The effect of sulfasalazine on bovine endothelial cell proliferation and cell cycle phase distribution. Comparison with olsalazine, 5-aminosalicylic acid, and sulfapyridine.

Sulfasalazine is used in the treatment of chronic inflammatory states, for example, in inflammatory bowel disease and to a lesser degree in rheumatoid arthritis. In chronic inflammation, the formation of new blood vessels may play a key role in maintaining the inflammatory state. This process is dependent on the activation and proliferation of the endothelial cells. To investigate the possible role of sulfasalazine and its metabolites, sulfapyridine and 5-aminosalicylic acid, we examined the effect of these drugs on vascular endothelial cell proliferation in vitro. Cultures of bovine aortic endothelial cells were incubated with sulfasalazine and its metabolites. At 24 hours of incubation, sulfasalazine inhibited tritiated thymidine incorporation and cell proliferation and had already slowed S-phase progression at a concentration greater than 0.125 mmol/L. After 3 hours of incubation, sulfasalazine inhibition of tritiated thymidine incorporation into the DNA of endothelial cells was observed. This inhibition was completely reversible 24 hours after the drug was removed. One of the possible mechanisms for the inhibition of endothelial cell proliferation is interference with the de novo synthesis of thymidine that depends on folate-dependent enzymes. The effect of deoxyuridine and tetrahydrofolate on tritiated thymidine incorporation into cellular DNA, as well as release of tritium to water by [5-3H]-labeled deoxyuridine on methylation to thymidine, were used as probes for the de novo synthesis of thymidine. Deoxyuridine and tetrahydrofolate, when added to cells either individually or together for 3 hours, suppressed incorporation of tritiated thymidine into DNA through an increase in de novo thymidine synthesis. Sulfasalazine, but not its metabolites, reduced this suppression.2+ culture is inhibited by sulfasalazine and olsalazine but not by their metabolites. This inhibition appears to depend partly on the reduction of de novo synthesis of thymidine that is folate dependent.

The effect of essential fatty acid deficiency on eicosanoid production in the inflamed rat colonic mucosa.

Eicosanoids are potent mediators of inflammation and are synthesized in increased quantity in active ulcerative colitis. To elucidate the role of prostaglandin E2, thromboxane A2, prostaglandin I2, and leukotriene B2 in acute chemical colitis induced by 4% acetic acid, we utilized an animal model which has a deficiency of arachidonic acid, the precursor of eicosanoids due to an essential fatty acid deficient diet. Forty-eight hours after colitis was induced, mucosal synthesis of the cyclooxygenase products, prostaglandin E2, thromboxane A2, and prostaglandin I2, was significantly decreased in essential fatty acid deficient rats compared to normal controls. However, the 5-lipoxygenase product, leukotriene B4, was not different between groups. The decrease in cyclooxygenase products did not correlate with any change in the severity of colonic inflammation as assessed by gross morphology, histology, or myleoperoxidase activity. Thus inhibition of formation of the cyclooxygenase products of arachidonate metabolism does not appear to improve the degree of inflammation under the experimental conditions employed in this study.

Inhibition of hepatic triglyceride secretion and exogenous triglyceride clearance in the cholestatic rat.

Extrahepatic biliary obstruction in humans and rats leads to hypertriglyceridemia. The observed hypertriglyceridemia could result from either a defect of plasma triglyceride (TG) catabolism or hepatic over-production of TG. To examine these questions we have used the rat model to determine hepatic TG secretion by the Triton WR-1339 methodology (inhibition of peripheral lipolysis) and exogenous TG clearance (after i.v. injection of Intralipid). Four groups of rats were studied: group OB--48 h post-operative--bile-duct obstructed; group DV--bile diverted; group SC--sham-operated controls; and group FC--48 h fasted, unoperated controls. The hepatic TG secretion rate for group OB rats was a factor of 7 lower than that of either group SC or FC, and 5 times lower than that for group DV. There were no differences between the hepatic TG secretion rates of groups DV and FC or SC. After i.v. injection of Intralipid, plasma TG decreased with first-order kinetics. The rate constant was taken as the exogenous TG clearance rate (ETGCR). Mean ETGCR for group OB was a factor of 3 lower than that for either control group; while the ETGCR for group DV was equivalent to the control groups. Thus biliary diversion does not affect hepatic TG secretion or the ETGCR. The apparent cause of the hypertriglyceridemia of cholestasis in the bile-obstructed rat is impaired plasma TG catabolism.

Histamine H2-receptor antagonists and high-density lipoproteins.

Epidemiologic evidence has shown that elevated levels of high-density lipoproteins (HDL) protect against the development of coronary heart disease (CHD). These observations have prompted the evaluation of various factors thought to affect HDL and its subspecies, HDL2 and HDL3. Numerous behavioral and physiologic factors have been shown to elevate HDL levels. These are currently being researched as potential tools in preventing CHD. Several pharmacologic agents are known to alter HDL levels. Studies show that patients with peptic ulcer disease treated with the H2-receptor antagonist cimetidine show significant elevations in their HDL, HDL2, and HDL3 profiles. In contrast, ranitidine has no effect, or may even decrease HDL levels. These divergent effects may be related to differences in pharmacologic activity unrelated to H2-receptor blockade. It should be noted that many of the variables affecting HDL levels were not controlled in these studies, and definite conclusions should not be extrapolated to the general population at risk for CHD. Currently, well-controlled trials to study the effect of cimetidine on HDL levels are in progress.

Correctable plasma lipoprotein abnormalities in infants with choledochal cysts.

Plasma lipoproteins from two female patients--patient A, 4 wk and patient B, 19 months--were examined prior to and at 1 and 5 wk after surgical correction of biliary obstruction due to choledochal cyst. The findings were correlated with standard indices of hepatic function, namely SGPT, GGTP, 5'nucleotidase, serum bile salts, and total and conjugated bilirubin. Prior to surgery in both patients plasma cholesterol, phospholipid, and triglyceride were elevated; cholesterol esters were low; high-density lipoprotein (HDL) cholesterol and apolipoprotein A-I, the major protein constituent of HDL, were subnormal; most of the plasma lipids were contained in the low-density lipoprotein density region; lipoprotein-X was present. Patient B had had a relatively brief obstruction and suffered little secondary hepatic injury. One week after surgery, plasma lipid concentrations returned to normal; apolipoprotein A-I increased in the HDL density region and a concomitant rise in cholesterol esters to near normal, 65%, was observed; plasma lipids were contained predominantly in HDL; hepatic function improved markedly. Patient A had had intrauterine obstruction and suffered major hepatic injury with cirrhosis. One week after surgery, plasma lipid concentrations, cholesterol esters, low-density lipoprotein lipid predominance, and hepatic function remained essentially unchanged. Five weeks after surgery, the lipoprotein levels and composition and hepatic function were near normal. In conclusion, children with biliary obstruction have lipoprotein abnormalities similar to those seen in adult patients. These alterations are rapidly reversible with surgical relief and may be used as prognostic indicators of outcome.

Plasma lipoprotein changes during abstinence in chronic alcoholics.

Plasma lipoprotein changes during a 2-week period of abstinence were followed in 6 male, chronic alcoholics without evidence of severe liver disease and with initial HDL cholesterol (HDL-CH) greater than 60 mg/dl. Fasting blood samples were obtained on days 1, 3, 7 and 14 of the study. Reference data were obtained from healthy non-alcoholic normolipidemic men who had abstained from alcohol for 10 days. At day 1, plasma- and HDL-CH (by the heparin-Mn2+ technique) were 17 and 69% higher, respectively, than those of controls. During abstinence, VLDL-CH increased 52% whereas HDL-CH decreased 30% compared to day 1 values. (IDL + LDL)-CH increased during abstinence to levels 33% higher than that of controls. Plasma, SF greater than 400, VLDL- and d greater than 1.006 g/ml-triglycerides (TG) were not significantly different from those of the control group. Only the d greater than 1.006 g/ml-TG showed a significant effect of abstinence, increasing by 37%. Initial plasma and d greater than 1.006 g/ml-phospholipid (PL) concentrations were 17 and 31% higher, respectively, than those of controls; and the latter was the only fraction to change significantly with abstinence, decreasing by 13%. Density gradient ultracentrifugation was employed to further resolve the d greater than 1.006 g/ml fraction into IDL, LDL, HDL and VHDL subfractions. Initial levels of IDL- and LDL-CH and -PL in the alcoholic group did not differ from those of controls. IDL-CH and -PL were invariant during abstinence, whereas LDL-CH and -PL levels increased 38 and 28%, during this period. Apo A-I, CH and PL contained in the 'lighter' density HDL region of the gradient (d = 1.063-1.125 g/ml) were 70-108% increased over the corresponding parameters for controls; and with abstinence decreased 30-40% between days 1 and 14. CH and PL in the 'heavier' density HDL region (d = 1.125-1.21 g/ml) were 30% increased over those of controls, while apo A-I levels were similar to controls. During abstinence, 'heavy' HDL-CH, -PL and -apo A-I decreased 25% compared to day 1 values. Inspection of 'difference' density gradient plots indicated that the dominant decrease in HDL occurred in the 'light' HDL subspecies in only one-half of the subjects. Other investigators have reported a preferential decrease in 'light' HDL subspecies during abstinence in chronic alcoholics, and have related the decrease to decreased TG turnover. We conclude that additional mechanisms related to the synthesis and/or catabolism of HDL may need to be considered in explaining decreases in HDL subspecies in alcoholics during abstinence.

Intestinal lipoproteins in the rat with D-(+)-galactosamine hepatitis.

D-(+)-galactosamine (GalN) induces severe reversible hepatocellular injury in the rat accompanied by lecithin: cholesterol acyltransferase (LCAT) deficiency, defective chylomicron (CM) catabolism, and accumulation of abnormal plasma lipoproteins (Lps), including discoidal high density lipoproteins (HDL). These abnormalities are presumed to result from hepatic injury alone, but the effect of GalN on intestinal Lps has not been studied. To assess possible effects on intestinal Lp formation and secretion, mesenteric lymph fistula rats were injected with GalN or saline. Twenty-four hours later a 2-hr fasting lymph sample was collected; this was followed by an 8-hr duodenal infusion of a lipid emulsion containing 17.7 mM [3H]triolein at 3 ml/hr. Fasting lymph and fat-infused lymph flow rates, 3H, triglyceride, and cholesterol output, residual 3H in intestinal lumen and mucosa, total 3H recovery, and d less than 1.006 g/ml Lp size and lipid composition were unchanged by GalN treatment, but d less than 1.006 g/ml Lps were depleted of apoE and C. Fat-infused lymph phospholipid (PL) output was higher in GalN rats due to PL-enriched d greater than 1.006 g/ml Lps. Electron microscopy of lymph and plasma LDL and HDL revealed spherical Lps in all samples. GalN plasma, fasting lymph, and fat-infused lymph also contained large abnormal LDL and discoidal HDL. Control lymph LDL and HDL did not differ in size from control plasma LDL and HDL. Control lymph LDL contained both apoB240K and B335K. However, spherical LDL and discoidal HDL in fasting lymph from GalN rats differed significantly in size from the corresponding plasma particles and became closer in size to the plasma particles with fat infusion. GalN lymph LDL contained only apoB240K and had a lower PL/CE than GalN plasma LDL. GalN fasting lymph HDL, depleted of apoC and having a PL/CE of 5, became enriched in apoE and the PL/CE increased to 10 with fat infusion to closely resemble GalN plasma HDL. GalN reduces apoE and C (mainly of hepatic origin) in d less than 1.006 g/ml gut Lps, which may contribute to the CM catabolic defect in GalN rats. Lymph LDL and HDL, especially in fasting lymph, may be partially gut-derived with increased filtration of plasma Lps into lymph with fat infusion. GalN fat-infused lymph HDL is enriched in apoE, but unable to transfer apoE to d less than 1.006 g/ml intestinal Lps. We conclude that GalN hepatitis is a model that allows study of intestinal Lps with normal lipid digestion and absorption in the face of severe hepatic injury and LCAT deficiency.

Alterations in lipoprotein composition associated with galactosamine-induced rat liver injury.

The apoprotein and lipid composition and the morphology of lipoproteins was determined in rats with D-(+)-galactosamine (GalN) hepatitis. Single intraperitoneal injections of GalN at several dose levels and postinjection exsanguination times resulted in depressed levels of cholesteryl esters, an index of plasma lecithin:cholesterol acyltransferase (LCAT) activity, and increased levels of phospholipids, unesterified cholesterol, and triglycerides. Plasma withdrawn from rats 24 hr after injection of 1000 mg/kg GalN was most deficient in cholesteryl ester and was studied further by sequential isolation of VLDL, LDL, HDL1, HDL2, and HDL3. The increased plasma triglyceride (TG) after GalN treatment accumulated in TG-rich VLDL which contained two types of particles: a large (mean diameter 193.6 +/- 48.3 nm) and rough-edged particle, and a smooth one with a mean diameter (63.4 +/- 13.2 nm) similar to control VLDL (69.4 +/- 20.2 nm). The increased phospholiThe increased plasma triglyceride (TG) after GalN treatment accumulated in TG-rich VLDL which contained two types of particles: a large (mean diameter 193.6 +/- 48.3 nm) and rough-edged particle, and a smooth one with a mean diameter (63.4 +/- 13.2 nm) similar to control VLDL (69.4 +/- 20.2 nm). The increased phospholiThe increased plasma triglyceride (TG) after GalN treatment accumulated in TG-rich VLDL which contained two types of particles: a large (mean diameter 193.6 +/- 48.3 nm) and rough-edged particle, and a smooth one with a mean diameter (63.4 +/- 13.2 nm) similar to control VLDL (69.4 +/- 20.2 nm). The increased phospholipids and unesterified cholesterol were predominantly in LDL, HDL1, and HDL2 which were largely rouleaux of flattened vesicles. Density gradient ultracentrifugation of d greater than 1.006 g/ml lipoproteins confirmed these results. GalN hepatitis appeared to decrease the larger apoB335K subspecies and the apoC-III0 and apoC-III2 content of VLDL. However, total apoB concentration as GalN VLDL was increased 2.6-fold over control. LDL and HDL were markedly enriched in apoE. LDL apoB concentration was decreased by 41% while HDL was deficient in apoA-I, A-II and A-IV, and C. These results demonstrate association of increased plasma triglycerides with particles of grossly abnormal apoprotein composition, and the association of increased plasma phospholipids and unesterified cholesterol with apoE-rich lipoproteins during the LCAT defect produced by GalN hepatitis. These abnormal lipoproteins may represent an abnormal level of normal LCAT substrates important in the transport and esterification of plasma cholesterol.

Plasma lipoprotein composition in alcoholic hepatitis: accumulation of apolipoprotein E-rich high density lipoprotein and preferential reappearance of "light'-HDL during partial recovery.

Abnormal lipoproteins accumulate in the plasma of alcoholic hepatitis patients in association with a deficiency of the cholesterol esterifying enzyme, lecithin:cholesterol acyl-transferase. Most of these abnormal lipoproteins are found in the d > 1.006 g/ml density fraction. To investigate the composition and morphology of the lipoproteins at various times during the illness in four patients, we have employed density gradient ultracentrifugation combined with analyses of gradient fractions by polyacrylamide gel electrophoresis, electroimmunoassay, and electron microscopy. At the onset of the illness, plasma cholesteryl esters ranged from 19-34% of total cholesterol; high density lipoprotein (HDL) cholesterol and apoA-I, the major HDL apoprotein, were <10% of normal; and most of the d > 1.006 g/ml triglycerides and phospholipids were found in the LDL density region. A linear correlation (r = 0.964, P < 0.001) was found between the d > 1.006 g/ml apoB concentration and the summation of the triglyceride and esterified cholesterol for that fraction, indicating a constant ratio of apoB to the summation of these two "core lipids". ApoA-I was primarily found in the fraction d > 1.18 g/ml (HDL(3) and VHDL) but not at all in the HDL(2) density range of the gradient. No cholesteryl esters were present in the apoA-I containing fractions. In contrast to normal, large amounts of apoE accumulated in lipoproteins isolated at d 1.055-1.114 g/ml. The apoE-rich fractions contained primarily phospholipids and unesterified cholesterol; they appeared by electron microscopy to be mixtures of spherical particles, vesicular particles, and chains of bilamellar discs. Analyses of the density gradient fractions by SDS polyacrylamide gel electrophoresis under reducing conditions indicated that apoA-II levels and distribution paralleled apoA-I, not apoE, providing evidence against appreciable concentrations of apoE-apoA-II complexes. During partial recovery from alcoholic hepatitis in three patients, the d > 1.006 g/ml unesterified cholesterol and triglyceride levels decreased, while esterified cholesterol, HDL-cholesterol, and apoA-I levels increased. The first HDL fractions to reappear were lipoproteins with HDL(2) density characteristics, as evidenced by simultaneous increases of apoA-I, apoA-II, cholesteryl esters and phospholipids. Lipoproteins with HDL(3) density characteristics appeared later. Long-term (6-10 months) follow-up studies indicated a substantial elevation of HDL cholesterol and apoA-I in three of the four patients that appeared to have resulted from further increases in their HDL(2)-like subspecies. The above results illustrate the diversity of abnormal lipoproteins in alcoholic hepatitis and the ability of density gradient ultra-centrifugation combined with lipid and apolipoprotein quantitation, electron microscopy, and polyacrylamide gel electrophoresis to partially resolve those lipoproteins in the d > 1.006 g/ml plasma fraction.-Weidman, S. W., J. B. Ragland, and S. M. Sabesin. Plasma lipoprotein composition in alcoholic hepatitis: accumulation of apolipoprotein E-rich high density lipoprotein and preferential reappearance of "light"-HDL during partial recovery.

Effects of insulin on plasma lipoproteins in diabetic ketoacidosis: evidence for a change in high density lipoprotein composition during treatment.

To determine the acute effects of insulin on lipoprotein metabolism, we have followed the plasma lipoprotein lipid and apolipoprotein levels during insulin therapy for the first 24 hr in 13 patients with diabetic ketoacidosis. Corrections were made for plasma volume changes during treatment. Before insulin treatment, mean plasma triglyceride and cholesterol levels were 574 mg/dl (range 53-2355) and 212 mg/dl (range 118-416), respectively. Insulin therapy resulted in rapid decreases in triglyceride-rich lipoproteins, chylomicrons, and very low density lipoproteins (VLDL), with most patients achieving plasma triglyceride levels below 150 mg/dl at 24 hr. Mean basal levels of intermediate density lipoproteins (IDL) and low density lipoproteins (LDL)-cholesterol were low (9.9 and 72 mg/dl, respectively) and were statistically invariant with therapy. Mean basal levels of high density lipoprotein (HDL) cholesterol were also low (26 mg/dl, range 5-48) and were invariant during the first 12 hr and increased significantly to 29 mg/dl by the 24th hr. Plasma apoprotein (apo) B levels were in the upper normal range (101 mg/dl) before treatment and decreased with therapy due to significant decreases in VLDL, but not IDL or LDL apoB. VLDL appeared to have a normal apoprotein composition which did not change with treatment. Mean apoA-I levels which were near normal in plasma and HDL before therapy, decreased significantly (16%) by 12 hr and subsequently increased towards basal levels between 12 and 24 hr. The ratio of apoA-I to cholesterol in HDL also fell significantly during the entire 24 hr. Density gradient ultracentrifugal analysis of the d > 1.006 g/ml fractions indicated a selective decrease in "lighter" density fractions of HDL-apoA-I during treatment. These results provide evidence that insulin may decrease the secretion of apoA-I into plasma or increase catabolism.-Weidman, S. W., J. B. Ragland, J. N. Fisher, Jr., A. E. Kitabchi, and S. M. Sabesin. Effects of insulin on plasma lipoproteins in diabetic ketoacidosis: evidence for a change in high density lipoprotein composition composition during treatment.

Normalization of triglycerides in type IV hyperlipoproteinemia fails to correct low levels of high-density-lipoprotein cholesterol.

High-density-lipoprotein (HDL) cholesterol protects against coronary heart disease, and ways to raise low HDL values are being sought. Cross-sectional population surveys have shown that HDL cholesterol is inversely related to plasma triglycerides, yet to our knowledge no longitudinal studies have shown that a decrease in elevated triglycerides will raise depressed HDL levels. We therefore used dietary therapy to lower the triglyceride levels of 29 men with Type IV hyperlipoproteinemia and evaluated the effects on HDL-cholesterol levels. Despite a reduction in triglyceride levels from 697 +/- 90 to 333 +/- 37 mg per deciliter (P < 0.01), initially low HDL-cholesterol values did not change (29 +/- 1 to 30 +/- 1 mg per deciliter). Even in a subgroup of 12 men whose triglyceride levels fell to normal (from 670 +/- 99 to 170 +/- 7, P < 0.01) and whose weight and triglycerides remained stable for two years, HDL cholesterol remained unchanged (29 +/- 1 vs. 32 +/- 1). The persistently low HDL-cholesterol level in the presence of normalization of triglycerides suggests that depressed HDL cholesterol may be an independent metabolic abnormality in Type IV hyperlipoproteinemia.

Lipolysis produces changes in the immunoreactivity and cell reactivity of very low density lipoproteins.

Smaller very low density lipoprotein (VLDL) remnants interact more readily with tissues than do larger "intact" VLDL. This may be related to changes in the availability of VLDL apoproteins on the surface of the lipoproteins. To test this hypothesis VLDL were incubated at 37 degrees C with bovine milk lipase (LPL), and the abilities of LPL-treated VLDL preparations to compete with (125)I-low density lipoproteins (LDL) for interaction with cultured normal human fibroblasts were measured. At the same time, the immunologic activities of these preparations were also tested by double antibody radioimmunoassay. Triglyceride (TG) contents of VLDL fell by 30-90% during incubation with LPL and, on zonal ultracentrifugation, VLDL of faster Svedberg unit of flotation (S(f1.063)) rates (>150) were gradually converted to smaller VLDL with lower S(f) rates (21-60). LPL-treated VLDL competed two to five times more effectively with (125)I-LDL for binding to cellular receptors than did control VLDL. Control VLDL incubated with heat-inactivated LPL at 37 degrees C, or with active LPL at 4 degrees C had unaltered cell reactivities and TG contents compared with VLDL incubated without any enzyme. The direct uptake and degradation of LPL-treated VLDL was also assessed by using VLDL (125)I-labeled in apoprotein (Apo)B. LPL-treated VLDL-(125)I-ApoB were taken up and degraded by fibroblast at greater rates than were control VLDL-(125)I-ApoB. Thus, hydrolysis of VLDL lipids was accompanied by an increased ability of VLDL to interact with fibroblasts. The immunoreactivity of ApoB in the same VLDL preparations, expressed as the "apparent ApoB contents" of LPL-treated VLDL, increased by 10-50% (P < 0.02) in those assays that contained anti-LDL antisera, but the ApoB of control VLDL remained constant. However, assays that contained antisera directed against ApoB isolated from VLDL did not distinguish between LPL-treated and control VLDL. Thus, VLDL lipid hydrolysis was accompanied by changes in the immunoreactivity of VLDL-ApoB, which probably reflect changes in the disposition of ApoB on the surface of VLDL. The altered disposition of ApoB on VLDL "remnants" may be related to their enhanced interaction with cells.

Hyperlipoproteinemia and multifocal neurologic dysfunction in systemic lupus erythematosus.

Two patients with systemic lupus erythematosus, including one with neuropathologic findings, had recurrent multifocal neurologic dysfunction and hyperlipoproteinemia. The lipoprotein disturbances were complex and variable over time. Deficient lipoprotein lipase was found in both patients and appeared to be related temporally to neurologic deterioration. One of these patients had neurologic disease and lipoprotein abnormalities 2 1/2 years before SLE could be documented serologically. These studies suggest that lipoprotein lipase deficiency may be a marker for the endothelial disorder causing cerebral vasculopathy in SLE.

Interaction of a paramagnetic analogue of oxaloacetate with citrate synthase.

Electron paramagnetic resonance studies have indicated that nitrosodisulfonate binds to pig heart citrate synthase. Titration of the enzyme with nitrosodisulfonate revealed several binding sites for the probe per subunit with one site (KD approximately 0.1 mM) having a greater affinity than the others. The substrate, oxaloacetate, competed very effectively for one of the nitrosodisulfonate binding sites (KD less than 10(-2) mM) at the same time eliminating the weaker probe binding sites. Citrate and (R)- and (S)-malates also displaced the probe. Failure to resolve low- and high-field shoulder in the high gain--high modulation electron paramagnetic resonance spectra of the enzyme--nitrosodisulfonate system indicated that the bound probe was "weakly immobilized". However, the electron paramagnetic resonance spectrum of the bound probe changed to one typical of a "strongly immobilized" nitroxide upon the addition of a saturating concentration of the substrate acetyl coenzyme A (acetyl-CoA) to the enzyme--nitrosodisulfonate system, indicating the formation of a ternary acetyl-CoA-enzyme-probe complex. Titration of the acetyl-CoA saturated enzyme with the probe indicated one binding site per subunit (KD = 0.37 mM). Thus, nitrosodisulfonate may be considered as a paramagnetic analogue of oxaloacetate in its interaction with citrate synthase. These results are compared with our previous studies with this enzyme, employing a spin-labeled acyl coenzyme A (acyl-CoA) derivative [Weidman, S. W., Drysdale, G. R., & Mildvan, A. S. (1973) Biochemistry 12, 1874--1883].