains I to IV with the cytochrome C3 motif. Region I is lacking in both one heme binding site, which would normally be heme binding site 2 in the cytochrome c3 pattern, and its corresponding His residue. Heme binding site 12 does not fit into the cytochrome C3 motif.

one of the heme Fe atoms in Hmc is coordinated to a Met residue, this would not be the first report of a c-type cytochrome containing both His-His- and His-Met-coordinated hemes; the tetraheme cytochrome c of the Rhodopseudomonas viridis photosynthetic reaction center has one His-His- and three His-Met-coordinated heme Fe atoms (6). The hmc gene does not encode any transmembrane sequences other than the NH2-terminal signal sequence, and its product is therefore expected to reside in the periplasmic

space.

Nucleotide sequences up- and downstream from hmc. An E. coli &70 consensus promoter is present at nt 209 (Fig. 2). A similar promoter sequence was found in the upstream region of the D. vulgaris Hildenborough cyc gene encoding cytochrome C3 (35). The -35 sequences of the two promoters are identical (TTGACA), as are four of the six nucleotides of the -10 sequences (TACCAT for cyc). The ribosome binding site at nt 460 contains the core AGGA sequence which is typical of all D. vulgaris Hildenborough genes sequenced to date except the rbo gene (2, 4, 32, 34, 35, 36). No hairpin loop transcription termination sequences have been detected downstream of the hmc gene, and sequencing of the downstream region, although incomplete, indicates that the hmc gene may be the first cistron of a redox operon, which is composed of at least three open reading frames (not shown). The metabolic role of Hmc has not been firmly established, and the sequencing of this operon may help to uncover the redox partners of Hmc. Probing of Desulfovibrio genomes with D. vulgaris Hildenborough hmc. Samples of total DNA from several Desulfovibrio species were digested with EcoRI, subjected to Southern blotting, and probed with the 2.2-kb SmaI fragment of plasmid pP6A (not shown). The region spanning nt 1 to 2109 of the sequence illustrated in Fig. 2 represents 94% of this 2.2-kb SmaI fragment. Therefore, this fragment is mainly representative of the hmc gene and its upstream region.

Autoradiography showed single bands for the following Desulfovibrio species: D. vulgaris Brockhurst Hill, Hildenborough, Monticello, and Wandle; D. desulfuricans Berre Sol, Canet 41, and Teddington R; and D. africanus Bhengazi and Walvis Bay. These results suggest that the hmc gene

may be present in these Desulfovibrio species. No hybridizing bands were observed for D. gigas, D. multispirans, D. salexigens, and D. desulfuricans El Agheila Z and Norway, which suggests that hmc is either not present or not sufficiently homologous for hybridization in these bacteria. None of the samples tested displayed multiple bands, which indicates that the hmc gene is present in single copy in the positive chromosomes. Expression of Hmc in D. desulfuricans G200. Electrophoresis in the presence of a reductive chelator such as 2-mercaptoethanol or dithiothreitol leads to the removal of iron from heme to yield the protoheme, which upon excitation with UV light fluoresces in the visible range (10, 15). Therefore, following electrophoresis under denaturing conditions in the presence of 2-mercaptoethanol, c-type cytochromes, which are covalently bound to their hemes, can be visualized under UV light, whereas other heme-containing proteins cannot since their hemes are not retained upon denaturation. The periplasmic protein fractions of D. vulgaris Hildenborough (Fig. 4, lanes 3) and the two exconjugants D. desulfuricans G200(pJRD215) (lanes 1) and D. desulfuricans G200(pBPHMC-1) (lanes 2) were subjected to SDS-PAGE. Directly upon completion of electrophoresis, the gel was analyzed for postelectrophoretic protoheme fluorescence (Fig. 4A). Comparison of lanes 1 and 2 shows that D. desulfuricans G200(pBPHMC-1) expressed the D. vulgaris Hildenborough hmc gene as a fluorescent protein of 66 kDa. Lanes 1 to 3 in Fig. 4A all share a fluorescent band corresponding to 15 kDa that is due to the presence of the cytochromes C3 native to both D. desulfuricans G200 (lanes 1 and 2; 35a) and D. vulgaris Hildenborough (lane 3; 26). The periplasmic fraction of D. desulfuricans G200 also contains a fluorescent protein of 45 kDa (lanes 1 and 2). The apparent absence of a fluorescent band corresponding to 66 kDa (mature Hmc has a molecular size of 65.5 kDa; see above) in lane 3 is due to the low concentration of Hmc in D. vulgaris Hildenborough. The clear presence of this band in lane 2 illustrates that the exconjugant was able to express the hmc gene product at higher levels than did the parent organism. Figure 4B, which shows the gel illustrated in Fig. 4A after staining with Coomassie blue, indicates that the

VOL. 173, 1991

2

*. : . :

D. VULGARIS HILDENBOROUGH hmc GENE

1

3

'.' .

.,, .'.

43.0

si ',

4-

..w3F

_

biochemical work presented here indicates that this assignment is incorrect. The hmc gene encodes an additional 31-amino-acid signal sequence (Fig. 2). Comparison of this sequence with the signal peptides of the periplasmic D. vulgaris Hildenborough cytochromes C3 and C553 (Fig. 5) shows that it is also a signal peptide with the same characteristic positive and hydrophobic regions as in the other two sequences. However, the Hmc signal sequence has a greater overall length, has three NH2-terminal Arg residues that are spaced out rather than adjacent, and has a signal peptide cleavage site that is not of the form A a A as is seen in the signal sequences of the other two procytochromes. Hmc can be purified from the periplasm of D. vulgaris Hildenborough in quantities similar to those obtained by using whole cells (see Materials and Methods). This protein was also found to be localized in the periplasm when it was overexpressed in D. desulfuricans G200 cells containing plasmid pBPHMC-1 (Fig. 4). The periplasmic localization of Hmc is in accord with an empirical rule which states that maturation of c-type cytochromes requires a membrane translocation step. This rule applies to both procaryotic and eucaryotic cytochromes, e.g., the mitochondrial, monoheme cytochromes c and cl (24, 25), for which it is known that the heme is covalently inserted by the enzyme heme lyase (8) following or during translocation of the apocytochrome.

2

:_

...':: .' '..:. ..

,,9

S.S.W,_

.:..

....

s -S -S£:f..

..A' 2i, .;

_

FIG. 4. Expression of D. vulgaris Hildehborough Hmc by D. desulfuricans G200 transformed with plasmid pBPHMC-1. (A) PostSDS-PAGE protoheme fluorescence analysis of the periplasmic fractions from D. desulfuricans G200(pJRD215) (lane 1), the recombinant exconjugant D. desulfuricans G200(pBPHMC-1) (lane 2), and D. vulgaris Hildenborough (lane 3). (B) Coomassie blue staining of the same gel. Positions of the size standards are marked on the left and are given in kilodaltons. Approximately 30 ,ug of protein was loaded in each lane.

absence of a fluorescent band at 66 kDa in lanes 1 and 3 of Fig. 4A was not due to differences in the amounts of protein loaded. A similar result was obtained for the cyc gene encoding cytochrome C3, which was expressed at a fivefoldhigher level in the D. desulfuricans G200 exconjugant than in D. vulgaris Hildenborough (35a). The presence of plasmid pBPHMC-1 in the recombinant exconjugant D. desulfuricans G200(pBPHMC-1) was verified by probing total DNA with the 2.2-kb SmaI fragment of pP6A (not shown). Periplasmic localization of Hmc. According to a predictive method, which involves a comparison of the mature NH2terminal amino acid residues of both known periplasmic and cytoplasmic proteins from Desulfovibrio spp. (16), Hmc would be classified as a cytoplasmic protein. The genetic and A

DISCUSSION The Hmc class of proteins. Cytochrome cc3, which has also been referred to as cytochrome C3 (Mr, 26,000) (18) and octaheme cytochrome C3 (17, 22), from D. vulgaris Hildenborough was reported to be an octaheme dimer composed of two identical tetraheme subunits, each with a molecular size of 20 kDa (Table 2). Partial amino acid sequence data obtained from this same cytochrome was used for the formulation of deoxyoligonucleotide probe P43 (see Materials and Methods), which was used to isolate the recombinant plasmid pP6A. Surprisingly, nucleotide sequencing (Fig. 2)

-1 +1 KALP

-10

-20

-30

MENGETLLBWAGVLAATAIIGVGGFWSQGTT MBELFFCGVLALAVAFALPVVA MzBVLLLSSLCAALSFGLAVSGVA

DYH Hmc

D.zH Cytochrome J3 DvH Cytochrome p553

B

1

DvH HMC

KALPEGPGEKRADLIEIGAMERFGKLDLPKV

L PEPTIDE:

IPLPGATGEQRADLVEIGVMAKFTNLELPKV

*

C DYH HMC

390

****

***

*

ADGA

*

*

****

400

KGTFDLNDIPEKVVIGSIA ***

S PEPTIDE:

**

APKA

30

20

10 **

225

****

*

**

KGTLPAVDIPEFVTIGVLK

FIG. 5. (A) Comparison of the NH2-terminal signal peptide sequences for three periplasmic, c-type cytochromes from D. vulgaris Hildenborough. All three sequences have been determined by nucleic acid sequencing of their respective structural genes. The cleavage site is marked by the vertical line, and + 1 marks the first amino acid residue of the mature polypeptide. Positively charged residues occurring in the signal sequence are underlined. (B and C) Comparison of the L (B) S (C) peptide sequences determined for D. vulgaris Miyazaki Hmc with their homologous regions in D. vulgaris Hildenborough Hmc. Numbers indicate positions of the amino acid residues in the Hildenborough sequence. The underlined Lys residues have not been sequenced but are assumed from the activity of lysyl-endopeptidase on these peptides. Direct homology is indicated by an asterisk.

226

POLLOCK ET AL.

J. BACTERIOL.

TABLE 2. Comparison of some physical properties and amino acid composition of cytochrome CC3 (18) and Hmc (13) with data obtained from the sequence of the gene for D. vulgaris Hildenborough Hmc (this study) Value obtained

Property Mr of holoform (103) Subunits Heme content/subunit pl of ferriform No. of residues Lys His Arg Asx Thr Ser Glx Pro Gly Ala Cys Val Met Ile Leu Tyr Phe Trp Total

Higuchi

Loutfi

et al. (18)

This work

et al. (13)

65.5

69"

70a

8lb

43.3c

2 x 20 kDad

1

1

16 NDe

16

54

44

31 23 47 24

31 25 46 23 22 51 33 41 68 27 27 11 18 27 3 15 ND

48 26 23 48 26 23 48 32 45 58 26 26 16 19 26 6 16 ND

512f

51Sf

24 44 33 37 63 32 27 12 18 25 5 13 2 514

9.2

4 9.2

a SDS-PAGE of holoform. b Gel filtration of holoform. I Analytical ultracentrifugation of holoform. d SDS-PAGE of apoform. e ND, Not determined. f Adjusted to 512 residues.

of the D. vulgaris Hildenborough chromosomal DNA insert carried by this plasmid indicated the presence of an open reading frame for a mature hexadecahemocytochrome c of 65.5 kDa rather than the expected tetrahemocytochrome c of 20 kDa. In addition to the nucleotide sequence data, the following evidence indicates that in D. vulgaris Hildenborough, cytochrome cc3 is a 65.5-kDa protein. (i) The amino acid sequence data obtained by direct peptide sequencing (polypeptides P1 to P12; Fig. 2) preclude the protein from being a 43.3-kDa, octaheme homodimer. Rather than the 4 heme binding sequences (C-X-X-C-H) expected in this model (18), polypeptides P1 to P12 contain 14 separate sequences C-X-X-C-H and cover 85% of the mature polypeptide sequence encoded by the hmc gene (Fig. 2). (ii) SDS-PAGE of preparations of holocytochrome CC3 indicates a major band at 70 kDa (18), whereas the apoprotein gives a major band at 56 kDa (not shown). These values correspond to the 65.5-kDa holoprotein and 55.7-kDa processed apoprotein encoded by the hmc gene. It is likely that the lowermolecular-weight bands observed in preparations of cytochrome CC3 from D. vulgaris Hildenborough are the result of limited proteolysis between different domains (Fig. 3). (iii) As discussed also by Yagi and Ogata (38), the amino acid composition and physical properties of D. vulgaris Hildenborough cytochrome CC3 (18) are similar to those of the high-molecular-weight cytochrome c from the same strain (13). These data are summarized in Table 2, which also

contains the amino acid composition derived from the gene sequence. Hmc and cytochrome CC3 are identical with respect to pl in the oxidized state, Mr by SDS-PAGE, and number of hemes per polypeptide and are found to have very similar amino acid compositions (Table 2). Although a cytochrome with the characteristics of Hmc (16 c-type hemes; Mr, 65.5 x 103) has not been isolated from other Desulfovibrio species, single hybridizing bands were observed when a Southern blot of EcoRI-digested chromosomal DNA from several Desulfovibrio species was probed with the nick-translated, 2.2-kb SmaI fragment of pP6A (see Results). This finding suggests that Hmc is not restricted to D. vulgaris Hildenborough in the genus Desulfovibrio. The data presented above clearly indicate that the model of D. vulgaris Hildenborough cytochrome CC3 as an octaheme homodimer of 43.3 kDa is incorrect. It is possible that this is also the case for the cytochromes CC3 identified in D. desulfuricans Norway (11) and D. gigas (3). However, these proteins differ from D. vulgaris Hmc in that both have never been reported to display a molecular size greater than 26 kDa. Additionally, neither D. desulfuricans Norway nor D. gigas chromosomal DNA was seen to hybridize to the 2.2-kb SmaI fragment of plasmid pP6A (see Results). Both of these observations suggest that the D. desulfuricans Norway and D. gigas proteins may be quite different from Hmc. A complete description of the amino acid sequence of these proteins will be required to settle this question. Cytochromes C3 and C553 from D. vulgaris Miyazaki are homologous to the corresponding proteins of the Hildenborough strain, and the Miyazaki strain also contains a cytochrome similar to Hmc (30a). Although this cytochrome is reported to have only 11 hemes per polypeptide (38), it does display similarity to the Hmc from the Hildenborough strain in two respects: (i) the holoform of this protein is also a monomer of 67 kDa by SDS-PAGE, and (ii) there is homology with Hmc in the limited amino acid sequence obtained for the Miyazaki protein both in the NH2-terminal peptide (L; 30a) and in an internal peptide fragment (S; Fig. 5). The Miyazaki and Hildenborough Hmcs are further compared below. Hmc protein structure. Cytochromes C3 display a distinct linear arrangement of heme binding sites and coordinating histidinyl residues (14). In D. vulgaris cytochrome C3, the elements of this pattern (Fig. 3) span approximately 100 amino acids and are, proceeding from NH2 to COOH termini: (i) a H-X-X-H sequence in which the first His residue coordinates to heme 1 and the second His residue coordinates to heme 3, (ii) a C-X-X-C-H heme binding sequence for heme 1, (iii) a single His residue that coordinates to heme 2, (iv) a C-X-X-X-X-C-H heme binding sequence for heme 2, (v) a single His residue that coordinates to heme 4, (vi) a C-X-X-C-H heme binding sequence for heme 3, and (vii) a C-X-X-X-X-C-H heme binding sequence for heme 4. The 15 non-heme binding site His residues of Hmc and the 16 heme binding sites are arranged relative to one another in a pattern that resembles a quadruplication of the same components as they are arrayed in cytochrome C3 (Fig. 3). There are some irregularities of possible consequence in this motif. The first domain (I; hemes and His 1 to 3 as in Fig. 3) is composed of only six of the eight elements present in the cytochrome C3 pattern. According to this pattern, the second heme binding site and its corresponding sixth-position His residue are absent. The other irregularity is related to the first in that there is an extra heme (number 12; Fig. 3) immediately following the third domain, and it is the only one that does not have a His residue, which may be related to it by the

VOL. 173,

1991

cytochrome C3 motif. By this reasoning, heme 12 is the one most likely not to have bis-histidinyl coordination. X-ray crystallographic data obtained for D. vulgaris Hildenborough Hmc have been presented (13) but are not of sufficient resolution for structural information. The pattern in Fig. 3 indicates a threefold symmetry for Hmc, since the 15 bis-histidinyl-coordinated hemes are arranged into three complete cytochrome c3-like domains (II to IV; Fig. 3), each of which comprises four hemes, and an incomplete domain (I; Fig. 3) with only three hemes. The single Met-Hiscoordinated heme may function as the electron exit point, interacting with a higher-potential redox protein. This exit heme could obtain its electrons from the three hemes in domain I (Fig. 3), which could each, in turn, receive electrons from one of the three remaining, tetraheme domains (II, III, and IV; Fig. 3). Domains II, III, and IV could each interact with a redox protein in addition to their interaction with heme 12 via domain I. NH2-terminal sequence data for the L and S peptides from D. vulgaris Miyazaki Hmc have been obtained (see above). The partial amino acid sequence of the S peptide (19.1 kDa) depicted in Fig. SC is homologous to a stretch of amino acids located between the third and fourth cytochrome c3-like domains of the Hmc from the Hildenborough strain. This Miyazaki peptide, cleaved from the holoprotein, has been reported to be free of heme, and all of the heme was to be found on the NH2-terminal L peptide (33.1 kDa). The discrepancy in heme content between the two Hmcs (11 and 16 hemes per 67 kDa for the Miyazaki and Hildenborough holoproteins, respectively) can be accounted for by a complete absence of heme binding sites from domain IV (Fig. 3), and this suggests that the Hmc class of cytochromes may display a high degree of variability in domain IV. ACKNOWLEDGMENTS W.B.R.P. acknowledges the support of a Graduate Scholarship from the Alberta Heritage Foundation for Medical Research. This work was supported by an operating grant from the National Sciences and Engineering Research Council of Canada to G.V. and by grant DE-FG02-87ER13713 from the Basic Energy Research Program of the U.S. Department of Energy awarded to J.D.W. We thank P. Codding for assistance with the interpretation of X-ray crystallographic data and J. Bonicel for help with the automated protein sequence analysis. REFERENCES 1. Bankier, A. T., and B. G. Barrell. 1983. Shotgun DNA sequencing, p. 1-34. In R. A. Flavell (ed.), Techniques in the life sciences, B5. Nucleic acid biochemistry. Elsevier Scientific Publishers Ireland Ltd., Shannon, Ireland. 2. Brumlik, M. J., and G. Voordouw. 1989. Analysis of the transcriptional unit encoding the genes for rubredoxin (rub) and a putative rubredoxin oxidoreductase (rbo) in Desulfovibrio vulgaris Hildenborough. J. Bacteriol. 171:4996-5004. 3. Bruschi, M., J. LeGall, C. E. Hatchikian, and M. Dubourdieu. 1969. Cristallisation et proprietes d'un cytochrome intervenant dans la reduction du thiosulfate par Desulfovibrio gigas. Bull. Soc. Fr. Physiol. Veget. 15:381-390. 4. Curley, G. P., and G. Voordouw. 1988. Cloning and sequencing of the gene encoding flavodoxin from Desulfovibrio vulgaris Hildenborough. FEMS Microbiol. Lett. 49:295-299. 5. Davison, J., M. Heusterpreute, N. Chevalier, H. T. Vinh, and F. Brunel. 1987. Vectors with restriction site banks. V. pJRD215, a wide-host range cosmid vector with multiple cloning sites. Gene 51:275-280. 6. Deisenhofer, J., and H. Michel. 1989. The photosynthetic reaction center from the purple bacterium Rhodopseudomonas viridis (Nobel lecture). Angew. Chem. 28:829-847.

D. VULGARIS HILDENBOROUGH hmc GENE

227

7. Dickerson, R. E., and R. Timkovich. 1975. Cytochromes c, p. 397-547. In P. D. Boyer (ed.), The enzymes, vol. 11, 3rd ed. Academic Press, Inc., New York. 8. Dumont, M. E., J. F. Ernst, D. M. Hampsey, and F. Sherman. 1987. Identification and sequence of the gene encoding cytochrome c heme lyase in the yeast Saccharomyces cerevisiae. EMBO J. 6:235-241. 9. Eaton, W. A., and R. M. Hochstrasser. 1967. Electronic spectrum of single crystals of ferricytochrome-c. J. Chem. Phys. 46:2533-2539. 10. Goodhew, C. F., K. R. Brown, and G. W. Pettigrew. 1986. Haem staining in gels, a useful tool in the study of bacterial c-type cytochromes. Biochim. Biophys. Acta 852:288-294. 11. Guerlesquin, F., G. Bovier-Lapierre, and M. Bruschi. 1982. Purification and characterization of cytochrome C3 (Mr 26,000) isolated from Desulfovibrio desulfuricans Norway strain. Biochem. Biophys. Res. Commun. 105:530-538. 12. Hanahan, D. 1985. Techniques for transformation of E. coli, p. 109-136. In D. M. Glover (ed.), DNA cloning, vol. 1. IRL Press, Washington, D.C. 13. Higuchi, Y., K. Inaka, N. Yasuoka, and T. Yagi. 1987. Isolation and crystallization of high molecular weight cytochrome from Desulfovibrio vulgaris Hildenborough. Biochim. Biophys. Acta

911:341-348. 14. Higuchi, Y., M. Kusunoki, Y. Matsuura, N. Yasuoka, and M. Kakudo. 1984. Refined structure of cytochrome C3 at 1.8 A resolution. J. Mol. Biol. 172:109-139. 15. Katan, M. B. 1976. Detection of cytochromes on sodium dodecylsulphate-polyacrylamide gels by their intrinsic fluorescence. Anal. Biochem. 74:132-137. 16. LeGall, J., and H. D. Peck, Jr. 1987. Amino-terminal amino acid sequences of electron transfer proteins from Gram-negative bacteria as indicators of their cellular localization: the sulfatereducing bacteria. FEMS Microbiol. Rev. 46:35-40. 17. Liu, M.-C., C. Costa, I. B. Coutinho, J. J. G. Moura, I. Moura, A. V. Xavier, and J. LeGall. 1988. Cytochrome components of nitrate- and sulfate-respiring Desulfovibrio desulfuricans ATCC 27774. J. Bacteriol. 170:5545-5551. 18. Loutfi, M., F. Guerlesquin, P. Bianco, J. Haladjian, and M. Bruschi. 1989. Comparative studies of polyhemic cytochromes c isolated from Desulfovibrio vulgaris (Hildenborough) and Desulfovibrio desulfuricans (Norway). Biochem. Biophys. Res. Commun. 159:670-676. 19. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 20. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3:208-218. 21. Moore, G. R., and R. J. P. Williams. 1977. Structural basis for the variation in redox potential of cytochromes. FEBS Lett.

79:229-232. 22. Moura, I., G. Fauque, J. LeGall, A. V. Xavier, and J. J. G. Moura. 1987. Characterization of the cytochrome system of a nitrogen-fixing strain of a sulfate reducing bacterium: Desulfovibrio desulfuricans strain Berre-Eau. Eur. J. Biochem. 162: 547-554. 23. Nakano, K., Y. Kikumoto, and T. Yagi. 1983. Amino acid sequence of cytochrome c-553 from Desulfovibrio vulgaris Miyazaki. J. Biol. Chem. 258:12409-12412. 24. Nargang, F. E., M. E. Drygas, P. L. Kwong, D. W. Nicholson, and W. Neupert. 1988. A mutant of Neurospora crassa deficient in cytochrome c heme lyase activity cannot import cytochrome c into mitochondria. J. Biol. Chem. 263:9388-9394. 25. Ohashi, A., J. Gibson, I. Gregor, and G. Schatz. 1982. Import of proteins into mitochondria. The precursor of cytochrome c1 is processed in two steps, one of them is heme-dependent. J. Biol.

Chem. 257:13042-13047. 26. Pollock, W. B. R., P. J. Chemerika, M. E. Forrest, J. T. Beatty, and G. Voordouw. 1989. Expression of the gene encoding cytochrome C3 from Desulfovibrio vulgaris (Hildenborough) in Escherichia coli: export and processing of the apoprotein. J. Gen. Microbiol. 135:2319-2328. 27. Postgate, J. R. 1984. The sulphate-reducing bacteria, 2nd ed., p.

POLLOCK ET AL.

228 32.

Cambridge University Press, Cambridge.

28. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. 29. Staden, R. 1984. Graphical methods to determine the function of nucleic acid sequences. Nucleic Acids Res. 12:521-538. 30. Staden, R., and A. D. McLachlan. 1982. Codon preference and its use in identifying protein coding regions in long DNA sequences. Nucleic Acids Res. 10:141-156. 30a.Tasaka, C., M. Ogata, T. Yagi, and A. Tsugita. Protein Sequences Data Anal., in press. 31. Van der Westen, H. M., S. G. Mayhew, and C. Veeger. 1978. Separation of hydrogenase from intact cells of Desulfovibrio vulgaris. J. Biotechnol. 12:173-184. 32. Van Rooien, G. J. H., M. Bruschi, and G. Voordouw. 1989. Cloning and sequencing of the gene encoding cytochrome C553 from Desulfovibrio vulgaris Hildenborough. J. Bacteriol. 171: 3575-3578. 33. Vieira, J., and J. Messing. 1982. The pUC plasmids, an M13mp7-derived system for the insertion mutagenesis and sequencing with synthetic universal primers. Gene 19:259-268. 34. Voordouw, G., and S. Brenner. 1985. Nucleotide sequence of the gene encoding the hydrogenase from Desulfovibrio vulgaris (Hildenborough). Eur. J. Biochem. 148:515-520.

J. BACTERIOL. 35. Voordouw, G., and S. Brenner. 1986. Cloning and sequencing of the gene encoding cytochrome C3 from Desulfovibrio vulgaris (Hildenborough). Eur. J. Biochem. 159:347-351. 35a.Voordouw, G., W. B. R. Pollock, M. Bruschi, F. Guerlesquin, B. J. Rapp-Giles, and J. D. Wall. 1990. Functional expression of Desulfovibrio vulgaris Hildenborough cytochrome C3 in Desulfovibrio desulfuricans G200 after conjugational gene transfer from Escherichia coli. J. Bacteriol. 172:6122-6126. 36. Voordouw, G., J. E. Walker, and S. Brenner. 1985. Cloning of the gene encoding the hydrogenase from Desulfovibrio vulgaris (Hildenborough) and the determination of the amino terminal sequence. Eur. J. Biochem. 148:509-514. 37. Weimar, P. J., M. J. van Kavelaar, C. B. Michel, and T. K. Ng. 1988. Effect of phosphate on the corrosion of carbon steel and on the composition of corrosion products in two-stage continuous cultures of Desulfovibrio desulfuricans. Appl. Environ. Microbiol. 54:386-396. 38. Yagi, T., and M. Ogata. 1990. Electron carrier proteins in Desulfovibrio vulgaris Miyazaki, p. 227-249. In J.-P. Belaich, M. Bruschi, and J.-L. Garcia (ed.), Proceedings of the FEMS symposium on microbiology and biochemistry of strict anaerobes involved in interspecies hydrogen transfer. Plenum Press, New York.