sn-1,2-diacylglycerol kinase of Escherichia coli. Diacylglycerol analogues define specificity and mechanism

A detailed structure/function analysis of the substrate specificity of Escherichia coli sn-1,2-diacylglycerol kinase was performed with three goals in mind: (a) to define the substrate specificity; (b) to discover inhibitors; and (c) to elucidate the specificity of diacylglycerol-dependent inactivation. Forty-seven structural analogues of sn-1,2-diacylglycerol were prepared and examined as substrates, inhibitors, and irreversible inactivators of the enzyme using mixed micellar assay methods. Modification of the acyl chains or the sn-2 ester affected the apparent Km but had only small effects on Vm; modifications of the sn-1 ester, sn-3 methylene, or sn-3 hydroxyl had large effects on the apparent Vm and smaller effects on Km. Consistent with these observations, diacylglycerol analogues modified only in the acyl chains or sn-2 ester were not diacylglycerol kinase inhibitors, whereas analogues with substitutions of the sn-1 ester or sn-3 hydroxyl frequently caused inhibition. A hydrogen bond-donating group was required for an analogue to be a diacylglycerol kinase inhibitor. Studies of diacylglycerol kinase inactivation by the various analogues were consistent with the previous conclusion that this process involves an interaction of diacylglycerols with an enzyme conformation different from that active in catalysis (Walsh, J. P., and Bell, R. M. (1986) J. Biol. Chem. 261, 15062-15069). Studies with a water-soluble diacylglycerol, sn-1,2-dibutyrylglycerol, allowed direct comparison of diacylglycerol kinase activity in mixed micelles with that in native membranes. The results are discussed in relation to the structural requirements of other diacylglycerol-dependent enzymes.     

Identification of a soluble diacylglycerol kinase required for lipoteichoic acid production in Bacillus subtilis

Diacylglycerol kinases (DagKs) are key enzymes in lipid metabolism that function to reintroduce diacylglycerol formed from the hydrolysis of phospholipids into the biosynthetic pathway. Bacillus subtilis is a prototypical Gram-positive bacterium with a lipoteichoic acid structure containing repeating units of sn-glycerol-1-P groups derived from phosphatidylglycerol head groups. The B. subtilis homolog of the prokaryotic DagK gene family (dgkA; Pfam01219) was not a DagK but rather was an undecaprenol kinase. The three members of the soluble DagK protein family (Pfam00781) in B. subtilis were tested by complementation of an E. coli dgkA mutant, and only the essential yerQ gene possessed DagK activity. This gene was dubbed dgkB, and the soluble protein product was purified, and its DagK activity was verified in vitro. Conditional inactivation of dgkB led to the accumulation of diacylglycerol and the cessation of lipoteichoic acid formation in B. subtilis. This study identifies a soluble protein encoded by the dgkB (yerQ) gene as an essential kinase in the diacylglycerol cycle that drives lipoteichoic acid production.     

Bacillus subtilis diacylglycerol kinase (DgkA) enhances efficient sporulation

The sn-1,2-diacylglycerol kinase homologue gene, dgkA, is a sporulation gene indispensable for the maintenance of spore stability and viability in Bacillus subtilis. After 6 h of growth in resuspension medium, the endospore morphology of the dgkA mutant by standard phase-contrast microscopy was normal; however, after 9 h, the endospores appeared mostly dark by phase-contrast microscopy, suggesting a defect in the spores. Moreover, electron microscopic studies revealed an abnormal cortex structure in mutant endospores 6 h after the onset of sporulation, an indication of cortex degeneration. In addition, a significant decrease in the dipicolinic acid content of mutant spores was observed. We also found that dgkA is expressed mainly during the vegetative phase. It seems likely that either the DgkA produced during growth prepares the cell for an essential step in sporulation or the enzyme persists into sporulation and performs an essential function.     

Distribution of distinct arachidonoyl-specific and non-specific isoenzymes of diacylglycerol kinase in baboon (Papio cynocephalus) tissues.

We investigated the diacyglycerol kinase species present in several baboon tissues using the substrates sn-1-stearoyl-2-arachidonoyl diacylglycerol and sn-1,2-didecanoyl diacylglycerol. Chromatography of octyl glucoside extracts of the baboon (Papio cynocephalus papio) tissues on hydroxyapatite columns revealed the presence of three diacylglycerol kinase species with different substrate preferences. One species markedly 'preferred' the substrate sn-1-stearoyl-2-arachidonoylglycerol, the two other species preferred sn-1,2-didecanoylglycerol. Measurement of the activity of the baboon brain diacylglycerol kinases toward diacylglycerols with a range of different fatty acid chains revealed a strict preference of the arachidonoyl diacylglycerol kinase for sn-1-acyl-2-arachidonoyl diacylglycerol, whereas the other enzymes showed no preference toward several long-chain-fatty-acid-containing diacylglycerols. The arachidonoyl diacylglycerol kinase was particularly abundant in brain and testis, whereas liver was practically devoid of this enzyme. The arachidonoyl diacylglycerol kinase from baboon brain was found to be predominantly associated with the particulate fraction and exhibited an apparent molecular mass of 130 kDa.     

sn-1,2-Diacylglycerol kinase of Escherichia coli. Mixed micellar analysis of the phospholipid cofactor requirement and divalent cation dependence

Efficient delivery of hydrophobic water-insoluble substrates and cofactors to membrane-bound enzymes is a recurring problem which has impeded kinetic analyses. Kinetic analysis of the Escherichia coli sn-1,2-diacylglycerol kinase, an extremely hydrophobic integral membrane protein of 122 residues, was facilitated by the development of a mixed micellar assay. beta-Octyl glucoside micelles quantitatively solubilized diacylglycerol kinase from membranes of strains which overproduced the enzyme up to 250-fold and provided an effective method to disperse and deliver the hydrophobic water-insoluble substrate, sn-1,2-dioleoyglycerol. Diacylglycerol kinase was active in mixed micelles containing octyl glucoside and dioleoyglycerol. Several phospholipids stimulated activity up to 6-fold, suggesting a cofactor function. Activation by phospholipids was not stereospecific and was mimicked partially by fatty acids. Half-maximal activation was observed at 1 mol % cardiolipin, suggesting that a small number of phospholipids are sufficient to activate the enzyme. Activity was dependent on the mole fractions of dioleoylglycerol and phospholipid in the mixed micelles, but independent of micelle number. Several lines of evidence indicated that the transfer of diacylglycerol between micelles was much more rapid than its utilization by the enzyme. Diacylglycerol kinase exhibited Michaelis-Menten kinetics with respect to diacylglycerol and MgATP. A second Mg2+ ion (in addition to MgATP) was required for activity. When Mg2+ was excluded from the assay, Mn2+, Zn2+, Cd2+, and Co2+ supported activity to lesser extents. These data establish a suitable system for in-depth kinetic analysis of the E. coli diacylglycerol kinase and its phospholipid cofactor requirements.     

Action of retinoic acid on the diacylglycerol-induced ornithine decarboxylase activity, reduction in EGF binding and protein kinase C activation in rat tracheal epithelial 2C5 cells

We have shown previously that in rat tracheal epithelial 2C5 cells the induction of ornithine decarboxylase (ODC) activity and the reduction in the binding of epidermal growth factor (EGF) by diacylglycerol is related to the activation of protein kinase C. In this paper we analyse the action of retinoic acid (RA) on these two parameters in order to determine whether RA acts on the level of protein kinase C. RA inhibits the induction of ODC activity by diacylglycerol (sn-1,2-dioctanoylglycerol) in a dose- and time-dependent manner. A biologically inactive analog of RA has no effect on this induction. RA does not affect the activation of protein kinase C by diacylglycerol in an in vitro assay. In contrast to the effect on ODC induction, RA does not counteract the reduction in EGF binding induced by diacylglycerol. These results are consistent with the concept that RA does not act at the level of protein kinase C and inhibits ODC induction during a stage following protein kinase C activation.     

Diacylglycerol metabolism and arachidonic acid release in human fetal membranes and decidua vera

Diacylglycerol lipase activity has been demonstrated in human fetal membranes and decidua vera tissues. The specific activity of the enzyme is highest in the microsomal fraction of decidua vera tissue. The acylester bond at the sn-1 position of 1,2-diacyl-sn-glycerol is hydrolyzed followed by release of the fatty acid at the sn-2 position. The diacylglycerol lipase activity present in the microsomal fraction of decidua vera tissue hydrolyzes preferentially a diacylglycerol containing an arachidonoyl group in the sn-2 position. Monoacylglycerol lipase activity was also demonstrated in these tissues. The specific activity of monoacylglycerol lipase was significantly greater than that of diacylglycerol lipase and catalyzed preferentially the hydrolysis of monoacylglycerols containing an arachidonyl group in the sn-2 position. Based on the subcellular distribution and the differential effects of various inhibitors, we suggest that the monoacylglycerol lipase and diacylglycerol lipase in decidua vera tissue are 2 distinct enzymes. Diacylglycerol kinase specific activity was examined also and was found to be 4-5 times greater in amnion than in either chorion laeve or decidua vera. The importance of diacylglycerol metabolism in the mechanism of arachidonic acid release and prostaglandin biosynthesis is discussed.     

Lipid-dependent membrane enzymes. Purification to homogeneity and further characterization of diacylglycerol kinase from Escherichia coli

1. Diacylglycerol kinase apoprotein was purified from membranes of Escherichia coli K12 by a six-step procedure that included HPLC. The proposed assignment of the enzyme to the dgkA gene [Lightner et al. (1983) J. Biol. Chem. 258, 10856-10861] could be supported by molecular mass determination (approximately 14 kDa), N-terminal sequencing (Met-Ala-Asn), cyanogen bromide fragmentation and amino acid analysis. As predicted, proline was absent. 2. The membrane-associated as well as the butan-1-ol-dissolved enzyme survived heating to 100 degrees C. 3. Alkylglycoside detergents were found to constitute an additional class of lipid activators. 4. The enzyme apoprotein in a non-activating substrate/detergent solution was capable of autocatalytic self-activation which was attributed to a novel feedback activation mechanism involving phosphatidic acid (diacylglycerol 3-phosphate).     

Immunological characterization of sn-1,2-diacylglycerol and sn-2-monoacylglycerol kinase from pig brain

Rabbit antisera were raised against diacylglycerol kinase purified from pig brain cytosol. Upon immunoblot analysis, the antibody was specifically reactive with the kinase (Mr = 79,000-80,000). Pig brain cytosol, microsomal, and synaptosomal fractions all contained the immunoreactive Mr = 80,000 polypeptide, thus showing that the same enzyme is present in the soluble as well as membrane fractions of the brain. The antibody could precipitate only 60% of the kinase activity present in the crude cytosol. Further, the antibody exhibited very little or no cross-reactivity toward liver cytosolic enzymes obtained from different animals including pigs. Immunostaining of brain tissues demonstrated that neurons, in particular, their nuclei, were positively stained, whereas glial cells were not stained. It is likely that there exists a tissue-and/or cell-dependent immunological multiplicity of diacylglycerol kinase. The enzyme activities phosphorylating sn-1 and sn-2 monoacylglycerols were co-precipitated by the antibody, indicating their identity with diacylglycerol kinase. The enzyme activity toward sn-1 monoolein was much lower than that obtained with sn-2 monoolein. Enzymic as well as chemical analyses of acyl isomers of the reaction products showed that even tested with pure (greater than 95%) sn-1 monoolein, about 70% of the formed lysophosphatidate was of the sn-2 acyl type. The results show that diacylglycerol kinase phosphorylates almost exclusively the sn-2 acyl type of monoacyl-glycerol.     

Membrane protein-lipid hydrogen bonding: evidence from protein kinase C, diglyceride, and tumor promotors

Membrane-bound proteins owe their retention and conformation in the lipid bilayer to hydrophobic peptide domains. Additional fixation, by protein-lipid hydrogen bonding, has been suggested, and recent reports on protein kinase C activation by diacylglycerol (DG) provide an unambiguous model for such bonding. The sn-1,2-diacylglycerol appears to donate a hydrogen bond from the sn-3 hydroxyl to the enzyme and to receive two hydrogen bonds, in the sn-1 and sn-2 ester CO groups, from the enzyme. This arrangement is confirmed in phorbol ester, a competitive inhibitor of DG for the kinase. This tumor promotor has a nearly identical spatial arrangement of hydrogen bond donor (9 alpha-OH) and acceptors (12 and 13 ester CO); so have two other tumor promotors, teleocidin and aplysiatoxin. There are reasons to believe that protein kinase C is not the only protein that is bound to membrane lipids by hydrogen bonding, and such bonding will have to be considered in membrane-associated events such as fusion, cross-membrane transport, or anesthesia.     

Diacylglycerol lipase and kinase activities in rabbit aorta and coronary microvessels

Diacylglycerol lipase and kinase activities were measured in particulate and soluble fractions from rabbit aorta (intima-media) and coronary microvessels. With rabbit aorta, the hydrolysis at the sn-1 position of 1-palmitoyl-2-oleoyl-sn-glycerol had a pH optimum of 5-6 and was greater than hydrolysis at the sn-2 position (pH optimum of 6.5). Only the 2-monoacylglycerol accumulated during incubations at pH 5 and 6.5. These results are consistent with an ordered two-step reaction sequence where the fatty acid at the sn-1 position is released first, followed by the hydrolysis of the fatty acid from the 2-monoacylglycerol by a monoacylglycerol lipase with a neutral pH optimum. Lipase activity (sn-2 hydrolysis) at pH 6.5 was greater than kinase activity at all substrate concentrations. The presence of arachidonate at the sn-2 position of the diacylglycerol increased kinase activity but had little effect on lipase activity. Kinase activity was mainly particulate, whereas 50-60% of diacylglycerol lipase and 50% of monoacylglycerol lipase activity were soluble. Diacylglycerol lipase and kinase were also present in coronary microvessel preparations. Diacylglycerol lipase (sn-2 hydrolysis) activity in coronary microvessels was not enhanced by preincubation of the enzyme preparation with cAMP-dependent protein kinase.     

Spk1, a new kinase from Saccharomyces cerevisiae, phosphorylates proteins on serine, threonine, and tyrosine.

A Saccharomyces cerevisiae lambda gt11 library was screened with antiphosphotyrosine antibodies in an attempt to identify a gene encoding a tyrosine kinase. A subclone derived from one positive phage was sequenced and found to contain an 821-amino-acid open reading frame that encodes a protein with homology to protein kinases. We tested the activity of the putative kinase by constructing a vector encoding a glutathione-S-transferase fusion protein containing most of the predicted polypeptide. The fusion protein phosphorylated endogenous substrates and enolase primarily on serine and threonine. The gene was designated SPK1 for serine-protein kinase. Expression of the Spk1 fusion protein in bacteria stimulated serine, threonine, and tyrosine phosphorylation of bacterial proteins. These results, combined with the antiphosphotyrosine immunoreactivity induced by the kinase, indicate that Spk1 is capable of phosphorylating tyrosine as well as phosphorylating serine and threonine. In in vitro assays, the fusion protein kinase phosphorylated the synthetic substrate poly(Glu/Tyr) on tyrosine, but the activity was weak compared with serine and threonine phosphorylation of other substrates. To determine if other serine/threonine kinases would phosphorylate poly(Glu/Tyr), we tested calcium/calmodulin-dependent protein kinase II and the catalytic subunit of cyclic AMP-dependent protein kinase. The two kinases had similar tyrosine-phosphorylating activities. These results establish that the functional difference between serine/threonine- and tyrosine-protein kinases is not absolute and suggest that there may be physiological circumstances in which tyrosine phosphorylation is mediated by serine/threonine kinases.     

Structural requirements for diacylglycerols to mimic tumor-promoting phobol diester action on the epidermal growth factor receptor

The structural requirements for diacylglycerols to mimic the action of tumor-promoting phorbol diesters on the epidermal growth factor (EGF) receptor of A431 human epidermoid carcinoma cells were investigated. Five biological effects were considered: inhibition of high affinity 125I-EGF binding, change in the phosphorylation state of the EGF receptor, inhibition of the EGF-dependent tyrosine phosphorylation of the EGF receptor, inhibition of [3H]phorbol 12 beta, 13 alpha-dibutyrate binding, and stimulation of calcium- and phospholipid-dependent protein kinase (C-kinase) in vitro. A marked effect of the acyl chain length, 3-10 carbons, of symmetric sn-1,2-diacylglycerols was observed on their ability to mimic the effect of 4 beta-phorbol 12 beta-myristate 13 alpha-acetate (PMA). sn-1,2-Dipropanoylglycerol did not mimic the effects of PMA, but sn-1,2-didecanoylglycerol potently mimicked PMA action. A correlation was found between the ability of these diacylglycerols to stimulate the activity of C-kinase in vitro and to mimic the effects of PMA on the EGF receptor in intact cells. Analogues of sn-1,2-dioctanoylglycerol in which the 3' hydroxyl group was substituted with hydrogen, thio or chloro moieties were inactive when assayed for their ability to stimulate C-kinase in vitro and mimic PMA action in intact cells. We conclude that the hydroxyl group of a diacylglycerol is vital for the interaction with the phorbol diester receptor. The stringent correlation between the potency of the 11 diacylglycerol analogues tested to modulate C-kinase in vitro and to mimic PMA action in vivo provides strong evidence for the hypothesis that C-kinase plays a central role in the regulation of A431 cell EGF receptors by tumor-promoting phorbol diesters.     

Stereospecificity of monoacylglycerol acyltransferase activity from rat intestine and suckling rat liver

The stereospecificity of monoacylglycerol acyltransferase from rat intestinal mucosa and suckling rat liver microsomes was examined using sn-1,2-diacylglycerol kinase from Escherichia coli. With 2-monooleoyl glycerol and palmitoyl-CoA, 88 and 87.9% of the diacylglycerol synthesized by the intestinal mucosa and suckling liver, respectively, was demonstrated to be the sn-1,2-isomer. Analysis of similar preparations of these diacylglycerol products by gas-liquid chromatography-mass spectrometry indicated that most of the remaining diacylglycerol was the 1,3-isomer that probably arose via acyl-migration. These results indicate that monoacylglycerol acyltransferase is stereospecific. Measurement of acyltransferase activities in microsomes using 1- and 2-monoacyl- and monoalkylglycerols as substrates indicated that the monoacylglycerol acyltransferases from suckling liver and intestinal mucosa have different substrate specificities.     

Regulation of one type of Ca2+ current in smooth muscle cells by diacylglycerol and acetylcholine

Electrophysiological recordings from freshly dissociated smooth muscle cells from the stomach of the toad Bufo marinus revealed two types of Ca2+ currents. One has a low threshold of activation and inactivates rapidly; the other has a high threshold of activation and inactivates more slowly. Acetylcholine (ACh) increased the high-threshold current but not the low-threshold current. The synthetic diacylglycerol analog sn-1,2-dioctanoylglycerol, an activator of protein kinase C (PKC), mimicked these effects of ACh on Ca2+ currents. However, another diacylglycerol analog, 1,2-dioctanoyl-3-thioglycerol, which has a closely related structure but does not activate PKC, failed to increase the Ca2+ current. The same was true of 1,2-dioctanoyl-3-chloropropanediol, an analog that even at high concentrations only minimally activates PKC. These results suggest that diacylglycerol may be the second messenger mediating the effects of ACh on one type of voltage-activated Ca2+ channel, possibly by activating PKC.     

Acyloxyacyl hydrolase, a leukocyte enzyme that deacylates bacterial lipopolysaccharides, has phospholipase, lysophospholipase, diacylglycerollipase, and acyltransferase activities in vitro.

Human acyloxyacyl hydrolase (AOAH) is a leukocyte enzyme that hydrolyzes acyloxyacyl bonds in the lipid A region of bacterial lipopolysaccharide (LPS), thereby detoxifying the LPS. We report here that the enzyme also acts in vitro on glycerophospholipids, lysophospholipids, and diacylglycerol. While AOAH preferentially removes palmitate or stearate from the sn-1 position of phospholipid and diacylglycerol substrates that have unsaturated acyl chains in the sn-2 position, it is able to cleave both palmitates from sn-1,2-dipalmitoylphosphatidylcholine and sn-1,2-dipalmitoylglycerol. This apparent preference for removing saturated (or shorter) acyl chains from glycerolipids is consistent with its ability to cleave laurate more rapidly than palmitoleate from lipopolysaccharide (Erwin, A. L., and Munford, R. S. (1990) J. Biol. Chem. 265, 16444-16449). AOAH also catalyzes acyl transfer from LPS and phosphatidylethanolamine to acceptor lipids; approximately equal amounts of laurate and myristate are transferred from LPS to monooleoylglyceryl ether, forming acyloleoylglyceryl ether. The demonstration that AOAH has phospholipase, lysophospholipase, diacylglycerol lipase, and acyltransferase activities in vitro suggests that the enzyme may have roles in addition to LPS deacylation (detoxification) in phagocytic cells.