Peroxidase-catalyzed halide ion oxidation

Abstract

The first complete mechanistic analysis of halide ion oxidation by a peroxidase was that of iodide oxidation by horseradish peroxidase. It was shown conclusively that a two-electron oxidation of iodide by compound I was occurring. This implied that oxygen atom transfer was occurring from compound I to iodide, forming hypoiodous acid, HOI. Searches were conducted for other two-electron oxidations. It was found that sulfite was oxidized by a two-electron mechanism. Nitrite and sulfoxides were not. If a competing substrate reduces some compound I to compound II by the usual one-electron route, then compound II will compete for available halide. Thus compound II oxidizes iodide to an iodine atom, I^·, although at a slower rate than oxidation of I^- by compound I. An early hint that mammalian peroxidases were designed for halide ion oxidation was obtained in the reaction of lactoperoxidase compound II with iodide. The reaction was accelerated by excess iodide, indicating a co-operative effect. Among the heme peroxidases, only chloroperoxidase (for example from Caldariomyces fumago) and mammalian myeloperoxidase are able to oxidize chloride ion. There is not yet a consensus as to whether the chlorinating agent produced in a peroxidase-catalyzed reaction is hypochlorous acid (HOCl), enzyme-bound hypochlorous acid (either Fe–HOCl or X–HOCl where X is an amino acid residue), or molecular chlorine Cl₂. A study of the non-enzymatic iodination of tyrosine showed that the iodinating reagent was either HOI or I₂. It was impossible to tell which species because of the equilibria:

I₂+H₂O=HOI+I^-+H⁺</ p>

I^-+I₂=I₃^-

The same considerations apply to product analysis of an enzyme-catalyzed reaction. Detection of molecular chlorine Cl₂ does not prove it is the chlorinating species. If Cl₂ is in equilibrium with HOCl then one cannot tell which (if either) is the chlorinating reagent. Examples will be shown of evidence that peroxidase-bound hypochlorous acid is the chlorinating agent. Also a recent clarification of the mechanism of reaction of myeloperoxidase with hydrogen peroxide and chloride along with accurate determination of the elementary rate constants will be discussed.     

Impact of cyanogen iodide in killing of Escherichia coli by the lactoperoxidase-hydrogen peroxide-(pseudo)halide system

In the presence of hydrogen peroxide, the heme protein lactoperoxidase is able to oxidize thiocyanate and iodide to hypothiocyanite, reactive iodine species, and the inter(pseudo)halogen cyanogen iodide. The killing efficiency of these oxidants and of the lactoperoxidase-H₂O₂-SCN^?/I^? system was investigated on the bioluminescent Escherichia coli K12 strain that allows time-resolved determination of cell viability. Among the tested oxidants, cyanogen iodide was most efficient in killing E. coli, followed by reactive iodine species and hypothiocyanite. Thereby, the killing activity of the LPO-H₂O₂-SCN^?/I^? system was greatly enhanced in comparison to the sole application of iodide when I^? was applied in two- to twenty-fold excess over SCN^?. Further evidence for the contribution of cyanogen iodide in killing of E. coli was obtained by applying methionine. This amino acid disturbed the killing of E. coli mediated by reactive iodine species (partial inhibition) and cyanogen iodide (total inhibition), but not by hypothiocyanite. Changes in luminescence of E. coli cells correlate with measurements of colony forming units after incubation of cells with the LPO-H₂O₂-SCN^?/I^? system or with cyanogen iodide. Taken together, these results are important for the future optimization of the use of lactoperoxidase in biotechnological applications.     

Availability of iodide and iodate to spinach (Spinacia oleracea L.) in relation to total iodine in soil solution

A greenhouse pot experiment was carried out to investigate the availability of iodide and iodate to soil-grown spinach (Spinacia oleracea L.) in relation to total iodine concentration in soil solution. Four iodine concentrations (0, 0.5, 1, 2 mg kg⁻¹) for iodide (I⁻) and iodate (IO₃⁻) were used. Results showed that the biomass productions of spinach were not significantly affected by the addition of iodate and iodide to the soil, and that iodine concentrations in spinach plants on the basis of fresh weights increased with increasing addition of iodine. Iodine concentrations in tissues were much greater for plants grown with iodate than with iodide. In contrast to the iodide treatments, in iodate treatment leaves accounted for a larger fraction of the total plant iodine. The soil-to-leaf transfer factors (TF_leaf) for plants grown with iodate were about tenfold higher than those grown with iodide. Iodine concentrations in soil solution increased with increasing iodine additions to the soil irrespective of iodine species. However, total iodine in soil solution was generally higher for iodate treatments than iodide both in pots with and without spinach. According to these results, iodate can be considered as potential iodine fertilizer to increase iodine content in vegetables.     

Peroxidase catalyzed reactions of iodide at low pH

Lactoperoxidase (LP) and horseradish peroxidase (HRP) catalyze the rapid oxidation of iodide to iodine at pH 3.6. One mole of peroxide reacts with 2 moles of iodide, producing 1 mole of iodine. Neither enzyme catalyzes the further oxidation of iodine. The turnover numbers for LP and HRP are 1.4 × 10⁵ and 2.2 × 10⁴ I₂ moles produced/min/enzyme mole, respectively.     

Transport of sulphate, thiosulphate and iodide by choroid plexus in vitro

—Isolated choroid plexuses of rabbits and cats were incubated in artificial cerebrospinal fluid medium containing [³⁵S]sulphate, [³⁵S]thiosulphate or [¹²⁵I]iodide and combinations thereof. After 1 hr incubation the mean ratio of tissue concentration to medium concentration was 2·46 for [³⁵S]sulphate, 2·39 for [³⁵S]thiosulphate, and 270 for [¹²⁵I]iodide. Uptake of all three anions was greatly reduced at 0° and by addition of dinitrophenol to the medium. Other inhibitors selectively reduced the uptake of particular anions; non-radioactive sulphate and thiosulphate reduced both [³⁵S]sulphate and [³⁵S]-thiosulphate uptake with much less effect on [¹²⁵I]iodide uptake, while non-radioactive iodide and thiocyanate greatly reduced [¹²⁵]iodide uptake with little or no effect on [³⁵S]sulphate or [³⁵S]thiosulphate uptake. It was concluded: (a) that sulphate and thiosulphate, like iodide, were accumulated by choroid plexus in vitro by active transport; (b) that sulphate and thiosulphate share and compete for a transport mechanism which is separate from the iodide transport mechanism; and (c) that the transport of sulphate out of cerebrospinal fluid demonstrated in vivo could occur at least in part in the choroid plexus.     

The hemoglobin of the crossopterygian fish, Latimeria chalumnae (Smith). Subunit structure and oxygen equilibria

There are five oxidation-reduction states of horseradish peroxidase which are interconvertible. These states are ferrous, ferric, Compound II (ferryl), Compound I (primary compound of peroxidase and H₂O₂), and Compound III (oxy-ferrous). The presence of heme-linked ionization groups was confirmed in the ferrous enzyme by spectrophotometric and pH stat titration experiments. The values of pK were 5.87 for isoenzyme A and 7.17 for isoenzymes (B + C). The proton was released when the ferrous enzyme was oxidized to the ferric enzyme while the uptake of the proton occurred when the ferrous enzyme reacted with oxygen to form Compound III. The results could be explained by assuming that the heme-linked ionization group is in the vicinity of the sixth ligand and forms a stable hydrogen bond with the ligand.The measurements of uptake and release of protons in various reactions also yielded the following stoichiometries: Ferric peroxidase + H₂O₂ → Compound I, Compound I + e^? + H⁺ → Compound II, Compound II + e^? + H⁺ → ferric peroxidase, Compound II + H₂O₂ → Compound III, Compound III + 3e^? + 3H⁺ → ferric peroxidase.Based on the above stoichiometries and assuming the interaction between the sixth ligand and heme-linked ionization group of the protein, it was possible to picture simple models showing structural relations between five oxidation-reduction states of peroxidase. Tentative formulae are as follows: [Pr·Po·Fe-(II) $\text{\$}\text{?}$PrH⁺·Po·Fe(II)] is for the ferrous enzyme, Pr·Po·Fe(III)OH₂ for the ferric one, Pr·Po·Fe(IV)OH^? for Compound II, Pr(OH^?)·Po⁺·Fe(IV)OH^? for Compound I, and PrH⁺·Po·Fe(III)O₂^? for Compound III, in which Pr stands for protein and Po for porphyrin. And by Fe(IV)OH^?, for instance, is meant that OH^? is coordinated at the sixth position of the heme iron and the formal oxidation state of the iron is four.     

Iodine uptake in Laminariales involves extracellular, haloperoxidase-mediated oxidation of iodide

Sporophytes of Laminaria digitata (L.) Lamour. were assayed for their content of accumulated iodine, which ranged from 0.4% of dry weight in adult plants up to 4.7% for young plantlets. Sporophyte tissue from Laminaria saccharina (L.) Lamour. and L. digitata took up iodide according to Michaelis-Menten kinetics. Hydrogen peroxide and various substances known to interfere with oxidative metabolism were shown to either inhibit or enhance the uptake of iodide, confirming that apoplastic oxidations play a key role in iodide uptake in Laminaria. Consistently, iodide uptake was triggered in L. saccharina protoplasts by incubation in the presence of hydrogen peroxide. Similarly, the uptake of iodide was enhanced in L. digitata gametophytes by addition of haloperoxidase, suggesting that this enzyme catalyses the oxidation of iodide by hydrogen peroxide and plays a key role in iodine uptake. Oxidative stress resulted in a marked efflux of the intracellular iodine. In both influx and efflux experiments, a marked proportion (10–30%) of the tracer was not accounted for, indicating volatilisation of iodine. The mechanism and possible functions of the accumulation of iodine by kelps are discussed. Received: 11 February 1998 / Accepted: 18 June 1998     

Effects of hydrogen peroxide on neutrophil ability to generate reactive oxygen and chlorine species and to secrete myeloperoxidase in vitro

In this work, the effects of H₂O₂ at concentrations of 10^?8–10^?2 mol/l on the neutrophil ability to generate reactive oxygen and chlorine species (ROCS) and to secrete myeloperoxidase (MPO) were studied, as well as the H₂O₂ damaging action on neutrophils. It was found that H₂O₂ at concentrations of 2 × 10^?3–10^?2 mol/l led to disturbances of neutrophil membrane barrier properties and to a lactate dehydrogenase release. Incubation of neutrophils with an addition of 10^?4–10^?7 mol/l H₂O₂ was accompanied by an increase of the cell ability to generate ROCS during phagocytosis and a decrease of neutrophil ability to secrete MPO and ROCS into the extracellular medium during adhesion. Mechanisms of the H₂O₂ action are coupled with arachidonic acid metabolism. Inhibition of the 5-lipoxygenase or cyclooxygenase metabolism pathways produced an enhancement of the H₂O₂ destructive effect. Block of 5-lipoxygenase pathway led to elimination of the H₂O₂ action on MPO and ROCS secretion and to an enhancement of the H₂O₂ effect on the neutrophil ability to generate ROCS during phagocytosis. The obtained data indicate a high blood neutrophil resistance to the H₂O₂ destructive action and confirm the H₂O₂ regulatory role with respect to the neutrophil functions.     

Dissociation of DNA double strand by hypohalous acids

Abstract

Calf thymus DNA was treated with authentic HOCl, and hypohalous acid-generating systems. This caused a decrease in fluorescence of ethidium-DNA complexes when ethidium bromide was subsequently added to the DNA. The fluorescence continued to decrease up to 30 min after adding HOCl. Loss in fluorescence was proportional to the concentration of HOCl and was complete when a 3-fold excess of HOCl was added to the DNA. No significant decrease in the fluorescence was observed when the chlorination was carried out in the presence of a concentration of monochlorodimedone (MCD) equivalent to that of HOCl. MCD is known to react stoichiometrically with HOCl. The decrease in fluorescence was completely inhibited by H₂O₂, ascorbate and glutathione (GSH). We have estimated the rate constant for the reaction of HOCl with H₂O₂ to be 1–2×10⁵ M^-1s^-1. When compared with authentic HOCl, HOCl-generating systems (Cl^-+H₂O₂+MPO or chloroperoxidase) were found to be inefficient in damaging DNA. This result most likely arises because the rate constant for reaction of HOCl with H₂O₂ is about 1000-fold faster than that for the reaction with DNA. HOBr and HOI generating systems also had a limited ability to damage DNA. We conclude that good chlorine acceptors and antioxidants protect DNA from hypohalous acid-induced oxidative damage.     

Trypanosoma cruzi dihydrolipoamide dehydrogenase is inactivated by myeloperoxidase-generated “reactive species”

Dihydrolipoamide dehydrogenase (LADH) from Trypanosoma cruzi was inactivated by treatment with myeloperoxidase (MPO)-dependent systems. With MPO/H₂O₂/NaCl, LADH lipoamide reductase and diaphorase activities significantly decreased as a function of incubation time. Iodide, bromide, thiocyanide and chloride effectively supplemented the MPO/H₂O₂ system, KI and NaCl being the most and the least effective supplements, respectively. LADH inactivation by MPO/H₂O₂/NaCl and by NaOCl was similarly prevented by thiol compounds such as GSH, L-cysteine, N-acetylcysteine, penicillamine and N-(2-mercaptopropionyl-glycine) in agreement with the role of HOCl in LADH inactivation by MPO/H₂O₂/NaCl. LADH was also inactivated by MPO/NADH/halide, MPO/H₂O₂/NaNO₂ and MPO/NADH/NaNO₂ systems. Catalase prevented the action of the NADH-dependent systems, thus supporting H₂O₂ production by NADH-supplemented LADH. MPO inhibitors (4-aminobenzoic acid hydrazide, and isoniazid), GSH, L-cysteine, L-methionine and L-tryptophan prevented LADH inactivation by MPO/H₂O₂/NaNO₂. Other MPO systems inactivating LADH were (a) MPO/H₂O₂/chlorpromazine; (b) MPO/H₂O₂/monophenolic systems, including L-tyrosine, serotonin and acetaminophen and (c) MPO/H₂O₂/di- and polyphenolic systems, including norepinephrine, catechol, nordihydroguaiaretic acid, caffeic acid, quercetin and catechin. Comparison of the above effects and those previously reported with pig myocardial LADH indicates that both enzymes were similarly affected by the MPO-dependent systems, allowance being made for T. cruzi LADH diaphorase inactivation and the greater sensitivity of its LADH lipoamide reductase activity towards the MPO/H₂O₂/NaCl system and NaOCl.     

H2O2 mediates nitrate‐induced iron chlorosis by regulating iron homeostasis in rice

The uptake of nitrate by plant roots causes a pH increment in rhizosphere and leads to iron (Fe) deficiency in rice. However, little is known about the mechanism how the nitrate uptake‐induced high rhizosphere pH causes Fe deficiency. Here, we found that rice showed severe leaf chlorosis and large amounts of Fe plaque were aggregated on the root surface and intercellular space outside the exodermis in a form of ferrihydrite under alkaline conditions. In this case, there was significantly decreased Fe concentration in shoots, and the Fe deficiency responsive genes were strongly induced in the roots. The high rhizosphere pH induced excess hydrogen peroxide (H₂O₂) production in the epidermis due to the increasing expression of NADPH‐oxidase respiratory burst oxidase homolog 1, which enhanced root oxidation ability and improved the Fe plaque formation in rhizosphere. Further, the concentrated H₂O₂ regulated the phenylpropanoid metabolism with increased lignin biosynthesis and decreased phenolics secretion, which blocked apoplast Fe mobilization efficiency. These factors coordinately repressed the Fe utilization in rhizosphere and led to Fe deficiency in rice under high pH. In conclusion, our results demonstrate that nitrate uptake‐induced rhizosphere alkalization led to Fe deficiency in rice, through H₂O₂‐dependent manners of root oxidation ability and phenylpropanoid metabolism.     

Non-enzymatic formation of sulfenyl iodide by solubilized thyroid particulate protein with peroxidase activity

Thyroidal particulate protein with peroxidase activity has been studied to determined wetherr it could be induced to form a sulfenyl iodide, postulated as a reactive intermediate in the iodination of tyrosine. The protein was solubilized with digitonin and purified by tryptic digestion and filtration through Sephadex G-200. When supplemented with H₂O₂ it catalyzed the oxidation of guaiacol iodide or thiourea at 37°C. With iodide as substrate the product was an iodoprotein. In the absence of H₂O₂ the protein did not bind ¹³¹I^? or thio[¹⁴C] urea unless it first had been dialyzed against [³⁶Cl] chlorinated buffer. During dialysis a portion of the ³⁶Cl from the dialysis medium was bound by protein. Subsequent binding of iodide or thiourea was accompanied by loss of protein-bound ³⁶Cl.Addition of iodide to dialyzed protein at 4°C resulted in formation of a yellow compound with maximum absorbance at 355 nm. It was postulated to be a sulfenyl periodide on the basis of its absorption spectrum and its behavior with thio[¹⁴C] urea and 2-mercaptoethanol. The stability of the colored species was dependent on temperature and concentration of iodide. Disappearance of color as the solution was warmed was accompanied by formation of iodo-protein. Predialysis of the protein against p-chloromercuribenzoate, but not 2-mercaptoethanol or bisulfite, prevented the formation of the yellow proteiniodide species, indicating that a reactive sulfhydryl group was involved in the reaction. It was concluded that a particulate protein closely asscociated with thyroid peroxidase could be induced by non-enzymatic means to form a species which has properties consistent with those of a sulfenyl iodide. Further investigation will be required to determine whether the same protein-iodine species can be identified during the peroxidase-catalyzed oxidation of iodide.     

An essential role of active site arginine residue in iodide binding and histidine residue in electron transfer for iodide oxidation by horseradish peroxidase

The objective of the present study is to delineate the role of active site arginine and histidine residues of horseradish peroxidase (HRP) in controlling iodide oxidation using chemical modification technique. The arginine specific reagent, phenylglyoxal (PGO) irreversibly blocks iodide oxidation following pseudofirst order kinetics with second order rate constant of 25.12 min^-1 M^-1. Radiolabelled PGO incorporation studies indicate an essential role of a single arginine residue in enzyme inactivation. The enzyme can be protected both by iodide and an aromatic donor such as guaiacol. Moreover, guaiacol-protected enzyme can oxidise iodide and iodide-protected enzyme can oxidise guaiacol suggesting the regulatory role of the same active site arginine residue in both iodide and guaiacol binding. The protection constant (K_p) for iodide and guaiacol are 500 and 10 M respectively indicating higher affinity of guaiacol than iodide at this site. Donor binding studies indicate that guaiacol competitively inhibits iodide binding suggesting their interaction at the same binding site. Arginine-modified enzyme shows significant loss of iodide binding as shown by increased K_d value to 571 mM from the native enzyme (K_d = 150 mM). Although arginine-modified enzyme reacts with H₂O₂ to form compound II presumably at a slow rate, the latter is not reduced by iodide presumably due to low affinity binding.The role of the active site histidine residue in iodide oxidation was also studied after disubstitution reaction of the histidine imidazole nitrogens with diethylpyrocarbonate (DEPC), a histidine specific reagent. DEPC blocks iodide oxidation following pseudofirst order kinetics with second order rate constant of 0.66 min^-1 M^-1. Both the nitrogens (, ) of histidine imidazole were modified as evidenced by the characteristic peak at 222 nm. The enzyme is not protected by iodide suggesting that imidazolium ion is not involved in iodide binding. Moreover, DEPC-modified enzyme binds iodide similar to the native enzyme. However, the modified enzyme does not form compound II but forms compound I only with higher concentration of H₂O₂ suggesting the catalytic role of this histidine in the formation and autoreduction of compound I. Interestingly, compound I thus formed is not reduced by iodide indicating block of electron transport from the donor to the compound I. We suggest that an active site arginine residue regulates iodide binding while the histidine residue controls the electron transfer to the heme ferryl group during oxidation.     

Kinetic simulation studies on the transient formation of the oxo‐iron(IV) porphyrin radical cation during the reaction of iron(III) tetrakis‐5,10,15,20‐(N‐methyl‐4‐pyridyl)‐porphyrin with hydrogen peroxide in aqueous solution

High‐valent oxo‐iron(IV) species are commonly proposed as the key intermediates in the catalytic mechanisms of iron enzymes. Water‐soluble iron(III) tetrakis‐5,10,15,20‐(N‐methyl‐4‐pyridyl)porphyrin (Fe(III)TMPyP) has been used as a model of heme‐enzyme to catalyse the hydrogen peroxide (H₂O₂) oxidation of various organic compounds. However, the mechanism of the reaction of Fe(III)TMPyP with H₂O₂ has not been fully established. In this study, we have explored the kinetic simulation of the reaction of Fe(III)TMPyP with H₂O₂ and of the catalytic reactivity of FeTMPyP in the luminescent peroxidation of luminol. According to the mechanism that has been established in this work, Fe(III)TMPyP is oxidized by H₂O₂ to produce (TMPyP)^·+Fe(IV)=O (k1 = 4.5 × 10⁴/mol/L/s) as a precursor of TMPyPFe(IV)=O. The intermediate, (TMPyP)^·+Fe(IV)=O, represented nearly 2% of Fe(III)TMPyP but it does not accumulate in suf?cient concentration to be detected because its decay rate is too fast. Kinetic simulations showed that the proposed scheme is capable of reproducing the observed time courses of FeTMPyP in various oxidation states and the decay pro?les of the luminol chemiluminescence. It also shows that (TMPyP)^·+Fe(IV)=O is 100 times more reactive than TMPyPFe(IV)=O in most of the reactions. These two species are responsible for the initial sharp and the sustained luminol emissions, respectively. Copyright © 2003 John Wiley & Sons, Ltd.     

A chelate complex‐enhanced luminol system for selective determination of Co(II), Fe(II) and Cr(III)

A determination method for Co(II), Fe(II) and Cr(III) ions by luminol‐H₂O₂ system using chelating reagents is presented. A metal ion‐chelating ligand complex with a Co(II) ion and a chelating reagent like ethylenediaminetetraacetic acid (EDTA) produced highly enhanced chemiluminescence (CL) intensity as well as longer lifetime in the luminol‐H₂O₂ system compared to metals that exist as free ions. Whereas free Cu(II) and Pb(II) ions had a strong catalytic effect on the luminol‐H₂O₂ system, significantly, the complexes of Cu(II) and Pb(II) with chelating reagents lost their catalytic activity due to the chelating reagents acting as masking agents. Based on the observed phenomenon, it was possible to determine Co(II), Fe(II) and Cr(III) ions with enhanced sensitivity and selectivity using the chelating reagents of the luminol‐H₂O₂ system. The effects of ligand, H₂O₂ concentration, pH, buffer solution and concentrations of chelating reagents on CL intensity of the luminol‐H₂O₂ system were investigated and optimized for the determination of Co(II), Fe(II) and Cr(III) ions. Under optimized conditions, the calibration curve of metal ions was linear over the range of 2.0 × 10^‐8 to 2.0 × 10^‐5 M for Co(II), 1.0 × 10^‐7 to 2.0 × 10^‐5 M for Fe (II) and 2.0 × 10^‐7 to 1.0 × 10^‐4 M for Cr(III). Limits of detection (3σ/s) were 1.2 × 10^‐8, 4.0 × 10^‐8 and 1.2 × 10^‐7 M for Co(II), Fe(II) and Cr(III), respectively. Copyright © 2012 John Wiley & Sons, Ltd.     

Leukocytic myeloperoxidase-mediated formation of bromohydrins and lysophospholipids from unsaturated phosphatidylcholines

Using MALDI-TOF mass spectrometry, we have shown that leukocytic myeloperoxidase (MPO) in the presence of its substrates (H₂O₂ and Br^?) does not induce any changes in saturated 1,2-dipalmitoyl-sn-glycero-3-phosphocholine. Incubation of liposomes prepared from mono-unsaturated phosphatidylcholine (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) with the (MPO + H₂O₂ + Br^?) system resulted in formation of bromohydrins as the main products. 1-Palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine (lysophosphatidylcholine) was the main product of the reaction of polyunsaturated phosphatidylcholine (1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine) with the (MPO + H₂O₂ + Br^?) system. The formation of lysophospholipids as well as of bromohydrins was not observed when the enzyme or one of its substrates (H₂O₂ or Br^?) was absent from the incubation medium, or if an inhibitor of MPO (sodium azide) or hypobromite scavengers (taurine or methionine) were added. Thus, it can be postulated that the formation of bromohydrins as well as lysophospholipids by the (MPO + H₂O₂ + Br^?) system results from reactions of hypobromite formed during MPO catalysis with double bonds of acyl chains of phosphatidylcholine. Such destructive processes may take place in vivo in membrane-or lipoprotein-associated unsaturated lipids in centers of inflammation.