J. Mol. Biol. (1976) 107, 179-182
Some Comparisons between two Crystallized Anaerobic Bacterial Rubredoxins from Desulfovibrio gigas and D. vulgaris Two bacterial rubredoxins isolated from Desulfovibrio vulgaris and D. gigas have been crystallized. The cell parameters and the space group are identical but comparison of the X-ray diffraction patterns and the sequence and reactivity data reveals structural differences. The solvent content of both crystals appears to be unusually low (,--27 %}. Rubredoxin, a small iron-containing protein, has been isolated and characterized from several species of anaerobic bacteria (Lovenberg & Sobel, 1965; Mayhew & Peel, 1966; Le Gall & Dragoni, 1966; Meyer et al., 1971) and in one variant form, an aerobic bacterium (Peterson et al., 1966). The rubredoxins from the anaerobic bacteria contain one iron atom, no labile sulfur and four eysteine groups as invariant features of the molecule. The molecular weight is about 6000 for a single polypeptide chain of 52 to 54 amino acid residues. The availability of several rubredoxins has led to interesting studies from the viewpoint of structural, functional and evolutionary relationships. The amino acid sequences of five anaerobic bacterial rubredoxins have been determined: Peptostrep. tococcu~ elsdenii (Bachmayer et al., 1968~), Clostridium pasteurianum (McCarthy, 1972), Peptococcus (formerly Micrococcus) aerogenes (Bachmayer et al., 1968b), Desulfovibrio vulgaris (Bruschi et al., 1976a), D. giga8 (Bruschi el al., 1976b). Alignment of these amino acid sequences shows 18 amino acid residues in identical positions (Vogel el al., unpublished data). These are of interest concerning the structural requirements for the residues binding the iron atom, for the specific residues involved in establishing the entaetic nature of the iron complex, and for the specificity (if any) in electron transfer. No differences are found in the position of the two peptides -Cys-X-Y-Cys-Gly- which are involved in the binding of the iron. The tertiary structure of rubredoxin from the bacterium C. pasteurianum has been determined to 1.5 A resolution (Watenpaugh el al., 1973). The iron atom of the structure is bonded to the four cysteinyl sulfurs and the polypeptide chain is folded in such a manner t h a t most of the aromatic groups are buried in the central core of the molecule. The Fe--(Cys)4 complex is located toward the exterior opposite the N- and C-terminal region of the molecule. I t is interesting to note t h a t one F e - - S bond appears to be shorter than the other three F e - - S bonds which m a y be a distinct feature of the complex. The sequences of the two rubredoxins isolated from the sulfate-reducing bacteria,
D. vulgaris and D. gigas, are similar. But m a n y more differences occur than would be expected from bacteria of the same genus. Thirty seven amino acids {about 70%) are found in identical positions; both have an N-terminal Met which is blocked in the D. gigas rubredoxin but not in the D. vulgaris rubredoxin. The N-terminal parts of these two proteins are identical between residues 4 and 16. 179
M. P I E R R O T E T AL.
180
The differences in the amino acid sequences are reflected in t h e i r functional relationships; an N A D H ~ H + rubredoxin oxido-reductase has been described b y Le Gall (1968) in ~). gigas. This enzyme is ten times more reactive with D. gigas rubredoxin t h a n with D. vulgaris rubredoxin and shows no activity toward rubredoxin from C. pasteurianum or P. aerogenes (H. D. Peck Jr, M. Odum, personal communication). I n order to explain such specific behavior we have undertaken the threedimensional structure determination of these rubredoxins. The proteins were isolated by the procedures of Le Gall & Dragoni (1966) for D. gigas and Bruschi & Le Gall (1972) for D. vulgaris. Both rubredoxins can be crystallized under the following conditions : 10 mg of protein were dissolved in 1 ml of sodium citrate buffer, 0.1 M at p H 4.0, solid a m m o n i u m sulfate was added, without stirring, to small samples (100/A) in order to cover a narrow range of salt concentration of about 25% (w/v). The closed tubes were maintained at 27~ and good-sized crystals (--~0.5 m m • 0.5 m m • 0-3 ram) developed within three days. X - r a y precession photographs of the diffraction pattern reveal t h a t crystals of the two rubredoxins belong to the monoclinic system, space group P21. The crystallographic unit cell parameters are listed in Table 1, together with those of rubredoxin C. pasteurianum (Watenpaugh et al., 1973). TABLE 1
Crystallographic unit cell parameters Source of rubredoxin
Space group
a(A)
b(A)
c(A)
fl or a(~
Z
V(A 3)
D. vulgaris
P21
20'00
41.50
24.41
107"6
2
19310
D. gigas
P21
19"75
41.70
24.38
108'1
2
19091
C. pasteurianum
R3
38.77
--
--
112.4
3
39187
Assuming two molecules of rubredoxin in the unit cell, one per asymmetric unit, the crystal volume per unit molecular weight (VM) is H I . 6 A 3 per dalton. This value for VM is very low as evidenced by a tabulation done by Matthews (1968), where the observed values of V~ range from 1.7 to 3.5/k 3 per dalton. VM for the C. pasteurianum rubredoxin is 2.2 A3/dalton which is close to the average value 2.37 A3/dalton determined b y Matthews' tabulation. With the plausible assumption t h a t the volume of the protein molecule is approximately the same for the three rubredoxins, the solvent cell content of rubredoxin in D. vulgaris and D. gigas ( ~ 2 7 % ) appears to be much lower t h a n t h a t for C. pasteurianum ( ~ 4 4 % ) . The high degree of similarity in the crystal cell parameters between rubredoxin from D. vulgaris and from D. gigas suggests t h a t the molecular packing arrangements in the two crystals are due to the involvement of invariant amino acid residues between the two tertiary structures. A comparison of the precession patterns of the principal zones (see Fig. 1) reveals substantial differences for m a n y of the diffracted intensities. The preliminary crystallographic data enhance the interest in determining the ~tructurr of both these rubredoxin~. Differences in terms of ~r as well ~s sequence
LETTERS
TO THE EDITOR
181
(a)
(b) FH~. 1. 16 ~ hk0 precession photographs of rubredoxin crystals. (a) D. giga,s; (b) D. vu/gar/8.
~82
M. P I E R R O T E T A L .
between the two proteins are probably related to local modifications in the poly:peptide side-groups producing the observed differences between the diffraction :patterns. Data sets for both rubredoxins have been collected to 1.5 A resolution on native protein crystals. Solution of the two structures will be attempted, separately, By rotation-translation functions of the structure of ruhredoxin from C. pasteurianum. .'It is not unreasonable to expect to solve the structure to near atomic resolution. 'Thus this will provide two independent determinations of the FeS4 complex geometry ,to compare with the FeS4 complex geometry of C. pasteurianum rubredoxin. These De~ulfovibrio rubredoxin crystals diffract extremely well, and it is anticipated that at least one of these structures will be determined to well beyond 1.0 A resolution. This work was supported by the Centre National de la Recherche Scientifique and by U.S. Public Health Services grant GM-13366 from the National Institute of Health. Centre des Mecanismes de la Croissance Cristalline Universit6 d'Aix-Marseille I I I Centre Saint J6rome 13397 Marseille/C6dex, France
M. PIERROT R . HASER M. FREY
Laboratoire de Chimie Bact6rienne CNRS Marseille, France
M. BRUSCHI J . LE GALL
Department of Biological Structure University of Washington, School of Medicine Seattle, Wash. 98195, U.S.A.
L. C. SIEKER L. H . J-ENSEN
Received 27May 1976 REFERENCES Bachmayer, H., Benson, A. M., Garrard, W. T., Whiteley, H. R. & Yasunobu, K. T. (1968a). Biochemistry, 7, 986-996. Bachmayer, H., Mayhew, S., Peel, J. & Yasunobu, K. T. (1968b). J. Biol. Chem. 243, 1022-1030. Bruschi, M. & Le Gall, J. (1972). Biochim. Biophys. Acta, 263, 279-282. Bruschi, M., Bonicel, J., Bovier-Lapierre, G. & Couchoud, P. (1976a). Biochim. Biophys. Aaa, in the press. Bruschi, M., Bonicel, J., Bovier-Lapierre, G. & Couchoud, P. (1976b). Biochem. Biophys. Res. Commun. in the press. Le Gall, J. (1968). Ann. Inst. Pasteur (Paris), 114, 109-115. Le Gall, J. & Dragoni, N. (1966). Biochem. Biophys. Res. Commun. 23, 145-149. Lovenberg, W. & Sobel, B. E. (1965). Prec. Nat. Acad. Sci., U.S.A. 54, 193-199. Matthews, B. W. (1968). J. Mol. Biol. 33, 491-497. Mayhew, S. G. & Peel, J. L. (1966). Biochem. J. 100, 80P. McCarthy, K. F. (1972). PhD. Thesis, George Washington University, Washington D.C. Meyer, T. E., Sharp, J. J. & Bartsch, R. G. (1971). Biochim. Biophys. Acta, 234, 266-269. Peterson, J. A., Basu, D. & Coon, M. J. (1966). J. Biol. Chem. 241, 5162-5164. Watenpaugh, K. D., Sieker, L. C., Herriott, J. R. & Jensen, L. H. (1973). Acta Crystallogr. sect. B29, 943-956.
Some Comparisons between two Crystallized Anaerobic Bacterial Rubredoxins from Desulfovibrio gigas and D. vulgaris Two bacterial rubredoxins isolated from Desulfovibrio vulgaris and D. gigas have been crystallized. The cell parameters and the space group are identical but comparison of the X-ray diffraction patterns and the sequence and reactivity data reveals structural differences. The solvent content of both crystals appears to be unusually low (,--27 %}. Rubredoxin, a small iron-containing protein, has been isolated and characterized from several species of anaerobic bacteria (Lovenberg & Sobel, 1965; Mayhew & Peel, 1966; Le Gall & Dragoni, 1966; Meyer et al., 1971) and in one variant form, an aerobic bacterium (Peterson et al., 1966). The rubredoxins from the anaerobic bacteria contain one iron atom, no labile sulfur and four eysteine groups as invariant features of the molecule. The molecular weight is about 6000 for a single polypeptide chain of 52 to 54 amino acid residues. The availability of several rubredoxins has led to interesting studies from the viewpoint of structural, functional and evolutionary relationships. The amino acid sequences of five anaerobic bacterial rubredoxins have been determined: Peptostrep. tococcu~ elsdenii (Bachmayer et al., 1968~), Clostridium pasteurianum (McCarthy, 1972), Peptococcus (formerly Micrococcus) aerogenes (Bachmayer et al., 1968b), Desulfovibrio vulgaris (Bruschi et al., 1976a), D. giga8 (Bruschi el al., 1976b). Alignment of these amino acid sequences shows 18 amino acid residues in identical positions (Vogel el al., unpublished data). These are of interest concerning the structural requirements for the residues binding the iron atom, for the specific residues involved in establishing the entaetic nature of the iron complex, and for the specificity (if any) in electron transfer. No differences are found in the position of the two peptides -Cys-X-Y-Cys-Gly- which are involved in the binding of the iron. The tertiary structure of rubredoxin from the bacterium C. pasteurianum has been determined to 1.5 A resolution (Watenpaugh el al., 1973). The iron atom of the structure is bonded to the four cysteinyl sulfurs and the polypeptide chain is folded in such a manner t h a t most of the aromatic groups are buried in the central core of the molecule. The Fe--(Cys)4 complex is located toward the exterior opposite the N- and C-terminal region of the molecule. I t is interesting to note t h a t one F e - - S bond appears to be shorter than the other three F e - - S bonds which m a y be a distinct feature of the complex. The sequences of the two rubredoxins isolated from the sulfate-reducing bacteria,
D. vulgaris and D. gigas, are similar. But m a n y more differences occur than would be expected from bacteria of the same genus. Thirty seven amino acids {about 70%) are found in identical positions; both have an N-terminal Met which is blocked in the D. gigas rubredoxin but not in the D. vulgaris rubredoxin. The N-terminal parts of these two proteins are identical between residues 4 and 16. 179
M. P I E R R O T E T AL.
180
The differences in the amino acid sequences are reflected in t h e i r functional relationships; an N A D H ~ H + rubredoxin oxido-reductase has been described b y Le Gall (1968) in ~). gigas. This enzyme is ten times more reactive with D. gigas rubredoxin t h a n with D. vulgaris rubredoxin and shows no activity toward rubredoxin from C. pasteurianum or P. aerogenes (H. D. Peck Jr, M. Odum, personal communication). I n order to explain such specific behavior we have undertaken the threedimensional structure determination of these rubredoxins. The proteins were isolated by the procedures of Le Gall & Dragoni (1966) for D. gigas and Bruschi & Le Gall (1972) for D. vulgaris. Both rubredoxins can be crystallized under the following conditions : 10 mg of protein were dissolved in 1 ml of sodium citrate buffer, 0.1 M at p H 4.0, solid a m m o n i u m sulfate was added, without stirring, to small samples (100/A) in order to cover a narrow range of salt concentration of about 25% (w/v). The closed tubes were maintained at 27~ and good-sized crystals (--~0.5 m m • 0.5 m m • 0-3 ram) developed within three days. X - r a y precession photographs of the diffraction pattern reveal t h a t crystals of the two rubredoxins belong to the monoclinic system, space group P21. The crystallographic unit cell parameters are listed in Table 1, together with those of rubredoxin C. pasteurianum (Watenpaugh et al., 1973). TABLE 1
Crystallographic unit cell parameters Source of rubredoxin
Space group
a(A)
b(A)
c(A)
fl or a(~
Z
V(A 3)
D. vulgaris
P21
20'00
41.50
24.41
107"6
2
19310
D. gigas
P21
19"75
41.70
24.38
108'1
2
19091
C. pasteurianum
R3
38.77
--
--
112.4
3
39187
Assuming two molecules of rubredoxin in the unit cell, one per asymmetric unit, the crystal volume per unit molecular weight (VM) is H I . 6 A 3 per dalton. This value for VM is very low as evidenced by a tabulation done by Matthews (1968), where the observed values of V~ range from 1.7 to 3.5/k 3 per dalton. VM for the C. pasteurianum rubredoxin is 2.2 A3/dalton which is close to the average value 2.37 A3/dalton determined b y Matthews' tabulation. With the plausible assumption t h a t the volume of the protein molecule is approximately the same for the three rubredoxins, the solvent cell content of rubredoxin in D. vulgaris and D. gigas ( ~ 2 7 % ) appears to be much lower t h a n t h a t for C. pasteurianum ( ~ 4 4 % ) . The high degree of similarity in the crystal cell parameters between rubredoxin from D. vulgaris and from D. gigas suggests t h a t the molecular packing arrangements in the two crystals are due to the involvement of invariant amino acid residues between the two tertiary structures. A comparison of the precession patterns of the principal zones (see Fig. 1) reveals substantial differences for m a n y of the diffracted intensities. The preliminary crystallographic data enhance the interest in determining the ~tructurr of both these rubredoxin~. Differences in terms of ~r as well ~s sequence
LETTERS
TO THE EDITOR
181
(a)
(b) FH~. 1. 16 ~ hk0 precession photographs of rubredoxin crystals. (a) D. giga,s; (b) D. vu/gar/8.
~82
M. P I E R R O T E T A L .
between the two proteins are probably related to local modifications in the poly:peptide side-groups producing the observed differences between the diffraction :patterns. Data sets for both rubredoxins have been collected to 1.5 A resolution on native protein crystals. Solution of the two structures will be attempted, separately, By rotation-translation functions of the structure of ruhredoxin from C. pasteurianum. .'It is not unreasonable to expect to solve the structure to near atomic resolution. 'Thus this will provide two independent determinations of the FeS4 complex geometry ,to compare with the FeS4 complex geometry of C. pasteurianum rubredoxin. These De~ulfovibrio rubredoxin crystals diffract extremely well, and it is anticipated that at least one of these structures will be determined to well beyond 1.0 A resolution. This work was supported by the Centre National de la Recherche Scientifique and by U.S. Public Health Services grant GM-13366 from the National Institute of Health. Centre des Mecanismes de la Croissance Cristalline Universit6 d'Aix-Marseille I I I Centre Saint J6rome 13397 Marseille/C6dex, France
M. PIERROT R . HASER M. FREY
Laboratoire de Chimie Bact6rienne CNRS Marseille, France
M. BRUSCHI J . LE GALL
Department of Biological Structure University of Washington, School of Medicine Seattle, Wash. 98195, U.S.A.
L. C. SIEKER L. H . J-ENSEN
Received 27May 1976 REFERENCES Bachmayer, H., Benson, A. M., Garrard, W. T., Whiteley, H. R. & Yasunobu, K. T. (1968a). Biochemistry, 7, 986-996. Bachmayer, H., Mayhew, S., Peel, J. & Yasunobu, K. T. (1968b). J. Biol. Chem. 243, 1022-1030. Bruschi, M. & Le Gall, J. (1972). Biochim. Biophys. Acta, 263, 279-282. Bruschi, M., Bonicel, J., Bovier-Lapierre, G. & Couchoud, P. (1976a). Biochim. Biophys. Aaa, in the press. Bruschi, M., Bonicel, J., Bovier-Lapierre, G. & Couchoud, P. (1976b). Biochem. Biophys. Res. Commun. in the press. Le Gall, J. (1968). Ann. Inst. Pasteur (Paris), 114, 109-115. Le Gall, J. & Dragoni, N. (1966). Biochem. Biophys. Res. Commun. 23, 145-149. Lovenberg, W. & Sobel, B. E. (1965). Prec. Nat. Acad. Sci., U.S.A. 54, 193-199. Matthews, B. W. (1968). J. Mol. Biol. 33, 491-497. Mayhew, S. G. & Peel, J. L. (1966). Biochem. J. 100, 80P. McCarthy, K. F. (1972). PhD. Thesis, George Washington University, Washington D.C. Meyer, T. E., Sharp, J. J. & Bartsch, R. G. (1971). Biochim. Biophys. Acta, 234, 266-269. Peterson, J. A., Basu, D. & Coon, M. J. (1966). J. Biol. Chem. 241, 5162-5164. Watenpaugh, K. D., Sieker, L. C., Herriott, J. R. & Jensen, L. H. (1973). Acta Crystallogr. sect. B29, 943-956.
