Vanadium Compounds

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on December 12, 2015 | http://pubs.acs.org Publication Date: December 10, 1998 | doi: 10.1021/bk-1998-0711.ch014

Chapter 14

Chloroperoxidase from Curvularia inaequalis: X-ray Structures of Native and Peroxide Form Reveal Vanadium Chemistry in Vanadium Haloperoxidases 1

1

2

Albrecht Messerschmidt , Lars Prade , and Ron Wever 1

Max-Planck-Institute für Biochemie, Am Klopferspitz 18 A, D-82152, Martinsried bei München, Germany E. C. Slater Institute, University of Amsterdam, Plantage Muidergracht 12, 1018 TV Amsterdam, The Netherlands

2

Crystal structures of different forms of the vanadium-containing chloro peroxidase from the fungus Curvularia inaequalis have been determined. The 2.03 Åcrystal structure (R = 19.7%) of the native enzyme reveals the geometry of the intact catalytic vanadium center. The vanadium is bound as hydrogen vanadate(V). It is coordinated by four nonprotein oxygen atoms and one nitrogen (NE2) atom from histidine 496 in a trigonal bipyramidal fashion. Three oxygens are in the equatorial plane and the fourth oxygen and the nitrogen are at the apexes of the bipyramid. In the 2.10 Åcrystal structure (R = 20.0%) of the azide enzyme complex the azide directly coordinates to the vanadium replacing the apical vanadium oxygen ligand. In the 2.24 Åcrystal structure (R = 17.7%) of the peroxide derivative the peroxide is bound to the vanadium in an η fashion after the release of the apical oxygen ligand. The vanadium is coordinated also by 4 non-protein oxygen atoms and one nitrogen (NE2) from histidine 496. The coordination geometry around the vanadium is that of a distorted tetragonal pyramid with the two peroxide oxygens, one oxygen and the nitrogen in the basal plane and one oxygen in the apical position. The 2.50 Åcrystal structure (R = 17.1%) of apo CPO shows that a water molecule is present at the position of the vanadate cofactor. Tungstate is bound in a similar fashion as the hydrogen vanadate(V) group in the native form as determinedfromthe 2.30 Åcrystal structure of tungstate CPO (R = 16.2 %). Amino acid sequence comparisons show that the fold of the C terminal halves of CPO and of vanadium containing haloperoxidases of marine algae and the architecture of their vanadium binding sites are very similar. A mechanism for the catalytic cycle has been proposed based on these X-ray structures and kinetic data. 2

186

©1998 American Chemical Society In Vanadium Compounds; Tracey, A., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

187 The bioinorganic chemistry of vanadium has been advanced by the recent first x-ray structure determination of a vanadium containing protein, a chloroperoxidase from the fungus Curvularia inaequalis (7). This chloroperoxidase (CPO) belongs to a group of vanadium containing haloperoxidases that oxidize halides (CI", Br', Γ) in the presence of hydrogen peroxide to the corresponding hypohalous acids according to : +

H 0 + Χ" + H -> H 0 + HOX. Downloaded by UNIV OF CALIFORNIA SAN DIEGO on December 12, 2015 | http://pubs.acs.org Publication Date: December 10, 1998 | doi: 10.1021/bk-1998-0711.ch014

2

2

2

The hypohalous acid, HOX, will react to a diversity of halogenated products if a convenient nucleophilic acceptor is present. The vanadium haloperoxidases are isolated primarily from marine algae, although they have been identified and characterized from a terrestrial lichen and fungi (2-5). Many of the organohalogens generated by these enzymes are biocidal and thus may provide defense functions. The marine haloperoxidases (for a review see Butler and Walker (6)) are widespread in the marine environment and are probably responsible for the formation of huge amounts of organohalogens like bromoform which are released to the atmosphere. The bromoperoxidase (BPO) from the seaweed Ascophyllum nodosum has been best characterized. The molecular mass of the monomer is ca. 65000 which is comparable to that of the monomer of the CPO of this study. BPO has been crystallized some times ago (7) and the crystal structure analysis is almost finished (Schomburg, D., University of Koln, personal communication, 1997.) but not published yet. Another haloperoxidase from the seaweed Corallina officinalis has been crystallized as well (8). The vanadium bromoperoxidases have been studied in great detail using a variety of biophysical techniques (9) including EXAFS (10, 11) and E S E E M (72, 75). Vanadium(IV) or (III) states have not been observed by EPR or K-edge x-ray studies in the presence of substrates or during turnover and the reduced enzyme is inactive. Apparently the redox state of the metal does not change during catalytic turnover and a model has been suggested (75) in which the vanadium site functions by binding hydrogen peroxide to yield an activated peroxo-intermediate which is able to react with bromide to produce HOBr. Similar models have now been proposed (14, 15) for a number of metal-catalyzed oxidations of bromide by hydrogen peroxide. A detailed kinetic study of the formation of HOC1 by the chloroperoxidase (16) revealed many similarities with the kinetics of the vanadium bromoperoxidase. Azide was found to be a weak inhibitor for the vanadium-containing marine bromo peroxidase (7 7). Overall Protein Structure of C P O from Curvularia inaequalis The crystal structure of CPO from Curvularia inaequalis in the complex with azide at an resolution of 2.1 Â (7 ) and of the native form at a resolution of 2.03 Â (18 ) have been published and revealed the architecture of the protein part and the coordination of the vanadium metal center. The structure solved by multiple

In Vanadium Compounds; Tracey, A., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1998.


188 isomorphous replacement (MIR) techniques reveals one molecule per asymmetric unit of the trigonal crystals (space group R3 with lattice constants a-b= 131.69 Â, c= 112.97Â, α = β = 90°, γ = 120°). The molecule has an overall cylindrical shape with a length of about 80 Â and diameter of 55 Â (Fig. 1). The molecules are arranged in the crystals as trimers around the crystallographic threefold axis. The secondary structure, which was analyzed with the program DSSP (20), is mainly helical (about 44% of the atomic model) consisting of 20 helices, a small part of βstructures ( 6 short β-strands, pairwise arranged as three antiparallel β-ladders) and the rest of extended strand and loop regions (Fig. 2 ). Two four-helix bundles (ahelices B , C, E, F and α-helices K , L , N , O) are the main structural motifs of the tertiary structure and are spatially related by a rotation of about 180° around an axis perpendicular to the sketch plane in Fig. 2. These helices are between 14 and 31 residues long. α-Helices G and S are each packed against one of the four-helix bundles and α-helices I and J are located on top of the four-helix bundles and approximately perpendicular to their helical axes. There are 8 short helices found in connecting segments. Four of them are α-helical (H, M , P, Q in Fig. 2) and the other four are 3 i helices (A, D, R, Τ in Fig. 2). At a first glance, the compact α-helical structure of CPO resembles those of the R2 subunit of ribonucleotide reductase from Escherichia coli (21) and the first domain of the α-subunit and the β-subunit of the hydroxylase protein of methane monooxygenase from Methylococcus capsulatus (Bath) (22), both non-heme di-iron proteins. However, the main motif in both non-heme di-iron proteins is an eight-helix bundle with a special topological sequence and arrangement of the α-helices different from that found here. Recently, the crystal structure of a bacterial bromoperoxidase (23) has been determined. This enzyme lacks cofactors and shows the general topology of the α/β hydrolase fold. Apparently, the vanadium chloroperoxidase and the bacterial bromoperoxidase lack any structural similarities and are very different enzymes. 0

Vanadium Binding Site The vanadium center is located on top of the second four-helix bundle (Fig. 1) and the residues forming the metal binding site are coming from helices J, L , Ν and Ο and from turns 290-294, 355-363, 395-406 and 492-502 (Fig. 3). Thus, all residues building up the metal center are from the C-terminal half of the molecule. The difference electron density map of the native form at the vanadium site (Figure 4) as reported in (18) obtained by refining the model of the azide form without the vanadate and azide groups against the observed structure factors of the native form clearly shows the vanadium bound to 4 oxygen atoms forming an orthovanadate group and to the NE2 nitrogen atom of histidine 496. The vanadium coordination geometry is trigonal bipyramidal with 3 nonprotein oxygen atoms in the equatorial plane (bond lengths about 1.65 Â), one nonprotein apical oxygen atom (bond length 1.93 Â) and the other apical nitrogen atom of histidine 496 (bond length 1.96 Â). The



189

Fig. 1. Ribbon-type representation of the CPO molecule (MOLSCRIPT (79)).



190

Fig. 2. Schematic drawing of the secondary structure elements of CPO. All helices are drawn as cylinders; β-strands are depicted as arrows. The Ν and C termini are indicated.



Fig. 3. Vanadium binding site of the native form of CPO with hydrogen bond pattern (MOLSCRIPT (79)).



Fig. 4. Difference electron density map of the native form of CPO around the vanadium binding site contoured at 3.0 σ. The difference electron density for the hydrogen orthovanadate(V) is the strongest feature in the whole map.


193 bond length of the apical oxygen OV4 to the vanadium is with 1.93 Â in the range of an O H ligand, indicating that the V 0 group is bound as hydrogen vanadate(V). The negative charge of the hydrogen vanadate(V) group is compensated by hydrogen bonds to surrounding positively charged or hydrophilic protein functions. Oxygen OV1 of the V 0 group makes hydrogen bonds to nitrogens NE1 of Arg360 (2.98 Â) and NE2 of Arg490 (3.03 Â), oxygen OV2 to nitrogen N Z of Lys353 (3.19 Â) and nitrogen Ν of Gly403 (2.98 A), oxygen OV3 to oxygen O G of Ser402 (2.79 A) and nitrogen N E of Arg490 (3.20 A). The apical oxygen OV4 forms three hydrogen bonds, two to water molecules (SOL 1440, 2.59 A; SOL 1262, 2.92 A), one to nitrogen ND1 of histidine 404 (2.97 A). The latter one seems to be important because histidine 404 has been identified as being necessary for the catalytic activity (see (7)). Binding of peroxide is inhibited when histidine 404 is doubly protonated. In the azide CPO complex (the crystallization buffer contained 2 m M azide for this crystal form), the azide coordinates directly to the vanadium replacing the apical oxygen atom OV4. A water molecule from solvent is hydrogen bonded to the azide nitrogen atom N l which coordinates to the vanadium. 4


4

The binding site for the hydrogen-vanadate(V) is a pocket preformed by the protein which is located at the end of a broad channel filled by solvent molecules. This channel supplies good access and release for the substrates and products of the enzymatic reaction. One contiguous half of its surface is mainly hydrophobic with Pro47, Pro211, Trp350, Phe393, Pro395, Pro396 and Phe397 as contributing side chains. The other half is predominantly polar with several main chain carbonyl oxygens and the ion pair Arg490 - Asp292. It is intriguing that the vanadium(V) is bound directly as hydrogen-vanadate(V) with one immediate protein ligand only. Peroxide Form of C P O The peroxide complex as reported in (18) was produced by soaking the crystals in mother liquor [2.0 M ( N H ) S 0 in 0.1 M Tris-H S0 , pH 8.0], containing 20 m M H 0 , for two hours. The X-ray measurements were done on a Hendrix/Lentfer X ray image-plate system (Mar-Research, Hamburg, Germany) mounted on a Rigaku rotating anode generator operated at 5.4 kW (λ = CuK„ = 1.5418 A). The peroxide crystals were measured at 100 K . For the cryo-measurement the peroxide soaked crystals were brought into a cryo-buffer (20 m M peroxide, 1.7 M ( N H ) S 0 in 0.1 M Tris-H S0 , pH 8.0, 30.1% glycerol), transferred into a cryo-loop, shock frozen and kept in a stream of evaporating liquid nitrogen using a cryo-device (Vectotherm, Karlsruhe) mounted on the image-plate system. The peroxide derivative diffracts to 2.24 A resolution with satisfying data completeness (83.8 %). The crystal structure was solved by difference Fourier techniques using the phases of the isomorphous native form and refined to a crystallographic R-factor of 17.7% (18). The difference electron density map of the peroxide form at the vanadium site reported in (18) is shown in Fig. 5. The phases used for the calculation of this map 4

2

2

4

2

4

2

4

2

2

4


4


Fig. 5. Difference electron density map of the peroxide form of CPO around the vanadium binding site contoured at 3.0 σ. The difference electron density for the peroxide metavanadate(V) is the strongest feature in the whole map.



195 have been obtained from an atomic model which did not contain the vanadate group and the two water molecules 1420 and 1377. The difference electron density map demonstrates that the peroxide has reacted with the vanadate group. The apical oxygen OV4 has been released and the peroxide binds side-on in the equatorial plane to the vanadium. Positive electron density is visible for two additional water molecules in the active site (1165, 1420). The vanadium peroxide complex has 5 direct ligands only. The coordination geometry is that of a distorted tetragonal pyramid. The apical ligand is oxygen OV3 (bond length about 1.60 Â) identifying this as a V=0 bond. The two peroxide oxygens OV2 and OV4 (bond lengths: V-OV2 and V-OV4, about 1.87 A ; OV2-OV4, 1.47 Â), oxygen OV3 (bond length 1.93 Â), and nitrogen NE2 from histidine 496 (bond length 2.19 Â) constitute the basal plane. One empty apical coordination site at the vanadium is sterically blocked by the side chain of Arg360. The empty coordination site generated by the release of OV4 seems to be predestinated to accept the chloride ion during continuation of turnover. The bound peroxo species may be addressed as a monoperoxo-metavanadate(V). The hydrogen bonding network around the vanadium site of the peroxide derivative is depicted in Fig. 6. His404 is no longer hydrogen-bonded to any oxygen function of the vanadium group. It forms a hydrogen bond to water molecule 1420. OV4 of the peroxide group is hydrogen-bonded to N Z of Lys353 and to the amide nitrogen of Gly403. OV2 of the peroxide group is linked via a hydrogen-bond to the same amide nitrogen. OV3 makes hydrogen-bonds to OG of Ser402 and N E of Arg490, and OV1 to NH1 of Arg360 and NH2 of Arg490. The empty vanadium coordination site generated by the release of OV4 is directed towards the active site pocket supplying good access for the second substrate from the solvent. Apo and Tungstate Forms of C P O The x-ray structures of the apo and tungstate form of vanadium chloroperoxidase (CPO) from the fungus Curvularia inaequalis have been solved by difference Fourier techniques using the atomic model of native chloroperoxidase (24). In the 2.50 Â crystal structure (R = 17.1 %) of apo CPO, a water solvent molecule is found in place of the vanadate cofactor. The 2.30 Â crystal structure of tungstate CPO (R = 16.2 %) reveals the binding of the tungstate to the metal-cofactor binding site. It binds in a similar fashion as the hydrogen vanadate(V) group in the native form. But the tungsten atom is not or only weakly bound to the NE2 nitrogen atom of His496 in contrast to native CPO where the vanadium atom forms a bond to the NE2 nitrogen atom of histidine 496. The overall protein structure and the metal-cofactor binding site of apo and tungstate CPO remain virtually unchanged. This means that the CPO protein matrix provides a rigid preformed metal-cofactor binding site. The binding mode of vanadate or tungstate can be described as rack-induced bonding according to Gray and Malmstrôm (25).



Fig. 6. Hydrogen bonding pattern of the peroxide form of CPO around the vanadium binding site (MOLSCRIPT (79)).


197


Relatedness of CPO with Other Vanadium Containing Haloperoxidases Vanadium containing haloperoxidases have been isolated from all classes of marine algae (see (6)). A n amino acid sequence alignment of the sequence of CPO with the recently determined partial sequence (236 residues) of the vanadium-containing bromoperoxidase from the brown alga Ascophyllum nodosum (26) has been carried out and the regions of high similarities are shown in Table I. There are three stretches of high similarity in the regions of residues providing the vanadium histidine ligand (His496), the vanadate-interacting residues (Lys353, Arg360, Ser402, Gly403, Arg490) and the catalytic His404. The alignment demonstrates that the metal binding sites in both vanadium-containing haloperoxidases and the overall protein fold in the C terminal half of the proteins will be very similar reflecting the close relationship of both enzymes. This will also be valid for the vanadium containing haloperoxidases from the other marine algae. Table I. Amino Acid Sequence Alignment of Stretches of High Similarities of CPO (27) and Vanadium Containing Bromoperoxidase from Ascophyllum nodosum (BRPO) (26). The alignment was carried out manually. Residue 24 of the BRPO partial sequence was published to be an Asn (26) but turned out to be a Lys after resequencing (Schomburg, D., University of Kôln, personal communication, 1997). *, Protein ligand directly bound to vanadium; Δ, residue hydrogen bonded to vanadate oxygen; • , catalytic histidine. CPO 343 TDAGIFSW K E K W E F- EFWRP LS-GV BRP 14 ELAQRSSWYQKWQVHRFARPEALGG Consensu A SW KW F RP G s Function Δ Δ CPO 393 FKPPFPAYPSGHATFGGAVFQMVRRY BRPO 90 GTPTHPSYPSGHATXNGAFATVLKAL Consensu Ρ Ρ YPSGHAT G A s Function ΛΛ + CPO BRPO Consensu s Function

480 152

WELMFENAIS RI F L G V H W R F D A A A KKLAVNVAFGRQMLGIHYRFDGIQ L A R L G H RFD Δ

*

It has recently been shown that the amino acid residues contributing to the active sites of the vanadium containing haloperoxidases are conserved within three families of acid phosphatases suggesting that the active sites of these enzymes are very similar (28).


198


Proposal of a Catalytic Mechanism for C P O The determination of the native form and the peroxide intermediate by X-ray structure analysis made it possible to propose a catalytic reaction scheme (18). Steady-state kinetic studies on CPO (16) had shown that in the pH range between 6 and 7 the enzyme mechanism is a ping-pong type. It has also been demonstrated that peroxide binds first. The proposed catalytic mechanism (18) is displayed in Fig. 7. We start from the native enzyme (panel 1). The apical oxygen is hydrogen-bonded to His404. This hydrogen-bond makes the apical OH-group more nucleophilic than a normally bound O H . The peroxide molecule approaches this apical O H and gets singly deprotonated. The generated apical water molecule is a weak vanadium ligand only and may leave the vanadium coordination sphere (panel 2). The hydroperoxide coordinates to the vanadium at this empty coordination site (panel 3) and the more nucleophilic oxygen OV2 abstracts the proton from the peroxide. The OH-ligand is displaced by the now more nucleophilic negatively charged peroxide oxygen (panel 4). The peroxide is now bound as in the peroxide complex which is observed in the crystal structure. At this stage the chloride ion binds to the empty vanadium coordination site (panel 5 and 6). In the original paper describing the structure of the azide CPO complex (1) a possible chloride binding site close to Trp350 and Phe397 in the active site and near the vanadium binding site has been discussed. Both hydrophobic side chains provide a hydrophobic environment, which seems to be necessary to stabilize the chloride binding. This has been found in chloride binding sites of various amylases where one side chain close to the chloride ion is donated from a hydrophobic residue (29). Similarly, in the haloalkane dehalogenase of Xanthobacter autotrophicus two tryptophan residues are present in halogen and halide binding (30). However, the coordination geometry of the peroxide intermediate in the peroxide CPO crystal structure makes a direct coordination of the chloride to the vanadium very probable. The bound peroxide is an oxidizing species and accepts two electrons from the chloride. The peroxide bond will be broken after the acceptance of the first electron and the O-Cl bond will be formed, and the V - 0 bond of the upper peroxide oxygen will be broken after uptake of the second electron. The OC1" molecule will take up a proton from a surrounding water molecule and will leave the active site as HOC1 (panel 7). A n alternative possibility is that the chloride ion reacts directly with one of the presumably polarized peroxide oxygen atoms. This results in vanadium bound OC1" which also takes up a proton from a water molecule. The generated OH" will coordinate to the vanadium site to rebuild the native state (panel 8). The formation of the peroxide intermediate should be impossible when His404 is doubly protonated. In this case the apical OH group could no longer form a hydrogen bond to His404 and would lose its ability to activate the peroxide by its deprotonation.


199

^N404

Φ Ν4Ω4

OV3


««η

Ν

ρ ·

·

0V2

^

Τ *

ΦΝ404

Ν404

-

['4

^ ^ ^ Η ^ - * -

*Ν 04 4

=

1

Ν

ΟΗ'

404

^Ν 04 4

^

Λ ^

Η

ο

α

Fig. 7. Proposal for the catalytic mechanism of CPO from Curvularia

inaequalis.


200


Acknowledgments. We thank Prof. R. Huber for supporting this project and for valuable suggestions, Dr. J.W.P.M. van Schijndel for supplying the purified CPO and Ms. E.G.M. Vollenbroek for her help in culturing fungi. A.M. thanks the Deutsche Forschungsgemeinschaft (Schwerpunktthema: Bioanorganische Chemie) for financial support. This work was also supported by grantsfromthe Netherlands Foundation for Chemical Research (SON) and was made possible by financial support from the Netherlands Organization for Scientific Research (NWO) and the Netherlands Technology Foundation (STW).

Literature Cited. 1. Messerschmidt, Α.; Wever, R. Proc. Natl. Acad. Sci. USA 1996, 93, 392. 2. Vilter, H. Phytochemistry 1984, 23, 1387. 3. de Boer, E.: Tromp, M.G.M.; Plat, H.; Krenn, B.E.; Wever, R. Biochim. Biophys. Acta 1986, 872, 104. 4. Plat, H.; Krenn, B.; Wever, R. Biochemical J. 1987, 248, 277. 5. Vollenbroek, E.G.M.; Simons, L.H.; van Schijndel, J.W.P.M.; Barnett, P.; Balzar, M.; Dekker, H.; Wever, R. Biochem. Soc. Trans. 1995, 23, 267. 6. Butler, Α.; Walker, J.V. Chem. Rev. 1993, 93, 1937. 7. Müller-Fahrnow, Α.; Hinrichs, W.; Saenger, W.; Vilter, H. FEBS Lett. 1988, 239, 292. 8. Rush, C.; Willetts, Α.; Davies, G.; Dauter, Z.; Watson, H.; Littlechild, J. FEBS Lett. 1995, 359, 244. 9. Wever, R.; Kustin, K. Adv. Inorg. Chem. 1990, 35, 81. 10. Arber, J.M.; de Boer, E.; Garner, C.D.; Hasnain, S.S.; Wever, R. Biochemistry 1989, 28, 7678. 11. Carrano, C.J.; Moham, M.; Holmes, S.M.; de la Rosa, R.; Butler, Α.; Charnock, J.M.; Garner, C . D . Inorg. Chem. 1994, 33, 646. 12. de Boer, E.; Keijzers, C.P.; Klaasen, A.A.K.; Reijerse, E.J.; Collison, D.; Garner, C.D.; Wever, R. FEBS Lett. 1988, 235, 93. 13. de Boer, E.; Boon, K.; Wever, R. Biochemistry 1988, 27, 1629. 14. Meister, G.E.; Butler, A. Inorg. Chem. 1994, 33, 3269. 15. Espenson, J.H.; Pestovsky, O.; Huston, P.; Staudt, S. J. Amer. Chem. Soc. 1994, 116, 2869. 16. van Schijndel, J.W.P.M.; Barnett, P.; Roelse, J.; Vollenbroek, E.G.M.; Wever, R. Eur. J. Biochem. 1994, 225, 151. 17. Vilter, H. Proc. Int. Conf. Coord. Chem. 17th, 1989, M62 (abstr.). 18. Messerschmidt, Α.; Prade, L.; Wever, R. Biol. Chem. 1997, 378, 309. 19. Kraulis, P. J. Appl. Crystallogr. 1991, 24, 946.



201 20. Kabsch, W.; Sander, C. Biopolymers 1983, 22, 2577. 21. Nordlund, P.; Sjöberg, B.-M.; Eklund, H. Nature 1990, 345, 593. 22. Rosenzweig, A.C.; Frederick, C.A.; Lippard, S.J.; Nordlund, P. Nature 1993, 366, 537. 23. Hecht, H.J.; Sonek, H.; Haag, T.; Pfeifer, O.; van Pée, K.-H. Nat. Struct. Biol. 1994, 1, 532. 24. Messerschmidt, Α.; Wever, R. Inorg. Chim. Acta 1998, in press. 25. Gray, H.B.; Malmström, B.G. Comments Inorg. Chem. 1983, 2, 203. 26. Vilter, H. In Metal Ions in Biological Systems; Sigel, H.; Sigel, Α.; Dekker: New York, 1995, Vol. 31; pp 326-362. 27. Simons, B.H.; Barnett, P.; Vollenbroek, E.G.M.; Dekker, H.L.; Muijers, A.O.; Messerschmidt, Α.; Wever, R. Eur. J. Biochem. 1995, 229, 566. 28. Hemrika, W.; Renirie, R.; Dekker, H.L.; Barnett, P.; Wever, R. Proc. Natl. Acad. Sci. USA 1997, 94, 2145. 29. Machius, M.; Wiegand, G.; Huber, R. J. Mol. Biol. 1995, 246, 545. 30. Verschueren, K.H.G.; Kingma, J.; Rozeboom, H.J.; Kalk, K.H.; Janssen, D.B.; Dijkstra, B.W. Biochemistry 1993, 32, 9031.


Vanadium Compounds

Recommend Documents