Lactoperoxidase from human colostrum

Dual binding mode of antithyroid drug methimazole to mammalian heme peroxidases - structural determination of the lactoperoxidase-methimazole complex at 1.97 Å resolution

Lactoperoxidase (LPO, EC 1.11.1.7) is a member of the mammalian heme peroxidase family which also includes thyroid peroxidase (TPO). These two enzymes have a sequence homology of 76%. The structure of LPO is known but not that of TPO. In order to determine the mode of binding of antithyroid drugs to thyroid peroxidase, we have determined the crystal structure of LPO complexed with an antithyroid drug, methimazole (MMZ) at 1.97 Å resolution. LPO was isolated from caprine colostrum, purified to homogeneity and crystallized with 20% poly(ethylene glycol)‐3350. Crystals of LPO were soaked in a reservoir solution containing MMZ. The structure determination showed the presence of two crystallographically independent molecules in the asymmetric unit. Both molecules contained one molecule of MMZ, but with different orientations. MMZ was held tightly between the heme moiety on one side and the hydrophobic parts of the side chains of Arg255, Glu258, and Leu262 on the opposite side. The back of the cleft contained the side chains of Gln105 and His109 which also interacted with MMZ. In both orientations, MMZ had identical buried areas and formed a similar number of interactions. It appears that the molecules of MMZ can enter the substrate‐binding channel of LPO in two opposite orientations. But once they reach the distal heme pocket, their orientations are frozen due to equally tight packing of MMZ in both orientations. This is a novel example of an inhibitor binding to an enzyme with two orientations at the same site with nearly equal occupancies.

Lactoperoxidase: structural insights into the function,ligand binding and inhibition

Lactoperoxidase (LPO) is a member of a large group of mammalian heme peroxidases that include myeloperoxidase (MPO), eosinophil peroxidase (EPO) and thyroid peroxidase (TPO). The LPO is found in exocrine secretions including milk. It is responsible for the inactivation of a wide range of micro-organisms and hence, is an important component of defense mechanism in the body. With the help of hydrogen peroxide, it catalyzes the oxidation of halides, pseudohalides and organic aromatic molecules. Historically, LPO was isolated in 1943, nearly seventy years ago but its three-dimensional crystal structure has been elucidated only recently. This review provides various details of this protein from its discovery to understanding its structure, function and applications. In order to highlight species dependent variations in the structure and function of LPO, a detailed comparison of sequence, structure and function of LPO from various species have been made. The structural basis of ligand binding and distinctions in the modes of binding of substrates and inhibitors have been analyzed extensively.

Inhibition of lactoperoxidase by its own catalytic product: crystal structure of the hypothiocyanate-inhibited bovine lactoperoxidase at 2.3-A resolution

To the best of our knowledge, this is the first report on the structure of product-inhibited mammalian peroxidase. Lactoperoxidase is a heme containing an enzyme that catalyzes the inactivation of a wide range of microorganisms. In the presence of hydrogen peroxide, it preferentially converts thiocyanate ion into a toxic hypothiocyanate ion. Samples of bovine lactoperoxidase containing thiocyanate (SCN(-)) and hypothiocyanate (OSCN(-)) ions were purified and crystallized. The structure was determined at 2.3-A resolution and refined to R(cryst) and R(free) factors of 0.184 and 0.221, respectively. The determination of structure revealed the presence of an OSCN(-) ion at the distal heme cavity. The presence of OSCN(-) ions in crystal samples was also confirmed by chemical and spectroscopic analysis. The OSCN(-) ion interacts with the heme iron, Gln-105 N(epsilon1), His-109 N(epsilon2), and a water molecule W96. The sulfur atom of the OSCN(-) ion forms a hypervalent bond with a nitrogen atom of the pyrrole ring D of the heme moiety at an S-N distance of 2.8 A. The heme group is covalently bound to the protein through two ester linkages involving carboxylic groups of Glu-258 and Asp-108 and the modified methyl groups of pyrrole rings A and C, respectively. The heme moiety is significantly distorted from planarity, whereas pyrrole rings A, B, C, and D are essentially planar. The iron atom is displaced by approximately 0.2 A from the plane of the heme group toward the proximal site. The substrate channel resembles a long tunnel whose inner walls contain predominantly aromatic residues such as Phe-113, Phe-239, Phe-254, Phe-380, Phe-381, Phe-422, and Pro-424. A phosphorylated Ser-198 was evident at the surface, in the proximity of the calcium-binding channel.

Active site structure and catalytic mechanisms of human peroxidases

Myeloperoxidase (MPO), eosinophil peroxidase, lactoperoxidase, and thyroid peroxidase are heme-containing oxidoreductases (EC 1.7.1.11), which bind ligands and/or undergo a series of redox reactions. Though sharing functional and structural homology, reflecting their phylogenetic origin, differences are observed regarding their spectral features, substrate specificities, redox properties, and kinetics of interconversion of the relevant redox intermediates ferric and ferrous peroxidase, compound I, compound II, and compound III. Depending on substrate availability, these heme enzymes path through the halogenation cycle and/or the peroxidase cycle and/or act as poor (pseudo-)catalases. Based on the published crystal structures of free MPO and its complexes with cyanide, bromide and thiocyanate as well as on sequence analysis and modeling, we critically discuss structure-function relationships. This analysis highlights similarities and distinguishing features within the mammalian peroxidases and intents to provide the molecular and enzymatic basis to understand the prominent role of these heme enzymes in host defense against infection, hormone biosynthesis, and pathogenesis.

A probe for capture and Fe3+-induced conformational change of lactoferrin selected from phage displayed peptide libraries

Linear pentadecamer and cyclic hexamer peptide phage libraries were used to isolate phage clones with binding affinity toward lactoferrins purified from human and bovine milk. Phage clones with high specificity toward lactoferrin were selected with different binding strengths depending on the sequence of the peptide displayed by the phage. Phages coated to a microtiterplate were able to capture lactoferrin from crude milk samples without prior treatment. One of the selected sequences, EGKQRR, failed to bind to lactoferrin. In contrast, a branched tree-peptide bearing 4 EGKQRR sequences did bind to lactoferrin (Kd approximately 29 microM) and was also capable of inhibiting the binding of the phage to lactoferrin (IC(50) approximately 17 microM), indicating that avidity was important. Unexpectedly, the affinity of the phage for lactoferrin was influenced by the amount of bound Fe(3+), with a much lower affinity when lactoferrin was saturated with Fe(3+) as compared with the iron-depleted or partially saturated (natural) lactoferrin. As the phage does not bind to the Fe(3+)-binding site, the difference in binding affinity is due to differences in conformation of lactoferrin induced by Fe(3+). These results demonstrate that avidity or multipoint attachment and Fe(3+)-induced conformational changes play an important role in the binding of the selected phage to lactoferrin. Thus, we could demonstrate that, by the use of selected phage clones, we are able not only to detect lactoferrin, but also to capture lactoferrin from crude milk samples. Furthermore, the extent of phage binding provides additional information about the iron content and the concomitant conformation of lactoferrin.

Oxidation of mitoxantrone by lactoperoxidase

The lactoperoxidase (LPO) catalysed oxidation of mitoxantrone, an anthraquinone type anti-cancer drug, was studied spectrophotometrically under turnover and single turnover conditions with a stopped flow apparatus. With Compound I and Compound II, mitoxantrone formed binding complexes that were deactivated with increasing substrate concentration. The productive second-order rate constants for reduction were 3.6 x 10(6) and 2.2 x 10(4) M(-1) s(-1) for Compound I and Compound II, respectively. Under turnover conditions, Compound II was the steady-state intermediate, but with increasing H2O2, Compound II reacted with H2O2 to form the catalytically inactive intermediate Compound III. Nitrite prevented formation of Compound III by reducing Compound II to the native state. It also modulated the pathway of mitoxantrone oxidation by increasing the level of oxidised metabolites such as MH2(2+) and the novel metabolite MH. The biological implication of drug activation by LPO with nitrite is discussed.

Purification and quantification of lactoperoxidase in human milk with use of immunoadsorbents with antibodies against recombinant human lactoperoxidase

BACKGROUND

Two heme-containing peroxidases, secretory lactoperoxidase and leukocyte-derived myeloperoxidase, which play host defense roles through antimicrobial activity, were previously identified in human colostrum. Within several days after the start of lactation, the relative contribution of myeloperoxidase to the peroxidase activity in milk was shown to decline as the number of milk leukocytes decreased.

OBJECTIVE

Our knowledge of lactoperoxidase in human milk is still limited. The objective of this study was to use specific antibodies as a means of simplifying the purification and quantification of lactoperoxidase.

DESIGN

Polyclonal antibodies were raised against recombinant human lactoperoxidase. Immunoglobulin G (IgG) was isolated by means of a protein A column and was characterized by immunoblotting. For the purification of lactoperoxidase from whey, a cation-exchange column and an immunoaffinity column with coupled IgG were used. The concentration of lactoperoxidase was determined by a sandwich enzyme-linked immunosorbent assay by using purified native lactoperoxidase as a standard. Native and biotinylated IgG were used as capture and detector antibodies, respectively.

RESULTS

Two bands with molecular masses of approximately 80 and 100 kDa were detected in an immunoblot of human whey. Similar heterogeneity was observed in the sodium dodecyl sulfate-polyacrylamide gel electophoresis profile of purified lactoperoxidase. The mean (+/-SD) concentration of lactoperoxidase in 26 whey samples was estimated to be 0.77 +/- 0.38 mg/L. The concentrations were positively correlated with the peroxidase activity detected in these samples.

CONCLUSION

Lactoperoxidase is commonly present in human milk throughout the lactation period and is likely to contribute to the protective effects of milk.

PCR cloning and baculovirus expression of human lactoperoxidase and myeloperoxidase

Lactoperoxidase (LPO) and myeloperoxidase (MPO) have been identified previously in human milk. These peroxidases have antimicrobial activity and presumably contribute to the protective functions of milk. In this study, we amplified genes encoding LPO and MPO from human mammary gland cDNA by the polymerase chain reaction (PCR). These genes were expressed in a baculovirus-insect cell system. Peroxidase activity was observed in the culture supernatant of Tricoplusia ni cells infected with the recombinant viruses and the levels increased upon addition of delta-aminolevulinic acid. Purified recombinant human LPO and MPO, both with a molecular mass of about 80 kDa, showed properties similar to bovine LPO and human MPO, respectively, in terms of absorption spectrum, sensitivity to dapsone, specificity for chloride ions, and reactivity with anti-bovine LPO or anti-MPO antibodies. Our data suggest that this expression system is useful for studying the catalytic mechanism and biological significance of these human peroxidases.

Identification of lactoperoxidase in mature human milk

Myeloperoxidase (MPO) derived from milk leukocytes and lactoperoxidase (LPO) secreted from the mammary gland have been identified previously in human colostrum. These peroxidases are known to play host defensive roles through antimicrobial activity. The goals of this study were to measure the peroxidase activity in mature human milk and to characterize the enzyme responsible for the activity. As determined using 3,3',5,5'-tetramethylbenzidine as substrate, whey prepared from human milk samples obtained 1 and 5 months postpartum showed levels of peroxidase activity equivalent to 0.13 +/- 0.18 and 0.24 +/- 0.21 microg/mL bovine LPO (bLPO; n = 13), respectively. Whey from early milk was fractionated into two peaks of peroxidase activity by cation-exchange chromatography; the peroxidase in the first peak was sensitive to dapsone, which is an inhibitor of LPO, whereas the second peroxidase was not. Whey from mature milk showed only the first peak. Purified bLPO and MPO showed chromatographic behaviors that were similar to the first and second peaks, respectively. The dapsone-sensitive peroxidase from mature milk was further purified (952-fold from whey) by hydrophobic interaction chromatography. This preparation showed two bands with molecular masses of 80 and 90 kDa by polyacrylamide gel electrophoresis and immunoblotting using an antibody against bLPO. After deglycosylation, two distinct proteins with lower molecular weights were observed. Amino acid sequencing indicated that both of these proteins are LPO. These results provide evidence that LPO is present in mature human milk and that it is responsible for most of the peroxidase activity in mature milk.

Identification of minor proteins of human colostrum and mature milk by two-dimensional electrophoresis

Two-dimensional electrophoresis (2-DE) followed by electroblotting and microsequencing is considered to be the most powerful method for the isolation and characterization of proteins. In this paper, we report the separation and determination of the N-terminal and/or internal amino acid sequences of the minor proteins of human colostral and mature milk by 2-DE and microsequencing. In order to analyze the minor proteins of human milk, we use immunoabsorbents to remove three major proteins, alpha-lactalbumin, lactoferrin and secretory immunoglobulin A. The major proteins removed by this process accounted for about 79 and 93% of the total whey proteins of mature and colostral milk, respectively. The remaining milk proteins were then separated by isoelectric focusing gel electrophoresis between pH 3 and 10, and subjected to 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Approximately 400 spots were detected in both colostral and mature milk by silver staining after 2-DE. Twenty-two major, well-resolved proteins (out of 400) were microsequenced (N-termini as well as internal). These include fatty acid binding protein, beta 2-microglobulin, complement C4, clusterin, alpha 1-antritrypsin, lysozyme C, alpha- and beta-casein, prealbumin, serotransferrin, fructose-bisphosphate aldolase A, and beta-casein fragments. No major differences in the protein patterns were observed between the minor proteins of colostrum and mature milk, indicating that the minor proteins remained relatively constant during lactation. These results suggest that the minor milk proteins are important for the health and development of breast-fed infants throughout lactation.

Molecular cloning and characterization of the chromosomal gene for human lactoperoxidase

Lactoperoxidase (LPO) is an oxidoreductase secreted into milk, and plays an important role in protecting the lactating mammary gland and the intestinal tract of the newborn infants against pathogenic microorganisms. In this study, the human LPO chromosomal gene was molecularly cloned, and its gene organization was determined. The human LPO gene was found to be arranged with the myeloperoxidase (MPO) gene in a tail-to-tail manner. Similar to the human MPO and eosinophil peroxidase (EPO) genes, the human LPO gene is split by 11 introns and spans 28 kb. Unlike most introns in mammalian gene, the 5' splice donor sequence of intron 11 starts with GC instead of GT. When the minigene comprised of exon 11, intron 11 and exon 12 of the human LPO gene was introduced into COS cells, the correct splicing of the intron was found, suggesting the intron 11 of the human LPO gene is functional. The coding sequence of human LPO consists of 2136 bp, and codes for a protein of 712 amino acids. The amino acid sequence of human LPO has 51% similarity with those of both human MPO and EPO, suggesting that these peroxidase genes have evolved from a common ancestral gene. On the other hand, the nucleotide sequences of the 5' promoter regions of these peroxidase genes exhibit no similarity among them, which agrees with their tissue-specific expression.

Characterization of sheep lacrimal-gland peroxidase and its major physiological electron donor

A soluble sheep lacrimal-gland peroxidase was purified to apparent homogeneity. It had a native molecular mass of 75 kDa with a subunit molecular mass of 82 kDa and an isoelectric point of 6.5. Western blotting showed that it shares some of the enzyme antigenic determinants in common with other soluble peroxidases. The enzyme exhibits a Soret peak at 410 nm which is shifted to 431 nm by 5 equiv. of H2O2 due to the formation of compound II. The latter is, however, unstable and gradually returns to the native state. The enzyme forms complexes with CN- and N3- and is reduced by dithionite showing a characteristic reduced peroxidase spectrum. Although the enzyme oxidizes I-, SCN- and Br- optimally at pH 5.5., 5.25 and 5.0 respectively, at physiological pH, it oxidizes I- and SCN- only. Since extracellular SCN- concentration is much higher than I-, SCN- may act as the major electron donor to the enzyme. The second-order rate constants for the reaction of the enzyme with H2O2 (k+1) and of compound I with SCN- (k+2) were 4 X 10(7) M-1 X s-1 and 8.1 X 10(5) M-1 X s-1 respectively. A plot of log Vmax against pH yields a sigmoidal curve consistent with a single ionizable group on the enzyme with a pK(a) value of 5.75, controlling thiocyanate oxidation. In a coupled system with the peroxidase, H2O2, SCN-, GSH, NADPH and glutathione reductase, peroxidase-catalysed SCN- oxidation by H2O2 could be coupled to NADPH consumption. The system is proposed to operate in vivo for the efficient elimination of endogenous H2O2.

Spectral analysis of lactoperoxidase. Evidence for a common heme in mammalian peroxidases

The identity of the non-extractable heme of mammalian lactoperoxidase (LPO) has remained unsolved for over 40 years. Accepted possibilities include a constrained heme b or an 8-thiomethylene-modified heme b. Recent studies of myeloperoxidase (MPO) (Fenna, R., Zeng, J., and Davey, C. (1995) Arch. Biochem. Biophys. 316, 653-656; Taylor, K. L., Strobel, F., Yue, K. T., Ram, P., Pohl, J., Woods, A. S., and Kinkade, J. M., Jr. (1995) Arch. Biochem. Biophys. 316, 635-642) suggest possible prosthetic group similarities between MPO and LPO. To address heme identity for LPO, we used comparative magnetic circular dichroism (MCD) spectroscopy of LPO versus myoglobin (Mb), horseradish peroxidase (HRP), and MPO. MCD spectra of native Fe3+-LPO and Fe3+-CN--LPO are approximately 10 nm red shifted from analogous forms of Mb and HRP, including the formate-Mb adduct. MCD spectra of native LPO and MPO are opposite in sign, and MCD spectra of their cyanoadducts also differ. These data indicate the LPO heme is distinct from heme b of Mb and HRP as well as from "heme m" of MPO. From this work and literature analysis, we suggest that the non-extractable "heme l" of LPO has the two vinyl groups of heme b but lacks the 2-sulfonium-vinyl linkage of heme m. The observed red shifts in LPO spectra may derive from ester linkages to protein as for MPO. Strong spectral analogies between LPO and mammalian peroxidases (e.g. from saliva, eosinophils, thyroid, intestine) indicate similar prosthetic heme moieties.

A fluorometric assay of peroxidase activity utilizing 2',7'-dichlorofluorescein with thiocyanate: application to the study of salivary secretion

A sensitive assay for secretory peroxidase activity has been developed utilizing the fluorogenic substrate 2',7'-dichlorofluorescein in the presence of thiocyanate. The assay has been characterized using bovine lactoperoxidase and used to determine the peroxidase activities of salivas and extracts obtained from rat submandibular glands. Comparison of the 2',7'-dichlorofluorescein-thiocyanate assay and the commonly used 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) colorimetric assay indicates that the new assay is approx. 50-fold more sensitive. This has enabled measurement of peroxidase activities present in parasympathetic saliva samples which were beyond the detection limit of the colorimetric assay. Despite great differences in the peroxidase activities and protein concentrations of parasympathetic and sympathetic salivas and tissue extracts, the activities per unit protein were very similar. Unlike most other published methods, prior dialysis of samples to remove interference by endogenous thiocyanate is not required. The assay is therefore convenient and will be particularly useful for applications in which sample volume or peroxidase activity is low.

Tissue distribution of constitutive and induced soluble peroxidase in rat. Purification and characterization from lacrimal gland

A thorough search for a soluble peroxidase in 31 different tissues of rat indicated the presence of a constitutive activity only in lacrimal, preputial and submaxillary gland. An induced soluble peroxidase activity was also detected in the lactating mammary gland and in the estrogen-induced uterine secretory fluid. The lacrimal gland was the richest source of the enzyme. No peroxidase activity was detected in the lactating mammary gland of mouse and hamster nor in the preputial gland of mouse and uterine fluid of hamster. The three constitutive and two induced soluble peroxidases of rat had a native molecular mass of 73 kDa by gel filtration and they showed a similar mobility in native PAGE. Lactoperoxidase of cow's milk and solubilized rat membrane-bound peroxidases of uterus, intestine and bone marrow showed in native PAGE a mobility which was distinctly different from that of rat soluble peroxidases. As the lacrimal gland of rat was the richest source of soluble peroxidase, the enzyme was purified from this gland to apparent homogeneity; SDS/PAGE then showed a single band of molecular mass 75 kDa which was similar to that obtained by gel filtration. Peroxidase also purified from preputial and submaxillary gland, as well as commercial lactoperoxidase, had a similar molecular mass on SDS/PAGE to purified lacrimal peroxidase. The visible spectrum of lacrimal peroxidase was similar to that of lactoperoxidase but different from membrane-bound peroxidase of rat neutrophils. On isoelectric focussing, purified lacrimal peroxidase resolved into about 14 multiple forms spanning a pI range of 6.5-3.5 while lactoperoxidase focussed at the cathode. Evidence presented suggests that the multiple forms are possibly due to differences in glycosylation. Immunodiffusion, immunoprecipitation and Western blot using antilacrimal peroxidase serum showed a similar interacting species for all five soluble peroxidases of rat while membrane-bound peroxidases showed no interaction. Although in immunodiffusion, the antiserum failed to cross-react with lactoperoxidase it did interact with lactoperoxidase on Western blot. The results indicate that the various constitutive and induced soluble peroxidases of rat tissues are similar to lacrimal peroxidase but are distinctly different from the known membrane-bound peroxidases of rat. However the lacrimal peroxidase shows both similarities as well as dissimilarities with bovine lactoperoxidase. This soluble peroxidase system of rat could be useful to study tissue-specific regulation of gene expression at the molecular level.

Purification and characterization of a soluble peroxidase of rat preputial gland: comparison with lactoperoxidase

A highly active soluble peroxidase (donor: H2O2 oxidoreductase EC 1.11.1.7) has been purified from the preputial gland of the rat by hydroxylapatite chromatography, ammonium sulfate fractionation, Sephadex gel filtration and affinity chromatography on con A-Sepharose. The enzyme shows apparent homogeneity when analysed by acid and alkaline-PAGE. Its molecular, spectral, kinetic and catalytic properties were compared with those of bovine lactoperoxidase (LPO). Preputial gland peroxidase (PPO) is a glycoprotein of molecular weight of 70-73 kDa slightly lower (78 kDa) than that of LPO. Using isoelectric focussing, PPO was resolved into eight different closely spaced protein species spanning a pI range of 5.4 to 6.4, while LPO focuses into several closely spaced protein bands in the pI range 8.5-9.3. PPO is similar to LPO in its spectral (Soret) and some kinetic properties, but it differs significantly from LPO in substrate (H2O2) tolerance and substrate inactivation. PPO also differs from LPO in showing differential inactivation by SDS. Both enzymes are glycoproteins and although concanavalin A (con A) showed a variable interaction with both enzymes, wheat germ agglutinin interacted specifically with LPO only. We suggest that PPO, the secretory peroxidase of the preputial gland, differs significantly from LPO in its molecular and catalytic properties.

Sensitive immunometric assays for secretory peroxidase and myeloperoxidase in human saliva

We have developed specific immunoassays for secretory peroxidase (SP) and for myeloperoxidase (MP) (polymorphonuclear leukocyte-derived peroxidase) in human saliva. Antibodies against SP and MP were produced using bovine milk lactoperoxidase (LP) and human MP as the immunogens, respectively. The methods developed are non-isotopic immunometric assays using biotinylated antibodies and avidin-enzyme conjugate. The detection limit was 0.1 ng/ml and the performance time less than 3 h for both assays. The determination ranges were 0.5-100 ng LP (SP)/ml and 0.5-200 ng MP/ml with intra- and interassay CVs of 4.3% and 15.6% for SP and 3.7% and 10.8% for MP, respectively. The mean analytical recoveries were 108.9% (SP) and 91.5% (MP). These assays correlated well (r = 0.849-0.871) with the colorimetric assays based on the oxidation of thiocyanate or chloride by peroxidases. However, compared to the colorimetric methods the new immunometric assays are much more sensitive and specific for salivary SP and MP. The assays are also more rapid since extensive dialysis to remove endogenous thiocyanate is not required.

Molecular cloning of cDNAs encoding bovine and human lactoperoxidase

Peptide sequences obtained from cyanogen bromide fragments of bovine lactoperoxidase (bLPO) were used to design oligonucleotide probes for library screening. These probes were used to screen a cDNA library constructed from bovine mammary tissue. Three overlapping clones were obtained, the longest of which (T3) contained a reading frame of 712 amino acid residues. The encoded amino acid sequence was homologous to those recently reported for myelo-, thyro-, and eosinophil peroxidases. Two possible amino termini of the mature enzyme were identified, and the predicted mature protein matched previous molecular weight estimates of 78,500. Of eight bovine tissues tested, transcription of T3 sequences were detected in mammary tissue only. Using the bLPO cDNA as a probe, a single hybridizing clone was found in a human mammary gland cDNA library. This clone (M1) encoded the carboxy-terminal 324 residues of a peroxidase distinct from the other three known human peroxidases, and was closely related to bLPO. This result confirms the presence of at least one distinct lactoperoxidase in humans.