The phospholipid-dependence of uridine diphosphate glucuronyltransferase. Effect of protein deficiency on the phospholipid composition and enzyme activity of rat liver microsomal fraction

After force-feeding a protein-free diet to male rats for 5-7 days a substantial (2.4-fold) increase in the specific activity of the liver microsomal enzyme UDP-glucuronyltransferase (EC 2.4.1.17) was observed. A similar activation of the enzyme occurred when rats were fed on a low-protein (5%, w/w, casein) diet for 60 days. Although both the short- and long-term protein-deficient diets decreased the contents of microsomal protein and phospholipid in liver tissue they did not significantly alter the ratio of these major membrane components. Protein deficiency profoundly altered the phospholipid composition of microsomal membranes. The most striking difference in microsomal phospholipid composition between control and protein-deficient rats was their content of lysophosphatides. Whereas microsomal membranes from protein-deficient rats contained significant proportions of lysophosphatidylcholine and lysophosphatidylethanolamine very little or no lysophosphatides were detected in control preparations. Pretreatment of microsomal fractions from normal rats with phospholipase A markedly increased their UDP-glucuronyltransferase activity as did their pretreatment with lysophosphatidylcholine. It is concluded that the quantities of lysophosphatides present in microsomal membranes from protein-deficient rats were sufficient to have caused the increased UDP-glucuronyltransferase activities of these preparations. Evidence is presented suggesting that these changes in microsomal phospholipid composition and UDP-glucuronyltransferase activity caused by protein deficiency reflect changes that occur in vivo. The possible physiological significance of these findings is discussed.     

The phospholipid dependence of UDP-glucuronyltransferase: Conformation/reactivity studies with purified enzyme

Highly-active purified UDP-glucuronyltransferase from guinea-pig liver microsomal membranes is associated with phospholipids. Removal of these phospholipids inactivated the transferase and caused profound changes in the enzyme's circular dichroism spectrum indicating that its secondary structure was drastically altered. Treatment of the delipidated fraction with phosphatidylcholine restored the enzyme to a much more helical, high reactivity conformation. These results show clearly that an intact phospholipid environment is required to maintain the transferase in a reactive conformation.     

Gonadotropin receptors in plasma membranes of bovine corpus luteum. II. Role of membrane phospholipids.

The role of phospholipids in the binding of 125I-choriogonadotropin to bovine corpus luteum plasma membranes has been investigated with the use of purified phospholipase A and phospholipase C to alter membrane phospholipids. The phospholipase C-digested plasma membrane preparation showed 85 to 90% inhibition of 125I-choriogonadotropin binding activity when 70% of the membrane phospholipid was hydrolyzed. Similarly treatment of plasma membranes with phospholipase A resulted in 45 to 55% hydrolysis of membrane phospholipid and almost 75% inhibition of receptor activity. Both these enzymes hydrolyzed membrane-associated phosphatidylcholine to a greater extent than phosphatidylethanolamine and phosphatidylserine. Phosphorylaminoalcohols of phospholiphase C end products were completely released into the medium, while phospholipase A by-products remained associated with plasma membranes. Addition of a phospholipids suspension or liposomes to plasma membranes pretreated with phospholipase A and C did not restore gonadotropin binding activity. Soluble phosphorylcholine, phosphorylethanolamine, and phosphorylserine and insoluble diglyceride products of phospholipase C action had no effect on receptor activity. In contrast, end products of the phospholipase A action, such as lysophosphatides and fatty acids, inhibited both on the membrane-associated and solubilized receptor activity. Lysophosphatidylcholine was the most effective end product inhibiting the binding of gonadotropin to the receptor, followed by lysophosphatidylethanolamine and lysophosphatidylserine. The inhibitory effects of phospholipase A or lysophosphatides were completely reversed upon removal of membrane-bound phospholipid end products by washing the membranes with defatted bovine serum albumin. However, phospholipase C inhibition could not be overcome by defatted albumin washings. Solubilization of plasma membranes with detergents which had been pretreated with phospholipase C partially restored the inhibited activity. It is concluded that the phospholipase-mediated inhibition of gonadotropin binding activity was due to hydrolysis and alterations of the phospholipid environment in the case of phospholipase C and by direct inhibition by end products in the case of phospholipase A.     

The effect of lipid peroxidation on the calcium-accumulating ability of the microsomal fraction isolated from chicken breast muscle.

The effect of lipid peroxidation on the Ca2+-accumulating and Ca2+-retaining abilities of the microsomal fraction from chicken breast muscle was investigated. At 25 degrees C, enzymic lipid peroxidation did not seriously affect either of these abilities unless ascorbic acid was present, when both were diminished. At 37 degrees C, Ca2+-concentrating ability was decreased further by the effects of heat damage to the membrane. Membrane lipid peroxidation did not affect microsomal adenosine triphosphatase activity unless the microsomal fraction was subsequently washed with albumin. This effect of albumin is possibly due to removal of lipid-breakdown products. Addition of soya-bean phospholipids to the peroxidized vesicles washed with albumin restored adenosine triphosphatase activity, demonstrating a non-specific phospholipid requirement.     

Enzyme and phospholipid asymmetry in liver microsomal membranes

The transverse distribution of enzyme proteins and phospholipids within microsomal membranes was studied by analyzing membrane composition after treatment with proteases and phospholipases. Upon trypsin treatment of closed microsomal vesicles, NADH- and NADPH-cytochrome c reductases as well as cytochrome b5 were solubilized or inactivated, while cytochrome P-450 was partially inactivated. When microsomes were exposed to a concentration of deoxycholate which makes them permeable to macromolecules but does not disrupt the membrane, the detergent alone was sufficient to release four enzymes: nucleoside diphosphatase, esterase, beta-glucuronidase, and a portion of the DT-diaphorase. Introduction of trypsin into the vesicle lumen inactivated glucose-6-phosphatase completely and cytochrome P-450 partially. The rest of this cytochrome, ATPase, AMPase, UDP-glucuronyltransferase, and the remaining 50% of DT-diaphorase activity were not affected by proteolysis from either side of the membrane. Phospholipase A treatment of intact microsomes in the presence of albumin hydrolyzed all of the phosphatidylethanolamine, phosphatidylserine, and 55% of the phosphatidylcholine. From this observation, it was concluded that these lipids are localized in the outer half of the bilayer of the microsomal membrane; Phosphatidylinositol, 45% of the phosphatidylcholine, and sphingomyelin are tentatively assigned to the inner half of this bilayer. It appears that the various enzyme proteins and phospholipids of the microsomal membrane display an asymmetric distribution in the transverse plane.     

Inhibition of glucose 6-phosphatase by pure and impure C-type phospholipases. Reactivation by phospholipid dispersions and protection by serum albumin.

1. Pure or impure C-type phospholipases hydrolysed rat liver microsomal phosphatides in situ at 5 degrees or 37 degrees C. At 5 degrees C mean hydrolysis of total phospholipids was 90% by Bacillus cereus and 75% by Clostridium perfringens (Clostridium welchii) C-type phospholipases. 2. Four degrees of inhibition of glucose 6-phosphatase (D-glucose 6-phosphate phosphohydrolase; EC 3.1.3.9) resulted. (a) At 37 degrees C inhibition was virtually complete and apparently irreversible. (b) At 5 degrees C phospholipase C inhibited 50-87% of the activity expressed by intact control microsomal fractions. (c) Bovine serum albumin present during delipidation alleviated most of this inhibition: at 5 degrees C phospholipase C plus bovine serum albumin inhibited by 0-35% (mean 18%):simultaneous stimulation by the destruction of its latency seems to offset glucose 6-phosphatase inhibition, sometimes completely. (d) If latency was first destroyed, phospholipase C plus bovine serum albumin inhibited 30-50% of total glucose 6-phosphatase activity at 5 degrees C. Only this inhibition is likely largely to reflect the lower availability of phospholipids, essential for maximal enzyme activity, as it is virtually completely reversed by added phospholipid dispersions. Co-dispersions of phosphatidylserine plus phosphatidylcholine (1:1, w/w) were especially effective but Triton X-100 was unable effectively to restore activity. 3. Considerable glucose 6-phosphatase activity survived 240min of treatment with phospholipase C at 5 degrees C, but in the absence of substrate or at physiological glucose 6-phosphate concentrations the delipidated enzyme was completely inactivated within 10min at 37 degrees C. However, 80mM-glucose 6-phosphate stabilized it and phospholipid dispersions substantially restored thermal stability. 4. It is concluded that glucose 6-phosphatase is at least partly phospholipid-dependent, and complete dependence is not excluded. For reasons discussed it is impossible yet to be certain which phospholipid class(es) the enzyme requires for activity.     

The phospholipid-dependence of uridine diphosphate glucuronyltransferase. Reactivation of phospholipase A-inactivated enzyme by phospholipids and detergents

Specific degradation of the phospholipid membrane of guinea-pig liver microsomal fraction with phospholipase A inactivated glucuronyltransferase. The inactivation was reversed by phosphatidylcholine and mixed microsomal phospholipid micelles at concentrations similar to those present in intact microsomal preparations. The other commonly occurring phospholipids did not reactivate phospholipase A-treated enzyme. Since the mixed microsomal phospholipids consisted mainly of phosphatidylcholine, it is concluded that the reactivation by phospholipids is phosphatidylcholine-specific. Reactivation was also achieved by low concentrations of the cationic detergents cetylpyridinium chloride and cetyltrimethylammonium bromide. Higher concentrations of these detergents inactivated the glucuronyltransferase activity of intact and phospholipase A-treated microsomal fractions. Anionic detergents were potent inactivators of the glucuronyltransferase activity of untreated and phospholipase A-treated microsomal fractions, whereas non-ionic detergents had little effect on the activity of either preparation. Measurements of the zeta-potentials of the micellar species used in this study showed that no obvious relationship existed between the zeta-potentials and the ability to reactivate glucuronyltransferase. However, high positive or negative zeta-potentials were correlated with the ability of the amphipathic compound to inactivate glucuronyltransferase.     

Phospholipid requirement for 2-acetamidofluorene N-and ring-hydroxylation by hamster liver microsomal cytochrome P-450 enzyme system.

A phospholipid requirement of 2-acetamidofluorene N- and ring-hydroxylation was investigated with partially delipidated microsomal fraction from livers of 3-methylcholanthrene-pretreated hamsters. Butan-1-ol extraction of microsomal fraction removed 90% of the total lipid content without any appreciable effect on microsomal proteins. Such extracted microsomal fractions had much lower capacity to N- and ring-hydroxylate 2-acetamidofluorene: 25 and 44% of control respectively. Addition of butan-1-ol-extracted total lipid restored both oxidations to some extent, whereas addition of phosphatidylcholine fraction restored both oxidations almost completely. Addition of synthetic phospholipid, dilauroyl phosphatidylcholine, restored both oxidations to a large extent, whereas synthetic dipalmitoyl or distearoyl phosphatidylcholine was ineffective in restoring these oxidations.     

The phospholipid-dependence of uridine diphosphate glucuronyltransferase. Temperature-dependence of microsomal enzyme activity and thermotropic changes in membrane structure.

Arrhenius plots of the non-latent UDP-glucuronyltransferase (p-nitrophenol acceptor) activity of guinea-pig microsomal membranes prepared with 154 mM-KCl were linear from 5 to 40 degrees C. Arrhenius plots for other microsomal preparations from guinea pig and rat liver that show various degrees of transferase latency, exhibited two linear regions intersecting at a sharp transition point near 20-25 degrees C. This discontinuity was abolished or greatly decreased when transferase latency was removed by treating the membranes with perturbants of phospholipid bilayer strucutre. The fluorescent probe N-phenyl-1-naphthyl-amine detected a thermotropic change in the fluidity of the phospholipid acyl chains of all the microsomal membrane preparations studied, at temperatures close to those of the Arrhenius-plot transitions. It is concluded that the thermotropic change in the structure of the membrane bilayer probably is a 'phase separation' or clustering of phospholipids, which affects a permeability barrier that restricts access of substrate to the transferase molecules.     

Phospholipid content and activity of pure uridine diphosphate-glucuronyltransferase from rat liver.

Rat liver phospholipids were radioactively labeled in vivo before purification of UDP-glucuronyltransferase to homogeneity. The pure enzyme contained very little phospholipid (approx. 0.7 mol of phospholipid/mol of protein). The solubilization detergent Lubrol 12A9 appeared to act as a phospholipid substitute, capable of supporting UDP-glucuronyltransferase activity. Phospholipase C did not inhibit the pure enzyme activity and pure UDP-glucuronyltransferase was stimulated by 40--100% by the addition of phospholipid dispersions.     

Effect on uridine diphosphate glucuronyltransferase activity of depleting and restoring phospholipids to guinea-pig liver microsomal preparations.

Phospholipid depletion substantially inhibited the maximum demonstrable activities of the forward (glucuronidation) and reverse reactions of UDP-glucuronyltransferase towards p-nitrophenol in guinea-pig liver microsomal preparations. Dispersions of liver phospholipids restored activity, whereas non-phospholipid amphipaths failed to do so effectively. These results suggest that the system is probably phospholipid-dependent rather than conformationally constrained by phospholipids.     

Changes in the phospholipid catabolism of mitochondria and microsomes during the development of rat liver.

The lipolytic activities of mitochondrial and microsomal fractions ('microsomes') isolated from foetal, suckling and adult rat liver were compared. The catabolism of endogenous phospholipids was followed by measuring the loss of phospholipids and the appearance of non-esterified fatty acids and lysophosphatides. The rate of mitochondrial phospholipid catabolism does not change significantly during development, but the rate of lipolysis of microsomal phospholipids increases 3-fold during development. Balance studies showed that, in mitochondria and microsomes of foetal, suckling and adult rat liver, fatty acid formation is greatly in excess of the fatty acids that can be accounted for by measuring phospholipid disappearance and lysophosphatide appearance. The hypothesis that this excess fatty acid formation resulted from the lipolysis of mitochondrial and microsomal triacylglycerols were tested and confirmed by preliminary experiments. Mitochondria and microsomes isolated from all developmental ages investigated had phospholipases with A1 and A2 activities. The degree of unsaturation of the fatty acids derived from the phospholipids of mitochondria did not vary significantly during development.     

Nucleotide metabolism by microsomal UDP-glucuronyltransferase and nucleoside diphosphatase as determine by 31P nuclear-magnetic-resonance spectroscopy.

31P n.m.r. spectroscopy was used to study the nucleotide kinetics of UDP-glucuronyltransferase and associated reactions in the liver microsomal fraction. The effects of Mg2+ and EDTA on these reactions were investigated qualitatively. It was found that the rabbit microsomal fraction has no nucleoside pyrophosphatase activity, that UDP was immediately hydrolysed and that it was released from the microsomal surface. Reverse glucuronyltransferase could be demonstrated. The results are discussed with reference to functional coupling of UDP-glucuronyltransferase to other enzymes and the effects of Mg2+ and EDTA on the system.     

Identification of increased amounts of UDP-glucuronyltransferase protein in phenobarbital-treated chick-embryo liver cells.

UDP-glucuronyltransferase activity of neonatal-chick liver or phenobarbital-treated chick-embryo liver catalysed the glucuronidation of 1-naphthol, 4-nitrophenol and 2-aminophenol. Only low transferase activity towards testosterone was detected, and activity towards bilirubin was not detectable. Liver microsomal transferase activity towards the three phenols was increased approx. 20-50-fold by phenobarbital treatment of chick embryos or by transfer of liver cells into tissue culture. A single form of UDP-glucuronyltransferase, which appears to catalyse the glucuronidation of these three phenols, was purified to near homogeneity from phenobarbital-treated chick-embryo liver microsomal fraction for the first time. The use of this purified enzyme as a standard protein facilitated the identification of this protein in chick-embryo liver microsomal fraction. Further, the accumulation of this microsomal protein was observed following phenobarbital treatment of chick embryos and during tissue culture of chick-embryo liver cells. The value of this model system for the study of the induction of UDP-glucuronyltransferase by drugs and hormones is discussed.     

Failure of phenobarbital to induce rat hepatic microsomal UDP-glucuronyltransferase toward phenolphthalein

Phenobarbital pretreatment was found not to induce Triton X-100 activated hepatic microsomal neonatal UDP-glucuronyltransferase activity towards phenolphthalein in male, female and pregnant female rats. This was not due to an insufficient concentration of the activating detergent or the choice of detergent. Other microsomal enzymes such as cytochrome P-450 were greatly induced by phenobarbital in the same experimental animals. Late fetal UDP-glucuronyltransferase activity induction by polycyclic hydrocarbons (β-naphthoflavone) was confirmed in this study. Phenobarbital inducibility of neonatal UDP-glucuronyltransferase may not be useful in discriminating between UDP-glucuronyltransferase isoenzymes.     

Involvement of distinct populations of phosphatidylglycerol and phosphatidylcholine molecules in photosynthetic electron-flow activities

When spinach thylakoid membranes were treated with pancreatic phospholipase A₂, phospholipids were degraded and the uncoupled non-cyclic electron-flow activity (from H₂O to NADP⁺) was progressively inhibited. To discriminate between the relative contributions of the hydrolysis products (free fatty acids and lysophospholipids) and of the phospholipid depletion per se to inhibit the activity, we made use of the known property of bovine serum albumin to remove such hydrolysis products from membranes. Using careful washings and adequate lipid extraction procedures, we could ascertain that all hydrolysis products generated by phospholipase A₂ were effectively removed from the thylakoid membrane by bovine serum albumin treatment. When bovine serum albumin was added to thylakoid membranes after various incubation times with the phospholipase A₂, the electron-flow activity was rapidly, but not completely restored. However, when phospholipid hydrolysis exceeded a certain extent (70–85%), the activity was totally inhibited and its restoration by albumin was no longer possible. Addition of EGTA to the phospholipase A₂-treated membranes blocked both the enzyme action and the progress of electron-flow inhibition. Under these conditions, the amplitude of the albumin-induced restoration of electron-flow rate did not depend on the time span between EGTA block and albumin addition. We show that phospholipid depletion of thylakoid membranes is entirely responsible for the irreversible (albumin-insensitive) inhibition of the electron flow from H₂O to NADP⁺ by phospholipase A₂. Plotting the extent (%) of this inhibition vs. the extent (%) of phospholipid depletion allowed us to distinguish three populations of both phosphatidylglycerol and phosphatidylcholine. The first one, which was easily accessible to the enzyme, did not support greatly the electron-flow activity (around 40% of each phospholipid destroyed vs. only 10% or less inhibition). On the other hand, the electron-flow activity strongly depended on the second, less accessible population of phospholipids (around 40% of each phospholipid destroyed vs. 90% inhibition). Finally, the third population of phospholipids was not involved in the uncoupled non-cyclic electron flow activity.     

The influence of fatty acid unsaturation and physical properties of microsomal membrane phospholipids on UDP-glucuronyltransferase activity.

The relationship between lipid composition, the physical properties of microsomal phospholipids and the kinetics of liver UDP-glucuronyltransferase was studied in microsomes from guinea pigs supplied with a normal or a fat-free diet for 28 days. Fatty acid deficiency did not modify either the cholesterol/phospholipid molar ratio or the polar head group composition, but exclusively redistributed the unsaturated fatty acid pattern, by partially exchanging oleic for linoleic acid. This phenomenon accounts for the decrease of both rotational and translational mobilities of the fluorescent probes 1,6-diphenyl-1,3,5-hexatriene (DPH) and pyrene respectively. When the thermotropic behaviour of the different systems was assessed, no transition temperature (gel-liquid-crystalline) between 10 and 40 degrees C was seen as a consequence of the lower degree of unsaturation, either in the microsomal membranes or in the total lipid or total phospholipid extracts from the treated animals. In spite of this, the polarization ratio of trans-parinaric acid and the fluorescence intensity of merocyanine 540 revealed that a significant lateral phase separation occurred at 20-22 degrees C in the extracted phospholipids, which was smoother in the total lipid fractions and in the native microsomal membranes. Fatty acid deficiency caused an upward shift of the midpoint temperature of the lateral phase separation. Furthermore, the phosphatidylcholine extracted from the 'normal' microsomes showed a lateral phase separation centred at a lower temperature than that extracted from 'fat-deficient' microsomes. In contrast, the Arrhenius plot of UDP-glucuronyltransferase from 'normal' microsomes exhibited a change in slope at a higher temperature than that from treated microsomes. These results would suggest that fatty acid deficiency in guinea-pig liver microsomes, while rigidizing the bulk lipids, would segregate the most unsaturated phosphatidylcholine molecules towards the UDP-glucuronyltransferase microenvironment, in accordance with our previous results with cholesterol incorporation [Castuma & Brenner (1986) Biochemistry 25, 4733-4738].     

Characteristics of renal UDP-glucuronyltransferase

Rat renal microsomes catalyzed the glucuronidation of l-naphthol, 4-methylumbelliferone and p-nitrophenol, whereas morphine and testosterone conjugation were not detected. In contrast, all five substrates were conjugated by hepatic microsomes; the activity was typically 5-10 times greater than with renal microsomes. Renal microsomal UDP-glucuronyltransferase toward l-naphthol was fully activated (six-fold) by 0.03% deoxycholate while the hepatic enzyme was fully activated (eight-fold) by 0.05% deoxycholate. Full activation of hepatic UDP-glucuronyltransferase occurred when microsomes had been preincubated at 0 C with deoxycholate for 20 min. This effect of preincubation was not observed with renal microsomes. The presence of 0.25M sucrose in the buffers during renal microsomal preparation resulted in a two-fold greater rate of l-naphthol conjugation in both unactivated and activated microsomes than renal microsomes prepared in phosphate buffers alone. Preparation of hepatic microsomes with or without 0.25M sucrose had no effect on UDP-glucuronyltransferase activity. Unactivated (-deoxycholate) renal enzyme was activated when incubations were done at a low pH (5.7), whereas fully activated (0.03% deoxycholate) renal microsomal UDP-glucuronyltransferase displayed a pH optimum at 6.5. Renal microsomal UDP-glucuronyltransferase activity toward l-naphthol, p-nitrophenol and 4-methylumbelliferone was induced by pretreatment of rats with beta-naphthoflavone and trans-stilbene oxide but not by phenobarbital or 3-methylcholanthrene. These data demonstrate that renal UDP-glucuronyltransferases are different from the hepatic enzymes with regard to biochemical properties, substrate specificity and in response to chemical inducers of xenobiotic metabolism.     

Perinatal developmental changes in hepatic UDP-glucuronyltransferase.

Postnatal developmental changes in hapatic microsomal UDP-glucuronyltransferase were studied in the rat. The previously reported postnatal decline in the capacity of microsomal fractions to glucuronidate p-nitrophenol was found to be observable in unperturbed preparations only at non-saturating concentrations of the substrate UDP-glucuronic acid. At saturating concentrations of UDP-glucuronic acid, activity is identical in newborns and adults. Kinetic analysis revealed that the enzyme from liver of newborns has a much higher affinity for UDP-glucuronic acid than does the enzyme in adults, but the same activity at Vmax. On the other hand, the enzyme from adult liver microsomal fractions can be activated by the physiological allosteric effector UDP-N-acetylglucosamine, whereas the enzyme from newborns is largely unaffected by it. Thus it appears that the number of enzyme active sites is not changing; rather, the enzyme is maturing to a more highly regulable form. There were also differences between the enzymes in newborns and adults in their response to perturbation of the membrane-lipid environment by detergent and phospholipase A. Possible interpretations of these differences are discussed.     

Altered sexual differentiation of hepatic uridine diphosphate glucuronyltransferase by neonatal hormone treatment in rats

The hepatic microsomal enzyme UDP-glucuronyltransferase undergoes a complex developmental pattern in which enzyme activity is first detectable on the 18th day of gestation in rats. Prepubertal activities are similar for males and females. However, postpubertal sexual differentiation of enzyme activity occurs in which male activities are twice those of females. Neonatal administration of testosterone propionate or diethylstilboestrol to intact animals resulted in lowered UDP-glucuronyltransferase activity in liver microsomal fractions of adult male rats, whereas no changes were observed in the adult females and prepubertal male and female animals. Neonatal administration of testosterone propionate and diethylstilboestrol adversely affected male reproductive-tract development as evidenced by decreased weights of testes, seminal vesicles and ventral prostate. Diethylstilboestrol also markedly decreased spermatogenesis. Hypophysectomy of adult male rats resulted in negative modulation of microsomal UDP-glucuronyltransferase and prevented the sexual differentiation of enzyme activity. In contrast hypophysectomy had no effect on female UDP-glucuronyltransferase activity. A pituitary transplant under the kidney capsule was not capable of reversing the enzyme effects of hypophysectomy, therefore suggesting that the male pituitary factor(s) responsible for positive modulation of UDP-glucuronyltransferase might be under hypothalamic control in the form of a releasing factor. Neonatal testosterone propionate and diethylstilboestrol administration apparently interfered with the normal sequence of postpubertal UDP-glucuronyltransferase sexual differentiation.