Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Immunomodulators that can improve both decreased and augmented immunological function are expected to provide fundamental treatment of rheumatoid arthritis. A series of thiol-containing immunomodulators such as D-penicillamine (Otomo et al., 1981; Nakaike et al., 1985), SA-96 (Fujimura et al., 1980), and tiopronin (Pasero et al., 1982) have exhibited various adverse effects, including thrombocytopenia, rash, and proteinuria (Craig and Buchanan, 1980; Suda et al., 1993). These toxic profiles are thought to be due to the high reactivity of thiol groups in the parent drugs or their metabolites with macromolecules in vivo (Coleman et al., 1988; Park and Kitteringham, 1990), and have resulted in the limited therapeutic potential of such agents.

Esonarimod (KE-298) [(±)-2-acetylthiomethyl-4-(4-methylphenyl)-4-oxobutanoic acid], has been shown to be an effective immunomodulator with inhibitory effects on inflammatory cytokine production (Kameo et al., 1988; Takeshita et al., 1988; Nagai et al., 1996; Takahashi et al., 1998). This compound contains a thioacetyl group and is easily deacetylated in vivo to form a pharmacologically active metabolite (thiol-containing deacetyl-esonarimod, M-I) that is a principal circulating metabolite (Yasuda et al., 1996; Yoshida et al., 1996). The toxicity of esonarimod is significantly lower than that of D-penicillamine in rats (unpublished data). It is suggested that the low toxicity may be closely related to the low reactivity of the thiol moiety of M-I with macromolecules in vivo (Yoshida et al., 1996). To support the low reactivity of M-I in vivo, the metabolic profile of esonarimod should be clarified. The objective of this study was to investigate the in vivo metabolism of esonarimod in rats by identifying its urinary and biliary metabolites using liquid chromatography/electrospray ionization-tandem mass spectrometry (LC/ESI-MS/MS¹) with postcolumn addition of 2-(2-methoxyethoxy)ethanol (2-MEE), a powerful signal-enhancing modifier (Yamaguchi et al., 1999).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. [¹⁴C]Esonarimod (Fig. 1) with a specific radioactivity of 762 kBq/mg was obtained from Amersham Pharmacia Biotech UK, Ltd. (Buckinghamshire, UK). Its radiochemical purity determined by radio-HPLC was not less than 98%. Unlabeled esonarimod and the following metabolites were synthesized at the Pharmaceutical Research Laboratories of Taisho Pharmaceutical Co., Ltd. (Saitama, Japan): KE-748 (M-II) [(±)-4-(4-methylphenyl)-2-methylthiomethyl-4-oxobutanoic acid]; KE-749 (M-III) [(±)-4-(4-methylphenyl)-2-methylsulfinylmethyl-4-oxobutanoic acid]; KE-767 (M-IV) [(±)-4-(4-hydroxymethylphenyl)-2-methylthiomethyl-4-oxobutanoic acid]; KE-768 (M-V) [(±)-4-(4-hydroxymethylphenyl)-2-methylsulfinylmethyl-4-oxobutanoic acid]; KE-766 (M-VII) [(±)-4-(4-carboxyphenyl)-2-methylsulfinylmethyl-4-oxobutanoic acid]; KE-298-19 (M-IX) [(±)-5-(4-hydroxymethylphenyl)-2,3-dihydro-3-thiophenecarboxylic acid]; and KE-764 (M-X) [(±)-5-(4-carboxyphenyl)-2,3-dihydro-3-thiophenecarboxylic acid]. 2-Methoxyethanol, 2-ethoxyethanol, and 2-MEE were obtained from Wako Pure Chemical Industries (Osaka, Japan). Methanol and acetonitrile (Wako Pure Chemical Industries) were of HPLC grade, and all other reagents were of analytical reagent grade.

View larger version (8K): [in this window] [in a new window]   Fig. 1.   Chemical structure of [14C]esonarimod. *, labeled position: ring-U-14C.

Animals. Male Wistar rats from Nihon SLC (Shizuoka, Japan), weighing between 190 and 210 g, were acclimated for at least 1 week in a 12-h light/dark cycle with free access to standard chow and water. They were fasted overnight before dosing and fed 4 h after dosing, and water was available freely during the fasting period.

Dosing and Sample Collection. ¹⁴C-Labeled or unlabeled esonarimod at a dose of 5 mg/kg was given orally to the animals as a suspension in 5% gum arabic. Three rats and one additional rat were administered [¹⁴C]esonarimod and the unlabeled drug, respectively, and urine specimens were then collected by housing rats individually in metabolic cages. For bile collection, [¹⁴C]esonarimod was administered to three bile duct-cannulated rats. Each such animal was then placed in a Bollman cage, and a bile specimen was collected over 24 h. In addition, a bile specimen was collected in the same manner from a rat administered unlabeled esonarimod. To prevent degradation of possible acyl glucuronide metabolites during sample collection (Benet and Spahn, 1988), biological fluids were collected into a plastic bottle immersed in ice, containing an appropriate volume of 0.5 M ammonium acetate (AcONH₄) buffer (pH 4.0), and then stored below 20°C until use.

Metabolite Identification.

Sample preparation An aliquot of the radioactive urine specimen, as a tracer, from a rat dosed with the labeled drug was mixed with an aliquot of the nonradioactive urine after the unlabeled drug administration to prepare a low-radioactive urine specimen for LC/ESI-MS/MS analysis (ca. 1000 dpm/µl). Similarly, a bile specimen with lowered radioactivity was obtained (ca. 140 dpm/µl). After filtration of these specimens with Millipore Ultrafree C3-LG tubes (0.22 µm, Tokyo, Japan), aliquots of the urine (20 µl) or bile (50 µl) were directly injected into an LC/ESI-MS/MS system.

LC/ESI-MS/MS. ESI-MS/MS with negative ion mode was carried out with a Finnigan MAT TSQ7000 triple-stage quadrupole tandem mass spectrometer equipped with a Finnigan MAT ESI source (San Jose, CA). The manifold temperature was set at 70°C, and ESI was performed at 4.0 kV with a heated capillary temperature of 200°C, a sheath gas (N₂) pressure of 70 psi, and an auxiliary gas (N₂) flow of 15 units. The product ion mass spectra of esonarimod and its metabolites were measured with a collision gas (Ar) pressure of 1.8 mTorr and a suitable collision offset voltage for each metabolite, ranging from 10 to 20 V. The apparatus was combined with an HPLC system constructed with components of Shiseido Nanospace SI-1 series (Tokyo, Japan), and a Raytest Ramona-2000 radioisotope detector equipped with a Raytest glass scintillation flow cell with a volume of 10 µl (Straubenhardt, Germany). Mobile phase was delivered with Pump 1 and/or Pump 2 at a flow rate of 100 µl/min. A postcolumn-adding modifier, at a flow rate of 40 µl/min, was delivered into the mobile phase from Pump 3 through a mixing tee (Upchurch Scientific, Oak Harbor, WA) that was fitted between the radioisotope detector and the ESI source.

Separating conditions-1. Mixtures of solvent A (20 mM AcONH₄ adjusted to pH 4.6 with AcOH delivered from Pump 1) and solvent B (acetonitrile delivered from Pump 2) were used as mobile phase. The linear gradient program was as follows: 0% solvent B for 2 min, and subsequent linear ramps from 0 to 10% (2 ~ 25 min), from 10 to 15% (25 ~ 70 min), from 15 to 20% (70 ~ 85 min) followed by a 15-min holding time (85 ~ 100 min), from 20 to 40% (100 ~ 120 min), and from 40 to 100% (120 ~135 min) followed by a 3-min holding time (135 ~ 138 min).

Separating conditions-2. Mixtures of solvent A (20 mM AcONH₄, pH 4.6, with AcOH) and solvent B (methanol) were used as a mobile phase. The linear gradient program was as follows: 2% solvent B for 10 min, and subsequent linear ramps from 2 to 10% (10 ~ 30 min), and from 10 to 20% (30 ~ 65 min).

Metabolite Quantification. For determination of metabolite composition, urine and bile specimens obtained after oral administration of [¹⁴C]esonarimod were subjected to radio-HPLC analysis, which was performed on a Jasco Gulliver HPLC system (Tokyo, Japan) with a Raytest Ramona-90 radioisotope detector equipped with a Raytest glass scintillation flow cell. Prior to analysis, the urine samples were filtered with the Ultrafree tubes. Separation was conducted with a YMC J'sphere ODS M-80 (150 × 4.6-mm i.d., 4 µm, Kyoto, Japan) with the following gradient elution (separating conditions-3). Mixtures of solvent A (20 mM AcONH₄, pH 4.6, with AcOH and solvent B (methanol) at a flow rate of 0.8 ml/min were used. The linear gradient program was as follows: 2% solvent B for 10 min, and subsequent linear ramps from 2 to 15% (10 ~ 25 min), from 15 to 20% (25 ~ 65 min), from 20 to 50% (65 ~ 75 min) followed by a 20-min holding time (75 ~ 95 min), and from 50 to 100% (95 ~ 125 min) followed by a 10-min holding time (125 ~ 135 min). After chromatography, the percentage of the radioactive metabolite in each peak, relative to the total urinary or biliary metabolites detected, was determined. The resulting value was multiplied by total excreted radioactivity in urine or bile, to calculate the percentage of dose excreted as each metabolite. In each case, recovery of radioactivity from the HPLC column was quantitative.

Measurement of Total Radioactivity. Radioactivity in urine and bile specimens was determined using a Beckman LS6000TA liquid scintillation counter (Fullerton, CA) with external standardization for quench correction. Each specimen was dissolved in a 10-ml aliquot of a Packard Instrument Co. Insta-Gel scintillation cocktail (Meriden, CT) prior to measurement.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

HPLC Profiles of Metabolites. After oral administration of 5 mg/kg [¹⁴C]esonarimod to rats, most (89%) of the radioactivity was excreted in urine over 24 h, whereas a small percentage (10%) of the radioactivity was recovered from bile, indicating that urinary excretion was the predominant route of elimination of this drug. Typical gradient reversed phase HPLC profiles of the urinary and biliary metabolites, excreted within 24 h after oral administration, with separating conditions-1 described under Materials and Methods are shown in Fig. 2, A and B, respectively. At least eight peaks were found in the urine, while only one predominant peak was detected in the bile. The radioactive peak designated peak 1 in Fig. 2A was completely separated into two peaks using separating conditions-2 (Fig. 3).

View larger version (37K): [in this window] [in a new window]   Fig. 2.   Typical HPLC radiochromatograms for rat urinary (A) and biliary (B) metabolites excreted within 24 h after oral administration of [14C]esonarimod (5 mg/kg). HPLC analysis was performed under separating conditions-1 described under Materials and Methods.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Typical HPLC radiochromatograms for rat urinary polar metabolites excreted within 24 h after oral administration of [14C]esonarimod (5 mg/kg). HPLC analysis was performed under separating conditions-2 described under Materials and Methods.

Optimization of ESI. To optimize ESI, the effects of postcolumn addition of various water-miscible organic solvents with higher boiling points (b.p.) than AcOH (b.p. 118°C) to the mobile phase on the sensitivity of KE-766, a putative metabolite of esonarimod, was examined (Fig. 4). Although higher modifier b.p. tended to increase signal-enhancing ability, the effects of 2-methoxyethanol (b.p. 124°C) and 2-ethoxyethanol (b.p. 135°C) were not significant compared with those of conventional modifiers, such as methanol (b.p. 64.7°C), acetonitrile (b.p. 81.6°C), and 2-propanol (b.p. 82.5°C). In contrast, 2-MEE, with a b.p. of 193°C, had 226 times the sensitivity of testing without addition of modifiers. Figure 5 compares the [M  H] intensity of KE-766 with those obtained with several AcONH₄ concentrations (0.5, 20, 50 mM) in the mobile phase, which were obtained without or with postcolumn addition of 2-propanol or 2-MEE. The postcolumn addition of 2-MEE prevented the decrease of the ESI responses resulting from increasing electrolyte concentrations in the mobile phase.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Effects of several postcolumn-adding modifiers on ESI responses of KE-766. The mobile phase was 20 mM AcONH4 with AcOH (pH 4.6), and each modifier was delivered at a flow rate of 40 µl/min. Data represent mean ± S.D. (n = 3).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Influence of AcONH4 content in the mobile phase (pH 4.6) on ESI responses of KE-766. The respective mobile phases with no organic solvent contained 0.5, 20, and 50 mM AcONH4. The experiment was carried out without (A) and with postcolumn addition of 2-propanol (B) or 2-MEE (C) at a flow rate of 40 µl/min. Data represent mean ± S.D. (n = 3).

View larger version (27K): [in this window] [in a new window]   Fig. 6.   ESI-mass chromatograms for rat urinary metabolites of esonarimod obtained in a Q1 full-scan mode without (A) and with (B) postcolumn addition of 2-MEE at a flow rate of 40 µl/min. The HPLC analysis was performed under separating conditions-1 described under Materials and Methods. Each arrow indicates the retention time of corresponding metabolite.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   ESI-mass chromatograms for rat urinary metabolites of esonarimod corresponding to peaks 2a and 2b obtained in Q1 full-scan mode without (A) and with (B) postcolumn addition of 2-MEE at a flow rate of 40 µl/min. The HPLC analysis was performed under separating conditions-2 described under Materials and Methods.

Metabolite Identification. The structures of metabolites were elucidated by their chromatographic behavior as well as mass spectral data summarized in Table 1.

Authentic esonarimod and KE-748. As references, the key collision-induced dissociation (CID) fragmentation patterns of authentic esonarimod and KE-748, a synthetic S-methyl metabolite of M-I (M-II), are shown in Fig. 8. Esonarimod gave an [M  H] at m/z 279 that underwent S-deacetylation subsequent to its dehydration by CID to yield m/z 237 (F1) and 219 (F2), respectively. The full-scan mass spectrum of M-II displayed an intense [M  H] at m/z 251, and the ion underwent elimination of CH₃SH and subsequent decarboxylation to form m/z 203 (F3) and 159 (F4), respectively. These characteristic product ions were used for structural analysis of the following in vivo metabolites.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Proposed key CID fragmentation schemes of esonarimod (A) and KE-748(B).

Metabolite M-III (peaks 4a and 4b). The full-scan mass spectrum from peak 4a gave an [M  H] at m/z 267, which was 16 unified atomic mass unit (u) greater than that of M-II. The product ion mass spectrum of the [M  H] included m/z 203 (F3) and 159 (F4), which were similar to those of M-II. In addition, m/z 63 corresponding to the -SOCH₃ moiety was detected. The mass spectral data for peak 4b were identical to those for peak 4a. Based on these results, the metabolite giving peaks 4a and 4b in reversed phase HPLC, i.e., M-III, was identified as the S-oxide metabolite of M-II, as a diastereomeric mixture. This was confirmed by comparison of the HPLC retention time and mass spectral data of M-III with those of synthetic standard (KE-749).

Metabolite M-IV (peak 6). The full-scan mass spectrum gave an [M  H] at m/z 267, similar to that of M-III, while in the product ion mass spectrum the ions corresponding to F3 and F4 were observed at m/z 219 and 175, respectively, which were 16 u greater than those of M-II. Based on these results, the metabolite giving peak 6, i.e., M-IV, was identified as the 4-hydroxymethylphenyl metabolite of M-II, and was confirmed by comparison of its retention time and mass spectral data of M-IV with those of synthetic standard (KE-767).

Metabolite M-V (peaks 2a and 2b). No ion peaks from peaks 2a and 2b were detected using separating conditions-1 due to peak overlapping with biological endogenous compounds. Separating conditions-2 were therefore used to achieve good separation, as shown in Fig. 7B. The full-scan mass spectrum for peak 2a gave an [M  H] at m/z 283, which was 16 u greater than those of M-III and M-IV. The product ion mass spectrum of m/z 283 demonstrated ions corresponding to F3 and F4 at m/z 219 and 175, respectively, which were similar to those of M-IV. The mass spectral data for peak 2b were identical to those for peak 2a. Based on these results, the metabolite giving peaks 2a and 2b, i.e., M-V, was identified as the 4-hydroxymethylphenyl metabolite of M-III, as a diastereomeric mixture. This was confirmed by comparison of the retention time and mass spectral data of M-V with those of synthetic standard (KE-768).

Metabolite M-VII (peaks 1a and 1b). The full-scan mass spectrum for peak 1a gave an [M  H] at m/z 297, which was 14 u greater than that of M-V. The product ion mass spectrum of m/z 297 gave an ion corresponding to F4 at m/z 189, which was 14 u greater than that of M-V. The mass spectral data for peak 1b were identical to those for peak 1a. Based on these results, the metabolite giving peaks 1a and 1b, i.e., M-VII, was identified as the 4-carboxyphenyl metabolite of M-III, as a diastereomeric mixture. This was confirmed by comparison of the HPLC retention time and mass spectral data of M-VII with those of synthetic standard (KE-766).

Metabolite M-IX (peak 5). The full-scan mass spectrum gave an [M  H] at m/z 235, which was 16 u greater than F2 (m/z 219) from esonarimod. The m/z 235 underwent decarboxylation by CID to yield m/z 191. Based on these results, the metabolite giving peak 5, i.e., M-IX, was identified as a 4-hydroxymethylphenyl-dihydrothiophen derivative, which was confirmed by comparison of its retention time and mass spectral data with those of synthetic standard (KE-298-19).

Metabolite M-X (peak 3). The full-scan mass spectrum included a weak [M  H] at m/z 249, 14 u greater than that of M-IX; however, an intense ion at m/z 205, which was probably formed by decarboxylation during ESI process, was detected. Under CID condition, the m/z 205 also underwent decarboxylation, indicating that this metabolite has two carboxyl groups. Based on these results, the metabolite giving peak 3, i.e., M-X, was identified as the 4-carboxyphenyl derivative of M-IX, which was confirmed by comparison of its retention time and mass spectral data with those of a synthetic standard (KE-764).

Metabolite M-II-Gluc (peak 7). The full-scan mass spectrum gave an [M  H] at m/z 427, which was 176 u greater than that of M-II. The product ion mass spectrum included an abundant ion at m/z 251 corresponding to the aglycon, as well as F3 (m/z 203) and F4 (m/z 159) ions that were similar to those of M-II. In addition, m/z 193 and 175, which were diagnostic ions for glucuronic acid moiety, were found. These results indicated that the metabolite giving peak 7 was the acyl glucuronide of M-II (M-II-Gluc).

Metabolite Quantification. The main urinary metabolite excreted within 24 h was M-III, accounting for 30.1 ± 0.8% of the dose, followed by M-V (16.6 ± 0.6%) and M-VII (13.2 ± 0.3%). In the case of bile, M-II-Gluc was predominant and accounted for 6.4 ± 0.3% of the dose. Each value represents mean ± S.E. for three rats.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our previous study demonstrated that [¹⁴C]esonarimod was rapidly and completely absorbed from the gastrointestinal tract after oral administration to rats, and a pharmacologically active deacetyl-esonarimod (M-I) and its S-methyl metabolite (M-II), which accounted for approximately 50 and 25%, respectively, of the plasma total radioactivity, were found with a small amount of the unchanged drug in plasma obtained at 30 min after dosing (Yoshida et al., 1996). In this present study, identification of the urinary and biliary metabolites of esonarimod was conducted to obtain further information on the in vivo metabolism of this drug in rats.

LC/ESI-MS/MS has been widely used in drug metabolism studies (Luffer-Atlas et al., 1997; Fernández-Metzler et al., 1999; Ramanathan et al., 2000). However, this technique has the serious drawback that analyte sensitivity is dependent upon the mobile-phase composition, such as organic solvent and electrolyte contents, as well as the nature of the analyte (Ikonomou et al., 1990; Jemal et al., 1998; Kamel et al., 1999). Our initial LC/ESI-MS/MS analysis in which a mobile phase containing AcONH₄/AcOH buffer, a common ESI-compatible buffer, was used failed to obtain any structural information on most of the esonarimod metabolites due to poor ESI responses. The ESI process involves the formation of highly charged droplets from the mobile phase, ion emission from the charged droplets into the gas phase, and gas-phase ion-molecular reaction (Gaskell, 1997). The poor ESI responses probably resulted from ion suppression caused by ion competition between the analyte and coexisting acetate anion (AcO) in the electrosprayed droplet surface during the ionization process (Witters et al., 1996). We recently demonstrated the usefulness of 2-MEE, an effective postcolumn-adding modifier, for negative ion ESI; this regent improves ion suppression by AcO (Yamaguchi et al., 1999). To optimize ESI for metabolite analysis of esonarimod, the effect of 2-MEE on sensitivity of synthetic KE-766, a putative metabolite of esonarimod, was examined. Unlike conventional signal-enhancing modifiers including 2-propanol, 2-MEE did not result in ion suppression, resulting in pronounced 226-fold signal enhancement. An effective modifier in improving ion suppression caused by AcO should possibly have a higher b.p. than that of AcOH (b.p. 118°C) as well as a surface tension-lowering property. The solvent 2-MEE (b.p. 193°C) has the desirable property, and probably makes possible the formation of smaller droplets containing lower percentages of water and AcO because of effective evaporation of AcOH from the droplets prior to analyte ion emission. This hypothesis may be supported by the b.p. dependence of the signal-enhancing ability of the surface tension-lowering solvents as indicated in Fig. 4. After method optimization, metabolite analysis was carried out. The postcolumn addition of 2-MEE allowed pronounced enhancement of ESI responses and structural elucidation for all metabolites without affecting chromatographic performance, indicating that this technique is very useful.

The results of the metabolite analysis demonstrated that esonarimod underwent extensive metabolism and was not excreted as the unchanged form. Moreover, M-I and M-II, which were circulating metabolites (Yoshida et al., 1996), were not clearly detected in either urine or bile specimens. The main urinary metabolite was M-III, followed by M-V and M-VII, which are S-methyl M-I-related metabolites. As minor components, dihydrothiophene derivative-related metabolites M-IX and M-X were identified, which were probably formed from M-I by intramolecular condensation of the thiol group with the C-4 carbonyl group and subsequent oxidation of the aromatic methyl group. In contrast, the biliary metabolite profile was qualitatively different from that for urine. One S-methyl M-I-related metabolite, M-II-Gluc, accounted for the majority of biliary radioactivity. On the basis of these findings, possible metabolic pathways of esonarimod were proposed as shown in Fig. 9. Esonarimod is rapidly biotransformed into an active metabolite M-I after drug absorption (Yoshida et al., 1996). This metabolite undergoes partial intramolecular condensation, but mainly S-methylation, followed by extensive metabolism by combination of S-oxidation and oxidative conversion of the aromatic methyl group.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Possible metabolic pathways of esonarimod in rats.

In general, thiol-containing drugs, such as D-penicillamine (Nozu et al., 1977; Pilkington and Waring, 1988), captopril (Yeung and Park, 1989), and SA-96 (Morikawa et al., 1985), tend to remain in the body due to the formation of disulfide bonds with macromolecules, resulting in various side effects (Yeung et al., 1983; Coleman et al., 1988; Park and Kitteringham, 1990). These thiol compounds are predominantly metabolized into disulfide metabolites such as dimers and cysteine-mixed disulfides via disulfide bonds (Bourke et al., 1984; Migdalof et al., 1984; Horiuchi et al., 1985), indicating high reactivity of their thiol moieties. In previous studies, esonarimod was not retained in tissues after oral administration of [¹⁴C]esonarimod to rats, and no disulfide metabolites were found in plasma (Yoshida et al., 1996). In the present study, moreover, no disulfide metabolites were found in excreta. These results strongly suggest that the reactivity of the thiol moiety of M-I with macromolecules in the body is much lower than that of common thiol-containing drugs, supporting the low toxicity of esonarimod in rats.

As shown in Fig. 9, M-I is considered not to be a simple thiol compound, but to exist as ketone-thiol and thiohemiacetal forms in a tautomeric equilibrium (Yoshida et al., 1996). The thiohemiacetal form may be much less reactive because of its masked thiol moiety in the molecule. Interestingly, M-I was relatively stable in plasma compared with D-penicillamine (Yoshida et al., 1996), but it was significantly unstable in phosphate buffer (pH 7.4) in which formation of its dimer occurred (data not shown). The plasma protein might fix the M-I tautomeric equilibrium to the thiohemiacetal form, resulting in the low reactivity of M-I in vivo. More detailed study of M-I reactivity should be performed.