the mitochondrial glycine cleavage system

THE JOURNAL OP Broworcr~. CHEMISTRY Vol. 255, No. 24, Issue of December 25, pp. 11664-11670. 1980 F’rinled in U.S.A.

The Mitochondrial Glycine Cleavage System PURIFICATION AND PROPERTIES OF GLYCINE DECARBOXYLASE FROM CHICKEN LIVER MITOCHONDRIA*

(Received for publication, December 26, 1979, and in revised form, June 25, 1980)

Koichi Hiraga and Goro Kikuchi From the Department of Biochemistry, Tohoku University School of Medicine, Sendai 980, Japan

Glycine decarboxylase, tentatively called P-protein and considered a constituent of the glycine cleavage system, was purified to apparent homogeneity from chicken liver mitochondria. P-protein is a homodimer having a iI& = -200,000 and consisting of identical subunits with M, = -100,000. Each subunit appears to contain an equimolar pyridoxal V-phosphate which is bound to the protein, possibly through a protonated aldimine linkage. The isoelectric point of P-protein was 7.2. P-protein could bind glycine, showing a & of 33 mu for it, and could catalyze glycine decarboxylation even though the rate of decarboxylation catalyzed by P-protein alone was extremely low. The product of glycine decarboxylation was methylamine and the K,,, for glycine was about 40 mu, which is close to the Kd for glycine. Methylamine could bind to P-protein, giv- ing a Kd value of 63 mu, and it inhibited the glycine decarboxylation. P-protein alone could also slightly catalyze the exchange of carboxyl carbon of glycine with CO2 and the exchange appeared to obey a ping- pong mechanism. Both glycine decarboxylation and the glycine-CO2 exchange catalyzed by P-protein were stimulated loo-fold or more by the addition of lipoic acid, which is a functional group of H-protein. We may define P-protein as glycine decarboxylase although P- protein alone exhibits only very low catalytic activities.

Glycine can be cleaved to carbon dioxide, ammonia, and methylene tetrahydrofolate by the glycine cleavage reaction in various animals (1,2), plants (3-5), and bacteria (6-lo), and the reaction has been demonstrated to be reversible in both the bacterial (11) and animal (12-14) systems. The glycine cleavage system, also called glycine synthase (EC 2.1.2.10), consists of four protein components which have tentatively been named P-protein (a pyridoxal phosphate-containing protein), H-protein (a lipoic acid-containing protein (15, 16) ini- tially called hydrogen carrier protein (17)), T-protein (a protein which catalyzes the tetrahydrofolate-dependent step of the reaction), and L-protein (a lipoamide dehydrogenase (18, 19)). P-protein, H-protein, L-protein, and T-protein are anal- ogous to P1, Pp, PB, and Pq from Peptococcus glycinophylus @O-23), respectiveljr. The glycine cleavage system in animals is confined to mitochondria, possibly as an enzyme complex (24) which is loosely bound to the mitochondrial inner mem- brane (25). A tentative scheme for the overall reaction of the reversible glycine cleavage has been presented (19,26,27).

* This work was supported in part by Grant 378068 from the Ministry of Education, Science and Culture, Japan. The cos& of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “c&uer. tisement” in accordance with 18 USC. Section 1734 solely to indicate this fact.

One of the most characteristic properties of the glycine cleavage reaction is that, although P-protein may belong to a class of pyridoxal phosphate-dependent amino acid a-decarboxylase, P-protein requires H-protein to significantly catalyze decarboxylation of glycine. Moreover, P-protein, when combined with H-protein, actively catalyzes the exchange of the carboxyl carbon of glycine with CO*. Also, it has been shown with the bacterial P-protein that lipoic acid could replace H-protein to some extent in both glycine decarboxylation (15) and in the glycine-COz exchange (15, 23). With the aim of further clarifying the features of the P-protein-catalyzed reactions, we attempted to purify P-protein from chicken liver mitochondria, taking advantage of the observation that the glycine cleavage activity is high in chicken liver mitochondria (2). P-protein has been purified from Arthrobacter glob- iformis (19), but the bacterial P-protein easily dissociates into pyridoxal phosphate and apoprotein and cannot be isolated in a holo form, while P-protein of animal origin stays fairly stable in a holo form. The present study provided evidence that P- protein is, by nature, glycine decarboxylase although P-protein alone exhibits only very low catalytic activities.

EXPERIMENTAL PROCEDURES

Materials-Fresh chicken livers were homogenized in 0.25 M sucrose and the liver mitochondria were prepared by the method of Schneider and Hogeboom (28) from 10% liver homogenate. Hydrox- yapatite was prepared by the method of Tiselius et al. (29). DEAE- cellulose was a product of Brown Co., Berlin, and CM-Sephadex C-25 was from Pharmacia, Uppsala. Sephadex G-200 was obtained from Sigma Chemical Co. [l-“CJGlycine, [2-‘“C]glycine, [‘4CJmethylamine, and Ba’%O, were obtained from New England Nuclear, Boston. [l- ‘%]Glycine and [2-‘%]glycine were purified by column chromatography on AG 5OW-X2 (200 to 400 mesh, Bio-Rad, Richmond) and on Dowex l-X2 (200 to 400 mesh) after being mixed with an appropriate amount of cold glycine. NaH%Os was prepared from Ba%Oa. Scin- tillamine hydroxide was a product of Wako Pure Chemical, Osaka. Ampholine and the apparatus required for isoelectric focusing were the products of LKB-Produkter AB, Bromma. Other chemicals were obtained commercially.

Assay of the Enzyme Activities-The activities of P-protein and H-protein were determined by measuring the amounts of [‘4C]bicar- bonate fmed to the carboxyl carbon of glycine during the exchange reaction. The reaction mixture contained, in a final volume of 1 ml, 50 pmol of potassium phosphate buffer (pH 6.2), IO pmol of glycine, 1 pmol of dithiothreitol, 0.1 pmol of pyridoxal phosphate, 20 pm01 of NaH”‘C03, 22.5 pg of H-protein for P-protein assay or 20 units of P- protein for H-protein assay, and an appropriate amount of the enzyme preparation to be tested; the final pHbf the reaction mixture was 6.6. The reaction was carried out in a test tube at 37°C. After a 3-min preincubation, the reaction was started with the addition of NaH’%Oa. At 15 min of incubation, an aliquot of the reaction mixture was transferred to a glass counting vial-which contained 50 ~1 of glacial acetic acid, and the mixture was dried on a hot plate. The dried sample was dissolved in 0.5 ml of distilled water and mixed with 5 ml of toluene-Triton X-100 scintillator (30) and the disintegrations per min were determined with an Aloka LSC 651 scintillation spec- trometer. One unit of P-protein and one unit of H-protein were

11664

The Mitochondrial Glycine Cleavage System 11665

defined, respectively, as the amount of P-protein and the amount of H-protein which bring about the fixation of 1 nmol of 14C02 into glycine/min under the conditions described above.

The glycine decarboxylase activity was determined by measuring “C02 formed from [l-’4C]glycine in a Warburg type flask with a double side arm. The reaction mixture contained, in a volume Of 1 ml, 20 pmol of potassium phosphate buffer (pH 7.0), 1 p o l ofdithiothre- itol, 0.1 pmol of pyridoxal phosphate, 50 pmol of [l-’4C]glycine (0.1 Ci/mol), and an appropriate amount of P-protein. The reaction was started with the addition of glycine from one side arm after a 3-min preincubation at 37°C and usually carried out for 90 min. It was terminated by adding 0.1 ml of 50% trichloroacetic acid from the other side arm and the I4CO2 formed was adsorbed into 0.25 ml of scintillamine hydroxide placed in the center well. Scintillamine hydroxide was quantitatively transferred to a glass counting vial and radioactivity was measured after addition of 10 ml of toluene scintillator.

Estimation of Molecular Weight and Examination of Subunit Structure of P-protein-The molecular weight of P-protein was determined by the sedimentation equilibrium method described by Edelstein and Schachman (31). The concentration gradient of the protein was recorded with Rayleigh interference optics equipped in a Hitachi analytical ultracentrifuge model 282, The molecular weight of P-protein was also examined by gel filtration using a column of Sephadex G-200 according to the method of Andrews (32) and by sucrose density gradient centrifugation according to the method of Martin and Ames (33). In gel filtration, catalase (195,000) (Ref. 32), lactate dehydrogenase (148,000), diaphorase (110,000), malate dehydrogenase (70,000), and myoglobin (1’7,800) were used as marker proteins; in sucrose density gradient centrifugation catalase (11.4 s, 250,000) and alcohol dehydrogenase (7.6 S, 150,000) were the reference proteins. The activities of catalase (34), alcohol dehydrogenase (35), lactate dehydrogenase (36) , diaphorase (37), and malate dehydrogenase (38) were determined spectrophotometrically according to the methods referenced, respectively.

The subunit structure of P-protein was examined by electrophoresis a t 8 mA/column (39) on 7.5% polyacrylamide gel (0.6 X 7 cm) containing 0.1% sodium dodecyl sulfate. The molecular size of the subunit of P-protein was estimated by using the following marker proteins, assuming their molecular weights as given in parentheses: ,&galactosidase (130,000), phosphorylase a (97,000) (Ref. 40), bovine serum albumin (68,000), catalase (58,000), glutamate dehydrogenase (53,000), ovalbumin (43,000), lactate dehydrogenase (36,000), and chymotrypsinogen (25,700).

Amino Acid Composition of P-protein and H-protein-Proteins were dialyzed against 2 liters of distilled water, and the external fluid was changed twice at 6-h intervals. P-protein (470 pg) and H-protein (440 pg), as estimated by the method of Lowry et aZ. (41), were lyophilized and subjected to hydrolysis with 6 N HC1 at ll0’C for 22 h. Proteins derivatized by the method of Moore (42) were also hydrolyzed in the same way as above. Individual samples were analyzed with a Hitachi amino acid analyzer, model 835. The content of tryptophan was determined spectrophotometrically (43) in 6 M gua- nidine/HCL Contents of half-cystine in the derivatized proteins were estimated by comparing the relative contents of other amino acids in the derivatized proteins and in the untreated proteins.

zsoelectrofocusing of Proteins-Isoelectrofocusing was performed using a LKB 8101 electrofocusing column (110 m l ) , according to the method of Vesterberg (44). For P-protein, a mixture of carrier am- phlytes with a specific pH range from 6 to 8 was used and for H- protein, a mixture with a pH range from 3.5 to 5 was used. After electrofocusing, individual fractions were examined directly for pH with a small electrode.

Determination of Protein Amount-For crude enzyme prepara- tions, the method of Lowry et al. (41) was used, with bovine serum albumin as standard. When the enzyme preparation contained dithiothreitol, the sample was pretreated in the following way to eliminate dithiothreitol. To the sample was added H202 (final concentration, 5 mM) and a NaOH-Na&Oa buffer (pH, -12; final concentration, 5 mM) and the mixture was heated for 2 to 3 min at 90°C in a water bath, then cooled to room temperature. Protein determination was not affected by this treatment. As for the purified P-protein and H- protein, protein amounts were determined on the basis of the results of quantitative amino acid analysis which indicated that the value of 0.1 in absorbance at 750 nm obtained by the Lowry method (41) corresponds to the concentrations of 22.4 pg/ml of P-protein and 19.6 p g / d of H-protein, respectively.

Preparation of H-protein-H-protein was purified from chicken

liver mitochondria essentially according to the method of Fujiwara et al. (E), starting usually with 600 g of fresh liver. The H-protein preparation which was finally obtained gave a single band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis and had a specific activity of 2,430 units/mg of protein in the glycine-C02 exchange when assayed under the standard reaction conditions as described above. The molecular weight of H-protein was about 14,500 when estimated by either sodium dodecyl sulfate-polyacrylamide gel electrophoresis or gel filtration with a column of Sephadex G-100. H- protein had an isoelectric point of 4.0. These results are in good agreement with those reported by Fujiwara et al. (16). Also, the amino acid composition of the purified H-protein was essentially the same as that reported by Fujiwara et al. (16).

A Hitachi recording spectrophotometer, model 232, was used for the examination of absorption spectra.

RESULTS

Purification of P-protein-A summary of a purifkation experiment to be described below is shown in Table I.

The mitochondrial preparation prepared from 1,500 g of chicken liver was suspended in 30 m~ Tris/HCl buffer (pH 7.8) containing 1 m~ EDTA to give a volume similar to that of the original liver, then homogenized once in a glass homog- enizer and centrifuged at 14,000 x g for 10 min. The supernatant, which contained about one-third of the total protein of the original mitochondrial preparation, was discarded, and the sediment was resuspended in 20 mM potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol and 1 mM pyridoxal phosphate to give a final volume of 750 ml. (All buffers used in the subsequent purification procedures contained 1 mM dithiothreitol and 0.1 mM pyridoxal phosphate unless stated otherwise.) The suspension was sonicated with a Kubota Insonator at 160 watts for 4 min, followed by centrifugation at 192,000 X g for 30 min. The supernatant obtained (crude extract) was subjected to hydroxyapatite column chromatography using six sets of the column (2.5 X 10 cm) equilibrated with the same buffer. Unadsorbed proteins were washed out with the same buffer, then proteins were eluted with a linear gradient provided with 400 ml of the same buffer and 400 ml of 500 m~ potassium phosphate buffer (pH 7.0). The activity of P-protein was eluted at about 250 mM in phosphate. The P-protein-active fractions were combined and dialyzed for 6 h against 15 volumes of 1 m~ dithiothreitol alone. Then, the fractions from six sets of the column were combined (first hydroxyapatite fraction) and applied to two sets of DEAE-cellulose column (3 x 38 cm), previously equilibrated with 20 mM potassium phosphate buffer (pH 7.0), containing 1 m~ dithiothreitol but no pyridoxal phosphate. The column was washed with 200 ml of the same buffer, then elution was performed with about 800 ml of a linear gradient of NaCl between 0 and 400 mM in the same buffer and 15-ml fractions were collected in test tubes in which 0.15 ml of 10 mM pyridoxal phosphate had been placed. The P-protein-

TABLE r Purification of P-protein

The purification was started with the mitochondria prepared from 1,500 g of fresh liver. Detailed procedures are described in text.

Fraction Total Total ac- Specific Recov- Purifica- protein tivity activity ery tion

mg units units’mg % -fold protein Crude extract 13,501 169,000 13 100 1 First hydroxyapa- 825 128,000 155 76 12

DEAE-cellulose 66 79,000 1,197 47 92 Second hydroxy- 38 6 4 , 0 0 0 1,W 38 130

CM-Sephadex 15 35,000 2,333 21 180 Sephadex G-200 8 33,000 4,125 20 317

tite

apatite

11666 The Mitochondrial Glycine Cleavage System

active fractions which emerged at about 140 mM in NaCl were combined (DEAE-cellulose fraction) and directly applied to the second hydroxyapatite column (2.5 X 10 cm) which had been equilibrated with 20 mM potassium phosphate buffer (pH 7.0). The bulk of the proteins were washed out by elution with 100 ml of 210 mM potassium phosphate buffer (pH 7.0), then P-protein was eluted with a linear gradient of 200 ml each of the same buffer and 500 mM potassium phosphate buffer (pH 7.0). The P-protein fractions which exhibited a greenish yellow color were combined and dialyzed against 20 m~ potassium phosphate buffer (pH 7.0) (second hydroxyapatite fraction).

The dialyzed fraction was adjusted to pH 6.6 with 20 m~ monobasic potassium phosphate, containing 1 mM dithiothre- it01 and 0.1 m~ pyridoxal phosphate, and was applied to a column of CM-Sephadex C-25 (3 X 38 cm) which had been washed with 20 m~ potassium phosphate buffer (pH 6.6). Unadsorbed proteins were washed out with the same buffer, then elution was performed with a linear gradient of 0 to 300 m~ NaCl in 400-ml amounts of the same buffer. The enzyme- active fraction (CM-Sephadex fraction) was concentrated with a small hydroxyapatite column (1.5 X 1.5 cm) by means of a single elution with 500 mM potassium phosphate buffer (pH 7.0), then applied to a Sephadex G-200 column (2.5 X 45 cm) equilibrated with 20 mM potassium phosphate buffer (pH 7.0). The P-protein-active individual eluate fractions after the gel fitration gave apparently the same specific activity of P- protein. The purified P-protein preparation (Sephadex G-200 fraction) could be stored at -20°C for several weeks without loss of activity.

Molecular Properties of the Purified P-protein-The pu- riiied P-protein gave a single protein band when subjected to SDS-polyacrylamide gel electrophoresis (Fig. 1) and the molecular weight of the peptide was estimated to be 100,000.

The molecular weight and the partial specific volume of P- protein were determined according to the theoretical equa- tions of sedimentation equilibrium of protein in the buffer systems composed of Hz0 and D20, respectively, as described by Edelstein and Schachman (31), where pH20 = 0.9990, pD2O = 1.0894, and k = 1.01328. The values of slope taken from Fig. 2 were 2.145 in Hz0 and 1.752 in D20. Then the molecular weight and the partial specific volume of P-protein were determined to be 200,230 and 0.719, respectively. On sucrose density gradient centrifugation, however, P-protein gave a M , = 208,000 or 10.1 S; on gel filtration, P-protein showed a value of 180,000 (data not shown).

L -

”

FIG. 1. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of P-protein. Electrophoresis was carried out using 5 pg of P-protein.

-I c FIG. 2. Sedimentation equilibrium of P-protein in DIO and

HIO solutions. P-protein (5.6 mg/ml) was diluted with 20 mM potassium phosphate buffer (pH 7.0) containing 1 rnM dithiothreitol and 0.1 rn pyridoxal phosphate in Hz0 or D20 (final, 85.7%) 80 as to give a P-protein concentration of 800 pg/ml and the samples were simultaneously centrifuged by using a six-sector cell. The rotor was maintained at 20°C and centrifuged at 13.000 rpm for 20 h. The data were plotted as the logarithm of the number of fringes against the square of the distance, r, from the axis of rotation. A, in H z 0 B, in Dz0.

I I 1 I

r n

FRACTION NUMBER FIG. 3. Isoelectric focusing of P-protein. 160 pg of P-protein

and Ampholine ranging from pH 6.0 to pH 8.0 were used. Electrofo- cusing was carried out for 30 h, then 3-ml fractions were taken and examined for P-protein activity ( 0 - 0 ) and pH (O-0).

Amino acid composition of P-protein, in terms of moles of amino acid residues/mol of subunit, was found to be as follows: Asp, 88; Thr, 43; Ser, 50; Glu, 94; Gly, 72; Ala, 78; %Cys, 23; Val, 53; Met, 28; Ile, 56; Leu, 80; Tyr, 30; Phe, 2 9 Lys, 44; His, 29; Arg, 54; Pro, 50, Trp, 4. The integral molecular weight of 905 constituent amino acid residues was 100,154. These results are consistent with the view that P-protein is a homodimer.

The isoelectric point of P-protein was 7.2 (Fig. 3). To examine the properties of the holo form of P-protein, a

purified enzyme preparation was concentrated with a small hydroxyapatite column in the same manner as described in the preceding section, and the sample was dialyzed succes- sively for 3 h against 500 ml of 20 m~ potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol and 0.1 m~ pyridoxal phosphate and for about 15 h against 1 liter of 20 m~ potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol but no pyridoxal phosphate. Some precipitates formed during the dialysis were eliminated by centrifugation. The clear supernatant obtained showed an absorption spec-


trum characteristic to the pyridoxal phosphate enzyme (Fig. 4, Curve I I ) . It gave two distinct absorptions with the absorption maxima at 335 nm and 428 nm, respectively, in addition to an absorption at 280 nm due to protein. The absorption spectrum of the holo form of P-protein was susceptible to the change in pH, but only slightly; for instance, the protein gave a little larger absorption at pH 6.0 (Curve I ) than at pH 7.8 (Curve 111). The molar extinction coefficients of the holo form of a P-protein subunit at 428 nm and 280 nm were determined to be 11,000 and 147,000 at pH 7.0, respectively, through several experiments with various concentrations of P-protein. When the holo form of P-protein was treated with an excess amount of sodium borohydride, the absorption maximum at 428 nm disappeared and, instead, a new absorption peak at 330 nm appeared (Curve IV).

The content of pyridoxal phosphate in the holo form of purified P-protein was examined, assuming that the molar extinction coefficient of pyridoxal phosphate in 0.1 N NaOH is 6,600 at 388 nm (45) and that of the phenylhydrazone of pyridoxal phosphate is 24,500 at 410 nm (46). The results are shown in Table 11. Apparently, 1 molecule of P-protein con- tains 2 molecules of pyridoxal phosphate. Since P-protein

0.03 I 1 1 1

0- 1 300 340 380 4 20 460 5 0 0

WAVELENGTH (nm) FIG. 4. Absorption spectra of P-protein under various con-

ditions. 0.2 ml of P-protein (1.1 m g / d in 20 mhi potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol) and 1.8 ml of 20 mM potassium phosphate buffer of pH 6.0, 7.0, or 8.0, containing 1 mM dithiothreitol were used. The fmal pH of the mixtures, designated by I, II, and ZII in the figure, were 6.0, 7.0, and 7.8, respectively. Curve IV represents the absorption spectrum after addition of 10 m o l of NaBK to the mixture which gave Curve 111.

TABLE I1 Pyridoxalphosphate content of the holo form of P-protein

P-protein samples were dialyzed overnight against indicated buffers containing 1 m~ dithiothreitol. The content of pyridoxal phosphate was calculated on the basis of the extinction coefficient of pyridoxal phosphate in 0.1 N NaOH being 6,600 m-’ cm” at 388 nm (45) and the extinction coefficient of the phenylhydrazone of pyridoxal phosphate being 24,500 m” cm” at 410 nm (46).

w w Pf 20 mM potassium

phosphate (pH 7.0) 2.7 6.2 2.3 4.5 1.7 (PH 6.6) 2.5 6.4 2.5 4.5 1.8 (pH 6.6) 2.4 5.3 2.6 4.1 2.1

20 m~ imidazole- 2.7 8.0 2.93 5.1 1.9 HCI (pH 6.6) PLP, pyridoxal phosphate.

appears to be a homodimer consisting of two identical subunits, each subunit may contain pyridoxal phosphate at a stoichiometric ratio of unity.

The absorption spectrum of the holo form of P-protein changed sigNficantly when glycine was added. The absorption maximum at 428 nm of P-protein shifted to 418 nm, the absorption slightly increased on addition of glycine, and a clear isosbestic point appeared at 437 nm, indicating the binding of glycine to the pyridoxal phosphate moiety of P- protein (Fig. 5). The results of a titration experiment are shown in Fig. 6. The double reciprocal plots of the data gave a straight line, and the apparent dissociation constant for glycine was estimated to be 33 m~ under the experimental conditions. These results also indicate that both subunits of P-protein may have the same glycine-binding affinity.

0.05 Glycine added (mM)

0 I I

300 340 380 4 20 460 500

WAVELENGTH (nm) FIG. 5. Binding of glycine to P-protein. The sample cuvette

contained 620 pg of P-protein in 2 ml of 20 mM potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol, and P-protein was omitted in the reference cuvette. Absorption spectra were recorded after addition of indicated concentrations (final) of glycine to both cuvettes; the spectra in the figure are those corrected for increased volume.

I 0.01 - I 1 1 I

-0 I O 20 30 40 50 GLYCINE (mM)

FIG. 6. Titration of P-protein with glycine. Sample and reference cuvettes contained 2 ml each of the holo- form of P-protein (190 pg/ml in 20 m~ potassium phosphate buffer (pH 7.0) containing 1 m~ dithiothreitol). 20- to 40-4 aliquots of glycine (0.2 M or 1 M) were added to the P-protein solution in the sample cuvette and the same volume of distilled water was added to the reference cuvette, and the difference in absorbance at 410 nm was determined. The absorbance was corrected for increased volume of the reaction mixture.

1166~1 The Mitochondrial Glycine Cleavage System

The holo form of P-protein could also bind methylamine; the absorption peak at 428 nm shifted to the shorter wavelength side and this was accompanied by a slight increase in absorption in a manner similar to that observed with glycine (data not shown). By applying the double reciprocal plot method to the degree of the spectral change as a function of the concentration of methylamine added, the dissociation constant for methylamine was estimated to be 63 mM.

Glycine Decarboxylation Catalyzed by P-protein Alone- An appreciable amount of 14C02 was obtained when relatively large amounts of P-protein were incubated with [ l-'4C]glycine. Formation of I4CO2 proceeded linearly for as long as 90 min (Fig. 7A), and the amount of 14C02 formed was proportional to the amount of P-protein used (Fig. 7B) . Also, the rate of glycine decarboxylation increased with the increased addition of [l-'*C]glycine (Fig. 8), and the K,,, for glycine was estimated to be about 40 mM. However, as can be seen from Fig. 7, the amount of I4CO2 formed was less than 4 times the amount of P-protein subunit used in terms of moles even when the reaction was carried out for 90 min with 50 mM glycine as substrate. The glycine decarboxylation was favored under acidic conditions although the effect of pH was relatively slight; the maximum activity was obtained at pH 6.0 and the activities at pH 6.5, 7.0, and 8.0 were 94, 89, and 48%, respectively, of the activity observed at pH 6.0. Glycine decarboxylation was competitively inhibited by methylamine with glycine; the reaction with 50 mM glycine was inhibited by about 40% by 50 mM methylamine.

The product of glycine decarboxylation was identified as methylamine in the following way: A mixture of equal

I I I I 1

INCUBATION TIME (mid

I I I I 1

OO L"---J 0.4 0.8 1.2 P-PROTEIN (nmol of subunit)

FIG. 7. Glycine decarboxylation catalyzed by P-protein alone. A, reaction mixtures contained, in a final volume of 1 rnl, 20 pmol of potassium phosphate buffer (pH 7.0), 1 pmol of dithiothreitol, 0.1 pmol of pyridoxal phosphate, and 55 pg (0.55 nmol of subunit) of P-protein. The reaction was carried out a t 37OC for the indicated periods. B, reaction mixtures were the same as in A except for the P- protein amounts employed (55 pg and 110 p g ) , The reaction was carried out for 60 min at 37°C.

amounts of [l-'4C]glycine and [2-'4C]glycine with the same specific radioactivity (0.075 Ci/mol) was incubated with P- protein, and the I4CO2 formed was trapped in scintillamine hydroxide. Then, the radioactive product remaining in the reaction mixture was separated from radioactive glycine and analyzed by the procedures described in the legend to Table 111. The radioactive product co-migrated with the authentic

I I I I

z 0 F 6 a ? /=

I ' I

0 20 40 60 GLYCINE (mM)

FIG. 8. Effect of glycine concentration on the rate of glycine decarboxylation catalyzed by P-protein alone. Reaction mixtures contained, in a final volume of 1 d, 20 pmo] of potassium phosphate buffer (pH 7.0), 1 pmol of dithiothreitol, 0.1 pmol of pyridoxal phosphate, 32 pg of P-protein, and indicated concentrations of [1-'4Clglycine. The reaction was carried out for 90 min a t 3 7 " ~ .

TABLE I11 Isolation of the product of glycine decarboxylation

A mixture of 35 pmol each of [1-I4C]glycine and [P-"C]glycine, both having a specific radioactivity of 0.075 Ci/mol, was incubated with 800 pg of P-protein for 90 min at 37'C. The reaction was terminated with the addition of 0.1 ml of 50% trichloroacetic acid containing 10 m~ HC1 and 200 mM methylamine, and the amount of I4CO2 formed was determined as described under "Experimental Procedures." Then, the reaction mixture was centrifuged and the supernatant obtained was treated with ether to eliminate trichloroacetic acid and neutralized with NaOH. The mixture was applied to a Dowex 1-X2 (OH- form, 200 to 400 mesh) column (0.4 X 8 cm) and the unadsorbed fraction was further processed with a AG 50W-X2 (H" form, 200 to 400 mesh) column (0.4 X 4 cm). The fraction adsorbed on AG 50W was eluted with 1 N NaOH and collected in a test tube, which contained 1 ml of 6 N HC1, then lyophilized. The residue was extracted with hot ethanol (about 60°C) to eliminate salts, and an aliquot of the ethanol extract was analyzed by ascending paper chromatography in the solvent system of the organic phase of a mixture of 1-butanol/ acetic acid/water (4:1:5). After an appropriate development at 2OoC, the paper was sprayed with a solution of 0.01 N HCl and then allowed to dry. Location of methylamine was visualized by means of ninhydrin. Then, the paper was cut into I-cm wide strips, and each strip was cut into pieces in a counting vial, followed by addition of 0.8 rnl of distilled water and 10 ml of toluene-Triton X-100 scintillator for determination of radioactivity. Blank values obtained in a control system without P- protein were subtracted from the values of the test sample. _ _ _ ~

Fraction Radioactivity of the product of glycine decarboxylation

recovered"

dPm Unadsorbed on Dowex 1-X2 (OH- form) 2948 Adsorbed on AG 50W-X2 (H' form) 2722 Ethanol extract 3103 Methylamine fraction after paper chro- 3040

matography O1 Radioactivity of 14C02 obtained in the test system was 3560 dpm.


0 0 0.05 0.1 0.15 0.2

'/GLYCINE (mM)

FIG. 9 (left). Double reciprocal plot of the rate of the glycine- C"OZ exchange reaction and the concentration of glycine. Reaction mixtures contained, in a final volume of 1 ml, 15 mM potassium phosphate buffer (pH 6.6), 1 mM dithiothreitol, 0.1 m~ pyridoxal phosphate, 120 pg of P-protein (1.2 nmol of subunit), 10 mM or 20 m~ NaH14COa, and varied concentrations of glycine. To avoid the change in pH of the reaction mixtures containing different concentrations of NaH"C03, the stock solution of NaHI4C03 (200 mM) had been neutralized to pH 6.6 with phosphoric acid (final concentration, 5 mM).

FIG. 10 (right). Stimulation of the glycine decarboxylation and the glycine-"CO% exchange by the addition of lipoate. Reaction mixtures for the glycine decarboxylation contained, in a

methylamine added as carrier on paper chromatography, showing a RF value of 0.2 in the solvent system of the organic phase of a mixture of 1-butanol/acetic acid/water (4:1:5). As shown in Table 111, the recovery of radioactivity after the paper chromatography, which is the final step of the isolation procedure, was 85% against the radioactivity of 14C02 obtained in the reaction. This would indicate the stoichiometric formation of COz and methylamine in the glycine decarboxylation reaction. A part of the radioactive product (a concentrated ethanol extract fraction described in the legend to Table 111) was also chromatographed in parallel with the authentic ['4C]methylamine on the same paper, and the dis- tribution of radioactivity of the product on the paper was found to be exactly coincided with that of the authentic ['4C]methylamine. Furthermore, a part of the product was dansylated at 60°C by a modifcation of the method described by Woods and Wang (47) and the dansylated sample was compared with that of the authentic ['4C]methylamine by thin layer chromatography on a polyamide plate, using ben- zene/acetic acid (91) as solvent (47). Both samples gave exactly the same RF value, 0.78.

Exchange of the Glycine Carboxyl Carbon with I4CO2 Catalyzed by P-protein Alone-P-Protein alone could also catalyze the exchange of glycine carboxyl carbon with C02 and, as shown in Fig. 9, a parallel set of lines was obtained when the double reciprocal plot method was applied to the data obtained with various amounts of glycine at f ied concentrations of NaHI4CO3. Apparent K , values for glycine obtained at 10 mM NaH14C03 and 20 mM NaH14C03 were 8.5 m~ and 15 mM, respectively. These values are lower than the K,,, for glycine obtained in the reaction of glycine decarboxylation. However, the rate of the exchange reaction was con- siderably lower than the rate of glycine decarboxylation (about one-fifth) when compared on the basis of the amount of P-protein used. This would suggest that the rate-limiting step of the exchange reaction may be the CO2 fmation to the aminomethyl moiety on P-protein. We incubated 50 mM methylamine, 20 m~ NaH14C03, and as much as 3300 units (800 pg as protein) of P-protein in either the presence or absence

L I POATE (mM)

final volume of 1 ml, 20 pmol of potassium phosphate buffer (pH 7.0), 1 p o l of dithiothreitol, 0.1 pmol of pyridoxal phosphate, 50 mM [l-'4C]glycine, 28 pg of P-protein, and indicated concentrations of lipoic acid. Reaction mixtures for the gly~ine-'~CO~ exchange contained, in a final volume of 1 ml, 20 pmol of potassium phosphate buffer (pH 6.2), 1 pmol of dithiothreitol, 0.1 pmol of pyridoxal phosphate, 50 pmol of glycine, 20 pmol of NaHI4CO3 (neutralized with phosphoric acid as described in the legend for Fig. 8), 28 pg of P- protein, and indicated concentrations of lipoate. Reactions were carried out for 90 min at 37°C. Enzyme activities in the glycine decarboxylation were expressed as nanomoles of "COZ formed (O-O), and those in the gly~ine-'~COZ exchange were expressed as nanomoles of [l-14C]glycine formed ( 0 - 0 ) .

of 105 pg of H-protein hoping to demonstrate the formation of glycine, but we failed to obtain positive indications. Free methylamine may be an extremely poor substrate for glycine synthesis.

Effect of Addition of Lipoic Acid on the P-Protein-catalyzed Reactions-Glycine decarboxylation catalyzed by P- protein was significantly stimulated by the addition of lipoic acid (Fig. 10). Under comparable experimental conditions, the reaction was stimulated about 100-fold by the addition of 20 m~ lipoic acid. The K,,, for lipoic acid was estimated to be 3.5 m~ by applying a double reciprocal plot to the data in Fig. 10. The K,,, for glycine, however, was not influenced by the addition of lipoic acid; a value of about 40 mM was obtained in the presence of 0.5 to 20 m~ lipoic acid. Lipoamide appeared to be equally effective as lipoic acid if the inhibition resulting from the addition of ethanol used to dissolve lipoamide was taken into account (data not shown).

Lipoic acid also stimulated the P-protein-catalyzed glycine- C02 exchange (Fig. 10); again, about 100-fold stimulation was obtained by the addition of 10 mM lipoic acid. The K,,, for lipoic acid in the exchange reaction was estimated to be about 3 m ~ , which is similar to the value obtained in the glycine decarboxylation reaction. The apparent K,,, for glycine in the exchange reaction was also not changed by the addition of lipoic acid; at any lipoic acid concentration tested, an apparent K,,, value of about 15 mM was obtained in the presence of 20 m~ NaHC03.

DISCUSSION

The present study indicated that P-protein of the chicken liver mitochondria is a homodimer which consists of two identical subunits each having pyridoxal phosphate at a stoichiometric ratio of unity. The spectral properties of the holo form of P-protein suggest that pyridoxal phosphate is bound to the apo-protein through a protonated aldimine linkage (e6 Ref. 48).

P-protein was shown to form a complex with glycine. P- protein alone could catalyze the glycine decarboxylation, yielding methylamine as the decarboxylation product, and the

11670 The Mitochondrial Glycine Cleavage System

K, for glycine in the decarboxylation reaction was close to the K d value for glycine. P-protein alone could also catalyze the exchange of the glycine carboxyl carbon with COz and the P-protein-catalyzed glycine-COz exchange appeared to obey the ping-pong mechanism of Cleland (49); a transient aminomethyl carbanion (cfi Refs. 50, 51) may be involved in the decarboxylation of glycine. At any rate, we may now define P- protein as glycine decarboxylase although P-protein alone exhibits only a very low decarboxylation activity. P-protein is unique as an amino acid a-decarboxylase in that P-protein catalyzes the exchange of the glycine carboxyl carbon with CO,. This unique property of P-protein would obviously pro- vide a basis for the reversibility of the whole process of the glycine cleavage reaction.

Lipoic acid has been shown to be able to replace H-protein in glycine decarboxylation (15) and in the gly~ine-'~COz exchange (15, 23) catalyzed by the bacterial P-proteins. More- over, when dihydrolipoic acid was employed, the whole process of glycine synthesis from formaldehyde, ammonia, and NaHC03 could proceed in the presence of only one enzyme, P-protein (15). The present study has revealed that lipoic acid also efficiently stimulates both glycine decarboxylation and the glycine-C02 exchange catalyzed by the P-protein purified from chicken liver. These observations add further support to the view that the lipoic acid moiety of H-protein functions as both an electron-pulling agent and a carrier of the aminomethyl moiety in the glycine cleavage reaction. The free lipoic acid, however, is far less effective than H-protein in stimulat- ing the P-protein-catalyzed reactions (15, 23, 52), suggesting that the protein moiety of H-protein may have a specific role. The functional association of P-protein and H-protein is the subject of the accompanying paper (52).

Acknowledgments-We thank Mr. K. Watanabe, Echo Poultry Farm, for his kind supply of chickens, Prof. K. Sat0 and Dr. T. Sato, Department of Biochemistry, Hirosaki University School of Medicine, for their kind help in ultracentrifugal analysis of P-protein, and Dr. M. Yoshida and Ms. A. Kikuchi, laboratory of Biochemistry, Faculty of Agriculture, Tohoku University, for amino acid analysis of P- protein and H-protein. We also acknowledge Mr. H. Kumagai for technical assistance and Ms. R. Torigoe for secretarial assistance.

REFERENCES 1. Richert, D. A., Amberg, R., and Wilson, M. (1962) J. Biol. Chem.

2. Yoshida, T., and Kikuchi, G. (1972) J. Biochem. (Tokyo) 72,

3. McConnell, W. B. (1964) Can. J. Biochem. 42,1293-1299 4. Cossins, E. A., and Sinha, S. K. (1966) Biochem. J. 101,542-549 5. Kisaki, T., Yoshida, N., and Imai, A, (1971) Plant Cell Physiol.

6. Sagers, R. D., and Gunsalus, I. C. (1961) J. Bacteriol. 81,541-549 7. Pitts, J. D., and Crosbie, G. W. (1962) Biochem. J. 83,35P 8. Jones, K. M., and Bridgeland, E. S. (1966) Biochem. J. 99,25P 9. Tait, G. H. (1970) Biochem. J. 118,819-830

10. O m , M., Liu, T. N., Cheung, I., Paulavicius, I., Wales, W.,

11, Kochi, H., and Kikuchi, G. (1969) Arch. Biochem. Biophys. 132,

12. Kawasaki, H., Sato, T., and Kikuchi, G. (1966) Biochem. Biophys.

13. Sato, T., Kochi, H., Motokawa, Y., Kawasaki, H., and Kikuchi, G.

237.99-103

1503-1516

12,275-288

Mehnert, D., and Blaise, D. (1977) J. Bacteriol. 129,926-933

359-369

Res. Commun. 23, 227-233

(1969) J. Bwchem. (Tokyo) 65,63-70

35. 36.

37.

38. 39.

40.

41.

42. 43. 44. 45.

46. 47.

48. 49.

50.

51.

52.

14. Motokawa, Y., and Kikuchi, G. (1969) J. Biochem. (Tokyo) 65,

15. Kochi, H., and Kikuchi, G. (1976) Arch. Biochem. Biophys. 173,

16. Fujiwara, K., Okamura, K., and Motokawa, Y. (1979) Arch. Bio-

17. Motokawa, Y., and Kikuchi, G. (1969) Arch. Biochem. Biophys.


19. Kochi, H., and Kikuchi, G. (1974) J. Biochem. (Tokyo) 76, 1113-

20. Klein, S. M., and Sagers, R. D. (1966) J. Biol. Chem. 241, 197-

21. Klein, S. M., and Sagers, R. D. (1967) J. Bwl. Chem. 242, 297-

22. Baginsky, M. L., and Huennekens, F. M. (1967) Arch. Biochem.

23. Robinson, J. R., Klein, S. M., and Sagers, R. D. (1973) J. Biol.

24. Hiraga, K., Kochi, H., Motokawa, Y., and Kikuchi, G. (1972) J.


26. Kikuchi, G. (1973) Mol. Cell. Biochem. 1,169-187 27. Motokawa, Y., and Kikuchi, G. (1974) Arch. Biochem. Biophys.

28. Schneider, W. C., and Hogeboom, G. H. (1950) J. Biol. Chem.

29. Tiselius, A., Hjerten, S., and Levin, 0 . (1956) Arch. Biochem.

30. Patterson, M. S., and Greene, R. C. (1965) Anal. Chem. 37,854-

31. Edelstein, S. J., and Schachman, H. K. (1967) J. Biol. Chem. 242,

32. Andrews, P. (1965) Biochem. J. 96,595-606 33. Martin, R. G., and Ames, B. N. (1961) J. Biol. Chem. 236, 1372-

34. Tolbert, N. E., Oeser, A., Kisaki, T., Hageman, R. H., and Ya-

71-75

71-81

chem. Biophys. 197,454-462

135,402-409

164,624-633

1127

205

300

Biophys. 120,703-711

Chem. 248,5319-5323

Biochem. (Tokyo) 72, 1285-1289

146,461-466

164,634-640

183, 123-128

Biophys. 65, 132-155

857

306-311

1379

mazaki. R. K. (1968) J. Biol. Chem. 243.5179-5184 Racker, E. (1950) J. Biol. Chem. 184,3131319 Pesce, A., McKay, R. H., Stolzenbach, F., Chan, R. D., and

Massey, V., and Veeger, C. (1961) Biochim. Biophys. Acta 48.33-

Kitto, B. (1969) Methods Enzymol. 13,106-116 Weber, K., Pringle, J. R., and Osborn, M. (1972) Methods En-

zymol. 26,3-27 Titani, K., Koide, A., Hermann, J., Ericsson, L. H., Kumar, S.,

Wade, R. D., Walsh, K. A., Neurath, H., and Fischer, E. H. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,4762-4766

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275

Moore, S. (1963) J. Biol. Chem. 238,235-237 Edelhoch, H. (1967) Biochemishy 6, 1948-1954 Vesterberg, 0. (1971) Methods Enzymol. 22,389-412 Peterson, E. A., and Sober, H. A. (1954) J. Amer. Chem. SOC. 76,

Wada, H., and Snell, E. E. (1961) J. Biol. Chem. 236,2089-2095 Woods, K. R., and Wang, K.-T. (1967) Biochim. Biophys. Acta

Metzler, D. E. (1957) J. Amer. Chem. SOC. 79, 485-490 Cleland, W. W. (1970) in The Enzymes (Boyer, P. D., ed) 3rd Ed,

Metzler, D. E., Ikawa, M., and Snell, E. E. (1954) J. Amer. Chem.

Mandeles, S., Koppelman, R., and Hanke, M. E. (1954) J. Biol.

Hiraga, K., and Kikuchi, G. (1980) J. Biol. Chem. 255, 11671-

Kaplan, N. 0. (1964) J. BWZ. Chem. 239,1753-1761

47

169-175

133,369-370

Vol. 2, pp. 1-65, Academic Press, New York

SOC. 76,648-652

Chem. 209,327-336

11676

the mitochondrial glycine cleavage system

Documents