Selenocysteine Lyase, a Novel Enzyme That Specifically Acts on Selenocysteine MAMMALIAN DISTRIBUTION AND PURIFICATION AND PROPERTIES OF PIG LIVER ENZYME*

We have found a novel enzyme that exclusively de- composes L-selenocysteine into L-alanine and HpSe in various mammalian tissues, and have named it sele- nocysteine lyase. The enzyme from pig liver has been purified to homogeneity. It has a molecular weight of approximately 85,000, and contains pyridoxal 5'-phos-phate as a coenzyme. Its maximum reactivity is at about pH 9.0. Balance studies showed that 1 mol of selenocysteine is converted to equimolar amounts of alanine and HzSe. The following amino acias are inert: L-cysteine, t-serine, L-cysteine sulfinate, selenocystea-mine, Se-ethyl-DL-selenocysteine, and L-selenohomo- cysteine. L-Cysteine (Ki, 1.0 m) competes with tsele-nocysteine (Km, 0.83 m ~ ) to inhibit the enzyme reaction. The enzyme is the first proven enzyme that specifically acts on selenium compounds. the of 0.1 ml 50% Alanine the reaction identified by amino acid analysis (retention time, 46.0 min) and gas chromatography of the N-heptafluorobutyryl n-propyl ester (retention time, 5.0 rnin). The mass spectrum of the N-heptafluorobutyryl n-propyl ester of the reaction product coincided with that of authentic alanine. We could not observe alanine the min), for DL-selenocysteine or rat liver homogenate, and with L-cysteine for DL-selenocysteine.

We have found a novel enzyme that exclusively decomposes L-selenocysteine into L-alanine and HpSe in various mammalian tissues, and have named it selenocysteine lyase. The enzyme from pig liver has been purified to homogeneity. It has a molecular weight of approximately 85,000, and contains pyridoxal 5'-phosphate as a coenzyme. Its maximum reactivity is at about pH 9.0. Balance studies showed that 1 mol of selenocysteine is converted to equimolar amounts of alanine and HzSe. The following amino acias are inert: L-cysteine, t-serine, L-cysteine sulfinate, selenocysteamine, Se-ethyl-DL-selenocysteine, and L-selenohomocysteine. L-Cysteine (Ki, 1.0 m) competes with tselenocysteine (Km, 0.83 m~) to inhibit the enzyme reaction.
The enzyme is the first proven enzyme that specifically acts on selenium compounds.
Selenium has been shown to be an essential micronutrient for mammals, birds, and several bacteria as reviewed by Stadtman (1) and Scott (2). The essentiality of selenium can be ascribed, at least partially, to the presence of enzymes that contain selenium as an integral component in these organisms. The selenium moiety of the following enzymes has been identified as a selenocysteine (2-amino-3-hydroselenopropionic acid) residue in the polypeptide chain: selenoprotein A of glycine reductase complex of Clostridium sticklandii (3), formate dehydrogenase of Methanococcus uannzelii (4), and glutathione peroxidase of rat liver ( 5 ) and of bovine erythrocytes (6). Evidence has been obtained for the participation of a selenocysteine residue in the catalytic processes (5, 6), but little attention has been paid to the biosynthesis and metabolism of selenocysteine. Recently, we showed that selenocysteine can be synthesized by the coupled reactions with cystathionine /?-synthase (EC 4.2.1.22) and cystathionine y-lyase (EC 4.4.1.1) purified from rat liver, and also by the reaction system with a rat liver homogenate (7). During   these enzyme reactions, we have found that the far less efficient selenocysteine formation with the homogenate system is due to the presence of a novel enzyme in the homogenate that decomposes specifically selenocysteine into alanine and H2Se. All the enzymes acting on selenium compounds so far studied inherently act on sulfur compounds (1). We have named the novel enzyme selenocysteine lyase and have puritied it to homogeneity from pig liver. We here describe distribution of the enzyme in mammalian tissues and the purification and properties of the pig liver enzyme.

Assay of Selenocysteine Lyase
All the reactions were performed in sealed tubes in which air was displaced by NB. Substrates and selenium-containing compounds such as DL-selenocysteine were freshly prepared from the corresponding diselenides by reduction with 5 mol of dithiothreitol per mol of diselenide in 0.2 M Tricine/NaOH buffer (pH 8.5) or 0.2 M sodium pyrophosphate buffer (pH 8.5) containing 33 p~ pyridoxal-P in a stream of Nz at room temperature (about 20 "C) for 10 min. The reaction was started by addition of enzyme to the above solution.
Method A: Assay of Crude Tissue Preparation-Selenocysteine lyase activity in the crude preparation of animal tissues was measured by determination of the alanine formed. The reaction mixture (final volume, 1.0 ml) contained 2 pnol of DL-selenocysteine, 5 pmol of dithiothreitol, 0.01 pmol of pyridoxal-P, 60 pmol of pyrophosphate buffer (pH 8.5), and a cell-free extract. Selenocysteine or enzyme was omitted in a blank. After incubation at 37 'C for 20 min, the reaction was terminated by addition of 0.1 ml of 50% trichloroacetic acid, followed by centrifugation. An aliquot of the supernatant solution was subjected to amino acid analysis (see below).
Method B: Standard Assay-The enzyme activity was measured by determination of H2Se2 with lead acetate. The standard reaction mixture contained 2 pmol of DL-selenocysteine, 5 pmol of dithiothreitol, 0.01 m o l of pyridoxal-P, 0.02 mg of bovine serum albumin (Sigma), 60 pmol of Tricine/NaOH buffer (pH 8.5), and enzyme in a final volume of 0.5 ml . The purified selenocysteine lyase is labile when the protein concentration is low. Bovine serum albumin (0.02 mg/0.5 ml) effectively protects the enzyme from inactivation. After incubation at 37 "C for 20 min, 3.5 d of 5 mM lead acetate solution in 0.1 N HC1 was added to the reaction mixture. Lead acetate solution was added to the reaction mixture prior to enzyme in a blank. Turbidity of yellowish brown PbSe colloid was measured at 400 nm within 15 min, because it diminishes at a rate of 0.3% per min. The complex of dithiothreitol with lead acetate (yellow precipitate) appears at neutral pH; dithiothreitol does not interfere with the determination of HzSe under the acidic conditions used. DL-Selenocysteine, L-selenohomocysteine, L-cysteine, and the other thiol compounds (2 pmol/0.5 ml of the reaction mixture) gave no precipitate under the conditions. Elemental selenium (5 mg) also did not react with lead acetate under the conditions used. However, when elemental selenium was added to the above reaction mixture (1 mg/0.5 ml), a large amount of colloidal PbSe appeared, indicating that elemental selenium is reduced to H2Se by dithiothreitol or selenocysteine. An apparent molar turbidity coefficient of colloidal PbSe at 400 nm was 2.36 X lo4 on the basis of determination of H*Se with 5,5'-dithiobis(2-nitrobenzoic acid) (E = 2.80 X IO4 at pH 7.4) (12). The amount of H2Se (0.015-0.15 pmol) added to the standard reaction mixture described above remained unaltered at least for 30 min when determined by the present method.

Determination of Chirality of Unreacted Selenocysteine with D -
Amino Acid Oxidase Selenocysteine remaining in the reaction mixture of selenocysteine lyase was derivatized into Se-ethylselenocysteine with ethyl iodide. The reaction mixture contained 5 pmol of dithiothreitol, 200 pmol of ethyl iodide, and 0.5 ml of the selenocysteine lyase reaction mixture in a final volume of 1.3 ml . After incubation at room temperature (about 20 "C) for 1 h, the reaction mixture was evaporated to dryness under reduced pressure, followed by dissolution in 0.5 ml of water. Selenocysteine was alkylated quantitatively under these conditions when the reaction was monitored by amino acid analysis. Oxidative deamination with amino acid oxidase was carried out in the following reaction mixture (a final volume, 1.0 m l ) at 37 "C for 20 min under vigorous shaking: 5 nmol of FAD, 0.14 unit of D-amino acid oxidase, 60 p o l of Tricine/NaOH buffer (pH 8.5), and the above sample solution (Sample I). The consumption of Se-ethylselenocysteine was measured by amino acid analysis.
Determination of Chirality of Alanine Formed with Alanine Dehydrogenase Alanine formed was oxidatively deaminated with alanine dehydrogenase to determine its chirality. The reaction mixture (0.55 ml) containing 3 p o l of NAD, 0.1 unit of alanine dehydrogenase, 0.1 unit of lactate dehydrogenase, 60 pmol of Tricine/NaOH buffer (pH 8.5), and 0.2 ml of the Sample I described above was incubated at 37 "C for 20 min to exhaust L-alanine. Alanine was determined by amino acid analysis before and after the alanine dehydrogenase reaction. *Although H2Se occurs as a dissociated form (HSeor Se2-) in aqueous solution, in particular at neutral and alkaline pH, we here express as HzSe for convenience.

Purification of Selenocysteine Lyase from Pig Liver
All steps were carried out at 4 "C unless otherwise stated. Potassium phosphate buffer (pH 7.4) containing 2 X M pyridoxal-P and 0.01% 2-mercaptoethanol was used as the standard buffer.
Step I : Preparation of Crude Extract-Fresh pig livers (6.5 kg) were minced with a meat mincer and subsequently homogenized a t full speed for 2 min with a Waring Blendor in 2 volumes of 0.05 M standard buffer. The homogenate was centrifuged at 10, OOO X g for 1 h.
Step 2: Heat Treatment-The supernatant solution was kept at 50 "C for 30 min and was centrifuged after cooling in ice.
Step 3: Ammonium Sulfate Fractionation-The supernatant solution (8,030 d) was brought to 25% saturation with solid ammonium sulfate and after standing for 20 min the precipitate was removed by centrifugation. Solid ammonium sulfate was slowly added to the supernatant solution to 45% saturation. The precipitate, collected by centrifugation, was dissolved in 900 ml of 0.01 M standard buffer and dialyzed against four changes of 10 liters each of the same buffer.
Step 4: DEAE-cellulose Column Chromatography-The dialyzed solution was applied to a DEAE-cellulose column (9.5 X 60 cm) equilibrated with 0.01 M standard buffer. After the column was washed with 5.0 liters of 0.01 M standard buffer, and then with 5.1 liters of the buffer containing 0.1 M KC1, the enzyme was eluted at a flow rate of 150 ml/h with 0.01 M buffer containing 0.14 M KCl. Fractions containing the enzyme were concentrated by the addition of ammonium sulfate (50% saturation), then dialyzed against 0.01 M standard buffer.
Step 5: First Hydroxyapatite Column Chromatography-The dialyzed solution was applied to a hydroxyapatite column (8 X 45 cm) equilibrated with 0.01 M standard buffer. The enzyme was eluted stepwise with 0.05 and 0.1 M standard buffer at a flow rate of 60 ml/ h. The active fractions were combined and concentrated with ammonium sulfate (50% saturation) and dialyzed against 0.01 M buffer.
Step 6: Second Hydroxyapatite Column Chromatography-The dialyzed enzyme was rechromatographed on a hydroxyapatite column (2.7 X 25 cm) equilibrated with 0.05 M buffer. The enzyme was eluted with 0.07 M buffer at a flow rate of 10 ml/h. The active fractions were combined and concentrated with ammonium sulfate (50% saturation) followed by dialysis against 0.01 M buffer. Step

Activity Staining
About 0.05 unit of the enzyme was subjected to disc gel electrophoresis by a modification of the procedure of Davis (19). After electrophoresis, the gel was soaked in a 6 times larger scale reaction mixture (3 ml) of Method B at 37 "C for 20 min. The gel was then washed briefly with water and soaked in 5 mM lead acetate solution in 0.1 N HC1. Active protein bands were colored yellowish brown. Yellow precipitste due to the complex of ditbiothreitol with lead acetate, which appeared on the gel surface, was rapidly diminished.

Enzymatic Cleavage of Selenocysteine
When DL-selenocysteine was incubated with a rat liver homogenate in the reaction mixture of Method A (see "Experimental Procedures"), the formation of alanine and HzSe was observed. HzSe was identified with 5,5'-dithiobis(2-nitrobenzoic acid) in a Thunberg tube containing the reaction mixture in the main compartment and a mixed solution (2 ml) of 20 pmol of 5,5'-dithiobis(2-nitrobenzoic acid) and 40 pmol of potassium phosphate buffer (pH 7.4) in the head compartment. After the tube was evacuated thoroughly, the reaction was started by tipping DL-selenocysteine into the main compartment and performed at 37 "C. The 5,5"dithiobis(Z-nitrobenzoic acid) solution in the head compartment was colored yellow (X,,, 412 nm) after 3 h, indicating that volatile HzSe formed from selenocysteine reduced 5,5'-dithiobis(2-nitrobenzoic acid). The production of HzSe was also confiied by the yellowish brown colloid formation with lead acetate under the acidic conditions (see "Experimental Procedures"). After 8 h, the reaction was terminated by the addition of 0.1 ml of 50% trichloroacetic acid. Alanine produced in the reaction mixture was identified by amino acid analysis (retention time, 46.0 min) and gas chromatography of the N-heptafluorobutyryl npropyl ester (retention time, 5.0 rnin). The mass spectrum of the N-heptafluorobutyryl n-propyl ester of the reaction product coincided with that of authentic alanine. We could not observe the alanine formation in the control experiments with boiled rat liver homogenate (boiled for 5 min), with water substituted for DL-selenocysteine or rat liver homogenate, and with L-cysteine for DL-selenocysteine.
Balance studies show that substantially equivalent amounts of alanine (0.93 pmol) and HzSe (1.04 pmol) are produced from DL-selenocysteine (2.0 pmol) with a rat liver homogenate (8.4 mg as protein) in the reaction mixture of Method B at 37 "C for 1 h. The alanine produced was reduced quantitatively with alanine dehydrogenase which specifically acts on L-alanine (9). However, the remaining selenocysteine was oxidized almost completely with D-amino acid oxidase after Se-ethylation with ethyl iodide. These results indicate that L-selenocysteine was converted into L-alanine and H2Se, and that the D-isomer of the DL-selenocysteine used as a substrate remained in the reaction mixture. 5-Mercaptoethanol and 2,3-dimercapto-1propanol can substitute for dithiothreitol in the reaction mixture of Method B. The following reducing agents were inert: NADH, NADPH, and L-ascorbic acid. We have found that these compounds cannot reduce selenocystine under the conditions used. These results show the occurrence of a new enzyme that catalyzes the degradation of selenocysteine into alanine and HzSe. This reaction apparently is a reduction, and we mentioned briefly the enzyme as "selenocysteine reductase'' in a previous paper (7). However, it is impossible to distinguish HzSe from Seo by any available method containing the present reactants. The presence of reductants in the reaction system is required to make selenocysteine from selenocystine, and to prevent selenocysteine from oxidation. Elemental selenium is reduced to H2Se by the reductants. In addition, selenocysteine, the substrate, serves also as a good reductant. Thus, we cannot decide whether H2Se or Seo is the actual product of the enzyme reaction. We have tentatively termed the enzyme selenocysteine lyase. A possible systematic name of the enzyme is selenocysteine-hydrogen selenide-lyase (alanine-forming) or selenocysteine-selenium-lyase (alanineforming).
Selenocysteine Lyase Activity of Seueral Mammalian Tissues The selenocysteine lyase activity was found in several animal tissues when assayed by the production of alanine from selenocysteine (Method A) ( Table I). We could not observe the formation of alanine from cysteine with all the rat tissues, bovine liver, and pig liver. We have also found that selenocysteine is a,P-eliminated too slowly with the homogenates of the above mammalian tissues to affect the amount of alanine formed from selenocysteine (0-5% of the selenocysteine lyase reaction rate). Thus, the formation of alanine through a transamination between pyruvate produced from selenocysteine and selenocysteine is not possible. We have found that selenocysteine lyase activities of livers and kidneys are generally higher than those of other tissues. Significant activity

Selenocysteine Lyase 4389
was found in rat pancreas, rat adrenal, and dog pancreas. However, no activity occurred in rat blood and rat fat. We have chosen pig liver for the purpose of purification of the enzyme in view of easy availability and abundance of enzyme activity.

Purification of Selenocysteine Lyase from Pig Liver
The pig liver enzyme was purified as described above and a summary of the purification is presented in Table 11. About one-half of the enzyme activity was lost by heat treatment of the enzyme at 50 "C for 30 min (Step 2 of the purification), but this treatment was essential to perform the subsequent column chromatographies effectively. The purified enzyme, which contains pyridoxal-P as described below, is completely devoid of the following activities that would possibly account for the selenocysteine lyase activity: cystathionine y-lyase, cystathionine /3-synthase, serine dehydratase, cysteine lyase, aspartate /3-decarboxylase, kynureninase, alanine aminotransferase, and serine hydroxymethyltransferase.
The purified enzyme can be stored a t -20 "C for a few weeks without loss of activity when protein concentration is more than 0.5 mg/ml, whereas it is inactivated significantly by freezing in a dilute solution (less than 0.06 mg/ml). However, the enzyme is fully stable in the presence of 20% sucrose, 20% glycerol, or 1% crystalline bovine serum albumin under the same conditions. Therefore, the enzyme was routinely stored in a deep-freeze (-20 "C) in the presence of 20% sucrose until use.

Properties of Purified Enzyme
Purity and Molecular Weight-The purified enzyme was found to be homogeneous by disc gel electrophoresis (Fig. 1). The single protein band stained with Amido black corresponded to the band stained by selenocysteine lyase activity. Homogeneity of the enzyme was demonstrated also by ultracentrifugation. The enzyme sedimented as a single and symmetrical peak during the sedimentation velocity run (Fig. 1). The sedimentation coefficient (smJ was calculated to be 5.5 S (20 "C; 10 mM potassium phosphate buffer (pH 7.4) containing 2 X M pyridoxal-P, 0.1 M KC1, and 0.01% 2mercaptoethanol; protein concentration, 2.0 mg). The molecular weight of the enzyme was estimated to be approximately 93,000 by the Sephadex G-200 gel fitration method. A molecular weight of 85,000 * 3,000 was obtained also by the sedimentation equilibrium method, assuming a partial specific volume of 0.74. Polyacrylamide gel electrophoresis in sodium lauryl sulfate gave a single major band that had an estimated molecular weight of 48,000. Therefore, the enzyme probably consists of two subunits with identical molecular weight.
Substrate Specificity-The ability of the enzyme to catalyze cleavage of various amino acids, in particular selenocysteine derivatives, was investigated. The enzyme was found to act specifically on selenocysteine to produce alanine and HpSe. From double reciprocal plots for the relationship between velocity and substrate concentration, the Michaelis constant for L-selenocysteine was calculated to be 0.83 mM. The following amino acids were inert: L-cysteine, L-serine, /3-chloro-Lalanine, L-cysteine sulfinate, S-methyl-L-cysteine, and Seethyl-DL-selenocysteine. Neither a-aminobutyrate nor H2Se (nor H&) was formed from the following amino acids or their derivatives: L-selenohomocysteine, L-homocysteine, and selenocysteamine. When dithiothreitol was omitted in the reaction mixture, no enzymatic cleavage was observed for DLselenocystine, L-cystine, L-selenohomocystine, and selenocystamine. L-Cysteine was found to behave as a competitive inhibitor against DL-SelenOCySteint? (Kt: 1.0 mM). None of the following compounds inhibited the enzyme reaction with 4 mM DL-selenocysteine: 5 mM L-serine, L-alanine, L-homocysteine, L-selenohomocysteine, and H2Se, and 10 mM glutathione.
Stoichiometry-Experiments to establish the stoichiometry of the enzymatic cleavage of selenocysteine were performed with a twice larger scale reaction mixture than that of Method B. The ratio of the amount of selenocysteine consumed to the amounts of alanine and H2Se formed was approximately 1:l:l ( Table 111). Amounts of dithiothreitol in 5 M excess over selenocysteine were routinely used to form selenocysteine as a substrate in the reaction mixture. In order to investigate the role of dithiothreitol in the enzyme reaction, we followed the reaction in the absence of extra dithiothreitol. When dithiothreitol was added in the reaction mixture with a molar ratio of 2:2 or 1.5:2 against DL-selenocystine, and determined with lead acetate prior to addition of enzyme, no reduced form of dithiothreitol remained in the reaction mixture. However, the enzyme reaction proceeded with virtually the same velocity as in the presence of 5 M excess amounts of dithiothreitol over selenocystine. Thus, the presence of extra dithiothreitol is not essential for the enzymatic cleavage of selenocysteine.
Role of Pyridoxal-P as a Cofactor-The absorption spectrum of selenocysteine lyase has an absorption maximum at 420 nm, which is characteristic of pyridoxal-P enzymes (Fig.  2). The enzyme, incubated with 10 mM hydroxylamine solution (pH 7.2), followed by dialysis against three changes of the standard buffer for 12 h, had no activity in the absence of added pyridoxal-P and no longer exhibited the absorption maximum a t 420 nm. Activity was about 94% restored by addition of 2 X M pyridoxal-P. The K , value for pyridoxal-P was estimated to be 3.3 X lo" M. The following vitamin Bs derivatives (2 X M) were not effective as a cofactor: pyridoxal, pyridoxamine 5'-phosphate, pyridoxine, and pyridoxine 5"phosphate. Reduction of the holoenzyme with sodium borohydride resulted in disappearance of the absorption  band at 420 nm with a concomitant increase in the absorbance at 330 nm. The reduced enzyme was catalytically inactive and the addition of pyridoxal-P did not reverse the inactivation. These results suggest that the borohydride reduces the aldimine linkage formed between the 4-formyl group of pyridoxal-P and an €-amino group of a lysine residue at the active site to yield the aldamine linkage as in other pyridoxal-P enzymes thus far studied.

DISCUSSION
Since selenocysteine was demonstrated as an essential constituent in polypeptide chains of several proteins, considerable attention has been given to the pathway of synthesis of the selenocysteine residue. Two possibilities were presented for the mechanism of selenocysteine residue synthesis of rat liver glutathione peroxidase: the post-translational Se incorporation to some precursor amino acid (e.g. dehydroalanine and cysteine) residue (23), and the direct incorporation of selenocysteine to the enzyme protein (24). The wide distribution of selenocysteine lyase demonstrated here may interfere with the exact observation of selenocysteine formation and incorporation into protein. Stadtman (25) has briefly reported a remarkable cleavage of selenocysteine by Clostridium sticklandii cells.3 Thus, it is requisite to take a breakdown of selenocysteine by selenocysteine lyase and some other enzyme into consideration for investigation of the mechanism of selenocysteine incorporation into protein.
Evidence has been obtained to show that selenocysteine is synthesized from selenomethionine derived from a diet (26) in the analogous pathway to the sulfur counterparts as follows: selenomethionine -+ Se-adenosylselenomethionine + Seadenosylselenohomocysteine + selenohomocysteine + selenocystathionine + selenocysteine (1,7,27), although other pathways of selenocysteine synthesis may be possible. Selenocysteine thus formed probably is incorporated into glutathione peroxidase or some other selenoprotein according to the direct incorporation hypothesis, and an excess of it is decomposed by selenocysteine lyase to alanine and H*Se, which was reported to be detoxified through methylation (28). Alternatively, according to the post-translational Se incorporation hypothesis, HzSe produced from selenocysteine by the enzyme is converted to the actual Se precursor to be incorporated into the proenzyme to form the selenocysteine resi- due. Based only on the K, value for L-selenocysteine (0.83 mM) and the K, value for L-cysteine (1.0 m~) obtained in vitro, the enzyme probably acts on selenocysteine very slowly in vivo, because the concentration of selenocysteine in the tissues is lower than the K,,, values4 However, the total activity of enzyme is most likely sufficient to metabolize a small amount of selenocysteine in the tissues. Further investigation, e.g. on the localization and compartmentation of the enzyme, the substrate, and the inhibitors, is needed to shed light on the physiological function of the enzyme.
The enzyme acting on sulfur compounds (e.g. mammalian cystathionine y-lyase (7) and bacterial L-methionine y-lyase (29)) work on the selenium analogs, although enzymes specific for selenium compounds have been considered (25). The indiscriminate catalytic action of enzymes on sulfur and selenium compounds probably is concerned at least partly in the toxicity of selenium compounds. Selenocysteine lyase, which does not act on cysteine at all, is unique with respect to the substrate specificity.
The selenocysteine lyase reaction is exceptional among those of the pyridoxal-P enzymes so far studied. The enzyme resembles the bacterial aspartate /3-decarboxylase (EC 4.1.1.12) (30) in the reaction mechanism where a moiety binding to CB of the substrate is cleaved to produce alanine. The following two mechanisms for the selenocysteine-lyase reaction are possible. Elemental selenium is released enzymatically from selenocysteine and reduced to H2Se nonenzymatically with selenocysteine or dithiothreitol contained in the reaction mixture. Alternatively, selenocysteine and dithiothreitol serve as reducing agents in the enzyme reaction and HzSe is produced inherently. The detailed mechanism of the enzyme reaction is currently under investigation.