Haem disorder in two myoglobins: comparison of reorientation rate.

The globins from sperm whale and from Aplysia limacina myoglobins were reconstituted by addition of stoichiometric ferric protohaem and the Soret c.d. was followed as a function of time. For both reconstituted proteins, the Soret c.d. changes with time, reflecting haem reorientation inside its pocket, as previously described [Aojula, Wilson & Drake (1986) Biochem. J. 237, 613-616] for sperm whale myoglobin. The time course of the c.d. transition is found to be approx. 10 times faster in Aplysia than in sperm whale myoglobin, a result which is in agreement with the known structural and physicochemical properties of the two myoglobins; furthermore, these results confirm that c.d. and n.m.r. data on haem orientation in haemoproteins reflect the same molecular phenomenon.     

Histidine and not tyrosine is required for the peroxide-induced formation of haem to protein cross-linked myoglobin

Peroxide-induced oxidative modifications of haem proteins such as myoglobin and haemoglobin can lead to the formation of a covalent bond between the haem and globin. These haem to protein cross-linked forms of myoglobin and haemoglobin are cytotoxic and have been identified in pathological conditions in vivo. An understanding of the mechanism of haem to protein cross-link formation could provide important information on the mechanisms of the oxidative processes that lead to pathological complications associated with the formation of these altered myoglobins and haemoglobins. We have re-examined the mechanism of the formation of haem to protein cross-link to test the previously reported hypothesis that the haem forms a covalent bond to the protein via the tyrosine 103 residue (Catalano, C. E., Choe, Y. S., Ortiz de Montellano, P. R., J. Biol. Chem. 1989, 10534 - 10541). Comparison of native horse myoglobin, recombinant sperm whale myoglobin and Tyr(103) --> Phe sperm whale mutant shows that, contrary to the previously proposed mechanism of haem to protein cross-link formation, the absence of tyrosine 103 has no impact on the formation of haem to protein cross-links. In contrast, we have found that engineered myoglobins that lack the distal histidine residue either cannot generate haem to protein cross-links or show greatly suppressed levels of modified protein. Moreover, addition of a distal histidine to myoglobin from Aplysia limacina, that naturally lacks this histidine, restores the haem protein's capacity to generate haem to protein cross-links. The distal histidine is, therefore, vital for the formation of haem to protein cross-link and we explore this outcome.     

Properties of monoclonal antibodies specific for determinants of a protein antigen, myoglobin

Monoclonal hybridoma antibodies specific for the protein antigen sperm whale myoglobin were produced using hyperimmune spleen cells from mice with the genetic trait of high responsiveness to myoglobin. Antibodies from the several clones tested were found to produce linear Scatchard plots, as predicted for homogeneous antibodies, and to possess high affinities for the immunogen (KA congruent to 10(9) M-1). None of the monoclonal antibodies tested reacted with either fragment (1-55) or fragment (132-153) of sperm whale myoglobin. Competitive binding assays using human and horse myoglobins suggested that several of these monoclonal antibodies, which can readily distinguish these myoglobins, recognize different antigenic determinants on the myoglobin molecule. Studies using additional myoglobin sequence variants as competitors should be able to more closely define these antigenic determinants.     

Transient spectroscopy of the reaction of cyanide with ferrous myoglobin. Effect of distal side residues

The reaction of cyanide metmyoglobin with dithionite conforms to a two-step sequential mechanism with formation of an unstable intermediate, identified as cyanide bound ferrous myoglobin. This reaction was investigated by stopped-flow time resolved spectroscopy using different myoglobins, i.e. those from horse heart, Aplysia limacina buccal muscle, and three recombinant derivatives of sperm whale skeletal muscle myoglobin (Mb) (the wild type and two mutants). The myoglobins from horse and sperm whale (wild type) have in the distal position (E7) a histidyl residue, which is missing in A. limacina Mb as well as the two sperm whale mutants (E7 His----Gly and E7 His----Val). All these proteins in the reduced form display an extremely low affinity for cyanide at pH less than 10. The differences in spectroscopy and kinetics of the ferrous cyanide complex of these myoglobins indicate a role of the distal pocket on the properties of the complex. The two mutants of sperm whale Mb are characterized by a rate constant for the decay of the unstable intermediate much faster than that of the wild type, at all pH values explored. Therefore, we envisage a specific role of the distal His (E7) in controlling the rate of cyanide dissociation and also find that this effect depends on the protonation of a single ionizable group, with pK = 7.2, attributed to the E7 imidazole ring. The results on A. limacina Mb, which displays the slowest rate of cyanide dissociation, suggests that a considerable stabilizing effect can be exerted by Arg E10 which, according to Bolognesi et al. (Bolognesi, M., Coda, A., Frigerio, F., Gatti, C., Ascenzi, P., and Brunori, M. (1990) J. Mol. Biol. 213, 621-625), interacts inside the pocket with fluoride bound to the ferric heme iron. A mechanism of control for the rate of dissociation of cyanide from ferrous myoglobin, involving protonation of the bound anion, is discussed.     

NMR studies of hemoproteins. VI. Acid-base transitions of ferric myoglobin and its imidazole complex.

220 MHz proton NMR was applied to the acid-base transition of ferric myoglobin and its imidazole complex. In horse and sperm whale ferric myoglobins: (1) pH-dependent shift of heme-ring methyl signals above p2H 10 was analyzed on the basis of rapid exchange between alkaline and acidic forms by the use of pK value 9.1 of acid-base transition in 1H20 solution; (2) limiting shifts of three methyl signals were reasonably determined for purely alkaline form. For the imidazole complex: (3) a drastic high field shift of each signal was observed above p2H 9.0, whereas N0methyl imidazole complex did not exhibit such a shift, which suggests the 2H+ dissociation from liganded imidazole greater than N2H. It is concluded thns.     

Unusually weak oxygen binding, physical properties, partial sequence, autoxidation rate and a potential phosphorylation site of beluga whale (Delphinapterus leucas) myoglobin

We purified myoglobin from beluga whale (Delphinapterus leucas) muscle (longissimus dorsi) with size exclusion and cation exchange chromatographies. The molecular mass was determined by mass spectrometry (17,081 Da) and the isoelectric pH (9.4) by capillary isoelectric focusing. The near-complete amino acid sequence was determined and a phylogeny indicated that beluga was in the same clad as Dall's and harbor porpoises. There were consensus motifs for a phosphorylation site on the protein surface with the most likely site at serine-117. This motif was common to all cetacean myoglobins examined. Two oxygen-binding studies at 37 degrees C indicated dissociation constants (20.5 and 23.6 microM) 5.7-6.6 times larger than horse myoglobin (3.6 microM). The autoxidation rate of beluga myoglobin at 37 degrees C, pH 7.2 was 0.218+/-0.028 h(-1), 1/3 larger than reported for myoglobin of terrestrial mammals. There was no clear sequence change to explain the difference in oxygen binding or autoxidation although substitutions (N66 and T67) in an invariant rich sequence (HGNTV) distal to the heme may play a role. Structural models based on the protein sequence and constructed on topologies of known templates (horse and sperm whale crystal structures) were not adequate to assess perturbation of the heme pocket.     

Structure-dynamics-function relationships in Asian elephant (Elephas maximus) myoglobin. An optical spectroscopy and flash photolysis study on functionally important motions.

In this work we report the thermal behavior (10-300 K) of the Soret band lineshape of deoxy and carbonmonoxy derivatives of Asian elephant (Elephas maximus) and horse myoglobins together with their carbon monoxide recombination kinetics after flash photolysis; the results are compared to analogous data relative to sperm whale myoglobin. The Soret band profile is modeled as a Voigt function that accounts for the coupling with high and low frequency vibrational modes, while inhomogeneous broadening is taken into account with suitable distributions of purely electronic transition frequencies. This analysis makes it possible to isolate the various contributions to the overall lineshape that; in turn, give information on structural and dynamic properties of the systems studied. The optical spectroscopy data point out sizable differences between elephant myoglobin on one hand and horse and sperm whale myoglobins on the other. These differences, more pronounced in deoxy derivatives, involve both the structure and dynamics of the heme pocket; in particular, elephant myoglobin appears to be characterized by larger anharmonic contributions to soft modes than the other two proteins. Flash photolysis data are analyzed as sums of kinetic processes with temperature-dependent fractional amplitudes, characterized by discrete pre-exponentials and either discrete or distributed activation enthalpies. In the whole temperature range investigated the behavior of elephant myoglobin appears to be more complex than that of horse and sperm whale myoglobins, which is in agreement with the increased anharmonic contributions to soft modes found in the former protein. Thus, to satisfactorily fit the time courses for CO recombination to elephant myoglobin five distinct processes are needed, only one of which is populated over the whole temperature range investigated. The remarkable convergence and complementarity between optical spectroscopy and flash photolysis data confirms the utility of combining these two experimental techniques in order to gain new and deeper insights into the functional relevance of protein fluctuations.     

On the difference in stability between horse and sperm whale myoglobins

The work in the literature on apomyoglobin is almost equally divided between horse and sperm whale myoglobins. The two proteins share high homology, show similar folding behavior, and it is often assumed that all folding phenomena found with one protein will also be found with the other. We report data at equilibrium showing that horse myoglobin was 2.1 kcal/mol less stable than sperm whale myoglobin at pH 5.0, and aggregated at high concentrations as measured by gel filtration and analytical ultracentrifugation experiments. The higher stability of sperm whale myoglobin was identified for both apo and holo forms, and was independent of pH from 5 to 8 and of the presence of sodium chloride. We also show that the substitution of sperm whale myoglobin residues Ala15 and Ala74 to Gly, the residues found at positions 15 and 74 in horse myoglobin, decreased the stability by 1.0 kcal/mol, indicating that helix propensity is an important component of the explanation for the difference in stability between the two proteins.     

Comparative oxygen affinity of fish and mammalian myoglobins

Summary Myoglobins from rat, coho salmon (Oncorhynchus kisutch), buffalo sculpin (Enophrys bison) hearts, and yellowfin tuna (Thunnus albacares) red skeletal muscle were partially purified and their O₂ binding affinities determined. Commercially prepared sperm whale myoglobin was employed as an internal standard. Tested at 20°C, myoglobins from salmon and sculpin bound O₂ with lower affinity than myoglobins from the rat or sperm whale. Oxygen binding studies at 12°C and 37°C suggest that this difference is adaptive, permitting myoglobins from cold-adapted fish to function at physiologically relevant temperatures. Taken together, purification and O₂ binding data obtained in this study reveal a previously unrecognized diversity of myoglobin structure and function.     

Contributions of residue 45(CD3) and heme-6-propionate to the biomolecular and geminate recombination reactions of myoglobin

Overall association and dissociation rate constants were measured at 20 degrees C for O2, CO, and alkyl isocyanide binding to position 45 (CD3) mutants of pig and sperm whale myoglobins and to sperm whale myoglobin reconstituted with protoheme IX dimethyl ester. In pig myoglobin, Lys45(CD3) was replaced with Arg, His, Ser, and Glu; in sperm whale myoglobin, Arg45(CD3) was replaced with Ser and Gly. Intramolecular rebinding of NO, O2, and methyl isocyanide to Arg45, Ser45, Glu45, and Lys45(native) pig myoglobins was measured following 35-ps and 17-ns excitation pulses. The shorter, picosecond laser flash was used to examine ligand recombination from photochemically produced contact pairs, and the longer, nanosecond flash was used to measure the rebinding of ligands farther removed from the iron atom. Mutations at position 45 or esterification of the heme did not change significantly (less than or equal to 2-fold) the overall association rate constants for NO, CO, and O2 binding at room temperature. These data demonstrate unequivocally that Lys(Arg)45 makes little contribution to the outer kinetic barrier for the entry of diatomic gases into the distal pocket of myoglobin, a result that contradicts a variety of previous structural and theoretical interpretations. However, the rates of geminate recombination of NO and O2 and the affinity of myoglobin for O2 were dependent upon the basicity of residue 45. The series of substitutions Arg45, Lys45, Ser45, and Glu45 in pig myoglobin led to a 3-fold decrease in the initial rate for the intramolecular, picosecond rebinding of NO and 4-fold decrease in the geminate rate constant for the nanosecond rebinding of O2. (ABSTRACT TRUNCATED AT 250 WORDS)     

Effects of amino acid substitutions outside an antigenic site on protein binding to monoclonal antibodies of predetermined specificity obtained by peptide immunization: Demonstration with region 15–22 (antigenic site 1) of myoglobin

Monoclonal antibodies (mAbs) of predetermined specificity were prepared by immunizing with a free (i.e., not conjugated to any carrier) synthetic peptide representing region 15–22 (site 1) of sperm whale myoglobin (SpMb). The cross-reactions of Mb variants with three mAbs were studied in order to determine whether such interactions are influenced by substitutions outsde the site. Finback whale Mb, which has no substitutions within region 15–22, showed lower cross-reactivity and relative binding affinity than the reference antigen, SpMb. Bottle-nose Atlantic dolphin myoglobin (BdMb) and badger myoglobin (BgMb), although they have identical substitutions within region 15–22 (Ala-15 to Gly and Val-21 to Leu), showed very different binding properties. The cross-reaction of BdMb was quite comparable to that of SpMb, while that of BgMb was much lower. Since the two proteins have identical structures in regions 15–22, the differences in their cross-reactivities are readily attributed to the effects of substitutions outside this region. Another pair of myoglobins, horse myoglobins (HsMb) and chicken myoglobin (ChMb), also have two identical substitutions (Ala-15 to Gly and Val-21 to Ile) within region 15–22, but possessed different cross-reactivity. The results indicate that the reaction of mAbs, whose specificity is precisely known and predetermined by the immunizing free peptide, can be markedly affected by substitutions outside the indicated binding region on the protein.     

Effects of amino acid substitutions outside an antigenic site on protein binding to monoclonal antibodies of predetermined specificity obtained by peptide immunization: demonstration with region 15-22 (antigenic site 1) of myoglobin.

Monoclonal antibodies (mAbs) of predetermined specificity were prepared by immunizing with a free (i.e., not conjugated to any carrier) synthetic peptide representing region 15–22 (site 1) of sperm whale myoglobin (SpMb). The cross-reactions of Mb variants with three mAbs were studied in order to determine whether such interactions are influenced by substitutions outsde the site. Finback whale Mb, which has no substitutions within region 15–22, showed lower cross-reactivity and relative binding affinity than the reference antigen, SpMb. Bottle-nose Atlantic dolphin myoglobin (BdMb) and badger myoglobin (BgMb), although they have identical substitutions within region 15–22 (Ala-15 to Gly and Val-21 to Leu), showed very different binding properties. The cross-reaction of BdMb was quite comparable to that of SpMb, while that of BgMb was much lower. Since the two proteins have identical structures in regions 15–22, the differences in their cross-reactivities are readily attributed to the effects of substitutions outside this region. Another pair of myoglobins, horse myoglobins (HsMb) and chicken myoglobin (ChMb), also have two identical substitutions (Ala-15 to Gly and Val-21 to Ile) within region 15–22, but possessed different cross-reactivity. The results indicate that the reaction of mAbs, whose specificity is precisely known and predetermined by the immunizing free peptide, can be markedly affected by substitutions outside the indicated binding region on the protein.     

1H NMR study of labile proton exchange in the heme cavity as a probe for the potential ligand entry channel in myoglobin

The exchange rates of heme cavity histidine nitrogen-bound protons in horse and dog metcyanomyoglobins have been determined at 40 degrees C as a function of pH by 1H NMR spectroscopy. They were compared to the results reported for the sperm whale homologue [Cutnell, J. D., La Mar, G. N., & Kong, S. B. (1981) J. Am. Chem. Soc. 103, 3567-3572]. The rate profiles suggest that the exchange follows EX2-type kinetics, and the relative rate values favor a penetration model over a local unfolding model. It was found that the behavior of protons located on the proximal side of the heme is similar in the three proteins. The distal histidyl imidazole NH, however, shows a highly accelerated hydroxyl ion catalyzed rate in horse and dog myoglobins relative to that in sperm whale myoglobin. NMR spectral and relaxational characteristics of the assigned heme cavity protons indicate that the global geometry of the heme pocket is highly conserved in the ground-state structure of the three proteins. We propose a model that attributes the different distal histidine exchange behavior to the relative dynamic stability of the distal heme pocket in dog or horse myoglobin vs. sperm whale myoglobin. This model involves a dynamic equilibrium between a closed heme pocket as found in metaquomyoglobin [Takano, T. (1977) J. Mol. Biol. 110, 537-568] and an open pocket as found in phenylmetmyoglobin [Ringe, D., Petsko, G. A., Kerr, D. E., & Ortiz de Montellano, P. R. (1984) Biochemistry 23, 2-4].(ABSTRACT TRUNCATED AT 250 WORDS)     

Reaction of nitric oxide with heme proteins and model compounds of hemoglobin

Rates for the reaction of nitric oxide with several ferric heme proteins and model compounds have been measured. The NO combination rates are markedly affected by the presence or absence of distal histidine. Elephant myoglobin in which the E7 distal histidine has been replaced by glutamine reacts with NO 500-1000 times faster than do the native hemoglobins or myoglobins. By contrast, there is no difference in the CO combination rate constants of sperm whale and elephant myoglobins. Studies on ferric model compounds for the R and T states of hemoglobin indicate that their NO combination rate constants are similar to those observed for the combination of CO with the corresponding ferro derivatives. The last observation suggests that the presence of an axial water molecule at the ligand binding site of ferric hemoglobin A prevents it from exhibiting significant cooperativity in its reactions with NO.     

Binding of alkylisocyanides with soybean leghemoglobin. Comparisons with sperm whale myoglobin

The binding of various linear and branched chain alkylisocyanides to soybean leghemoglobin has been studied with respect to association and dissociation kinetics and the results compared with those obtained in parallel on sperm whale and horse heart myoglobins; the linear ligands used (methyl to n-heptyl) cover a greater distribution of chain lengths than hitherto used. The association rate constants are much higher for leghemoglobin than for myoglobin, while the dissociation rates are slower. For a given protein, the dissociation rate constants are not much different when different isocyanides are used (except for methyl), whereas the association rates show complex behavior in relation with the alkyl chain length; singular differences are observed between leghemoglobin and sperm whale myoglobin in this regard. For myoglobin, the binding rate constants decrease from methyl to n-propyl, but remain approximately the same when the ligand carries a still longer alkyl chain. In contrast, for leghemoglobin, although the rate constants decrease from methyl to n-propyl, they show a progressive and important rise with longer alkyl substituents: n-butyl and n-pentyl.     

Titration behavior of individual tyrosine residues of myoglobins from sperm whale, horse, and red kangaroo.

The titration behavior of individual tyrosine residues of myoglobins has been studied by observing the pH dependence of the chemical shifts of Czeta and Cgamma of these residues in natural abundance of 13C Fourier transform NMR spectra (at 15.18 MHz, in 20-mm sample tubes, at 37 degrees) of cyanoferrimyoglobins from sperm whale, horse, and red kangaroo. A comparison of the pH dependence of the spectra of the three proteins yielded specific assignments for the resonance of Tyr-151 (sperm whale) and Tyr-103 (sperm whale and horse). Selective proton decoupling yielded specific assignments for Czeta of Tyr-146 of the cyanoferrimyoglobins from horse and kangaroo, but not the corresponding assignment for sperm whale. The pH dependence of the chemical shifts indicated that only Tyr-151 and Tyr-103 are titratable tyrosine residues. Even at pH 12, Tyr-146 did not begin to titrate. The titration behavior of C zeta and Cgamma of Tyr-151 of sperm whale cyanoferrimyoglobin yielded a single pK value of 10.6. The pH dependence of the chemical shift of each of the resonances of Tyr-103 of the cyanoferrimyoglobins from horse and sperm whale could not be fitted with the use of a single pK value, but was consistent with two pK values (about 9.8 and 11.6). Furthermore, the resonances of Czeta and Cgamma of Tyr-103 broadened at high pH. The titration behavior of the tyrosines of sperm whale carbon monoxide myoglobin and horse ferrimyoglobin was also examined. A comparison of all the experimental results indicated that Tyr-151 is exposed to solvent, Tyr-146 is not exposed, and Tyr-103 exhibits intermediate behavior. These results for myoglobins in solution are consistent with expectations based on the crystal structure.     

A comparative study on axial coordination and ligand binding in ferric mini myoglobin and horse heart myoglobin

The absorption and resonance Raman spectra and the azide binding kinetics of ferric horse heart myoglobin (Mb) and mini myoglobin (a chemically truncated form of horse heart Mb containing residues 32-139) have been compared. The steady-state spectra show that an additional six-coordinated low-spin form (not present in entire horse heart Mb, which is purely six-coordinated high spin) predominates in mini Mb. The distal histidine is possibly the sixth ligand in this species. The presence of two species corresponds to a kinetic biphasicity for mini Mb that is not observed for horse heart Mb. Azide binds to horse heart Mb much more slowly than to sperm whale Mb. This difference may result from a sterically hindered distal pocket in horse heart Mb. In both cases, the rate constants level off at high azide concentrations, implying the existence of a rate-limiting step (likely referable to the dissociation of the axial sixth ligand). The faster rate constant of mini Mb is similar to that of sperm whale Mb, whereas the slower one is similar to that of entire horse heart Mb.     

Identification of the myoglobin tyrosyl radical by immuno-spin trapping and its dimerization

5,5-Dimethyl-1-pyrroline N-oxide (DMPO) spin trapping in conjunction with antibodies specific for the DMPO nitrone epitope was used on hydrogen peroxide-treated sperm whale and horse heart myoglobins to determine the site of protein nitrone adduct formation. The present study demonstrates that the sperm whale myoglobin tyrosyl radical, formed by hydrogen peroxide-dependent self-peroxidation, can either react with another tyrosyl radical, resulting in a dityrosine cross-linkage, or react with the spin trap DMPO to form a diamagnetic nitrone adduct. The reaction of sperm whale myoglobin with equimolar hydrogen peroxide resulted in the formation of a myoglobin dimer detectable by electrophoresis/protein staining. Addition of DMPO resulted in the trapping of the globin radical, which was detected by Western blot. The location of this adduct was demonstrated to be at tyrosine-103 by MS/MS and site-specific mutagenicity. Interestingly, formation of the myoglobin dimer, which is known to be formed primarily by cross-linkage of tyrosine-151, was inhibited by the addition of DMPO.     

Electron spin resonance spectra of free radicals formed in the reaction of metmyoglobins with ethylhydroperoxide

Free radicals of myoglobins were measured at room temperature with an ESR spectrometer equipped with a flow apparatus. When horse heart MetMb was mixed with an equimolar amount of ethyl hydroperoxide (EtOOH), a well resolved ESR spectrum with 6 lines and a shoulder was observed. It reached a maximum in a few seconds and decayed with a half-life of about 10 s when the final concentrations of MetMb and EtOOH were 200 microM. This decay rate was the same at a MetMb concentration of 50 microM. The maximum molar radical concentration amounted to about half of the total myoglobin. In the case of sperm whale myoglobin, a similar 6-line spectrum reached a maximum in 1 s and decayed with a half-life of a few seconds. In this case, however, a small and poorly resolved doublet spectrum remained, the half-life of which was about 8 min. An effect of O2 on the signal decay was evident for horse heart myoglobin, but not for sperm whale myoglobin.     

Reaction of haem containing proteins and enzymes with hydroperoxides: the radical view

The reaction between hydroperoxides and the haem group of proteins and enzymes is important for the function of many enzymes but has also been implicated in a number of pathological conditions where oxygen binding proteins interact with hydrogen peroxide or other peroxides. The haem group in the oxidized Fe3+ (ferric) state reacts with hydroperoxides with a formation of the Fe4+=O (oxoferryl) haem state and a free radical primarily located on the pi-system of the haem. The radical is then transferred to an amino acid residue of the protein and undergoes further transfer and transformation processes. The free radicals formed in this reaction are reviewed for a number of proteins and enzymes. Their previously published EPR spectra are analysed in a comparative way. The radicals directly detected in most systems are tyrosyl radicals and the peroxyl radicals formed on tryptophan and possibly cysteine. The locations of the radicals in the proteins have been reported as follows: Tyr133 in soybean leghaemoglobin; alphaTyr42, alphaTrp14, betaTrp15, betaCys93, (alphaTyr24-alphaHis20), all in the alpha- and beta-subunits of human haemoglobin; Tyr103, Tyr151 and Trp14 in sperm whale myoglobin; Tyr103, Tyr146 and Trp14 in horse myoglobin; Trp14, Tyr103 and Cys110 in human Mb. The sequence of events leading to radical formation, transformation and transfer, both intra- and intermolecularly, is considered. The free radicals induced by peroxides in the enzymes are reviewed. Those include: lignin peroxidase, cytochrome c peroxidase, cytochrome c oxidase, turnip isoperoxidase 7, bovine catalase, two isoforms of prostaglandin H synthase, Mycobacterium tuberculosis and Synechocystis PCC6803 catalase-peroxidases.