Mechanism of the glycine cleavage reaction: retention of C-2 hydrogens of glycine on the intermediate attached to H-protein and evidence for the inability of serine hydroxymethyltransferase to catalyze the glycine decarboxylation

Glycine is converted to carbon dioxide and an intermediate attached to a lipoic acid group on H-protein in the P-protein-catalyzed partial reaction of the glycine cleavage reaction [K. Fujiwara and Y. Motokawa (1983) J. Biol. Chem. 258, 8156-8162]. The results presented in this paper indicate that the decarboxylation is not accompanied by the removal of a C-2 hydrogen atom of glycine and instead both C-2 hydrogens are transferred with the alpha carbon atom to the intermediate formed during the decarboxylation of glycine. The purified chicken liver cytosolic and mitochondrial serine hydroxymethyltransferase preparations could not catalyze the decarboxylation of glycine in the presence of either lipoic acid or H-protein. The decarboxylation activity of the serine hydroxymethyltransferase preparation purified from bovine liver by the method similar to that of L. R. Zieske and L. Davis [(1983) J. Biol. Chem. 258, 10355-10359] was completely inhibited by the antibody to P-protein, while the antibody had no effect on the activity of the phenylserine cleavage. Conversely, D-serine inhibited the activity of phenylserine cleavage but the activity of the decarboxylation of glycine was not affected by D-serine. Finally, the two activities were separated by the chromatography on hydroxylapatite. The results clearly demonstrate that serine hydroxymethyltransferase per se cannot catalyze the decarboxylation of glycine.     

The mitochondrial glycine cleavage system

Summary The glycine cleavage enzyme system is composed of four different proteins tentatively called P-protein, H-protein, T-protein and L-protein, and catalyzes the following reaction reversibly: Glycine + tetrahydrofolate + NAD⁺ 5, 10-methylene-tetrahydrofolate + NH₃ + CO₂ + NADH + H^– Glycine decarboxylase, tentatively called P-protein, is able by itself to catalyze glycine decarboxylation, yielding methylamine as product, but at an extremely low rate. P-Protein alone is also able to catalyze slightly the exchange of carboxyl carbon of glycine with CO₂. However, the rates of the P-protein-catalyzed reactions are greatly increased by the co-existence of aminomethyl carrier protein, a lipoic acid-containing enzyme tentatively called H-protein. Several lines of evidence suggest that H-protein brings about a conformational change of P-protein which may be relevant to the expression of the decarboxylase activity of P-protein and that the functional glycine decarboxylase may be an enzyme complex composed of both P-protein and H-protein. H-Protein seems to play a dual role in the glycine decarboxylation; the one as a regulatory protein of P-protein, and the other as an electron-pulling agent and concomitantly as a carrier of the aminomethyl moiety derived from glycine. The idea that H-protein functions as a modulator of P-protein was further supported by the study of a patient with nonketotic hyperglycinemia. The primary lesion in this patient appeared to consist in structural abnormality in H-protein; the H-protein purified from the liver of this patient was apparently devoid of functional lipoic acid. Nevertheless, H-protein from the patient could stimulate the P-protein-catalyzed exchange of the carboxyl carbon of glycine and CO₂, although only to a limited extent. The observed activity should be independent of the functioning of lipoic acid and would be a reflection of a conformational change in P-protein brought about by H-protein.P-Protein was inactivated when it was incubated with glycine in the presence of II-protein, and the inactivation was completely prevented when bicarbonate was further added so as to allow the glycine-CO₂ exchange to proceed. The inactivation was accompanied by a spectral change of P-protein. The inactivation of P-protein seemed to take place as a side reaction of the glycine decarboxylation and to reflect the formation of a ternary complex of P-protein, H-protein and aminomethyl moiety of glycine through a Schiff base linkage of the H-protein-bound aminomethyl moiety with the pyridoxal phosphate of P-protein.     

The mechanism of dextransucrase action. Direction of dextran biosynthesis

Appropriate combinations of purified components of the reversible glycine cleavage system of rat liver catalyze three partial reactions: (1) decarboxylation of glycine or its reverse reaction catalyzed by P- and H-protein, (2) condensation of one carbon substrate and ammonia or its reverse reaction catalyzed by T- and H-protein, and (3) oxidation and reduction of active disulfide of H-protein catalyzed by L-protein. Reactions (1) and (2) give the same product which is bound to H-protein. The protein-bound product was isolated by gel filtration and converted to glycine by incubation with P-protein and CO₂ or degraded further to one carbon unit and ammonia by incubation with T-protein and tetrahydrofolate. The data are consistent with the conclusion that the enzyme-bound product is an intermediate in the reversible glycine cleavage reaction. A scheme is presented for the reactions catalyzed by the enzyme system.     

Glycine metabolism by rat liver mitochondria. Reconstruction of the reversible glycine cleavage system with partially purified protein components

An enzyme system which catalyzes the degradation of glycine to one carbon unit, ammonia, and carbon dioxide and the synthesis of glycine from these three substances has been isolated from rat liver mitochondria. The reversible glycine cleavage system is composed of four protein components named as P-, H-, L-, and T-protein, respectively. A procedure is described for the purification of P-protein which catalyzes the decarboxylation of glycine or its reverse reaction in the presence of H-protein, and for T-protein which participates in the formation of one carbon unit and ammonia or the reverse reaction. The procedure described leads to the isolation of a nearly homogeneous form of T-protein but P-protein still is heterogeneous. The molecular weight of T-protein, estimated by molecular sieve chromatography, is 33,000. Properties of the synthesis and cleavage reactions and the exchange of carboxyl group of glycine with bicarbonate are also presented.     

The mitochondrial glycine cleavage system: differential inhibition by divalent cations of glycine synthesis and glycine decarboxylation in the glycine-CO2 exchange

The exchange of glycine carboxyl carbon with CO2 catalyzed by the combination of chicken liver glycine decarboxylase (P-protein) and aminomethyl carrier protein (H-protein) was markedly inhibited by various divalent cations, although extents of inhibition by individual metal ions varied considerably. Cu2+ and Zn2+, at 100 microM, inhibited the reaction almost completely, and the inhibitions by Co2+ and Ni2+ were also significant, while Mg2+ and Mn2+ did not appreciably affect the reaction. The inhibition by Zn2+ was competitive with both bicarbonate and H-protein and non-competitive with glycine. Of the two reactions involved in the glycine-CO2 exchange, decarboxylation of glycine yielding the H-protein-bound aminomethyl moiety was not significantly affected by 100 microM Zn2+ or Cu2+, but carboxylation of the H-protein-bound aminomethyl moiety to form glycine was strongly inhibited by either Zn2+ or Cu2+. Various degrees of inhibition of the glycine-CO2 exchange by other divalent metal ions could also be accounted for by the inhibition of the carboxylation step of the exchange reaction. The primary site of the action of divalent metal ions is likely to be not P-protein but H-protein, and the binding of metal ions with the H-protein-bound intermediate of glycine decarboxylation was assumed to account for the observed marked inhibition.     

Properties of recombinant glycine decarboxylase P- and H-protein subunits from the cyanobacterium Synechocystis sp. strain PCC 6803

The multi-enzyme complex glycine decarboxylase is important for one-carbon metabolism, essential for the photorespiratory glycolate cycle of plants, and comprises four different polypeptides, P-, H-, T-, and L-protein. We report on the production and properties of recombinant P-protein from the cyanobacterium Synechocystis and also describe features of recombinant H-protein from the same organism. The P-protein shows enzymatic activity with lipoylated H-protein and very low activity with H-apoprotein or lipoate as artificial cofactors. Its affinity towards glycine is unaffected by the presence and nature of the methyleneamine acceptor molecule. The cyanobacterial H-protein apparently forms stable dimers.     

Interaction between the lipoamide-containing H-protein and the lipoamide dehydrogenase (L-protein) of the glycine decarboxylase multienzyme system. 1. Biochemical studies.

Lipoamide dehydrogenase or dihydrolipoamide dehydrogenase (EC 1.8.1. 4) is the E3-protein component of the mitochondrial 2-oxoacid dehydrogenase multienzyme complexes. It is also the L-protein component of the glycine decarboxylase system. Although the enzymology of this enzyme has been studied exhaustively using free lipoamide as substrate, no data are available concerning the kinetic parameters of this enzyme with its physiological substrates, the dihydrolipoyl domain of the E2 component (dihydrolipoyl acyltransferase) of the 2-oxoacid dehydrogenase multienzyme complexes or the dihydrolipoyl H-protein of the mitochondrial glycine decarboxylase. In this paper, we demonstrate that Tris(2-carboxyethyl)phosphine, a specific disulfide reducing agent, allows a continuous reduction of the lipoyl group associated with the H-protein during the course of the reaction catalysed by the L-protein. This provided a valuable new tool with which to study the catalytic properties of the lipoamide dehydrogenase. The L-protein displayed a much higher affinity for the dihydrolipoyl H-protein than for free dihydrolipoamide. The oxidation of the dihydrolipoyl H-protein was not affected by the presence of structurally related analogues (apoH-protein or octanoylated H-protein). In marked contrast, these analogues strongly and competitively inhibited the decarboxylation of the glycine molecule catalysed by the P-protein component of the glycine decarboxylase system. Small unfolded proteolytic fragments of the H-protein, containing the lipoamide moiety, displayed Km values for the L-protein close to that found for the H-protein. On the other hand, these fragments were not able to promote the decarboxylation of the glycine in the presence of the P-protein. New highly hydrophilic lipoate analogues were synthesized. All of them showed Km and kcat/Km values very close to that found for the H-protein. From our results we concluded that no structural interaction is required for the L-protein to catalyse the oxidation of the dihydrolipoyl H-protein. We discuss the possibility that one function of the H-protein is to maintain a high concentration of the hydrophobic lipoate molecules in a nonmicellar state which would be accessible to the catalytic site of the lipoamide dehydrogenase.     

Reduction of the level of the glycine cleavage system in the rat liver resulting from administration of dipropylacetic acid: an experimental approach to hyperglycinemia.

Prolonged administration of dipropylacetic acid, a branched-chain fatty acid, reduced the glycine cleavage activity in the liver of rat to about one-third of the activity in the control rat. The reduction of the activity appeared to be due mainly to reduction of the level of P-protein, a pyridoxal phosphate enzyme, which is responsible for the first step of the glycine cleavage, although dipropylacetic acid was also found to inhibit the activity of P-protein in vitro in a noncompetitive but partially competitive manner with respect to glycine. The rat treated with dipropylacetic acid may provide an experimental approach for the biochemical study of hyperglycinemia which accompanies to metabolic disorders of branchedchain keto acids. In the dipropylacetic acid-treated rat, however, the glycine concentration in the serum was not appreciably elevated and this may be accounted for by the fact that the activities of both the glycine cleavage system and serine dehydratase are considerably higher in the rat liver as compared with those in other animals including human.     

The primary structure of human H-protein of the glycine cleavage system deduced by cDNA cloning

A full-length cDNA encoding the human H-protein of the glycine cleavage system has been isolated from a lambda gt11 human fetal liver cDNA library. The cDNA insert was 1091 base pairs with an open reading frame of 519 base pairs which encoded a 125-amino acid mature human H-protein with a 48-amino acid presequence. Human H-protein is 97%, 86%, and 46% identical to the bovine, chicken, and pea H-protein, respectively.     

Expression of mature bovine H-protein of the glycine cleavage system in Escherichia coli and in vitro lipoylation of the apoform.

H-protein, a component of the glycine cleavage system with lipoic acid as a prosthetic group, was expressed in Escherichia coli using a T7 RNA polymerase plasmid expression system. After induction with 25 microM isopropyl-beta-D-thiogalactopyranoside, bacteria harboring the recombinant plasmid expressed mature bovine H-protein as a soluble form at a level of about 10% of the total bacterial protein. Little of the H-protein was lipoylated in E. coli cultured without added lipoate, but when the cells were cultured in medium supplemented with 30 microM lipoate, about 10% of the recombinant protein expressed was the correctly lipoylated active form, 10% was an inactive aberrantly modified form, presumably with an octanoyl group, and the remaining 80% was the unlipoylated apoform. Each of the three forms was purified to homogeneity and shown to have the same NH2-terminal amino acid sequence as that of native bovine H-protein. The specific activity of the lipoylated form of H-protein expressed was consistent with that of H-protein purified from bovine liver. The purified recombinant apo-H-protein was lipoylated and consequently activated in vitro with lipoyl-AMP as a lipoyl donor by lipoyltransferase purified 150-fold from bovine liver mitochondria. The lipoylation was dependent on lipoyl-AMP, apo-H-protein, and lipoyltransferase. The partially purified lipoyltransferase had no lipoate-activating activity. These results provide the first evidence that in mammals two consecutive reactions are required for the attachment of lipoic acid to the acceptor protein: the activation of lipoic acid to lipoyl-AMP catalyzed by lipoate-activating enzyme and the transfer of the lipoyl group to an N epsilon-amino group of a lysine residue to apoprotein by lipoyl-AMP:N epsilon-lysine lipoyltransferase.     

Structure of P-protein of the glycine cleavage system: implications for nonketotic hyperglycinemia

The crystal structure of the P-protein of the glycine cleavage system from Thermus thermophilus HB8 has been determined. This is the first reported crystal structure of a P-protein, and it reveals that P-proteins do not involve the alpha(2)-type active dimer universally observed in the evolutionarily related pyridoxal 5'-phosphate (PLP)-dependent enzymes. Instead, novel alphabeta-type dimers associate to form an alpha(2)beta(2) tetramer, where the alpha- and beta-subunits are structurally similar and appear to have arisen by gene duplication and subsequent divergence with a loss of one active site. The binding of PLP to the apoenzyme induces large open-closed conformational changes, with residues moving up to 13.5 A. The structure of the complex formed by the holoenzyme bound to an inhibitor, (aminooxy)acetate, suggests residues that may be responsible for substrate recognition. The molecular surface around the lipoamide-binding channel shows conservation of positively charged residues, which are possibly involved in complex formation with the H-protein. These results provide insights into the molecular basis of nonketotic hyperglycinemia.     

The impaired expression of glycine decarboxylase in patients with hyperglycinemias

Glycine decarboxylase, a constituent of the glycine cleavage system, in patients with either nonketotic or ketotic hyperglycinemia (NKH and KH) was examined using an anti-chicken glycine decarboxylase antibody. Patients with NKH who have lesion in glycine decarboxylase are differentiated by its expressed level in the liver. One group is cases of the neonatal onset type who have neither activity of the enzyme nor protein reactive to the antibody. The other is a case of the late onset type who shows low but detectable activity of the enzyme and the desirable amount of the immunoreactive material. In the liver of a patient with KH not showing the appreciable activity of H-protein, ubiquitous amount of protein reactive to anti-H-protein IgG is detected and amount of glycine decarboxylase has also been lowered. It is suggested that several mechanisms may be involved in determining the expressed level of glycine decarboxylase in patients with hyperglycinemias.     

Cloning of cDNA encoding human H-protein, a constituent of the glycine cleavage system

A cDNA that encodes human H-protein, a constituent protein of the glycine cleavage system, was cloned with anti-rat H-protein antibody as a probe from a human liver cDNA library constructed with an expression vector, lambda gt11. The longest size of cDNA of the isolated clones was about 750 base long (lambda HH15B9). On the other hand, we determined the primary structure of human H-protein from the amino terminal Ser by the 12th Val, including a hexapeptide, -Glu-Lys-His-Glu-Trp-Val-. In addition to the finding that most cDNA inserts cloned hybridized with the synthetic DNA probe composed of the possible sequences for the hexapeptide, we confirmed that lambda HH15B9 encodes the partial primary structure of H-protein in an open reading frame.     

Decarboxylation of glycine by serine hydroxymethyltransferase in the presence of lipoic acid

Serine hydroxymethyltransferase and the glycine cleavage system are both present in liver mitochondria and both bind glycine to form a pyridoxal 5'-phosphate carbanionic quinoid species. Lipoic acid has been shown to have the ability to intercept the carbanionic intermediate formed from the binary complex of serine hydroxymethyltransferase and glycine and form an intermediate adduct which is ultimately processed to yield CO2 and a methylamine adduct. Kinetic studies have shown that the lipoic acid-dependent decarboxylation of glycine catalyzed by serine hydroxymethyltransferase proceeds through a sequential mechanism. This lipoic acid-dependent decarboxylation catalyzed by serine hydroxymethyltransferase is similar to the initial reaction of the glycine cleavage system and to the lipoic acid-dependent decarboxylation of glycine by the P-protein alone suggesting that both enzymes could serve in lieu of each other.     

Control of photosynthesis in barley plants with reduced activities of glycine decarboxylase

A mutant (LaPr 87/30) of barley (Hordeum vulgare L.) deficient in glycine decarboxylase (GDC; EC 2.1.2.10) was crossed with wild-type plants to generate heterozygous plants with reduced GDC activities. Plants of the F₂ generation were grown in air and analysed for reductions in GDC proteins and GDC activity. The leaves of heterozygous plants contained reduced amounts of H-protein, and when the content of H-protein was lower than 60% of the wild-type, the P-protein was also reduced. The contents of the other two proteins of the GDC complex, T-protein and L-protein were not affected. Glycine decarboxylase activities, measured as the decarboxylation of [1-¹⁴C]glycine by intact mitochondria released from protoplasts, were between 47% and 63% of the wild-type activity in heterozygous plants and between 86% and 100% in plants with normal contents of H-protein. The enzyme activity was linearly correlated with the relative content of H-protein. Plants with reduced GDC activities developed normally and did not show major pleiotropic effects. In air, the reduction in GDC activity had no effect on the leaf metabolite content or photosynthesis, but under conditions of enhanced photorespiration (low CO₂ and high light), glycine accumulated and the rates of photosynthesis decreased compared to the wild-type. The accumulation of glycine did not lead to a depletion of amino donors or to the accumulation of glyoxylate. The lower rates of photosynthesis were probably caused by an impaired recycling of carbon in the photorespiratory pathway. It is concluded that GDC has no control over CO₂ assimilation under normal growth conditions, but appreciable control by GDC becomes apparent under conditions leading to higher rates of photorespiration. Received: 24 November 1996 / Accepted: 23 January 1997