chem. Rev. 1998, 93, 1937-1944

Marine Haloperoxidases

Alison Butler’ and J. V. Walker

oepemnmt of ~%~II/SIIY. (~vasllyof cawom$. anta &&am. CaBlanh, 93106

Reodved A@ 20, 1993 (Revlsed Manusmt Racslved May 20, 1993)

Contents I. Introduction 1937 A. Halogenated Compounds in the Marine 1937 Envlronment p: ? E. Marine Habperoxidases 1938 6 1. Types of Marine Haioperoxidases 1938 8. 2. Standard Assay of Habperoxidase 1939 Activity 11. Vanadium Bromoperoxidase 1939 A. Characterlstics of V-ErPO 1939 1. Isoiatlon and Purlflcation of VgrpO 1939 2. The Vanadium Sne 1939 ...... :.;...... E. Reactivity Of V-ErpO 1940 . . :. .. . :. . [.. 1. Halogenation and HalldeAssisted 1940 AII~~utler received her &A. from Reed College in 1977 and her Disproportionation of Hydrogen P~.D.from the University of California. San Diego. in 1962. Alter Peroxide pwMoaaal stays at UC Los Angeles and Caitech. she joked the 2. Specific Activltles 1941 faculty at UC Santa Barbara in 1986. She is a recent reciplent of an American Cancer Society. Junior Faculty Research Award, 1941 3. Enzyme Kinetics an Alfred P. Sbn Foundatlon Fellowship, and the UCSE PIOUS 4. Peroxide Selectivity 1941 TeacherlSchoiar Award. Her research interestsarein bioinaganic 5. Mechanistic Considerations 1941 chemistry and metalloblochemistry. including that of the marlne environment. C. Stability of V-ErPO 1942 D. Comment on the Putative Non-i-bma Iron 1942 Eromoperoxidases 111. Fell“ Bromoperoxidase 1942 A. Characteristics of FeHeme 1942 Eromoperoxbse E. Reactivity of FeHeme Bromoperoxidase 1942 1. Biosynthetic Investigations 1943 2. Other FeHema Haloperoxidases 1943 1V. Future Directions 1944

I. Introductlon

A. Halogenated Compounds in the Msrlne Environment Jenylalne Vanessa Walker was born in St. John’s. Antigua, West Indles.in 1967. Shereceived her6.S. in ChemisbyfromCalifmia Ocean water is approximately 0.5 M in chloride, 1 State University, Bakersfield. in 1989. She is currently working on mM in bromide, and 1 pM in iodide. Given the high her Ph.D. at the University of California. Santa Barbara, under tlm halogen content, it is not surprising that marine dlrectlon of Prof. Alison Butler. Ourlng her time at UCSB she has received a University Regents Fellowship, a UCSE Graduate organisms have developed means to incorporate halo- Opportunity Fellowship. and a UCSB Graduate Reaearch Mentwship gens into their metabolites. Many of these halogenated Fellowship. Her general research interests include biolnorganic compounds are thought to be involved in chemical chemistry, natural products chemistry, and marine biochemistry. defense roles to keep predators away from a particular organism. Inmany cases these compounds are also of volatile halohydrocarbons produced in very large pharmacological interest due to their biological activ- quantities, such as bromoform, dibromomethane, meth- ities, which include antifungal, antibacterial, antine- yl iodide, bromo- and chloroanisoles, etc.14 Others oplastic, antiviral (e.g., anti-HIV), antiinflammatory, include phenolic derivatives in which the biosynthesis and other activities. of these compounds can be readily envisaged, as in The halogenatedmarine natural products encompass aeroplysinin-1 (see 1 in Figure I), an antimicrobial a very wide range of compounds. Some are simple metabolite from sponges? or 14-debromoprearaplysillin (2) from the marine sponge Druinella purpurea? both To whom mrrespondencs should be addread. of which could originate from tyrosine (see below), or 0009-2665/93/0793-1937$12.00/0 0 I993 American Chemical SocW 1938 Chemical Reviews, 1993, Vol. 93, NO. 5 Butler and Walker

OMe Br Br bromoperoxidase catalyzes the oxidation of bromide and iodide by hydrogen peroxide, while iodoperoxidase catalyzes the oxidation of only iodide by hydrogen peroxide. A better system of nomenclature would be based on the enzyme’s physiological role, although at this point, that is hard to define given that the 1 concentrations of halides, hydrogen peroxide, and nature of the endogenous organic substrate are not known in many marine organisms.

1. Types of Marine Haloperoxidases Haloperoxidases have been isolated from all classes of marine algae and many other marine organisms (Table I); in addition haloperoxidase activity has been detected in many other species of algae13 and in other marine organisms. Two types of marine haloperoxi- 6 dases have been identified: (1) vanadium bromoper- 5 oxidase (V-BrPO),a non-heme enzyme, and (2) FeHeme Figure 1. Examples of marine natural products with bromoperoxidase (FeHeme-BrPO). In some cases, a important biological activities. marine haloperoxidase has been isolated, but the vanadium or FeHeme content has not been determined, the antimicrobial compound 2-(2’-bromophenoxy)- as in the marine snail Murex truncuZus.30 In addition 3,4,5,6-tetrabromophenol(3)from the sponge Dysidea.’ to these enzymes, other terrestrial haloperoxidases are Numerous chiral chlorinated and brominated terpen- known, including the first discovered haloperoxidase, oids have also been isolated, including solenolide E (4), chloroperoxidase, the FeHeme enzyme from the fungus a diterpene which is an antiinflammatory and antiviral Caldaromyces f~mago,~~mammalian haloperoxidases compound isolated from a gorgonian.8 One of the more such as eosinophil peroxidase32 and myeloper~xidase~~ famous brominated marine natural products is tyrian found in human white blood cells, salivary and lac- purple (5), isolated from a marine mollusk. Other toperoxidase found in saliva and tearsF2 and bacterial brominated and chlorinated indoles, e.g., 5-bromo-NJV- haloperoxidases (bromoperoxidase from Streptomy- dimethyltryptamine (6), from a marine ~ponge,~are ces33 and chloroperoxidase from Pseudom0nads3~)for also known. A comprehensive summary of halogenated which the identity of the active-metal species, if any, compounds is contained in a series of reviews of marine has not been determined. Recently a new enzyme, natural products by Faulkner.lo methyl transferase, was isolated from marine alga Marine natural products chemists have long invoked (Endocladia muricata) that catalyzes the formation of the role of haloperoxidases in the biogenesis of the methyl chloride.35 Methyltransferase uses S-adenosyl- halogenated marine natural products. Bromoperoxi- methionine as the carbon source. dase and iodoperoxidase activity have been well doc- umented in marine algae since the turn of the centu- Table I. Sources of Marine Haloperoxidases ry,l1J2 Haloperoxidases have been found in virtually ref(s) all classes of marine organisms. Early on, Hewson and ~~ Hager established that many species of marine algae vanadium bromoperoxidase have bromoperoxidase activity in aqueous algal ex- I. Rhodophyta (red algae) Corallina officinalis 14 tract~.~~Rhodophyta were the richest in both bro- Corallina pilulifera 15,16 moperoxidase activity and lipid halogen content; Pheo- Corallina uancouveriensis 17 phyta were the poorest. Marine haloperoxidases have Ceramium rubrum 18 now been isolated and purified to homogeneity; thus 11. Phaeophyta (brown algae) the scope of this review article will be the reactivity Alaria esculenta 19 Ascophyllum nodosum 20 and characterization of the marine haloperoxidases. Chorda filum 21 Fucus distichus 22 Laminaria digitata 23 B. Marlne Haloperoxldases Laminaria sacchirina 21 Macrocystis pyrifera 22 Haloperoxidases are enzymes that catalyze the ox- 111. Chlorophyta (green algae) idation of a halide (i.e., chloride, bromide, or iodide) by Halimeda sp. 24 hydrogen peroxide, a process which results in the FeHeme bromoperoxidases concomitant halogenation of organic substrates. I. Rhodophyta Cystoclonium purpureum 25 Rhodomela larix 26 Org-H + X- + H20, + H+ - X-Org + 2H20 11. Chlorophyta Penicillus capitatus 21,28 Penicillus lamourouxii 28 The nomenclature for the haloperoxidases has tra- Rhipocephalus phoenix 28 ditionally been based on the most electronegative halide 111. Marine worm which is able to be oxidized by HzO2 catalyzed by the Notomastus lobatus 29 enzyme. Thus chloroperoxidasecatalyzes the oxidation Ptychodera flavin laysanica 26 of chloride, bromide, and iodide by hydrogen peroxide; Thelepus setosus 26 Marine Haloperoxidases Chemical Reviews, 1993, Vol. 93, No. 5 1090 be achieved by addition of excess vanadate and sub- sequent removal of adventitiously bound vanadium(V) by dialysi~.~~?~~~~~While the physical characteristics of V-BrPO isolated from marine algae are all very similar, some differences in reactivity have been observed, such Figure 2. Halogenation of monochlorodimedone (MCD) as specific activity (see below). V-BrPO (Ascophyllum catalyzed by haloperoxidases. nodosum) has been crystallized, although refined structural data have not been reported yet.39 The 2. Standard Assay for Haloperoxidase Activity crystals defract to 2.4-A resolution. Four molecules are The standard assay for haloperoxidase activity is the present per asymmetric subunit. halogenation of monochlorodimedone (MCD; 2-chloro- Isozymes of vanadium bromoperoxidases from A. 1,&dimedone) using hydrogen peroxide as the oxidant nodosum have been isolated which differ in carbohy- of the halide (Figure 2).31 drate ~ontent.'~~~~The more abundant bromoperoxi- The halogenation of MCD is followed spectropho- dase based on isolated yield, V-BrPO-I, was found in tometrically at 290 nm, which monitors the loss of MCD the thallus and the other bromoperoxidase, V-BrPO- in the enolform (t = 20 000M-l cm-l). Bromoperoxidase 11, was reported to be present on the thallus surface.37 activity is expressed as micromoles of MCD brominated A previous study reported that V-BrPO is present in per minute per milligram of enzyme (i.e., units/mg). two different locations of A. nodosum, one in the cell The early work on V-BrPO employed the oxidation of walls of the transitional region between the cortex and iodide by hydrogen peroxide,20 forming triiodide (I3-) medulla of the thallus and the other in the cell wall of which was followed spectrophotometrically at 353 nm the thallus surface.40 More recent experiments dem- (t = 26 400 M-l cm-l). However, this reaction is less onstrate that vanadium-dependent bromoperoxidase desirable for quantitation of haloperoxidase activity activity is present in both the cortical and surface because of competing side reactions, such as the protoplasts of Macrocystis pyrifera,4l Laminaria sac- nonenzymatic oxidation of iodide by hydrogen peroxide charina, and Laminaria digit~ta.~~The biosynthesis and reduction of triiodide by hydrogen peroxide. of V-BrPO in the protoplasts of Laminaria saccharina The formation of bromochlorodimedone and dichlo- has been shown using [35S]methionine.42 rodimedone have been verified by chromatographic analysis and comparison with an authentic sample of 1. Isolation and Purification of V-BrPO Bromochlorodimedoneis not stable at neutral Isolation and purification of V-BrPO from A. no- pH over long periods of time, but it is stable at low pH. dosum begins by homogenization of the macroalga in Thus the bromoperoxidase reactions run at neutral pH an equalvolume of 0.2 M Tris-sulfate (pH 8.3), followed were adjusted to pH 3 at the end of the reaction in by centrifugati~n.l~?~~@~The pellet is collected and order to isolate bromochl~rodimedone.~~ reextracted until the enzymatic activity of the reex- tracted pellet has decreased significantly. The extracts II. Vanadium Bromoperoxidase (typically 20-60 L for 1-4 kg of algae) are then pooled and the alginic acid precipitated by making the solution A. Characterlstlcs of V-BrPO 0.1 M in calcium chloride. The calcium alginate is Vanadium bromoperoxidases are all acidic pro- conveniently removed by centrifugation using a con- tein~~~~~with very similar amino acid composition3 tinuous flow centrifuge; the supernatant volume is then (see Table 11), molecular weight, charge (PI 44,and reduced in a hollow fiber concentrator prior to protein vanadium content. The subunit molecular weight of precipitation by 80 % ammonium sulfate. The ammo- V-BrPO is ca. 65 000. As isolated V-BrPO contains only nium sulfate precipitate is then collected and dialyzed about 0.4 vanadium atoms per subunit. However, a against 0.1 M Tris-sulfate (pH 8.3) before batch loading content of one gram-atom of vanadium per subunit can onto an anion exchange column (e.g., DEAE). Active fractions from the anion exchange column are pooled Table 11. Percent Amino Acid Composition of Marine and subjected to a series of gel filtration columns (e.g., Haloperoxidases sepharose C1-6B, sephacryl S200). The activity and amino A. nodosum L. saccharina C. ilulifera P. capitatus purity of the enzyme are monitored throughout the acid V-BrPO V-BrPO f-BrPO FeHeme-BrPO course of the purification procedure by the MCD assay Asx 14.0 12.9 11.4 12.8 and SDS-PAG electrophoresis, respectively. An im- Thr 6.9 6.4 4.9 7.1 proved isolation procedure based primarily on a two- Ser 5.0 5.8 6.7 5.4 Glx 10.2 9.1 10.2 11.9 phase extraction system (i.e., poly(ethy1ene glycol) and Pro 4.9 5.2 5.3 0.5 aqueous potassium carbonate) has been described for GlY 9.8 8.9 7.8 4.1 V-BrPO from Laminaria digitata and Laminaria Ala 11.1 8.0 10.0 6.0 Val 6.2 5.5 6.5 8.3 sacch~rina.~~~*This procedure, however, is not effec- Met 1.0 2.0 1.4 0.7 tive in the isolation of V-BrPO from A. nodosum,which Ile 4.0 2.7 5.9 6.3 has been the principal source of V-BrPO for the Leu 8.5 9.4 9.2 10.1 mechanistic studies that have been reported to date. Tyr 1.7 3.1 2.2 2.3 Phe 5.7 3.5 5.7 9.9 LYB 1.6 3.7 4.3 1.1 2. The Vanadium Site His 1.2 1.7 1.4 1.2 4.0 4.7 1.4 7.1 The resting oxidation of vanadium in V-BrPO is CYS 1.7 5.5 5.1 4.6 vanadium(V). From extended X-ray absorption fine Trp nd nd 0.8 nd structure (EXAFS) analysis, the vanadium site is ref 63 63 63 27 believed to be a distorted octahedron coordinated by 1940 Chemlcal Reviews, 1993, Vol. 93, No. 5 Butler and Walker organic substrates or react with another equivalent of hydrogen peroxide, forming dioxygen, as depicted in Scheme I for bromide."^^'

Scheme I. Summary of the General Reactivity of V-BrPO with Bromide V-BrPO Br' + H,O, -Br* (eg., HOBr, Br,, Br3., Figure 3. The proposed active site of V-BrPO from EXAFS4 Enz-Br) a single terminal oxide ligand at 1.61 A,three unknown light-atom donors ca. 1.72 A, and two nitrogen donors at 2.11 A (possibly histidine nitrogen^^^) (Figure 3). The EXAFS spectrum of V-BrPO is independent of In the case of bromide, the dioxygen formed is in the the presence of H202 and Br-. V-BrPO can be reduced singlet excited state (lO2,lAg) which was identified to V02+-BrP0as shown by the appearance of an ESR spectroscopically by the near-infrared emission char- signal.44 The EXAFS spectrum of the reduced enzyme acteristic~.'~Singlet oxygen (lAg) is a well-established is consistent with formation of vanadyl-BrPO in which product of the oxidation of H202 by HOBr, HOC1," the V(V) site is coordinated by a single terminal oxide and bromamines.55 Recent H21802 tracer studies now at 1.63 A,and 5 O(N) ligands, of which three are at 1.91 establish that each atom of oxygen in 02originates from A and two are at 2.11 A4 (Figure 3). Electron spin echo the same molecule of H202 which is also consistent with results of the VIV-BrPO derivative indicates that a singlet oxygen production.24 nitrogen ligand is in the equatorial plane.45 V-BrPO is The rate of dioxygen formation in the absence of an ESR silent under turnover conditions (i.e., in the organic substrate and the rate of bromination of presence of Br- and H202). monochlorodimedone (MCD; at >75 CLM)are the same, The standard procedure for removal of active-site which supports the mechanism in Scheme I; the reactive vanadium(V) has been to incubate V-BrPO in 0.1 M intermediate is common to both pathways and is formed phosphate-citrate buffer pH 3.8, containing 10 mM in a rate-limiting step.50 Moreover kl[MCDI is com- EDTA. These conditions remove over 95% of the petitive with k2[H2021, because the sum of kl[MCDl vanadium, producing the inactive apo-BrPO deriva- and kz[H202] in the presence of MCD is equal to tive.20148 The essential component of the apoprotein kz[H2021 in the absence of an organic halogen acceptor." preparation is the phosphate, without which vanadium V-BrPO also catalyzes peroxidative chlorination is not completely removed and the enzyme is not reactions and the chloride-assisted disproportionation completely ina~tivated.~~In fact phosphate in the of hydrogen peroxide.53The chloroperoxidase reactivity absence of EDTA is sufficient for preparation of apo- of V-BrPO was not recognized rig in ally,^^ presumably BrP0;47inactivation by phosphate is much faster at due to the much slower rate of chloride oxidation than low pH (pH = 4) than at neutral or higher pH's. bromide oxidation catalyzed by V-BrPO. Like the Interestingly, phosphate inactivation does not occur in enzyme-catalyzed reactions with bromide, the rates of the presence of hydrogen peroxide;47this result suggests chlorination of MCD and the chloride-assisted dioxygen that the mechanism of vanadium09 release is through formation reaction are equivalent, except in the pres- a phosphate-vanadate complex of which the formation ence of amines (primary, secondary, and tertiary).63The would be inhibited by peroxide coordination to V(V). rate of dioxygen formation is much slower in the The activity of the apoderivative can be fully restored presence of amines due to the formation of chloramines by addition of vanadate.& Addition of other metal ions which can be observed spectrophotometrically by their did not restore bromoperoxidase activity.20 Vanadi- characteristic UV absorption maxima: um(V) is only fully incorporated in the absence of pho~phate.~~~~Like phosphate, molybdate, arsenate, NH2R + C1- + H202+ H+ - NHRCl + 2H,O tetrafluoroaluminate (AlF4-),and tetrafluoroberrylate Chloramine formation is observed because chloramines (BeF42-)are also reported to be competitive inhibitors do not readily oxidize hydrogen peroxide. Identical of the coordination of vanadium(V) to ap0-BrP0.~~ rates of MCD chlorination in the presence and absence of amine are observed53 (Table 1111, but because B. Reactivity of V-BrPO chloramines chlorinate very rapidly, it cannot be determined whether MCD chlorination proceeds 1. Halogenation and Halide-Assisted Disproportionation through a chloramine intermediate. of Hydrogen Peroxide One might also ask if bromamines can be observed V-BrPO catalyzes peroxidative halogenation reac- in the bromoperoxidase reaction. Bromamine forma- tions20122749and the halide-assisted disproportionation tion would not be observed using hydrogen peroxide of hydrogen peroxide." In the first step, the enzyme because bromamines are rapidly reduced by hydrogen catalyzes the oxidation of the halide by hydrogen Table 111. Comparison of the Rates of MCD peroxide producing an intermediate which is a two- Chlorination and C1--Assisted Dioxygen Formation in electron-oxidized halogen species; for bromide hypo- the Presence and Absence of an Amine bromous acid, bromine, tribromide, or an enzyme-bound "bromonium ion equivalent" are all consistent with the [taurine], mM -d[MCD]/dt, phllmin d[Ozl/dt, pM/min in vitro reactivity of the enzyme. In the second step, 0 4.62 (0.07) 4.61 (0.01) the oxidized intermediate can halogenate appropriate 1 4.62 (0.07) 1.31 (0.09) Marine HaloperoxMases Chemical Reviews, 1993, Vol. 93, No. 5 1041 peroxide, forming singlet oxygen and bromide.55 By dioxygen formation reaction and the MCD bromination replacing hydrogen peroxide with acyl peroxides (e.g., reaction agree within a factor of ca. 2, providing further peracetic acid and others; see below), bromamine evidence that the rate-limiting steps are the same in formation can be detected in the absence of organic both the bromination of MCD and the bromide-assisted substrates. Because acyl peroxides are not efficient disproportionation of hydrogen peroxide reactions.51 two-electron reductants, the Br+NHR- species can build Bromide inhibition is strongest at pH 5-5.5 for V-BrPO up in solution22(Scheme 11). Bromamines have been from the three sources e~amined.~l*s~The inhibition constants for hydrogen peroxide (KiiHz0z,KhHzOz) could Scheme 11. Summary of the Reactivity of only be determined for the bromide-assisted dioxygen Bromamines with Hydrogen Peroxide and Acyl formation reaction since high concentrations of hy- Peroxides drogen peroxide were required. Under these conditions BrPO the kinetics of MCD bromination were complicated by H202+ Br' + NH2R + H+- BrNHR + H20 the competing dioxygen formation reaction. Hydrogen fast peroxide inhibition is strongest at pH 7-8. H202 BrNHR + H202 - '02+ Br' + H+ + NH2R inhibition is noncompetitive at all pH's and KiiHzoz is equal to KisHz02, showing that bromide does not affect BrNHR + RCOOH - N.R or very slow rxn. the inhibition by H202.24 Fluoride is a competitive inhibitor of H202 binding in both the MCD bromination proposed as possible intermediates in other haloper- reaction and the bromide-assisted disproportionation oxidase reactions.5156 of hydrogen peroxide.s1 Fluoride inhibition (KiF) is uncompetitive with respect to bromide.51 2. Specific Activities The steady-state kinetic studies also showed that an The specific chloroperoxidase, bromoperoxidase, and active-site residue with a pK, of ca. 5.7 is required to iodoperoxidase activities differ substantially and de- be deprotonated prior to binding of hydrogen per0xide.m pend on pH, halide concentration and hydrogen per- The group was attributed to histidine or possibly a oxide concentrations. The pH for maximum specific bound water molecule. activity for bromination of MCD by V-BrPO is generally higher than that for chlorination, but direct comparisons 4. Peroxide Selectivity are difficult because the pH maximum can be shifted The catalytic activity of V-BrPO is supported by over several pH units by varying the ratio of halide to peracids (i-e., peracetic acid, m-chloroperoxybenzoic hydrogen peroxide and because both of these substrates acid, p-nitroperoxybenzoic acid, and phenylperacetic can also act as inhibitors. The specific activity of MCD acid) to catalyze the bromination of organic substrates.= bromination for V-BrPO isolated from A. nodosum is Unlike the reactions using hydrogen peroxide, the 170 units/mg (at pH 6.5,2 mM H202,O.l M BF, 50 pM uncatalyzed rate of bromide oxidation by peracetic acid MCD, 0.2 M Na2S04). The specific activities for the at pH 6.5, i.e., the pH range used for V-BrPO studies, M pyrifera and F. distichus enzymes are 1730 units/ is appreciable, however, the V-BrPO-catalyzedreactions mg (pH 6, other conditions same as above) and 1580 are much faster, particularly in the higher pH range.22 units/mg (pH 6.5, other conditions same as above), Dioxygen is not formed from peracids because peracids respectively;22the reasons for the significantly higher do not readily reduce the oxidized bromine species (e.g., specific activities of these enzymes compared to V-BrPO HOBr, Brz, Bra-, BrNHR) in the time frame of the from A. nodosum is not yet known. The specific enzymatic reactions (Le., several minutes) (see Table chloroperoxidase activity is 0.76 unit/mg (under con- 111). It has been established that peracid reactivity ditions of 1 M KC1,2 mM H202,50 pM MCD in 0.1 M actually arises from direct use of peracid and not from citrate buffer, pH 4.5) which is ca. 200 times less than hydrogen peroxide which could be formed by the the maximum specific bromoperoxidase activity.53 hydrolysis of the peracids.22 Alkyl hydroperoxides, including ethyl hydroperoxide, cuminyl hydroperoxide, 3. Enzyme Kinetics and tert-butyl hydroperoxide are not used by V-BrPO Detailed steady-state analyses of the rate of MCD to catalyze bromination reactions.22 bromination51956 and dioxygen f~rmation~l~~~catalyzed 5. Mechanistic Considerations by V-BrPO from A. nodosum, M. pyrifera, and F. distichus fit a substrate-inhibited bi-bi ping-pong Spectrophotometric analysis of the V-BrPO reaction kinetic mechanism, in which the substrates bromide has provided evidence that hydrogen peroxide binds and hydrogen peroxide are also noncompetitive inhib- first to V-BrPO, most likely by coordination to the itors at certain pH's (see Table IV).51124 The kinetic vanadium.58 Tromp et al. have reported a small parameters (KmBr,KmHz02, KiiBr, KisBr) obtained in the absorbance decrease (0.03 absorbance units for 20 pM V-BrPO) between 300 and 340 nm upon addition of Table IV. Summary of Kinetic Constants for V-BrPO H202 to V-BrPO;68the original UV spectrum reappears (A.nodosum) upon addition of bromide, i.e., turnover conditions. In DH ref a different set of experiments, we have established that ~ KmHflP 284 pM 5.25 51 stoichiometric bromination of MCD does not occur by KmB= 10.5 mM 5.25 51 addition of excess bromide to relatively high concen- Kip 332mM 4.55 51 trations of V-BrPO (i.e., 20 pM) and 50 pM MCD. Thus KbBr 730 mM 4.55 51 V-BrPO does not oxidize bromide in the absence of a KiiHflz 63 pM 7.0 24 peroxide source (hydrogen peroxide or an acyl perox- KbHaOa 67 pM 7.0 24 ide).24 Taken together, these two results indicate that KiF 1.11 mM 6.5 51 hydrogen peroxide coordinates first, followed by bro- 1942 Chemical Revlews, 1993, Vol. 93, No. 5 Butler and Walker Scheme 111. Catalytic Cycle for V-BrPO Showing Table V. Stability of Haloperoxidases on Turnover of Coordination of Hydrogen Peroxide before Oxidation Second Aliquot of Hydrogen Peroxide in the of Bromides Bromide-Assisted Disproportionation of Hydrogen Peroxide Producing Singlet Oxygen" 0 control 1 '02 yield from 2nd enzyme 1st aliquot HzOz aliquot HzO~ V-BrPO 1.00 (0.01) 1.07 (0.01) FeHeme-LactoPO 1.00 (0.05) 0.00 (0.01) FeHeme-CIPO 1.00 (0.05) 0.47 (0.02)

BrPO does not lose activity when stored at room temperature for a month in 60% (v/v) acetone, meth- t IO2 + Br' anol, or ethanol or 40% (v/v) l-pr~panol.~~V-BrPO also retains activity after immobilization on solid The intermediate, VvO,"OBr", is not meant to indicate whether support media, e.g., photocross-linked to DEAE cel- the oxidized bromine species is enzyme bound or released. lulofine.64 mide oxidation (Scheme 111). One can envisage several mechanisms of halide oxidation by the peroxovanadi- D. Comment on the Putative Non-Heme Iron um(V) species. The halide could coordinate to vana- Bromoperoxidase dium before oxidation or the halide could attack bound peroxide. A non-heme iron bromoperoxidase was reported to The detailed role of the active-site vanadium ion, be isolated from certain species of Rhodophyta.66v66An Le., whether it functions as an electron-transfer catalyst unusual characteristic was the low specific activity of bromide oxidation or a Lewis acid catalyst (and which has now been shown to be due to a small amount remains in the 5+ oxidation state), is not known. There of vanadium contained in these enzymesl6J6 despite is considerable precedence for both types of reactions, the presence of excess iron and magnesium. In these as in the two-electron reduction of vanadium(V) tet- enzymes, the metal ions can be completely removed raglyme by bromide (electron-transfer catalysis)sgand and the apoenzyme can be fully reconstituted by in the oxygenation of organic substrates by oxygen atom addition of vanadate.l6J7 The original enzyme isolation transfer from vanadium(V) peroxides (Lewis acid procedure was carried out in phosphate buffer.@ Given catalysis).m The vanadyl-BrPO state is probably not what is now known about phosphate displacement of an important component because V02+-BrP0does not bound ~anadium(V),4~the low specific activity of the have bromoperoxidase activity and because an EPR Corallina enzymes can be understood in terms of signal is not observed during turnover conditions. Most phosphate displacement of endogenous vanadium(V). other peroxidases are FeHeme-containing systems, which function as two-electron redox catalysts (see I I I. FeHeme Bromoperoxldase below). Hydrogen peroxide oxidizes the FeHeme moiety by two electrons, forming compound I.61 Com- A. Characteristics of FeHeme Bromoperoxidase pound I oxidizes the halide ion, forming the active halogenating species. This mechanism cannot be FeHeme bromoperoxidase (FeHeme-BrPO) was iso- operative in V-BrPO because the vanadium is already lated from several marine sources (see Table I) in the in its highest accessible oxidation state. Moreover, early 1980s by Baden and Corbett,28 Ahern,26 and native V-BrPO (protein-Vv=O), which might be viewed Manthey and Hager,27v36t61and more recently from a analogous to compound I (Heme+FeN=O) does not marine snaiPO and wormm (see below). The most oxidize bromide directly. 5lV NMR studies of func- thoroughly characterized of these enzymes is the one tional model compounds show that a long-lived reduced isolated from the green alga, Penicillus capitat~s.27~~*6~ oxidation state is not observed.@ FeHeme-BrPO is a dimeric protein (MW 97 600 de- termined by equilibrium ultracentrifugation) comprised C. Stability of V-BrPO of two identical subunits (55 000 determined from SDS PAG electrophoresis) each containing a tightly bound V-BrPO has been shown to be an exceptionally stable ferriprotoporphyrin IX moietyaZ7The FeHeme moiety enzyme which is not inactivated by high concentrations is present in a high-spin electronic configuration.61The of singlet oxygen or oxidized bromine species.17 V-BrPO amino acid composition of FeHeme-BrPO shows that is capable of turning over multiple aliquots of hydrogen it is rich in the acidic residues (Table 11), as was also peroxide in the bromide-assisted catalase reaction observed for V-BrPO and other peroxidases.27 producing stoichiometric yields of singlet oxygen (102,1Ao)without inactivation (Table V). By contrast the FeHeme haloperoxidase enzymes, lactoperoxidase B. Reactivity of FeHeme Bromoperoxidase and fungal chloroperoxidase, are inactivated in the FeHeme-BrPO catalyzes peroxidative halogenation halide-assisted disproportionation of hydrogen peroxide reactions, halide-assisted disproportionation of hydro- (Table V). The inactivation of the FeHeme haloper- gen peroxide, direct catalase activity, and free-radical oxidases is due to the species produced during turnover peroxidation reaction^.^^ In the peroxidative haloge- (ie., l02, HOBr/Brz/Br3-) primarily although a small nation reactions, the FeHeme moiety functions by amount of inactivation occurs by hydrogen peroxide in cycling between native enzyme (FeIIIHeme) and com- the absence of halide.17 pound I (Heme+-FeIV=O; the two-electron oxidized V-BrPO also has appreciable thermal stability and state above the native enzyme) as shown in Scheme IV. stability in many organic solvents. For example V- Compound I oxidizes a halide (I-, Br, or Cl-36)by two Marine Heioperoxldases Chemical Reviews, 1993, Vol. 93, No. 5 1943 Scheme IV. Summary of the General Reactivity of OMe OMe OMe FeHeme-BrPO with Bromide Br-MCD,MCD heme-Fe"' \ 0 -0-8 y2 y2 YH2 HOBr=Brz=Br3' CH.NH,+ CN CHO ', J c0,- A B C Figure 4. Oxidation of p-methoxytoluene (A), to p-meth- oxyphenylacetonitrile (B),and p-methoxyphenylacetaldehyde ((2). The order of addition of reactants in the MCD electrons, which in the case of Br- forms an equilibrium bromination reaction was found to affect the overall mixture of HOBr, Br2,and Br3-. Kinetic studies show reaction profile.6l Initiating the reaction with bromide that the rate of Br3- formation catalyzed by FeHeme- showed linear steady-state MCD bromination assays BrPO is twice as fast as the rate of MCD br~mination~~over time. Initiation of the reaction with hydrogen which indicates that the active brominating species is peroxide showed a fast initial rate followed by a slower not enzyme bound, but rather released into solution, steady state rate equal to that of the bromide-initiated as depicted in Scheme IV. reactions.61 The slower rate was shown to be due to the Saturation kinetics were not observed for bromide formation of the inactive compound I11 (Heme-FeII-02 which was interpreted by the authors as indicative of or Heme-FeI1*-02-). Compound I1 (Heme-FeN=O) is the absence of a specific bromide binding site or one formed first by the one-electron oxidation of MCD by with a very low affinity.36 This kinetic behavior differs compound I; compound I11 is then formed by reduction from V-BrPO in which bromide is also an inhibitor; of compound I1 with excess hydrogen peroxide.61 thus both bromide saturation rate behavior and bromide Compound I11 can be reduced to native enzyme.6' inhibition are observed.S1*NThe steady-state kinetics 1. Biosynthetic Investigations of MCD bromination catalyzed by FeHeme-BrPO as a function of pH show that protonation of an active site Hager has investigated biosynthetic origins of halo- residue with a pK, of 6.4 and deprotonation of another hydrocarbon formation.67 Crude extracts of P. capi- residue with a pK, of 5.3 are required for optimal tutus as well as the purified FeHeme-BrPO from this activity. Fluoride is a partial inhibitor of FeHeme- alga catalyze the formation of l-bromo-Zheptanone, BrPO.= l,l-dibromo-2-heptanone,and l,l,l-tribromo-2-hep- Initially it was thought that FeHeme-BrPO could tanone from 3-oxooctanoicacid, hydrogen peroxide, and not catalyze peroxidative chlorination reactions. How- bromide. l,l,l-Tribromo-2-heptanonethen hydrolyzes ever, peroxidative chlorination and the chloride-assisted in a nonenzymatic reaction, producing bromoform. disproportionation of hydrogen peroxide is observed36 Halogenated ketones have been isolated from numerous but in a low pH regime where FeHeme-BrPO does not species of algae (for a review see ref 4). Bromomethane catalyze bromide oxidation.27 Chlorination was estab- and dibromomethane, which are found in the red alga lished by isolation of dichlorodimedone. The nature Bonnemaisonia hamerifera, are not produced by crude of the oxidized chlorine intermediate, as enzyme-bound extracts of P. capitatus acting on 3-oxooctanoic acid. or released, has not been established.36 FeHeme-BrPO from P. capitatus has been shown to FeHeme-BrPO also catalyzes the direct dispropor- catalyze the conversion of a-amino acids and peptides tionation of hydrogen peroxide, forming dioxygen (i.e., to the decarboxylated nitriles and aldehydes.6 The direct catalase activity),27although the overall activity conditions required are those of the normal enzyme is only 1%of that of catalase. Like V-BrPO, halide- activity, i.e., bromide, hydrogen peroxide, and a-amino assisted catalase activity is observed in the presence of acid. This reaction was discovered by presence of bromidez7 and chl0ride.~6 Alkyl peroxides and acyl endogenous isobutyronitrile in crude bromoperoxidase peroxides can be used in place of hydrogen peroxide to preparations. Nieder and Hager showed that isobu- support halogenation reactions;27dioxygen formation tyronitrile could arise from valine via the formation of is not observed with these peroxides contrary to the the N-bromo and NJV-dibromo amino acid derivatives.6 reactivity of fungal FeHeme CIP0.27 In fact ethyl p-Methoxytoluene (A), is also converted to p-meth- hydroperoxide, peracetic acid, and m-chloroperoxy- oxyphenylacetonitrile (B), and p-methoxyphenyl- benzoic acid have been used to generate stable solutions acetaldehyde (C) (Figure 4). Such reactivity may form of the compound I state.6l the basis of the biogenesis of aeroplysinin-1 (see 1 in In addition to haloperoxidase activity, FeHeme-BrPO Figure l),a brominated compound isolated from certain catalyzes classical peroxidase reactions, such as the marine sponges. single-electron oxidation of phenolic compounds (py- rogallol, guaiacol, and hydroquin~ne),~~o-diani~idine?~ 2. Other FeHeme Haloperoxidases and MCD6' in the absence of halide, contrary to the Recently a new FeHeme chloroperoxidase was iso- reactivity of V-BrPO which does not carry out these lated from the marine worm, Notomastus lobatus.29 reactions. In these reactions, compound I is reduced This protein (MW 174000) is comprised of four by one electron to compound I1 (Heme-FeIV=O) and flavoprotein subunits (MW 30 000), tentatively iden- the organic radical. Compound I11 (the oxyferrous or tified as containing FAD, two FeHeme subunits (MW ferric superoxide states) is also observed by reaction of 15 500) and two other FeHeme subunits (MW 11 500). compound I1 with excess hydrogen pero~ide.~~*6l All components are required for activity, since neither 1844 Chemlcal Reviews, 1993, Vol. 93, No. 5 Butler and Walker

the FeHeme or flavoproteins have chloroperoxidase (15) Krenn, B. E.; Izumi, Y.; Yamada, H.; Wever, R. Biochim. Biophys. Acta 1989, 998, 63-8. activity alone. (16) Yu, H.; Whittaker, J. W. Biochem. Biophys. Res. Commun.1989, 160,87-92. (17) Everett, R. R.; Kanofsky, J. R.; Butler, A. J. Biol. Chem. 1990,265, I V, Future Directions 4908-14. (18) Krenn, B. E.; Plat, H.; Wever, R. Biochim. Biophys. Acta 1987, One of the most interesting, yet unsolved problems 912, 287-91. in the area of marine halogenation, is the biogenesis of (19) Wever, R.; Olafsson, G.; Krenn, B. E.; Tromp, M. G. M. Abstracts the chiral halogenated marine natural products. An of the 32nd IUPAC Conference, Stockholm, 1991; no. 210. (20) Vilter, H. Phytochemistry 1984,23,1387-1390. attractive feature of an enzyme-bound haloamine (21) de Boer, E.; van Kooyk, Y.; Tromp, M. G. M.; Plat, H.; Wever, R. (chloroamine or bromoamine), or other enzyme-bound Biochim. Biophys. Acta 1986,869,48-53. (22) Soedjak, H. S.; Butler, A. Biochemistry 1990,29, 7974-81. halogenated moiety, would be the possible generation (23) Jordan, P.; Vilter, H. Biochim. Biophys. Acta 1991,1073,98-106. of chiral halogenated compounds, through direct ha- (24) Soedjak, H. S. Ph.D. Dissertation, UC Santa Barbara, 1991. logenation by an enzyme-bound halonium ion. One (25) Pederson, M. Physiol. Plant 1976, 37, 6. (26) Ahren, T. J.; Allan, G. G.; Medcalf, D. G. Biochim. Biophys. Acta factor affecting the stability of haloamines is the 1980,616,329. presence of excess halide. For example bromonium (27) Manthey, J. A.;Hager,L. P. J.Biol. Chem. 1981,256,11232-11238. ions are readily displaced by bromide forming bromine (28) Baden. D. G.: Corbett. M. D. Biochem. J. 1980. 187. 205-211. (29) Chen, Y. P.; Lincoln, D. E.; Woodin, S. A.; Lovell, C. R. J. Biol. or tribromide, particularly at lower pH. The critical Chem. 1991,266,23909-15. factor would be the reactivity of the putative enzyme- (30) Jannun, R.; Coe, E. L. Physiol., B Comp. Biochem. 1987, 884 911-22. I bound halogenated intermediate toward halogenation, Hager, L. P.; Morris, D. R.; Brown, F. S.;Eberwein, H. J. Biol. hydrolysis, or displacement by excess halide. Another Chem. 1966,241, 1769-1777. possibility to account for chiral halogenation via HOBr Neidleman, S. L.; Geigert, J. Biohalogenation; Ellis Horwood Ltd Press: New York, 1986; p 203. (or equivalent) would be if the precursor substrate was van Pee, K. H.; Sury, G.; Lingens, G. Biol. Chem. Hoppe Seyler bound by the haloperoxidase or another enzyme such 1987,368,1225. that only one face of the substrate could be halogenated. Wiesner, W.; van Pee, K. H.; Lingens, F. J.Biol. Chem. 1988,263, 13725. Previously proposed mechanisms of the biosynthesis Wuosmaa, A. M.; Hager, L. P. Science 1990,249,160-162. of certain chlorinated compounds have invoked elec- Manthey, J. A.; Hager, L. P. Biochemistry 1989, 28, 3052-3057. trophilic bromination of alkenes followed by passive Krenn, B. E.; Tromp, M. G. M.; Wever, R. J.Biol. Chem. 1989,264, 19287-92. chloride attacks3*While this mechanism could explain Wever, R.; Krenn, B. E.; de Boer, E.; Offenberg, H.; Plat, H. Prog. the origin of adjacent brominated and chlorinated Clin. Biol. Res. (OxidasesRelat. Redox Syst.) 1988,274,477-93. Muller-Fahmow, A.; Hinrichs, W.; Saenger, W.; Vilter, H. FEBS carbons, it does not readily account for chlorine-only Lett. 1988,239,292-294. containing compounds. Thus, with the discovery of Vilter, H.; Glombitza, K.-W. Bot. Mar. 1983, XXVI, 341-344. chloroperoxidase activity of the vanadiums3 and Fe- Butler, A,; Soedjak, H. S.; Polne-Fuller, M.; Gibor, A.; Boyen, C.; Kloareg, B. J. Phycol. 1990, 26, 589-92. Hemez7 enzymes, the origin of specific chlorinated Jordan, P.; Kloareg, B.; Vilter, H. J.Plant Physiol. 1991,137,52& marine natural products can now be addressed. 524. Vilter, H. Bioseparation 1990,1,283-292. Arber, J. M.; de Boer, E.; Garner, C. D.; Hasnain, S. S.;Wever, R. A ckno wledgments Biochemistry 1989,28,7968-73. de Boer, E.; Keijzers, C. P.; Klassen, A. A. K.; Reijerse, E. J.; Collison, A.B. gratefully acknowledges support from the Na- D.; Garner, C. D.; Wever, R. FEBS Lett. 1988,235,93-97. de Boer, E.; Boon, K.; Wever, R. Biochemistry 1987,27,1629-35. tional Science Foundation (DBM90-18025; for inves- Soedjak, H. S.; Everett, R. R.; Butler, A. J. Ind. Microbiol. 1991, tigations on vanadium bromoperoxidase). A.B. is an 8,37-44. Alfred P. Sloan Research Fellow. Partial support for Tromp, M.; Tran, T. V.; Wever, R. Biochem. Biophys. Acta 1991, 1079, 53-6. this work is also sponsored by NOAA, National Sea Wever, R.; Plat, H.; de Boer, E. Biochim. Biophys. Acta 1985,830, Grant College Program, Department of Commerce, 181-6. under Grant NA89AA-D-SG138, project R/MP-53 Everett, R. R.; Butler, A. Inorg. Chem. 1989,28, 393-5. Everett, R. R.; Soedjak, H. S.; Butler, A. J. Biol. Chem. 1990,265, through the California Sea Grant College Program. 15671-9. Foote, C. S. In Singlet Oxygen;Wasserman, H. H., Murray, R. W., Eds.; Academic Press: Orlando, FL, 1979; pp 139-71. References Soedjak, H. S.; Butler, A. Znorg. Chem. 1990,29,5015-17. Wever, R.; Kustin, K. Adu. Inorg. Chem. 1990,35,81-115. Gschwend, P. M.; MacFarlane, J. K.; Newman, K. A. Science 1985, Kanofsky, J. R. Arch. Biochem. Biophys. 1989, 274, 229-234. 227, 1033-1035. de Boer, E.; Wever, R. J. Biol. Chem. 1988,263,12326-32. Walter. B.: Ballschmiter, K. Chemosphere 1991.22, 557-67. Soedjak, H. S.; Butler, A. Biochim. Biophys. Acta 1991,1079,l-7. Manley, S. L.; Goodwin, K.; North, W. J. Limnol. Oceanogr. 1992, Tromp, M. G. M.; Olafsson, G.; Krenn, B. E.; Wever, R. Biochim. 37, 1652-1659. Biophp. Acta 1990,1040,192-198. McConnell, 0. J.; Fenical, W. Tetrahedron Lett. 1977,48,4159- Neumann, R.; Assael, I. J. Am. Chem. SOC.1989, 111, 8410. 4162. Mimoun, H.; Mignard, M.; Brechot, P.; Saussine, L. J. Am. Chem. Nieder, M.; Hager, L. Arch. Biochem.Biophys. 1985,240,121-127. SOC.1986,108,3711-3718. James, M. D.; Kunze, H. B.; Faulkner, D. J. J. Natl. Prod. 1991, Manthey, J. A.; Hager, L. P. J.Biol. Chem. 1985,260,9654-9659. 54, 1137. de la Rosa, R. I.; Clague, M. J.; Butler, A. J. Am. Chem. SOC.1992, Salva, J.; Faulkner, J. J. Natl. Prod. 1990,53, 757. 114,760-761. Groweiss, A.; Look, S.; Fenical, W. J. Org. Chem. 1988,53, 2401. de Boer, E.; Plat, H.; Tromp, M. G. M.; Wever, R.; Franssen, M. Djura, P.; Stierle, D. B.; Faulkner, D. J.; Arnold, E.; Clardy, J. J. C.;vanderPlas,H. C.;Meijer,E. M.;Schoemaker,H.E.Biotechnol. Org. Chem. 1980,45,1435. Bioeng. 1987, 30,607-10. Faulkner, D. J. Nut. Prod. Rep. 1984,1,551-598.Faulkner, D. J. Itoh, N.; Cheng, L. Y.; Izumi, Y.; Yamada, H. J. Biotechnol. 1987, Nut. Prod. Rep. 1986, 3, 1-33. Faulkner, D. J. Nut. Prod. Rep. 5, 29-38. 1987,4,539-576.Faulkner, D. J. Nut. Prod. Rep. 1988,5,613-663. Itoh, N.; Izumi, Y.; Yamada, H. Biochem. Biophys. Res. Commun. Faulkner, D. J. Nut. Prod. Rep. 1990, 7,269-309. Faulkner, D. J. 1985, 131, 428-435. Nut. Prod. Rep. 1991,8,97-147.Faulkner, D. J. Nut. Prod. Rep. Itoh. N.; Izumi, Y.: Yamada, H. J. Biol. Chem. 1987,262,11982- 1992,9,323-364. Faulkner, D. J. Oceanus 1992,35,29-35. 11987. Atkins, W. R. G. Sci. Proc. R. Dublin SOC.1914, 14, 199. Beissner, R. S.; Guilford, W. J.; Coates, R. M.; Hager, L. P. Reed, G. B. Bot. Gaz. 1915, 407. Biochemistry 1981,20, 3724-3731. Hewson, W. D.; Hager, L. P. J. Phycol. 1980,16,340-345. de Boer, E.; Tromp, M. G. M.; Plat, H.; Krenn, B. E.; Wever, R. Sheffield,D.J.;Harry,T.; Smith,A. J.; Rogers,L. J.Phytochemistry Biochim. Biophys. Acta 1986,872,104-15. 1993, 32, 21. Jordan, P.; Vilter, H. Electrophoresis 1990,Il,653-655.