Reduction properties of nitrated naphthalenes: relationship between electrochemical reduction potential and the enzymatic reduction by microsomes or cytosol from rat liver.

The nitroreductase activities of rat liver microsomes and cytosol towards various nitrated naphthalenes (1-, 2-mononitro-, 1,3-, 1,5-, 1,7-, 1,8-dinitro-1,3,5- and 1,3,8-trinitronaphthalenes) were characterized as follows. (1) The rates of reduction of nitrated naphthalenes in either microsomal or cytosolic incubation were found to increase in the order of trinitro- > dinitro- > mono-nitronaphthalene, although, in the case of microsomal nitroreduction, trinitronaphthalenes were reduced more rapidly than in cytosol. (2) The effective cofactors, electron donors, in the nitroreduction of nitrated naphthalenes in cytosol were NADH and hypoxanthine, but not NADPH. (3) The nitrated naphthalenes with a nitro group at a beta-position appear to be more easily reduced among the various isomers. The cytosolic nitroreductase activities towards the nitrated naphthalenes were closely related to the single-electron reduction potentials measured by cyclic voltammetry and hence, there was a good relationship between the logarithm of nitroreductase activities and the electrochemical reduction potentials. In microsomes, nitroreductase activities were rather less well related to electrochemical reduction potentials.     

Nicotinamide nucleotide synthesis in regenerating rat liver

1. The concentrations and total content of the nicotinamide nucleotides were measured in the livers of rats at various times after partial hepatectomy and laparotomy (sham hepatectomy) and correlated with other events in the regeneration process. 2. The NAD content and concentration in rat liver were relatively unaffected by laparotomy, but fell to a minimum, 25 and 33% below control values respectively, 24h after partial hepatectomy. NADP content and concentration were affected similarly by both laparotomy and partial hepatectomy, falling rapidly and remaining depressed for up to 48h. 3. The effect of injecting various doses of nicotinamide on the liver DNA and NAD 18h after partial hepatectomy was studied and revealed an inverse correlation between NAD content and DNA content. 4. Injections of nicotinamide at various times after partial hepatectomy revealed that the ability to synthesize NAD from nicotinamide was impaired during the first 12h, rose to a peak at 26h and fell again by 48h after partial hepatectomy. 5. The total liver activity of NAD pyrophosphorylase (EC 2.7.7.1) remained at or slightly above the initial value for 12h after partial hepatectomy and then rose continuously until 48h after operation. The activity of NMN pyrophosphorylase (EC 2.4.2.12) showed a similar pattern of change after partial hepatectomy, but was at no time greater than 5% of the activity of NAD pyrophosphorylase. 6. The results are discussed with reference to the control of NAD synthesis in rapidly dividing tissue. It is suggested that the availability of cofactors and substrates for NAD synthesis is more important as a controlling factor than the maximum enzyme activities. It is concluded that the low concentrations of nicotinamide nucleotides in rapidly dividing tissues are the result of competition between NAD synthesis and nucleic acid synthesis for common precursor and cofactors.     

Role of sulfhydryl groups in benzoquinone-induced Ca2+ release by rat liver mitochondria

Incubation of rat liver mitochondria with benzoquinone derivatives in the presence of succinate plus rotenone has been shown to cause NAD(P)H oxidation followed by Ca2+ release. Further investigation revealed: (1)p-Benzoquinone-induced Ca2+ release was not initiated by a collapse of the mitochondrial membrane potential. However, Ca2+ release and subsequent Ca2+ cycling caused limited increased membrane permeability. (2) p-Benzoquinone-induced NAD(P)H oxidation and Ca2+ release were prevented by isocitrate, 3-hydroxybutyrate, and glutamate but not by pyruvate or 2-oxoglutarate. (3) Inhibition of pyruvate and 2-oxoglutarate dehydrogenases by p-benzoquinone was attributed to arylation of the SH groups of the cofactors, CoA and lipoic acid. Isocitrate dehydrogenase was also inhibited by p-benzoquinone, but the cofactors NAD(P)H and Mn2+ protected the enzyme. Glutamate dehydrogenase was not inhibited by p-benzoquinone. (4) Arylation of mitochondrial protein thiols by p-benzoquinone was associated with an inhibition of state 3 respiration, which was attributed to the inactivation of the phosphate translocase. In contrast, state 4 respiration, and the F1.F0-ATPase and ATP/ADP translocase activities were not inhibited. It was concluded that inhibition of mitochondrial NAD(P)H dehydrogenases by arylation of critical thiol groups will decrease the NAD(P)+-reducing capacity, and possibly lower the NAD(P)H/NAD(P)+ redox status in favor of Ca2+ release.     

Amelioration of D-galactosamine-induced acute liver injury in rats by dietary supplementation with betaine derived from sugar beet molasses

The effects of betaine supplementation on D-galactosamine-induced liver injury were examined in terms of hepatic and serum enzyme activities and of the levels of glutathione and betaine-derived intermediates. The rats induced with liver injury showed marked increases in serum enzyme activity, but those receiving dietary supplementation of 1% betaine showed enzyme activity levels similar to a control group without liver injury. Administration of betaine also increased both hepatic and serum glutathione levels, even following D-galactosamine injection. The activity of glutathione-related enzymes was markedly decreased following injection of D-galactosamine, but remained comparable to that of the control group in rats receiving 1% betaine. The concentrations of hepatic S-adenosyl methionine and cysteine showed similar trends to that observed for hepatic glutathione levels. These results indicate that 1% betaine has a hepatoprotective effect by increasing hepatic and serum glutathione levels along with glutathione-related enzyme activities in rats.     

Induction of intestinal ATP-binding cassette transporters by a phytosterol-derived liver X receptor agonist

The nuclear receptors liver X receptor (LXR) alpha and LXRbeta serve as oxysterol receptors and regulate the expression of genes involved in lipid metabolism. LXR activation induces the expression of ATP-binding cassette (ABC) transporters, such as ABCG5 and ABCG8, which inhibit intestinal absorption of cholesterol and phytosterols. Although several synthetic LXR agonists have been generated, these compounds have limited clinical application, because they cause hypertriglycemia by inducing the expression of lipogenic genes in the liver. We synthesized derivatives of phytosterols and found some of them to act as LXR agonists. Among them, YT-32 [(22E)-ergost-22-ene-1alpha,3beta-diol], which is related to ergosterol and brassicasterol, is the most potent LXR agonist. YT-32 directly bound to LXRalpha and LXRbeta and induced the interaction of LXRalpha with cofactors, such as steroid receptor coactivator-1, as effectively as the natural ligands, 22(R)-hydroxycholesterol and 24(S),25-epoxycholesterol. Although the nonsteroidal synthetic LXR agonist T0901317 induced the expression of intestinal ABC transporters and liver lipogenic genes, oral administration of YT-32 selectively activated intestinal ABC transporters in mice. Unlike T0901317 treatment, YT-32 inhibited intestinal cholesterol absorption without increasing plasma triglyceride levels. The phytosterol-derived LXR agonist YT-32 might selectively modulate intestinal cholesterol metabolism.     

Occurrence of hydroxysteroid oxidoreductases in liver of turtles

1. Hydroxysteroid oxidoreductases have been partially purified from the cytosol fraction (105,000 g supernatant) of liver from a fresh-water turtle (Podocnemis expansa) and a sea-water turtle (Chelonia mydas mydas) by precipitation with ammonium sulphate (AS, 10-80% saturation). 2. The following enzymes were detected (substrates in brackets): 3 alpha-hydroxysteroid oxidoreductase (androsterone), 3 beta-hydroxysteroid oxidoreductase (DHEA) and 17 beta-hydroxysteroid oxidoreductase (testosterone, oestradiol-17 beta). NAD as well as NADP were effective as cofactors. 3. In fresh-water turtle, highest activities of the 3 alpha-enzyme were measured in the 20% AS fraction (cofactor NAD), of the 3 beta-enzyme in the 60% AS fraction (cofactor NAD) and of the 17 beta-enzyme in the 40% AS fraction (cofactor NADP). 4. In sea-water turtle, highest activities were observed for all three enzymes in the 60% AS fraction. 5. Generally, enzyme activities were higher in sea-water turtles than in fresh-water turtles. The most active enzyme in both turtles was found to be the 3 alpha-hydroxysteroid oxidoreductase, followed by the 17 beta- and the 3 beta-hydroxysteroid oxidoreductases.     

Mutagenicity of dimethylnitrosamine to mammalian cells as determined by the use of mouse liver microsomes.

The mutagenicity of dimethylnitrosamine (DMN) to mammalian cells was investigated using a metabolic activation system. Mutation from 8-azaguanine (8AG) sensitivity to resistance in FM3A cells, a cell line derived from C3H mouse mammary carcinoma, was found only in the presence of dimethylnitrosamine, mouse liver microsomes and cofactors. The different inducibility of the mutation was shown by the use of liver microsomes from different strains of mouse.     

Formation of glycine and aminoacetone from L-threonine by rat liver mitochondria

Threonine is a precursor of glycine in the rat, but the metabolic pathway involved is unclear. To elucidate this pathway, the biosynthesis of glycine, and of aminoacetone, from L-threonine were studied in rat liver mitochondrial preparations of differing integrities. In the absence of added cofactors, intact mitochondria formed glycine and aminoacetone in approximately equal amounts from 20 mM L-threonine, but exogenous NAD+ decreased and CoA increased the ratio of glycine to aminoacetone formed. In intact and freeze-thawed mitochondria, the ratio of glycine to aminoacetone formed was markedly sensitive to the concentration of L-threonine, glycine being the major product at low L-threonine concentrations. Disruption of mitochondrial integrity by sonication (1 min) decreased the ratio of glycine to aminoacetone formed, and in 20000 X g supernatant fractions from sonicated (3 min) mitochondria, aminoacetone was the major product. The main non-nitogenous two-carbon compound detected when intact mitochondria catabolized L-threonine to glycine was acetate, which was probably derived from deacylation of acetyl-CoA. These results suggest that glycine formation from L-threonine in rat liver mitochondria occurred primarily by the coupled activities of threonine dehydrogenase and 2-amino-3-oxobutyrate CoA-ligase, the extent of coupling between the enzymes being dependent upon a close physical relationship and upon the flux through the dehydrogenase reaction. In vivo glycine synthesis would predominate, and aminoacetone would be a minor product.     

Unusual estrogen-binding liver protein in rats and the metabolism of sex steroids

A study was made of the effect of a highly purified preparation of an unusual estrogen-binding protein (UEBP) of rat liver on 3H-estradiol (3H-E2) and 3H-testosterone (3H-T) metabolism by female rat liver homogenates. The UEBP decreased the rate of metabolic conversions of 3H-E2 and 3H-T in a dose-dependent and specific manner. The effectiveness of the inhibitory action of the UEBP declined with increase of metabolic activities of homogenate enzymes. The effects of the UEBP were reduced fully or partly by the ligands specifically binding to the UEBP, estrone (E1), and estriol (E3) respectively. The latter fact was used to demonstrate the effectiveness of the endogenous male rat liver homogenate UEBP in the control of 3H-E2 metabolism. The UEBP did not exhibit any oxidoreductase activity on the use of 3H-E1, 3H-E2, 3H-E2, 3H-T and 3H-androstenedione with NAD+, NADH, NADP+, and NADPH as cofactors. These data present the first direct experimental evidence of the regulatory function of the UEBP in the time course of changes in sex steroids.     

Nonregenerative stimulation of hepatocyte proliferation in the rat: variable effects in relation to spontaneous liver growth; a possible link with metabolic induction

Three procedures were used to stimulate hepatocyte proliferation in the rat without reducing liver mass, resulting in a supplementary growth which differs from the regenerative growth observed after loss of liver mass by hepatectomy or toxic necrosis. They were: (a) the ingestion of cyproterone, a cytochrome P450 inducing drug (b) the injection of an irritant which provokes glycogenesis and synthesis of acute-phase proteins (c) the injection of albumin-bound bilirubin leading to elimination of glucuronated bilirubin in bile.
This ensuing supplementary growth was studied in the rat under several conditions of hepatic proliferation:

The highest level of stimulation occurred when the liver growth and the hepatocyte proliferation were already high. This suggests that these stimulants are not complete mitogenic stimuli and need cofactors which are present during the spontaneous growth or, alternatively, that the effect of stimulants is opposed by an inhibitory mechanism present in the adult rat.     

Release of mutagenic metabolites of benzo[a]pyrene from the perfused rat liver after inhibition of glucuronidation and sulfation by salicylamide

The role of glucuronide and sulfate conjugation in presystemic inactivation of benzo[a]pyrene (BP) metabolites was investigated with rat livers perfused with BP (12 mumol). Comparisons were made between metabolite profiles and mutagenicity of medium from perfusions with and without salicylamide, a selective inhibitor of glucuronide and sulfate conjugation. After 4 h perfusion in the presence of salicylamide, certain BP metabolites (diols, quinones, phenols, and metabolites more polar than BP-9,10-diol) were significantly increased at the expense of quinones and phenols in the glucuronide fraction. Mutagenicity of medium (detected by the Ames test, using tester strains TA98 and TA100) was low in perfusion without salicylamide. Mutagenicity detected with tester strain TA98 was significantly increased in perfusions with salicylamide. Involvement of glucuronidation in BP inactivation was also observed at the subcellular level; when cofactors of glucuronidation were added to liver homogenates along with the NADPH regenerating system in the Ames test, BP mutagenicity was markedly decreased. Both the activation of BP to mutagenic metabolites and the inactivation of BP metabolites by glucuronidation was much more pronounced with liver homogenates from 3-methylcholanthrene-treated rats than with those from phenobarbital-treated animals or untreated controls. The results suggest an important role for glucuronidation and sulfation in the inactivation and elimination of polycyclic aromatic hydrocarbons.     

Isolation and characterization of glycolic acid dehydrogenase from human liver.

Glycolic acid dehydrogenase has been purified over 800-fold from human liver by (NH4)2SO4 fractionation and column chromatography with DEAE-cellulose and hydroxyapatite. The enzyme catalyzes the direct oxidation of glycolate to oxalate without forming glyoxylate as a free intermediate. Activity is found only in the liver in the soluble fraction. The enzyme is specific for glycolate and inhibits no activity towards glycine or glyoxylate. Glyoxylate and DL-phenyllactate exhibit the enzyme. Optimum activity occurs sharply at pH 6.1 and the Michaelis constant for glycolate was 6.3.10(-5)M. Molecular oxygen does not appear to be the electron acceptor and no requirement for cofactors has been demonstrated, althoug flavin mononucleotide, ascorbate and cytochrome c stimulate activity. The isolation of this enzyme which may account for a significant part of the normal oxalate excretion in man, provides a more complete understanding of the pathways of oxalate biosynthesis and must be taken into account when considering possible methods for controlling disorders of oxalate metabolism.     

Determination of cytochrome P450 1A activities in mammalian liver microsomes by high-performance liquid chromatography with fluorescence detection

A sensitive method for the determination of cytochrome P450 (P450 or CYP) 1A activities such as ethoxyresorufin O-deethylase (EROD) and methoxyresorufin O-demethylase (MROD) in liver microsomes from human, monkey, rat and mouse by high-performance liquid chromatography with fluorescence detection is reported. The newly developed method was found to be more sensitive than previous methods using a spectrofluorimeter and fluorescence plate reader. The detection limit for resorufin (signal-to-noise ratio of 3) was 0.80 pmol/assay. Intra-day and inter-day precisions (expressed as relative standard deviation) were less than 6% for both enzyme activities. With this improved sensitivity, the kinetics of EROD and MROD activities in mammalian liver microsomes could be determined more precisely. EROD activities in human and monkey liver microsomes, and MROD activities in liver microsomes from all animal species exhibited a monophasic kinetic pattern, whereas the pattern of EROD activities in rat and mouse liver microsomes was biphasic. In addition, the method could determine the non-inducible and 3-methylcholanthrene-inducible activities of EROD and MROD in rat and mouse liver microsomes under the same assay conditions. Therefore, this method is applicable to in vivo and in vitro studies on the interaction of xenobiotic chemicals with cytochrome CYP1A isoforms in mammals.     

Appearance of magnesium guanylate cyclase activity in rat liver with sodium azide activation.

Native soluble and particulate guanylate cyclase from several rat tissues preferred Mn2+ to Mg2+ as the sole cation cofactor. Wtih 4mM cation, activities with Mg2+ were less than 25% of the activities with Mn2+. The 1 mM NaN3 markedly increased the activity of soluble and particulate preparations from rat liver. Wtih NaN3 activation guanylate cyclase activities wite similar with Mn2+ and Mg2+. Co2+ was partially effective as a cofactor in the presence of NaN3, while Ca2+ was a poor cation with or without NaN3. Activities with Ba, Cu2+, or Zn2+ were not detectable without or with 1 mM NaN3. With soluble liver enzyme both manganese and magnesium activities were dependent upon excess Mn2+ or Mg2+ at a fixed MnGTP or MgGTP concentration of 0.4 mm; apparent Km values for excess Mn2+ and Mg2+ were 0.3 and 0.24 mM, respectively. After NaN3 activation, the activity was less dependent upon free Mn2+ and retained its dependence for free Mg2+, at 0.4 mM MgGTP the apparent Km for excess Mg2+ was 0.3 mM. The activity of soluble liver guanylate cyclase assayed with Mn2+ or Mg2+ was increased with Ca2+. After NaN3 activiation, Ca2+ had no effect or was somewhat inhibitory with either Mn2+. After NaN activation, Ca2+ had no effect or was somewhat inhibitory with either Mn2+ or Mg2+. The stimulatory effect of NaN2 on Mn2+-and Mg2+-dependent guanylate cyclase activity from liver or cerebral cortex supernatant fractions required the presence of the sodium azide-activator factor. With partially purified soluble liver guanylate cyclase and azide-activator factor, the concentration (1 mjM) of NaN3 that gave half-maximal activation with Mn2+ or Mg2+ was imilar. Thus, under some conditions guanylate cyclase can effectively use Mg2+ as a sole cation cofactor.     

Androgen metabolism by hepatic and renal tissues of the fetal rhesus monkey

Liver and kidney from fetal monkeys (day 125 of gestation) were fractionated into low speed pellets, microsomal and cytosolic fractions. Liver cytosols converted as much testosterone (T) to 5 beta-androstane-3 alpha,17 beta-diol (5 beta-diol) at 0 degrees C as at 4 degrees-45 degrees C without exogenous cofactors. The principal product formed from 5 alpha-dihydrotestosterone (5 alpha-DHT) was 5 alpha-diol. A 1000-fold molar excess of radioinert 5 beta- or 5 alpha-DHT inhibited 5 beta-diol formation from [3H]T by cytosols and increased 5 beta-DHT formation. Similarly, using 5 alpha-DHT as substrate, 5 alpha-diol formation was inhibited. Microsomal and low speed pellets with added cofactors formed products which recrystallized with either etiocholanolone or androsterone from [3H]T or [3H]DHT, respectively. Little product was formed without cofactor. Whole liver homogenates produced 5 beta-reduced products from [3H]T in the presence of an NADPH generating system whereas kidney homogenates produced 5 alpha-reduced products. These data provide new information on the capacity of fetal monkey liver and kidney to metabolize androgens. The 3 alpha-reductases are cytosolic. The 5 alpha- and 5 beta-reductases are mostly in the low speed pellet but are sufficiently represented in cytosols to mediate diol formation. The 17-hydroxysteroid dehydrogenases are in the microsomal fraction. Our results suggest that 5 alpha-DHT is the active androgen in fetal liver since testosterone is metabolized to 5 beta-DHT and 5 beta-diol which are inactive androgens.     

Matrix metalloproteinase-2 and -9 in bile as a marker of liver metastasis in colorectal cancer

Matrix metallproteinases (MMP)-2 and -9 are associated with cancer invasion and metastasis. MMP-2 and MMP-9 activities have never been assayed in bile. In the present study we investigated whether MMP-2 and -9 activities in the bile could be a marker for evaluation of liver metastasis in colorectal cancer. Fifty-three patients underwent colorectal resection for histologically verified adenocarcinoma. Twenty-six patients had colorectal cancer without liver metastasis and 27 patients had metastatic liver tumor. Six patients were studied as carcinoma-free control. MMP-2 and MMP-9 activities were assayed in bile using gelatin zymography and quantitated. Active MMP-2 activity of colorectal cancer with liver metastasis group (24.1 +/- 2.5 pixel count) was significantly higher than that of colorectal cancer without liver metastasis group (11.4 +/- 1.3 pixel count) (P < 0.001) or of control group (6.4 +/- 1.0 pixel count) (P < 0.001). Active MMP-9 was not detected in bile. ProMMP-9 activity of colorectal cancer with liver metastasis group (530.3 +/- 127.5 pixel count) was significantly higher than that of colorectal cancer without liver metastasis group (213.9 +/- 33.2 pixel count) (P = 0.008). This is the first report showing that the levels of active MMP-2 and proMMP-9 in bile were significantly higher in liver metastasis of colorectal cancer than in metastasis-free colorectal cancer. The results suggest that activities of active MMP-2 and proMMP-9 in the bile may be useful markers for predicting liver metastasis in colorectal cancer.     

Characterization of L-threonine deaminase activity of guinea pig liver induced by a high protein diet

In a foregoing paper, we demonstrated that under equilibrated diet conditions, guinea pig liver L-threonine deaminase activity should be allocated to two distinct enzymes: a specific L-threonine deaminase without activity toward L-serine and a L-serine deaminase having a secondary activity toward L-threonine. In the present work, we observed that a high protidic diet caused an elevation of total threonine deaminase activity. Thus purification of guinea pig liver L-threonine deaminase was attempted, using ultracentrifugation, salt precipitation, heat treatment, ion exchange chromatography on DEAE Sephacel, Sephadex G 200 molecular sieve, 2 amino-2 methyl-1 propanol linked CH 4B Sepharose chromatography. The weak variations of the ratios of specific activities respectively toward L-threonine and L-serine observed at each stage of the purification procedure indicated that both activities are very likely supported by a single enzyme preexisting in the liver of guinea pigs fed an equilibrated diet. No isoenzyme was evidenced by polyacrylamide gel electrophoresis or DEAE Sephacel chromatography. Moreover, our purification procedure demonstrated that not only inducible L-threonine deaminase guinea pig liver activity was due to L-serine deaminase, but also that an initially existing specific L-threonine deaminase activity paradoxically disappeared with a protein rich diet.     

The FLS mouse: a new inbred strain with spontaneous fatty liver.

BACKGROUND AND PURPOSE: A new strain of mouse, named FLS (fatty liver Shionogi), which develops spontaneous fatty liver without obesity, was established by inbreeding. Morphologic, physiologic, and genetic characterization of the strain was done. METHODS: Characteristics of male FLS mice were compared with those of the sister strain, dd Shionogi (DS), which does not develop spontaneous fatty liver. A genetic cross experiment was performed by mating FLS with C3H/He/Shi mice. RESULTS: The hepatocytes of neonatal FLS mice contained fine lipid droplets throughout the lobules, and large lipid droplets appeared as mice aged. Liver triglyceride concentrations of FLS mice were fivefold higher than those of DS mice, but serum lipid concentrations and the lipoprotein profile did not indicate abnormalities. Higher plasma aspartate transaminase and alanine transaminase activities in FLS, compared with DS mice, suggested hepatocellular lesions. The genetic cross experiment suggested that the fatty liver formation is a complex polygenic trait. CONCLUSION: The FLS mice develop a progressive hepatic steatosis without obesity and diabetes. The FLS mouse might be a good model for investigating hepatic disorders accompanied by fatty liver unrelated to alcoholism or obesity.     

Identification of inducible calmodulin-dependent nitric oxide synthase in the liver of rats.

A calmodulin-dependent nitric oxide synthase was significantly induced in the liver of rats treated intravenously with heat-killed Propionibacterium acnes and 5 days later with Escherichia coli lipopolysaccharide. The apparent calmodulin-dependent and -independent isozymes were separated by Mono Q column chromatography after their partial purification by 2',5'-ADP-agarose affinity chromatography. Both enzymes had a molecular weight of 125,000 as determined by SDS-polyacrylamide gel electrophoresis and required NADPH, tetrahydrobiopterin, and dithiothreitol as cofactors. Their activities were completely inhibited by the specific nitric oxide synthase inhibitors NG-monomethyl-L-arginine and N omega-nitro-L-arginine at 80 and 800 microM, respectively. The peptide maps of these two isozymes with lysylendopeptidase and their reverse-phase column chromatographic profiles were indistinguishable. In the presence of bovine calmodulin, the purified calmodulin-dependent isozyme behaved as a calmodulin-independent isozyme on Mono Q column chromatography. The purified calmodulin-independent isozyme was converted to a calmodulin-dependent isozyme by EDTA and EGTA. Calmodulin blot analysis using 125I-calmodulin showed that the two isozymes bound calmodulin equally efficiently.     

Formation of glycine and aminoacetone from l-threonine by rat liver mitochondria

Threonine is a precursor of glycine in the rat, but the metabolic pathway involved is unclear. To elucidate this pathway, the biosynthesis of glycine, and of aminoacetone, from l-threonine were studied in rat liver mitochondrial preparations of differing integrities. In the absence of added cofactors, intact mitochondria formed glycine and aminoacetone in approximately equal amounts from 20 mM l-threonine, but exogenous NAD⁺ decreased and CoA increased the ratio of glycine to aminoacetone formed. In intact and freeze-thawed mitochondria, the ratio of glycine to aminoacetone formed was markedly sensitive to the concentration of l-threonine, glycine being the major product at low l-threonine concentrations. Disruption of mitochondrial integrity by sonication (1 min) decreased the ratio of glycine to aminoacetone formed, and in 20 000 × g supernatant fractions from sonicated (3 min) mitochondria, aminoacetone was the major product. The main non-nitogenous tow-carbon compound detected when intact mitochondria catabolized l-threonine to glycine was acetate, which was probably derived from deacylation of acetyl-CoA. These results suggest that glycine formation from l-threonine in rat liver mitochondria occured primarily by the coupled activities of threonine dehydrogenase and 2-amino-3-oxobutyrate CoA-ligase, the extent of coupling between the enzymes being dependent upon a close physical relationship and upon the flux through the dehydrogenase reaction. In vivo glycine synthesis would predominate, and aminoacetone would be a minor product.