J. Biochem, 109, 909-917 (1991)

Cysteine Residues in the Active Site of Corynebacterium Sarcosine Oxidase Haruo Suzuki and Yasuko Kawamura-Konishi Department of Biophysical Chemistry, Kitasato University School of Medicine, Kitasato 1-15-1, Sagamihara, Kanagawa 228 Received for publication, December 25, 1990

Sarcosine oxidase from Corynebacterium sp. U-96 is inhibited by iodoacetamide (IAM) and the inhibition is prevented by the substrate analog, sodium acetate. To elucidate the mechanism of inhibition of the enzyme by IAM, we determined the amino acid sequences around the IAM-reactive cysteine residues, and the effects of the modification on the enzyme activity and the oxidation-reduction of the FAD moieties of the enzyme. The enzyme was specifically labeled with [ UC] IAM, and the labeled subunit B was digested with trypsin and chymotrypsin. The HPLC profiles of the proteolytic digests showed mainly two radioactive peaks. The "C-labeled peptides were purified, and their N-terminal sequences were determined to be Cys-Gly-Thr-Pro-Gly-Ala-Gly-Tyr (TC-1) and Ala-Gly-Ile-Ala-CysXaa-Asp-Xaa-Val-Ala-- (TC-2). Peptide TC-2 contains a covalent FAD-binding sequence [Asx-His-Val-Ala; Shiga et al. (1983) Biochem. Int., 6, 737]. [uC]IAM-incorporation into the TC-1 sequence was strongly inhibited by sodium acetate. The N-terminal amino acid sequence of the CNBr fragment containing the TC-1 sequence (65 residues) was determined. According to the secondary structure predictions, Gly-Thr-Pro-Gly-Ala-Gly of the TC-1 sequence is located between the $ sheet and a helix of the sequence, indicating the presence of an AMP-binding site in the TC-1 region. The activity of the enzyme treated with IAM in the presence and absence of sodium acetate was not inhibited by sodium sulfite, which is known to react specifically with covalent FAD. The time courses of reduction of the flavin moieties of the IAM-treated enzymes by sarcosine showed fast and slow phases. In the presence and absence of sulfite, the rates of the fast phase were almost the same for the enzymes treated in the presence and absence of sodium acetate. But the fractions for the fast phase for the enzymes treated in the presence and absence of acetate were 55 and 20% of the total absorbance change (the oxidized minus the reduced form) in the case of the native enzyme, respectively. The rates of oxidation of the IAM-treated enzymes were approximately half of that in the case of the native enzyme. These data suggest that at least two cysteine residues in the enzyme are located at different sites on subunit B, one at the sarcosine-binding site and the other at the covalent FAD-binding site, and that IAMtreatment of the enzyme impairs the activity of the covalent FAD and thus the noncovalent FAD mainly functions in the oxidation of sarcosine.

Sarcosine oxidase [sarcosine:oxygen oxidoreductase (demethylating), EC 1.5.3.1] catalyzes the oxidative demethylation of sarcosine to yield formaldehyde, glycine, and hydrogen peroxide. The enzyme isolated from Corynebacterium sp. U-96 consists of four nonidentical subunits (A, Mr 110,000; B, Mr 44,000; C, Mr 21,000; and D, M10,000), and contains both covalently bound and noncovalently bound FAD in a molar ratio of 1 : 1 (1, 2). The different roles of theflavinsin catalysis have been postulated from the result of the titration of the enzyme with dithiothreitol (3) and the substrate, sarcosine (4). Jorns (4) proposed that the noncovalent FAD functions as a "dehydrogenase1' flavin and the covalent FAD as an "oxidase" flavin. We studied the enzyme reaction as to the rapid reaction kinetics and explained the data by assuming the different roles of the flavins (5).

The enzyme contains 8 sulfhydryl groups per mol and sodium acetate inhibits the enzyme competitively with respect to sarcosine (6, 7). It was also reported that iodoacetamide (IAM) inactivates the enzyme and the inactivation is prevented by the substrate analog, acetate (6). These findings suggest the presence of a sulfhydryl group at the substrate-binding site. Using [14C]IAM, we showed that subunit B was specifically labeled with IAM and that the labeling was prevented by acetate (8). Therefore we proposed that the noncovalently bound FAD is also bound to subunit B (8). This paper describes the modification of the enzyme with [UC]IAM and determination of the amino acid sequences around the labeled cysteine residues. We also compare the reactivity of sarcosine with the bound FAD of the native and IAM-treated enzymes in the presence and absence of

Abbreviations: IAM, iodoacetamide; IAM ( + A) enryme and IAM S^ 1 " 1 1 s u l f i t e . 8 m c e 8 u l f i t e r e a c t s specifically with the (-A)eniyme, enzymes treated with IAM in the presence and absence covalent FAD (4). The data suggest that one sulfhydryl of sodium acetate, respectively. group is located at the substrate-binding site (possibly the Vol. 109, No 6, 1991

909

H. Suzuki and Y. Kawamura-Konishi

910 noncovalent FAD-binding site), and the other at the covalent FAD-binding site, and that LAM modification reduces the reactivity of the covalent FAD and thus that the noncovalent FAD mainly functions in the oxidation of sarcosine.

h at 110'C in 6 N HC1-2 mM phenol in the vapor phase using a Waters Work Station. The hydrolysate was derivatized with phenylisothiocyanate as described by Holmes et al. (11). Phenylthiocarbamoyl amino acids were separated on a PICO-Tag d 8 column as described in a Waters Associates mnnnal

MATERIALS AND METHODS

Materials—LAM was obtained from Wako Pure Chemical, and [1-UC]LAM (53mCi/mmol), from Amersham, was diluted with the LAM. The [ UC]LAM used had a specific activity of 17.9 dpm/pmol. Other chemicals used were of the highest grade of purity commercially available. Corynebacterium sarcosine oxidase was purified as described (1) and kindly supplied by Dr. Masaru Suzuki, Noda Institute for Scientific Research. The concentrations of the enzyme and the subunit were determined by measuring the absorbance at 280 or 455 nm, using the reported absorption coefficients (1). Buffer A comprised 20 mM pyrophosphate-HCl (pH 8.0)-0.1 mM EDTA. Assaying of the Enzyme Activity—The activity of the native and LAM-treated enzymes was assayed by measuring the oxygen uptake, using an oxygen electrode (Yellow Springs Instruments, U.S.A.) in the presence of 1 mM sarcosine in buffer A at 25'C. Labeling of the Enzyme with [ U C]LAM—The enzyme was incubated with 1 mM [14C]LAM for 30 min at 30'C in the presence and absence of 50 mM sodium acetate in buffer A, and then LAM was added to the final concentration of 25 mM. After boiling the labeled enzyme in 31 mM Tris-HCl(pH 6.8), 1% SDS, 5% glycerol, and 2.5% mercaptoethanol for 5 min, the labeled subunit B was separated from the other subunits by SDS/PAGE. The subunit was eluted from the gels electrophoretically (9) and then concentrated with a Centricon 10 (Amicon). SDS remaining in the solution was removed by the addition of potassium phosphate as described (20). Fragmentation of the Labeled Subunit B—The labeled subunit B was digested with chymotrypsin and trypsin for 24 h at 37'C in 1% NH4HCO3. The protein to enzyme ratio was 20 : 1 (w/w). The subunit was also cleaved with CNBr as follows. The subunit prepared as above was precipitated by the addition of acetone and then dried. The dried precipitate ( 230 //g) was dissolved in 70% formic acid and then CNBr in 70% formic acid was added to 2%. The whole mixture (100 //I) was kept at room temperature for 24 h. CNBr used was purified by sublimation just before use. Purification of the Labeled Peptides by HPLC—Peptides were purified by HPLC using a JASCO Trirotor ILL High Pressure Liquid Chromatograph (Japan Spectroscopic). The u C-labeled peptides in the proteolybc digests of the labeled subunit B were purified on HPLC columns of Unisil Q18 (5 //m, 4.6 X 250 mm) and then Inertsil ODS-2 (5 //m, 4.6x250 mm) from GL Sciences (Tokyo), as described in the legends to the figures. The subunit B treated with CNBr was directly applied on a HPLC column of TSK gel (ODS120T, 6.0 X 150 mm, 5 //m ) from Tosoh, Tokyo, and then the peptides were eluted from the column as described in the relevant figure. FAD-containing fractions were detected qualitatively by irradiating the fractions with UV light (Mineralight Lamp, model UVGL-58; San Gabriel, California, U.S.A.). Amino Acid Analysis—Peptides were hydrolyzed for 22

Amino Acid Sequence Analyses of Peptides—Automated sequence analyses of the purified peptides were carried out using a gas-phase sequencer (Applied Biosystem, model 470A) as described previously (12). Cys residues were determined by measuring the radioactivity of the fractions after each Edman cycle after the addition of 3 ml of Aquasol-2 (New England Nuclear). Spectrophotometry—Anaerobic reduction of the enzyme with 0.25 mM sarcosine was monitored by obtaining a spectrum in the visible wavelength region using a Thunberg-type cuvette under anaerobic conditions. Anaerobisis was attained by degassing and flushing with nitrogen gas 5 times. A double beam spectrophotometer, type Ubest-50, from Japan Spectroscopic was used. Flow Experiments—The rates of oxidation and reduction of the FAD moieties of the enzyme were determined by monitoring the absorbance change at 455 nm, using a rapid reaction analyzer, USP-516, from Unisoku-ScientificInstruments, Japan. The enzyme was treated with 1 mM LAM for 30 min at 30'C in buffer A and then the whole mixture was dialyzed against buffer A at 4"C. The rate of reduction of the treated enzymes was measured by mixing them with 1 mM sarcosine, using a stopped-flow apparatus, under anaerobic conditions. Anaerobisis was attained by using a glucose-glucose oxidase-catalase system (5). For measurement of the rates of oxidation of the treated enzymes, 250//M sarcosine was added to each enzyme solution in buffer A under a nitrogen atmosphere. The solution was kept for a few min at 25"C to obtain the reduced form of the enzyme. When we used a lower concentration of sarcosine to consume the dissolved oxygen and to reduce the enzyme, we could not get the reduced form of the enzyme. The reduced form of the enzyme was rapidly mixed with 120//M O2 and then the absorbance change at 455 nm was monitored. As a control, similar experiments were performed with the native enzyme. The first-order rate was estimated by the Damping GaussNewton method (13). RESULTS Amino Acid Sequences of the IAM-Reactive Cysteine Residues—We previously reported that sarcosine oxidase from Corynebacterium sp. U-96 was inactivated by IAM, that the inactivation was linearly related to the incorporation of [ 14C]IAM into subunit B and that almost no incorporation into other subunits was observed (8). When the native enzyme was incubated with [14C]IAM for 30 min at 30"C, as described under "MATERIALS AND METHODS," 1.7 mol of IAM was incorporated into subunit B, and by extrapolation to complete inactivation of the enzyme, approximately 2 mol of IAM was calculated to be incorporated into subunit B. To characterize the sequences around the IAM-reactive Cys residues, we labeled the enzyme with l m M [MC]IAM for 30 min at 30"C. Then the labeled subunit B was obtained and digested with proteolytic enzymes, as described under "MATERIALS AND METHODS." J Bwchem.

911

Active Site SH Groups of Sarcosine Oxidase The uC-labeled subunit B was digested with chymotrypsin and trypsin extensively. The HPLC profile of the digests is shown in Fig. 1A. Two radioactive peaks were mainly observed, the yield of the two peak materials being 73% of the radioactivity applied. The ratio of the TC-1 and TC-2 peak materials was 1:1.35. The first radioactive peak material (TC-1) was collected and the radioactive peptide was further purified on another HPLC column, Inertsil ODS-2 (Fig. IB). The yield of the radioactive peptide was 83%. The N-terminal sequence of the TC-1 peptide was analyzed with a gas-phase sequencer (Fig. 2A). The 14 C-labeled peptide in the second peak material (TC-2) was further purified by HPLC on a column of Inertsil ODS-2 (Fig. 1C). The yields of the TC-2 and TC-(3 + 4) peptides were 44 and 29%, respectively. The N-terminal amino acid sequence of the labeled and purified peptide (TC-2) was analyzed (see Fig. 2B). The amino acid composition of the TC-1 peptide agreed well with the sequence data, but the amino acid composition of the TC-2 peptide did not explain the sequence data well (see the legend to Fig. 2). This may be ascribable to the fact that the sequence determined did

not represent the full length of the TC-2 peptide (see the sequence of CNBr-1 in Table II) and partially due to a contaminating peptide. The second peak material (TC-2 in Fig. 1A) contained a small amount of 14C-labeled peptides which were different from the TC-2 sequence in Fig. 2. Though we could not purify these peptides well by using another HPLC column, their possible N-terminal sequences are Val-Asn-Cys-Gly- (TC-3 in Table I) and GlyVal-Cys-Pro-Ile-Asn-Thr-Gly-Asp-Asn-Ile- (TC-4 in Table I). It would be interesting to know if there is any difference in the extent of IAM modification of these Cys residues. To determine the relative 14C-incorporation into these Cys residues, the specific radioactivities of TC peptides were calculated by measuring the radioactivity of fractions collected during Edman degradation. The amount of PTHcarboxyamidated Cys was calculated by using the repetitive yield of the degradation determined from the plot as in Fig. 2. The specific activities of TC-1, 2, 3, and 4 were calculated to be 10.3, 12.8, 7.2, and 2.7 dpm/pmol, respectively. Therefore, the relative incorporation into TC-1, 2, Fig 1. A) HPLC profiles of the tryptic and chymotryptic peptides of the "C-labeled subunit B. The 14C-labeled subunit B (27,000 dpm) was digested with trypsin and chymotrypsin as described under "MATERIALS AND METHODS." The digests were applied on a HPLC column of Unisil and the eluates were collected in 0 5 ml fractions. A 20 p 1 aliquot of each fraction was counted in 2 ml Aquasol-2 (•). The background count of 20 dpm was subtracted. The fractions indicated by bars were collected for further purification The yields of the TC-1 and TC-2 peak materials were 31 and 42%, respectively B, C) Rechromatography of the "C-labeled peptides separated on a Unisil column as in A. The TC-1 and TC-2 fractions from the several chromatographic runs were collected separately and purified on an Inertsil ODS-2 column as described under "MATERIALS AND METHODS." The HPLC patterns of the TC-1 (B) and TC-2 fractions (C) are shown. A 20//I aliquot of each fraction was taken to count the radioactivity (•). The TC-1 and TC-2 fractions indicated by bars were used to analyze the sequence. The yield of the TC-1 peptide was 83% (B), and those of TC-2 and TC-(3 + 4) were 50 and 33%, respectively. , acetonitnle (AcCN).

A)

OTC-2

B)TC-I

100 JAuc

0.032 A

-130

£200,-

/ Hh

/

jj'lOO 0

20

Vol. 109, No. 6, 1991

40 Froctton

- - 60 No.

25g 80 No.

912

H. Suzuki and Y. Kawamura-Konishi

3, and 4 was 0.8 : 1 : 0.56 : 0.2. IAM-incorporation into Cys residues in the TC-4 sequence may not be important, since the relative UC-incorporation was low. Shiga et al. (14) reported the sequence around the covalently bound FAD to be -Asx-His(FAD)-Val-Ala-. It is therefore concluded that peptide TC-2 contains the covalent FAD, since 1) peptide TC-2 contains the sequence of Asp-Xaa-Val-Ala and a small amount of His was observed as Xaa during the sequence analysis, and 2) peptide TC-2 fluoresced under UV light. [ltC]IAM-Labeling of the Enzyme in the Presence and Absence of Acetate—The labeling of the enzyme with IAM is prevented by acetate, a competitive inhibitor with respect to the substrate, sarcosine (6, 8). To determine if there is any differences in the reactivity of the Cys residues in the sequences of the TC-1 and TC-2 peptides, we compared the labeling of these two peptides in the presence

and absence of sodium acetate. The enzyme was incubated with 1 mM [ U C] IAM for 30 min at 30'C in the presence and absence of 50 mM sodium acetate, and then the labeled subunit B was prepared as described under "MATERIALS AND METHODS." uC-Incorporation into the subunit B in the presence of acetate amounted to 55% of the control level, confirming the previous work (8). After complete digestion of the labeled subunit B with trypsin and chymotrypsin, the proteolytic digests were analyzed by HPLC. As Fig. 3 shows, 14C-incorporation into the TC-1 peptide was strongly inhibited (26% of the control level) and that into the TC-2 peptide did not change so much (82% of the control level). These data indicate that the Cys residue in the TC-1 sequence is located at the sarcosine-binding site, while the Cys residue of the TC-2 peptide is not. CNBr Cleavage of u C-Labeled Subunit B—As described above, it is highly possible that the TC-1 sequence is located at the sarcosine-binding site and the TC-2 sequence at the

A) Control

10

12

Fig. 2. Edman degradation of the TC-1 (A) and TC-2 peptides (B) indicated in Fig. 1. The amounts analyzed were 3,100 dpm for the TC-1 peptide and 790 dpm for the TC-2 peptide. The radioactivity (•) at each Edman cycle detected was measured as described under •MATERIALS AND METHODS." The results of amino acid analyses of these peptides were- TC-1 peptide: Ala (1.0), Gly (3 4), Pro (0 9), Thr (0.78), and Tyr (0.1), TC-2 peptide Asp (1 1), Glu (3.3), Ser (1.27), Gly (1.5), Ala (2.4), Val (1.0), He (0.6), Leu (1.0), Phe (0.6), and Lys (1.1). The numbers in parentheses are the relative contents of the individual amino acids.

20

40 Froctlon

80 IOO 60 No. Fig. 3 HPLC profiles of the tryptic and chymotryptic digests of the "C-labeled subunit B. The enzyme was labeled with 1 mM [HC]IAM in the presence (B) and absence (A) of 50 mM sodium acetate in buffer A. After incubation for 30 min at 30'C, the reactions were stopped by adding 20 mM mercaptoethanol and proteins were sedimented by the addition of 4 volumes of acetone. The precipitates were collected and dried. The precipitates were thus suspended in 1% NH.HCOj, and digested with trypsin and chymotrypsin as described under "MATERIALS AND METHODS.' The digests were analyzed as in Fig. 1. The eluates were collected directly in scintillation vials to count the radioactivity (•). , acetonitnle (AcCN)

TABLE I. N-Terminal amino add sequences of the various peptides obtained on proteolysis of subunit B. TCI Cys-Gly-Thr-Pro-Gly-Ala-Gly-Tyr TC-2 Ala-Gly-Ile-Ala-Cys-Xaa-Asp-Xaa-Val-AlaTC-3 Val-Asn-Cys-Gly 18

TC-4

10

54

Gly-Val-Cys-Pro-ne-Asn-Thr-Gly-Asp-Asn-ne31

21

23

19

13

23

8.6

7

75

8

T-l Thr-Pro-Ile-Gln-Asn-Uu-Tyr-Val-Asn-Cys-Gly-Xaa-Gly-Xaa-Gly-Gly-PheC-l Ak-Leu-Glu-XaaPhe-Glu-Thr-Gly-His-Leu-ue-Asp-Glu-His-Gly-Xaa-Ala-Ala-Val-Ala The numbers under the sequences indicate pmol PTH amino acids detected at each cycle of Edman degradation Cys residues were detected as PTH-[uC]carboxyamidated Cys, except that Cys in T-l was detected as PTH-['H]carboxymethylated Cys. Tryptic peptide T-l and chymotryptic peptide C-1 were obtained during the experiments on the chemical modification of subunit B J. Bwchenu

Active Site SH Groups of Sarcosine Oxidase

913

covalent FAD-binding site. As it has been proposed that the noncovalently bound FAD reacts with sarcosine (4), it would therefore be interesting to know whether the sequence around the TC-1 contains the characteristic flavin-

No.

Fig 4. HPLC profiles of the CNBr fragments of the [UC] labeled submit B. The subunit B (230 ^g) was incubated with 2% CNBr in 100 }i\ of 70% formic add as under "MATERIALS AND METHODS." After 24 h, the mixture was directly applied onto a HPLC column of TSK-ODS which had been equilibrated with 25% acetonitrile-0 1% tnfluoroacetic acid. Peptides were eluted at 1 ml/mm with a linear gradient of acetomtrile (AcCN) containing 0.1% trifluoroacetic acid, as shown ( ). The eluate was collected in 0.5 ml fractions and 10 li 1 of each fraction was taken to count the radioactivity (•) The upper figure shows the absorbance profile at 210 nm and the bar indicates the fluorescent fractions. The fractions designated as CNBr-1 and CNBr-2 were collected for the sequence analyses.

binding site of other flavoproteins, whose three-dimensional structure is known (15-17). So we determined the longer amino acid sequence around the IAM-reactive SH groups. The u C-labeled subunit B was treated with CNBr and the peptide fragments were fractionated by HPLC. As shown in Fig. 4, two radioactive peaks were mainly observed. The fractions denoted as CNBr-1 and CNBr-2 in Fig. 4 were collected separately, and then the N-terminal sequences of the materials were analyzed. Fortunately, a single PTH-amino acid was detected at each Edman cycle to cycle 26 for the CNBr-1 fragment and to cycle 65 for the CNBr-2 fragment. The N-terminal sequences of the CNBr fragments of CNBr-1 and CNBr-2, respectively, are shown in Table II. The sequence of the TC-2 peptide was found in the CNBr-1 sequence, and unknown amino acids at positions 6 and 8 of the TC-2 peptide were found to be Lys and possibly His, respectively. The TC-1, TC-3, T-l, and C-l sequences shown in Table I had common amino acid residues to the CNBr-2 sequence, as shown in Table II. Spectral Properties of the IAM-Treated Enzymes—The enzyme was treated with 1 mM IAM in the presence and absence of 50 mM sodium acetate in buffer A for 30 min at 30'C as described under "MATERIALS AND METHODS." Then each whole mixture was dialyzed against buffer A at 5'C overnight and then spectra were measured under various conditions. As shown in Fig. 5, the spectra of the oxidized forms of the IAM( —A) and IAM( + A) enzymes were the same as that of the native enzyme. But the addition of sodium sulfite and/or sarcosine changed the spectra greatly. That is, the spectrum of the IAM (—A) enzyme after the addition of sarcosine revealed the partial reduction of the bound flavins. The extent of reduction estimated from the absorbance change at 455 nm was the same irrespective of the presence or absence of 21 mM

CNBr-1

5

10



1 5 + 1 8

Cly-Ala-Thr-Tyr-Cln-Pro-Tyr-Ala-Gly-IIe-AIa-Cys-Ly3-A3P-(Hia)-Val-Ala-Thr125 150 15 65 70 65 68 68 50 49 55 50 I TC-2 Xaa Xaa 1 20 25 27 Ala-Phe-Ala-Asn-Lys-Ala-Asn-Glu-HeL 20 28 7 28 18 CNBr-2 5 10 15 • Asp-Ala-Ser-Pro-Ile-Ile-Ser-Lys-Thr-Pro-Ile-Cln-Asn-Leu-Tyr-Val-Asn-Cys300 300 10 130 115 140 10 100 13 50 50 40 40 38 35 30 35

,

JJ^

I.... 20 25 « 30 36 Gly-Trp-Gly-Thr-Gly-Gly-Phe-Phe-Cys-Gly-Thr-Pro-Gly-Ala-Cly-Tyr-Thr-Leu28 30 3 8 16 10 8 4 2 7 13 13 13 7 1.5 5 Xaa Xaa 1 I TC~ 1 1 ... I 40 45 50 54 Ala-(Glu)-Thr-I le-Ala-Xaa-Asp-GI u-l'ro-Lys-Lys-Leu-Asn-Ala-Pro-Phe-Ala-Leu8 0 . 8 5 6 8 4 4 0 . 5 1 2 3 2 . 5 1 1 2 1 I 55 60 65 Glu-Xaa-Phe-Glu-Xaa-G1y-Xaa-Xaa-1 1e-Asp-G1u1 1 1.5 0.5 1 1 0.3 £^1 Thr His-Leu H is-Gly-Xaa-Xaa-Ala-Val-Ala-|

Vol. 109, No. 6, 1991

TABLE H. N-Termlnal amino acid sequences of the CNBr fragments of the 14C-labeled subunit B. The underlined sequence 14 to 17, of the CNBr-1 fragment agreed with the reported sequence of the covalent FAD-binding site (14). The Met residue of CNBr-1 fragment was identified as PTH-homoserine The positions of the sequences of TC-1, 2, and 3, T-l, and C-l are shown under the CNBr sequences. The numbers above the CNBr sequences are the residue numbers from the N-terminal and the numbers under the sequence indicate pmol PTH ammo acids detected at each cycle of Edman degradation *, the IAM-reactive Cys residues; +, His residue of the binding site of the covalent FAD.

914

H. Suzuki and Y. Kawamura-Konishi

reduction of the enzyme by 1 mM sarcosine was followed by fast and slow phases. The absorbance change between the oxidized form and the slow phase extrapolated to zero time of reduction was estimated from the time course. Then the relative fraction of the absorbance change of the fast phase against the difference in absorbance between the oxidized and reduced forms of the native enzyme was calculated. In the presence and absence of sulfite, the fractions of the fast phase as to that of the native enzyme were estimated to be 20% for the IAM(-A) enzyme and 55% for the IAM(+A) enzyme, and the apparent rate constants of reduction in the fast phase were 7.4 and 6.3 s"1 for the enzymes treated in the presence and absence of acetate, respectively. These facts indicate that IAM modification of the SH group in the TC-2 sequence almost completely inhibited the reduction of the covalent FAD by sarcosine under the conditions employed. Moreover, it is suggested that modification of the Cys residue in the TC-1 sequence inhibited the reduction of the noncovalent FAD by sarcosine and that acetate prevented the modification. Oxidation of the Bound FADs by Oxygen—As described above, reduction of the bound FADs by sarcosine was inhibited by IAM treatment of the enzyme. So it would be interesting to know whether the FAD moieties of the enzyme can be oxidized by molecular oxygen. The IAMtreated enzymes were reduced with sarcosine and then mixed with molecular oxygen using a rapid mixing apparatus, as described under "MATERIALS AND METHODS." Figure 6C shows the time course of the absorbance change at 455 nm after mining with oxygen. Analysis of the absorbance changes showed that the oxidation followed 300 4OO 500 600 300 400 500 600 simple first-order reaction kinetics. The apparent first Wavelength, nm order rates constants for the native, IAM( —A) and IAM( + A) enzymes were 43, 22, and 25 s"1, respectively. Fig. 5. Visible absorption spectra of the native and IAMtreated eraymes. The spectrum of the enzyme (24 //M in 500^1) in Activity of the IAM(-A) and IAM(+A) Enzymes in the a Thunberg-type cuvette was measured in buffer A under anaerobic Presence of Sodium Sulfite—The results of the spectral and conditions. Then 10^1 of 1 M sodium sulfite was added, followed by stopped-flow experiments indicate that only the noncothe addition of 5//I of 25 mM sarcosine. o, the oxidized form of valent FAD functioned in the oxidation of sarcosine on the enzyme; s, after the addition of sodium sulfite; r, after the addition of IAM modification of the enzyme. That is, sarcosine reduced sarcoaine; sr, after the addition of sulfite and then sarcosine Left the noncovalent FAD and the reduced noncovalent FAD was column (A,C,E), control; right column (B,D,F), with added sulfite. A and B, native enzyme, C and D, IAM ( — A) enzyme, E and F, IAM- directly oxidized by oxygen. If this is the case, sodium ( + A) enzyme sulfite may not inhibit the oxidation of sarcosine with these treated enzymes under the conditions employed, since sulfite is known to react only with the covalent FAD (4). As Fig. 7 shows, the native enzyme was almost completely sodium sulfite (compare C with D in Fig. 5). This indicates inhibited by 21 mM sodium sulfite, but the IAM ( — A) and that only the noncovalent FAD was reduced by sarcosine, IAM( + A) enzymes were not inhibited by 21 mM sodium since sulfite was reported to inhibit the reduction of the sulfite at all. covalent FAD by sarcosine (4). For the IAM( +A) enzyme, the spectra of the enzyme reduced with sarcosine were The relative concentration of the free enzyme was similar to that of the reduced form of the native enzyme calculated by using the dissociation constant of the (compare E and F with A in Fig. 5). This indicates that the enzyme-sulfite complex. The dissociation constant of the IAM( + A) enzyme must be nearly completely reduced complex was determined from the plot shown in Fig. 7B. during the approximately 5 min when the mixture was The activity-[sulfite] profile of the native enzyme agreed processed for the spectral measurement. So to monitor the well with the [free enzyme]-[sulfite] profile, but those for early stage of the reduction by sarcosine, stopped-flow the IAM-treated enzymes did not (Fig. 7A). This indicates experiments were performed. that the covalent FAD of the IAM-treated enzymes was not involved in the oxidation of sarcosine. Reduction of the Bound FADs by Sarcosine—As described above, one of the IAM-reactive Cys residues is located at the substrate-binding site. Therefore, the reacDISCUSSION tivity of the bound flavins with sarcosine should change on the treatment of the enzyme with IAM. To examine this, we Corynebacterium sarcosine oxidase consists of 4 nonidenperformed flow experiments, using a rapid mixing apparatical subunits, and contains both one covalent and one tus under anaerobic conditions. As shown in Fig. 6, the noncovalent FAD. IAM specifically reacts with subunit B Control

• Sulflte

J Bwchem.

915

Active Site SH Grvups of Sarcosine Oxidase C) Oxidation

A) Reduction, control •—ox

IAM(-A)

IAM(-A) JA

100 msec

0 025

IAM(*A)

Reduction

native

1

red lav*

Oxidarion

B) Reduction, +sulfife IAM(-A)

f

Fig. 6. Reduction and oxidation of the bound FAD8 of the native and IAMtreated enzymes. A, B) Time course of the absorbance decrease at 455 nm upon mining of the oxidized forms of the enzymes with 1 mM sarcosine in the presence (B) and absence (A) of 20 mM sodium sulfite C) The enzymes were reduced with 250/iM sarcosine under anaerobic conditions and then rapidly mixed with 120 ^M 0,. The time courses of the absorbance increase at 455 nm for the reduced enzymes are shown. Buffer A was used. Temperature, 25'C

^A 0 025 IAM(«A)

native

Sodium sulfite, M; IAM

Fig 7. A) Effect of sodium Bulflte on the activities of the enzymes treated with LAM as B) described under "MATE• 0.15 6 RIALS AND METHODS." The activity of each enzyme C was assayed at 25'C in buffer A 100 The sarcosine concentration ZE was 1 mM Sodium sulfite was dissolved in deionized water 4 0.15 a? just before use. The enzyme activity in the absence of sulfite £ was 4.8 s"1 for the native (•), 2 3 s-' for the IAM (+A) (o), and 1.1 s"1 for the IAM (-A) D native 2 0 05 (A) enzyme The concentrations • IAM(-A) 3 of the enzyme were 2 1 x 10*7 — M for the native one, 1.5 x 10~7 O IAM(*A) M for the IAM (+A) one, and / 1.2xlO"TM for the IAM (-A) 10 Sodium sulfite, one. •• • •, , the relative 0 concentration of the free en0 005 O.I zyme, which was calculated by Sodium sulfite, M ; native using dissociation constants for the covalent FAD-sulfite complex of 1 52xlO" 3 M for the native enzyme and 6.0xl0~ J M for the IAM-treated enzymes (Fig 7B), respectively B) The dissociation constants for the enxyme-sulfite complex were determined by titrating the enzyme with sodium sulfite, the changes in the absorbance at 455 nm being measured at 25"C in buffer A. Then [sulfiteJ/zHmj was plotted against [sulfite]. The crossing point of the plot on the abscissa gives the negative value of the dissociation constant The concentrations of the native and IAM-treated enzymes were 14.8 x 10"' and 7.9 x 10"' M, respectively. 0

i

02

e

(Mr 44,000) and the inhibition of enzyme is dependent on the IAM incorporation into subunit B (S). In this work the partial amino acid sequences around the IAM-reactive Cys residues of the subunit were analyzed and the functional changes in the enzyme-bound FADs on IAM-modification were studied. The reaction of the Cys residue in the TC-1 sequence with IAM was inhibited by sodium acetate (Fig. 3), a competitive inhibitor with respect to sarcosine. So there is a possibility that the Cys residue is located at the sarcosine-binding site. The 14C-incorporation into the TC-2 peak material was slightly inhibited by acetate (Fig. 3). The Cys residue in the TC-3 sequence is located near the TC-1 sequence, so this inhibition may be ascribable to the inhibition of l4C-incorporation into the TC-3 sequence in CNBr-2 (Table II). Jorns proposed that the noncovalent FAD reacts with sarcosine and functions as a "dehydrogenase" flavin (4). It is therefore possible that this Cys Vol. 109, No 6, 1991

O.I

residue may be located at the noncovalent FAD-binding site. It is known that flavoproteins have a common sequence characteristic of the AMP-binding site, that is, Gly-XaaGly-Xaa-Xaa-Gly (15-18). As shown in Table HI, the CNBr-2 peptide contains this sequence (CNBr-2: Gly at residues 19, 21, and 24). If this sequence is located at the AMP-binding site, it should be present between the P structure and the a helix (15-17), so we estimated the secondary structure of the sequence of the CNBr-2 peptide by the method of Chou and Fasman (19), using GENETYX genetic information software (Software Development, Tokyo). Though we determined 65 residues of the CNBr-2 peptide (Table II), residue 42 could not be identified. So we estimated the secondary structure from residue 1 to 40 (Fig. 8). Unexpectedly, the predictions indicated that this Gly-Trp-Gly-Thr-Gly-Gly sequence is present between the P structures, and that the sequence around the IAMreactive Cys residue (residue 27) has a bend between the yS

916

H. Suzuki and Y. Kawamura-Konishi TABLE IV. Sequences of the covalent FAD-binding sites of flavoproteins. * His residue at the binding site of the covalent FAD. CNBr-1, CNBr fragment 1 (residues 1-26, Table I); FR, fumarate reductase (residues 31-56, Ref. 20), SDH, succinate dehydrogenase (residues 30-55, Ref 21); 6HN0, 6-hydroxy-D-nicotine oxidase (residues 56-81, Ref. 22). CNBr-1 GATYQPYAGIACKDHVATAFANKANE FR AKIALISKVYPMRSHTVAAEGGSAAV SDH QTCALLSKVFPTRSHTVSAQGGITVA 6HN0 ACDNGLEISVRSGGHNPNGYATNDVA * FAD

FR

Kinetic parameters for the IAM-treated enzymes v/e (s-) /red Calculated Observed IAM ( + A) 7.4 25 0.55 3.1 2.3 IAM (-A) 1.3 0.20 1.1 6.3 22

TABLI! V

__A/VWY FAD

C SDH

6HN0

Fig. 8 Secondary structure predictions for the CNBr fragments of subunit B and those around the covalent FAD-bindlng sites of other flavoproteins shown in Table IV. , random coil; A A , P sheet; JUL, a helix; ^ > , p turn. The arrows indicate the IAM-reactive cysteine residues. TABLE III. AMP binding site homology. A one letter code is used for amino acids. CNBr-2, CNBr fragment 2 (residues 19-46 in Table II); Gr, glutathione reductase (residues 21-45, Refs. 15 and 26); PHBH, p-hydroxybenroate hydroxylase (residues 3-27, Ref. 17). * , consensus sequence of Gly-Xaa-Gly-Xaa-Xaa-Gly; -t-, consensus sequence element proposed (18). CNBr-2 GWGTGGFFCGTPGAGYTLAETIAXDEPK * * *

GR PHBH

YDYLVIGGGSGGLASARRAAELGAR TQVAIIGAGPSGLLLGQLLHKAGID * *

*

+

+

sheet (residues 24 to 29) and a helix (residues 35 to 40), as shown in Fig. 8. The secondary structure of residues 24-40 is characteristic of the AMP-binding sites of other flavoproteins (15-17). So it is possible that the Cys residue, residue 27, of the CNBr-2 peptide (Table II) is located at the noncovalent FAD-binding site. It is conceivable that IAM-modification of this Cys residue changed the conformation of the site and reduced the reactivity of the noncovalent FAD with sarcosine. A conformational transition around the flavin moieties is supported by our previous

observation that the CD spectrum in the region of 350 to 500 run changed with IAM treatment of the enzyme (8). We also determined the sequence around the covalent FAD-binding site (CNBr-1 in Table II). We compared the sequence with the reported sequences of other flavoproteins whose His residue is covalently bound to FAD (2022). We could not find any homology among these sequences (see Table TV), so their secondary structures were estimated according to Chou and Fasman (14). The secondary structure of the CNBr-1 peptide and the sequences around the binding site (His) of the covalent FAD of other flavoproteins (listed in Table TV) were estimated. One similarity in the secondary stuctures of these flavoproteins is that the His residue occurs within 3 residues from the ft turn (Fig. 8). It is not clear at present if this fact has any relevance to the flavinylation of these flavoproteins. The native sarcosine oxidase was not completely inhibited by sodium sulfite, the extent of inhibition being 90% at 21 mM sodium sulfite (Fig. 7). Similar data were reported by Jorns (4). Using the binding constant for that of sodium sulfite to the covalent FAD, Kd = 1.52xlO- 3 M (see the legend to Fig. 7B), the inhibition of the enzyme by sulfite was explained well by assuming that the enzyme-sulfite complex had no enzymatic activity. This supports Jorns' proposal (4) that electrons are transferred from sarcosine to oxygen as follows, sarcosine-

noncovalent FAD -

• covalent FAD • oxygen.

However, on modification of the enzyme with IAM, both the reactivity of the covalent FAD with oxygen and the electron transfer from the noncovalent to the covalent FAD must be impaired, and mostly the noncovalent FAD functions in the oxidation of sarcosine. This idea is supported by the observations that 1) the fractions of the fast phase of the reduction of the bound FADs of the IAM(-A) and IAM(+ A) enzymes were the same irrespective of the presence or absence of sulfite (Fig. 6), 2) the activities of the IAM(—A) and IAM( + A) enzymes were not inhibited by sodium sulfite (Fig. 7A), and 3) the IAM-reactive Cys residue is located at 3 residues N-terminus from the His residue of the covalent FAD-binding site (Table II). Therefore we like to explain our data by assuming that on J. Biochem

917

Active Site SH Groups of Sarcosine Oxidase IAM modification of the enzyme only the noncovalent FAD functions in the oxidation of sarcosine, that is, the noncovalent FAD is reduced by sarcosine and the reduced FAD is reoxidized by oxygen. One may argue that the loss of inhibition of the IAMtreated enzyme by sulfite is due to the decreased reactivity of the enzyme with sulfite (see B, D, and F in Fig. 5). However, at 21 mM sodium sulfite, the IAM-treated enzymes should be inhibited by sulfite if the covalent FAD is involved in the enzyme reaction, since the concentration of the free enzyme was calculated to be 75% of the control level at 21 mM sulfite (see the dashed line in Fig. 7A). Moreover, the decreased reactivity of the IAM-treated enzymes with sulfite may also support the idea that the covalent FAD was not reduced by sarcosine rapidly (Figs. 5F, 6AandB). Reduction of the bound FADs by sarcosine showed fast and slow phases. The apparent rates of reduction (

Cysteine residues in the active site of Corynebacterium sarcosine oxidase.

Sarcosine oxidase from Corynebacterium sp. U-96 is inhibited by iodoacetamide (IAM) and the inhibition is prevented by the substrate analog, sodium ac...
797KB Sizes 0 Downloads 0 Views