Coordination chemistry of iron(III)±porphyrin±antibody complexes In¯uence on the peroxidase activity of the axial coordination of an imidazole on the iron atom Solange de Lauzon 1 , Daniel Mansuy 1 and Jean-Pierre Mahy 2 1 Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, UMR 8601 CNRS, Universite  Rene  Descartes, Paris, France; 2 Laboratoire de Chimie Bioorganique et Bioinorganique, FRE 2127 CNRS, ICMO, Ba à t. 420, Universite  Paris-Sud XI, Orsay, France An arti®cial p eroxidase-like h emoprotein has been obtained by associating a monoclonal antibody, 13G10, and its iron(III)±a,a,a,b-meso-tetrakis(ortho-carboxyphenyl)por- phyrin [Fe(ToCPP)] hapten. In this antibody, about two- thirds of the porphyrin moiety is inserted in the binding site, its ortho-COOH substituents being recognized by amino- acids of the protein, and a carboxylic acid side chain of the protein acts as a general acid base catalyst in the heterolytic cleavage of the O±O bond of H 2 O 2 , but no amino-acid res- idue is acting as an axial ligand of the iron.We here show that theironof13G10±Fe(ToCPP)isabletobind,likethatof free Fe(ToCPP), two small ligands such as CN ± , but only one imidazole ligand, in contrast to to the iron(III) of Fe(ToCPP) that binds two. This phenomenon is general for a series of monosubstituted imidazoles, the 2- and 4-alkyl-substituted imidazoles being the best ligands, in agreement with the hydrophobic character of the antibody binding site. Com- plexes of antibody 13G10 w ith less hindered i ron(III)± tetraarylporphyrins bearing only one [Fe(MoCPP)] or two meso-[ortho-carboxyphenyl] substituents [Fe(DoCPP)] also bind only one imidazole. Finally, peroxidase activity studies show that imidazole inhibits the peroxidase activity of 13G10±Fe(ToCPP) whereas it increa ses that of 1 3G10± Fe(DoCPP). This could be interpreted by the binding of the imidazole ligand on the iron atom which probably occurs in the case of 13G10±Fe(ToCPP) on the less hindered face of the porphyrin, close to the catalytic COOH residue, whereas in the case of 13G10±Fe(DoCPP) it can occur on the other face of the porphyrin. The 13G10±Fe(DoCPP)±imidazole complex thus constitutes a nice arti®cial peroxidase-like hemoprotein, with the axial imidazole ligand of the iron mimicking the proximal histidine of p eroxidases and a COOH side chain of the an tibody acting as a general acid- base catalyst like the distal histidine of peroxidases does. Keywords: catalytic antibody; p eroxidase; arti®cial hemo- protein; porphyrin; imidazole. The production of monoclonal antibodies raised against transition state a nalogs has proven to be a powerful strategy to obtain antibodies that are able to catalyze a wide range o f reactions [1±8]. However, as most o f these cata lytic antibodies have modest catalytic ef®ciencies, several other strategies have been envisioned. A ®rst strategy involves the production of antibodies directed toward the idiotype of antienzymes antibodies. This strategy has led to antibodies that display an acetylcholine esterase activity, with the highest ef®ciency (1.35 ´ 10 5 M )1 ás )1 ) ever reported for catalytic antibodies [9], or a b-lactamase activity [10]. A second strategy is based on the association of antibodies with cofactors such a s inorganic cofactors [11,12], natural cofactors [13], metal ions [14±17], or metal cofactors [18±40]. In particular, antibodies raised against porphyrin deriva- tives have received in the last few years considerable attention as models for hemoproteins of biological impor- tance s uch as cytochromes P450 [41] and heme peroxidases [42]. Antibodies have thus been elicited against meso- carboxyaryl substituted- [19,23,28,31±33,36,38], N-substi- tuted- [20,21,27,29,30,34,39], and Sn- [22,24] or Pd- [25,26] porphyrins. Five of the obtained antibodies [21,27,28,31,34] were found to have a signi®cant peroxidase a ctivity with k cat /K m values ranging between 3.8 ´ 10 3 and 2.9 ´ 10 5 M )1 ámin )1 . Three metalloporphyrin±antibody complexes were f ound to have a cytochrome P450-like activity: two had a weak catalytic activity for the epoxida- tion o f styrene [22,39] and, more recently, a monoclonal antibody raised against a water soluble Sn(IV) porphyrin containing an axial a-naphthoxy ligand, was found to be able, in the presence of a Ru(II) porphyrin cofactor, to catalyze the stereoselective sulfoxidation o f a romatic s ul- ®des by iodosylbenzene [43]. In previous papers [31,36,38], we reported the production of two monoclonal antibodies, Correspondence to J P. M ahy, Laboratoire de Chimie Bioorganique et Bioinorganique, FRE 2127 CNRS, ICMO, Baà t. 420, Universite  Paris-Sud XI, 91405 Orsay Cedex, France. Fax: + 36 1 01 69 15 72 81, E-mail: jpmahy@icmo.u-psud.fr Abbreviations: ToCPP, meso-tetrakis(ortho-carboxyphenyl)porphyrin; DoCPP, meso-di(ortho-carboxyphenyl)diphenylporphyrin; MoCPP, meso-mono(ortho-carboxyphenyl) triphenylporphyrin; ABTS, 2,2¢-azinobis(3-ethylbenzothiazoline-6 sulfonic acid); ImH, imidazole; KLH, keyhole limpet hemocyanin; BSA, bovine serum albumin; ELISA, Enzyme linked immunosorbent assay. Enzymes: cytochrome P-450 (EC 1.14.14.1); horseradish peroxidase (EC 1.11.1.7). (Received 27 July 2001, revised 7 November 2001, accepted 14 November 2001) Eur. J. Biochem. 269, 470±480 (2002) Ó FEBS 2002 13G10 and 14H7, which not only bound the hapten, iron(III)-a,a,a,b-meso-tetrakis(ortho-carboxyphenyl)por- phyrin [Fe(ToCPP)] (Fig. 1) w ith a high af®nity (K d  10 )9 M ), but also exhibited in i ts presence an interesting peroxidase activity with k cat  540 min )1 and k cat /K m  3.2 ´ 10 4 M )1 ámin )1 [38]. Measurements of the binding constants for various porphyrins [36], together with pH dependence studies of the kinetics of t he peroxidase reaction [38] associated with chemical modi®cations of the antibody protein [36,38] have shown that: (a) approximately two-thirds of the porphyrin moiety was inserted in the antibody pocket, three of the ortho-carboxylate substituents of the meso-phenyl rings being recognized by th e side chains of amino acids of the antibody [36]; (b) in the case of 13G10, one carboxylic acid residue of the protein could participate in the catalysis of the heterolytic cleavage of the O±O bond of peroxides [38]; (c) unfortunately, no amino-acid residue was coordinating the iron atom. We have thus undertaken studies of the coordination chemistry of the iron(III) of Fe(ToCPP) bound or not to antibody 13G10 with two objectives: ®rst, to get more precise information about the topology of the binding site of the antibody and particularly to appreciate the size of the cavity left around the iron atom and which kind of ligands it can accommodate; second, to measure the in¯uence of an axial ligand of the iron atom such as imidazole on the catalytic activity of the Fe(ToCPP) ±IgG complex. In the present paper, we report the results obtained by absorption spectroscopy studies which show that: (a) the iron atom is able to bind two CN ± ligands in Fe(ToCPP) alone as well as in its complex with antibody 13G10; (b) in contrast, whereas the iron(III) of Fe(ToCPP) alone is able to bind two imidazole ligands, that of the Fe(ToCPP))13G10 complex is able to bind only one imidazole ligand; (c) the binding of one imidazole to the iron atom inhibits the peroxidase activity of the Fe(ToCPP))13G10 complex whereas it e nhances that of the complexes of 13G10 with iron(III)-mono- and di-ortho-carboxyphenyl substituted tetraaryl porphyrins, Fe(MoCPP) and Fe(DoCPP). Finally, this paper shows that the association of an anti-porphyrin Ig 13G10 with a F e(III)-di-ortho-carboxyphenyl-porphyrin and imidazole provides an accurate arti®cial peroxidase- like hemoprotein, with the axial imidazole ligand of the iron mimicking the proximal histidine o f peroxidases and a COOH side ch ain of the antibody acting as a general acid- base catalyst like the distal histidine of peroxidases does. EXPERIMENTAL PROCEDURES Chemicals Sodium azide and sodium isothiocyanate were from Sigma. Potassium cyanide, imidazole, 1-methylimidazole, 1-benzy- limidazole, 2-methylimidazole, 4-methylimidazole, and 2-ethylimidazole were from Fluka. 2,2¢-azinobis(3-eth yl- benzothiazoline-6 sulfonic acid) (ABTS), and H 2 O 2 from Sigma. Synthesis of iron(III)± ortho -carboxyphenyl substituted tetraarylporphyrins The synthesis of the four atropoisomers of Fe(ToCPP) as well as those of Fe(MoCPP) and of Fe(DoCPP) has been made in three steps as described in a previous paper [36]. The ortho-carboxymethyl substituted tetraaryl porphyrins were ®rst synthesized by reaction at room temperature of ortho-carbomethoxybenzaldehyde with pyrrole in CH 2 Cl 2 in the presence of BF 3 -etherate as catalyst according to an already described procedure [36,43]. The atropoisomers have then been separated on a silicagel column and Fig. 1. Structure and nomenclature of the various porphyrins used in this work. Ó FEBS 2002 Peroxidase-like Fe±porphyrin±antibody complexes (Eur. J. Biochem. 269) 471 identi®ed by absorption, 1 H NMR and mass spectroscopies [36]. The iron atom was then inserted b y reaction of the isolated atropoisomers with Fe(CO) 5 in the presence of I 2 in toluene at room temperature to avoid isome rization [44]. Finally, the ortho-carboxy substituted t etraarylporphyrin isomers were subsequently obtained by saponi®cation of the ortho-methyl ester substituents in 2 M KOH in 80% EtOH at room temperature [45]. Production of monoclonal antibodies The generation of monoclonal antibodies has been reported i n d etail in previous p apers [31,36,38]. Fe(ToC- PP) was activated by N-hydroxysuccinimide and cova- lently attached to keyhole limpet hemocyanin and BSA in phosphate buffered saline p H 7.5. The conjugates were then puri®ed by chromatography on Biogel P10 and four 5-week-old, female BALB/c mice w ere i mmunized con - ventionally with the Fe(ToCPP)±KLH conjugate. The spleen cells of the m ouse showing the best immune response were fused with PAI myeloma cells acc ording to Ko È hler & Milstein [46]. The supernatants from the hybridoma cells were screened by ELISA for binding to the h apten±BSA c onjugate p eroxidase linked goat anti- (mouse Ig) Ig [47]. Positive hybridoma were cloned twice and propagated in ascites. Antibodies were then puri®ed from ascite ¯uid by protein A af®nity chromatography and their purity and homogeneity were checked by SDS gel electrophoresis. Absorption spectroscopy measurements Absorption spectra were recorded at 19  0.1 °Cusingan UVIKON 860 UV/visible spectrophotometer as follows. The sample cuvette contained either 2 l M Fe(ToCPP) or Fe(ToCPP) preincubated with 3 l M 13G10 in 50 m M phosphate buffer pH 7.0, the reference cuvette only co n- tained 50 m M phosphate buffer pH 7.0. Equal amounts of ligand L (L  imidazole, mono-substituted imidazole, CN ± , SCN ± ,N 3 ± )(2 M in the same buffer) were then added in both cuvettes and difference spectra were recorded between 350 and 650 nm. In most cases, the spectral evolution observed involved the formation of well de®ned isobestic points indicating the presence of two absorbing species. The reaction could then be represented by: PFe III  nL  PFe III L n 1 where P  ToCPP, 13G10±(ToCPP) and L  imidazole, mono-substituted imidazole, CN ± ,SCN ± ,N 3 ± . According to Brault & Rougee [48], it could then be analyzed by means of the standard equation 1aDA  1aDA I  K d aDA I  1aL n 2 where DA  A ) A 0 , DA I  A ) A I and A 0 , A I ,andA are the absorbances of the initial, ®nal and mixed species, respectively. The linearity of the graph representing 1/DA as a function of 1/[L] n wasthenassayedwithn  1andn  2 and K d and A I could be determined graphically. I t is noteworthy that when n  1, C 50  K d ,whereaswhenn  2, C 50  K 1a2 d ,withC 50 representing the concentration of ligand for which half o f the starting Fe(ToCPP) or Fe(ToCPP))13G10 complex has been converted into (ToCPP)Fe(L) n or 13G10±(T oCPP) Fe(L) n . Assay of peroxidase activity To assay the peroxidase activity of the various iron(III)± ortho-carboxy substituted tetraarylporphyrins and their complexes with antibody 13G10, the oxidation of ABTS by H 2 O 2 was performed at 19  0.1 °Cin0.1 M citrate/ 0.2 M phosphate buffer, pH 5, containing 0.2% dimethyl- sulfoxide. The absorbance was monitored at 414 nm using an UVIKON 860 UV/visible spectrophotometer. T he initial rates of oxidation were determined from the slope at the origin of the curve representing the variations of the absorbance at 414 nm as a function of time, using an e value of 28 000 M )1 ácm )1 [21]. In a ®rst set of experiments, ABTS (0.2 m M ) was oxidized by H 2 O 2 (0.7 m M ) in the presence of 0.4 l M Fe(ToCPP) or Fe-a,a-1,2- or -a,b-1,2-(DoCPP) preincubated or not with 0.6 l M 13G10 as catalysts. In a second set of experiments, ABTS (0.03 m M )was oxidized by H 2 O 2 (5 m M ) in the presence of 0.4 l M Fe(MoCPP) preincubated 60 min with 0.2 l M 13G10 or 0.2 l M Fe(ToCPP) or Fe-a,a-1,2- or -a,b-1,2 -(DoCPP) preincubated with 0 .4 l M 13G10. The in¯uence of imidazole on the kinetic parameters of the oxidation of ABTS by H 2 O 2 in the presence of Fe± porphyrin±antibody complexes was examined as follows. The catalysts were ®rst prepared by preincubation of 0.4 l M Fe(ToCPP) or a,a-1,2-Fe(DoCPP) or a,b-1,2-Fe(DoCPP) with 0.6 l M 13G10 for 60 min at 19 °C. For the reactions with imidazole, a further 15 min incubation at 19 °Cwith 50 m M imidazole was done; 0.2 m M ABTS was then added and the reaction was s tarted by the addition of H 2 O 2 at concentrations ranging b etween 0 and 10 m M . The initial rates of oxidation were then measured as above mentioned and the k cat and K m were calculated in all the cases from Lineweaver-Burk plots. RESULTS Binding of cyanide to Fe(ToCPP) and to 13G10±Fe(ToCPP) The reactions of SCN ± ,N 3 ± and CN ± with the iron(III) of Fe(ToCPP) and its complex with antibody 13G10 were examined by UV/visible spectroscopy in 0.1 M phosphate buffer, pH 7 at 19  0.1 °C as described in experimental procedures. SCN ± and N 3 ± failed to react with both complexes [data not shown] but, when increasing amounts of potassium cyanide, up to 11 m M ,wereaddedtoa2l M solution of Fe(ToCPP), the initial spectrum characteristic of a high spin iron(III) species was gradually replaced, with isobestic points at 4 07, 479 and 548 nm, by a new spectrum with maxima of absorption at 417 and 549 nm (Fig. 2A). Such a spectrum is similar to that already described for tetraaryl-Fe III ±CN complexes [49]. As in addition, 1/DA 417 varied linearly with 1/[CN ± ] ( Fig. 2A, inset), it is clear that the ®rst reaction observed was the binding of CN ± ligand to the iron(III) of Fe(ToCPP) (Eqn 3), with a calculated K d value of 3.70  0.06 m M (Table 1). FeToCPPCN À ToCPPFe III À CN 3 472 S. de Lauzon et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Further addition of potassium cyanide, up to 50 m M , resulted in the appearence, with isobestic points at 422, 486 and 586 nm, of a new spectrum with peaks at 426, 565 and 600 nm (Fig. 2B), characteristic of a [(tetraarylporphy- rin)Fe III (CN] 2 ] ± species [49]. Accordingly, when 1/ DA 426 was plotted vs. 1/[CN ± ], a straight line was obtained and a K d value of 19.5  0.3 m M could be determined graphically (Table 1). ToCPPFe III À CN  CN À ToCPPFe III CN 2  À 4 When the same e xperiment was carried o ut with the Fe(ToCPP))13 G10 comp lex (2 l M ), a ®rst species absorb- ing at 420 and 555 nm was formed with isobestic points at 409, 477 and 552 nm for concentrations of CN ± below 6m M (Fig. 3A). A second species absorbing at 429, 541 and 608 nm was obtained, with isobestic points at 428, 485 and 594 nm, for concentrations of CN ± higher than 10 m M (Fig. 3B). As, respectively, 1/DA 420 and 1/DA 429 varied linearly as a function of 1/[CN ± ] (Fig. 3, insets), it is clear that those two species were, respectively, the 13G10± (ToCPP)Fe±CN and 13G10-[(ToCPP)Fe(CN) 2 ] ± complexes. Accordingly, when (ToCPP)Fe-CN and [(ToCPP) Fe(CN) 2 ] ± were inserted into the antibody protein, the characteristic bands in their visible spectrum were shifted toward higher wavelengths to g ive spectra that were similar to those obtained upon direct reaction of potassium cyanide with 13G10±Fe(ToCPP) (data not shown). This shows that the binding of the two cyanide ligands to the iron atom of 13G10±Fe(ToCPP) actually occurred inside the binding pocket of the antibody. The value of the dissoci- ation constant calculated for 13G10±(ToCPP)Fe±CN (0.39  0.01 m M ) was about 10-fold lower than th at calcu- lated for (ToCPP)Fe±CN (3.70  0.06 m M )(Table1), whereas that of 13G10±[[ToCPP]Fe[CN] 2 ] ± (16.90  0.13 m M ) was only slightly lower than that of Fig. 2. Addition of cyanide to Fe(ToCPP). (A) Spectral evolution observed for the addition of 0±11 m M CN ± to 2 l M Fe(ToCPP) in 0.1 M phosphate buer, pH 7 at 19 °C. Inset: corresponding values of 1/DA 417 plotted against 1/[CN ± ]. (B) Spectral evolut ion observed for the addition of 11±50 m M CN ± to 2 l M Fe(ToCPP) in 0.1 M phosphate buer, pH 7 at 19 °C. Inset: correspondin g values of 1/DA 417 plotted against 1/[CN ± ]. Table 1. Visible characteristics and C 50 values of the complexes of Fe(ToCPP) and 13G10- Fe(ToCPP) with cyanide in 50 m M phosphate buer, pH 7.0 at 20 °C. Complex Visible bands kmax (nm) C 50 a (m M ) (ToCPP)Fe III ±CN 417, 549 3.70  0.06 13G10±(ToCPP)Fe III ±CN 420, 555 0.39  0.01 ((ToCPP)Fe III (CN] 2 ) ± 426, 565, 600 19.5  0.3 (13G10±(ToCPP)Fe III (CN) 2 ) ± 429, ±, ± 16.90  0.13 a When only one CN ± is bound to Fe, C 50  K d and when two CN ± are bound to Fe, C 50  K d 1/2 . Fig. 3. Addition of cyanide to the Fe(ToCPP))13G10 complex. (A) Spectral evolution observed for the addition of 0±6 m M CN ± to 2 l M 13G10±Fe(ToCPP) in 0.1 M phosphate buer, pH 7 at 19 °C. Inset: corresponding values of 1/DA 420 plotted against 1/[CN ± ]. (B) Spectral evolut ion observed for the addition of 10±50 m M CN ± to 2 l M 13G10±Fe(ToCPP) in 0.1 M phos- phate buer, pH 7 at 19 °C. Inset: corresponding values of 1/DA 420 plotted against 1/[CN ± ]. Ó FEBS 2002 Peroxidase-like Fe±porphyrin±antibody complexes (Eur. J. Biochem. 269) 473 [(ToCPP)Fe(CN) 2 ] ± (19.5  0.3 m M ) (Table 1). The bind- ing of the ®rst CN ± ligand to the iron was thus more easy in the hydrophobic binding pocket of the antibody than the binding of the second one, most probably because of the steric hindrance brought by the protein around the iron atom of the porphyrin. Binding of monosubstituted imidazoles to Fe(ToCPP) and to its complex with antibody 13G10 Upon addition of increasing amounts of imidazole (ImH), up to 14 m M ,toa2l M solution of Fe(ToCPP), the initial spectrum of the high spin iron(III)±(ToCPP) was gradually replaced, with isobestic points at 408,475, 551 and 600 nm, by a new spectrum with maxima of absorption at 417, 549 and 580 nm (Fig. 4). Such a spectrum is similar to that already described for tetraaryl-Fe III [ImH] 2 complexes [50]. As on addition, 1/DA 417 varied linearly with 1/[ImH] 2 but not with 1/[ImH] (Fig. 5, inset), it is clear that the reaction observed was the binding of two ImH ligands to the iron(III) of Fe(ToCPP) (Eqn 5), with a C 50 ( K 1a2 d ) value of 2.70  0.04 m M (Table 2). FeToCPP2ImH ToCPPFe III ImH 2 5 When the same experiment was carried out with the Fe(ToCP P))13G10 complex (2 l M ), a species absorbing at 419, 552 and 587 nm was formed, with isobestic points at 407 and 538 nm for concentrations of ImH up to 200 m M (Fig. 5). As 1/DA 419 varied linearly with 1/[ImH] but not with 1/[ImH] 2 (Fig. 5, inset), it is clear that contrary to free Fe(ToCPP), the 13G10±Fe(ToCPP) complex was only able to bind one imidazole ligand (Eqn 6). 13G10 ÀToCPPFe III  ImH  13G10 ÀToCPPFe III ImH6 In addition, the C 50 ( K d ) value calculated in this case (21.3  0.3 m M ) is about 10-fold higher than that obtained for the formation of the (ToCPP)Fe III (ImH) 2 complex. The same reaction was performed with several mono- substituted imidazoles. In all the cases, the iron of free Fe(ToCPP) was able to bind two monosubtituted imidazole ligands whereas that of Fe(ToCPP) complexed with anti- body 13G10 was able to bind only one, the absorption spectra being similar to those obtained with non substituted imidazole (Table 2). In the case of 1-substituted imidazoles, the C 50 ( K d ) values observed with 13G10±Fe(ToCPP) were higher than the C 50 ( K 1a2 d ) values observed in the case of free Fe(ToCPP). Indeed for 1-methyl- and 1-benzylimidazole, C 50 ( K d ) values of, respectively, 4.30  0.06 m M and 14.5  0.2 m M were observed in the case of 13G10± Fe(ToCPP) whereas C 50 ( K 1a2 d ) values of, respectively, 2.60  0.04 m M and 0.63  0.01 m M were observed in the case of free Fe(ToCPP) (Table 2). In the case of 2 - and 4-substituted imidazoles, the iron o f free Fe(ToCPP) was also able to bind two ligands, w ith much higher C 50 ( K 1a2 d ) values (10- to 90-fold) than those calculated for imidazole and 1-substituted imidazoles (Table 2). Indeed, C 50 values of 26.2  0.04 m M , 56.0  0.8 m M and 54.0  0.8 m M could be calculated for 4-methyl-, 2-methyl- a nd 2-ethylimidazole, respectively Fig. 4. Addition of imidazole to iron(III)-a,a,a,b- meso-tetrakis(ortho- carboxyphenyl)porphyrin (Fe(ToCPP)). Spec tral evolution observed for the ad dition of 0±14 m M ImH to 2 l M Fe(ToCPP) in 0.1 M phosphate buer, pH 7 at 19 °C. In set: c orrespondin g values o f 1/DA 417 plotted against 1/[ImH] and 1/[ImH ] 2 . Fig. 5. Ad dition of imida zole to the Fe(ToCPP))13G10 complex. Spectral evolution observed for t he addition of 0±200 m M ImH to 2 l M Fe(ToCPP) in 0.1 M phosphate buer, pH 7 at 19 °C. Inset: corresponding values of 1/DA 419 plotted against 1/[ImH] and 1/[ImH] 2 . 474 S. de Lauzon et al. (Eur. J. Biochem. 269) Ó FEBS 2002 (Table 2). Those values were also 10±15-fold higher than the C 50 ( K d ) values found for the 13G10±(ToCPP)Fe)4- methyl- (2.80  0.06 m M ), 2-methyl (4.10  0.08 m M ) and 2-ethylimidazole (3.10  0.05 m M )(Table2). Binding of imidazole to various iron(III)- ortho -carboxy substituted tetraarylporphyrins and to their complexes with antibody 13G10 We also examined the binding of imidazole to iron(III)- mono- and di-ortho-carboxyphenyl substituted tetraaryl- porphyrins, Fe(MoCPP) and Fe(DoCPP) (Fig. 1), which were previously shown to form complexes with antibody 13G10 with, respectively, a 50-fold lower and an almost equal af®nity than Fe(ToCPP) [36]. The addition of increasing amounts of imidazole to 2 l M solutions of a,b- 1,2-Fe(DoCPP) or Fe(MoCPP) in 50 m M phosphate buffer pH 7.4, at 19 °C, led to the formation of the corresponding porphyrin±Fe III (ImH) 2 complexes, characterized by absorp- tion spectra with bands around 420, 550 and 580 nm (Table 3). However, owing to the low solubility of those complexes in the reaction medium, their K d values could not be determined accurately. When the same reaction was performed with a,a-1,2-Fe(DoCPP)-, a,b-1,2-Fe(DoCPP)- or Fe(MoCPP))13G10 complexes, new complexes, absorb- ing around 420, 550 and 580 nm were obtained (Table 3). In all the cases, 1/DA 420 was a linear function of 1/[ImH], which showed that in those cases, as in the case of Fe(ToCPP), only one imidazole ligand bou nd to the iron atom. In addition, the C 50 ( K d ) values c ould b e calculated (Table 3) and it appeared that the C 50 values obtained in the case of 13G10-a,a-1,2-Fe(DoCPP) and 13G10-a,b-1,2- Fe(DoCPP) were threefold and twofold lower than that obtained with 13G10±Fe(ToCPP). In contrast, a much higher C 50 value was found for 13G10±Fe(MoCPP) (236  4m M )(Table3). In¯uence of imidazole on the peroxidase activity of various iron(III)- ortho -carboxy substituted tetraarylporphyrins and their complexes with antibody 13G10 The in¯uence of the binding of imidazole to the iron atom of iron(III)-ortho-carboxy s ubstituted tetraarylporphyrins and their complexes with antibody 13G10 on their peroxidase activity was studied. The rate of oxidation of 0.2 m M ABTS by 0.7 m M H 2 O 2 was then measured at 19 °C in the presence of increasing concentrations of imidazole, using as catalyst e ither 0.3 l M iron(III)±por- phyrin or 0.3 l M iron(III)±porphyrin previously incubated with 0.6 l M 13G10. The reactions were performed in 50 m M phosphate buffer pH 5 as it had previously been shown that the peroxidase activity of the 13G10±Fe(ToC- PP) complex was optimal around this pH value [38]. We ®rst of all checked by absorption spectroscopy that both Fe(ToCPP) and 13G10±Fe(ToCPP) still bound imidazole at pH 5 with C 50 values similar to those calculated at pH 7 (data not shown). The peroxidase activity of Fe(ToCPP) alone was then assayed in the presence of concentrations of imidazole increasing from 0 to 150 m M (Fig. 6). With this catalyst, the rate of oxidation of ABTS by H 2 O 2 increased from an initial value of 0.16 l M ABTS oxidized per min to Table 2. Visible characteristics and C 50 values of the complexes of Fe(ToCPP) and 13G10-Fe(ToCPP) with various monosubstituted imidazoles in 50 m M phosphate buer, pH 7.0 at 20 °C. L P  ToCPP P  13G10-ToCPP n Visible bands kmax (nm) C 50 a (m M ) n Visible bands kmax (nm) C 50 a (m M ) ImH 2 417, 549, 580 (sh) 2.70  0.04 1 419, 552, 587(sh) 21.3  0.3 1-CH 3 -Im 2 418, 550, 580 (sh) 2.60  0.04 1 419, 548, 580(sh) 4.30  0.06 1-Bz-Im 2 417, 546, 582 (sh) 0.63  0.01 1 419, 548, 590(sh) 14.5  0.2 4-CH 3 -Im 2 418, 554, 582 (sh) 26.2  0.4 1 420, ±, - 2.80  0.06 2-CH 3 -Im 2 418, ±, ± 56.0  0.8 1 421, ±, - 4.10  0.05 2-C 2 H 5 -Im 2 419, ±, ± 54.0  0.8 1 421, ±, - 3.10  0.05 a n and K d for Fe III (L) n complexes were determined as described in Experimental procedures: when n  1, C 50  K d and when n  2, C 50  Kd1/2 . Table 3. Visible characteristics and C 50 values of the complexes of various iron(III)-ortho- carboxy-substituted-tetraarylporphyrins with imidazole in 50 m M phosphate buer, pH 7.0 at 20 °C in the presence or not of antibody 13G10. Porphyrin (Porphyrin)Fe III (ImH) 2 Visible bands kmax (nm) 13G10-(Porphyrin)Fe III (ImH) Visible bands kmax (nm) C 50 a (mM) a,a,a,b-Fe(ToCPP) 417, 549, 580(sh) 419, 552, 587(sh) 21.3  0.3 a,b-1,2-Fe(DoCPP) 417, 552, 581(sh) 417, 548, 580(sh) 13.0  0.2 a,a-1,2-Fe(DoCPP) ± 419, 547, 580(sh) 7.7  0.1 Fe(MoCPP) 420, 545, 582(sh) 420, 549, 580(sh) 236  4 a n and K d were determined as described in Experimental Procedures, C 50  K d . Ó FEBS 2002 Peroxidase-like Fe±porphyrin±antibody complexes (Eur. J. Biochem. 269) 475 a plateau value of 0.72 l M ABTS oxidized per min for a concentration of imidazole of 50 m M (Fig. 6). For con- centrations o f imidazole higher than 1 00 m M the r ate of oxidationofABTSstartedtodecrease(Fig.6).With 13G10±Fe(ToCPP) a s catalyst, the rate of oxidation of ABTS sharply decreased from 1.62 to 0.79 l M ABTS oxidized per min. in the p resence of concentrations o f imidazole increasing from 0 to 20 m M and then decreased more slowly to reach a plateau value of about 0.40 l M ABTS oxidized per min. for a concentration of imidazole of 150 m M (Fig. 6). Thus, w hereas the addition of imid- azole to Fe(ToCPP) was found to increase its peroxidase activity with a A 50 of 16 m M , it inhibited the peroxidase activity of the 1 3G10±Fe(ToCPP) with an I 50 of about 19 m M . In addition, the activity of Fe(ToCPP) was even higher than that of 13G10±Fe(ToCPP) for concentrations of imidazole higher than 50 m M , as shown by the curves representing the variations of the rates of oxidation of ABTS by H 2 O 2 observed, respectively, for those two catalysts (Fig. 6). The peroxidase activity of the two atropoisomers of a,a-anda,b-1,2-Fe(DoCPP) (Fig. 1), which were previously found to have also a high af®nity for antibody 13G10 [36], was also assayed in the presence of increasing concentra- tions of imidazole and compared to that of the a,a-and a,b-1,2-Fe(DoCPP ))13G10 complexes. With both a,b-and a,a-1,2-Fe(DoCPP), the rate of oxidation of ABTS by H 2 O 2 increased with increasing concentrations of imidazole, from initial values of, respectively, 0.08 and 0.10 l M ABTS oxidized per min to respective plateau values of 0.47 and 0.36 l M ABTS oxidized per min in the presence of more than 50 m M imidazole ( Fig. 6). When the reaction was performed in the presence of a,a-anda,b-1,2-Fe(DoC- PP))13G10 complexes, the rate of oxidation of ABTS by H 2 O 2 increased sharply in the presence of increasing amounts of imidazole, from an initial value of 0.37 l M ABTS oxidized per min to respective maximum values of 3.68 and 2.43 l M ABTS oxidized per min in the presence of more than 50 m M imidazole (Fig. 6). Thus, contrary to what occurred with the 13G10±Fe(ToCPP) complex, t he addition of imidazole to a,a-anda,b-1,2-Fe(DoC- PP))13G10 complexes was found to increase largely their peroxidase activity with respective A 50 values of 15 and 25 m M . The kinetic parameters for the oxidation of 0.2 m M ABTS by H 2 O 2 , in the presence of either Fe(ToCPP) or Fe(ToCPP)- and Fe(DoCPP))13G10 complexes as cata- lyst, were measured at pH 5 without imidazole and in the presence of 50 m M imidazole (Table 4). It appeared that in all the cases the addition of 50 m M imidazole had a major effect on the k cat value: in the case of Fe(ToCPP)) 13G10, it ca used a de crease of the k cat value by a factor of % 4, from 109  10 min )1 to 32  3min )1 whereas in contrast, with both a,a-anda,b-1 ,2-Fe(DoCPP))13G10 complexes, it caused an increase the k cat value by a factor of % 5±6, respectively, from 32  3min )1 to 152  10 min )1 and from 16  2min )1 to 96  9min )1 .The addition of 50 m M imidazole had a more moderate effect on the K m value that only slightly decreased from 29  3m M to 19  2m M inthecaseofFe(ToCPP)) 13G10, whereas it decreased by a factor 3, respectively, from 34  3m M to 10  1m M and from 18  2m M to 7  1m M with a,a-anda,b-1,2-Fe(DoCPP)) 13G10. As a consequence, the addition of 50 m M imida- zole caused a two fold decrease of the k cat /K m value from 3.8  0.7 ´ 10 3 M )1 ámin )1 to 1.7  0.4 ´ 10 3 M )1 ámin )1 inthecaseofFe(ToCPP))13G10 as catalyst, whereas on the contrary, it caused an about 15-fold increase of the k cat /K m value, respectively, from 0.9  0.2 ´ 10 3 M )1 ámin )1 to 15.2  2.5 ´ 10 3 M )1 ámin )1 and 0.9  0.2 ´ Fig. 6. In¯uence of the addition of imidazole on the peroxidase activity of Fe(ToCPP), a,a-1,2- and -a,b-1,2-F e(DoCPP) and their complexes with 13G10. Variations of the initial rate of oxidation of 0.2 m M ABTS by 0.7 m M H 2 O 2 as a function of the concentration of imidazole in the presence 0.4 lm catalyst: (s) Fe(ToCPP) (d) 13G10±Fe(ToCPP) (h) a,a-1,2-Fe(DoCPP) (j)13G10-a,a-1,2-Fe(DoCPP) (n) a,b-1,2- Fe(DoCPP) (m) 13G10-a,b-1,2-Fe(DoCPP). Table 4. In¯uence of imidazole on the kinetic parameters of the oxidation of ABTS by H 2 O 2 catalyzedbyFe(ToCPP)±andFe(DoCPP))13G10 complexes at pH 5. Catalyst Without ImH + 50 m M ImH k cat (min )1 ) K m (m M ) k cat/ K m ( M )1 ámin )1 ) k cat (min )1 ) K m (m M ) k cat/ K m (M )1 ámin )1 ) Fe(ToCPP) 68  737 4 1.8  0.3 ´ 10 3 71  7 8.5  1 8.3  1.5 ´ 10 3 Fe(ToCPP))13G10 109  10 29  3 3.8  0.7 ´ 10 3 32  319 2 1.7  0.4 ´ 10 3 a,a-1,2-Fe(DoCPP)) 13G10 32  334 3 0.9  0.2 ´ 10 3 152  10 10  1 15.2  2.5 ´ 10 3 a,b-1,2-Fe(DoCPP)) 13G10 16  218 2 0.9  0.2 ´ 10 3 96  97 1 13.7  2.8 ´ 10 3 476 S. de Lauzon et al. (Eur. J. Biochem. 269) Ó FEBS 2002 10 3 M )1 ámin )1 to 1.7  0.4 ´ 10 3 M )1 ámin )1 with a,a-1,2- Fe(DoCPP)- an d a,b-1 ,2-Fe(DoCPP))13G10 as catalysts (Table 4). DISCUSSION Binding of cyanide to the iron(III) of Fe(ToCPP) and Fe(ToCPP))13G10 First of all, the aforementioned results show that the iron(III) of the Fe(ToCPP))13G10 complex is able to bind, like that of free Fe(ToCPP), two cyanide ligands. Indeed, like in the case of free Fe(ToCPP), the addition of increasing amounts of cyanide to Fe(ToCPP))13G10 leads to the formation of two successive complexes (Fig. 3): a ®rst one absorbing at 420 nm for CN ± concentrations below 6 m M and a second one absorbing at 429 nm for CN ± concentra- tions higher than 10 m M . The 13G10±(ToCPP)Fe±CN and 13G10±[(ToCPP]Fe(CN) 2 ] ± structures were strongly sug- gested for those two complexes as: (a) their spectra of absorption were similar to those previously reported f or (porphyrin)Fe±CN and [(porphyrin)Fe(CN) 2 ] ± complexes [49], their maxima of absorption being only 3 nm redshifted; (b) both 1/DA 420 and 1/DA 429 varied linearly as a function of 1/[CN ± ] (Fig. 3, insets); (c) when (ToCPP)Fe-CN and [(ToCPP)Fe(CN) 2 ] ± were reinserted into apo-13G10, spec- tra similar to those already oberved for 13G10±(ToCPP)Fe± CN and 13G10±[(ToCPP)Fe(CN) 2 ] ± were obtained, which showed that the binding of the two cyanide ligands on the iron did occur inside the binding pock et o f the antibody. This is totally diff erent f rom what w as reported b y Kawamura-Konishi et al. [29] for the anti-(N-methyl mes- oporphyrin IX) Ig 2B4. Indeed, in this case, the iron(III) of the 2B4±Fe(mesoporphyrin IX) complex was found to be able to bind only one CN ± ligand, which was interpreted as a side-on binding of the porphyrin inside the antibody pocket, leaving only one of its faces accessible to ligands. Conse- quently, it is more likely that in our case, a n edge-on binding of Fe(ToCPP) occurs inside the binding pocket of 13G10 (Fig. 7) that allows 2 CN ± ligands to bind on the iron atom, one on each face of the porphyrin. This hyp othesis is in agreement with the binding site topology w hich we proposed recently, and in which approximately two-thirds of the porphyrin moiety was inserted in the antibody pocket, three of the ortho-carboxylate substituents of the me so- phenyl rings being bound to side chains of amino acids such as arginine [36]. In this respect, it is noteworthy that the X-ray structure of a complex of iron(III)±mesoporphyrin IX with an anti-(N-methyl-mesoporphyrin IX) Ig [37] showed that in this case also, approximately two-thirds of the porphyrin moiety was inside the antibody pocket with three pyrrole rings packed tightly against r esidues of the V H domain and two pyrrole rings packed against tyrosine residues of the V L domain. Binding of imidazoles to the iron(III) of Fe(ToCPP) and Fe(ToCPP))13G10 The second part of our results concerns the binding of imidazole d erivatives on the iron of Fe(ToCPP) either alone in solution or inside the binding pocket of antibody 13G10. It then appears that Fe(ToCPP) alone forms bis-imidazole complexes with imidazole and its deriva tives: 1-methyl-, 1-benzyl-, 2-methyl-, 2-ethyl- and 4-methylimidazole. This was shown particularly by: (a) the UV/visible spectra obtained after addition of imidazole (Fig. 4) or its deriva- tives (Table 2) to Fe(ToCPP), which were very similar to those already reported for (tetraarylporphyrin)Fe III (ImH) 2 complexes [50]; (b) the linear dependence of 1/DA 417 as a function of 1/[ImH] 2 when increasing amounts of imidazole were added to Fe(ToCPP); and (c) the C 50 ( K 1a2 d ) found for the complexes of Fe(ToCPP) with imidazole derivatives (Table 2), which were in good agreement with the equilib- rium constant b 2 measured for the formation of (porphy- rin)Fe(ImH) 2 complexes [51]. This of course is not very surprising as it has been reported extensively in the literature that Fe-tetraarylporphyrins were able to form bis-imida- zole±iron(III) complexes [50±53] and the X-ray structure of some of these complexes has been determined [52,53]. More surprising, ho wever, i s the ®nding that the iron of the Fe(ToCPP))13G10 complex is able to bind only one imidazole ligand as shown by the fact that 1/DA 419 varies linearly as a function of 1/[ImH] but not as a function of 1/[ImH] 2 (Fig. 5, inset). This suggests that there is not enough space left around the iron atom inside the antibody pocket to accommodate two imidazole ligands and, as the C 50 value calculated in this case (21.3  0.3 m M ) is about 10-fold higher than that calculated in the case of Fe(ToCPP) (2.70  0.04 m M ) (Table 2), that even the formation of the mono-imidazole complex of 13G10±Fe(ToCPP) is more dif®cult than the formation of the bis-imidazole complex of Fe(ToCPP). A likely explanation for this is that one imidazole ligand is able to bind on the less hindered face of the porphyrin bearing only the b-carboxyphenyl group Fig. 7. Various possibilities for the binding of ligands on the iron of Fe(porphyrin))13G1 0 complexes: (A) binding of H 2 O 2 on the iron of 13G10-(Fe(ToCPP), (B) bind ing of two CN ± ligands on the iron of 13G10-(Fe(ToCPP), (C) binding of H 2 O 2 and imidazole on the iron of 13G10-(Fe(ToCPP), (D) binding of H 2 O 2 and imidazole on the iron of 13G10-a,a-1,2-Fe(DoCPP). Ó FEBS 2002 Peroxidase-like Fe±porphyrin±antibody complexes (Eur. J. Biochem. 269) 477 but a second imidazole is then unable to bind on the other more hindered f ace of the porphyrin that bears the three a-carboxyphenyl groups and is stacked against the antibody protein (Fig. 7). This hypothesis is furth er sustained by t he C 50 values measured for the binding of various substituted imidazoles on the iron of Fe(ToCPP) and 13G10±Fe(ToCPP) (Table 2). First, with 1 -substituted imidazoles, which bear an hydrophobic substituent on the nitrogen atom opposite to th at which binds to th e iron, the C 50 values measured were lower than those measured for imidazole both with F e(ToC- PP) and 13G10±Fe(ToCPP). Indeed, for Fe(ToCPP), C 50 values of 2.60  0.04 m M and 0 .6 3  0.01 m M were found, respectively, for 1-CH 3 - and 1-benzylimidazole and 2.70  0.04 m M for imidazole whereas for 13G10±Fe(ToC- PP), C 50 values of 4.30  0.06 and 14.5  0.2 m M were found, respectively, for 1-CH 3 - and 1-benzylimidazole and 21.3  0.3 m M for imidazole. This could be explained by a greater hydrophobicity of the two 1-substituted imidazoles with respect to that of imidazole. Second, like in the case of imidazole, the C 50 values are higher with the antibody± Fe(ToCPP) complex than in the case of Fe(ToCPP) alone, which con®rms that the binding of one 1-substituted imidazole on the iron of 13G10±Fe(ToCPP) is even more dif®cult than the binding of two 1-substituted imidazoles on the iron of free Fe(ToCPP). In addition, the in¯uence of the nature of the substituent was different in both cases. In the case of Fe(ToCPP), it did not cause any steric hindrance for the binding and 1-methylimidazole had the same af®nity for the iron th an imidazole, whereas 1-benzylimidazole had a better af®nity than imidazole. In constrast, in the case of 13G10±Fe(ToCPP), the C 50 value for 1-benzylimidazole was about fourfold higher than that for 1-methylimidazole, which could arise from a more important steric interaction of the 1-benzyl substituent with the antibody protein than that with the 1-methyl substituent. In the case of 2- and 4-substituted imidazoles, which bear an alkyl substituent on the carbon next to the nitrogen atom binding the iron, the C 50 values obtained for Fe(ToCPP) were 20- to 4 0-fold higher than the one for i midazole (Table 2). This was due to an important steric interaction between the 4- and 2-alkyl substituent with the plane of the porphyrin [48,52]. For 13G10±Fe(ToCPP), much lower C 50 ( K d ) values, of, respectively, 2.80  0.06 m M , 4.10  0.05 m M , and 3.10  0.0 5 m M for 4-methyl-, 2-methyl-, and 2-ethylimidazole were observed (Table 2). This could be explained by the sum of three effects: (a) a higher hydrophobic c haracter of 2- and 4-substituted imidazoles with respect to imidazole; (b) the absence of steric interaction between the 2- a nd 4-alkyl substituent and the protein; and (c) ®nally, in the antibody±Fe(ToCPP) complex, as only one 2- or 4-substituted imidazole was bound to the iron, the steric hindrance due to the substituent could be balanced by a distortion of the porphyrin ring and a shift of the iron atom outside the plane of the porphyrin. Binding of imidazole to the iron(III) of Fe(ToCPP), a,a-1,2- a,b-1,2-Fe(DoCPP), Fe(MoCPP) and to their complexes with antibody 13G10 The UV/visible s tudies reported above s howed that the addition of increasing amounts of imidazole to a,a-1,2- and a,b-1,2-Fe(DoCPP ) and to F e(MoCPP ), and to their complexes with 13G10 led to results that were similar to those observed f or Fe(ToCPP): iron(III)±bis-imidazole complexes were formed in the case of free Fe±porphyrins whereas mono imidazole±iron(III) complexes were formed in the case of Fe±porphyrin±antibody complexes (Table 3). In addition, in the particular c ase of a,b-1,2- and a,a-1,2- Fe(DoCPP))13G10 complexes, the C 50 values found were, respectively, twofold and threefold lower than that found for Fe(ToCPP) (Table 3). This could be due to an easier access of the imidazole to the iron in those complexes of 13G10 w ith t wo less hindered d i-ort ho-carboxyphenyl substituted tetraarylporphyrins. In¯uence of imidazole on the peroxidase activity of the Fe±porphyrin±antibody complexes In heme peroxidases, such as horseradish peroxidase, the iron atom is bound to the apoprotein by a proximal histidine [42] and it has been reported that this axial ligand has an important role in the modulation of the redox potential of the iron [54] and thus has a great in¯uence on the catalytic activity of those enzymes. Because our studies on the binding of imidazole to the iron(III) of our Fe±porphyrin±antibody complexes have shown that, in all the cases, only one imidazole was able to bind to the iron atom, this suggested that the association of Fe±porphyrin± antibody complexes with imidazole could constitute a very good biomimetic system for peroxidases. Consequently, the peroxidase activity of the iron(III)-ortho-carboxy substi- tuted tetraarylporphyrins and their complexes with anti- body 13G10 was measured in the presence of varying concentrations of imidazole ( Fig. 6, Table 4). First of all, the addition of increasing amounts of imidazole, from 0 to 150 m M , caused a slight increase of the peroxidase activity of Fe(ToCPP), a,a-1,2- and a,b-1,2- Fe(DoCPP) ( Fig. 6). This w as not surprising as it was previously reported that imidazole increased the ability of iron(III)± and Mn(III)±porphyrin complexes to catalyze the oxidation o f substrates such as sul®des [55], alkanes and alkenes [56,57] by H 2 O 2 . Second, when increasing amounts of imidazole were added to the Fe±porphyrin±antibody complexes, two different effects were observed depending upon the nature of the porphyrin. In the case of 13G10± Fe(ToCPP), the progressive addition of up to 150 m M imidazole was found to inhibit strongly the peroxidase activity with an I 50 of about 19 m M (Fig. 6). As this value is close to that of the C 50 calculated for the formation of the 13G10±(ToCPP)Fe±ImH complex, it is clear that this inhibition was due to the binding of the imidazole ligand on the iron atom. In addition, measurement of the kinetic parameters for the oxidation of ABTS by H 2 O 2 catalyzed by Fe(ToCPP))13G10 in the presence of 50 m M imidazole, showed that the k cat /K M value was decreased by a factor of % 2 (Table 4). Contrary to what was observed in the case of 13G10±Fe(ToCPP), the addition of increasing amounts of imidazole to a,a-1,2- and a,b-1,2-Fe(DoCPP))13G10 complexes led to a large increase of the peroxidase activity with A 50 values of about 15 and 25 m M , respectively, the activity being optimal for a concentration of imidazole of 50 m M (Fig. 6). Such a concentration o f imidazole was found to cause a 15-fold increase of the k cat /K m value with both a,a-1,2-Fe(DoCPP)- and a,b-1,2-Fe(DoCPP))13G10 complexes. 478 S. de Lauzon et al. (Eur. J. Biochem. 269) Ó FEBS 2002 We propose a likely explanation for t he above men- tioned results based on the possible active site topology of antibody 13G10 presented in Fig. 7 and which also takes into account our previously published observations [36,38]: (a) about two-thirds of the porphyrin moiety is inserted inside the antibody active site, a nd (b) a carboxylic acid residue of the protein participates to the catalysis of the heterolytic cleavage of the O±O bond of H 2 O 2 .Inthecase of 13G10±Fe(ToCPP) it is likely that there is room enough in the antibody active site to allow the binding of two CN ± ligands on the iron atom (Fig. 7B). In contrast, there is probably not enough space t o accommodate two imidazole ligands on the iron atom, and the only one that enters the active site bind on the less hindered face of the porphyrin which bears only one ortho-COOH substituent (Fig. 7C). The inhibition of the peroxidase activity could then arise from the fact that this imidazole binds on the same face of the porphyrin as the catalytic COOH residue, thus preventing it from acting as a g eneral acid±base catalyst, whereas H 2 O 2 can only bind on the opposite, more hindered face of the porphyrin. This probably does not occur in the case of the complexes of antibody 13G10 with the less hindered a,a-1,2- and a,b-1,2-Fe(DoCPP), as the addition of increasing amounts of imidazole to those complexes causes an increase of their peroxidase activity. It is then likely that in those complexes, the imidazole can bind on either face of the porphyrin a nd, in the more favorable conformation, H 2 O 2 binds to the iron on the same face of the porphyrin as the catalytic COOH residue, the imidazole binding to the iron on the opposite face of the porphyrin (Fig. 7D). An optimal catalytic effect can then be obtained as the COOH residue can act as a general acid±base catalyst and the imidazole ligand can modulate the redox potential of the iron atom [54]. Finally, the present work has led t o a new arti®cial hemoprotein or h emoabzyme that displays an interesting peroxidase-like a ctivity. The complex is composed of a robust protein, a monoclonal anti-porphyrin I g, with an iron(III)-DoCPP cofactor and imidazole as an axial ligand of the iron which, respectively, mimick the heme cofactor and the axial histidine ligand of the iron in peroxidases, whereas a COOH side chain of the antibody acts as a general acid±base catalyst in the same manner as the distal histidine of peroxidases. REFERENCES 1. Lerner, R.A., Benkovic, S.J. & Schultz, P.G. (1991) At the crossroads of chemistry and immun ology: catalytic antibodies. Science 252, 659±667. 2. Schultz, P.G. & Lerner, R.A. (1995) From molecular diversity to catalysis: lessons from the i mmune system. Science 269, 1835±1842. 3. Wentworth, P. & Janda, K.D. (1998) Catalytic antibodies. Curr. Opin. Chem. Biol. 2, 138±144. 4. Blackburn,G.M.,Datta,A.,Denham,H.&Wentworth,P.Jr (1998) Catalytic antibodies. Adv.Phys.Org.Chem.31, 249±292. 5. Reymond, J.L. (1999) Catalytic antibodies for organic synthesis. Top. Curr. Chem. 200, 59±93. 6. Hilvert, D. (2000) Critical analysis of antibody catalysis. Annu. Rev. Biochem. 69, 751±793. 7. Blackburn, G.M. & GarcË on, A. (2000) Catalytic antibodies. In Biotechnology, 2nd edn. (Kelly, D.R., eds), pp. 403±490. Wiley- VCH, Weinheim. 8. Stevenson, J.D. & Thomas, N.R. (2000) Catalyt ic antib odies and other biomimetic catalysts. Nat. Prod. Report 17, 535±577. 9. Izadar,L.,Friboulet,A.,Remy,M.H.,Roseto,A.&Thomas,D. (1993) Monoclonal anti-idiotypic antibodies as functional internal images of enzyme active sites: production of a catalytic antibody with a cholinesterase activity. Proc. Natl Acad. Sci. USA 90, 8876±8880. 10. Avalle, B., Thomas, D. & Friboulet, A. (1998) Functional mim- icry: elicitation of a monoclonal anti-idiotypic antibody hydrol- izing beta-lactams. FASEB J. 12, 1055±1060. 11. Hsieh, L.C., Yonkovich, S., Kochersperger, L. & Schultz, P.G. (1993) Controlling chemical reactivity with antibodies. Science 260, 337±339. 12. Hsieh,L.C.,Stephans,J.C.&Schultz,P.G.(1994)Anecient antibody-catalyzed oxygenation reaction. J. Am. Chem. Soc. 116, 2167. 13. Shokat, K.M., Leuman, C.J., Sugasawara, R. & Schultz, P.G. (1988) An antibody-mediated redox reaction. Angew. Chem. Int. Engl. Ed. 27, 1172±1174. 14. Roberts, V.A., Iverson, B.L., Iverson, S.A., Benkovic, S.J., L erner, R.A., Getzo, E.D. & Tainer, J.A. (1990) Antibody remodeling: a general solution to the design of a metal coordination site in an antibody binding pocket. Proc. Natl Acad. Sci. USA 87, 6654±6658. 15. Wade,W.S.,Koh,J.S.,Han,N.,Hoekstra,D.M.&Lerner,R.A. (1993) Engine ering metal coordination sites into the antibody light chain. J. Am. Chem. Soc. 115, 4449±4456. 16. Wade, W.S., Ashley, J.A., J ahangiri, G.K., McElhane y, G., Janda, K.D. & L erner, R.A. (1993) A h ighly selective metal-acti- vated catalytic antibody. J. Am. Chem. Soc. 115, 4906±4907. 17. Crowder, M.W., Stewart, J.D., Roberts, V.A., Bender, C.J., Tevelrakh, E., Peisach, J., Getzo, E.D., Ganey, B.T. & Ben- kovic, S.J. (1995) Spe ctroscop ic studies on the design ed metal- binding sites of the 43C9 single chain antibody. J. Am . Ch em. Soc . 117, 5627±5634. 18. Iverson, B.L. & Lerner, R.A. (1990) Sequence-speci®c peptide cleavage catalyzed by an antibody. Science 243, 1184±1187. 19. Schwabacher, A.W., Weinhouse, M.I., Auditor, M.M. & Lerner, R.A. (1989) Metalloselective anti-porphyrin monoclonal anti- body. J. Am. Chem. Soc. 111, 2344±2346. 20. Cochran, A.G. & Schultz, P.G. (1990) Antibody catalyzed porphyrin metallation. Science 249, 781±783. 21. Cochran, A.G. & Schultz, P.G. (1990) Peroxidase activity of an antibody-heme complex. J. Am. Chem. Soc. 112, 9414±9415. 22. Keinan, E., Sinha, S.C., Sinh a-Bagchi, A., Benory, E., Ghozi, M.C., Eshhar, Z. & Green, B.S. (1990) Towards antibody- mediated metallo-porphyrin chemistry. Pure Appl. Chem. 62, 2013±2019. 23. Harada, A., Okamoto, K. & Kamachi, M. (1991) Cooperative binding of porphyrin by anti-porphyrin antibodies, Chem. Lett., 953±956. 24. Keinan, E., Benory, E., Sinha, S.C., Sinha-Bagchi, A., Eren, D., Eshhar, Z. & Green, B.S. (1992) Catalytic antibodies: circular dichroism and ultraviolet-visible studies of antibody±metallopor- phyrins interactions. Inorg. Chem. 31, 5433±5438. 25. Savitsky, A.P., Demcheva, M.V., Mantrova, E.Y. & Ponomarev, G.V. (1994) Monoclonal antibodies against metalloporphyrins. Speci®city of interaction with structurally dierent metallopor- phyrins. FEBS Lett. 355, 314±316. 26. Savitsky, A.P., N elen, M.I., Yatsmirsky, A .K., Demcheva, M.V., Ponomarev, G.V. & Sinikov, I.V. (1994) Kinetics of oxidation of o-dianisidine by hydrogen peroxide in the presenc e of antibody complexes of iron (III)-coproporphyrin. Appl. Bioch em. Biotech- nol. 47, 317±327. 27. Feng,Y.,Liu,Z.,Gao,G.,Gao,S.J.,Liu,X.Y.&Yang,T.S. (1995) Study of the abzyme with catalytic peroxidase activity. Ann. NY Acad. Sci. 750, 271±276. Ó FEBS 2002 Peroxidase-like Fe±porphyrin±antibody complexes (Eur. J. Biochem. 269) 479 [...]... (1987) Unusual orientation of axial ligands in metalloporphyrins Molecular structure of low-spin bis (2-methylimidazole) (mesotetraphenylporphyrinato) iron (III) perchlorate J Am Chem Soc 109, 1963±1968 54 Rietjens, I.M., Osman, A.M., Veeger, C., Zakharieva, O., Antony, J., Grodzicki, M & Trautwein, A.X (1996) On the role of the axial ligand in heme based catalysis of the peroxidase and cytochrome P450... S., Watanabe, Y & Fujimori, K (1982) Biomimetic oxidation of organic sul®des with TPPFe (III) Cl /imidazole/ hydrogen peroxide Tetrahedron Lett 23, 1189±1192 56 Mansuy, D & Battioni, P (1989) Alkane functionnalisation by cytochromes P-450 and by model systems using O2 or H2O2 In Activation and Functionnalisation of Alkanes (Hill, C.L., ed.), pp 195±218 John Wiley and Sons, Inc., New York 57 Battioni, P.,... K.J & Hatano, K (1980) Preparation and characterization of the anionic complex potassium dicyano (mesotetraphenylporphyrinato) iron (III) bis acetone J Am Chem Soc 102, 3017±3021 50 Quinn, R., Nappa, M & Valentine, J.S (1982) New ®ve- and sixcoordinate imidazole and imidazolate complexes of ferric tetraphenylporphyrin J Am Chem Soc 104, 2588±2595 51 Walker, F.A., Lo, M.W & Ree, M.T (1976) Electronic e€ects... M.W & Ree, M.T (1976) Electronic e€ects in transition metal porphyrins The reactions of imidazoles and pyridines with a series of para-substituted tetraphenylporphyrins complexes of chloroiron (III) J Am Chem Soc 98, 5552±5560 52 Scheidt, W.R., Osvath, S.R & Lee, Y.J (1987) Crystal and molecular structure of bis (imidazole) (meso-tetraphenylporphyrinato) iron (III) chloride A classic molecule revisited... Study of abzymes with cytochrome P450 activity Ann NY Acad Sci 750, 273±275 40 Nimri, S & Keinan, E (1999) Antibody-metalloporphyrin catalytic assembly mimics natural oxidation enzymes J Am Chem Soc 121, 8978±8982 41 Ortiz de montellano, P.R (1995) Cytochrome P450: Structure, Mechanism and Biochemistry, 2nd edn Plenum Press, New York 42 Marnett, L.J & Kennedy, T.A (1995) Comparison of the peroxidase activity. .. hapten and exhibiting a high anity for ironporphyrins FEBS Lett 395, 73±76 32 Khoda, K., Kakehi, M., Ohtsuji, Y., Takagi, M & Imanaka, T (1997) Studies of high thermostability and peroxidase activity of recombinant antibody L chain-porphyrin Fe (III) complex FEBS Lett 407, 280±284 33 Harada, A., Fukushima, H., Shiotsuki, K., Yamaguchi, H., Oka, F & Kamachi, M (1997) Peroxidation of pyrogallol by antibodymetalloporphyrin... speci®city Nature 256, 495±497 47 de Lauzon, S., Desfosses, B., Moreau, M., Le Trang, N., Rajkovski, K & Cittanova, N (1990) Comparison of monoclonal antibodies to estradiol obtained from structurally di€erent immunogens Hybridoma 9, 481±491 48 Brault, D & Rougee, M (1974) Binding of imidazole and 2-methylimidazole by hemes in organic solvents Evidence for ®ve -coordination Biochem Biophys Res Commun 3, 654±659... antibodymetalloporphyrin complexes Inorg Chem 36, 6099±6102 34 Kawamura-Konishi, Y., Asano, A., Yamasaki, M., Tashiro, H & Suzuki, H (1998) Peroxidase activity of an antibody-ferric porphyrin complex J Mol Catal B: Enz 4, 181±190 35 Blackwood, M.E Jr,, Rush, T.S., Romesberg, F., Schultz, P.G & Spiro, T.G (1998) Alternative modes of substrate distortion in enzyme and antibody-catalyzed ferrochelation reactions Biochemistry... 195±218 John Wiley and Sons, Inc., New York 57 Battioni, P., Renaud, J.P., Bartoli, J.F., Reina-Artiles, M., Fort, M & Mansuy, D (1988) Monooxygenase-like oxidation of hydrocarbons by H2O2 catalyzed by manganese porphyrins and imidazole selection of the best catalytic system and nature of the active oxygen species J Am Chem Soc 110, 8462±8470 ... catalyzed insertion of metal ion into porphyrin J Biochem 119, 857±862 30 Kawamura-Konishi, Y., Neya, S., Funasaki, N & Suzuki, H (1996) Speci®c orientation of porphyrin at the binding site of catalytic antibody Biochem Biophys Res Commun 225, 537±544 31 Quilez, R., de Lauzon, S., Desfosses, B., Mansuy, D & Mahy, J.P (1996) Arti®cial peroxidase- like hemoproteins based on antibodies constructed from a . Coordination chemistry of iron( III)±porphyrin±antibody complexes In¯uence on the peroxidase activity of the axial coordination of an imidazole on the iron. accommodate; second, to measure the in¯uence of an axial ligand of the iron atom such as imidazole on the catalytic activity of the Fe(ToCPP) ±IgG complex. In the